Twenty Years after Chico Mendes

Permanent Link: http://ufdc.ufl.edu/UFE0023907/00001

Material Information

Title: Twenty Years after Chico Mendes Extractive Reserves' Expansion, Cattle Adoption and Evolving Self-Definition among Rubber Tappers in the Brazilian Amazon
Physical Description: 1 online resource (231 p.)
Language: english
Creator: Gomes, Carlos
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009


Subjects / Keywords: amazon, brazilian, cattle, chico, extractive, mendes, ranching, reserves, rubber, tappers
Geography -- Dissertations, Academic -- UF
Genre: Geography thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: In the 20 years since the murder of rubber tapper leader Chico Mendes and subsequent creation of Extractive Reserves (ERs), ERs have become a popular conservation and development tool in the Brazilian Amazon. However, ER residents have faced challenges to improving their livelihoods based on sustainable use of the forest, leading to a shift towards cattle ranching and a questioning of rubber tappers status as sustainable forest managers. This dissertation focuses on three core themes. The first examines overall ER policy evolution over two decades in the Amazon. The second examines issues of land-use/land-cover change and changing livelihoods among forest extractivists in the Chico Mendes Extractive Reserve (CMER) in Xapuri/Acre compared with cattle adoption by smallholder colonists. The third explores the implications of livelihood changes for the evolving self-definition of rubber tappers. This dissertation is guided by theoretical perspectives of political ecology, environment and development, and land-use/land-cover science. It analyzes socioeconomic data gathered through multiple household surveys and government agencies, qualitative information from key informant interviews, and land-cover change detection. Sixty-four ERs have been created in the Brazilian Amazon since 1990, spanning over 12 million hectares. Different waves of ER establishment responded to political and regional development dynamics at multiple scales that favored or constrained their implementation. In Xapuri, the size of cattle herds rose by 200% from 1990-2005, and throughout the CMER, land cover change detection from 1986-2003 showed some communities nearly surpassing the allowable limits of deforestation. Rubber tapper households have begun pursuing diverse livelihood activities that include not only extractivism, but also small-scale market cultivation, and increasingly, cattle production. Self-identification as a rubber tapper was positively correlated to larger landholding size, more knowledge of management rules, and participation in community activities. Conversely, more recent occupancy in the CMER and, surprisingly, involvement in the official social movement were negatively correlated with respondent s self-identification as a rubber tapper. This dissertation contributes to the debate of ERs as an environmental policy in Amazon. Despite advances, challenges remain for ERs to promote sustainable development, particularly in light of changing livelihoods and continued pressure at development frontiers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carlos Gomes.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Smith, Nigel J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0023907:00001

Permanent Link: http://ufdc.ufl.edu/UFE0023907/00001

Material Information

Title: Twenty Years after Chico Mendes Extractive Reserves' Expansion, Cattle Adoption and Evolving Self-Definition among Rubber Tappers in the Brazilian Amazon
Physical Description: 1 online resource (231 p.)
Language: english
Creator: Gomes, Carlos
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009


Subjects / Keywords: amazon, brazilian, cattle, chico, extractive, mendes, ranching, reserves, rubber, tappers
Geography -- Dissertations, Academic -- UF
Genre: Geography thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: In the 20 years since the murder of rubber tapper leader Chico Mendes and subsequent creation of Extractive Reserves (ERs), ERs have become a popular conservation and development tool in the Brazilian Amazon. However, ER residents have faced challenges to improving their livelihoods based on sustainable use of the forest, leading to a shift towards cattle ranching and a questioning of rubber tappers status as sustainable forest managers. This dissertation focuses on three core themes. The first examines overall ER policy evolution over two decades in the Amazon. The second examines issues of land-use/land-cover change and changing livelihoods among forest extractivists in the Chico Mendes Extractive Reserve (CMER) in Xapuri/Acre compared with cattle adoption by smallholder colonists. The third explores the implications of livelihood changes for the evolving self-definition of rubber tappers. This dissertation is guided by theoretical perspectives of political ecology, environment and development, and land-use/land-cover science. It analyzes socioeconomic data gathered through multiple household surveys and government agencies, qualitative information from key informant interviews, and land-cover change detection. Sixty-four ERs have been created in the Brazilian Amazon since 1990, spanning over 12 million hectares. Different waves of ER establishment responded to political and regional development dynamics at multiple scales that favored or constrained their implementation. In Xapuri, the size of cattle herds rose by 200% from 1990-2005, and throughout the CMER, land cover change detection from 1986-2003 showed some communities nearly surpassing the allowable limits of deforestation. Rubber tapper households have begun pursuing diverse livelihood activities that include not only extractivism, but also small-scale market cultivation, and increasingly, cattle production. Self-identification as a rubber tapper was positively correlated to larger landholding size, more knowledge of management rules, and participation in community activities. Conversely, more recent occupancy in the CMER and, surprisingly, involvement in the official social movement were negatively correlated with respondent s self-identification as a rubber tapper. This dissertation contributes to the debate of ERs as an environmental policy in Amazon. Despite advances, challenges remain for ERs to promote sustainable development, particularly in light of changing livelihoods and continued pressure at development frontiers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carlos Gomes.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Smith, Nigel J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0023907:00001

This item has the following downloads:

Full Text




2 2009 Carlos Valrio Aguiar Gomes


3 To the memory of my father, Carlos Maria da Silva Gomes. I know you were so proud of me. I wish I would have had the opportunity to give you this good news.


4 ACKNOWLEDGMENTS I have com pleted this journey with the encouragement, guidance, and support of many people and institutions. I am very thankful to the members of my dissertation committee: Dr. Nigel Smith, Dr. Marianne Schmink, Dr. Stephen Perz and Dr. Michael Binford for their support and valuable guidance along the way. Dr. Nige l Smith, my chair, has been by my side at every step, sustaining my ideas and encour aging me to move forward through our many conversations that help put my research within a broader perspective of historical and current natural resource use in the Amazon. Dr. Perz has al ways been a source of new ways to approach understanding livelihood transformations of r ubber tappers communities in Acre, and his guidance was critical in helping me structure my dissertation. Dr. Binford has given me much constructive criticism and has chal lenged me to think critically about my research questions and methods for land use science and to understand how exciting and the broader field of Geography can be. Finally, Dr. Schmink has been a wonderful and constant source of support and encouragement since I first t hought about coming to study in the United States. During these years in graduate school, she has given me guidance that goes behind the wall of academia. I express my heartfelt thanks for the professi onal inspiration she has given me and for her friendship. I gratefully acknowledge the s upport of the Brazilian Scienc e Council (CNPq-Ministry for Science and Technology of Brazil), Instituto Internacional de Educao do Brasil (IEBPrograma BECA) and the Tropical Conservation and Development Program /Center for Latin American Studies-UF for fellowships that funde d my doctoral studies at the University of Florida. I am also grateful for financial support for fieldwor k granted by the Center for Environmental Research and Conservation (CERC) Overbroo k Fellowship at Columbia University, International Foundation for Scien ce (IFS), World Wildlife Funds Prince Bernard


5 Fellowship for Nature Conservation and th e Charles A. and Anne Morrow Lindbergh Foundation. I also am grateful to colleagues at UFs Center for Latin American Studies/Tropical Conservation and Development Program. Special thanks go to Dr. Charles Wood, Jon Dain, Dr. Karen Kainer, Hannah Covert, Patricia Sampaio, Wanda Carter, and Margarita Ganda. I also thank Desiree Price in the Department of Ge ography for always being helpful and friendly. I am also grateful for the many people and organiza tions in Acre, Brazil that helped me in the implementation of my fieldwork. I first thank the Research and Extension in Agroforestry Systems Group (PESACRE) for providing institu tional and logistical support during the study period. Special thanks at PESACRE go to Eduar do Amaral (Cazuza), Adair Duarte, Hilza Arcos and Juraci Pacheco. Similarly, the Zoobotanical Park at the Federa l University of Acre provided me with a place to work while in Acre. The re search group there helped smooth out logistical arrangements and made me feel welcome in th eir work space. Special thanks go to Dr. Irving Foster Brown who was my undergraduate advisor at UFAC, and since then has been a long term source of support and advice. I would also like to thank my colleagues in Acre, Willian Flores, Silvia Brilhante, Elza Mendoza, Monica de los Ri os, Andrea Alechandre, Arthur Leite, Ricardo Dantas and Kleder for their s upport and friendship. Thanks also go to my colleagues Francisco Carlos at Embrapa-Acre, Christiane Ehringhaus at Cifor in Belm, Mauro Pires at MMA in Braslia for sharing conversations ideas and data for my research. Many thanks go also to Amy Rosenthal for all her great help on editing my writing and Doriam Borges for helping me with statistical analysis. While at UF, I am very grateful to have ma de many friends that pr ovided me with support and many enriching and memorable moments. To name a few, Geraldo Silva has been a fountain


6 of friendship and support, always being there. Thanks also go to Lucimar Souza, Ane Alencar, Alain Michel, Diana Alvira, Al fredo Rios, Marcos Lentini, Leonardo Martinez, Leonardo Pacheco, Christopher Baraloto and Kara, Simone Athayde, Rafael Rojas, Andrea Chavez, Jeffrey Hoelle, and Iran Rodrigues. I am flattered by the warmth of the rubber tapper families in the Chico Mendes Reserve who took me into their homes and shared thei r knowledge and concerns with incredible generosity. I am greatly indebted with them. During my two years in the fieldwork, I had the opportunity to collaborate with Jacqueline Vadj unec. We decided to join our dissertation proposal together and thus spent countless hours in the field together. What could have been a one time attempt on innovative models of graduate students collaboration in the field, have fostered a bigger sense of mutual collaborati on and strengthened commitments for further and long-term collaboration among us. As my colleague and friend she has been of great support through this dissertation. I woul d like to thank Amy Duchelle for both supporting me through this dissertation and also for pulling me away from it with her love and carinho. Last but definitely not least, I would like to express my love and thanks to my family for continuously providing me with love and unde rstanding. This accomplishment would not be possible without the strength you give me wh en I am back home in Par, Brazil.


7 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................11 LIST OF FIGURES.......................................................................................................................11 ABSTRACT...................................................................................................................................12 CHAP TER 1 INTRODUCTION..................................................................................................................14 Research Questions............................................................................................................. ....19 Overview of Methods.............................................................................................................20 Structure of the Dissertation...................................................................................................23 Extractive Reserve and the People and Parks Debate............................................................ 29 2 EVOLUTION OF EXTRACTIVE RESE RVES AS A C ONSERVATION AND DEVELOPMENT STRATEGY IN THE BRAZILIAN AMAZONIA.................................. 34 Introduction................................................................................................................... ..........34 Evolution of the ERs Establishment: Contemporary Importance................................... 36 Analytical Framework: A Politic al Ecology Mu lti-Scale Analysis................................ 40 Data and Methods...................................................................................................................42 Results.....................................................................................................................................42 Cross State Analysis: Federal and State Extractive Reserves ......................................... 42 The Evolution of Extractive Re serves Policy: Three Stages ...........................................44 Current Trends: the First Wave States vs the Second W ave........................................... 48 Regional Development Dynamics Favoring or Constraining ER .................................. 50 Continued Demands and New Frontiers for ERs............................................................ 57 Conclusions.............................................................................................................................59 3 ADOPTION OF CATTLE RANCHING AMONG COLONIST SMALLHOLDERS AND FORE ST EXTRACTIVISTS IN THE BRAZILIAN AMAZON: HISTORICALCULTURAL CONTRASTS AND ECONOMIC EXPLANATIONS................................... 62 Introduction................................................................................................................... ..........62 Cattle Ranching in the Brazilian Amazon.............................................................................. 63 Study Cases.............................................................................................................................68 Uruar..............................................................................................................................70 Xapuri..............................................................................................................................73 Field Survey Methods and Data.............................................................................................81


8 Uruar..............................................................................................................................82 Xapuri/Chico Mendes Ertractive Reserve....................................................................... 84 Cattle Adoption Among Smallholder Col onists and Forest Ex tractivists.............................. 85 Socioeconomic Comparisons.......................................................................................... 86 Comparisons of Natural Resource Management............................................................. 93 Comparisons of Pasture and Cattle Managem ent............................................................ 97 Discussion.............................................................................................................................102 Comparative Analysis........................................................................................................... 105 Conclusions...........................................................................................................................108 4 LAND-USE/LAND-COVER CHANGE AMONG RUBBER TAPPERS IN THE CHICO MENDES EXTRACTIVE RESERVE, ACRE, BRAZIL ...................................... 109 Introduction................................................................................................................... ........109 Study Area............................................................................................................................114 Chico Mendes Extractive Reserve................................................................................. 114 Land Use History and Livelihood Trajecto ries Among Rubber Tappers in Acre .........116 Methods................................................................................................................................120 Site Selection.................................................................................................................120 Remote Sensing Analysis..............................................................................................121 Household Surveys........................................................................................................126 Results...................................................................................................................................128 Remote Sensing Results................................................................................................128 Household Survey.........................................................................................................132 Discussion.............................................................................................................................140 Conclusions...........................................................................................................................144 5 WHAT MAKES A RUBBER TAPPER IN THE BRAZILIAN AM AZON?: CULTURAL, LIVELIHOODS AND IN STITUTIONAL-ORGANIZATIONAL FACTORS UNDERLYING SE LF-DEFINITION...............................................................146 Introduction................................................................................................................... ........146 Rubber Tappers Historical Trajecto ries and Definition Cons truction.................................147 Data and Methods.................................................................................................................156 Selection of the Study Sites........................................................................................... 156 Household Surveys........................................................................................................157 Operational Definitions, Measurement a nd Hypothesized Effect of Explanatory Variables ....................................................................................................................158 Socio-economic indicators..................................................................................... 160 Institutional context indicators............................................................................... 162 Land-use indicators................................................................................................ 164 Location indicators................................................................................................. 164 Reserve wide land-use rules................................................................................... 165 Logistic Multiple Regression Model............................................................................. 166 Results...................................................................................................................................167 Analysis of Descriptive Findings.................................................................................. 167 Analysis of the Multivariate Regression Model............................................................171


9 Discussion.............................................................................................................................174 Conclusions...........................................................................................................................181 6 CONCLUSIONS:................................................................................................................. 184 APPENDIX A CORRELATION ANALYSIS.............................................................................................190 B QUESTIONNAIRE USED IN THE CHIC O MENDES EXTRACTIVE RESERVE ......... 195 LIST OF REFERENCES.............................................................................................................212 BIOGRAPHICAL SKETCH.......................................................................................................231


10 LIST OF TABLES Table page 2-1 Number and area of federal and state ERs in the Brasilian Amazon ................................. 44 3-1 Cattle herds and pasture areas in the legal Am azon and Brazil, 1985-1995...................... 65 3-2 Crop cultivation, cattle ranching, and fo rest extractivism in Uruar and Xapuri.............. 73 3-3 Socioeconomic indicators, farm households in Uruar, Par............................................ 88 3-4 Socioeconomic indicators, seri ngal households in the CMER, Acre ................................ 89 3-5 Natural resource management, fa r m households in Uruar, Par...................................... 95 3-6 Natural resource management, seri ngal households in the CMER, Acre ..........................96 3-7 Cattle and pasture management, farm households in Uruar, Par, and s eringal households in the CMER, Acre.........................................................................................99 4-1 Characteristics of the study sites...................................................................................... 121 4-2 Land-use classification schema........................................................................................ 124 4-3 Percent deforestation in six seringais from 1986-2003.................................................... 130 4-4 Percent secondary succession in six s eringais from 1986-2003 ......................................131 4-5 Area and percent remaining in se con dary succession and later reused (swidden/fallow cycle) between 1986-2003....................................................................131 4-6 Percent of households involve d in m arket activitis in 2003............................................133 4-7 Summary statistics from household surv eys: land-use and livelihood statistics .............135 4-8 Livelihood index, deforestati on, and land-use correlations ............................................. 137 5-1 Operational definitions and hypothesized relation ships of explanatory and outcome variables...........................................................................................................................159 5-2 Descriptive statistics for socioeconomic aspects, institutional context, land-use practices, location and land-use rules, am ong households in the CMER........................ 168 5-3 Multivariable model of household defin ition as ru bber tapper regressed on household background, institutional context, land-use practices and regulations in the CMER...... 172


11 LIST OF FIGURES Figure page 2-1 Number and age of federal and st ate extractive reserves in Am azonia............................. 45 2-2 Spatial distribution and age of federal and state E R in Amazonia.................................... 46 2-3 Evolution of land under ER in the first and second wave states of ER system .................50 3-1 Map of the two study cases in the Brazilian Amazon........................................................69 3-2 Change in crop cultivation, cattle ranc hing, and forest extractivism Uruar, PA............ 74 3-3 Change in crop cultivation, cattle ranc hing, and forest extractivism Xapuri, AC............ 82 4-1 Study area and study si tes within the CMER ................................................................... 115 4-2 Location of the study communities within the CMER.................................................... 120 4-3 Trajectories of LUCC in the seringa is in Assis Brasil within the CMER ....................... 129 4-4 Trajectories of LUCC in the seri ngais in Brasilia within the CMER ............................129 5-1 Location of the CMER..................................................................................................... 157


12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TWENTY YEARS AFTER CHICO MENDES: EXTRACTIVE RESERVES EXPANSION, CATTLE ADOPTION AND EVOLVING SELF-DEFINITION AMONG RUBBER TAPPERS IN THE BRAZILIAN AMAZON By Carlos Valrio Aguiar Gomes May 2009 Chair: Nigel Smith Major: Geography In the 20 years since the murd er of rubber tapper leader Chico Mendes and subsequent creation of Extractive Reserves (ERs), ERs have become a popular conservation and development tool in the Brazilian Amazon. Howeve r, ER residents have faced challenges to improving their livelihoods based on sustainable use of the forest, leading to a shift towards cattle ranching and a questioning of rubber tappers status as sust ainable forest managers. This dissertation focuses on three core themes. The fi rst examines overall ER policy evolution over two decades in the Amazon. The second examines issues of land-use/land-cover change and changing livelihoods among forest extractivists in the Chico Mendes Extractive Reserve (CMER) in Xapuri/Acre compared with cattle adoption by smallholder colonists. The third explores the implications of live lihood changes for the evolving self -definition of rubber tappers. This dissertation is guided by theoretical perspectives of political ecology, environment and development, and land-use/land-cover science. It analyzes socioeconomic data gathered through multiple household surveys and government agencies, qualitative information from key informant interviews, and land-cover change detection.


13 Sixty-four ERs have been created in the Brazilian Amazon since 1990, spanning over 12 million hectares. Different waves of ER esta blishment responded to political and regional development dynamics at multiple scales that fa vored or constrained their implementation. In Xapuri, the size of cattle herds rose by 200% from 1990-2005, and throughout the CMER, land cover change detection from 1986-2003 showed some communities nearly surpassing the allowable limits of deforestation. Rubber ta pper households have begun pursuing diverse livelihood activities that include not only extractivism, but also small-scale market cultivation, and increasingly, cattle producti on. Self-identification as a rubber tapper was positively correlated to larger landholding si ze, more knowledge of management rules, and participation in community activities. Conversely, more recen t occupancy in the CMER and, surprisingly, involvement in the official social movement we re negatively correlated with respondents selfidentification as a rubber tapper. This dissertation contributes to the debate of ERs as an environmental policy in Amazon. Despite advances, challenges remain for ERs to pr omote sustainable development, particularly in light of changing livelihoods and continued pressure at development frontiers.


14 CHAPTER 1 INTRODUCTION This dissertation focuses on provocative subjec ts in the debate about forest dwellers conservation and developm ent strategies, changi ng livelihoods conditions, and evolving selfdefinition in the Brazilian Amazon. These subjects are: 1) the evolut ion of the Extractive Reserve as a conservation and development po licy; 2) land-use/land -cover change changes patterns, highlighting the adopti on of cattle ranching by forest ex tractivists; and 3) the evolving self-definition of forest extractivists as sustai nable forest managers in the context of their historical and current deve lopment trajectories. The Amazon contains the largest remaining contiguous forest in the tropics, but also faces some of the strongest development pressures and one of the highest deforestation rates in the world. Development pressures are likely to increase with new large infrastructure projects such as highways and dams, and expansion of agribusiness due to increased demand for cattle, soy, and emerging biofuels. Access to and control of fore st areas will continue to be highly contested by individual and organized smallholders, cattle ranchers, loggers, a nd agribusinesses. Finding solutions that protect forests and forest communities by guarding against both deforestation and displacement is one of the biggest challenges for bala ncing conservation and development in the Amazon, as well as in other tr opical forest regions. The state of Acre in the sout hwestern Brazilian Amazon is the birthplace of the rubber tapper movement, which was the first grassroots m ovement in Brazil to advocate conservation of Amazonian forests through the es tablishment of Extractive Reserves (ERs). Chico Mendes was the major force behind the movement and due to his campaign against forest destruction was killed by cattle ranchers in 1988. ERs, federal lands where non-i ndigenous forest smallholders hold usufruct rights to forest resources, were conceptualized as the rubber tapper agrarian


15 reform (CNS 1985). The ER concepts initial assumption was that an economy based on nontimber forest products could increase the economic value of forests and rural income, offering a variety of subsistence resources for extractivist communities, while ensuring ecological sustainability. In the past twenty years, ERs ha ve gained a solid foothold in Amazonian forest policy and have be seen as a conservation model th at closes off the frontier in contested forest areas by securing spaces for forest-dependent sm allholders and stemming the advance of largescale deforestation processes, th ereby reducing the costs of regional development to forests and people. The ER has come to be seen as the u ltimate, ultra-flexible conservation tool in the Brazilian Amazon. As a result, ER area has increa sed significantly ; it encompasses a diversity of rural social groups with a variet y of forest-based livelihood system s, a range of ecossistems, and strikingly different state political contexts. For example, both sm allholder colonist farmers (e.g., Transamazon Highway) and traditional rive rbank people in the eastern Amazon have successfully lobbied for ERs due to the continue d pressure of large-scale cattle ranching and soybean production at development frontiers. Although large-scale cattle ranc hing has historically been on e of the most important land uses in the Amazon, leaving widespread deforestat ion in its wake and gi ving rise to the rubber tapper struggle, cattle ranching practices have co ntinued to evolve in the region. Today, ranching is not limited to large-scale processes, but has become a practice among smallholders who employ diverse land tenure systems and come from a variety of cultural backgrounds including in Extractive Reserves. While ERs have beco me pivotal to Amazonian conservation policy, cattle ranching has emerged among rubber tapper households in key ER areas, including the Chico Mendes Extractive Reserve (CMER), a symbo lic site of the early struggles and social


16 movements success. This marks an ironic departure from the ERs original role in closing the frontier and preventing large-scale cattle ranching expansion in the region. During the past two decades of ER experi ence, rubber tapper communities views of development and livelihood opti ons have also evolved, conti nuing to respond to changing land uses and development trends occurring outside ER areas. ER populations have faced repeated challenges to improving their livelihoods based sole ly on sustainable use of the forest. Despite several political and development experiments to improve the economic viability of ERs by creating markets for new products extracted from th e forest, forest communities continue to face economic hardship and encounter difficulties in both establishing market opportunities for extractivist products and maintain ing consistent income throughout the year. Due to diminishing returns from extractivism and increasing pressure to diversify production and stabilize seasonal yield, rubber tappers are increasi ngly driven to small-scale cattle ranching and cash-crop agriculture practices, adopting land-use activities that have more destructive impacts on forest cover and exacerbating deforestation in ER areas Cattle ranching in rubber tapper communities have become a major driver of land-cover change in ERs, despite the fact th at it runs counter to the ecologically sustainable vision of the ER and the cultural hist ory of rubber tapper communities in the region. Cattle ranching among rubber tappers might re present the most important dilemma for Amazonian conservation because of its polit ical implications for both the extractivist communities and for the ER model as a conser vation strategy. Given its conflict with the theoretical underpinnings and ecolo gical sustainability of ERs, cattle ranching challenges the conservation objectives of the ER model. Furthe rmore, it throws into question rubber tappers traditional notion as successful forest managers living sustainably and peacefully with the forest,


17 and articulators of a powerful social and envi ronmental justice movement in defense of the forest. This rubber tapper definition was most ac tively vocalized when rubber tapper groups fought together against cattle ranchers and propos ed the ER concept. As a marginalized social group, rubber tappers were relegate d to an isolated, local arena for their grievances, but the urgent threat of large-scale cattle ranching sh arpened their traditional definition and provided them a platform with which to lead a dive rse group of allies across a mobilized global community committed to preserving the environment. In this new context, in which non-timber forest products have failed to be the panacea that was anticipated, the expectations assigned to the rubber tapper movement need to be revisited and potentially rede fined. The stereotype of the rubber tappers as "forest stewar ds" could be called into questio n with their adoption of cattle ranching, but it could only be que stioned with thoughtful consid eration of the socio-economic and development factors that they are faced in Amazonia today. Despite two decades of ER implementation a nd its centrality to people-based conservation, there has yet to be a region-wide analysis of this model. In order to fill this gap, this dissertation first examines ER evolution across space and time in the Amazon by analyzing the expansion of Extractive Reserves over a 20-year period with a multi-scale approach at the federal and state levels. This work also provides an analysis of the political and regi onal development dynamics, showing how these dynamics have affected ER policy and either favored or constrained the establishment of ERs. Given the growing importance of cattle ranching in Amazonia and its expansion among diverse social groups, including many inhabitants of ERs, this dissertation then analyzes the emergence of cattle ranching among smallholders in two ways. First, I conduct a comparative


18 analysis of cattle ranching adoption among colonist smallholders and forest extractivists in the Amazon, exploring why livelihood strategies am ong heretofore distinct social actors are converging in a highly heterogeneous socioeconomic region. To do this, I perform a comparative socio-economic study across space and time of two municipa lities located in the eastern and western Brazilian Amazon, respectively: Uruar, in Pa r state, comprised primarily of colonist farmers, and Xapuri, the birthplace of the rubber tappers movement in the state of Acre. I address the shift toward cattle ranching among these two hi storically contrasting social groups, focusing on economic market forces and ot her socio-economic factors as explanations for expanding cattle raising among groups not pr eviously noted for ranching in Amazonia. Second, I link land-use practices among rubber tappers to remote sensing data, addressing landuse/land-cover changes in the CMER that appear to be the result of forest dwellers changing livelihoods, illustrating that as cattle ranching plays an increasingly important role in livelihood strategies, corresponding growth in pasture is trigging defo restation in the Reserve. Finally, I consider the implications of ch anging livelihoods and land-use practices for rubber tapper definition in ERs. Since cattle ranching have emerged to challenge both the ecological sustainability of ERs and the popular cultural definition of rubber tappers, this dissertation addresses the histor ical-cultural trajectories and current livelihood practices of rubber tappers. I highlight hist orical-cultural factors that shape rubber tapper definition, systematically addressing a series of indicators that represent historica lly important strategies and contemporary changing socio-economic circumstances in the very place where the ER concept originated twenty years ago.


19 Research Questions This dissertation exam ines key questions th at are strongly linked to the central and evolving themes regarding the relationship between forest conservation and livelihood development of rural people in the Brazilian Amazon. The first theme examines a major public policy for sustainable development by offering a temporal and spatial analysis of Extractive Re serve as a conservation and development strategy, along with current outcomes for c ontinuing ER policy in the region. The second theme highlights livestock as the main driver of deforestation as land is occupied in frontier areas, from which emerged the Extractive Reserve concept as an alternate deve lopment strategy. While debates over cattle ranching in the Amazon have ofte n focused on large properties, observers increasingly recognize that social actors asid e from capitalist interests are buying cattle and converting forested land to past ure. As forest extractivists and small colonist farmers increasingly engage in cattle ranching, it is crucial to understa nd the factors that stimulate the expansion of cattle production among diverse groups and the role that cattle production has in providing revenues for either subsistence or a ccumulation objectives am ong these groups. The third theme refers to the complex self-defintion of rubber tappers, which is shaped by their recognition as sustainable forest managers, but faces current controve rsy as land uses and livelihoods continue to evolve in Amazon Extrac tive Reserves. The central questions addressed in this dissertation are: (i) Given nearly two decades of Amazon Ex tractive Reserves, how has the ER model evolved in light of social movements, as we ll as federal and state policies, political and regional development dynamics and interactions among different actors at distinct levels? How have these elements favored or constrai ned the establishment of ERs in different states at different moments in the re cent history of the Brazilian Amazon? (ii) Why do livelihood strategies converge toward cattle ranching among small-farm colonists and forest extractivists despite thei r profound historical and cultural differences


20 and far-flung locations in a highly hetero geneous socio-economic region such as the Brazilian Amazon? (iii) What are the current dominant land use pr actices and livelihoods of extractivists in the Chico Mendes Extractive Reserve, and how do they affect the amounts and rates of deforestation in the reserve? (iv) Given their complex hist orical livelihood trajectories an d notable accomplishments as articulators of a successful social and environmental movement, which institutionalorganizational, cultural and livelihoods factors currently shape rubber tapper self-definition in the face of evolving socioeconomic and development circumstances in the Amazon? Addressing this studys research questions demands the adoption of an eclectic approach that relies on diverse disciplinary perspect ives, a combination of qualitative and quantitative methodologies. In the following I will present an overview of both the research site and methodological procedures developed throughout each chapter of the dissertation. Overview of Methods This dissertation seeks to understand questions about soci o-environm ental processes underway at different scales in the Brazilia n Amazon. The study draws on research methods from the social and land-use/land-cover scie nces, involving multi-sca lar qualitative and quantitative approach. The evolution of an ER model in Amazonia, i nvolved a region-wide approach of data analysis. The comparison of li velihoods among colonist sm allholders in Uruar (Par) and forest extractivist in Xapuri (Acr e), required the aggregation of multi-sited and temporal household surveys. The dynamics driv ing land-use land-cover changes and livelihood transformation within the CMER, relied on the analysis of household survey and quantitative approaches (key informant interviews), as well as land classifications and change detection. Social Methods In order to provide a region-wide analysis of the evolution of the ERs, I created/assembled a database of ERs created in Amazonia from 1990 to 2007. Information for the database was collected through a variety of s ources: the Brazilian Environmenta l Institute (IBAMA) and other


21 state environmental agencies, and re view of the literature news reports and and internet sources. The database is composed of the following inform ation about ERs: year of creation, status as state or federal, size, location, population, and main economic resources. It also contains a GIS based set of information of shape files. For a comparative analysis of livelihoods between colonist families in Uruar and forest homesteads in Xapuri, this work draws on data for two time point s in both sites, and featured households that received follow-up interviews, f acilitating temporal comparisons. Survey data for Uruar refer to 1996 (261 households) and 20 02 (143 households). For Xapuri, survey data refer to 2000, with an initial sample of 66 households, and 2004/2005 with a more extensive questionnaire and a larger sample, totaling 149 ho useholds. Panels were defined on the basis of whether a household was interviewe d at both time points in the two data sets for both sites. The greater detail in the survey da ta and the ability to compare acr oss sites and over time affords a richer portrayal of the circumstances and dynamics of livelihoods transformations among small farm colonists and forest extractivists in isol ated areas of the Brazilian Amazon. More specific details about the survey and statistical tec hniques applied can be found in Chapter 3. While this dissertation used multiple datasets (chapter 2 and 3), I only carried out fieldwork research in the CMER, wh ich is explored in comparative terms in Chapter 3, and is the central theme for the next chapters (4 and 5) Those chapters includes more concentrated analyzes of the 2004/2005 survey data and are the re sult of my intensive fi eldwork carried out in eight seringais (rubber estates) within three municipa lities of the CMER: Assis Brasil (So Francisco, Icuri, Paraguau), Brasilia (Huma ita, Porongaba, Filipinas) and Xapuri (So Joo do Iracema, Indpendencia). Seringais were chosen with the assistance of reserve leaders, and were stratified based on traditional extractiv e orientation versus non-traditional market


22 orientations. The survey data we re collected over almost two year s of intensive fieldwork in the CMER between January 2003 and July 2005. A to tal of 149 household interviews were completed within the eight seringais by the author and field colla borator, J. Vadjunec, in conjunction with a local non-governmental orga nization (NGO), the Group for Research and Extension in Agroforestry Systems in Acre (P ESACRE). Each interview took about three hours to complete, and contained open-ended and struct ured questions regarding specific land-use production and sale of agricultu ral and extractive products, sma ll animal and cattle production, as well as specific questions re garding prospective scenarios, motivations for cattle ranching. Additionally, I gathered detailed da ta regarding labor and capital investments, as well as off-farm income from wage labor, government jobs, and soci al services. I also collected information about household characteristics and history such as hou sehold composition, gender, age, origin, wealth, and access transportation and markets. Finally, I recorded opinions of the householders perceived advantages and disadvant ages as residents of an Extrac tive Reserve, past and current roles of social movement organizations a nd potential development projects for improving householders well-being. Field research in these seringais required intensive fieldw ork and comprised several interwoven activities. Field trips last for 8-12 days at a time with overnight homestays, which always facilitated the applicati on of a mixed-methods quantitativ e and qualitative approach. The qualitative approach involved observations and in formal conversation about controversial issues over land-use preferences, govern ment agencies and institutions, and economic, social, political opportunities and constraints faced by residents of the CMER, as well as others aspects of daily life in the seringal. These c onversations often provided importa nt insight for the structured questions on the questionnaire. Various aspects of these mixed-methods approach inform each of


23 the three stand-alone articles. A series of qua ntitative methods are employed for data analysis, and qualitative information is used to provide a ri cher basis for the interpretation. The data were entered into an excel database and a codebook was cr eated to facilitate sta tistical analysis with SPSS program. In the codebook, approximately 400 variables were defined. More specific details about the survey and methods of data analysis can be found in Chapters 3, 4, and, 5. A copy of the questionnaire utilized for the 2004/20 05 household survey is provided in appendix C. Remote Sensing Linking remote sensing and social sciences is increasingly important due to its potential to understand processes of land-cover change over time and their relationship to socio-economic transformations on the ground. Remote sensi ng techniques were performed to address environmental implications of livelihood transformations on fo rest landscapes by explicitly looking at land-use change trajectories taking pla ce in the CMER. Satellite imagery was used to analyze the amount and rates of forest change at the 30m pixel level in the study areas using a hybrid classification approach developed at the Anthropological Ce nter for Training and Research on Global Environmental Change (ACT). Landsat 5-TM and 7-ETM+ imagery was obtained during the dry season (June-September) for 1986, 1992, 1996, 1999, and 2003. The remote sensing portion of the research is part of an on-going collaboration with J. Vadjunec, and was therefore carried out on the footprint level for Row 02/ Path 67 and Row 03/ Path 67. The remote sensing portion of this research informs the land-use paper. Details can be found in chapters 4. Structure of the Dissertation This disse rtation is presente d as four separate papers (Chapters 2-5), presented in publication style for submission to academic jour nals. Each paper is therefore a stand-alone document, addressing different aspects of ERs a nd forest dweller issues utilizing both unique


24 and overlapping methodologies. Preceding these ar ticles, the current ch apter (Chapter 1) provides an introduction to the research problem and questions, an overview of the research site, and an overview of methods of data collection. It also outlines the theoretical foundations supporting the dissertation highlighting the debate of people-based models of conservation, in which the ER model has featured as a princi pal strategy, but could be undermined by cattle ranching, which is considered the main driver of land-use change in Amazonia, transforming landscapes and traditional livelihoods systems. Chapter 6 presents the main conclusions of this dissertation and explores furthe r directions for research focu sed on Amazonian ERs and their inhabitants. Collaborations, co-aut horship, and target j ournals are explained below, followed by a summary of each article. The first article (Chapter 2), Evolution of Extractive Reserves as a Conservation and Development Strategy in the Brazilian Amazon is the result of on-going research I started as a TA for the class Environment, Developmen t and Social Movements: An Amazonian Perspective, taught by visiting professor Dr. Mary Allegretti at the Center for Latin American Studies at UF. In cooperation with other professors (e.g., Marianne Schmink, Christopher Baraloto) we established a group of researchers inte rested in Extractive Reserve issues, called the Extractive Reserve Network (h ttp://reservasextrativistas.blogspot .com/). I created a database of information about ERs in Amazonia, and this ch apter is a result of this effort. Drs. Marianne Schmink, Mary Allegretti and Christopher Bara loto are co-authors. Target journals are World Development or Development and Change. The second article (Chapter 3), entitled Adoption of Ca ttle Ranching Among Colonist Smallholders and Forest Extractivists in the Br azilian Amazon: Historical-Cultural Contrasts and Economic Explanations, is also led by me w ith co-authorship by Dr. Stephen Perz and Dr.


25 Jacqueline Vadjunec, which resulted from my part icipant in projects coor dinated by Dr. Perz in Amazonia. We analyze cattle ranching by two diffe rent groups of rural groups smallholder colonists and forest extractivists, combining our empirical research sites in two distinct part of Amazonia. Target journal is Economic Geography The third article, (Chapter 4) entitled Land-Use / Land-Cover Change Among Rubber Tappers in the Chico Mendes Extractive Reserve, Acre, Brazil, is the result of an on-going collaboration on land-use land-cover research with Dr. Jacqueline Vadjunec, initiated while we both were developing doctoral fiel d research in Acre (Gomes et al. 2004). First a doctoral student at Clark and then a professor in the Department of Geography at Oklahoma State University, Dr. Vadjunec and I developed our field research toge ther and combined efforts on field surveys, database and land-use land-cover an alyses in the CMER. We also shared ideas, collaborated with local institutions to generate results for local ne eds, and committed to publish papers together as a result of our joined dissertati on efforts in the field. Dr. Vadjunec took a lead role on this paper and I am a second author. An early version of this paper was presented as a chapter of her dissertation (Vadjunec 2007) and a reviewed ve rsion was recently (06/08) submitted to the Journal of Land Use Science. The last stand-alone paper (C hapter 5), What Makes a Rubber Tapper in Southwestern Brazilian Amazon?: Cultural, Livelihoods and In stitutional-Organizational Factors Underlying Self-Definition, is also a resu lt of the collaboration (descr ibed above) with Dr. Jacqueline Vadjunec and Dr. Stephen Perz, both of whom are co-authors to my lead authorship, due to our shared interest in the intrigui ng question we have debated togeth er about the idea of what it means to be a rubber tapper in Amazonia today.


26 As noted, this dissertation is the result of a collective effort with contributions from my professors at UF and colleagues in the field fr om whom I have been fortunate to have the opportunity to share ideas and co llaborative possibilities that ha ve greatly improved the quality of this dissertation. Evolution of Extractive Reserves as a Co nservation and Development Strategy in the Brazilian Amazon. The Brazilian Amazon contains the larg est remaining contiguous forest in the tropics, but it also faces strong developmen t pressures and one of the highest rates of deforestation in the world. In the nearly 20 years since the murder of rubber tapper leader Chico Mendes and the subsequent creation of Extractive Reserves, ERs ar e alive as a conservation and development strategy that strives to secure lands for forest-depe ndent smallholders and stem the advance of large-scale deforestation. To date, 64 federal and state ERs have been created in the Brazilian Amazon, spanning over 12 million ha. De spite two decades of ER implementation and its centrality to people-based c onservation, there has yet to be a region-wide analysis of this model. In order to fill this gap, we analyze th e Brazilian ER experience, showing the growth of ERs over time by state. We emphasize the first a nd second waves of ER policies in the Brazilian Amazon, documenting the land area in each st ate under ER protection and focusing on the federal and state-level political and institutional dynamics that may have favored or limited the growth of the ER model. Although ERs have an imp ressive lineage and remain in the spotlight of environmental policy debates in Brazil, challeng es exist to combat continued pressure at development frontiers. This paper contributes fo r understanding the current condition of the ER model and provides lessons for its future implementation in the Brazilian Amazon. Adoption of Cattle Ranching Among Colonist Smallholders and Fo rest Extractivists in the Brazilian Amazon: Historical-Cultur al Contrasts and Economic Explanations. In


27 recent years, the growth of ca ttle ranching among small-scale ag riculturalists in the Brazilian Amazon has become a focus of policy debates about how to control deforestation. This paper engages this debate by addressing the shift toward cattle ranching by two historically contrasting social groups in the Brazilian Amazon: small-farm colonists and fore st extractivists. We performed a comparative socio-economic study across space and time in two Amazonian municipalities: Uruar, in Par state, comprised primarily of co lonist farmers, and Xapuri, the birthplace of the rubber tappers movement in th e western state of Acre. These two populations differ greatly with regard to their livelihood strate gies, due in part to th eir geographic locations and distinct histories and cu ltures. Survey data from 1996 and 2002 were analyzed for 250 colonist households in Uruar, and from 2000 and 2004 for 150 households in Xapuri. Results reveal some similarities between colonist househ olds and forest homesteads, including family size and association membership, bu t also differences, such as l onger durations of residence in the older eastern frontier in Urua r. Although cattle ranching appear to be more predominant in Uruar when measured in pasture area and cattle herd size, there has been a recent expansion in cattle ranching at both sites: ar ea of pasture increased by 20-30% in both Uruar and Xapuri over the respective study periods. However, in Xapuri the size of cattle herds rose by 200%, whereas in Uruar herds grew by 60%. Th ese findings show that to promote sustainable development in the Amazon, policy makers will need to have greater understanding of cattle ranching dynamics among different smallholder groups, particular ly ranching activity incentives for forest extractivists. Land-Use / Land-Cover Change Among Rubber Tappers in the Chico Mendes Extractive Reserve, Acre, Brazil. The Extractive Reserve System is championed as a win-win model of sustainable development in which rubbe r tappers serve to protect the forest, while


28 improving economic growth and resident well-bei ng. Recently, reserves are being questioned in terms of their environmental sustainability, as many rubber tappers increasingly turn to market agriculture in times of economic duress and instab ility. This study explor es land-use /land-cover change (LUCC) in six rubber estates within the Chico Mendes Extractive Reserve, using both remote sensing analysis and household surveys, a nd addresses the differences in deforestation by livelihood trajectories. A remote sensing an alysis between 1986 and 2003 shows that some communities are close to surpassing the allowable limits of deforestation. Rubber tapping plays less an important role in liveli hood strategies, as welfare is linked to non-extractive activities. Households pursue diverse livelihood activities including extractivism, small-scale market cultivation, animal rear ing, and cattle production. The results suggest that LUCC is highly dynamic in the reserve. What Makes a Rubber Tapper in the Braz ilian Amazon?: Cultural, Livelihoods and Institutional-Organizational Factors Underlying Self-Definition. The Brazilian Extractive Reserve System has been promoted as a sustai nable development alternative for traditional rubber tappers. This paper explores how land-us e, socio-economic and institutional dimensions within the Chico Mendes Extrac tive Reserve (CMER) impact tr aditional rubber tappper selfdefinition. We draw on household survey data to construct statistical models of who selfidentifies as a rubber tapper, a nd employ qualitative information from key informant interviews to substantiate our interpretations of the models. The quantitative resu lts reveal that larger size of land-holding, greater knowledge of management rules, and partic ipation in community activities make identification as a rubber ta pper more likely, while more recent occupancy in the CMER and involvement in the official movement have a negative impact on id entification as a rubber tapper. The qualitative findings in dicate that rubber tapper self -definition is a complex and ever-


29 changing idea, created over time, but strengthened at the height of the ru bber tappers movement in the 1980s and continuously evolving as landuse and livelihood issues continue to emerge. Extractive Reserve and the People and Parks Debate ERs represent one of the few m odels that are created not despite local people but because of them, being premised on and legitimized by the presence of local people and their social institutions (Ehringhaus 2005). ERs therefore lend themselves especi ally well to an analysis of people-based conservation that aims to secure resource access and local livelihoods as well as conserve natural resources. ERs have been a cr itical productive conserva tion strategy used by the Brazilian government and a preferred approach by the social movement, particularly in development frontier regions. Despite its appeal and the large scale of im pact of ERs, the model generated a polemic debate among promoters (Allegretti 1989; Alle gretti 1990; Allegretti 1994; Anderson 1992; Clay 1992; Daly 1990; Schwartzman 1989; Schwartzman 1992) and critics of the model (Browder 1990a; Browder 1990b; Browder 1992a; Browde r 1992b; Homma 1989; Homma 1992; Homma 1993; Salafsky et al. 1993) from the moment of its conception. Mo re recently, ERs have also featured in the resurgence of the global debate regarding the legitimacy of indigenous and other local populations as stewards of protected la nds (Oates 1999; Pimm et al. 2001; Terborgh 1999; Terborgh 2000; Wilkie et al. 2006), the impacts of protected areas on lo cal communities (Adams 2004; Brockington et al. 2006; Cern ea & Schmidt-Soltau 2006; Chap in 2004; Hutton et al. 2005; Schmidt-Soltau 2004; Schmidt-Soltau & Broc kington 2007) and the conservation value and effectiveness of human-inhabited protected area s (e.g., Nepstad et al. 2006a; Schwartzman et al. 2000b; Schwartzman & Zimmerman 2005). Within this context, ERs have been repeated ly used to illustrate points both by supporters of people-based conservation (Campos & Nepstad 2006; Schwartzman & Zimmerman 2005;


30 Schwartzmann et al. 2000a; Schwartzmann et al. 2000b), as well as defenders of more strict conservation areas (Redford et al. 2006; Re dford & Sanderson 2000; Sanderson & Redford 2004; Terborgh 2000). Yet, the global debate continues to be highly dic hotomous, with repeated cycles of irresistible dialec tic (Redford & Painter 2006) pitting strict conservation on the one hand against human wellbeing on the other. On the one hand conservation biologists ar gue that people-oriented approaches to conservation have largely failed to achieve their main goal the protection of biological diversity and advocate for increased efforts towards cr eating large areas for strict protection of biodiversity (e.g., Brandon 1998; Kr amer et al. 1997; Oates 1 999; Peres 2005; Terborgh 1999). These authors have embraced the demystificatio n of ecologically friendly local communities and argue that increased population pressure, access to technolog y, as well as cultural and economic changes have led to practices that are less en vironmentally beneficial. These arguments clash with defendants of indigenous and local peop les land rights (e.g., Br osius 2004; Ghimire 1994; Schmidt-Soltau & Brockington 2007; West et al 2006; Wilshusen et al. 2002) and of the conservation value of human inhabited area s (Colchester 2000; Nepstad et al. 2006a; Schwartzman et al. 2000a; Schwartzman et al. 2000b). Defenders of indigenous and other human-inhabited area argue that protectionists un dervalue the political strength of grassroots organizations to demand conservationist policies (Schwartzmann et al. 2000b) and ignore key social aspects of social and po litical processes that shape how conservation interventions happen in specific contexts (Wilshusen et al. 2002). These critics also argue that this approach neglects not only the history of local comm unities but also the history of the landscape in which people play a central role. Many authors stress the larg e-scale conservation value of indigenous areas, ERs, (Goeschl & Igliori 2004; Nepstad et al 2006a; Peres 2005; Schwartzman & Zimmerman


31 2005) and other human-inhabited areas including private lands (West & Brockington 2006), although some of these authors te nd to downplay the potential e nvironmental impacts of rural communities through hunting, small-scale logging and agriculture. Specifically hunting has been the focus of the debate for the critics who are mostly concerned with effects on game, while defenders argue that these human-inhabited protec ted areas are avoiding much larger problems, such as deforestation. The strong divide between hopeful promoters and cynical opponents of ERs erected steep walls of dichotomies between s aving and losing the tropi cs and between success and failure (Browder 1990b; Salafsky et al. 1993), leaving little r oom for the discussion of intermediate possible outcomes and complex r ealities in which conservation and development efforts are situated. Recognizing the common goals of fighting agai nst the larger threats from extensive agriculture, industrial forestry, plantati on establishment, and broadscale fires to both biodiversity and marginalized forest dwellers, authors are increasingly trying to escape this highly dichotomous debate (Redford & Painter 2006) around whether people should be included or excluded from protected areas. An interdisciplinary middle-field is emerging th at addresses the social, political, cultural and ecological complexities and contradictions in conservation and development efforts and attempts to eschew extreme simplifications (Berkes 2004; Brosius 2004; Redford & Brosius 2006; Redford & Painter 2006; Redford et al 2006; Sanderson & Redford 2004; West & Brockington 2006). Those critical analyses of both conventional and community-based conservation are moving towards a more differe ntiated understanding of social-ecological interactions and the inevitable trade-offs betw een human use and biodiversity conservation and the equitable distributio n of both benefits and costs of conservation.


32 While opinions on ERs on this debate have al ways been strong, few publications (Brown & Resende 2000a; Brown & Resende 2000b; Ruiz-P erez et al. 2005) have actually included on the ground, empirical research that evaluates ERs and only one looked at ERs in a comparative context (Nepstad et al. 2006a). Most of the protagonists of the large debates have never conducted research in ERs and base much of their argumentation on both outdated information and scientific hearsay, resulting in selective and skewed represen tation. Also, the recent debates of ERs refer almost exclusively to pre-extractive rese rve papers (Ehringhaus 2005), not taking in account the now 20-year history of ERs and the si gnificant changes in the process. In view of this dearth of published information and the influence that ERs had on people-based conservation debates, an update of the current state of the ER mode l is long overdue and critical for further discussions in the broader cont ext of the people an d parks debate. More recently, a significant body of literature representing a diversity of themes based on empirical studies has slowly documented the tran sformations of the ERs, revisiting old themes and emerging challenges surrounding forest dwelle rs economic and forest resource management practice through the ER experience in Amazonia. Some of those aspects include transformation of traditional livelihoods strategies in ERs, and how these changes affect traditional livelihoods systems (Gomes 2001; Souza 2006; Wallace 2004) ; ecology and management of non-timber forest products, especially brazi l nut, given its socioeconomic importance for ERs areas (Cotta et al. 2008; Kainer et al. 1998; Kain er et al. 2007; Wadt et al. 2005; Wadt et al. 2008); communitybased timber management as an emerging economic option surrounding forest extractivist livelihoods strategies (Humphries & Kainer 2006 ; Rockwell et al. 2007a; Rockwell et al. 2007b; Stone 2003); in addition to the broader context of how changes in the la rger political-economic context have affected rubber tappers in ER, the roles of grass-roots institutions in ER


33 management and governance, government incenti ves and political empowerment (Ehringhaus 2005; Vadjunec 2007). Most of those studies have b een carried out on the first ERs created; the ERs that played a pivotal role in the people-based conservation debate. As so, those empirical studies are significant not only to inform the main protagoni sts on the debate but al so to provide a more realistic picture of transformations and rules th at ERs have played out over the 20-year as a strategy for conservation and development of Amazonia. As ER area has increased significantly in recent years, it encompasses a diversity of populations with a variety of forest-based livelihood systems, a range of ecological niches and varied state political contexts. In this context, there is a great need for further resear ch on ER areas and to synthesize the contribution of the diverse empirical studies carried out more recently and how they contribute to the needed region-wide analysis of the evolvi ng ER model. Chapter 2 of this di ssertation is an effort in that direction, as it analyzes the grow th of ERs over twenty years, drawing on an analysis of the evolution of the model and discussing the divers e contexts in which ERs have been implemented in Amazonia. Chapter 3, 4 and 5 analyses emergi ng themes related to livelihoods transformations and land-use land-cover changes within the Ch ico Mendes Extractive Re serve as issues of central importance for the ER de bate due to their potential la ck of harmony with both the theoretical underpinnings and eco logical sustainability of ER.


34 CHAPTER 2 EVOLUTION OF EXTRACTIVE RESE RVES AS A C ONSERVATION AND DEVELOPMENT STRATEGY IN THE BRAZILIAN AMAZONIA Introduction Brazil, as owner of both the largest portion of th e world' s rainforests and the highest absolute deforestation rate, is de facto a leader in both conservation and destruction of these forests. The state of Acre in the southwestern Brazilian Amazon is the birthplace of the rubber tapper movement, which originated in the late 1980s when a group of rubber tappers fought to protect their land against encroaching large scale cattle ranchers. This wa s the first grass-roots movement in Brazil to advocate conservation of Amazonian forests throu gh the establishment of Extractive Reserves (ERs). ERs are forest area s inhabited by extractive populations granted longterm usufruct rights to resources which they manage (Allegretti 1989). Chico Mendes was the major force behind the movement and because of his campaign against forest destruction, he was killed by cattle ranchers in 1988. Th e creation of ERs as one of the first formalized systems of people-based protected areas in the Brazilian Amazon has marked an unprecedented success of both social movement mobilization and enviro nmental policy-making in the Brazilian Amazon (Allegretti 1994; Schwartzman 1989) and has been promoted as a major strategy for forest conservation while simultaneously providing sust ainable economic returns to local people. Twenty years since the assassi nation of Chico Mendes and crea tion of the first ERs in the region, the model has gained a solid foothold in Amazonian forest policy, while at the same time evolving and diversifying significantly beyond the original con cept. The question of what happened after establishment of first ERs is complex and needs to be addressed in a contemporary context. Since the creation of the first ER in1990 in the Western Amazon state of Acre, ERs today encompass a great diversity of populations, with a vari ety of forest-based livelihood systems, and living in a range of ecolo gical niches and under va ried state political


35 contexts. The model has been implemented in all Amazonian states as both federal and state level strategies, which have been used at different points in tim e and undergone important changes. More importantly, the ER model still maintains its key principal of land conflict resolution and has become a major land tenure strategy advocated by different socio-cultural groups in distinct Amazon ecosystems. ERs, are one of the few models created not despite local people but because of them (Ehringhaus 2005). More th an in any other conservation and development approach, ERs are premised on and legitimized by the presence of local people and their social institutions. Environmental legislation in Brazil has dram atically improved over the last few decades. The creation of ERs represented a huge change in environmental law in the Brazilian Amazon, and environmental legislation, governance and ma jor governmental investment in conservation has improved substantially in the region with innovative policies observed at both federal and state levels. While the Brazilia n national government takes a broad approach for conservation and development issues and has developed a major program for creating and supporting conservation units, state governments have specific development and environmental agendas that have both constrained and supported the ER model in the region. As federal and state environmental policies are increasingly integrated in the Brazilian Amazon, it has become important to understand how the ER model has evolved at different scales, and what role ERs have played in a more comprehensive conser vation and development policy in Amazonia. Although considered a victory for forest dwellers, the evolving ER model has been fraught with challenges, and political support is key to its long-term success. In view of almost two decades of existence of ERs, an analysis of their evolution and an update of the current state of the ER model is long overdue and critical for the further


36 discussions of the implication of current outcome s for continuing ER policy in the region. In this article we use a political ecology approach to explore what is happeni ng in ERs by focusing on the evolution of the ER model through a multi-scale analysis at the federal and state levels. We offer an analysis of ERs evolution and a discus sion of the political and regional development dynamics of each Amazonian state. We analyze how the ERs have evolved in light of social movements and political support that have favored or constrained the establishment of ERs in each state. Evolution of the ERs Establishment: Contemporary Importance Since the establishm ent of Brazils first ER on the Upper Juru River in Acre, the western most state of the Brazilian Amazon, the concep t has spread east thr oughout the entire Amazon, stretching to the mouth of the Am azon River and the Atlantic coas t in Par. The first ER, Alto Jura, was created by presidential decree, imbuing it with a speci al status. The federal decree number 98. 987 (01/30/1990) was the first legal instrument to recognize ERs in Brazil, in which there was a co-management of the ERs by th e Brazilian government and the resident associations, regulated by a Utilization Plan. The legal instruments of ER management evolved to the broad environmental regulatory law that now includes all conservati on units in Brazil. In 2000, the National System of Protected Areas law (SNUC law 9.985) was created, bringing together the diverse models of protected areas at the federal, state and municipal levels. These units were classified by two major categories: 1) the strictly prot ected model, with biodiversity conservation as the principal objective; and 2) the sustainable use model, which allows for varying forms and degrees of exploitation, a nd has biodiversity protection as a secondary objective (MMA 2000; MMA 2002; Ryla nds & Brandon 2005; Silva 2005).1 The SNUC law 1 The strictly protected areas (as defined by SNUC) include national parks, biological reserves, ecological stations, natural monuments and wildlife refuges. There are 111 of these federal areas totaling 28, 245.720 ha (42% of all


37 clearly differentiates when specific protected areas should focus on preserving areas of high biological importance or satisfy social demands, especially as development frontier expands in the Amazon. The terms strictly protected or sustainable use are also substitutes to the terms used before creation of the SNUC law: conservation units of indirect use (strictly protected) versus those of direct use (sustainable use). The ERs fall under the direct use model, or as some prefer to label them, people-based conservation model. The SNUC law is innovative in its establishment of management pl ans and deliberative councils as the major instruments for regulation and decision making within ERs. These mechanisms bring together a diverse set of local and regional stakeholders, which provides a broad development perspective, but also diminish the role of local ER commun ities in resident associations. From their conception, ERs have redefined the conventional goals of conservation. Over time, their implementation has brought local people to the forefront of c onservation in protected areas and led to a restructuring of the nationa l and state environmental institution apparatus, establishing the traditional peoples category as legitimate stakeholders in environmental policy and conservation strategies. The ER model has faced strong opposition, especially during its initial stages. The most stri dent critiques of ERs were produced by Browder (1990a; 1990b; 1992a; 1992b) and Homma (1989; 1993), and continue to be cited as grounds for criticism, despite the lack of updated information. Despite criticism, for the initial strongest proponents of the ER model, the Brazilian anthropologi st Allegretti (1989; 1990; 1994; 1995) and the American sociologist Schwartzman (1989; 1991; 1992) who publicized the rubber tappers federally protected areas) The sustai nable use areas include environmental protection areas, areas of particular ecological interest, national forests, extractive reserves fauna reserves, sustainable development reserves, and private natural heritage reserves; there are 141 federal prot ected areas for sustainable us e totaling 30,194.984 ha or 58% of all federally protected areas (Rylands and Brandon 2005; Silva 2005). Those numbers show a balance in terms of area protected between those two categories. But th ere is a major difference when they are compared across biomes of the country, especially in the Pantanal the Cerrado and the Atlantic Forest. Only in Amazonia is there an approximate balance between strict (49%) protection and sustainable use (51%) (Rylands and Brandon 2005).


38 cause, ERs continue to represent a very vibran t example of a innovative policy that balances conservation and development among traditiona l communities (Allegretti 2002; Schwartzman & Zimmerman 2005). ERs were first thought of as a federal m odel of agrarian reform while promoting conservation. But soon after the first federal ERs were created, a state-level approach was created following the same principles as the federal model. The principal difference in the approach concerned the institut ional arrangements for establis hment and monitoring of ERs. While the federal ERs are create d by presidential degree and the Ministry of the Environment plays a major role in their administration, St ate ERs are created by st ate governments with approval of the State House of Representativ es. Both approaches, however, depend on the mobilization of social movements and their capacity to negotiate at both state and federal political arenas. In both cases, local politic al will is necessary to stabilize the ER. In the Brazil Amazon, the ER model has pave d the way for the creation of other peoplebased protected areas, as well as hybrid land tenure models (Ehr inghaus 2005), that are being implemented by various institutions at both the federal and state level, and respond to changing political contexts and opportunities. These sustainable-use prot ected areas include marine ERs (in Amazonia and along the Atlantic coast of Br azil), Extractive Settle ment Projects (PAE) which are forest-based, land reform areas that pr eceded ERs, Sustainable Development Reserves (RDS), Sustainable Settlement Projects (PDS) an d more recently Forest Settlement Projects (PAF). Although the ER model a ssociated with the Brazilian Institute of Environmental Protection (IBAMA) the implementation arm of the Ministry of the Enviro nment (MMA), is still the most well-known type of sustainable use areas today many state and federal institutions, including the National Agrarian Reform Institute (INCRA) of the Brazilian Ministry of Agrarian


39 Development (MDA), are involved in the architectu re of institutional arrangements in defining and creating new sustainable use areas. In 2007, the MMA created the Chico Mendes Institute a government branch responsible for creating an d managing all conservation units in Brazil, which seems to be a clear demonstration of the ro le ERs have played in influencing conservation policy in Amazonia, and demonstrates the pol itical empowerment of grassroots social movements in the region. Despite the institut es naming after a rubber tapper martyr, the transformation of the ER model to a governmental institution can al so be seen as a confused mix of government and social movement partnership configurations. Furthermore, the ER model has been expande d and been applied to a diverse range of ecological and social contexts. It has moved beyond forest envir onments to riverine floodplains and marine ecosystems and now encompasses dive rse social groups with distinct historical backgrounds. For instance, while originally only r ubber tappers were entitled to ERs, today the concept has been applied to ot her extractivist rural workers in Amazonia. Given the increasing number of ERs of different types, there is a great diversity in si ze, population density and background, ecological and cultura l context, livelihood strategies, social organization, market access and development pressure among the different ER sites. Aside from proposing the ER model, the rubbe r tapper movement has evolved from being a movement of powerless workers to a powerhouse influencing both environmental policy and land reform with its activists operating in a nd contributing to local, regional and national governments, legitimizing its philosophy of social justice. Rubber tapp er institutions are represented in Amazonian states and have helped to establish a broad network of rural workers organizations, including the powerful Grupo de Trabalho Amazonico (Amazon Working Group),


40 created in 1992 with twenty grassroots organi zations. Today, this group has over five hundred small organization members based on eighteen branches in Amazonia. The rubber tapper movements success has come as a result of its efforts to pioneer strategies to open new political spaces, create lo ng-lasting partnerships with other social groups, to establish the successful and dynamic concept of ER, and develop the capacity and flexibility to adapt to diverse social and political c ontexts in Amazonia. In 2005, the rubber tapper movement celebrated twenty years, with the Congr ess of Extractivist Peopl es of Amazonia. With the participation of ove r three hundred extractiv ists and representati ves of a variety of institutions, the three-day meeting hold in Manaus focused on intense discussion about the experience and the future of ERs and other related protecte d model areas in Amazonia. Addressing old and new themes, the meeting addresse d five central topics: (i) land tenure issues, (ii) production, credit and techni cal assistance, (iii) social or ganization and management, (iv) management plans, and (v) payment for envi ronmental services. Th e discussion surrounding those themes demonstrates that in those twenty years, new and old challenges are intertwined within an agenda of a social movement still y earning for success, despite its accomplishment in those two decades. Analytical Framework: A Politic al Ecology Multi-Scale A nalysis Advances and challenges to the ERs in Brazi l have been the product of complex forces interacting over time at divers e levels. This complexity calls for a political ecology framework of analysis that considers the wide interactions of actors with each other and the environment (Bryant & Bailey 1997: 191). Poli tical ecology acknowledges the human production of nature and political forces behind such production. We present a multi-scale analysis of the ER mo del, focusing first on social movements that support them, as well as federal and state policies, international pressures, and how actors at


41 different levels interact in th e creation of ERs. We take into account th e changing dynamics of politics, policy, and social move ment strength in explaining the different forms ERs have taking in different states at differe nt moments in recent history. Previous studies have shown the fundamental importance of social movements in creating the original ERs, as well as in proposals for other innovative social-e nvironmental policies in Brazilian Amazonia (Allegretti 2002; Allegretti & Schmink forthcoming; Becker et al. 1990; Hall 1997). Our analysis therefore examines to what extent the implementation of ERs has been associated with strong social movements in different states at different periods. Favorable political support at federal, state, and local levels, as they interact with international political pressures, also plays an important role in determ ining policy outcomes. We therefore play close attention to power relations and to the role of the state, over time, in promoting development initiatives that may directly or indirectly lead to success or failure of ER policy. Many of the outcomes of Amazon development policy can be understood as the outcome of social conflicts and disputes over deve lopment models and practices (Schmink & Wood 1992). ERs were created because of people and ha ve been implemented in a very politicized environment that often involved disputes and power relations among distinct political forces and social actors. The study therefore focuses on ERs as part of emerging negotiations among different interests regarding the future of Amazonian territories. We explore the evolution of th ree key themes related to th e political ecology of ERs in Brazil social movement strength; political will at diverse levels, and emerging forms of negotiation as they help to e xplain the evolution of ER policies in different Amazonian states over the past two decades. The findings suggest that ER policies have contributed not only to the


42 land tenure security of forest peoples in Amazonia, but also to the emergence of new forms of dialogue over Amazonian policy. Data and Methods We elaborated a database of ERs created in Am azonia from 1990 to 2007. Information for the database was collected through a variety of sources: requests for information from the Brazilian Environmental Institute (IBAMA) and state environmental agencies, attendance at local meetings and workshops, and review of the literature, news reports and data found on the internet. In addition, in terviews with key informants we re conducted, including government officials and social movement leaders. The database is composed of a series of information about ERs: year of creation, status/state or federal, size, location, population, main economic activities among others. Geo-referenced information (sha pe files) was also collected through online database of the IBAMA and database from the Institute for Amazonian Research (IPAM). We analyze the evolution of ERs temporally, in addition to spatially across all states in the Brazilian Amazon. We first provide an overall cross state analysis of growth and distribution of ER. Then, we further disaggregate the data thro ugh examining time of ER creation focusing our analysis on temporal progression of ER implem entation. We discuss the evolution of ER over three time periods: 1) First wave of ERs (19901995); 2) Transition period (1996-2000); and 3) Second wave of ERs (2001-2007). Results Cross State Analysis: Federal a nd State Extractive Reserves The Brazilian legal Am azon, a concept created for economic development planning in the region, includes all states in north ern Brazil, as well as the state of Mato Grosso and a portion of Maranho. This corresponds to approximately 5 million km2, which constitute 60% of the total Brazilian territory. As of January 2008, there we re 39 federal Extractive Reserves in Amazonia


43 covering an area of almost 10 million ha and 25 state reserves covering approximately 2.4 million ha. This is a total of 64 ERs that encompass an area of more than 12 million ha in eight states of the Brazilian Amazon, as observed in Table 2-1. This figure, which shows the distribution of ERs by state, reveals the following : Acre has 5 federal ERs, Amap has 1 federal ER, Rondnia has 22 state and 4 federal ERs, Maranho has 4 federal ERs, Tocantins has 1 federal ER, Mato Grosso has 1 state ER, Par has 17 federal ERs, and Amazonas has 2 state and 7 federal ERs. This figure shows a varying appr oach of ER establishment in the region, as observed through the creation of 22 state level ERs in Rondnia and 2 in Amazonas, as well as the creation of Marine ERs on the Atlantic coast of Pa r and Maranho states. Moreover, this figure shows the absolute ar ea under ERs in each state and what this represents in percentage of state territory. It reveals that th e state of Par w ith approximately 3.8 million of ha under federal ERs, has the highest absolute amount of land under ERs, followed by Acre with 2.7 million of ha of federal ERs. The states of Amazonas and Rondnia, which initiated the creation of ERs at the state level, appear with similar figures in terms of land protected under ERs, with 2.5 million and 2.4 million ha respectively. Nonetheless, Rondnia has the greatest amount of area unde r state level ER protection (over 1.9 million ha), while in Amazonas the vast majority of ER lands are federal (over 2.1 million ha). This figure changes when comparing which states have the highest pe rcentage of their territo ry under ERs. The state of Acre, the birthplace of the ER model, appears with 17.7% of its territory in ERs, the largest percentage of all Amazon states, followed by th e state of Rondnia with 10.1%, and Amap with 3.4%. Those are all relatively sm all states in the Brazilian Amaz on and where the ER model was first implemented, starting in 1990. Par and Amaz onas, with 3% and 1.6% respectively, on the other hand, are the two largest Amazonian states and only in the later 1990s did the ER model


44 expand to those states. These features do not al low for further discussion of the evolution and recent trends of ERs establishment in the region and thus require further disaggregation of the data through examining time of ER creation. Table 2-1. Number and area of federa l and state ERs in the Brazilian Amazon Amazonia States Level Number State area (%) Area (ha) Acre Federal 5 17.73 2,704,353 Amap Federal 1 3.37 481,650 Rondnia State 22 8.25 1,959,473 Federal 4 1.87 444,011 Maranho Federal 3 0.09 29,628 Marine 1 0.56 185,186 Tocantins Federal 1 0.03 9,280 Mato Grosso State 1 0.05 49,029 Par Federal 7 2.64 3,288,229 Marine 10 0.36 443,451 Amazonas State 2 0.23 367,339 Federal 7 1.38 2,163,529 Sub-total State 25 2,375,841 Federal 39 9,749,317 Marine 11 628,637 Forest 53 11,496,521 Total 64 12,125,158 The Evolution of Extractive Reserves Policy: Three Stages In this section, we focus on the tem poral progression of ERs implementation throughout the Brazilian Amazon. Figure 2-1 s hows the ages of federal and state ERs, and Figure 2-2 gives a temporal and spatial distribution of ERs across st ates over three times peri ods. The first wave of ERs (1990-95) is viewed as an innovative, bo ttom-up policy process that, despite the debate stemming from this conservation and developmen t strategy, the potential of the ER model did not reach a large scale. The first wave period re presented an initial push in 1990 and led to the creation of four federal ERs in three different states; two in Acre (the Upper Juru River and Chico Mendes), one in Amap (Rio Cajari) and one in Rondnia (Rio Ouro Preto), which cover approximately 2.2 million ha. After this first pu sh, the total area of land under federal ERs


45 changed very little, despite the crea tion of four new federal ERs in two states in 1992. In the state of Maranho, three small ERs were created, cove ring a total of approxi mately 30.000 ha, while in the state of Tocantins an ER of approxima tely 10.000 ha was created. The biggest change during those five years came with the adoption of the model to create the first state level ER in Rondnia, which in 1995 alone, created 22 state ERs covering approximately two million ha. Therefore, the first five years of ER experi ence were characterized by an initial push of establishment of four ERs in three different st ates, and an adaptation of the federal model, represented by the creation of state level ERs in Rondnia. Figure 2-1. Number and age of fe deral and state extractive reserves in Amazonia: abbreviations of the names of Amazonian states are as follows: Acre-AC, Amap-AP, RondniaRO, Maranho-MA, Tocantins-TO, Mato Gr osso-MT, Par-PA, and Amazonas-AM. The F or S following the abbreviations sta nds for federal or state-level ERs.


46 Figure 2-2. Spatial distribution and age of fede ral and state extractive reserves in Amazonia During the transition period (1996-2000), an ad ditional four federal ERs were created, covering an area of over 1.2 million ha. This period of ERs is characterized by the application of ER systems in the two biggest states in Amazonia where no previous ERs had been before. In 1997, the state of Amazonas created its first ER, M dio Juru with an area of over 250 thousands ha. Then, in 1998, the 674,000 ha Tapajs-Arapiuns ER was created in the state of Par. Another characteristic of this period was that in 2000, completing ten years of ER experience in Amazonia, the federal model was applied again in th e first time in two of the states where it was first established, when the Alto Tarauac ER with approximately 180,000 ha was created in Acre


47 and the Lago Cuni ER with over 50,000 ha in R ondnia. This period is considered a transition, because there was little change in ER area in the first wave states, but the model began to be applied on the two largest states in the region, ma king a huge contribution to the expansion of the ER model during the following peri od. In the transition period, th e social movements created an agenda for the broad application of ERs in the region. This was a period in which the concept spread and generated discussion among actors on multiple scales, from traditional communities to political decision-makers. The second wave of ERs (2000-07) represented a huge increase in ER areas with the establishment of 29 ERs covering approximately 6.9 million ha of land. This represented an increase of over 50% land under ERs system on the previous two periods. During this period, only in 2007 were neither Federa l nor State ERs implemented in the region, while in all other years there was significant establ ishment. 2003 represents the year in which the least amount of land (216,874 ha) was designated as ERs, with the implementation of only one state level ER in the state of Amazonas, and 2004 shows the high est amount with over 2.5 million ha of land designed as ERs. This represented an increase of over 1.1 million ha of land designed as ERs every year during this period. It can be observed that the state of Pa r alone provided an important contribution to this large increase: 16 federal ER s were created covering an area of approximately 3.1 million ha. This included the creation of 10 Marine ERs in mangroves of approximately half million ha. Nine of Pars ERs were created between 2005-07, representing more than half its total ERs thus far. Following the trends of Par state is the state of Amazona s, although with less extension. During the second wave period, th ere were six federal ERs created, covering an area over 1.9 million ha. In addition, Amazonas state took a stat e level approach, as formerly observed in


48 Rondnia and created two state le vel ERs, covering an area of 368,000 ha. This increase in ER land during that period was also a result of tw o new ERs established in Acre, Cazumb-Iracema and the Riozinho da Liberdade, which were created respectively 2002 and 2005. This increased the state area under federal ER by over one millio n ha. Small changes were also observed in the state of Rondnia with the establishment of two small federal ER s in 2001, covering an area of 183,578 ha. Despite of almost two decades of ER experience in Amazonia, the average ER is only four years old, demonstrating that th e concept has just recently taken off. During this final period, we obser ve a widespread application of the ER model in the states of Par and Amazonas, where the model did not or iginate, but incorporat ed the policy during the transition phase. It is also obser ved that ER concept gained diffe rentiation in its application as seen through the creation of ma rine ERs in Par. The approach of ER implementation was increasingly top-down despite cont inued engagement of the social movement in the process. This is likely a result of the populari zation of ERs model at multiple scales, from regional grassroots organizations to governmental acceptance of th e ER model as a productive conservation and development strategy for the region. Current Trends: the First Wave States vs the Second Wave Figure 2-3 shows that in seventeen years of ER s policy in the first wave states a total of 5,862,610 ha have been designated under ER, while th e second wave states with one decade of ER policy history have reached a total of 6,262,549 ha of land unde r the ER system. While the concept is very much alive in the first wave st ates, such as Acre, the biggest expansion of the model took place in the second wave states, especi ally in the last six years. In terms of percentage of land under ER, the first wave stat es show a greater per centage than the second wave states. When it comes to absolute number of area under ERs, the second wave states have a greater number than the first wave states. This information gives us insight into future trends for


49 ER growth. It is unlikely that ER s will continue to grow in number in the first wave states, such as in Acre where almost 20% of its territory is under ER area. Time (alm ost two decades) and the political context in Rondnia give no indication of a continuing establishment of ERs in this state. The ER model in the small, first wave st ates could be exhausted, while in the second wave states, the model will continue to be consider ed. Despite the impressive advantage of absolute land area, the second wave states have recently cr eated, especially in Par, it is far below what the state of Acre has already done in dedicating a larger percentage of its territory under the ER system. This discussion of ERs does not intend to in fer about which state in the region has made a greater contribution to the system of conservation units in the Brazilian Amazon, but instead to compare where and when ERs have been demanded. For the former debate, a much broader perspective on all models of conservation units under the SNUC law would need to be taken under consideration. The ER system is a pioneer model under the directuse protected areas, however there has also been an advancement of strict protected areas in the region (Silva 2005, Rylands and Brandon 2005, Ribeiro et al 2005, and Borges et al 2007). However, under the direct-use protected areas the ER system as a pioneer has paved the way for the creation of other direct use conservation unites. In this context, if ER is not exactly the choice model of direct conservation units in some states, it provided the foundation for the creation of several other conservation unit mode ls in the region. This is an impo rtant advantage of the ER system, but can also represent a loss for what ERs predic t for sustainable use of forest resources. The state of Amazonas, for example, has made an important contribution in the expansion of the Sustainable Development Reserve (SDR), whic h was created under the foundation of the ER system, with a focus on state level initiatives. Si nce 2003 the state of Amazonas has created


50 seven SDRs covering an area of approximately 4.7 m illion ha of forest. This represents an area 86% bigger than all the federal and state level ERs in the state, although the ER system has a longer history in the state. This is a clear demons tration that the ER mode l is not the only option available, and that there is a competition of dir ect-use protected areas occurring and that state governments in the region are also making political choices for designation of a specific model. Figure 2-3. Evolution of land under ER in the fi rst and second wave states of ER system Regional Development Dynamics Favoring or Constraining ER The ER m odel was initially impl emented in the first wave stat es due to land use conflicts provoked by sponsored government programs. While in Acre, the state government promoted the occupation of former rubber tapper estates to migrant cattle ra nchers from southern Brazil, Rondnia was under alarming deforestation rates a nd land conflict with indigenous peoples and 0 1 2 3 4 5 6 7 8 9 10 11 12199019911992199319941995199619971998199920002001200220032004200520062007yeararea (mi ha) First wave of ERs Second wave of ERs Accumulated Transition Second wave First wave 1990 1995 2000 2007


51 rubber tapper population due to the paving of the BR-364 highway funded by the World Bank. Meanwhile, on the opposite side of the Brazilian Amazon, Brazil nu t collectors in Amap were being threated by project Jari, a major cellulose production project in th e southern portion of the state where the greater major ity of rural households as well as the greater expanses of Brazil nut groves were located the major forest extractivist product in the state. Of the three first wave states of ER establishment, Acre has shown the strongest co mmitment to the ER policy. While the ER policy has evolved over the past 15 year s, in Rondnia it shows there was only a brief period of ER establishment, as well as in Amap, where the establishment of ERs did not evolve at all. The political success of the rubber tapper m ovement in creating the concept of ERs transformed politics in the state of Acre. Jorge Viana, a forester and one-time close adviser to Chico Mendes, was elected governor in 1998 and reelected in 2002. In the last decade, Acre government launched innovative policies to strength en the extractiv ist economy within the state (Kainer et al. 2003). Support of ERs are a majo r component of goverment approach, and thus from 2000 to 2006, a new cycle of establishment of ERs blossomed as three new federal ERs were created. Acres forest government highlight ed the social component of local and regional development, as evidenced in the governments innovative forest conservation and development paradigm Florestania or forest citizensh ip, built around a recogniti on and appreciation for local knowledge systems. Acre policies greatly increased representation of forest peoples in regional politics, involved rubber ta ppers and other extractivists populations more directly in the administration of forest resources and developm ent initiatives, and integrated rubber tapper culture into regional society, creating a spirit of participation, collaborati on and social awareness of what is becoming a shining example of governance sparked by the movement initiated by


52 Chico Mendes. It is fair to say that Acre exemplifies the most vivid scenario of the ERs model in the Amazon. It is unlikely that the focus of the new policies will be in creating new ERs areas, but rather to improve livelihood conditions of the extractivist population living in established areas. Rondnia took steps of not only becoming the fi rst state to adopt the state-level ER approach, but also designated a re latively larger area for this fi rst initiative. In Rondnia, this initial government policy apparen tly demonstrates that the Rondnian state government of the mid-1990s was committed to the development goals of ERs. One must ask what factors in political development brought about this remarkable course of action, since then there has been no indication of continuity in the process of the esta blishment of reserves. These state ERs were created in the contex t of the Planafloro Program (1993-2002) which was funded by the World Bank as a response to pr evious major investments by the Bank that had major environmental and social impacts in the region2. The major goal of the program was to protect biodiversity thro ugh a zoning program and create a diverse system of conservation units: the state ERs were one of the models proposed by the social movement. The option for state-level reserves was also a source of debate; even though the areas designed for the state reserves were federal la nd, the decision to statelevel approach was due Federal ER would take longer to be implemented and grassroots organization could lose the political momentum of social movement pr essure supported by external donors on the state government for the legal designation of those areas. By adopting an innovative state-level model, the local social movement guaranteed that at least the rubber tapper co mmunities would acquire 2 In the late 1980's, the Bank's POLONOROESTE road co nstruction (highway BR 364) project that opened the western Brazilian Amazon to slash-and-burn agriculture, logging, and cattle ranching gained international attention due to the environmental and social damage it provoked in the region and is widely considered today one of the worst ecological disasters ever supported by the Bank.


53 immediate rights to the land. As Ademir de Melo Uchoa, a leader of th e Organization of the Rubber Tapper of Rondnia, stated in a personnal interview, if we had not created those state reserves, today all this area would already be pa sture land and our families would be leaving in slums in urban areas. Despite an active social movement resistan ce, there is a lack of local governments commitment with ER policy, leav ing little room on the local political development agenda for the rubber tappers demands, which are often s een by local economic el ite and government as antagonist to the state-wide development goals. This, in turn, has resulted in a political and institutional fragility that have made them di fficult to consolidate ove r the long term. ERs in Rondnia have likely had to face the greatest pr essures of all the ER s in the Amazon. Rubber tapper leaders are still living in an atmosphere of sustained rural conflicts, constantly harassed by local opposition pushing for other land uses, resulti ng in constant illegal practices (specially logging), and deforestation in th e states ERs (Euler et al. 200 8; Ribeiro et al. 2005). Rondnias experience with creating and managing state ERs is a valuable lesson for federal and state policy makers. Local political dynamics show that state level ERs are more vulnerable in terms of applying and consolidating the concept of the ERs as sustainable environmental policy. The rural workers union of Amap led an im portant social movement against the Jari Project a major cellulose production project in the southern part of the state, which resulted creation of the Extractivist Settlement Projects in 1988, and then creating the Cajari ER in 1990. The creation of the Rio Cajari reserve was st rongly supported by pressure of the social movement in the region, especia lly the Brazil nut collectors. Amap is one of the smallest states in th e region, located on the Amazon estuary and the forest product economy is more centered in the sout hern part of the states Because of its social


54 and ecological characteristics, Amaps southern region became the cente r of the direct-use conservation units in the state. Three protected areas dominate the states extractivist production are located in the south. The central region is the Cajari ER, the 860,000-ha Iratapur River State Sustainable Development Reserve (RDS), esta blished in 1996, on the northern border of the Cajari ER. The 600ha Extractivist Settlement Project of Marac was created in 1988, on what was to become the Cajari ERs southern border. Moreover, Amap is not located within frontier expansion areas, which in turn, may have led to reduced social movement pressure and land tenure conflicts over resources. Nevertheless, it seems that the social movement in the region has received political support for establishment extractivist deve lopment initiative, although not advocating for creation of new ERs over the last decades. According to Pedro Ramos (65), an importa nt extractivist leader when Joo Alberto Capiberibe was elected gover nor of Amap (1994/1998/2002), the concept of sustainable development came to governmental agenda, a nd the government launch ed the Sustainable Development Program of Amap (PDSA), creatin g Sustainable Development Reserves (SDRs) rather than ERs. Today with the new government (Valdez Goes, 2006) there is increasing endorsement of a broader policy that indicates continued suppor t for extractivist development, which led to greater support for extractive commun ities living in ERs and other models of directuse conservation units. Recently, the state govern ment launched a planning strategy called the Biodiversity Corridor, which advances an integrated management policy for conservation units, respecting the goals of each model and prom oting the creation of protected areas, in which new ERs are planned. The unchangeable scenery of ERs establishment in Amap since the creating of the Rio Cajari in 1990, does not illustrate the picture of scenarios of extractive


55 development in Amap; it merely shows that the ER model has not grown in terms of number. Amap shows an interesting scenery of extractiv ist development initiatives, one in which the ERs has not being applied, but has borrowed its central principals. The evolution of ERs in the second wave st ates in the late 1990s is a reflection of improvement of environmental policy in Amazoni a. Although the ERs are in their preliminary stages in Par, they are likely to be a result of a strong social movement mobilizing in the region. This social movement, which stated in early 198 0s (Campos & Nepstad 2006), has only recently found the right arena to propose its ideas for fore st conservation and rural development based on sustainable agriculture and extr activist livelihood systems. Local extractivist communities and colonist small farmers suffering from an absence of state assistance developed an important resistance with support of the Catholic Church and Rural Workers Unions creating the Movimento Pelo Desenvolvimento da Transamaznica e Xingu(MDTX); a coalition of grassroots organi zations of sustainable development groups, involving grassroots organizations of rural worker unions, colonist organizations and organizations of forest dwellers. This move ment has a broad green agenda that includes demands for technical assistance, credits, market access, infrastructure improvement, and basic social services (FVPP 2000). The agenda clamored for a development plan that reconciles forest conservation with rural development in the eixo North and South of the Transamaznica highway. In the northern part, the ER Verde Para Sempre (Forever Green), the largest ER in Amazonia today, was created, while in the sout hern part the Terra do Meio conservation mosaic was created, which includes other small ERs (eg., ER of Iriri) an d other restricted use conservation units (Nepstad et al. 2006a).


56 The MDTX constructed a common agenda among different groups and stgrengthened its local and regional alliances in order to demand concrete change s in government policies in the region. The Fundao Viver, Produzir and Pr eservar-FVPP (Foundation for Living, Producing and Preserving) connected over fifty grassroots organizations and became the operational arm of the movement, in addition to bringing financia l support and technical assistance for the groups. Its strongest period in build ing bridges to consolidate its voice took place through the mobilization for the paving of the Cuiab-Santarm Highway (BR-163), a road connecting soy bean producers from the state of Mato Grosso to the port of Santarm Its increased social network resulted in bigger role and pressure for the federal government to develop effective territorial planning actions to mitigate defore station in the region. The MDTX network and articulation process in addition to the join support of environmental NGOs were fundamental to the proposal of creating a conservation unit mosaic in the region, which came together with the governmental initiatives to mitig ate deforestation and solve dras tic land tenure conflict in the Terra do Meio region. This resulted in a soci o-environmental zoning program for the region, bringing together diverse stakeh olders and institutions in an innovative forum of debates on development and forest conservation, and envi ronmental governance in the region (IPAM 2004; ISA 2004). Establishment of federal forest ERs in Par represents a complex context of negotiation in Amazonia, one which has allevi ated long standing land tenure problems and agrarian conflicts3. It also represent a impor tant example of increased environmental governance through dialogue and negotiation between different development interests, continuing coalition 3 Tarcsio Feitosa da Silva (35) a key leader of the MDTX who was awarded one of six 2006 Goldman Environmental Prices for his work on defending local communities against land grabs, illegal illegal logging and human rights was emphatic when asked what were the advantages today of the creation of those ERs in the region:I am going to answer your question in a very si mple way and will use a phase from senhor Manelito when he found out that the ER of the Iriri was created in June/0 6: now I can sleep with tran quility in the ER of Iriri. I never expected that the government would give this area for me and my family. Now we have to continue work here to solve our problems related to economy, education and health. We need also to continue to guarantee our organization here.the threats are gone.


57 of social movements, and the increased presen ce of federal government institutions in frontier areas. Continued Demands and New Frontiers for ERs The dem and for ERs by local populations and their supporters has in creased significantly, demonstrating the popularization of ERs and the acceptance of this model as an important conservation and land tenure strategy used by forest peoples to gain rights to land and resources. Nevertheless, the process for es tablishment of federal ER is complex and can take years.4 As of January 2008, there were five federal ER processes concluded and waiting for federal decree to be officially created in Amazonia, all of them in the second wave state of ER systems. In the state of Par, the Montanha-Mangabal ER in the municipa lity of Itaituba, in the region of the upper Tapajs river; and the Mdio Xingu ER, in the municipality of Altamira, in the basin of the Xingu river will be created. In the state of Amazonas the Mdio Purus ER and the Ituxi ER will be created, in the municipalities of Lbrea and Pauini respectively; as will the Baixo Rio Branco-Jauaperi ER in the border of the state of Amazonas with Roraima in the municipalities of Rorainopolis and Novo Airo. This is an indication that the ER system will soon have important increment in land in those second wave states. In addition, th ere are several other ongoing processes for creation of ER in both first a nd second wave states as following: Acre (4), Amap (3), Rondnia (3), Tocantins (13), Amazonas (9 ), and Par (15). This suggest that in the next coming years ER will continue to growth, as it is expected to be created 23 ER in the first wave state and 29 in the second wave states. 4 It starts with an official request of population organization for the government. Then, a land tenure inventory, followed by a socio-economic survey of the region coordinated by the government. Then, a public consult is carried out at the municipalities where the ER will be created. The final process depends on the federal decree signed by the Brazilian President.


58 Moreover, the ER system has star ted to be implemented in the cerrado, or savanna, biome in central Brazil. This is an important move fo rward that will likely represent a new frontier for ER expansion in the future. In central Br azil, the ER model is being claimed by the cerrado extractivists groups that face similar problems to the Amazonian groups, which are related to land conflicts and maintenance of an extractivis t economy due to the expa nsion of agricultural frontier in the region. The cerrado is the second biggest biome in Brazil with an area of approximately 2 million km and is one of the most threatened in the country as it is Brazils principal area for agricultural expansion. It is a biome that en compasses eleven states in the country, including vast areas of wh at is considered the legal Amaz on, particularly in the state of Mato Grosso, Tocantins and Maranho. The cerrado had recently become an immense agricultural frontier, which in some aspects has substituted Amazonia as open space for economic expansion. Contrary to Amazonia, wh ere resource preservation has become an international concern, the cerrado has been overlooked as an impor tant biome. In this manner both the cerrado and the Amazon were seen as empty economic space to be explored, with the cerrado offering the advantage that its openness has led to faster economic occupation. From an internal and external political point of view, the environmental issues in the cerrado ecosystems have not appeared polemic, whereas in the Am azon imperative environmental concern is at the forefront. In 2006, the first two ERs created in the cerrado biome, the Recanto das Araras da Terra Ronca with 11.964 ha and Lago do Cedro with 17,337, both located at the state of Gois. In 2007, the Chapada Limpa ER with 11,971 ha was crea ted in the state of Maranho, the first ER on cerrado ecossytem in the state. The adaptation of ER in the cerrado has been thought to be a


59 strategy for extractivist groups a nd a barrier to the expansion of the agricultural frontier in specific parts of the region. The Brazilian Environmental Agency has re ceived eighteen requests from extractivist groups for continued establishment of new ER are in the cerrado region. The creation of ERs in central Brazil, incite a debate in the cerrado that was initiated twenty years ago in Amazonia, but has only recently gained political power as obs erved with the creation of many ERs in Amazonia in the last few years. Whether or not ERs will take off as an instrument for conservation in central Brazil is still in doubt, but its implementation, at least in the short run, will have an immediate impact on halting more destructive land us e practices, in addition to preserving traditional extractvist land use practices that have been lost to large-scale agricultural activities. With the implementation of the first ERs in the cerrado, it is also expected that government and other social groups will learn and unmask much of what is still little known about the traditional extractivist economy in the Central region of Brazil. Conclusions What have ERs contributed to conservation an d developm ent in Amazonia in the past two decades? And what will the coming decades hold fo r the Extractive Reserve system? In the first wave states, despite stagnation in Rondnia, bot h Rondnia and Acre have placed a significant percentage of their territory unde r ERs. Rondnias state reserves may be the most paradigmatic in their application of the model after twelve years; a typical case of struggle between the economic development and environmental protection camps, in which the winner is usually determined by the group that has mo re political weight. In sum, the status of ERs in Acre appears to be solid, in terms of territory and political s upport while in Rondnia, it is stagnated in growth and there is a profound lack of government support for ERs. Another first wave state, Amap,


60 appears stagnated in terms of numbers, despite the fact that extractivism development and environmental protection concerns have a pr esence in the state governmental agenda. The situation in the second wave states looks promising if recent advances continue in coming years; it is very likely they will. Par and Amazonas are far from reaching some of the first wave states in percentage of land under ERs, but have demonstrated a consistent process of establishing ERs since th e early 2000s, representing the tre nd for continuing growth of ERs areas. ERs as a public policy is widely considered on e of the options to simultaneously decrease deforestation rates in th e region while responding to social gr oup demands, especially in frontier areas. Most of the states in the Brazilian Amaz on have adopted this policy, especially in recent years. The creation of ERs should not be seen as static or synonymous of forest protection, rather, it is protection for a much longer time. Taking into considera tion that most ERs are created in frontier regions, three elements s hould be essential to all ER efforts: 1) the improvement of living conditions for the extractive population; 2) a strong monitoring process; and 3) an active and continuous social movement in the area. The biggest challenges might be not overcome with the establishmen t of an ER, but occur in the years following its creation. In many cases, the creation of an ER means the immediate resolution to land conflict. However, ER residents cannot be left alone w ith little or no governmental polic ies/support in place to help them improve their economic and social well-being, as have been the case in several of the first ER implemented in the region. This is especially likely if the social movement views the creation of an ER as an end of an end in itself. Federal ERs appear to be less vulnerable to state political interferences, while the statelevel approach seems to depend on closer ali gnment with the goals of state governments


61 development agenda. The general focus of both social movement institutions and government has been to create new ERs areas and pay less attention to ela borating public policies for socioeconomic growth within ERs. This perspective need to be reviewed in order to guarantee sustainable future for already established ERs in Amazonia.


62 CHAPTER 3 ADOPTION OF CATTLE RANCHING AMONG COLONIST SMALLHOLDERS AND FOREST EXTRACTIVISTS IN THE BRAZI LIAN AM AZON: HISTORICAL-CULTURAL CONTRASTS AND ECONOMIC EXPLANATIONS Introduction The recen t explosion of research on the envir onmental and economic costs and benefits of land use and land cover change in the Amazon has often featured the centr al role of cattle ranching activities (e.g., Barreto et al. 2005; Fa minow 1998; Nepstad et al. 2006b; Smeraldi & May 2008; Veiga et al. 2004; Wood & Porro 2002). Pastures cover the large majority of deforested land in agricultural use in the Braz ilian Amazon (Arima et al. 2006; IBGE 1998), and substantial areas that appear to be secondary vegetation when viewed in satellite images are actually dirty or degraded pastures (e.g., Perz & Walker 2002). While debate over cattle ranching in the Amazon has often centered on the large properties that contain a considerable share of the pastures in the region, observers increasingly recognize that so cial actors aside from capitalized interests are buying cat tle and converting other land uses to pasture (e.g., Barreto et al. 2005; Mertens et al. 2002; Tourrand & Veiga 2003; Walker et al. 2000). This shift toward cattle, a process called pecuarizao or cattle-ization, raises questions about why livelihood strategies among heretofore dist inct social actors would converg e in a highly heterogeneous socio-economic region such as the Amazon. Two historically contrasting social gr oups who are increasingly engaged in cattle ranching are small-scale colonists and forest extractivists. Initia lly upon arrival in frontier areas of the Amazon, colonists focused on producing f ood crops and engaged in ranching mainly for subsistence purposes (Fearnside 198 6; Moran 1981). For different r easons, forest extractivists such as rubber tappers also ha d little history of raising catt le (e.g., Dean 1987; Weinstein 1983). But by the 1990s, the livelihood systems and land use profiles of both groups were changing.


63 Cattle ownership among small-farm colonists in the Amazon grew faster than herds on large properties (e.g., Perz 2002). And rubber tappers, who in the 1970s and 1980s constituted a grassroots social movement to forestall defo restation and the expans ion of ranching, were adopting cattle and forming pastures (e.g., Gomes 2001; Gomes 2005). This parallel trend toward cattle ranching occurred despite profound historical and cultural differences among colonists and rubber tappers in different portions of the Amazon. This paper addresses the shift toward ranching among small-farm colonists and rubber tappers in the Brazilian Amazon. We first prov ide a background discussion on the debate over the environmental and economic benefits and costs of ranching in the Amazon. We then introduce the two study cases: Uruar, a small farm colony located in the Br azilian state of Par, and Xapuri, the birthplace of th e rubber tappers movement a nd gateway to the Chico Mendes Extractive Reserve (CMER) in the state of Ac re. We emphasize how the two populations differ deeply with regard to their live lihood strategies due to their far-f lung locations and their distinct histories and cultures. The third part of the paper provides docume ntation of the continuing shift from other land-use activities toward cattle ranc hing in the two sites si nce the late 1990s. We draw on survey data for two time points from bot h study sites to explore reasons for the common trend toward cattle. While economic explanations are important, other f actors also emerge as relevant to cattle adoption in the two sites. We conclude by di scussing the implications of the shift toward cattle ranching by small farm colonist s and forest extractivists for the economy and environment in the contemporary Amazon. Cattle Ranching in the Brazilian Amazon A large bod y of literature has emerged since the 1980s on the expanding ranching sector in the Brazilian Amazon. Here we briefly review the impetus behind frontier expansion in Brazil and the consequent discussions con cerning cattle ranching in Amazonia.


64 In the 1960s, Brazils then-military government prioritized frontier expansion as a means of integrating the Amazon with the rest of the country while al so stimulating economic growth via resource extraction (Guimares 1991; Maha r 1979). On the one hand, the government sought to avoid agrarian reform in l ong-settled regions by instituti ng colonization projects in the Amazon; on the other, the state motivated capita l investment in frontie r areas with fiscal incentives (Schmink & Wood 1992). By the 1970s, mi gration by farm families into colonization areas proceeded alongside the establishm ent of large ranches by investors. In this context, the ranching sect or in the Brazilian Amazon expanded.1 From 1985 to 1996, the cattle herd in the Brazi lian Amazon roughly doubled in si ze from about 18.5 million to 35.2 million heads. While the overall Brazilian herd also grew, the non-Amazon herd grew very little during this period; most growth in Brazils cattle herd occurred in the Amazon. As a result, the Amazon herd rose from 14% to 24% of Brazils overall herd during 1985-1995 (IBGE 1998). During the same period, the land area under natu ral and planted pastur e also grew in the Amazon, from roughly 42 million to 51 million h ectares. Meanwhile, overall pasture area in Brazil stagnated in area, and pasture in non-Amaz onian Brazil actually declined. Consequently, the Brazilian Amazons percentage of pasture also rose, from 24% to 29% of the total pasture area in Brazil. These data (Table 3-1) indicate the spatial expa nsion of ranching, and especially rising stocking densities, and the growing importa nce of cattle ranching in the Amazon for Brazil overall. The growth of the cattle herd and the prev alence of pasture in the Brazilian Amazon stimulated considerable debate about the social and ecological consequences of ranching on the frontier. Socially, ranching in the Amazon and elsewhere in Brazil has long been associated 1 By Brazilian Amazon we refer to the Legal Amazon, a state planning region that encompasses nine states in the Amazon biome in Brazil as well as portions of the savannah (cerrado) along the Amazonian fringe (IBGE 1998).


65 Table 3-1. Cattle herds and pasture areas in the legal Amazon and Brazil, 1985-1995 Cattle (000s heads) Pasture (000s ha) State 1985 19951995 / 1985 19851995 1995 / 1985 Rondonia 771 3,937 5.11,1012,922 2.7 Acre 334 847 2.5326614 1.9 Amazonas 425 734 1.7476529 1.1 Roraima 306 400 1.31,2471,543 1.2 Par 3,487 6,080 1.76,6127,456 1.1 Amap 47 60 1.3479245 0.5 Tocantins 3,604 5,218 1.410,65111,078 1.0 Maranho 2,973 3,619 1.25,0275,311 1.1 Mato Grosso 6,547 14,330 2.216,40521,281 1.3 Brazilian Amazon 18,494 35,226 1.942,32350,978 1.2 Brazil 128,042 153,058 1.2179,188177,700 1.0 Brazil NonAmazon 109,548 117,832 1.1136,865126,722 0.9 Pct. Amazon 14.4 23.0 1.623.628.7 1.2 Sources: IBGE agricultural censuses (IBGE 2005). with concentrated land ownership, rural exodus and urban poverty be cause ranching is landextensive but requires little labor (Hecht 1985; Hech t & Cockburn 1990). Where landless families arrived alongside land companies and other capitalized interest groups, land conflicts have often arisen. These groups often encountered indigenous tr ibes and traditional forest extractivists in the process, l eading to conflicting land claims a nd rural violence (Almeida 1995; Branford & Glock 1985; Schmink & Wood 1992). Th e Parrots Beak in Eastern Amazonia is one such region, and it became infamous for rural violence by gunmen and police supported by well-connected ranchers who sought to defend la rge land claims. This and other parts of the Amazon have a record of human rights abuses, in cluding assassinations of rural labor leaders, clergy, and others who sought to resist land concentration in large cattle ranches (Simmons et al. 2007). This stimulated debate over the question of whether the expansion of extensive cattle ranching on large properties constitutes social development in the Amazon. A second social


66 consequence has concerned the economic issue of productivity per hectare in extensive ranching, as opposed to alternative land uses such as non-timber forest produc tion, agroforestry, and intensified agricultural systems. More diversified agricultural systems produce higher incomes per hectare (Perz 2004; Tourrand & Veiga 2003), as do intensive systems that focus on multicropping, improved fallows, and the like (Almeida 1996). This has raised questions about the economic wisdom of extensive ranching as co mpared to alternative land uses that could produce more value on less deforested land. Debate has also transpired over the ecologi cal impacts of cattle ranching in the Amazon. One key issue has concerned doubts about pastur e sustainability due to soil degradation. Whereas early research suggested that clearing forest for pasture increased soil fertility (Serro et al. 1978), more recent work indicated that the in crease was due to burning vegetation and its effects were thus temporary (Barbosa & Fear nside 1996; Fearnside & Barbosa 1998). As a consequence, pasture productivity declines over time, such that 815 years later, pastures must either be abandoned due to weeds and soil compaction or renovated via tillage and investments in improved pasture grasses. By the early 1990s, 50% of pastures in the Amazon were considered degraded, and in many parts of the region, pastures had very low stocking rates and beef productivity per hectare (S erro & Homma 1993). The relianc e on fire to clear secondary vegetation has the economic advantage of being low-cost, but carries the disadvantages of destroying seed stocks for future forest regeneration and facilitating regrowth of fire-adapted plant species which prompts the need for more pasture burning. Another ecological problem with cattle ranching concerns microclimatic change (Buschbacher 1986). Temperatures of surface soils under pasture are much higher than in pr imary or secondary forest, which destroys microorganisms important for soil health (Nepstad et al. 1998). Further, heat rising from pastures


67 can generate locally strong winds and may be altering local precipitation (Wu et al. 2000). Finally, pastures support much less biodiversity than primary or secondary forest. Many species of vertebrates will not cross pastures, and many plant species will not grow in pastures if pioneer species are prevented from colonizing and in itiating the process of succession. Furthermore, fragmentation or edge effects have been s hown to impact even greater areas beyond those deforested, expanding the areas negatively a ffected beyond the pastures themselves (e.g., Laurance et al. 2002; Laurance et al. 2000; Skole & Tucker 1993). In the context of these doubts, pasture and cattle management in the Amazon changed considerably in the 1990s, and revisionist perspectives about cattle ranching in the region consequently emerged to question the conven tional wisdom. A key change has involved improvements in pasture management, partly due to landholder experimentation, and partly due to considerable extension research on pasture su stainability (e.g., Rueda et al. 2003; Smith et al. 1995; Valentim & Andrade 2005). The incorporation of new and improved grass varieties better suited to the Amazonian biophysical environment has afforded pastures a longer use life and hindered soil degradation and erosion, as well as weed invasions. At the same time, the urbanization of Amazonian populations and the li mited stocking densities in other regions of Brazil have impelled both regional and national demand for beef from the region (Faminow 1998). Such demand allowed beef prices to remain stable even as Brazils ranching sector has expanded, motivating the expansion of ranchi ng in frontier areas of the Amazon. This increasingly calls into question older arguments that ranching in the Amazon could only persist with the support of state tax breaks and subsidies. Another factor has been the enormous effort to control hoof-and-mouth and other cattle diseases via vaccina tion programs (MAPA 2005). Even in frontier areas of the Amazon, there are catt le vaccination and hygiene programs, part of


68 Brazils objective of increasing the exporting of beef products once they meet international sanitation standards (Lima et al. 2005; Smeraldi & May 2008). The prospect of beef exports from the Amazon is of fundamental importance, since Amazonian beef is cheaper to produce than that elsewhere in Brazil and most of the rest of the world, making it eminently competitive globally (Nepstad et al. 2006b). Since Brazil has become one of the top exporters of beef worldwide, Brazilian investor s are increasingly looking to th e Amazon frontier for continued expansion as well as improvement of cattle ranching. In this context, cattle ranching is incr easingly being adopted by groups other than corporate and individual large landholders in the Amazon. Smallhol der colonists, who originally came to northern Brazil to engage primarily in production of annual and perennial crops, are increasingly turning to cattle (Ludewigs 2006; Pacheco 2005; e.g., Wa lker et al. 2000). Similarly, forest extractivists, often the desce ndants of rubber tappers during the Rubber Boom a century ago, are also cleari ng forest and breeding cattle and other livestock (e.g., Gomes 2001 and 2005; Ehringhaus 2005;Vadjunec 2007). C onsequently, despite the Amazons land abundance, cattle ranching is no longer the purview of large capitalized op erations. This raises questions about the specific economic processe s at work that prompt cattle adoption for smallholder colonists and forest ex tractivists have very different cu ltures and histor ies, and rather different avenues for market access. In the next section, we focus on two concrete study cases, one involving smallholder colonists and the other featuring forest extractivists. Study Cases Our two study cases are the m unicipalities of Ur uar, in the state of Par in the eastern Brazilian Amazon, and Xapuri, in the state of Acre in the west ern Brazilian Amazon (Figure 31). While such choices might seem capricious or arbitrary, we selected th ese two sites for four reasons.


69 Figure 3-1. Map of the two study cases in the Brazilian Amazon First, we have fairly recent multi-temporal data about land use for both sites, which makes possible the analysis we pursue. Second, the two sites are located far apart in opposite ends of the Brazilian Amazon. This means that they are, in effect, independe nt of one another, such that the changes they incur can be viewed as occurring in parallel, whether influenced by a common external factor or not. This ma kes it possible to argue that similar processes such as cattle adoption are not due to one case influencing the other. Third, as a result of their geographic separation, Uruar and Xapuri have very different histories, bei ng founded at different times for contrasting purposes and attract ing distinct interest groups. And fourth, the cultures and


70 production systems in these two locations differ subs tantially in terms of the use of forests and cleared land. Thus, it is possible to document contrasts in past land use and livelihoods in the two sites in order to argue that their similarity in terms of cattle adoption is a culturally unexpected phenomenon. That in turn prompts our attention to economic and other factors as explanations for expanding cattle and pasture among groups not previously noted for cattle ranching. In this section, we provide br ief historical overviews of our two cases, highlighting their differences. Whereas Uruar is a young roadside frontier town, Xa puri has a history that extends back over a century and began as a rivers ide entrepot for forest product marketing. We highlight contrasts in resource manageme nt in these two study areas by paying close attention to the current dynamics of key economic products. First, in order to facilitate some initial comparison between our study sites, we rely on secondar y data sources from Brazilian censuses and production annuals for munici pal-level data (IBGE 1998; IBGE 2007). Uruar Uruar is a c olonist community in the Brazilia n state of Par situated on the Transamazon highway (IDESP 1990). Uruar was founded in th e early 1970s as a roadside colonization project to resettle rural families from the Brazili an Northeast. The state agency for colonization and land titling, INCRA, surveyed an d distributed lots of 100 hectar es (ha) to a first wave of colonists, who began to develop small farms. Co lonists first planted a nnual crops, later followed by perennials. This formed a relatively homogene ous social structure in Uruar, where smallscale family farms still predominate. By the e nd of the 1970s, local churches provided support to incipient small producer organizations (Toni 2003). In the early 1980s, Brazils economic crisis le d the state to effectively abandon official colonization, leaving the smallholder colonists to fend for th emselves. But in the mid-1980s, perennials such as cocoa and black pepper commanded high regional prices, which prompted


71 households to expand their clearings for cash crops (IDESP 1990). This stimulated a second wave of in-migration, raising the munici palitys population to 25,000 by 1991 (IBGE 1996). Growth during the 1980s prompted church lead ers and producer groups to lobby for the colony to become a new municipality, which o ccurred in 1987 (IDESP 1990; Toni 2003). The municipality of Uruar covers an area of about 10,066 km2. Economic dynamism in Uruar in the 1980s gave way to difficulties in the 1990s (Nascimento and Drummond 2003). Pests attacked cocoa and black pepp er, reducing cash crop production. This occurred at the same time as perennial crops also incurred price declines, which reduced agricultural incomes. These difficulties s timulated political mobilization in Uruar, as small producer organizations as well as busin ess interests and local politicians sought new directions for community development (Toni 2003) This period also witnessed the emergence of the MPST (Movimento pela Sobrevivencia da Tr ansamazonica, or Movement for Transamazon Survival), which helped form alliances am ong municipal producer groups (Nascimento and Drummond 2003). In Uruar, a series of municipal workshops resulted in the PGDU, or Global Plan for Uruar Development, which identifies va rious strategies for sustainable development as via agroforestry and reforestation (FUNDASUR 1996). At the same time, the Amazon Development Ba nk, BASA, made available a special credit line for small producers, FNO (Fundo Constituci onal de Financiamento do Norte). Many local organizations moved quickly to obtain FNO funds, and used the financing for pasture expansion and livestock purchases (Toni 1999). As a result cattle ranching expanded in Uruar during the 1990s. Smallholders shifted to cattle due to the more stable prices than those for perennials, the ability to sell cattle year-r ound instead of at harvest as with perennials, and easier commercialization due to trucking lines from slaughterhouses with refrigeration (Tourrand &


72 Veiga 2003). Establishment of a vaccination program in Uruar improved cattle health and survival, and made ranc hing more profitable. During the 1990s, Uruar's population continued to grow, reaching 45,000 by 2000 (IBGE 2000). By 2000, 23% of the forests in the m unicipality had been cleared (Nepstad et al. 2000, cited in Nascimento and Drummond 2003: 126). The 2000 elections again shifted the political aff iliations of those in local power back to supporters of the PGDU (Toni 2003). This coinci ded with a series of new (or belatedly implemented) initiatives such as for soil recuperation, forestation with economically valuable species (timber and non-timber). As a result, since 2000, Uruar has exhibited more complex dynamics in terms of its land use and livelihood systems. Table 3-2 summarizes changes in 5-year increments for se lected indicators of key economic activities in Uruar from 1990 to 2005 (IBGE 2007). Similarly, Figure 3-2 shows trend lines for annual changes over the same period. In brief, annual crops (i.e., subsistence food crops such as rice, beans, corn, manioc, etc.) declined and then rose, but did not return to their 1990 level by 2005. Preliminary figures for 2006 (not pr esented) indicate a rapi d decline to a level below 1990. Perennial crops (i.e. tr ee crops for commercial products, such as cocoa, coffee, and black pepper) rose in importance, but most of th e rise occurred in the early 1990s, as Figure 3-2 makes evident. However, it was the cattle herd which grew fastest, from roughly 43,000 head in 1990 to over 250,000 in 2005. There was a surge in the early 1990s that corresponds with difficulties in crops, but most of the expansi on in cattle in Uruar has happened since 2000. Timber extraction, performed by local sawmills in cooperation with landowners, also rose, from roughly 16,000 m3 in logs in 1990 to 56,000 in 2005. It is worth noting, however, as shown in Figure 3-2, that this value was a decline from 2004, as 2005 witnessed government inspections


73 and closings in sawmills due to illegal timber extraction. Non-timber forest products associated with traditional extractiv ism (such as rubber and Br azil-nuts) were insignif icant or in terminal decline in Uruar Together, these indicators por tray a young frontier community with a declining focus on annual crops, slow growth in pere nnials, but rapid growth in cattl e, an important timber sector perhaps with an uncertain future, and little by wa y of non-timber forest extractivism. As Figure 3-2 shows, from 1990 to 2005, it is cattle that grew the most in Uruar. Table 3-2. Crop cultivation, cattle ranching, and fo rest extractivism in Uruar and Xapuri, 19902005 Annual Perennial Cattle Timber Rubber Brazil Nut Crops (ha) Crops (ha) (heads) (m3) (T) (T) Uruar 1990 11,663 7,771 43,200 16,500 0 350 1995 7,014 9,923 110,000 40,410 0 28 2000 7,851 9,573 95,345 37,970 0 55 2005 9,348 11,465 250,739 56,100 0 4 Xapuri 1990 8,083 310 61,171 10,210 773 1,270 1995 3,325 255 61,204 10,930 638 1,019 2000 2,594 334 76,200 11,430 302 636 2005 3,791 241 190,986 12,720 479 2,007 Source: IBGE agricultural production annuals (IBGE 2005). Xapuri Xapuri is located in the far western Brazilian state of Acre, in the southeastern or Alto Acre portion of the state at the convergence of the Acre and Xapuri Rivers. Historically, the municipality of Xapuri covered about 8,137 km2 until 1999, when it was divided and reduced to 4,705 km2 (Toni & Souza 2003). Xapuri was founded in 1905 and is the second-oldest city in Acre after the state capital, Rio Branco. Both towns emerged as market centers along rivers, which served as the main means of transport during the Rubber Boom (1880-1920). As a port for


74 rubber and a longstanding town along the river, Xapuri came to be known as the little princess of Acre ( a Princesinha do Acre ). 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 1990199119921993199419951996199719981999200020012002200320042005 Annuals (ha) Perennials (ha) Cattle (heads) Timber (m3) Rubber (T) Brazil-nut (T) Figure: 3-2. Change in crop cultivation, cattle ranching, and forest extractivism, Uruar, PA 1990-2005 (1990=100). Source: IBGE Agricu ltural Production A nnuals (IBGE 2005) Xapuri was once inhabited by indigenous tribes such as the Xapuri, Catiana and Meneteri, who were joined and then replaced in the 19th century by migrants mainly from the Brazilian northeast, who came to the region attracted by the great promise of wealth and prosperity from the rubber industry (Rancy 1992). Th e high density of rubber trees found in the Alto Acre region made Xapuri a zone for extraction and a star ting point in the chain of rubber production downstream for export to North Atlantic economie s. During the rubber bo om, real estate in Xapuri was highly valued and among the lands c ontested between Brazil and Bolivia in the 1902


75 Acre Revolution, in which a brief conflict resulted in Brazil ac quiring Acre and the official control of important rubber ta pping territories (Bakx 1988). A key hallmark of rubber tappi ng during the boom years was the aviamento system of debt-peonage between seringueiros (rubber tappers) and their patres (rubber barons) (Weinstein 1983). Rubber barons claimed large land areas called seringais (rubber estates) which they divided up among rubber tappers, w ho each had access rights to a colocao (forest homestead) amidst several trails in the fo rest that passed by rubber trees. Rubber tappers arrived in Acre, however, in debt to the rubber baron for the pric e of their passage, and were prohibited from growing their own food crops or breeding their own livestock. This went hand-in-hand with the additional requirement that rubber tappers c ould only buy subsistence goods from the rubber baron on whose land they worked. Furthermore, rubber tappers could only sell their rubber to their particular rubber baron. These rules, along with falsified bookkeeping, ensured continued debt for rubber tappers and a focus on rubber extractivism to the exclusion of agriculture or ranching. Brazils monopoly on rubber production was broken when an English botanist smuggled rubber seeds out of the Brazilian Amazon to the Royal Botanic Garden in Kew, outside London, which were later introduced in Malaysia (Dean 1987). Such plan tations were free of pests of rubber endemic in the Amazon and yielded pr oduction at much lower prices. Malaysian plantations exhibited rapidly rising production in the 1900s, and by the 1910s, rubber from the Amazon was being outcompeted, causing the boom to go bust. With the collapse of the rubber trade in Amazonia, by the 1920s, Xapuri town a nd the residents of th e nearby rubber estates entered into a deep economic crisis (Tocantins 1 979). This resulted in the total abandonment of many rubber estates by the rubber barons, leaving the rubber tappers in th e forest to fend for


76 themselves. The rubber economy experienced a shor t resurgence in Brazil during World War II, when the United States and its allies were cut off from the Malaysian rubber trade when the Japanese occupied Malaya (Co rra 1967). Xapuri saw a short peri od of growth during this time. It appeared that the old Princess of Acre was re-awakening as new migrants from the Northeast began to flock to the region to work on the rubber estates as rubber soldiers recruited by the Brazilian government for the war effort (Martinello 1988). After the war, however, the rubber economy in the region once ag ain collapsed, leaving the rubber tappers all but forgotten by Brazili an society. Without many livelihood options, but having established a stro ng identity with forest extractivis m, rubber tappers responded to the crisis by staying into the forest. However, these boom-bust dynamics, along with a newfound freedom from the aviamento system, led to the diversific ation of livelihoods among rural communities around Xapuri and other parts of th e Acre. During the post-boom period, other forest products such as Brazil-nuts became more important, and rubber tappers began to engage in small-scale subsistence agri culture (e.g., Campbell 1996). By 1960, the municipality of Xapuri had a population of 13,256 (IBGE 1973). Xapuri and the rest of eastern Acre were to undergo further changes in the 1970s, when Brazils government built highways such as the Transamazon to facilitate the establishment of frontier towns like Uruar. Another highw ay, the BR-364 through neighboring Rondnia, connected Acre to southern Brazil. While migrants from the south flooded into Rondnia by the thousands beginning in the 1970s, the governor of Acre, Wanderlei Dantas (1971-74), gave a speech before investors in So Paulo, extolli ng the virtues of Acres relatively cheap and extensive land, long since abandoned by the rubber barons (Bakx 1988; Silva 1990). Overlooked, however, were the forest commun ities that formed after the rubbe r barons departed. The arrival


77 of roads and investors led to conflicts over the organization of space as defined by key natural resources. Whereas rubber tappers still viewed the area around Xapuri in terms of rubber estates and forest homesteads defined by the location of rubber trees, large ranc hers defined properties by geographic boundaries in terms of cleared la nd with good access to roads. By the 1980s, forgotten rubber soldiers found themselves confr onting cattle ranchers who sought evictions of forest-based communities in oste nsibly empty rubber estates. Rubber tappers fought back, mobilizing forest communities and rural workers to engage in non-violent forms of resistance beginning in Xapuri, making it the birthplace of the rubber tappers movement (Calaa 1993; Sobrinho 1992). In 1977, a rural workers union formed in Xapuri to resist encroachment by ranchers and other outsiders. In the 1980s, rural violence elsewhere in the Amazon proceeded alongside intens ifying land conflicts in Acre. The murder of Chico Mendes on December 22 1988, a year of unprecedented levels of deforestation and burning, catapulted the rubber tapper s into internationa l headlines as defe nders of the Amazon rain forest (Allegretti 1989; Hecht & Cockburn 1990; Schwartzman 1989). In this context, growing recognition of the social and ecological consequences of deforestation for large-scale ranching prompted the creation of federal Extractive Reserves (ERs) (Allegretti 1990; Schwartzman 1992) Since the late 1980s, the ER concept first gained currency in Brazil and today there are 64 ERs in Amazonia covering an area of over 12 million hectares. The ER model has paved the way for the creati on of people-based protec ted areas (Ehringhaus 2005) that are being implemented by various institu tions at both the federa l and state levels in response to changing political contexts and opportunities for Amazon conservation. ERs recognized the productive activities of rubber tapp ers, thus allowing for the productive use of standing forest per traditiona l livelihoods. Rubber tapper commun ities collectively manage areas


78 corresponding to old rubber estate s, whereas each rubber tapper household retains access rights to the rubber trails around their homestead. Fede ral ER land belongs to the federal government, but families have rights to use the forest for th eir livelihoods. In general, ER families engage in sustainable activities such as the extraction of non-timber fore st products (NTFPs), including rubber and Brazil-nuts. ER rules impose a deforest ation limit of 10% of the total area of a reserve, along with a 5% limit on pasture area, and commercial timber extraction is prohibited, though debated. The first federal ER was the Chico Mendes Extractive Reserve, or CMER, established in 1990, which currently en compasses 930,985 ha, or nearly 1,000 km2 (GovernmentofAcre 2000). The CMER spans several municipalities in Acre, but its main market gateway is the town of Xapuri. Now linked by both river and road, Xapuri remains a market center for forest-based production despite roadside deforestation in cattle ranches that border the CMER. By 2000, Xapuri had a population of 11,956, roughly the size it had 40 years before. Establishment of the CMER opened a new chapte r in Xapuris history, namely that of the struggle to make the ER model economically as well as socially and environmentally viable (Cavalcanti 2002; Rego 1999). Rubber pr ices went into decline in the 1980s, and this continued into the 1990s. Low market pri ces and organizational difficultie s threatened the viability of administering the ERs and supporting forest communities (Anderson 1994). The killing of Chico Mendes seemed to fracture the rubber tapper movement and its battle against cattle ranchers, who believed they would go unpunish ed. But Mendes ideas led to continued social mobilization during the 1990s th at generated new organizations to support forest communities in Xapuri and other munici palities encompassing the CMER. By 2000, the agroextractive cooperative in Xapuri, AMOREX, experienced substantial success, having grown to 360 members with a strong political voice in organizing extractivism at the municipal level


79 (Michelotti 2000). In addition, Julio Barbosa de Aquino, a rubber tapper who stood shoulder to shoulder with Chico Mendes in the confrontations with ranchers, won office as mayor of Xapuri in 1996. He was reelected in 2000. In the 1998 state and national elections, movement leaders won seats in the state and national congress, and a forester and former political advisor of Mendes, Jorge Viana, was elected Acres governor (1998-2006). The daughter of a rubber tapper from Acre, Marina Silva, was elected Senator and became Brazils Minister of the Environment after the Presidential election in 2002, when Luis Incio Lula da Silv a, a union labor leader from the south who had supported Chico Mendes to organize rubber tappers in Acre, became president of Brazil. In light of these victories, support of forest extractivism in Acre has grown considerably. The new state government in Acre, known as the Forest Government, pursued numerous policy initiatives to capitalize on Acres comp arative advantage forest resources by improving infrastructure, subsidizing the proc essing and commercialization of NTFPs, and strengthening certification of sustainably harvested timber (GovernmentofAcre 2005; GovernmentofAcre 2006; Viana 2004). In the 20 00 municipal elections, politicians supported by social movements won several offices in Xapuri, consolidating their gr ip on local power. Xapuri politicians and movement leaders have since engaged in local planning in support of forest extractivism. These e fforts resulted in the PDLIS ( Plano de Desenvolvimento Local Integrado e Sustentavel de Xapuri ), a sustainable development plan for the municipality (Toni & Souza 2003). Xapuri has consequently become the focus of several green development projects. The forest government has invested in the lo cal processing of Brazil-nut production. It has financed a condom factory in Xapuri to use loca lly produced green latex to supply condoms in conjunction with the Federal Ministry of Health in an attempt to revitalize the local rubber


80 industry. The long story of forest extractivism and political struggle in Acre was recently portrayed in Brazil with the Globo networks telenovela (an extended nightly television miniseries) celebrating the rich history of the region, featuring the town of Xapuri. But regardless of these many chan ges in support of forest extractivism, cattle ranching is expanding in Acre (Toni 2007; Valentim et al. 2002; see also Table 1). What is more, this land use activity is not confined to larger-scale landowners but is growing among colonists and erstwhile rubber tappers (Gomes 2001; Ludewi gs 2006; Rgo et al. 2003). During the 1990s, deforestation rose in the CMER (Sassagawa 1999) Analysis of land-use change in six rubber estates with high deforestation rates indicate that during the pe riod of 1986 to 2003, deforestation accelerated over time, mostly for pasture creation. Cattle ranching is expanding in the municipa lity of Xapuri, where the rubber tapper movement first emerged. Table 3-2 and Figure 3-3 provide indicators of key rural economic activities in the municipality of Xapuri, from 1990 to 2005, allowing comparisons and contrasts over time and with the case of Uruar. During this period, the land area in Xapuri under annual crops declined, while perennials varied without e xhibiting a distinct trend. The number of cattle did not change during the early 1990s, but in the late 1990s and especially since 2000, growth has accelerated. As Figure 3-3 shows that by 200 5 Xapuri had more than three times the number of cattle as it had in 1990 and more than twice the number in 2000. At the same time, timber extraction showed a moderate ri se. Rubber, on the other hand, d eclined in importance during the 1990s, but has risen again since 2000, perhaps a re sult of rubber subsidie s provided by the local forest government. Similarly, Brazi l-nuts went into decline, but recovered by 2005, in part due to local government support for the commercialization of local cooperatives, a nd especially due to the rising price for Brazil-nut on the international market. What emerges from a review of rural


81 economic products in Xapuri is a picture somewh at different than in Uruar. Uruar has a population roughly four times that of Xapuri. Pe rennial crops are more important in Uruar, whereas NTFPs are more important in Xapuri. But in both places, cattle is expanding rapidly, and faster than any other of the key economic products reviewed here. The timing of the expansion of cattle in Xapur i raises very interesting questions. Many analysts, including ourselves, have ar gued that the rise in cattle was due to declines in prices (and therefore extraction and commercialization) fo r rubber, Brazil-nuts, and other NTFPs (e.g., Gomes 2001; Toni 2007). But as Table 3-2 and Fi gure 3-3 show, cattle herds in Xapuri have expanded since 2000 alongside the expansion of both rubber and Br azil-nut extraction. It is likely that livelihood diversification among forest communities in the ER, particularly in the CMER, may help explain the dynamics of th ese products (e.g., Ehringhaus 2005; Vadjunec 2007; Wallace 2004). However, the municipal-leve l census data do not allow a breakdown of cattle ownership and the other i ndicators among specific social gr oups such as rubber tappers in the CMER. This limitation motivates our analysis of field survey data, pursued for both small farm colonists in Uruar and families in the CMER. Field Survey Methods and Data In this section we provide a m ore in-depth exploration of rural households and their livelihoods, comparing colonist families in Uruar to forest homesteads in the CMER in Xapuri and neighboring municipalities. In particular, we draw on data for two time points in both sites, and feature households that were re-interviewed, which facilita tes temporal comparisons. The greater detail in our survey da ta and the ability to compare acr oss sites and over time affords a richer portrayal of the circumstances and dynamics of cattle adoption among small farm colonists and forest extractivists, even in far-flung portions of the Brazilian Amazon.


82 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 1990199119921993199419951996199719981999200020012002200320042005 Annuals (ha) Perennials (ha) Cattle (heads) Timber (m3) Rubber (T) Brazil-nut (T) Figure 3-3. Change in crop cultivation, cattle ranching, and forest extractivism, Xapuri, AC 1990-2005 (1990=100). Source: IBGE Agricultural Production Annuals (IBGE 2005) Uruar Survey data for Uruar refer to 1996 and 2002. In June and July 1996, the second author participated in a nine-m ember research team c onsisting of North American and Brazilian social and agricultural scientists from the University of Texas-Austin, Florid a State University, and EMBRAPA/CPATU who administered a survey questionnaire to farm households in Uruar (Perz et al. 2006; e.g., Walker et al. 2000). The questionnaire wa s divided into two components, where the first addressed household characteristic s and the second concerned the lot(s) held by households. The household component included items such as family age composition, sources of income, and material wealth. The lot component included items such as land use, access to credit, use of agricultural technologies, and distance to market. It also included numerous


83 questions on resource use, incl uding agriculture and ranching, which in turn included questions on pasture formation as well as cattle management. Systematic sampling of farm lots proved intractable because not all lots had houses. Moreover, systematic sampling of houses encoun tered was problematic because residents were sometimes absent. We therefore sampled by first opportunity of residents encountered on their lot. We employed a cadastral map of Uruar from the Par state office of Brazils agricultural research agency, EMBRAPA/ CPAT U, as our sampling frame, to ensure that sampling was not clustered spatially or selective of households by socioeconomic status2 .The 1996 sample includes 261 households, or 12% of all rural establishments in Uruar at the time (IBGE 1998). The sample also includes 347 lots, as 25% of house holds held more than one lot, and the same questions were asked about each lot. In 2002, another team of North American and Br azilian social and agri cultural scientists from Michigan State University, IMAZON, and th e Federal University of Bahia administered a similar questionnaire in Uruar (Aldrich et al. 2006). This questionnaire had many of the same items as the 1996 instrument, but was more exte nsive and included additional questions about pasture and cattle, including commercializati on of cattle. A key goal of the 2002 fieldwork was to locate lots and households sampled in 1996 in order to constitute a panel for temporal comparisons. While a lot is a geographic entity, a household may possess one or more lots that in turn comprise a property. The 2002 sample includes 143 households, which held 170 lots in 1996 and 221 lots in 2002. The difference is due to sale s and especially purchases of lots during the interim. We defined the panel on the basis of whether a lot had been in the 1996 sample. This 2 The 1996 Brazilian population count (IBGE 1998b) and 19 95/96 Brazilian agricultural census (IBGE 1998a) allow for comparisons to assess sampling bias. In terms of household size and land allocation, the survey data are very similar to census data for Uruar sample. For more, see e.g. Perz, et al. Perz, S. G., Walker, R. T. & Caldas, M. M. 2006 Beyond population and environment: Household demographic life cycles and land use allocation among small farms in the Amazon. Human Ecology 34 829-849..


84 therefore includes lots in a gi ven property in both 1996 and 2002, as well as lots in one property in 1996 and another in 2002. While this hinders the correspondence somewhat, it helps capture dynamics on both lots and properties fr om the 1996 sample to the 2002 sample.3 Xapuri/Chico Mendes Ertractive Reserve The first and third authors conducted interv iews with families in the CMER in 2000 (Gomes 2001) and 2004 as part of their gra duate research (Gomes 2008; Vadjunec 2007; Vadjunec et al. nd). In May th rough August of 2000, the first au thor employed a questionnaire containing items on migration history, association memberships, land use and forest extractivism, and assets and wealth. The in itial sample included 66 households in four seringais (old rubber estates) of the CMER in Xapuri and two neighbor ing municipalities, Brasileia and Assis Brasil. These locations were selected for variability in terms of their distance to Xapuri town and various levels of deforestation (S assagawa 1999). While this does not guarantee representativeness of all households in the CMER it does capture much of the heterogeneity in the reserve.4 3 One might argue that because the lots included in the panel for the two dates are different, the panel is not comparable. But if we link data for the two dates in terms of production systems, in the presence of an active land market, the result will necessar ily be different lots in th e household panel at differen t moments, which itself is necessary to capture in order to observe the changes in pr oduction systems over time. See Aldrich, et al. 2006 for more. 4 In terms of basic variables such as household size and re source management activities, the data reported here are broadly similar to data collected by others in the CMER for similar dates. For example, government data for the entire CMER indicate that forest home steads average 671.7 ha per household, a figure somewhat higher than the averages for our full CMER samples and but similar to the CMER panel IBAMA/CNPT. 1999 Projeto RESEX: Um Futuro Sustentvel para a Amaznia, pp. 30. Brasilia: Ministrio do Meio Ambiente MMA; IBAMA/CNPT.. Data from 1996/1997 for another sample from the CMER Cavalc anti, F. d. S. 2002 A poltica ambiental da Amaznia: Um estudo sobre as reservas extrativistas Campinas, SP: Universidade Estadual de Campinas, Rgo, J. F., Costa Filho, O. S. & Braga, R. R. A. 2003 Anlise econmica dos sistemas de produo familiar rural da regio do Vale do Acre, 1996/1997 Rio Branco, Acre, Brazil: UFAC/SEBRAE/Ford Foundation. indicate average brazil nut extraction of 2178 kg per household, which is similar to our values for 2000, especially the panel data. Rubber extraction reported in the same sources for 1996/1997 was around 633 kg, higher than we report in 2000, but that is consistent with the ongoing decline during the 1995-2000 peri od as reported in those sources as well as in state data (see Figure 3-3). Finally, the other sample reports an average of 3.6 heads of cattle per CMER household, a value lower than we report, but also consistent with state data indicating a rapid rise in cattle since the mid-1990s (Figure 3-3).


85 In 2004 and 2005, the first and third authors con ducted interviews with households in the CMER. They jointly designed a more extensiv e questionnaire with additional items on commercialization of agricultural production, fore st extractivism, and livestock, as well as land tenure rules for resource use in the CMER. These researchers also obtained a larger sample, totaling 149 households in eight seringai s (four of the same seringais of 2000) in the reserve in Xapuri, Brasilia, and Assis Brasil. Selection crite ria was based on the pre-processing of satellite imagery, and discussions with local leaders and NGOs operating in the region (Vadjunec et al. nd). We targeted a variety of communities that had both real and perceive d differences regarding deforestation and produc tive activities ranging from NTFPs to commercial agriculture and cattle ranching. The 2004/2005 sample includes some of the same households from the 2000 research, affording a panel for temporal comparisons. We compared the names of forest homesteads and their owners in the two data se ts and identified 35 homesteads interviewed at both time points. Because this panel is smaller than both of the full samples for 2000 and 2004/2005, we will present data for both full samples as well as th e panel homesteads at the two time points. This affords temporal comparisons among the panel ho mesteads as well as between the panel and the larger samples. Cattle Adoption Among Smallholder Colonists and Forest Extractivists We divided our analysis of th e field survey data into three parts. First, we present socioecono mic indicators for the samples and panels in Uruar and the CMER to characterize the households. This provides some baseline comparis ons across the sites and over time. The second part of the analysis then evalua tes natural resource management in the two sites in terms of land use, forest extractivism, agri cultural production, and cattle ranching. This enables comparisons across the sites and over time in terms of land use and resource management dynamics. The third


86 and final part of the analysis focuses on management of cattle and pasture. Here we examine the most recent data and compare the sites in term s of the prevalence of cattle ranching, use of inputs, herd composition, purchases and sales of cat tle; we also evaluate certain site-specific issues, namely obedience to land -use restrictions (in the CMER) and future plans for ranching (in Uruar) and sustainable forestry (in the CMER). Socioeconomic Comparisons Table 3-3 presents full sam ple and panel da ta for the Uruar surveys in 1996 and 2002, and Table 3-4 does the same for the CMER surveys in 2000 and 2004/2005. Both tables provide basic and largely comparable indicators of household background, institutional context, and market access across the sites and survey years. Each table also provides p-values from t-tests comparing means for the two time points for whic h we have panel data. Moreover, Table 3-4 provides p-values for t-tests comparing the most recent samples for the two sites.5 By household background, we refer to lengt h of residence, whether the household interviewed was the first occupant on that property or homestead, whether the household interviewed had acquired the site via inheritance, and the number of residents. Together these variables help characterize the te nure of the family on their property/homestead as well as the age of the place relative to the length of a fa milys occupation. We would expect families in Uruar, a relatively new frontier area, to have shorter residenc e durations and to be living on younger properties than the homesteads on rubbe r estates in the CMER. We also expect durations of residence to rise from the first time point of the surveys to the second, and for first occupancy to decline and inheritance to rise. Fi nally, we expect there to be large families 5 All t-tests are two-tailed tests that do not assume equal variances in the two samples being compared.


87 (several members), due to the importance of family labor in smallscale rural production systems. Tables 3-3 and 3-4 largely b ear out these expectations. Fo r Uruar, Table 3-3 compares the 1996 full and panel sample data to the 2002 pane l data. It is worth noting that the values for the household background variables are similar between the full and pa nel data in 1996, though the panel is somewhat selective of older and larger households. But there is a statistically signi ficant rise in the length of re sidence in the panel from 1996 to 2002, though inheritance declined (albeit from an already very low level), and family size declined. Other research suggests that decreasing family size is due to aging and relocation of children, whether to Uruar town or their own properties (e.g., Perz et al. 2006). Table 3-4 presents the CMER comparisons for 2000 and 2004/2005. There, the panel also appears very similar to the full samples for both time points, though as in Uruar, the CMER panel appears somewhat selective of older, la rger households. But whether we compare the full samples or the panel across time points, we find that length of residence rose over time, which perhaps is a positive indicator of the viability of ERs as a successful land tenure model, with little outmigration in the CMER. First occupancy also exhibits an increase over time, perhaps due to the division of homesteads in order to establish new family homest eads. Inheritance declined though the difference is not significant. Finally, the number of resident s rose significantly, perhaps due to in-migration, as living conditions have improved in the CMER, especially in the rubber estates bordering towns. Many family members, especially th e younger ones, who may have prev iously moved to town looking for job opportunities and/or education, ha ve returned to live in the CMER.


88 Table 3-3. Socioeconomic indicators, farm households in Uruar, Par Uruar 1996 1 Uruar 1996 Uruar 2002 p(t-test), Full Sample Panel Sample Panel Sample 1996-2002 Household Background Length of Residence (Years) 12.05 (6.78) 2 13.49 (7.14) 17.81 (13.56) <0.01 First Occupancy (0=No, 1=Yes) 0.33 (0.47) 0.36 (0.48) NI N/A Acquired via Inheritance (0=No, 1=Yes) 0.02 (0.15) 0.03 (0.16) 0.00 (0.00) 0.08 Residents on Property (Persons) 7.23 (4.58) 7.47 (5.01) 5.81 (4.80) 0.04 Institutional context Community Assoc. Member (0=No, 1=Yes) 0.32 (0.47) 0.26 (0.44) 0.60 (0.49) <0.01 Labor Union Member (0=No, 1=Yes) NI NI 0.45 (0.50) N/A Credit (0=No, 1=Yes) 0.57 (0.50) 0.70 (0.46) NI N/A Receiving Retirement Income (0=No, 1= Yes) 0.23 (0.42) 0.23 (0.42) NI N/A Market Access Distance to Town (km) 28.09 (13.87) 25.12 (11.82) 24.98 (13.18) 0.93 Notes. 1. The full Uruar sample contains 261 households and 347 lots; the Uruar panel contains 143 ho useholds with 170 lots in 1996 and 221 lots in 2002. 2. Values shown are either proportions (for binomial variables) or arithmetic means (for contin uous variables). Values in parentheses are standard deviations.


89 Table 3-4. Socioeconomic indicators, seringal househol ds in the Chico Mendes Extractive Reserve, Acre 2000 1 2000 2004/2005 2004/2005 p(t-test), p(t-test), Full Sample Panel Sample Panel Sample Full Sample 2000-2004 Uruar Household Background Length of Residence (Years) 12.12 (10.24)2 13.36 (10.87) 15.45 (9.49) 13.75 (12.43) 0.38 0.02 First Occupancy (0=No, 1=Yes) 0.09 (0.29) 0.11 (0.32) 0.14 (0.36) 0.27 (0.45) 0.69 N/A Acquired via Inheritance (0=No, 1=Yes) 0.23 (0.42) 0.22 (0.42) 0.11 (0.32) 0.19 (0.40) 0.23 <0.01 Residents on Property (Persons) 5.52 (2.76) 6. 22 (2.65) 7.89 (2.87) 5.86 (2.75) 0.01 0.91 Institutional context Community Assoc. Member (0=No, 1=Yes) 0.65 (0.48) 0.64 (0.49) 0.68 (0.47) 0.61 (0.49) 0.74 0.75 Labor Union Member (0=No, 1=Yes) 0.55 (0.50) 0.53 (0.51) 0.70 (0.47) 0.69 (0.46) 0.15 <0.01 Credit (0=No, 1=Yes) 0.36 (0.48) 0.47 (0.51) 0.51 (0.51) 0.49 (0.50) 0.72 N/A Receiving Retirement Income (0=No, 1=Yes) 0.15 (0.36) 0.17 (0.38) NI NI N/A N/A Market Access Distance to Town (km) 40.13 (59.45) ?? ?? 43.81 (22.42) ?? ?? Notes. 1. The CMER panel contains 35 households, each with one colocao The full CMER sample in 1999 includes 66 households, and the full sample in 2004 includes 149 households, each with one colocao. 2. Values shown are either proportions (for binomial variables) or arithmetic means (for contin uous variables). Values in parentheses are standard deviations.


90 This may be because the government created subsidies for rubber production, Brazil nut prices increased substantially, there were improvements in tran sportation via opening of unpaved feeder roads in the CMER, and huge improvements in education programs for both younger and older residents. Comparing the two sites, we find distin ct durations of residence (though actually somewhat longer in Uruar), and greater first o ccupancy in Uruar than the CMER (as expected, though the lack of comparable households for si milar time points hamper this comparison), no significant difference in inheritance (perhaps less in Uruar, due to an active land market there and the land tenure system of the CMER in wh ich the government owns the land), and similar numbers of family members, though with different dynamics (a decline due to aging in Uruar and a rise perhaps due to retu rning migrants in the CMER). In terms of household background, the two sites differ in many ways, though not in labor availability. We also make socioeconomic comparisons in terms of the institutional context in which families find themselves. By institutional contex t we mean whether a household is a member of a community or neighborhood association, whether it has one or more rural labor union members, whether it had received bank credit, and whether it received retirement income. We view the first two indicators in terms of soci al capital via solidarity and ties to social movements (Vadjunec 2007); we see the last two as key sources of fina ncial capital, which can co mplement labor assets and alter resource use. In general we expect so cial capital to be importa nt in both sites, but especially in the CMER given its history of successful grassroots social mobilization, though we expect access to capital to be gr eater in Uruar due to the impo rtance of FNO-e in the 1990s. Table 3-3 summarizes the institutional variab les for Uruar and Table 3-4 does the same for the CMER. In Uruar, some differences app ear between the full 1996 sample and the panel,


91 which suggests less membership in community or ganizations and more access to credit. Some lack of comparability hinders temporal comp arisons, but it is evident that community organization membership rose significantly ove r time. In 1996, there was widespread but not universal access to credit, and roughly one in fi ve households sampled were receiving retirement income. In the CMER, some differences also ap pear between the full samples and the panel, where there appears greater membership as well as credit. Community association membership levels were high and steady ove r time, labor union membership was somewhat less common but rose somewhat, and retirement income is infre quent (at least in 2000). It is important for the rubber tappers to be associated with the co mmunity association for economic and social purposes. Due to their dispersed location in the forest, it is of utmost economic importance to access the market through the association. Gove rnment incentives for extractivism require association membership to access the subsidized prices for rubber. Comparing the two sites, it appears that asso ciation membership is similar, at least in more recent data. The same holds for credit, and again this may reflect a convergence in more recent data. Retirement income appears slightly more important in Uruar, at least for the earlier dates of the surveys. Those characteristics seem to illustrate interesting historical and cultural perspectives. While in Uruar colonists have mo re social benefits, the rubber tappers from Xapuri, on the other hand, are more marginalized and isolated in the forest and depend more on social institutions to access government benefits. This helps to explain why the data show strong social organization among the extractivists. The l ack of basic citizenships benefits of the CMER extractivist population has been a major concern fo r the Forest Government, which highlights the social component of the local and regional development paradigm, Florestania, or forest citizenship (GovernmentofAcre 2005).


92 Finally, we compare the two sites in terms of market access. By market access we refer to distance to the nearest town; in Uruar that is th e town of Uruar, and in the CMER, it is usually Xapuri but in some cases Brasilia or Assis Bras il. Market distance allows a proxy appraisal of the costs of transport to market (Caldas et al. 2007), and thus the feasib ility of commercializing cattle and other products. In Uruar (Table 3-3) the panel appears se lective of properties closer to town for 1996, but a comparison of the panel case s with 2002 indicates a slight increase in market distance. This is due to acquisition of new lots somewhat farther from town, a strategy for expanding production systems fostered by unofficial road extensions (Perz et al. 2007). In Xapuri (Table 3-4), we observe a minor incr ease in market distance when comparing full samples from 2000 and 2004/2005. This increase is due to households from one seringal in the 2004/2005 sample, Icuri, having a much greater distance to market than other households in the sample, with a distance of about 80 kilometers of road to access the closes t city, Assis Brasil. As a result of the geographically-dispersed characte ristics of extractivist households, they depend on multiple means of transportation, usually by anim al from the homestead to the association nucleus with access to road, and then by truck to access markets in the city. Such difficulties for selling forest products influences households experimentation in ne w land uses, such as cattle ranching, which offers advantages in marketing access.6 In brief, a socioeconomic comparison of colonist households in Uruar to forest homesteads in the CMER reveals several simila rities, such as family size and association 6 For the 2000 data, household distance to market was calculated by adding the time spent on transportation by animal from each colocao to a central point (associa tion nucleus) with road access and distance from the association nucleus to nearest town. Both distances were measured in time trav eled, with subsequent conversion to kilometers. Time on trails from each co locao to association nucleus was tr ansformed considering 1 hour animal transport = 5km; time from association nucleus to town wa s transformed considering 1 hour in truck = 40 km. For the 2004/2005 data set, distance from the colocao to the association nucl eus was calculated in km using UTM coordinates collected with a GPS unit. From this point to the nearest town, distance was measured in km using our car odometer and/or GPS unit.


93 membership. However, there are also many differe nces, which largely follow expectations based on the distinct historical contex ts in the two locations. Given these differences, we turn to a comparative analysis of natural resource management in the two places, focusing on cattle ranching. Comparisons of Natural Resource Management Table 3-5 presents indicators of natural re source m anagement for households in Uruar, and Table 3-6 does the same for the CMER samples. In both, we first present data on land use: total area claimed, and land under primary forest crops, pasture, and secondary vegetation. All land use data are based on areas reported by respondents. In Urua r (Table 3-5), the panel in 1996 shows larger properties than for the full sa mple, an indication of some selectivity of households with more lots. But in both, primar y forest covered roughly 60% of properties in 1996, and of the remaining area, the largest land use category is pasture for cattle production, which averaged roughly 30 ha. By 2002, property sizes forest cover, and cropland in the panel had not changed significantly, but pasture area and secondary gr owth had both risen significantly. In the CMER (Table 3-6), homestead s claimed much larger areas, a reflection of the extensive nature of forest extractivism, where rubber tappers originally extracted from trees along 4-6 trails, each covering roughly 100 ha (e .g., Gomes 2001). A comp arison of panels to full samples indicates (as in Uruar) some select ivity of larger homest eads. CMER homesteads in the panel did not exhibit significant changes over time in total area, forest cover, or cropland. Interestingly, the apparent rise in pasture area was not statistically signi ficant, though there was a rise in secondary growth.7 These dynamics in the panel are no t as pronounced as in the Uruar 7 The results for secondary growth in the CMER may be an artifact of differences in the data. We estimated secondary growth for 2000, based on the area under crop land and assumptions about cultivation time (3 years) relative to fallowing durations (8 years), derived from ob servations by the first and third authors during fieldwork


94 panel, and caution is necessary about pastur e expansion based on the CMER panel, though the full samples for 2000 and 2004 provide some additional though limited support for this interpretation. Tables 3-5 and 3-6 also presen t indicators of forest extractivism, annual and perennial crop production, and cattle. In Uruar (Table 3-5), th ere is minimal extractivism, and never rubber extraction. On the other hand, despite some selectiv ity in the panel for la rger production systems, there is considerable annual crop production (as indicated by rice and beans) as well as production of perennial crop comm odities (indicated by cacao, coffee, and black pepper). From 1996 to 2002, according to the panel, rice production declined significantly, though bean production did not. But there were al so significant declines in coff ee and especially black pepper, though not in cocoa. Conversely, cattle herds ro se significantly from 1996 to 2002, among cattle on a households own properties as well as in terms of cattle grazi ng on other properties. Overall, the average cattle herd per property in the Ur uar panel more than doubled, from roughly 33 heads in 1996 to 80 in 2002. In the CMER (Table 3-6), forest extractivis m is very important. There are differences between the full samples and panel and the resulti ng trends run in the oppos ite directions, so we emphasize the panel when making temporal co mparisons. In brief, Brazil nut extraction8 rose while rubber declined, but neither change was statis tically significant. Inte restingly, extraction of both Brazil nut and rubber cont inued as of 2004/2005. Production of annuals also exhibits a and from contacts with knowledgeable informants who have conducted extension work in the CMER. Data on secondary vegetation for 2004/2005 are areas reported by respondents. 8 Note that brazil nut extraction is typically measured in l atas. We follow Wadt, et al. Wadt, L. H. O., Kainer, K. A. & Gomes-Silva, D. A. P. 2005 Population structure and nut yield of a Bertholletia ex celsa stand in Southwestern Amazonia. Forest Ecology and Management 211, 371-384. in converting latas to kg by assuming 11 kg per lata.


95 Table 3-5. Natural resource management, farm households in Uruar, Par Uruar 1996 1 Uruar 1996 Uruar 2002 p(t-test), Full Sample Panel Sample Panel Sample 1996-2002 Land use (ha) Total Area Claimed 133.44 (83.03) 3 151.37 (98.41) 151.78 (119.35) 0.85 Primary Forest 86.10 (67.49) 95.18 (82.07) 84.11 (63.92) 0.25 Annual, Perennial Crops 8.85 (8.80) 10.51 (9.11) 10.34 (13.30) 0.90 Cattle Pasture 29.48 (27.06) 34. 79 (26.75) 46.92 (76.44) 0.02 Secondary Vegetation 9.02 (11.04) 10.89 (13.72) 16.50 (21.52) <0.01 Forest Extractivism (kg) Brazil Nut NI NI 5.58 (29.43) N/A Rubber 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 1.00 Annual Crops (kg) Rice 3007.92 (3669.43) 3326.02 (4140.03) 1159.42 (2131.41) <0.01 Beans 284.83 (481.32) 264.69 (441.08) 184.96 (619.69) 0.32 Perennial Crops (kg) Cocoa 816.59 (2537.62) 1196.55 (3212.84) 1147.05 (3757.70) 0.85 Coffee 394.63 (1304.90) 560.75 (1691.10) 524.89 (1657.41) 0.01 Black Pepper 658.40 (1241.93) 973.25 (1507.91) 366.93 (748.23) <0.01 Cattle (heads) On Own Property 23.52 (38.30) 28.68 (32.33) 68.56 (150.04) <0.01 On Other Properties 2.49 (8.04) 3.40 (10.14) 12.49 (48.99) 0.02 Notes. 1. The full Uruar sample contains 261 households and 347 lots; the Uruar panel contains 143 ho useholds with 170 lots in 1996 and 221 lots in 2002. 2. Values shown are either proportions (for binomial variables) or arithmetic means (for contin uous variables). Values in parentheses are standard deviations.


96 Table 3-6. Natural resource management, seringal househol ds in the Chico Mendes Extractive Reserve, Acre CMER 2000 CMER 2000 CMER 2004/5 CMER 2004/5 p(t-test), p(t-test), Full Sample 1 Panel Sample Panel Sample Full Sample 2000-2004 Uruar Land use (ha) Total Area Claimed 566.67 (319.29) 2 633.33 (360.16) 634.29 (335.14) 491.89 (459.89) 0.99 <0.01 Primary Forest 551.21 (318.04) 3 616.80 (357.62) 612.41 (330.71) 469.09 (458.71) 0.87 <0.01 Annual, Perennial Crops 2.28 (1.59) 2.35 (1 .78) 1.83 (0.99) 2.26 (1.60) 0.14 <0.01 Cattle Pasture 7.10 (10.06) 7.93 (9.96) 9.17 (7.48) 8.80 (7.62) 0.55 <0.01 Secondary Vegetation 6.08 (4.23) 6.26 (4.76) 11.19 (8.88) 12.04 (16.85) <0.01 0.05 Forest Extractivism (kg) Brazil Nut 2499.67 (3134.72) 2070.44 (2732.79) 2441.06 (2665.33) 1686.25 (2680.94) 0.56 <0.01 Rubber 268.61 (418.36) 250.56 (378.43) 200.60 (312.91) 142.68 (269.91) 0.55 <0.01 Annual Crops (kg) Rice 1248.79 (1342.09) 1378.33 (1687.85) 948.00 (832.86) 937.78 (664.75) 0.18 0.24 Beans 495.61 (834.96) 554.44 (965.73) 537.14 (648.17) 393.21 (431.71) 0.93 <0.01 Perennial Crops (kg) Cocoa 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 1.00 <0.01 Coffee NI NI 49.03 (207.60) 45.11 (194.58) N/A <0.01 Black Pepper 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 1.00 <0.01 Cattle (heads) On Own Property 8.35 (10.32) 10.50 (12.10) 16.57 (17.48) 14.20 (17.57) 0.09 <0.01 On Other Property NI NI 1.71 (5.62) 1.27 (3.70) N/A <0.01 Notes. 1. The CMER panel contains 35 households, each with one colocao The full CMER sample in 1999 includes 66 households, and the full sample in 2004 includes 149 households, each with one colocao. 2. Values shown are arithmetic means. Numbers in parentheses are standard deviations. 3. Values for primary and secondary forest in 1999 are estimat es based on swidden area and assumptions about fallowing practice s.


97 somewhat mixed picture, with ri ce in decline (though not significantly, at leas t in the panel) and beans holding steady. Cocoa, coffee, and black peppers are of little im portance in the CMER, and this did not change after 2000. Cattle, on the other hand, rose in importance over time (though this was of marginal signifi cance in the panel, a similar change appears if we compare the full samples for the two time points), and in 2004, we have evidence of households running cattle on other homesteads, a indication of rubber tappers awareness regarding land-use regulations and the political repe rcussions of increasing cattle st ocks in their homesteads. As their cattle stocks increase and th ere is a greater need to increas e pasture land in the homesteads, some rubber tappers prefer to not increase their pa sture land, which can lead to substantial fines, and instead transfer part of thei r stock to other homesteads, usually occupied by a family relative who may have pasture land but little cattle. Overall, resource management in the two sites was very different, as indicated by significant differences in the most recent data sets. In Uruar, annual and perennial crops are relatively more important, while in the CMER, fo rest extractivism is much more important. While cattle ranching appears more important in Uruar than in the CMER, whether measured in terms of pasture area or cattle herd s, in both sites, we find ranchi ng in expansion, and at a faster rate than other productive activities involving natural resource management. Whereas pasture area in both sites rose by 20-30% over roughly five years, cattle herds increased faster, from 60200%. Comparisons of Pasture and Cattle Management The first part of our analysis docum ented a ri se in cattle herds in the municipalities most closely tied to our study sites (Table 3-2 and Figures 3-2 and 3-3); the second part provided additional data specifically for colonist a nd forest extractivist households, and documented


98 pasture expansion as well as gr owing cattle herds (Tables 3-5 and 3-6). The third and final part of the analysis now focuses on practices related to pasture and cattle management, in order to provide more insight into the processes at work and gain some inkling of the prospects for cattle ranching among these small-scale producers. We focus on the most recent survey data from the two sites, which refer to the 2002 panel in Uruar and the full sample for 2004/2005 in the CMER. Table 3-7 presents a comparison of pastur e and cattle management indicators, broken down into six groups: general, i nputs and practices, cattle herd composition, purchases and sales of cattle, along with site-specific indicators fo r compliance with deforestation rules (in the CMER) and future planning (in Uruar). Our gene ral indicators are simply whether a household had cattle, and (for the CMER only) the time since first acquisition of cattle. The data indicate widespread adoption of cattle, in fact slightly greater in the CMER than among colonists in Uruar, though the difference is insignificant. The onset of cattle acquisition in the CMER is about 7-8 years before the 2004/2005 survey, recent compared to th e average length of residence, which was roughly 14 years (Table 3-4), suggesti ng again that ranching is a relatively recent activity among residents. Technological inputs and management practic es in Uruar and Xapuri reveal some key characteristics about ranching among these subp opulations. For one thin g, there is somewhat limited capitalization, measured here in terms of whether a househol d owned a chainsaw. Roughly half did in Uruar, whereas only a bout one-third did in the CMER. Significant differences also appear for inve stments in enclosed pastures: hous eholds in Uruar had far more fencing than those in the CMER and roughly half of households in Uruar had corrals while


99 Table 3-7. Cattle and pasture management, farm households in Uruar, Par, and seringal households in the Chico Mendes Extractive Reserve, Acre Uruar 2002 CMER 2004/2005 p(t-test), Panel Sample 1 Full Sample Site Comparison Cattle Raising (0=No, 1=Yes) 0.83 (0.37) 2 0.87 (0.34) 0.29 Cattle Onset (Years Ago) NI 7.46 (7.54) N/A Inputs and Practices Chainsaw (0=No, 1=Yes) 0.53 (0.50) 0.34 (0.47) <0.01 Fencing (km) 39.92 (105.35) 0.25 (0.56) <0.01 Corral (0=No, 1=Yes) 0.49 (0.50) 0.29 (0.46) <0.01 Fire Maintenance (0=No, 1=Yes) 0.50 (0.50) 0.73 (0.44) <0.01 Pasture Planting (0=No, 1=Yes) 0.80 (0.40) 0.92 (0.27) <0.01 Pasture Fertilizers (0=No, 1=Yes) 0.04 (0.19) NI N/A Vaccinations (0=No, 1=Yes) 0.82 (0.39) 0.78 (0.42) 0.35 Mineral Salt (0=No, 1=Yes) 0.81 (0.39) NI N/A Cattle Herd Composition (Heads) Bulls NI 0.62 (0.67) N/A Cows NI 5.57 (7.59) N/A Calves NI 8.01 (10.24) N/A Total (Own Property Only) 67.56 (150.04) 14.20 (17.57) <0.01 Purchases and Sales, Previous Year (Heads) Total Cattle Purchased NI 0.64 (1.70) N/A Bulls Sold 10.38 (71.68) 0.19 (0.60) 0.10 Cows Sold 4.81 (14.72) 0.45 (1.32) <0.01 Calves Sold 28.42 (151.96) 2.17 (3.99) 0.04 Total Cattle Sold 43.62 (168.30) 2.84 (4.77) <0.01 Deforestation Rules Agree with 10% Deforestation Rule (0=No, 1=Yes) N/A 0.81 (0.39) N/A Agree with 5% Pasture Rule (0=No, 1=Yes) N/A 0.62 (0.49) N/A Future Planning Plans for Pasture (0=Other, 1=Expand) 0.59 (0.49) NI N/A Plans for Cattle (0=Other, 1=Expand) 0.77 (0.42) NI N/A 1. The Uruar panel in 2002 includes 143 households with 221 lots; the Chico Mendes full sample in 2004 includes 149 households, each with one colocao. 2. Values shown are either proportions (for binomi al variables) or arithmetic means (for continuous variables). Values in parentheses are standard deviations.

PAGE 100

100 about 30% did in the CMER. The greater degr ee of capitalization in Uruar reflects the somewhat more established state of ranching am ong colonists (more pasture area, more cattle, etc.). That said, differences do not appear as great for management practices. More households relied on fire for pasture maintenance in th e CMER than in Uruar, but pasture planting ( roagem ) was common in both places, and the use of fertilizers for pasture maintenance was rare in Uruar and unknown in the CMER. In addition, the use of cattle vaccinations was virtually the same among households in the tw o locations, and the use of mineral salts (in Uruar, at least) occurred on a comparable frequency as vaccinations.The high proportions for use of vaccinations is likely a result of govern ment efforts to eradicate hoof-and-mouth disease in the Amazon (Smeraldi & May 2008). The pi cture that emerges here is one where capitalization is related to the scale of ranchi ng activities (i.e., chainsaws are more common in Uruar, where pastures and herds are larger), while the use of specific management practices among these distinct groups does not appear to vary as much, in part due to co-occurrence in the same national context as concerns beef production. Table 3-7 also presents data on cattle herd composition. Unfortunately, this was documented more systematically in the CMER th an Uruar. That said, herd composition in the CMER and that anecdotally observe d in Uruar appear fairly typi cal of breeding systems for beef that are constituted by herds with a few bulls, many cows and numerous calves (Tourrand & Veiga 2003; Veiga et al. 2004). There is considerable variability in herd size among households in both of the sites, and in the CMER this also applies to each category in the herd. More illuminating are the data for purchase s and especially sales of cattle. Whereas we have anecdotal evidence of cons iderable cattle purchasing in Ur uar during the late 1990s (e.g., Toni 1999), data from the CMER suggest very li mited acquisition of cattle via purchase (in the

PAGE 101

101 year prior to the 2004/2005 survey ). Similarly, whereas households in Uruar sold several bulls, a few cows, and many calves, totaling roughly half the herd size, in the CMER, sales were a relatively smaller proportion of the herd, especially for cows and calves. This suggests that cattle ranching among small farm colonists in Uruar is very focused on breeding for sales, whereas ranching in the CMER may be more focused on a ccumulation (where cattle is an investment) than liquidation per se. This is also an indication that cattle in the CMER is undergoing an initial stage of establishment having less economic im portance for the household income composition than in Uruar. Nevertheless, it represents a st age of evolution in which cattle are increasingly gaining in economic importance in relation to the total annual income of a typical rubber tapper householder. Cattle are sold mostly in critical moments of household decision-making, including sickness of family members, housing improvement s, equipment acquisition, to pay off a debt, among others. These findings raise interesting ques tions about buyers of cattle in the two sites, especially Uruar, where there is some indicati on that small-scale operati ons sell to buyers who then sell to larger operations that fatten cattle for the slaughterhouse (Veiga et al. 2004). In the CMER there are few middlemen operating inside the CMER, usually someone with connections to larger-sca le cattle ranches in the region that fatten cattle for the slaughterhouse, like in Uruar. Middle-men offer advantages to rubber tappers, main ly freeing them of responsibility to pay trans portation costs, in addition to cash on receipt. The last two sections of Table 3-7 deal with issues specific to the two sites. In the CMER, a key issue concerning pasture formation for cattle ranching are the regulations limiting deforestation (up to 10% of the area of the CMER) and pasture (up to 5% of the CMER). While all available data indicate risi ng deforestation and expanding past ure areas in the CMER, a large majority of respondents in 2004/2005 supported th e 10% deforestation rule (Vadjunec 2007).

PAGE 102

102 However, a somewhat smaller majority supported the 5% pasture rule. Remote sensing data for the CMER (Vadjunec et al. nd) suggest that clearing (at least up until the late 1990s) had not yet approached the 10% limit, but if pasture continues to expand as a proportion of the area cleared, it was, in some instances, likely surpassing the 5% limit by the mid-2000s, which may help explain the lower degree of s upport for the 5% pasture rule. In Uruar, a key issue concerns the futu re of ranching as a livelihood option, given problems with alternatives, access to special credit lines in the 1990s, and the rapid expansion of ranching there since 2000 (e.g., Table 3-2). In the 1996 survey, roughly half of the households interviewed (as well as half of the households re-i nterviewed and therefore included in the panel) indicated an intent to expand their cattle herd (not shown in Ta ble 3-7). In the 2002 survey, such intentions had increased in frequency to 60-80% especially for expanding cattle herds. This suggests that liquidation may be proceedi ng alongside expansion in herd size. Discussion These results raise m any ques tions about cattle ranching among colonist smallholders and forest extractivists. Whereas ranching among co lonists has already re ceived attention (e.g., Walker et al. 2000; Tourrand and Veiga 2003), we focus here on interpreting the motives and process for the expansion of cattle herds in the CMER, followed by a comparative analysis of the processes at play in the Uruar colonization case. Despite the theoretical economic potential of NTFPs (e.g., Godoy et al. 1993; Grimes et al. 1994; Peters et al. 1989), the traditional livelih ood system of forest people in the region suffer from limited options for extractiv ist diversification and develo pment of value-added non-timber forest products. Most initiatives involving NTFPs have run into market obstacles that limit their economic benefit for local communities (e.g., Godoy 2000; Godoy et al. 1997). These profitreducing barriers include: absence of demand, gr eat distance to markets, lack of efficient

PAGE 103

103 marketing structures, weak social institutions, and poor development incentives. These activities are increasingly abandoned once mo re rewarding alternatives beco me available and as increasing pressure on household labor resources make these low-value, labor-intensive activities no longer attractive. Given the difficulties securing market s for value-added extractivist products, forest people find themselves losing faith in extractivis m where they fail to see economic gains. As a result, forest residents have turned toward the short-term economic benefits of cattle raising. It appears that decisions about cattle adoption among extrac tivists in the CMER reflect the economic value of cattle, the ability to avoid transportation and market access difficulties, and the lack of enforcement of restrictions on deforestation a nd pasture formation. Cattle are considered a means of storing and accumulating wealth, enabling extractivists to sell when cash is needed. This perpective is shared by an extrac tivist from the rubber esta te Paraguau, who sees his cattle as a life insurance polic y for himself and his children: E xtractivists are investing in cattle because its the easiest product to sell a nd it can guarantee securi ty for the producers, which you cant get with extrac tivist or agricultur al products. They dont give you anywhere close to the same price or market [demand]. Ranching is an attractive economic opportuni ty for CMER communities because of the easy market access for livestock. As the president of the Rural Workers Union of Assis Brasil put it: The Reserve didnt present the opportunity for rubber tappers to extract more profit from the forest. Cattle did, and they leave the forest on their own legs, so th ere are no problems with transportSo a lot of people found that th ey could make a living through cattle. In addition, competing land tenure systems in the region influence forest extractivists decision for cattle raising activit ies. Much of the CMER is surrounded by large cattle ranches and colonization projects with extensive pastur es. Ranching is practically the sole activity

PAGE 104

104 practiced in those neiboghbori ng areas, and forest extractiv ists perceive their neighbors practicing ranching with access to transportation, electricity, a nd housing, among other material goods, giving extra proof of the benefit of cattl e for their economic well-being. Such perceptions, and weak enforcement of forest clearing and pa sture establishment restrictions in the CMER, have thus also fostered adoption of cattle ranching there. As the president of the Rubber Tappers Association in Brasilia explaine d: The rubber tapper has thought a lot about his quality of life. They thought at first about their land that cattle ranchers were taking away from them. They thought about the land, but not about rubber prices The rubber tappers were seeing this happen and seeing the condition they themselves were in and what did they thin k? They thought about cattle ranching. Why not? The cattle ranchers and colonists raise cattle outside the reserve, and the rubber tappers who saw this began ranching, too. Thinking about it today, what makes the extractivists deforest is the developmen t of the reserve that never took off. While raising a small number of cattle has always been defended by extractivist communities since cattle provide milk and tr ansportation, herd size has steadily grown among forest people. As one Rubber Tappers Associat ion leader explained, How was the idea of raising cattle born for the extr activist? When everyone was engaged in tapping rubber and the price was very low, a rubber tapper had to tap 15 kilos of rubber to buy a jug of milk ( lata de leite ) in the city for his child. Ev eryone already has thoughts of k eeping a few heads for milk and transportation. They then start to feed on this idea and begin to like rais ing cattle. Herds start becoming large, and the extractivist starts to dedi cate himself to raising mo re cattle because of its economic attractiveness. Pasture formation provides opportunities that do not necessary depend on cash surplus. When a householder has pasture land available, capital constraint s can be overcome to initiate

PAGE 105

105 ranching by renting the land through informal c ontract to a neighboring householder with growing herd size but limited pasture. A neighbor with an established cattle herd might also place cattle under the care of another household who performs the work of a meeiro one who raises anothers cattle, and will be responsible for the feeding, maintenance and care of the animals. Often these cattle are young females, wh ich are mated with bulls, and the offspring are divided equally between the cattl e owner and the pasture landowne r. In addition, initial ranching might begin around family relations. For example, a newly married son is customarily granted a few heads of cattle if the family househol der head is engaged in the activity. Cattle trade is a year-around activity pr oviding sustained cash influx for the CMER residents, although more intense dur ing the dry season since transpor t is easier. Residents of the CMER also know that when they need cash they can go to the city to sell their cattle at any time. Another method that facilitates th e cattle trade market all year around is through agreements that middlemen have developed with some residents of the reserve, usually someone who owns a larger cattle herd in specific areas. These residents buy cattle for the middlemen whenever a neighbor puts them up for sale. The middlemen provide some advance cash for their purchases and thereby dominate the market among householders. It is uncertain if forest extractivists will continue to increase ranching in the extractive reserve, as its c ontroversial aspects have not yet been openly discussed among different actors of extractive reserve management. Although both deforestation and small-scale cat tle ranching are on the rise in the CMER, this does not mean that traditional livelihoods systems offer no pr omise, as local govern mental agendas have developed several experiments to st rengthen the state extractivist economy. Comparative Analysis Regional m arket prices largely explain cat tle adoption among smallholder colonists and forest extractivists, despite their cultural differences and contrasting historical trajectories.

PAGE 106

106 Whereas Uruar colonists were settled in th e region with the opening of the Transamazon Highway, forest extractivists in Xapuri were organizing against th e very same policies of road construction and agricultural land settlement for cattle ranching N onetheless, roads have played an important role connecting col onists as well as extractivists to local and regional markets, which has motivated shifts toward products easil y transported and marketed, such as beef. While there are important contrasts in the productive activit ies and the economic difficulties facing colonists in Uruar and extractivists in the CMER, they also found similar economic opportunities involving cattl e ranching. The cyclical adap tation among colonists is not observed among forest extractivis ts, who practice subsistence ag riculture, almost no perennial crop production for marketing, and have managed to maintain their traditional production system over the years. Cattle ranching nonetheless beca me an economically attractive livelihood option. Forest extractivists faced an optio n: to be part of the economic pr ocess with a different level of well-being in which forest products have a lesser role. Such contrasts alongside the convergence around cattle ranching raise additional comparative questions about prospective change s among colonists in places like Uruar and extractivists such as those in the CMER. Smallhol der colonists in Uruar have existed for three decades and have seen many changes in land use practices and economic fortunes (Perz 2003). With diversified market oriented land use practic es, colonists have acquired important strategies to respond to fluctuating regional markets. W ith continued improvement of market links, agricultural technologies, and land consolidation, colonists tend to improve diversification with mechanized agriculture and intensive cattle pr oduction. The cattle sector will continue to increase in the regional economy, playing a larger role in the household economy. Land consolidation is likely to benefit households th at intensify cattle rais ing, and colonists will

PAGE 107

107 gradually turn into larger la nd holders, with pasture for cat tle as the predominant land use (Aldrich, et al. 2006). Government development policies in Acre are also encouraging the extractivist population to diversify modes of production, locally called neo-extrac tivism, which refers to an alternative vision of forest use and conservati on that builds on the cu ltural, political, and economic context of local people (Rego 1999). Ne o-extractivism argues that to improve the economic situation of extractivis t families, alternativ e technologies that in crease the marketing return for traditional non-timber forest products n eed to be combined with the incorporation of new extractive products. The government of Acre cr eated a series of incen tives for entrepreneurs to establish sustainable logging activities and for community-based timber management operations. Acre recently became the first state in Brazil to achieve Forest Stewardship Council (FSC) certification in a community-managed fore st, to improve their access to national and international markets. Advocates of sustaina ble logging have argued that timber-management can play an important role in income genera tion, slowing down cattle expansion in the CMER, thereby rendering viable forest extractivism. Howe ver, the proposal for lo gging in the CMER has generated a polemic debate among ex tractivists due to concerns about damage to the forest and prohibitions on logging in the ERs. Furthermore, without proper management and monitoring, logging is unlikely to be an sust ainable alternative. Any alternat ive to cattle expansion in the reserve has to build on a comprehensive understandi ng of the processes that have contributed to it, including limitation of dive rsified extractivist production, and the short-term economic benefits that cattle ranching provides. No single solution is li kely to change this current economic trend of forest extractivists economic development. There is an undercurrent of political anxiety in policy-makers and rubber tapper leadership a bout the growth and uncertainty

PAGE 108

108 of cattle ranching in ERs, reflected in the feelin g of symbolic discontinuity that cattle ranching represents for forest extractivism as a culture and the historical trajectory out of which it developed. While ranching is a reality among the productive activities pr acticed in the CMER, cattle has caused its own political controversy among the extractivist population. The debate over diversified extractivism with forw ard linkages under the neo-extractivist concept is likely to limit cattle expansion in Extractive Reserves over the long-run. But its current adoption shows immediate needs to specify the ways in which di versified development appr oaches affect forest conservation and the socio-economic well-being of communities in the region. Conclusions Showing sim ilar trends regard ing the growth of cattle rais ing in Xapuri and Uruara, the results suggest a regional push toward cattle ra nching expansion among colonist smallholders and forest extractivists in the Brazilian Am azon. Cattle ranching among colonist and forest extractivists are driven by market forces, but al so have distinct politic al development contexts. Smallholder colonists present a more diversifie d economy with greater market links in which cattle ranching is consolidated as a economic practi ce, while forest extractivists, regardless of the potential of forest products to improve local livelihoods, have more limited economies in which cattle represent a viable option under an uncertain economy, and they have not yet confronted the political and development controversies of cattle raising in Extractive Reserves. Cattle ranching among forest extractivists in Xapuri is a challenge and may represent an opportunity to promote a broader debate to review forest res ources co-management pacts by government institutions and grass -root organizations for Extractiv e Reserve administration in the very place where it was negotiated twenty years ago.

PAGE 109

109 CHAPTER 4 LAND-USE/LAND-COVER CHANGE AMONG RUBBER TAPPERS IN THE CHICO MENDES E XTRACTIVE RESERVE, ACRE, BRAZIL Introduction Tropical deforestation is arguably the most significant type of land-use/cover change (henceforth, land change) underway globally, give n its multi-scalar and -dimensional elements affecting a wide range of ecosyst em services from landscape-level precipitation, to river basin hydrology, and to the functioning of the Earth sy stem, including global climate change (Myers 2000, Watson et al. 2000, Steffen et al 2004, Bunker et al. 2005, Lambin and Geist 2006). No region of tropical deforestation has drawn more concern than the Amazon basin, especially the Brazilian Amazon. The Amazon remains a hots pot for tropical deforestation (Achard et al 2002; Asner et al. 2005) with profound, documented impacts on a full range of ecosystem to earth system concerns (Nobre et al 1991, Salati and Nobre 1991, Gash et al 1996, Laurence et al 2000; 2002, Ferraz et al 2003, Houghton et al 2005). Deforestation follows the expansion of the logging and agricultural fr ontier, both increasingly driven by global markets (Faminow 1998, Mertens et al. 2001), while traditional people struggl e over access to land and resources (Schmink and Wood 1992). Pan-Amazon and especia lly Brazilian policies, therefore, have the challenge of balancing environmental concerns with land reform while ensuring economic growth and development (Hecht and Cockburn 1990, Smith 2001, Kainer e t al. 2003). The Brazilian Extractive Reserve System (ER) is one form of land use and entitlement seeking this balance. Threatened by cattle ra nching and large-scale agricultural projects, traditional extractivists in the regionthose who live mainly off of the extraction of non-timber forest products (NTFPs) such as rubber, Brazil nuts and hearts -of-palm have fought for the legal recognition of their land ri ghts. The ER was born out of th e increasingly recognized need for sustainable development (WCED 1987, IUCN 1995), within the context of economic and

PAGE 110

110 social justice, particularly for Chico Mendes and the Rubber Tappers Union (CNS 1985). Under this system the government holds the title to the land, while extractivis t associations hold concessions to its use. The ER, therefore, ha s two complementary goals: (1) the conservation of important ecosystems, such as forests, and (2 ) the preservation of traditional extractive economies and cultures through sustainable development (MMA 2004). Since its creation in 1990, ER has been cham pioned by many as a win-win (sustainable development) model in which reserve residents cons erve the forest in a form of sustainable use, which is expected to lead to economic grow th and improved well-bei ng (Murrieta and Rueda 1995). ER thus joins a litany of efforts worldw ide to use working preserves or parks to maintain forest cover and ecosystem integrity th rough maintenance of traditional uses (Alcorn 1992, Gomez-Pompa and Kaus 1992, Terborgh and Van Schaik 2002, Tucker 2004, Zarin 2004). Recognizing that many forest lands are co-e volved landscapes (Alcorn and Lynch 1992, Clement et al. 2003), the win-win character of these efforts, especia lly regarding the material development of the peoples involved, remains an open question, with critics of parks citing problems of corruption, bureaucracy, political instability and the fu rther marginalization of the poor, among others (Terborgh and Peres 2002, Van Schaik 2002, Van Schaik and Rijksen 2002). This concern is muted by some advocates of th e ER who, in their attempt to legitimize the rubber-tappers movement and alternative forms of development, may have overly romanticized both the rubber tappers and the main objectives of the extractive reserve model. Since his assassination in 1988, rubber tapper leader, Chico Mendes has himself reached almost mythic proportions, being described as an ecologist, a rain-forest Ga ndhi, an eco-martyr, and one of the most important environmental thi nkers in history (Gale 1998, Palmer 2002, Maxwell 2003, Revkin 2004). To others, he was a hard-hitting communist party labor leader and a

PAGE 111

111 radical political militant (Hecht and Cockburn 1990, Maxwell 2003). Such extreme descriptions are not limited to Chico Mendes but to the rubber ta ppers themselves, and even to descriptions regarding the sustai nability of the ER model. A polemicized debate concerni ng the sustainability of the ER model has emerged with little common ground between advocates and critic s. Proponents argue that ERs serve to lower deforestation rates, protect biod iversity and ecosystem integrit y, and as buffers preventing or inhibiting land-cover destruction by fire, while preserving the cu lture and ensuring the economic livelihood of traditional forest-dwelling pe ople (Allegretti and Schwartzman 1989, Anderson 1990, Hall 1997, Nepstad et al. 2006). Brazilian environmental policy makers, environmentalists, and the Nationa l Council of Rubber Tappers (CNS), often backed by financial support from the international community, have fa vored this type of green development and press for the creation of ERs and other conser vation units to cover 10 % of the Brazilian Amazon, an estimated 500,000 square kilo meters, or roughly the size of Spain.1 By February 2008, 39 federal Extractive Reserves, and 25 state Extractive Reserves, co vering an area of over 12 million hectares, had already been created in the Brazilian Amazon. Critics question the reserves in terms of their economic, social, and environmental sustainability (Fearnside 1989, Browde r 1992). Specifically, the reserves have been criticized on issues of biodiversitythat extractivists and ex tractive activities may have a negative, predatory affect on both flora (Moegenburg and Levey 2002) and fauna (Peres 2005). Furthermore, some economists suggest that, historica lly, extractivism has not been able to alleviate poverty, and this exclusive kind of economy will eventually disapp ear, as economically valuable forest products 1 Increasingly, the concept of extractive reserves is expanding, being applied to a diverse range of ecological and social contexts outside of Amazonia, including non-rainfor est environments, such as marine areas on the Atlantic coast of Brazil, and increasingly in savanna ecosys tems where soy-bean expa nsion threatens both local environments and traditional populations (Theulen 2006).

PAGE 112

112 inevitably end up being grown on plantations, at less cost and of higher quality (Homma 1989, Browder 1990, Homma 1994). Much of this debate is theoretical in na ture, lacking systematically collected socioeconomic data collected on ERs. More recent studies in the Chico Mendes Extractive Reserve (CMER) have tried to ground these arguments by focusing on field-based data collection and analysis of both livelihood traj ectories and the ecological impact s of such livelihood decisions by residents of the CMER (Sassagawa1999, Gomes 2001, Wallace 2004, Eringhaus 2005, Vadjunec 2007). A deeper understanding of this debate, aided by the results of research being carried out in the CMER, is increasingly important because criticisms surrounding the ER system have been re-ignited. In August 2005, runaway fires ravaged the CMER for three weeks before they could be contained. The fact that it was the rubbe r tappers themselves who accidentally started the fires outraged much of the public, sparking tens ions by some regarding the legitimacy of the entire ER model. As a result, the reserves abi lity to perform as a conservation unit is currently under fire in the Brazilian nati onal press, with a few critics accusing many ER parks as a conservation lie, grounded in pseudo science, and offering as proof examples of (nonFederal, non-ER) state-run ERs in Rondnia, where between 60-100 % of forest cover has already been lost (Brito 2005b).2 The CMER has witnessed a substantial growth in pasture and small-scale cattle ranching triggering deforestation in some areas (Gomes 20 01), although the scale of deforestation is low compared to the surrounding region. A three pe rcent total forest-cove r loss between 1986 and 2 While we recognize that state-run extractive reserves in Rondnia are currently undergoing extreme land-use pressures (Euler et al. 2008), we argu e that they should be considered an exception, not the norm, and addressed separately from the ER system.

PAGE 113

113 1999 was recorded in the reserv e by two remote sensing analyses (CNPT 1999), one of which suggests that some communities in the CMER are close to reaching their allowable limits of deforestation and would reach them by 2003 (Sassagawa 1999). While variation among rubber tapper communities is addressed in this last assessment, less attention has been given to explaining the reasons why such variation exists (Gomes 2001) and virtually no attention has been given to assessments of deforestation as associated with the variation in community livelihoods in the CMER. In f act, considerable variation in community economies has emerged in the reserve, with some communities focused mo re on extractivist activities and others more on agriculture and cattle rearing. This paper compares deforestation in the co mmunities in the CMER with their diverse livelihoods practices. This exercise is importa nt because it helps to illuminate the relative impacts on forest conservation and household income of different human-environment conditions existing in the rese rve, thus providing insights about the ideal conditions for sustainability of the CMER considering its dual purposes of forest cons ervation and sustainable development (economic growth and resident well-b eing). In order to explor e this relationship, six communities are examined in the CMER, four of which were projected to surpass their legal deforestation limits by 2003. Using remote sensing and household surveys, this study asks the following sub-questions: (i) What are the land-use practices in the six communities and how are they changing? (ii) What are the amounts and rates of deforestation and secondary regrowth within the six communities between 1986-2003? (iii) Have any of the six communities surpasse d allowable deforestation limits within the reserve, thus compromising the rese rves role as a conservation unit?

PAGE 114

114 Study Area Chico Mendes Extractive Reserve The CMER, located in s outheastern Acre (als o known as the Alto Acre Region), is the most famous of all the ERs. It was created on March 12, 1990 by Federal decree, comprises 970,570 hectares, and is located approximately between latitude 10 05 41 S and 11 00 00 S and longitude 67 56 W and 69 48 W (Figure 4-1). Approximately 12,000 people (about 2,000 households) live within 46 different seringais or rubber tapper estates, within the reserve (CNPT 1995). The BR-317 highway runs along the eastern and southern bou ndaries of the reserve, connecting the capital of Rio Branco to Acres other major urban centers of Xapuri, Brasilia, and Assis Brasil, and co nnecting Brazil to both Peru an d Bolivia. The more isolated regions of the reserve can be access ed by the Acre, Xapuri and Iaco rivers. The CMER is, for the most part (73%), an open tropical forest biome high in both palm and bamboo species diversity with an average annual rainfall of 2,200 mm, a mean temperature of 26 C, and a marked dry season (from May to A ugust) where average temperatures climb to about 38 C. The remaining 27 % of the reserve is clas sified as dense tropical forest, which can mainly be found in the municipality of Brasilia (C NPT 1999). This type of forest is often richer than open tropical forests in many key extractive species, such as aai an d the Brazil-nut tree, thus providing some communities with inherently greater extractive potential. Topography of the region is slowly undulating, or h illy, and elevation varies between 100-300 m. In general, the reserve is dominated by a mix of mainly eutrophic and dystrophic soils, which, according to state authorities, provides workab le soils and marked agricultural pot ential with the adoption of proper management practices (GovernmentofAcre 2000a). Although the reserves title is held collectively by the reserves associations, individual resource parcels are given to each household, based on a traditional land-use system known as

PAGE 115

115 Figure 4-1. Study area:and study sites within the CMER: (1) Icur i, (2) So Francisco, (3) Paraguau, (4) Humait, (5) Porongaba, (6) Filipinas the colocao where the emphasis is on resources in the parcels, not on the land itself. A typical colocao contains the rubber tappers house and other farm buildings, water well or small creek, patio garden, fruit or chard, agricultural plot and pasture, if a ny. Historically, the most important asset in a colocao is the number of estradas de seringa, or rubber trails, held by each household. A household does not hold de jure title to the land, but instead possesses the right to use the res ources found within the colocao and rubber trails (mainly Brazil nut and rubber trees, historically holding the most value). At the reserves creation, each rubber tapper was guaranteed at leas t three rubber trails, which is estimated to be roughly equivalent to 30 0 ha, although some families have more or less after trading, buying, or divi ding up properties. In most seringais there is no communal forest as in a traditional common property regime; all forest is claimed by some household or another.

PAGE 116

116 The Utilization Plan (PU) for the CMER, which wa s made by the rubber tappers themselves, sets the allowable deforestation limits at 10 % of the colocao or theoretically 30 ha, with 5 % (or 15 ha) allowed for pasture conversion. The nature of the colocao makes it difficult to map as rubber trails often intersect and overlap. It is important to note that households do not have traditional geometrical plots that are easily ma pped and bounded. Furthermore, the CMER is a common property system where enforcement and regulation occurs mainly through the social organization at the community level. As a result, most monitoring of the 10 and 5 % rules occurs at the community level, instead of the house hold level, if at all (Sassagawa 1999, Gomes 2001, Ankersen and Barnes 2004). Recently, the efficacy of the reserve has been questioned due to the expansion of small-scale cattle ranching with in the reserve (Gomes 2001). In order to understand the controversy surrounding this recent land-use and livelihood change, it is important to look to the past. In the next section, we discuss the history of rubber in Acre, resulting in the CMER today, and the changes in land-use currently underway. Land Use History and Livelihood Traject ories Among Rubber Tappers in Acre Starting with the rubber boom of the late 19th century, the state of Acre has had a long and varied history of rubber tapping and latex produ ction (Barnham and Coomes 1996). By the late 1880s, the rubber trade was so economically impor tant to the local, regional, and global economies that, accordingly (Hecht and Co ckburn 1990: 80, Stokes 2000: 232), Acre was claimed to be one of the most valuable and important pieces of commercial real estate on earth, where at its peak in 1912 wa s exporting approximately 1,200 tons a year to French, Belgian, British and American firms (Colho 1982: 68). This booming economy required many workers in a place where populations were sparse, the landscape was unforgiving, and the labor supply scarce. As a result, the peasant poor from the drought-stricken areas of the Brazilian Northeast were encouraged by the Brazilian

PAGE 117

117 government to migrate into the rubber territories of the Amazon. In the drought of 1877 alone, an estimated 100,000 North-easterners migrated into the Amazon region (Hecht and Cockburn 1990: 80). The rubber barons suppl ied the rubber tappers with trade goods, food, work supplies and even the cost of their passage on credit, thus keeping the tapper indebt ed to continue tapping rubber for a relatively low price. Rubber tappers were regularly discouraged and even banned from practicing agriculture in their colocao resulting in both the protection of precious extractive species and historically low defo restation rates, while keeping the tappe rs dependent on the rubber barons and middle-men w ho delivered supplies. The first rubber bust occurred in 1915 with the development of the Southeast Asian rubber plantations, when rubber exports for the whole of Amazonia fell to just under 30,000 tons annually (Colho 1982: 70). Not only was the co st of rubber production lower in Asia; Asian plantations also received a better price for a higher quality pro duct, thus making it difficult for Brazilian-based firms to compete. By 1932 the market all but bottomed out, with the whole of Amazonia producing less than 5,000 tons of rubber annually (Dean 1987: 169). The second and short-lived rubber boom came with the advent of the Second World War, and an urgent need for latex among the United St ates and its allies (C orra 1967; Martinello 1998). The Brazilian government was called upon to triple its latex production, creating the Rubber Reserve and calling for soldados de borracha, or rubber soldiers, to fulfill their patriotic duties, offering up promises of a soldie rs respect as well as a pension (only recently realized) in return for their rubber tapping serv ices. By 1947, the US and Brazilian governments had pulled funding for the Rubber Reserve, an d once again, the Amazonian rubber economy went bust, causing many rubber barons to abandon their rubber estates, le aving rubber tappers to fend for themselves.

PAGE 118

118 While many rubber tappers fled to Acres urban centers, many tappers stayed in the abandoned rubber estates supplying local and regional markets with latex (otherwise banned for importing into Brazil) and engaging in hunting, Brazil-nut collecting, and subsistence farming. The rubber tappers were left alone until the 1970s when the military government officially opened the Brazilian Amazon for development. In Acre, this new development policy often led to violent land conflicts between rubber tappe rs and cattle ranchers where rubber tappers were evicted and former rubber estates were cleared by cattle ranchers. It was under the context of such conflicts, along with in creasing environmental concerns, that both the CMER and the entire ER system were created (Allegretti 2002). According to a reserve-wide study comp leted immediately following the reserves creation, rubber was the main source of income for residents of the reserve, representing 44.7 % of the total household income, while 24.6 % of the household income came from Brazil-nut collection (in other words, al most 70 % of a households income was derived from NTFP extraction) (Feitosa 1995: 70, see also CNS 1992) Agricultural production represented 21 % of the household income, while small animal breed ing and cattle rearing represented 4.8 and 4.2 % of the total income respectively (Feitosa 1995: 70 ). At this time, not only was rubber production the most important economic activity, the CMER produced more rubber than any other extractive reserve, producing on av erage 714 kg of rubber per family per year (Feitosa 1995: 70). Ironically, in the mid-1990s, just as the rubbe r tappers themselves were gaining the long overdue recognition they deserved, and as the CMER was being formalized, Brazilian policies ended the forced import bans and national subsidie s placed on rubber. The rubber tappers finally gained land security, but at the same time lost many of the protections and incentives placed on their main source of livelihood. They also, during this time, gained autono my to make their own

PAGE 119

119 land-use decisions. As a result, many rubber tappers have abandone d latex production altogether and are increasingly drawn to small-scale agri culture, animal husbandry, and even cattle production, resulting in the deforesta tion of concern w ithin the CMER. With the creation of the reserve came th e long overdue recognition and inclusion of rubber tappers in Amazonian development polic y. On the one hand, rubber tappers are faced with long-term market challeng es regarding latex production and other NTFPs. On the other hand, with the creation of the re serve, they also have increas ed road access and therefore opportunities to participat e in more diverse markets. The dominant market opportunities available in Acre, despite the attempts of the pro-forest government to build a sound sustainable development policy, unfortunately, are often not based on sustainabl e strategies, but instead are influenced by local and regiona l economies operating outside of the reserve, which privilege cattle, and to a lesser extent, agriculture, over rubber and NTFPs. Overall, cattle ranching is now one of the most important economic activities in Ac re and contributes more than 60% of the state tax revenue (Valentim et al. 2002) Furthermore, recent studies suggest that smallholders are responsible for roughly 62 percen t of the new deforestation in Ac re (Barreto et al. 2006). Activities of large-hold ers are increasingly co nstrained by the large am ount of protected areas in the region. This constraint, however, provides new opportunities for historically neglected smallholders to participat e in these new market econom ies. As a result extractivists ar e increasingly turning to more intensive land uses, such as market agriculture a nd cattle raising. This brief overview of land use trajecto ries among the rubber ta pper population points out the reality that the political economic cond itions under which the original rubber tapper land holdings and economic livelihood orie ntations were established ma y no longer exist. Many of the rules and regulations applied to their seringais remain, however, based on the assumption

PAGE 120

120 that the extractive reserve model can improve re sident well-being while protecting the majority of the forested areathe win-win solution. Methods Site Selection Research was carried out in a total of six rubber es tates within the tw o m unicipalities of Assis Brasil and Brasilia (three estates each) of the reserve (Figure 4-2). The base characteristics of each rubber estate ar e provided in Table 4-1. Figure 4-2. Location of the study communities within the CMER: (1 ) Icuri, (2) So Francisco, (3) Paraguau, (4) Humait, (5) Porongaba, (6) Filipinas Communities were chosen with the help of th e municipal level reserve associations based on the following criteria: (i) Rubber estates were open and inviting to outsi de research. This was an important factor as the reserve is considered Fe deral property and outside research requires an invitation from the

PAGE 121

121 communities and a research license from the National Centre for the Sustainable Development of Traditional Populations (CNPT). (ii) Among those rubber estates chosen by the associations, estates were stratified based on perceptions of the reserve leaders on traditiona l extractive orientation versus non-traditional market orientations. Table 4-1. Characteristic s of the study sites Brasilia Assis Brasil Filipinas (6) Porongaba (5) Humait (4) Paraguau (3) So Francisco (2) Icuri (1) Estimated No. households 72 26 32 40 73 72 No. households surveyed 26 17 19 20 19 29 Ave. family size 6.96 4.82 5.56 8.15 5.15 4.61 Av. holding size (ha) 608.00 429.00 432.00 560.00 416.00 300.00 Total Area Seringal (ha) 29.908 8,937 14,487 19,511 29,933 64,405 Forest canopy type Dense Dense Dense Dense/Open Open Open Impact by division of watershed? No No No No No Yes Main form of transport Road Road Road Road Road Water/ Road Distance from community base to nearest urban center 36 km 23 km 30 km 10 km 22 km 75 km Livelihood Emphasis NTFP Cattle/NTFP Cattle/NT FP Cattle/Ag NTFP/Ag Cattle/Ag Remote Sensing Analysis Our rem ote sensing analysis follows from a collaboration between researchers with different research objectives and, in one case, study sites. Theref ore, satellite image analysis was carried out on the footprint level for the Alto Acre region (10 scenes ). LANDSAT TM and ETM+ images were obtained for 1986, 1992, 1996, 1999, and 2003 for Path/Row 02/67 and 03/67 (Image Source: Tropical Rain Forest Informa tion Center of Michigan State University). Chronic cloud and haze problems common to the region required that the image synchronization match the short dry season, between mid-June a nd early September, rather than more narrowly

PAGE 122

122 circumscribed dates. We started our time series in 1986 in order to get a glimpse of land-use and land-cover before the reserve was created, and used subsequent years to see patterns in change, but limited our timeframe to 2003 in order to match our remote sensing data to our survey data, thus capturing more specific explan ations of land-use and land-cover change at a particular point in time. Ground data were collected between October 2003 and August 2005 for image georeferencing, as well as over two hundred add itional training samples of various land-cover types. A training-sample protocol was de veloped based on the specific ecology and landuses/covers of the region and a training-sample data base was created, including information on the geographic coordinates of the sample us ing a Global Positioning System (GPS), the landuse/land-cover type, sketch maps, and digital photos. We also collected brief land-use histories given by the land-owners to aid in the developm ent of training sample data over various time frames (other than 2003). Given the spatial reso lution of our imagery (i .e. 30m), samples were limited to areas one hectare or greater. Geometric rectification with a linear first-order polynomial fit was carried out on the 2003 images based on ground control points collected in the region using a GPS, using roads and intersections, and when a good geographic distri bution was necessary and roads were lacking, both rivers and waterways. All resampling was done using a nearest neighbor resampling algorithm. The 2003 images were then used as a base to which the remaining scenes were resampled. On average, 12 ground c ontrol points (GCPs) were used for each image, with at least one point in each quadrant of the im age, resulting in a root mean square error less than 0.5 pixels (or < 15 m) in each case.

PAGE 123

123 Radiometric calibration (RADCAL) and dark subtraction (ATCORR) model developed by Green et al. (2002) was used to correct images for comparation. Although this method helped to clean up the images in general, it did not clear up all atmospheric problems, particularly scattering effects on the 1992 02/67 image. We adopted a hybrid unsupervised/supervised approach, which allowed us to create a more am ple classification schema, even in years where training sample data were lacking. An unsupervised classification was carried out on each image using a clustering algorithm to define land-cover classes based on spectral statis tics with a standard nu mber of 100 classes, 12 maximum iterations, and a convergence threshol d of 0.95. Our 100 classes were gradually collapsed into 12 classes: dense forest, open/ bamboo forest, initial secondary succession, intermediate secondary succession, pasture with trees, managed clean pasture, unmanaged dirty or scrubby pasture, agriculture/bare soil, high density urban, water, cloud, and cloud shadow (see Table 4-2 for description of classes). Class assignment involved three assessments: (i) Visual interpretation of the image sequence (geographically linked) using expert knowledge and vectorized ground data and land -use histories collected in th e field to guide our decisions regarding class ownership, and analysis of the sp ectral behavior of indivi dual pixels belonging to known classes using a spectral profile tool; (ii) co mparative analysis of the spectral behavior of the signatures for each of the 100 classes as well as groups of similar classes using the signature mean plot module; and (iii) th e interpretation of un supervised classes ba sed on their spectral seperability statistics generated us ing a transformed divergence module. Based on these criteria, the initial 100 cla sses were individually analyzed and assigned, with the help of the vectorized ground data and land-use histories we collected in the field, to a final classification schema, spectrall y similar classes were merged w ith like classes, while others,

PAGE 124

124 such as those representing noise, were eliminated altogether from the cla ssification. Due to the large number of initial classes, this was done in several stages producing a series of intermediary signature files for each year: thus, a 100 class sign ature file might be reduced to 60 classes, then 30, and once again until the final desired 12 classe s had been reached. In order to ensure maximum spectral seperability of our classes, Steps 1-3 were completed and evaluated various times throughout the intermediary stages. Table 4-2. Land-use classification schema ID Class Description 1 Dense Tropical Forest Dense, closed tropical forest common to the region 2 Mixed Open/Bamboo Forest Open semi-deciduous forest intermingled with patches of bamboo and/or palm species 3 Secondary Succession 1 Herbaceous, seedling and sapling vegetation up to 5 meters high, typical of agricultural land left to fallow or nonmanaged pastures 4 Secondary Succession 2 Intermediate secondary stage with woody vegetation up to 15 meters high, closed canopy 5 Pasture with Trees Managed pasture intermingled with trees 6 Managed Clean Pasture Managed clean pasture, open grasslands with substantial green cover 7 Tall grass/Wetlands Tall grass/wetlands, lush green vegetation typical of areas with poor drainage or grassy wetlands 8 Agriculture/Bare Soil/Overgrazed Past ure At peak of dry season, areas lacking green vegetation cover, and/or showing patches of bare soil 9 High Density Urban Presence of asphalt and human-made surfaces with high reflectance (found outs ide of reserve only) 10 Water Rivers, creeks, ponds and other water bodies 11 Clouds Cloud 12 Cloud Shadow Cloud Shadow The final signature files representing 12 land-use/land-cover classes for each image constituted the input for a supervised classification using a probability-based Maximum Likelihood algorithm. After classi fication and evaluation, the classe s were further reduced to a more simple and realistic clas sification schema given the cons traints with such a long image sequence: Forest, Secondary Succession, a combined Agricultural class (i.e., pasture, agriculture, bare-soil), Water, and Cloud/Shad ow. Aggregation helped eliminate sources of error among our

PAGE 125

125 anthropogenic classes, in partic ular, given that all images co rrespond to the dry season, when vegetation is commonly under hydrological stress. To deal with common problems caused by atmospheric scattering due to cloud cover problems in our 1992 image specifically and misc lassifications and c onfusion caused by other anomalies (e.g., tree shadow and natural bamboo d ie-off in one instance ), post-classification techniques, such as a 3 x 3 kernel filer, and to a lesser extent, polygon fil ling, were used in some instances and in specific area of interests (AOI s), followed by careful in spection and the visual comparison of problem areas with the original ba nds geographically linked to the filtered image. An independent hold-out sample of roughly 150 ground reference points were digitized into polygons and used in an accuracy assessment for both the 02/67 and 03/67 footprints for our 2003 images (representing both 895 and 1,264 pixels respectively). In bo th instances, we achieved an accuracy level of 85 % (p-value of 0.05). Land-cover transition matri ces were generated using a matrix or cross-tabulation module. Starting with the 1986 and 1992 images for each footprint, the matrix module was run successively adding each year in the sequence until a final transition matrix representing all years between 1986 and 2003 was complete. These transition images follow the state of any given pixel and its change from year to year. Sepa rate transition matrices were produced to track deforestation, secondary regrow th, and a final matrix that attempted to capture the swidden/fallow cycle within each rubber estate, resulting in 6 final images, 3 for each footprint. Given uncertainties due to the spectral similari ties between anthropogenic land uses as well as the problematic nature of having such a long (18 yr) timeframe, pasture and cultivation classes were aggregated together to minimize classification error. Our transiti on matrices, therefore, used a simple forest-nonforest-secondary forest classification, thus mi nimizing the chances of

PAGE 126

126 error common between anthropogenic classes. As a result, it is impossible to distinguish through imagery analysis if any of these rubber estates ha ve greater than five pe rcent of their allowable area in pasture, focusing instead on the efficacy of the 10% rule. Lastly, digital boundaries of the six seringais were then used to create masks producing 18 final images (capturing deforestation, regrowth, the swidden/fallow cycle) for each community, allowing the calculation of the rates and amounts for each transition type between any given time period. Household Surveys A total of 130 household interviews were com p leted within the six rubber estates between March 2004 and December 2004 by two of the authors (Vadjunec and Gomes) and two field assistants in conjunction with a local nongovernmental organization (NGO), the Group for Research and Extension in Agroforestry Systems in Acre (PESACRE). As often as possible, the interviews were completed with both the male a nd female head of the household. Each interview was conducted in Portuguese, took about three hour s to complete, and cont ained structured and open ended questions regarding specific land-use, production a nd sale of agricultural and extractive products, small animal and cattle produ ction regarding production and yields for the year previous (2003), as well as household charac teristics and history, and questions regarding the rules (both formal and informal) and mana gement of the reserve, the history of the households involvement in the rubber tappers m ovement, and the social organization within each seringal or rubber estate. Additionally, sinc e many households operate on a mainly subsistence or recently transitioning mixedsubsistence and market livelihood system, we constructed a livelihood index to show what we feel is a more realistic picture of overall economic development and well being. The index to tals the presence or absence of 30 social, economic, and health-related development indicators (e.g., overal l house construction, possession of important household appliances, ag ricultural equipment, access to an electric

PAGE 127

127 generator or other power source, water pump or well, access to safe drinking water or water treatment, trash and sewage disposal, acces sible healthcare and education, mode of transportation, overall satisfaction with the reserve etc.). Although the original intent was to choose households at random, difficult access and incomplete residency lists caused us to adopt a more opportunistic sampling strategy. Few roads exist in the reserve and some communities required a three day wa lk to reach. The only residency list available to use was outdated, and furthermore, was compromised by lack of information on households that are not officially members of the associations. Sometimes outsiders and non-members of the movement have purchased land, often illeg ally; they therefore may not be registered residents, but they no-doubt bring different la nd-use practices and economic orientations to the reserve. Of the 130 households selected, 65 are involved in the government sponsored environmental program ProAmbiente, which, although still in initial phases, proposes to pay small producers for environmental services. Th ese households are officially recognized by and participate in the municipal-le vel extractive associations. Th ese households were selected through a series of open public meetings held by the Rubber Tappers Union to generate interest in ProAmbiente. The remaining 65 households participatin g in this study were chosen based on both maximizing the geographical distribution of households selected within each community to the greatest extent possible, and with the help of key informants, emphasis was placed on representing the range of house holds involved in the various ex tractive/non-extr active activities that were brought to light in our initial co mmunity meetings (i.e., cattle, extractivism, agriculture, or a mix of activities). After two years of field experience, we estimate that 40% of the total households were interviewed in each of the study sites.

PAGE 128

128 Results Remote Sensing Results Figure 4-3 and 4-4 illustrates patterns and spatia l/tem poral trajectories of land-use/land-cover change in studied Seringais from 1986 to 2003. Seringal Humait had the largest percentage of total area deforest ed by 2003, totali ng 9.20 % of the seringais (Table 4-3). Porongaba and Paraguau fell between (7.84 % and 6.20 % respectively). Filipinas and So Fr ancisco had the next lowest amounts of deforestation (3.80 % and 4.93 % re spectively). Lastly, Icur i, had the lowest total percent deforestation (3 .58 %) among all of the six rubber estates. Deforestation rates were on average highe st during the initial 1986-1992 time period, accounting for between 17-39 % of the to tal area deforested by 2003, averaging 24 % of the total area deforested per seringal This period was marked by extreme changes in the lifestyles of rubber tappers, including part of the initial violent conflicts between rubber tappers and cattle ranchers and the subsequent creation of the reserve in 1990, followed by the discontinuation of rubber subsidies that occurred in the 1990s. The periods between 1992-1996 and 1996-1999 indicate a general overall drop in deforestation rates within each of these six estates. An overall drop in deforestation rates during this time fram e was also observed elsewhere in the region (Ludewigs 2006). This overall drop in deforestation rates is perhap s attributable to the initial optimism of the rubber tappers in securing the reserve, initial international funding and support, as well as the end of violen t land conflicts throughout the region. Between 1999 and 2003 this trend changed toward more deforestation, for the mo st part, with the excep tion of So Francisco, which continues to see a drop in deforestation ra tes. Contrary to prediction, however, none of the study areas have surpassed the allowable limits of deforestation as set by the 10 % rule in the Utilization Plan.

PAGE 129

129 Figure 4-3. Trajectories of LUCC in the seringais in Assis Brasil within the CMER: (1) Icuri, (2) So Francisco, (3) Paraguau Figure 4-4. Trajectories of LUCC in the seringais in Brasilia within the CMER: (4) Humait, (5) Porongaba, (6) Filipinas 1 2 3 6 4 5

PAGE 130

130 Table 4-3. Percent deforestation in six seringais from 1986-2003 Brasilia Assis Brasil Filipinas (6) (ha/%) Porongaba (5) (ha/%) Humait (4) (ha/%) Paraguau (3) (ha/%) S. Francisco (2) (ha/%) Icuri (1) (ha/%) Forest 2003 28,777 (96.20) 8,413 (92.18) 12,480 (90.79) 15,838 (93.80) 29,264 (95.07) 63,368 (96.42) Deforestation 1999-2003 289 (0.97) 167 (1.83) 248 (1.80) 132 (0.78) 196 (0.64) 646 (0.98) Deforestation 1996-1999 171 (0.57) 63 (0.69) 147 (1.07) 113 (0.67) 438 (1.42) 443 (0.67) Deforestation 1992-1996 233 (0.78) 188 (2.06) 282 (2.05) 162 (0.96) 318 (1.03) 397 (0.60) Deforestation 1986-1992 201 (0.67) 188 (2.06) 494 (3.59) 506 (3.00) 477 (1.55) 724 (1.10) Area deforested (1986) 242 (0.81) 107 (1.18) 94 (0.69) 134 (0.79) 88 (0.29) 144 (0.22) Total Deforested 1,136 (3.80) 714 (7.84) 1,265 (9.20) 1,047 (6.20) 1,517 (4.93) 2,354 (3.58) Transition matrices following secondary su ccession between the 1986-2003 period further reveal variations in land-use trajectories and sc hemas between households within each of the six communities (Table 4-4). Porongaba and Filipina s had among the highest total percentage of area in secondary succession by 2003, with 4.65 % and 4.62 % resp ectively. Seringal Humait, with the highest percentage of total area defore sted, also had one of the higher percentages in secondary succession by 2003 (3.59 %). In general, the seringais located in the municipality of Brasilia have considerably higher rates of seco ndary succession than their counterparts in Assis Brasil (Paraguau 1.22 %, So Francisco 1.87 %, a nd Icuri 1.47 %). Filipinas, with the second lowest amount of deforestation, also had the se cond highest rate in s econdary succession (4.62 %), while So Francisco had both a low rate of de forestation and a much lower rate of secondary succession (1.87 %), suggesting different land-use decisions and intensif ication among all rubber estates.

PAGE 131

131 Table 4-4. Percent secondary succession in six s eringais from 1986-2003 Filipinas (6) (ha / %) Porongaba (5) (ha / %) Humait (4) (ha / %) Paraguau (3) (ha / %) S. Francisco (2) (ha / %) Icuri (1) (ha / %) Secondary Succession 1999-2003 141 (1.03) 66 (0.72) 134 (0.97) 111 (0.36) 183 (0.59) 362 (0.55) Secondary Succession 1996-1999 195 (1.42) 234 (2.57) 214 (1.56) 41 (0.13) 80 (0.26) 72 (0.11) Secondary Succession 1992-1996 89 (0.65) 45 (0.49) 67 (0.49) 46 (0.15) 64 (0.21) 178 (0.27) Secondary Succession 1986-1992 58 (0.42) 27 (0.30) 34 (0.24) 41 (0.13) 66 (0.21) 215 (0.32) Area in secondary succession (1986) 152 (1.10) 52 (0.57) 46 (0.33) 139 (0.45) 184 (0.60) 145 (0.22) Total Secondary 635 (4.62) 425 (4.65) 495 (3.59) 379 (1.22) 577 (1.87) 972 (1.47) Lastly, a final transition matrix was made fo r each of the six communities to capture the swidden/fallow cycle used among small-scal e agriculturalists in Amazonia (Moran 2000, Ludewigs 2006). The final matrix quantified area s that shifted from deforested to secondary succession and remained in secondary successi on throughout the transitions between the 19861992, 1992-1996, 1996-1999, and 1999-2003 images. The result is an estimate of the percent of areas in secondary succession that remain in secondary succession by 200 3 versus the percent that are re-used and deforested again throughout the 18 year transi tion (see Table 4-5). Table 4-5. Area and percent remaining in secondary succession and later reused (swidden/fallow cycle) between 1986-2003 Filipinas (6) Porongaba (5) Humait (4) Paraguau (3) S. Francisco (2) Icuri (1) Total area secondary succession b/t 19862003 (ha) 635.45 428.07 495.23 378.81 576.63 972.00 Area remaining in secondary succession by 2003 (ha) 269.92 175.58 214.55 57.15 109.27 195.93 Percent remaining in succession by 2003 42.48 41.02 43.32 15.09 18.78 20.16 Percent reusedswidden/fallow cycle by 2003 57.52 58.98 56.69 84.91 81.12 79.84

PAGE 132

132 More than half of the area in secondary succession has been re-used within each of the seringais by 2003. These numbers suggest a high amount of intensifica tion among rubber tapper households within each seringal. The rubber estates in the municipality of Assis Brasil, on average, have lower rates of deforestation, along with lower rates of long-term secondary succession, as well as a higher percentage of ar eas appearing to capture the swidden/fallow cycle. As a result, Paraguau, So Francisco, and Icuri appear to be areas undergoing less extensification, with greater preservation of prim ary forests. The seringais in Brasilia, on the other hand, can be classified as areas with genera lly higher levels of deforestation along with higher rates of secondary succession, apparently emphasizing extensificat ion and other options over more traditional swidden techniques. In th e next section, we use the household data to explain the driving forces of su ch land change currently underwa y in the CMER, as well as to explore its relationship to resident well-being. Household Survey The results of our household surveys show a wi de range of both subsistence and m arket activities among rubber tappers with in the various rubber estates su rveyed. The ma in extractive market activities are the colle ction of Brazil-nuts and con tinued rubber tapping for latex production (despite the overall decline of rubber production in the region). Increasingly, smallscale agriculture (beans, rice, manioc, coffee and corn), anim al husbandry (mainly chickens, swine, ducks, sheep and goats), and cattle ranc hing are seen by rubber tapper households as necessary income-generating alte rnatives, given the market in stabilities associated with traditional NTFPs. As a result, households are beco ming more involved in some sort of market activity beyond traditional extractivism (Table 4-6). Multivariate Analysis of the Variance (ANOVA) reveals ma rked between group difference regarding the percent of households involved in extractivism and agri culture per rubber estate (p

PAGE 133

133 0.05) (Table 4-6). The most dramatic differe nce is regarding extrac tive activities. All households surveyed within Brasilia are involved in some sort of extractivism. The households within the municipality of Assis Brasil are involved to a much lesser extent. Conversely, the majority of households in Assis Brasil are involved in market ag riculture, compared to less than half of those surveyed in Br asilia. There are no between group differences regarding the number of households involved in small animal production and cattle ranching for market sale, although the extent a nd intensity varies. Table 4-6. Percent of households involved in market activitis in 2003 (n=households) Brasilia Assis Brasil Filipinas (6) Porongaba (5) Humait (4) Paraguau (3) S. Francisco (2) Icuri (1) Extractivism* 100.00 (26) 100.00 (17) 100.00 (19) 65.00 (13) 47.37 (9) 10.00 (3) Agriculture* 38.46 (10) 58.88 (10) 47.37 (9) 90.00 (18) 73.68 (14) 72.41 (21) Small Animal Production 84.62 (22) 70.59 (12) 73.68 (14) 85.00 (17) 57.89 (11) 68.97 (20) Cattle 53.85 (14) 41.18 (7) 68.42 (13) 55.00 (11) 52.63 (10) 79.31 (23) *Statistically significant at 0.05 level, ** 0.10 level (ANOVA test of significance) A closer look at the means and standard deviations between seringais reveal marked differences and emphases in land-use and land-c over strategies among the households surveyed (Table 4-7). ANOVAs indicate statistically significant betwee n group differences regarding extractive activities between rubber estates (p 0.05). There are important differences in Brazilnut production with the households in Brasili a producing upwards of 10 times the amount of Brazil-nuts produced in the seringais found in Assis Brasil. Brazil-nut prices are consistent and regulated by the larger market, so the distin ction in question is not the result of market differences between the two municipalities. Rather, it may be the result of ecological differences

PAGE 134

134 in terms of forest types between rubber estates. The dense clos ed canopy forest in the region of Brasilia is richer in Br azil-nut trees than are the forests of the other seringais Likewise, the number of rubber trails on each property and the amount of rubber production are significantly differe nt between rubber estates (p 0.05). Porongaba and Icuri (23.59 and 33.93 kg/household, respectively) produce the least la tex, while Filipinas and So Francisco produce the most (223.85 and 258.95 kg/household, respectively), thus being the most traditional of the six seringais in our study. Note, however, th e figures for the two high-latex seringais are well below the 1995 reserve-wide average reported as 714 kg of rubber produced per household (Feitosa 1995: 72). Although the percent of total income from extractivism remains significantly high, at least in Brasilia, the percent of the tota l household income derived from rubber has plummeted from the 44.7 % report ed by Feitosa (1995: 71 ) to 8.55 % in the six rubber estates reported here, maki ng it one of the least important economic activities within our six seringais. Not only does rubber production remain highest in our traditional seringais they remain fairly dependent on it, deriving 11.44 percent of the total househol d income from the activity in Filipinas and 20.80 % in So Francisco. In general, these differences can be attributed to various responses by rubber tappers to the last r ubber bust, where some households responded by diversifying their production base, creating new economic opportuni ties, while others, due to either a strong sense of traditi on or insufficient resources or other opportunities, continue to invest in rubber production. In one instance, however, low levels of r ubber production may be attributed to local species abundances and economic inertia. Seringal Icuri occupies portions of two watersheds

PAGE 135

135 Table 4-7. Summary statistics from household surveys: land-use and livelihood statistics (Mean and S. D.) Brasilia Assis Brasil Porongaba (n=17) Filipinas (n=26) Humait (n=19) Paraguau (n=20) S. Francisco (n=19) Icuri (n=29) Land Holding Rubber Trails on Property (n)* 4.29 (2.71) 6.08 (3.32) 4.32 (1.83) 5.60 (3.27) 4.16 (2.27) 1.69 (2.80) Brazil-nut Trees on Property (n)* 213.00 (245.82) 319.42 (345.31) 358.75 (315.19) 92.25 (91.96) 36.95 (113.94) 2.24 (5.50) Pasture (ha)** 9.44 (4.13) 7.00 (6.38) 12.03 (12.47) 7.70 (7.95) 7.29 (5.37) 11.72 (7.22) Swidden Agriculture (ha)* 2.94 (2.34) 2.16 (1.49) 1.61 (0.76) 2.25 (1.01) 2.34 (1.83) 3.16 (1.60) Extractivism Brazil Nut (latas)* a 318.71 (224.00) 367.54 (375.97) 254.63 (166.92) 22.80 (46.84) 0.00 (0.00) 0.14 (0.52) Rubber (kg)* 23.59 (75.23) 223.85 (256.55) 125.26 (176.40) 174.00 (349.88) 258.95 (432.14) 33.93 (124.76) Agricultural Production Rice (kg) 1030.59 (617.62) 1156.15 (834.40) 738.95 (534.64) 916.00 (647.80) 934.21 (481.33) 916.07 (652.64) Beans (kg)* 382.94 (344.72) 268.85 (599.58) 228.95 (213.77) 750.00 (580.06) 457.89 (391.67) 395.71 (284.01) Corn (kg)* 537.94 (546.33) 666.92 (745.57) 557.89 (393.09) 995.00 (945.61) 1121.05 (1258.56) 1239.29 (1010.44) Manioc flour (kg) 574.12 (532.81) 599.42 (735.51) 273.63 (418.85) 628.00 (795.18) 331.58 (381.59) 253.93 (195.57) Caf (kg)** 20.88 (52.63) 1.00 (4.05) 37.11 (126.00) 173.80 (386.70) 76.00 (320.70) 25.21 (67.92) Animal Production (n) Chicken 87.35 (136.29) 38.65 (22.29) 39.68 (31.72) 53.10 (27.15) 61.58 (60.09) 63.43 ( 32.33) Swine 7.58 (11.04) 9.44 (17.03) 6.16 (9.25) 9.95 (8.25) 12.89 (15.96) 8.64 (9.06) Ducks 0.94 (2.14) 3.42 (5.81) 1.05 (2.86) 1.10 (2.55) 2.95 (3.92) 3.54 (6.00) Sheep/Goats* 3.00 (4.87) 3.41 (5.23) 7.16 (12.81) 5.55 (9.72) 2.63 (5.59) 9.35 (10.00) Cattle** 14.18 (13.39) 8.19 (8.31) 19.47 (29.78) 16.55 (20.17) 7.74 (7.36) 17.45 (13.31) Dependency on Activity (% of total income) Extractivism (%)* % of total rubber 63.06 (29.10) 0.69 74.31 (19.55) 11.44 62.84 (24.31) 5.68 22.95 (27.96) 10.41 20.80 (29.09) 20.80 2.95 (11.89) 2.95 Agriculture (%)* 13.60 (17.59) 8.11 (16.00) 7.60 (14.43) 30.64 (30.00) 27.69 (24.23) 35.15 (30.56) Small Animal Production (%)* 6.61 (6.89) 9.64 (10.02) 11.26 (14.68) 24.76 (26.13) 11.92 (18.27) 9.38 (12.77) Cattle (%)* 16.73 (25.70) 7.90 ( 8.52) 18.46 (22.32) 21.60 (24.29) 18.54 (24.17) 45.30 (31.42) Livelihood Index*b 10.71 (3.29) 9.04 (2.63) 11.42 (3.83) 7.90 (3.16) 9.63 (4.57) 11.71 (4.22) Statistically significant at *.05 level, **0.10 level (ANOVA test of between gr oup diff erence) a) A lata, or can, is the traditional system of measuring Brazil Nuts in their brute form for sale (i.e., with shells on). 1 lata is equivalent to a 20 liter bucket, and norma lly yields about 11 kgs of Brazil Nuts. b) Based on 30 individual indicators (e.g., overall house constr uction, appliances, agricultural equipment, access to an electr ic generator, water pump, safe drin king water, trash and sewage disposal, accessibl e healthcare, mode of transportation, etc.). and that located closest to the main community is not rich in rubber trees Historically, this community served as a hub for the sale of region al latex and, given the proximity to the Iaco River, the transport of latex to urban centers. Many of the residents in this seringal did not tap

PAGE 136

136 rubber but instead worked in the counting houses, supply good st ores, and with transportation logistics regarding the rubber tr ade. Since both the paving of the BR-317 and changing market conditions, traditional water trans port routes for rubber have been either eliminated or have changed, leaving the residents of Ic uri with less of a role to play in the rubber trade, and fewer livelihood options, mainly agriculture and cattle ranching. While the local and regional markets have a ffected rubber tappers land-use choices, the density of Brazil-nut and rubber tr ees appear to do so as well. Pearsons correlation coefficients (Table 4-8) reveal a significant positive correla tion between the percent of total income derived from extractivism and the numbe r of Brazil-nut trees in the colocao (r = 0.467, p-value 0.01). Similarly, there is a signifi cant positive correlation between the percent of total income derived from extractivism and the number of rubber trails controlled (r = 0.454, p-value 0.01). Similarly, there is a significant positive correlation between the percent of total income derived from extractivism and the number of rubber trails controlled (r = 0.454, p-value 0.01). In contrast, there is a signifi cant negative correlation between both the number of Brazil-nut trees and rubber trails controlled by a household and th e percent of total income derived from both cultivation and cattle, suggesting that land-users may be choosi ng livelihood stra tegies based on ecological constraints due to differences in NTFP potential among forest types. There are marked differences in agricultural practices among rubber esta tes that appear to be more subtle than other activities, with different seringais emphasizing different products. Outside of manioc flour production, the three rubber estates in Assi s Brasil produce greater amounts of beans, corn, and coffee (p-value 0.10), and other than Porongaba, have among the highest amount of land under agricu ltural production. Furthermore, the seringais in Assis Brasil

PAGE 137

137 Table 4-8. Livelihood index, deforestati on, and land-use correlations (n=130) Livelihood Index % Income Agriculture % Income Small Animals % Income Cattle % Income NTFPs Total Deforestation (ha) Area Pasture (ha) Agricultural Plot (ha) Brazil-nut Trees (n) Rubber Trails (n) Livelihood Index 1 -0.047 0.073 0.348* -0.183** 0.519* 0.491* 0.227* -0.045 -0.216** % Income Agriculture -0.047 1 -0.079 -0.030 -0.548* -0.023 -0.060 0.101 -0.252* -0.225** % Income Small Animals 0.073 -0.079 1 -0.170 -0.178** -0.124 -0.149 -0.006 -0.107 -0.003 % Income Cattle 0.348* -0.030 -0.170 1 -0.527* 0.384* 0.370* 0.192* -0.202** -0.255* % Income NTFPs -0.183** -0.548* -0.178** -0.527* 1 -0.152*** -0.120 -0.218* 0.467** 0.454* Total Deforestation (ha) 0.519* -0.023 -0.124 0.384* -0.152 1 0.981* 0.329* 0.080 0.100 Area in Pasture (ha) 0.491* -0.060 -0.149 0.370* -0.120 0.981* 1 0.235* 0.111 0.120 Agricultural Plot (ha) 0.227* 0.101 -0.006 0.192** -0.218** 0.329* 0.235* 1 -0.184** -0.068 Total Brazil-nut Trees (n) -0.045 -0.252* -0.107 -0.202* 0.467* 0.080 0.111 -0.184** 1 0.394* Rubber Trails (n) -0.216** -0.225** -0.003 -0.255* 0.454* 0.100 0.120 -0.068 0.394* 1 *Correlation is significant at the 0. 01 level, **0.05 level, ***0.10 level

PAGE 138

138 derive at least twice as much of their income from market agriculture than in Brasilia. A Pearsons correlation coefficient shows a slight positive relationship between the size of an individual households ag ricultural plot and pastur e size (r = 0.235, p-value 0.01). Rules that allow a landholder to deforest tw o hectares a year (1 hectare ma ture, 1 hectare secondary forest) for their agricultural plots may be inadvertently encouraging pastur e formation, as it is easier to transform an old agricultural plot into pasture, than to go from mature forest to pasture. In a traditional swidden system, each y ear a household deforests a small area for agriculture, which, in the pa st would be left to fallow for an average of three to four years. Increasingly that fallow is being replaced by pasture and new forest is cut for cultivation. In fact, 56.6 % of the households surveyed leave an aba ndoned agricultural plot to fallow, 13.1 % turn that plot immediately to pasture, 19.7 % plant Pueraria phaseoloides, or tropical kudzu, and 10.6 % follow some combination of these options. Tropical kudzu, a legume (nitrogen fixing) and weed suppressant species, is often planted for forage for cattle and is used in pasture development. The majority of households survey ed in the rubber estates in Assis Brasil leave their abandoned agricultural plots for fallow. Th e majority of households surveyed in Brasilia, on the other hand, plant either tr opical kudzu or pasture following agriculture, suggesting various land management strategies be tween the two municipalities. Pasture and agricultural deve lopment have a positive relation ship with deforestation, while extractivism a negative one. Interestingly, there ap pears to be no correlation between either the total area deforested or the total area in pasture, and the number of Brazil-nut trees and rubber trails on household property (see Table 4-8). These results lead us to be lieve that the in fluence on land-use decisions of changing markets an d ecological conditions notwithstanding, pushing land-holders in one direction or another, decisions regardin g the number of cattle to poss ess and the amount of land

PAGE 139

139 to deforest may be based on other social, institutional, or political factors, such as access to capital funds, rubber tapper cu lture, or participatio n in the movement. It is impo rtant to note th at cattle are seen by many rubber tapper s as a savings account, an asset to be used in times of distress (Feitosa 1995). This follows a simi lar rationale amongst coloni st smallholders elsewh ere in Acre (Ludewigs 2006) and can be fo und throughout Amazonia (Faminow 1998, Mertens et al. 2002). For other households, cattle ar e becoming a major ma rket activity (Gomes 2001). Regardless, the investment in cattle pays off as there is a significant positive correlation between the percent of total household income derived from cattle and livelihood welfare (r = 0.384, p-value 0.01), while there is a negative correlation between the percent of total in come derived from extractivism and quality of life (livelihood index) (r = -0.183, p-value 0.05). Agriculture and small animal production appears to have no significant relationship to livelihood, but a signifi cant positive relationship exists between deforestation and a hous eholds livelihood index (r = 0.519 p-value 0.01). Overall the seringal -level household resu lts suggest that Sering al Filipinas and So Francisco are the most traditional rubber esta tes with the highest latex production and lowest amounts of both cattle (8.19 and 7. 74) and pasture (7.00 and 7.29 ha) per househol d. Households in Humait have the greatest amounts of both past ure (12.03 ha) and cattle (1 9.47), but also remain involved to some extent in a vari ety of extractive activi ties. Icuri has the second highest reported amounts of both pasture (11.72 ha) and cattle (17.45). While Humait and Icur i appear to be the least traditional of the rubber estates, the communities of Porongaba and Paraguau reside somewhere in between these two nod es in terms of both pasture (9.44 ha and 7.70 ha) and cattle (14.18 and 16.55 ). With the exce ption of Icuri, these results are consistent in explaining the deforestation trends outli ned previously in the remote sens ing analysis.

PAGE 140

140 Discussion The results suggest that the label rubber tappers for the oc cupants of the CMER m ay be a misnomer. In fact, among the seringais surveyed here, households pursue diverse livelihood practices including extractivism, small-scale market cultivati on, small animal rearing, and increasingly, cattle production. Rubber tapping plays increasingl y less an important role in livelihood strategies. Still, the majority of households surveyed identify themselves, for historical, cultural, and political reasons, as ru bber tappers first and foremost, rather than colonists, small-scale ranchers, or agriculturalists. Indeed, so me evidence suggests that rubber tappers involvement in agriculture and small-scale cattle rearing is not just a recent development, but always existed, although to a much lesser extent, since the rubber barons abandoned their rubber estates in the 1960s and 1970s (Murrieta and Rueda 1995). The main change is not the adoption of such activities; it is the extent of their use and relative proportion of household income. Rubber tappers are no doubt being influenced by development forces operating outside the CMER. In Acre, rubber has descended from the most important economic activity to one of the least, while cattle and cultivation have become increasingly important. Thus cattle have a ready market outsi de of the reserve, while many rubber and extractive products simply do not. Understanding the land dynamics in the reserve re quires the expansive interpretation of the rubber-tapper label that is held by the occupa nts and central to their land-use decisions. According to the remote sensing analysis, none of the six seringais surveyed here have surpassed the 10 % limit set on the amount of allowable defore station by 2003, although at least one trend projection indicated that this limit would be exceeded by that date. Humait and Porongaba are, however, close to reaching this limit. It is also noteworthy that some seringais

PAGE 141

141 may have passed the 5 % allowable for pasture (roughly 15 hectares per household).3 Compared to other land-users in Amazonia, deforestati on rates remain low, however. Evidence of high amounts of land in secondary succession in Bras ilia and increasing in tensification among landusers in Assis Brasil suggests th at rubber tappers may be undergoing a process of adaptation that may alleviate deforestation pressures with pr oper management. The reserve provides other important environmental services that must also be considered. The CMER currently serves as a firewall, buffering the rain forest (located betw een highly deforested cat tle ranches and largely forested indigenous communities), and offeri ng biodiversity protection (Vadjunec 2007). Pasture for cattle and crop development are th e greatest drivers of deforestation within our study areas, and appear to be nega tively correlated with extractive activities. The rubber estates with the highest amount of cattle also have th e highest amount of deforestation. The only exception to this trend was Seringal Icuri, but this may be due to inconsistencies with both population dynamics and boundary issues. Cropping pl ays more of an important role in the rubber estates of Assis Brasil wher e natural forest differences re sulting from the division of the watershed limit NTFP extractive potential. Furt hermore, cultivation is linked to deforestation because increasingly households convert their aba ndoned agricultural plots to pasture instead of leaving them to fallow. As expected, Filipinas and So Francisco, the two most traditional extractive seringais (in terms of sustained emphasis on latex producti on), had fewer cattle, maintaining among the lowest rates of deforestation. This result suggests that investme nts in non-timber forest projects might be beneficial to forests. As curre ntly practiced, however, emphasis on NTFP production 3 Due to the common property aspects of the reserve, the difficulty mapping at the household level, and the fact that many households actually create pasture together, while other households rent their pasture out, we rely on the remote sensing results to test the efficacy of the deforest ation rules, and use the hous ehold level data to explore relationships between land-use, deforestation, and well-being.

PAGE 142

142 does not appear to better livelihood conditions. Fo r extractivism to be a serious contender, these traditional seringais press for the return of long-term and realistic subsidies, supporting rubber collection as well as heavy investments in other green products such as aai fruit, honey, vegetal leather products and handicrafts. The pro-forest government of Acre has already started to invest in these areas, with the creation of a condom factory in Xapuri for example, but it remains to be seen whether it will get residents b ack to rubber tapping, let alone working in other NTFP activities. Most of the debate surroundi ng cattle in the CMER focuses on deforestation for pasture, but such arguments need to be expanded to in clude considerations of both secondary regrowth and the swidden/fallow cycle. The remote sensing analysis shows that much of the deforestation in the six seringais surveyed here is permitted to move into secondary succession, and over half of the area in secondary succession is re-used at a later time period. Thus a significant proportion of the disturbed land cover in the stud y area remains in some disturbed condition and is not permitted to reach mature stages of forest regrowth. Rubber tapper households appear to be changing their production systems. A lthough most rubber tapper households are not expanding their agricultural plots per se, they may be using the -hectare per year rule to augment their pasture size on a yearly basis Overa ll, the results suggest th at land-use and landcover change, and even the responses to such change, is dynamic in the CMER. Aside from changing market and land-use emphases and the re sulting variable rates of deforestation, there are also increasing differences in production sy stems and technologies ad opted by rubber tappers as their main source of livelihood begin to chan ge, suggesting not only ex tensification but also intensification.

PAGE 143

143 Lastly, it is obvious from this exercise that new property boundaries of some seringais are needed. Icuri is a case in point, a community in which the official digital rubber estate boundaries appear to be inconsistent with both th e communitys and local officials definition of the actual seringal with some officially recognized resi dents appearing to live outside of both the reserve itself, as well as the community In addition, various boundaries exist for a few of the seringais in the CMER that have size differences of as much as 15 %. Such variances no doubt affect the results of the deforestati on amounts and rates reported here at the seringal level. When the question of which boundary to use ar ose, we opted for the boundaries that most incorporated the households surv eyed within the specific comm unity, relying on both GPS points and field experience in the region. If remote sensing is to be considered a serious tool for monitoring the sustainability of the CMER, it is crucial that these issues be resolved. The discrepancy with Icuri between the high amount of land reported by households to be in pasture and the low levels of actual defo restation found in the remote sensing analysis points to yet another issue; ma inly, extenuating ecological circ umstances caused by the division of the watershed, making much of the seringal uninhabitable. Unlike the five other rubber estates in this study, where all land is claimed by some hous ehold or another, much of Icuri is without water (or Brazil-nut trees), and consequently without households. These vast tracks of barren land create an unfair advantage for the residents of Icuri when monitoring the 10 % rule at the community level. Exceptions like Icuri reveal the limits to a remote sensing approach when monitoring the adherence of house holds to the rules outlined in the PU, and show why it is important to monitor the reserves efficacy on the ground. Given these new boundary issues in an historically common property system where prop erty is defined by access to trees and trails, rather than geometric space per se, we argue that in order for monitoring to be effective with the

PAGE 144

144 adoption of new land-use practices, authorities really need to inve st in mapping properties at the household ( colocao ) level. This will ensure that eff ective and timely monitoring can, indeed, take place. Conclusions The search f or a win-win scenario in the CMER and other Extractive Reserves in Brazil appears to be far more complicated than th e simplified notion that rubber tapping and other NTFP activities can provide improved material well being while preserving tropical forest. Furthermore, the majority of rubber tappers themse lves, at least in this study and most likely for the CMER, define their rubber tapper label as including a wide-range of activities that require the forest to be cut, perhaps increasingly so. Overall, when compared to outside colonization projects and ranching operati ons, the deforestation occurri ng within the CMER can be considered quite low, but when looke d at as a conservation unit, the six seringais examined are undergoing surprisingly rapid land-us e and land-cover changes with ca ttle and agriculture as the main drivers of such change. Such changes ar e inextricably linked to current outside market influences, which privilege cattle to NTFP extractivism. As of yet, however, none of these six seringais have surpassed the 10 % limit set on the amount of allowable deforestation. While some communities may surpass allowable deforest ation limits in the near future, evidence of intensification among land-users suggests that rubber tappers may be undergoing a process of adaptation that alleviates deforestation pressure s while providing improved incomes. After all Holt (2005: 209) argues: conservati on develops as a result of e xperience and learning, sparked by negative changes in resource characteristics. The enviro nmental changes now underway may be capturing this learning and adjustment process, rather than a long-term trend suggesting system unsustainability.

PAGE 145

145 Much of the current criticism questioning the sustai nability of the CMER and the entire ER model, may reflect biases or misc onceptions regarding the character of rubber tapping livelihoods as well as the pristineness of Amazonian forests. The murder of Chico Mendes ha s been enlarged to equate cattle with ev il livelihoods. As a resu lt of such caricaturizing, the mere idea that the rubber tappers champi oned by Mendes are now raisi ng cattle themselves is me t with little enthusiasm by some proponents of the extractive reserve c oncept. It must be reme mbered that Chico Mendes and the rubber tappers ma y have been fighting for the forest, but also, and more importantly, for a right to both a better livelihood a nd land security. Cattle and market agricu lture are and will likely remain part of a rubber tappers land-use system, and consequently need to be addresse d realistically in terms of appropriate policy development for the rubber tapper populations in th e region. Rubber tappers want technical support regarding agricultural de velopment, particularly pasture formation and its sustained use. Cattle, however, remains a dirty word in the CMER, and r ubber tappers continue to find little support, information, or new policy and re gulations regarding such matters. The result s of this study show that cattle are positively linked to both livelihood welfare and deforestation, while dependence on extractivism (at least as cu rrently practiced) is linked to a decrease in welfare, as well as a decrease in deforestation. Our suggestion, therefore, is twofold. Policy directions fo r both the CMER and the ER model need to focus on cattle regulation and sustainable past ure management, as well as a serious re-investment in NTFP marketing and development in order to ma ke extractivism more viable.

PAGE 146

146 CHAPTER 5 WHAT MAKES A RUBBER TAPPER IN THE BRAZILIAN AM AZON?: CULTURAL, LIVELIHOODS AND INSTITUTIONAL-OR GANIZATIONAL FACTORS UNDERLYING SELF-DEFINITION Introduction The Amazon Rubber Boom, which spanned th e late nineteenth and early twentieth centuries, introduced a new social actor to Amazonia the rubber tapper a forest extractivist at the bottom of a mercantile networ k linking local, regional and globa l economies. Over the course of a century, rubber tapper iden tities originated a nd centered on their cu ltural heritage as impoverished migrant peasants with roots in th e semi-arid Brazilian northeast (Santos 1980), laborers coerced by rubber estate owners in Amazonia under the dept peonage systems (Cunha 1946, Weinstein 1983, Bunker 1985), and more recently evolved to emphasize successful natural resource managers living at peace with the forest (Anderson 1990, 1992 and 1994, Fearnside 1992, Nepstad and Schwartzman 1992, Allegretti 1995) and as articulato rs of a victorious social and environmental justice movement in defense of the forest that have di rectly influenced policy in the Amazon through the Extractive Rese rve system (Allegretti 1989, 1990 and 1994, Schwartzman 1989, Fearnside 1989). With a recent history of notable accomplis hments, rubber tappers perceptions of livelihood options have become broader and are st arting to show contrasting trajectories (Gomes 2001, Vadjunec 2007). Official recognition of rubber tapper land claims led many observers to expect forest extractivism to continue, indeed to flourish. Forest ex tractivism, however, has increasingly faced economic difficulties that have prompted changes in livelihood strategies, even to the point of the growi ng adoption of cattle ranching, an activity long associated with rancher adversaries of r ubber tappers (Gomes 2001 and 2005, Ehringhaus 2006, Vadjunec 2007, Toni 2007).

PAGE 147

147 Since their historic achievements in the 1980s, there has been lit tle attention to the cultural definition rubber tappe rs (Wallace 2004, Vadjunec 2008). Given rapid socio-economic, institutional and political transformations in the Amazon, it is important to examine rubber tappers individuals self-d efinition, after almost three decades of their historical struggle. Given the circumstances in which they are now situated, we explore the idea of what it means to be a rubber tapper in Amazonia today. To answer this question, it is important to understand the evolution of rubber tapper live lihood trajectories to comprehend how rubber tapper definition was historically constructed and how indi viduals currently define themselves. Rubber Tappers Historical Trajectori es and Definition Construction During the rubber boom the definition of a rubber tapper was straightforward: a migrant worker tapping rubber in the Amazon. The collap se of the Amazonian rubber industry prompted by the appearance of plantation-grown rubber in Southeast Asia (Weinstein 1983), however, forever altered this definition. The Amazonian rubber economy was briefly revitalized by the Brazilian and United States governments during the Second World War when Southeast Asian rubber supplies areas were controlled by the Ja panese, blocking the U.S and its allies access to rubber (Martinello 1988). Although the state assumed economic and social control of the rubber industry during this period, long-ti me rubber tappers and new migran ts continued to be subjected to a highly exploitative relationship under the same historic debt peonage systems (called aviamento ), despite the fact that a grassroots reaction to these systems became increasingly more evident (Bakx 1988, Oliveira 1985, Ma rtinello 1988). After the s econd short-lived rubber boom, the Amazonian rubber economy once again collaps ed, causing many rubber barons to abandon their rubber estates, leav ing the rubber tappers to fend themselves. Despite the failure of rubber to establish long-term economic development of the region (Coomes and Barnham 1994), the

PAGE 148

148 rubber tapping industry indelibly impacted its culture (Tocantins 1979, Santos 1980, Bakx 1988, Rancy 1992). In the state of Acre in the southwestern Brazilian Amazon, rubber tapping was once at the core of the state economy, bringing thousands of people to the re gion and shaping local culture. After the rubber economy crash, while many rubber tappers in the state fled to urban centers, others stayed in the abandoned rubber estates supplying local and region al markets with latex (otherwise banned for importing into Brazil). The establishment of strong ties with the forest along with a new found freedom from the aviamento system, led to livelihood diversification among rubber tapper communities. Other forest products such as Brazil nuts became more important, and rubber tappers began to engage in small-scale subsis tence agriculture (e.g., Campbell 1996). Such extractive activities, however, only held a peripheral importance in regional politics extractivism was generally viewed as a back ward, economically-stagnant, and peripheral activity, and forest dw ellers were considered as obstacles to development in Amazonia (Barbosa 2000). After the final rubber bus t, the largely forgotten rubber tappers were virtually invisible on the na tional scene (Almeida 2004). In the 1960s, Brazils then-military government prioritized regional development policies aimed at opening up the Amazon frontier as a way to integrate the region w ith the rest of the country, including major infrastructure programs, colonization projects, and stimulation of economic growth via fiscal incentives for capit al investment (Mahar 1978, Hecht and Cockburn 1989, Moran 1981, Oliveira 1991, Schmink and W ood 1992). By the 1970s, migration by farm families into colonization areas proceeded alongside the establishment of large ranches by investors. As a result, ranching ac tivities began to be consolidate d, while forest extractivism was forgotten by the national government.

PAGE 149

149 In Acre, where extractivism was still the basis of the state economy, the 1970s were marked by a major federal campaign to attract inve stors from the south of Brazil. The isolated lands targeted for settlement were historic rubber estates, which at the time were still occupied by rubber tappers unable to prove land rights under Brazilian law. Nonetheless, their land was rapidly transferred by rubber estate owners to Paulistas ranchers originating from So Paulo and other states in southern Brazil. Widespread land speculation occurr ed rapidly, and in few years around of one third of the land in Acre was sold to Paulistas and former rubber estates were cleared by cattle ranchers (Bakx 1988, Silva 1990, Calaa 1993). The arrival of roads and investors alongside with a rapid shift in landownership, led to violent conflicts along the development frontie r (Nunes 1991, Sobrinho 1992, Silva, 2003, Paula, 2005, Paula and Silva 2006). The clash of legacy of the rubber boom, with its history of a strong forest-based culture, and heavily influenced by the Rural Workers Union and the Catholic Church, rubber tappers mobilized as a social movement in Acre in the late 1970s. Chico Mendes was a major force behind the movement. Throug h non-violent collective action, Mendes and the rubber tappers fought for recogniti on of their historic land claims in light of abandonment of rubber estates by their owners. The rubber tapp ers movement gradua lly evolved, receiving increasing political clout for thei r agrarian reform via recognition of their traditional land claims. During the rubber tappers fight against cattl e ranchers, the traditional notion of rubber tapper definition was actively portrayed and vocalized. Grassroots mobilization in defense of forests and their inhabitants won admirers among international environmental organizations who emphasized the ecological benefits of forest-based extractivism, propelling rubber tappers onto the world stage through a discourse that linked human rights abuses and environmental destruction. Rubber tapper s articulated an enviro nmental image of extractivism to support their

PAGE 150

150 agrarian reform agenda. The success of the rubber tappers movement in protecting their land and livelihoods in addition to their evolving strategy of political negotiation and agendabuilding, drastically changed the conceptualization that rubber ta ppers might have previously elaborated, requiring a more complex definition. As a result, being a rubber tapper came to be associated with social movement for agrarian reform to recognize traditio nal land rights based on long-term use of forest for sustainable livelihood s practices. It was under these circumstances, as has been well-documented (see Revkin 2004, Hecht and Cockburn 1990), that the traditional definition of rubber tapper evolved to be associated with greenness, emphasizing the environmental dimension of forest extractivism. Growing recognition of the social and ecologi cal consequences of deforestation and the murder of Chico Mendes in a year of record fo rest burning in the Amazo n prompted the creation of federal extractive re serves system (Allegret ti 1990; Schwartzman 1992) Extractive reserves are forest areas inhabited by extr active populations granted long-term usufruct rights to resources which they collectively manage (A llegretti 1989). The creation of th e first extractive reserves in early 1990s marked an unprecedented success of e nvironmental policy-making in the Brazilian Amazon (Allegretti 1989; Schwartzman 1989), and has been promoted as a focal policy tool for promoting forest conservation while simultaneously increasing the economic value of forests and rural income. Despite this potential and promotion as a model for extractivism development in Amazonia, historically only two non-timber forest products (NTFPs), rubber and brazil nut, have been extensively managed. Ironically, at the onset of ER implementation in the 1990s, federal rubber subside for rubber was cut, and the alr eady low rubber prices plummeted. Almost two decades after the establishment of the federal Extractive Reserves in Amazonia, however, the

PAGE 151

151 model still faces challenges of putting these obje ctives into practice as economic, social, and ecological constraints have emerged (e.g. Anderson 1992, Browder 1990 and 1992, Homma 1992, Hall 1997). Extractivist communities are still f aced with a current pattern of low-income generation based solely on forest activities, and the goal of economic growth is a major driver forcing livelihood change in Extractive Reserves. As the pressure to diversify production and to increase income grows, rubber tappers are increasingly moving to ward cash-crop agriculture, and small-scale cattle ranching as important co mponents of their liveli hoods strategies to maintain and/or increase household income (e.g. Gomes 2001, Wallace 2004, Ehringhaus 2006, Vadjunec 2007). Nevertheless, despite the widesp read decline of rubber tapping, the category of rubber tapper ( seringueiro ) has become and continues to be culturally and politically powerful within the social movement and regional politics. In 1998, allies of the rubber tappers movement came to power in the state of Acre and launched new experiments in regional developmen t policies (Kainer et al 2003). Jorge Viana, an environmentalist, an academically trained forest er, and a close associate of Chico Mendes, was elected governor for the state in 1998 and re-elected in 2002. In the 2006 election, Arnbio Marques (Binho), Vianas vice-g overnor, was elected governor and continues the development policies initiated by his predecessor. The Forest Government as it is locally called, operates under broadly defined concepts of forest development. The driving principal of diversified extractivist development emphasizes the social asp ect of local and regional development and is evidenced in the governments latest forest cons ervation and development concept Florestania, or forest citizenship. Florestania is built on respect for the environment and recognition and appreciation of local knowledge and cultural traditions of forest dwellers, honoring the legacy of Chico Mendes and the rubber tappers movement (Government of Acre 2005).

PAGE 152

152 The Forest Government has gained growing support in Brazil and abroad and has been successful in rallying financial support for its ground-breaking development approach. This legacy and the platform the government gave to alternative development ideas helped them to advance major programs funded by the Brazilia n national government, the National Bank for Social and Economic Development (BNDES), and multilateral cooperation agencies, such as the Inter-American Development Bank (IDB). Forest dwellers are encouraged to increase pr oduction of the twin pillars of the traditional Acrean extractivist economy: rubber tapping and Brazil nut harvesting. An important fiscal incentive was the creation of the Chico Mende s Law in 1999 a rubber subsidies program to encourage a return to rubber tapping (Kainer et al. 2003). To offer value-added income for rubber tappers families, the state government has cr eated a series of incentives for forest-based entrepreneurs, which include access to green markets and processing of forest-based production. Xapuri, where the rubber tapper movement arose, has become an industrial complex for diverse forest-based development initia tives, which include community-b ased logging, an eco-friendly furniture industry, a Brazil nut processing plant, and a condom factory that uses local latex to supply the Brazilian Health Ministrys program w ith distribution of condoms in medical centers throughout the country. In this context, rubber tapper defini tion has become strongly embedded in Acreano State history and pride, becoming part of a larger, more complex re gional concept over time. Yet, their traditional definition and knowledge remain alive, complex, and extremely dynamic. Little research has been done in Amazonia regarding pe asant self-definition. Nugent (1993) argues that traditional Amazonian peasantry is historical ly wrongly defined by the other and remains relatively invisible, yet quite complex. Harris (2007: 7) concludes that Amazonian river

PAGE 153

153 dwellers define themselves based on shared skil ls and experiences, adapted around survival to changing market and environmen tal conditions. Becker (2004: 9) criticizes the seemingly uniform category of Latin American peasantry, and instead argues that many social groups have hybrid identities based on a complex comb ination of their own culture, the prevailing dominant culture, land struggles, and political and economic forces. Here, we explore this complex definition of peasant as it refers to rubber tappers. While local discourse sometime s points in this direction, th ere is no golden age of rubber tapping to return to and regardless of these many changes in support of forest extractivism, cattle ranching is continuous expanding in Acre (Valentim et al. 2002; T oni 2007). What is more, this land use activity is not confined to larger-scale landowners but is growing among other groups and erstwhile rubber tappers (Gomes 2001). A lthough cattle ranching runs counter to the ecologically sustainable precepts of the extractive reserve concep t, rubber tapper communities in the Chico Mendes Extractive Reserve (CMER), the symbolic site of th e early struggles are increasingly pursuing small-scale cattle ranching as a sound income-generating activity. This trend toward cattle ranching occurred despite pr ofound historical and cultu ral differences with rubber tappers livelihoods trajectories, but it respond to regi onal market demand operating in the region, and dominate economy outside extractiv e reserve areas. We therefore argue that traditional definition of rubber tapper need to be revisited, and a more dynamic view that recognizes the wider range of economic possibiliti es within extractive re serves discussed. The image of forest guardian with in herently environmental practices is dangerous in the context of changing pressures and practices an d the idealized guardian can be recast as the forest villain, fueling agendas of hardcore protectionists and developers.

PAGE 154

154 Chico Mendes has himself become a legend of almost mythic propor tions, being described as an ecologist, a rain-forest Gandhi, an eco-martyr, and one of the most important environmental thinkers in history (Gal e 1998, Palmer 2002, Maxwell 2003, Revkin 2004). On the 20th anniversary of his murder (dec/2008), his memory was celebrated in many ways. In Acre, the state government officia lly dedicated 2008 as the Year of Chico Mendes; the Brazilian Congress dedicated a solemn special session to re member his name and struggle; his story was retold throughout major newspapers and television networks (e.g., BBC, Discovery Channel, Globo); and the Brazilian President, Luiz Ignacio Lula da Silva, led tributes to Chico Mendes in his weekly radio address. Diverse political coalitions ha ve coalesced around Chico Mendes legacy, developing a dialectic around the importa nce of the forest-dwellers that Mendes represented; however, in many cases the coalition member groups have each tried to set a different agenda for rubber tappers, each claiming to be a direct philosophical descendent of Chico Mendes legacy of forest conservation. For all this celebration and attention, when rubber tappers do not behave as expected, they are criticized. Ironically, while everyone was celebrating Chico Mendes memory, the Brazilian Environmental Agency (IBAMA) began an opera tion to expel illegal occupants from the extractive reserve that carries Chico Mendes name. IBAMA took the stance that those raising cattle were acting illegally by not obeying the reserves rules ag ainst increasing cattle raising activities, and 350 families were commanded to l eave the reserve. The president of the Rural Workers Union of Xapuri, the same Union Ch ico Mendes was president when assassinated, however, challenged IBAMAs action against these reserve families, arguing that small-scale cattle ranching has become an important economic activity for rubber tapp ers and the lack of

PAGE 155

155 government investment in extractive reserves is to blame for such livelihoods transformations and its environmental impacts in the Reserve. The problem described here represents an opportunity for broader debates on forest resource co-management pacts by government institutions and grassroots organizations for economic options and administration of extrac tive reserves in the very place where Chico Mendes argued for this right twen ty years ago. It is time to as k questions about how these new contradictions might shape r ubber tapper definition today. The past one hundred years or so have constituted rubber tapper as a distinct social category. During this time, rubber tapper definition has evolved. That is of course based on neocolonial comparative advantages of the region, tied to rubber, albeit under highly exploitative labor relations. That changed via the autonomy and the agrarian reform movements, which evidence attachment to the r ubber estates, which were adam antly defended by rubber tappers, which also helped constitute regional identity in Acre as opposed to the Paulistas from southern Brazil, and originated a strong envi ronmental movement in the region. Campaign all call on invocation of seringueiro in specific ways th at were not always consonant. Rubber tapper definition have ch anged over time and often been imposed by outsiders, whether critics or s upporters. And now, all of this is back in play, because the historical trappings of expl oitative labor relations has b een swept away, the economic importance of rubber is in questi on, and the quality of life in towns looks in many ways better than that in Extractive Reserve. There are curre ntly divergent pressures on people in the CMER as regards their productive activit ies, via supports for extractivi sm alongside market incentive for cattle ranching, and consequently this pulls various directions at how rubber tappers define themselves. If all these are shorn from rubber tappers as a basis for their definition, we are left

PAGE 156

156 with regionalism, which is somewhat diffuse si nce it does not just apply to the CMER but to Acre more generally. This study seeks to shed light on rubber tapp ers definition reflecting on their historical livelihood trajectories while cons idering their current evolving socioeconomic and development circumstances as residents of Extractive Reserve. This paper uses quantitative data collected from a household survey, and qualitative data collected from key-inform ant interviews from the Chico Mendes Extractive Reserve to explore the followi ng questions: (1) Does everyone self-identify as a rubber tapper today? (2) If not, which cultu ral, institutional organizational and livelihoods characteristics help explain who does?, and (3) Wh at additional qualitative evidence can help us interpret what it now means to identify as a rubber tapper? Data and Methods Selection of the Study Sites This study was carried out with rub ber tapp er communities of the Chico Mendes Extractive Reserve (CMER) in southwestern Brazilian Amaz onian state of Acre (Figure 5-1). The eastern part of Acre lies at the westernmost limit of the Amazonian development frontier that experienced massive changes in land use since the 1970s, constituting part of the complex land use mosaic. The CMER lies just at this limit of the frontier. Forest dwel ling communities in this region live at the threshold of forest based ec onomies and competing la rge-scale types of landuse in which the CMER is nested within and al ong the limits of the deforestation front and within different land tenure t ypes and thus, different land uses In many ways, the study region and its unique political ecology are particularly appropriate for the study of the changing perception of what it means to be a rubber tapper today. The CMER is the second largest extractiv e reserve in Amazonia with an area of approximately 970,570 ha (9,705 square kilometers). It is divided into the historically important

PAGE 157

157 rubber tapper estates ( seringais), stretching across seven muni cipalities and bordering the Brazilian highway leading to the Pacific. Research was carried out in eight rubber estates within three municipalities of the CMER: Assis Brasil (So Francisco, Icuria, Paraguau), Brasilia (Humaita, Porongaba, Filipinas) and Xapuri (So Joo do Iracema, Indpendencia). Seringais were chosen with the assistance of reserve lead ers, and were stratifie d on the basis of their location and orientation towards perceived traditional extractive and non-traditional rapidly changing communities. ACRE CMER BRAZIL w 69o30 w 69o00 w 68o30 w 68o00 s 11o00 s 10o30 s 10o05 Xapuri Brasilia Assis Brasil Figure 5-1 Location of the Chico Mendes Extractive Reserve Household Surveys A total of 14 9 household interviews were co mpleted within the eight communities between March 2004 and January 2005. Each interview took about three hours to complete, and contained

PAGE 158

158 open and closed ended questions regarding general land-use a nd household characteristics and history, as well as specific questions regarding individuals self-definiti on issues. Randomization, in this case, was not possible. Difficult access in the rainy season and incomplete residency lists caused the research team to adopt a more opportuni stic approach. Few roads exist in the reserve, and some communities are 50-75 km from the neares t urban center. As a result, attention focused on ensuring adequate geographic coverage while stratifying for different livelihood orientations and emphases (i.e., extractivist, agriculture, small animal producti on and cattle). Data from the surveys are used in descriptive statistics and in the multivariate regression model. Results of the data are enriched by open-ended in terviews with key-informants. Operational Definitions, Measurement and Hypothesiz ed Effect of Explanatory Variables This section presents operational definitions for a respondents se lf-definition outcome variable as well as a set of explanatory variable s that may influence a respondents self-definition as a rubber tapper. These variables consider both indicators related to rubber tappers historical trajectories as well as explore the contradictor y nature of changing land use observed in the Chico Mendes Reserve. Table 5-1 contains definitions and the operationalization for al l explanatory variables used in the analysis, in addition to their expected effects on a household heads self-definition as a rubber tapper. We focus on a single outcome binom ial variable: a respondents self-definition as a rubber tapper. To account for variation in this outcome variable, the analysis includes five groups of explanatory variables: socioeconomic characteristics of the households heads, institutional and organizational contexts, landuse practices, location, and knowledge about land use rules in the reserve. These i ndicators provide a measure of th e diverse factors that may lead to a household heads self-definition as a rubbe r tapper, given the transformations currently occurring in the extractivis t economy throughout the region.

PAGE 159

159 Table 5-1. Operational definitions and hypothesized relationships of explanatory and outcome variables Variables Operational definition/Unit Expected effects on individual definition as rubber tapper Outcome variable: Household heads self-definition rubber tapper (0=no, 1 = yes) Independent Variables: 1 Socioeconomic Indicators A Household Heads Background First occupant of the ho mestead (0=no, 1 = yes) + Years of residence y ears in the homestead + Age of household head years + Schooling of household head years Previous residence 0 = in the reserve; 1=outside + Relatives living in this seringal 0=no, 1 = yes + B Assets Household owns a tv 0=no, 1 = yes Household owns gas 0=no, 1 = yes Household owns a refrigerator 0=no, 1 = yes Manioc processor house 0=no, 1 = yes Household owns a solar panel 0=no, 1 = yes Household owns a chainsaw 0=no, 1 = yes Household has a wood house 0=no, 1 = yes + Household owns a phone 0=no, 1 = yes 2 Institutional Context indicators Assoc. at municipal level Membership ( 0=no, 1 = yes) + Assoc. at community level Membership ( 0=no, 1 = yes) + Assoc. w/ producers cooperative Membership ( 0=no, 1 = yes) + Assoc. w/ rural workers syndicate Membership ( 0=no, 1 = yes) + Participated in empate ( 0=no, 1 = yes) + Participates in multiro ( 0=no, 1 = yes) + Household head received gov. credit ( 0=no, 1 = yes) 3 Land-Use Indicators C Forest Extractivism Rubber tapping ( 0=no, 1 = yes) + Total area hac + Rubber production measured in kilos + Total number of brazil nut trees number + Total brazil nut production measured in kilos + D Annual Crops sold Rice production sold kilos sold Corn production sold kilos sold

PAGE 160

160 Table 5-1 Continued. Variables Operational definition/Unit Expected effects on individual definition as rubber tapper Bean production sold kilos sold Manioc production sold kilos sold E Cattle ranching and pasture Raise cattle ( 0=no, 1 = yes) Household started to raise cattle measured in years Total number of head of cattle number Total number of cattle head sold number Household started to make pasture measured in years Total size of pasture area measured in ha Plan for # of cattle in 10 years number 4 Location Household at municipalities 1=Xapuri + 2=Brasileia + 3=Assis Brasil 5 Land use rules indicators Knows the utilization plan ( 0=no, 1 = yes) + Know the 10% deforestation rule ( 0=no, 1 = yes) + Know the 5% pasture rule ( 0=no, 1 = yes) + Socio-economic indicators Household heads socio-econom ic indicators have been used elsewhere in modeling smallholder deforestation in Amazonia, in re lation to wealth, and productive practices, among other characteristics (Brondizi o & Siqueira 1997; Godoy 1997; Godoy et al. 1997; Brondizio 1999; Brondizio & al. 2001; Perz 2001b, 2001a), but assumptions regarding the classification (e.g., small farmer, colonists, caboclos) are take n for granted and little effort is made to understand how these rural people in fact see themselves. The first group of variables presents socioeconomic indicators divided by househol d heads background and assets. By background (table 5-1a), we refer to house hold heads first occupancy of th e homestead, previous and length of residence in the homestead, as well as age and kinship ties.

PAGE 161

161 Many homesteads in the CMER are very ol d, dating from the rubber boom period, while more recent new settlements are being opened through sub-division of those old settlements (Gomes 2001). We have noted that there are re cent changes in migration dynamics within the CMER, as many families constantly seeking out a home with better road access and closer to cities. Being a first occupant and years of residence in a homestead may give an indication of household heads origin and/or iden tification with the history of th e region. As the rubber market has drastically decreased in the region, we have observed that rubber tapping is often practiced to a greater extent by older househol d heads that have a vast kn owledge regarding extraction and processing techniques, often preferring to tap ru bber over some other activity despite market constraints, while younger household heads tend to be engaged in other activ ities, due to a lack of historical commitment to extractivism (Gomes 2001; Vadjunec et al. nd). We, therefore, expect that household heads who are first occupant of a homestead with a longer residence in the reserve, and older household heads should be more pr one to define themselves as rubber tappers. By assets (table 5-1b), we consider a householder access to larger, more expensive secondary material goods. As individuals in the region are increasingly integrated into regional and, even global markets, we have also seen impo rtant transformations in their lifestyles, having more access to material goods that until recen tly were seldom seen in the reserve (e.g. refrigerator, TV, solar panels, cell phones, elec tric generators). These changes may directly influence diet, lifestyle and means of communi cation in the CMER. For example, until recently, communication was more based on a collective leve l, with news being spread via the local association either through word of mouth or two-way radio. Incr easingly, individuals in some areas now have their own cell phone, thus el iminating the need for collective means of communication. Changes are also observed in terms of living condition as respondents

PAGE 162

162 increasingly buy chainsaws or pay others an hourly rate for the use of it, and we have noted changes in their housing quality (finished wood houses vs. traditional rubber tapper construction). With those transformations, rubber tappers may reject the rubber tapper lifestyle for its perceived backwardness, desiring instead, a seemingly more modern counterpart. We expect that this increased access to material goo d will also impact social status, and therefore may have a negative affect on a respondent s self-definition as a rubber tapper. Institutional context indicators The second group of variables represents inst itutional context and social organization indica tors. Institutional variables include me mbership association at the municipal and community level, participation in the workers union, cooperativism, empates multiro, and access to credit. Rubber tapper co mmunities in the CMER have several forms of nested institutional-organizational affiliation structures, operating within and outside their communities. (Cardoso 1992; Vadjunec 2007). Individuals firs t institutional affiliation occurs at the community level as families living close to each other create association, in which important collective decisions are made re lated to shared concerns abou t the community (e.g., maintenance of varadouros or feeder roads, marketing transportation, health care, and education). A second level of association affiliation operates at the municipalities co mpassing the reserves territory, as a strategy to coordinate th e reserve management in different municipalities, making the link between rubber tapper community associations, the Brazilian government and other agencies in the region. Other institutional measurements are related to affiliation with institut ions not restrict to the CMERs residents but direct linked to the ex tractive reserve roots. Cooperative affiliation, a more direct market oriented inst itution, plays an important role fo r the reserves residents, such as the marketing of extractive goods and helping ease transportation costs. Moreover, the rural

PAGE 163

163 workers syndicates were historically key enti ties articulating the initi al needs of the rubber tappers movement (Hecht & Cockburn 1989; Paula 1991; Sobrinho 1992), and thus, we, consider the reserve residents a ffiliation to those institutions. A major characteristic of the rubber tapper m ovement in the 1980s was its organizational capacities and strong sense of co mmunity. Living isolated and geographically dispersed in the forest, they slowly began to organize against th e major causes of deforest ation that occurred in the area, and developed a resistance movement by the means of empates or stand-offs, blocking people and machinery with their bodies as a strategy to pr otect the forest against cattle ranchers (Paula 1991; Calaa 1993; Almeida 2002). H ouseholders in the CMER are in the core area where such strategies developed. Indeed, se veral of the individuals in our sample reported that they participated in empates while others, especially younge r household heads, have older family members who were directly involved in or knowledgeable of those historical processes, which are still referred to in the current day. Another means of mutual collaboration was known as the mutiro a still frequent way to help each other at the homestead level, such as clearing trails, building bridges to facilitate transportation, taking care of a neighbors agricultural field in case of sickness, among other collective tasks. Participation in such institutions and traditi onal forms of mutual coll aboration is indelibly related to shared concerns in the CMER that arise due to co-residence in forested areas traditionally associated with extractivism. Overa ll, we expect that individual affiliation with these institutions together with participation on tr aditional practices of mutual collaboration that are very important on rubber tapp er culture, given their collective nature, should strengthen traditional livelihoods systems and thus, we exp ect, will portray a positive influence on an individual respondents self-d efinition as rubber tapper.

PAGE 164

164 Land-use indicators The third set of indicators is related to re source m anagement, grouped into three main components of livelihoods systems in the CMER: forest extractivism, an nual crops and cattle ranching. This set of indicators provides a means of measuring the effects of livelihoods changes, and are expected to have mixed results on an individual self-definition. Extractivist production (table 5-1c) have been historically important to rubber tappers livelihoods, and is in the core of the extractive reserve concept (Allegretti 1989; Schwartzman 1989), and thus is expected to be significant in defining individua ls self-definition as rubber tappers. Recent studies, have identified livelihood changes in the CMER (e.g., Sassagawa 1999; Gomes 2001; Ehringhaus 2005; Vadjunec 2007). Given that subs istence agriculture has long s een mostly as a subsistence land use practice among rubber tapper, we measured agriculture in term of actual agricultural products sold (table 5-1d). Among those changes, small-scale cattle raising represents the most questionable for what was initially envisioned as a sustainable livelihood system for extractivist populations. Cattle ranching and pasture-related indi cators (table 5-1e) are considered due to its recent importance as a short-term income source fo r families as well its contradictory nature of changing livelihoods with the hist orical struggle and political di scourse of the rubber tappers movement. Given the history of the movement as anti-ranching, today, cattle and pasture development, in addition to increased commercial agriculture, is likely to constrain and reduce individual self-definitio n as rubber tappers. Location indicators We also attempt to capture differences through location of households across municipalities. The reserve is spr ead out into thre e municipalities with different histories in the articulation of the rubber tappers movement and current differen t levels of support for extractive reserve development initiatives. There are al so different ecological zones in the reserve

PAGE 165

165 potentially affecting access to forest resources (i.e. some areas do not have access to rubber and/or Brazil nut). Given thes e differences, we measured a re spondents self-definition across three municipalities, in a attemp t explore if they may differentia te across location despite being within the same extractive reserve. We expect that in the municipa lity of Assis Brasil, which has little forest extractive resources, location will have a negative impact on individuals selfdefinition as rubber tappers. Reserve wide land-use rules A final group of indicators is about land-use ru les in the reserve. Th e reserve was created as the m ain result of the rubber tappers move ment, forging a new land tenure system with comanagement responsibilities divided between th e Brazilian government and the rubber tapper organizations. As the CMER was created, the Brazilian government along with rubber tappers created a series of rules combining traditional practices with Brazilian environmental laws which all the reserve residents have to obey, referre d to as the Utilization Plan. We collected individuals knowledge about the utilization plan extensively (see Vadjunec 2007), but for the purpose of this analysis we measure, only three asp ects. First, if the resp ondent was aware of the existence of the utilization plan regulating land use in the rese rve. Then, we used two more specific questions related to the most eviden t transformations we have observed, that is, increased deforestation for pasture land for cattl e raising. We asked if the respondents recognize the rule that states that they can deforest only 10% of their property. Lastl y, if they are aware of this rule, we asked them if they were aware that out of that 10% permitted, only 5% can be allocated for pasture creation. This gives evidence if they were in compliance with the rules created and, in general terms, th eir commitment to the extractive reserve model as a means of sustainable development. Despite a household h eads potential knowledge of those rules, we cannot necessarily assume that they always obey th e rules. Evidence, however, exists that even

PAGE 166

166 mere knowledge of the rules positively impact s a community, creating social cohesion and decreasing environmental impacts (Ostrom 1990; Moran & Ostrom 2005). We expect that respondents knowledge about the rules in the reserve should have a positive impact on a respondents self-definiti on as a rubber tapper. Logistic Multiple Regression Model The variab le that we used to represent rubber tapper individuals self-definition is a binary variable. The most appropriate mode l, in this case, is therefore a binary logistic regression. Such a model is used in instances of binary variables when the desire is to estimate the probability of the occurrence of an event (Agresti 1990), or in our case, the probability of household heads self-identifying as rubber tappe r or not. In our model, Yi is defined so that: 1= self-identification as being a rubber tapper, and 0=no self-identification as rubber tapper. Logistic regression model allows estimation of the likelihood of the occurrence of an event e xpressed as the probability of the event occurring relative to the probability of the event not occurring, where pi is the likelihood of the occurrence of the event i for each individual. Logist ic regression allows estimation of coefficients b for explanatory variables X (1n). The coefficients can be exponentiated to yield odds ratios which indicate the change in the likelihood of Y=1. Results from logistic regression thus indicate effects of the independe nt variables on the likelihood of self-identifying as a rubber tapp er, including the variables which make such self-identification more likely. In order to define the final model and mi nimize the number of variables used that maximizes the precision of the model, we test ed one by one variable until the final model was found. This allowed us to start with a high num ber of inter-related va riables (explained above) and systematically eliminate all variables that pr oved to be insignificant, to give us the best

PAGE 167

167 possible regression model for our study. For more detail on the operationalization details of the dependent and independent variables, see table 5-1. Results Analysis of Descriptive Findings A responden ts self-definition as a rubber ta pper may reflect, among other things, ones background as a traditional forest -dweller of a specific socioec onomic class, ones degree of participation in the rubber ta ppers institutions and social movement, ones main livelihood practices, the degree of integr ation into market for both extractivist and non-extractivist production, in addition to ones recognition of the current land tenure regulation. Table 5-2 presents descriptive statistics for those charac teristics among household heads in the CMER. We find it surprising that 78 percent of responde nts still identify as rubber tappers. Among the block of variables that assess socio-economic aspects for household heads background, only 27% of them were the first residents on the prop erty, while the average duration of residence was over 13 years and the mean age of household heads was 40 years old. These averages suggest generally older household heads and a relativel y long duration of residence; however, the standard deviations reveal considerable variation among respondents in the sample. It also shows that some assets related to improved livelihood needs and production activities are become more common, as over 50% of the individuals surveyed own a finished wood house and have access to gas for cooking, wh ile over 30% own a chainsaw and a house to process manioc. Other assets more related to communication that usually depend on surplus cash, such as the possession of a TV or a phone, are still limited (possessed by less than 15% of the individuals in the sample). These findings we argue are important, because they are relatively new material goods, implying a recent and differentiated socio-economic status for some rubber tapper families in the reserve.

PAGE 168

168 Table 5-2. Descriptive statistics for socioeconom ic aspects, institutional context, land-use practices, location and land-use rules, among households in the Chico Mendes Reserve Variables Mean Standard N deviation Outcome variable: Self-definition as rubber tapper (0=no, 1 = yes) 0.78 0.41 147 Independent Variables: 1 Socioeconomic indicators A Household heads Background First occupant (0=no, 1 = yes) 0.27 0.45 143 Years of residence 13.17 10.28 148 Age of household head 40.69 13.46 149 Years of schooling of household head 2.5 1.71 149 Previous residence 0.29 0.45 147 Relatives living in this seringal 0.87 0.32 147 B Assets (0=no, 1 = yes) Household owns a tv 0.11 0.31 148 Household owns gas 0.52 0.5 148 Household owns a refrigerator 0.02 0.16 148 Manioc processor house 0.38 0.48 148 Household owns a solar panel 0.15 0.36 148 Household owns a chainsaw 0.33 0.47 148 Household has a wood house 0.75 0.43 149 Household owns a phone 0.08 0.27 148 2 Institutional indicators (0=no, 1 = yes) Association at municipal level 0.92 0.26 148 Association at community level 0.62 0.49 148 Association producers cooperative 0.14 0.35 148 Association w/ rural workers syndicate 0.69 0.46 147 Household head participated in empate 0.25 0.43 143 Household participates in multiro 0.76 0.42 143 Household head received gov. credit 0.48 0.5 147 3 Land-Use Indicators C Forest extractivism (kg) Rubber tapping (0=no, 1 = yes) 0.41 0.49 146 Total area (hectares) 426.17 303.67 149 Rubber production 142.68 269.91 149 Total number of brazil nut trees 186.24 286.58 149 Total brazil nut production (kg) 1686.25 2680.94 149 D Annual crops sold (kg) Rice production sold 248.04 387.83 149 Corn production sold 81 339.43 149

PAGE 169

169 Table 5-2 Continued. Bean production sold 162.1 268.04 149 Manioc production sold 173.55 799.35 149 E Cattle ranching and pasture Raise cattle (0=no, 1 = yes) 0.86 0.34 149 Total number of head of cattle 14.04 17.46 149 Year household starte d to raise cattle 8.52 7.77 135 Total number of cattle head sold 3.48 9.19 149 Total size of pasture area 8.79 7.62 149 Year household started to make pasture 8.57 7.14 131 Plan for # of cattle in 10 years 84.55 172.09 138 4 Location Household at Municipalities 1=Xapuri 0.19 0.39 149 2=Brasilia 0.35 0.48 149 3=Assis Brasil 0.46 0.49 149 5 Land use rules (0=no, 1 = yes) Knows the utilization plan 0.79 0.4 147 Know the 10% deforestation rule 0.81 0.38 146 Know the 5% pasture rule 0.72 0.44 146 1. Values shown are either pr oportions (binomial variable) or arithmetic means (continuous variable) Regarding institutional-organizational context indicators, overall, 60% of household heads were affiliated with a grass-roots institutions, ex cluding cooperatives; 92% were affiliated with the reserve-wide association, demonstrating the importance of the association in recognizing individuals as official reserve residents, divvy ing up land and resources. Individual affiliation with a cooperative was much lower, only 14%, implying that they access other mechanisms in marketing products. Nearly 50% of the household heads had received government access to credit in some form. This strikes us as somehow surprising, as only recen tly have rubber tappers even begun to receive formal government credits. These results imply an increased governmental recognition for rubber tappers, a result of their con tinued political coal ition and broader development agenda in the region. In addition, pa rticipation in traditional forms of cooperation were high with 76% of household heads having participated in a mutiro and 25% having

PAGE 170

170 participated in an empate implying that such households were directly engaged in the movement for some time, twenty years ago when the last empate took place. Forest extractivist practices and the intensity of extractivism depend mo stly on the size of land traditionally measured by the number of rubber trails (01 rubber trail = 100 hectares), as well as resource (rubber and Brazil nut) dispersion in different ecological zones of the reserve (Vadjunec 2007). Thus, total area of a household s howed an average of 400 hectares, or about four rubber trails per households. Despite this land availability for tapping rubber, only about 40% actually tap rubber, and the average production is about 142 k ilos per year, an insignificant outcome for people that traditionally had their main livelihood economy based on rubber production. Obviously, a clear sign of the stagnation of the rubbe r economy in the region. The number of brazil-nut is the next most important measurement of forest extractivist production, and individuals are very aware of how many brazil nuts trees they hold as a means of projecting annual brazil nut production and income, as well as calculating the value of their land. The average number of brazil nut trees per propert y was close to 200 trees, while annual production was over 1600 kilos per household. Overall, forest extractivism indicators showed substantial standard deviations, indicating differential produ ction, likely to be a result of substantial variation regarding extractivist resources across different areas of the reserve. The descriptive statistics also show low m eans of agricultural production sold, with rice representing the most important cash crop--wit h less than 300 kilos sold per household. This suggests mainly subsistence agriculture, and de monstrates an overall low market link among respondents based on agricultural production, but standard deviati ons are also high, implying that for some individuals agricultural production might represent an important income generator. Cattle ranching appear as a widespread activity, as 86% of the respondents raise cattle, with an

PAGE 171

171 average of 14 heads of cattle and a mean of a bout 9 hectares of land unde r pasture per household. The high percentage of cattle rais ing combined with the fact that both the number of years since a individual started to raise cattle, and number of years started to create pasture area average less than 9 years, indicates that ranching a relatively young activity, having increased substantially only recently. This is supported by the average h ousehold heads plan for cattle in the next tenyear period, averaging over 80 heads. Analysis of the Multivariate Regression Model For a better understanding of the complex in teraction between vari ables, this section presents the multivar iate model in which the variab les ultimately were found to be significant in the model tested. Table 5-3 presen ts the results of the logistical regression model regarding respondents self-definition regr essed against indicators of ho usehold heads background and assets, institutional-organizat ional context, livelihood practic es and land-use regulation indicators. The first determinant of self-definiti on involves a household heads background. In contrast to our expectation, the odds ratio indicates that individuals that were first occupants of the homestead were 0.3 times more likely to decl are themselves a rubber tapper than those that are not first occupants. In other words, the firs t occupant have less chances of self-define as rubber tapper, as indicated by a odds ratio less th an 1. We suspect, homestead turnover is not due to outsiders coming in, but ra ther reflects increasing sub-di vision among family members for more secure usufruct rights to their land and r ecognition as a new CMER re sident. Even so, most such young household heads tends to neglect traditional forms of extractive livelihoods, but may be linked to extractive cultural iden tities through their pa rents history and experiences. None of the other household heads bac kground variables show significant net effects on individual self-

PAGE 172

172 definition, which suggests that the way they define themselves is complex, regardless most of the background variables measured. Table 5-3. Multivariable model of household self-definition as rubber tapper regressed on household background, institutional context, land -use practices and regulations in the Chico Mendes Extractive Reserve Variable B Odds Ratios Wald ChiSquare df Sig. Intercept -1.884 4.194 1 .041 Household heads background First occupancy of the homestead No -1.182 0.3 4.647 1 .031 Yes 0.000 Institutional context Affiliation w/ community association No -1.347 0.3 5.001 1 .025 Yes 0.000 Participation in mutiro No 1.390 4.0 4.901 1 .027 Yes 0.000 Land use Total area household landholding (ha)None (ref.) 2.788 4.9 8.019 1 .005 100 < 300 1.189 9.5 1.911 1 .167 300 < 600 0.538 16.2 0.318 1 .573 > 600 0.000 Reserve regulation Knows the utilization plan No 1.000 2.7 3.112 1 .078 Yes 0.000 Two institutional context variables exert strong effects on a household heads selfdefinition. Individuals affiliated with a commu nity level association, diverging from our expectation, tend to abandon the tr aditional notion of rubber tapper, as the odds ratios indicate that individuals affiliated with community level association were 0.3 times more likely to define themselves as rubber tappers than those not aff iliated. That is, individu als affiliation with the association tend to decrease th e chance of self-definition as ru bber tapper. In addition, household heads participation in a mutiro shows a significant positive effec t: respondents that participated in a mutiro are 4 times more likely to declare themselv es a rubber tapper than those that did not participate.

PAGE 173

173 Land use variables reveal a limited array of si gnificant factors. We expected that land use practices related to extractivist activities would positively influence individual self-definition as a rubber tapper, while other predicators relate d to cattle ranching and pasture would have a negative effect. Our expectation was not met. Th e only predictor that exhibits the anticipated effects was the total land area of the household, a finding that supports arguments that selfdefinition is not dependent on land-use practic es, but on the size of the land hold. Traditional land area definition is based on the number of r ubber trails in a households land holding, but in a few parts of the reserve (espec ially in the municipality of A ssis Brasil), rubber trees are not present, and thus this traditional definition is problematic. Individual with areas with 300 hectares or three rubber trail (01 rubber trai l = 100 ha) are 4.9 times more likely to declare themselves as rubber tappers than those where rubber trail is not an appropriate definition of land measurement. This chance increas e according to the size of land holding as individuals with 300600 hectares and individuals holdi ng over 600 hectares (or 06 rubber trail) are respectively 9.5 and 16.2 times likely to declare themselves as r ubber tappers than those with land not formally defined based on the number of rubber trails. Although we expected th at size of landholding would be important because it implies more potentia l extractivist resources available to the landuser, this effect was not observe d when looking at actual land use variables, suggesting that size of the area a key factor for ex tractivist production is key fo r household heads self-definition as a rubber tapper but does not contribute to supp ort land use practices as a significant predictor for self-definition. Household heads general knowledge about the utilization plan also presents important significance in their self-definition as rubber tapper. Those individua ls that responded that they knew about the reserve utilization plan were 2.7 time s more likely to define themselves as rubber

PAGE 174

174 tapper than those who did not know the utiliza tion plan. More detailed knowledge about two specific rules regarding land use did not show significant net effects on a respondents selfdefinition Discussion These findings bear im portant implications for (i) understanding of the factors that differentiate among who self -identifies as rubber tappers and w ho does not, (ii) the recent history of the rubber tapper struggle for fo rest conservation, (iii) the role of institutions and land use practices on household head livelih oods improvement and (iv) polic ies for extractive reserves as an important environmentally sound strategy in the region. The multivariate model provides empirical support that reveals only a few factors exert significant effects on a household heads self-definition, showi ng some support for what is commonly expected, but more importantly not su pporting general expectations of what a rubber tapper is expected to be in the region, accord ing to their historical social movement and traditional livelihoods trajectories With the exception of being the first occupant, socioeconomic factors did not exert significant effects on respond ent self-definition, contrary to our expectation. Yet, first occupancy, as observed in table 5-3 reveals a significance contrary to what we expected as being the first occupant in creases the likelihood of those households heads not considering themselves a rubber tapper. Homesteads are traditionally older-aged settlements (Gomes 2001). But we have observed since the reserves establishment, especially on those areas close to cities, that old homesteads have been increasingly sub-divided for the fo rmation of new homesteads by family members. This might represent a shift suggesting that ne w occupants may have th e tendency to reject traditional notions of rubber tapper definition. New families may be opting for a modern

PAGE 175

175 lifestyle following general trends of the regi on, refusing the idea of a traditional lifestyle imposed on them by the traditional and narrow historic definition of a rubber tapper. Only two out of seven institutional variables showed significance in individual respondent self-definition as presented on tabl e 5-3, with one (affiliatio n with the community level association) working against traditional ru bber tapper definition, as the likelihood of household heads declaring themselves as rubber tapper was higher for those not affiliated. This finding seems to represent a rupture to traditiona l rubber tapper definitio n in the core of the smallest social organization stru cture that was very important to the coalition of the well-known tappers movement for land and social justice merely twenty years ago. It is easy to see the early success of these organizational proce sses and expect that this is maintained in the core of their social structure. But, with the creation of the re serve their main goal (right for land) was finally reached, and their agenda as a social group, reside nts of the reserve, appears to have shifted, becoming more about development and less about the history, unity and success of the social movement (Ehringhaus 2005). In many instances the real or historic lead ers of the movement have left the reserve, participating instead in political discussions outside of the reserve. Vadjunec (2007) cites increasing dissatisfaction of resident s with the community level associations, suggesting that thes e associations are no longer articul ating the needs of traditional rubber tappers. The rubber tapper movement is very alive on the development agenda in Acre with supporters from NGOs and government instit utions. But residents of the reserve, by refusing a self-definition as a rubbe r tapper in the bottom of the dive rse social struct ures, seem to challenge those institutions that historically supported and capitalized on the image of the traditional rubber tapper an definition that seemed more appropriate to win over policy makers by emphasizing the reserve and the movement in term s of a conservation initiative, rather than a

PAGE 176

176 economic development initiative. As expected, participation in a multiro has a positive impact on respondent self-definition as rubber tapper, su ggesting that rubber tapper definition is more tied to sources of social community building for mutual practical collaboration than for their current formal institutions at the community level. Livelihoods changes seem to be the obvious in dicators revealing the loss of traditional lifestyles among the reserve residents. Yet, not one land-use indicator measured exerted significance in a household heads self-definition. In contrast to our expectations, neither traditional land use practices (rubber and brazil nut extractivism), nor commercial agriculture, nor cattle ranching were important in a household h eads self-definition as a rubber tapper. That is to say, individual self-definiti on cannot be predicted by current land-use ac tivities. Despite the widespread decline of rubber tapping or recently increased small-scale ca ttle ranching practices, the category of rubber tapper continues to be cult urally important, as descriptive analyses show that over 70% of the individuals in terviewed consider themselves as rubber tapper. That is to say, despite livelihood transformations underway, most respondents still refer to themselves as rubber tappers, illustrating the cultural importance of the title, and perhaps the perceived political benefits that go along with it. If rubber tapping can arguably be linked to rubber tapper historical definition, is seems to be problematic to consid er individuals that do not tap rubber as rubber tappers, today. In fact, one importa nt challenge for the rubber tappers today is the recasting of their traditional defini tion during a time practically devoid of rubber, followed by increasing cattle ranching practices. Cattle ranching among rubber tapper are neith er a fabricated argument nor outsiders speculation against traditional per ception of rubber tapper. Thus, it appears as the main factor in this contrasting transformation and one that easil y finds opponents. In fact, much of the initial

PAGE 177

177 success of the rubber tappers movement came from not being a cattle rancher, or being exactly the opposite of a cattle rancher. Rubber tappers were the foil to cattle ranchers it was a powerful way to create political allies with environmentalists and among the international community. The inspirations behind the rubber tapper movement against encroaching cattle ranching in Acre are well expres sed in their cultural manifesta tion involving music, religion, and art, as can be seen in the rubber tapper anthem.1 Cattle ranching expansion among the CMER residents has not been formally addressed by governmental institutions either to discuss potentially better ways to enforce current regulati on or to sustainably manage inside the CMER. Leaders of the reserves resident associations are open to discussing the issue, and possess a strong sense of reality about the situati on; as a leader from Brasilia says: Today, people have different visions. It is no t that same vision of the trade unionists from 20 years ago. The vision of some officials in our association is differe nt. They are sons of the native or historic rubber ta ppers, but have differe nt vision of their father. So, this is a contradiction. Even more, today almost everybo dy is in favor of the raising of livestock. We can no longer be against cattle this time is over. All of us have cattle. What we can say now is that we need to create new rules fo r cattle raising and past ure creation inside the reserve, we intend to do this; or better yet, we need to do this. We need proper technology in order to modify the kind of cattle we raise, not only with the idea of raising cattle for sale. Beyond this, we need these animals for pulling cargo and for milk for our children. Our leaders stayed very quiet du ring this entire time, and extractivists tried to conceal what is going on, what they were doing, creating pasture behind their seconda ry parcels. Some directors of the associations still avoid th e question when someone asks them, how is deforestation being controlled in side the reserve. I am agains t deforesting large clearings, but in those areas that are already cleared, lets raise cattle. The situation will only change if a plan provides strong economic alternativ es for extractivists, only this will decrease deforestation for pasture on the CMER. The results in cattle ranching and pasture areas indicators lead us to believe that it is not seen by individual respondents as problematic for their self-definiti on as rubber tappers. However, the analyses show that the likelihoods of residents to consider themselves as rubber 1 The rubber tapper anthem (hino do seringueiro): Lets give value to the rubber tappers/ lets give value to this nation/ for it is through their work/ that cars and airplanes ti res are made/ Bicycle tires are not made of cheese/ It is not cattle leather that truck tires are made of/ it is not ca ttle horns that erase letters, no/ these are all products of rubber, made by our hands. (By J.S. Arajo First president of the Rubber Tapper National council).

PAGE 178

178 tappers is higher when they know the existen ce of the utilization plan, although variables measuring specific rules about ca ttle ranching and pastur e regulation (e.g. size of land allowed for deforestation and for pasture) did not show significance in sh aping individuals self-definition. Cattle ranching remains a dirty word in the CM ER, being politically se nsitive and can be a source of conflict among neighboring, and regulation of cattle raisi ng and pasture area has been part of a debate among individuals with differe nt development perspectives. During fieldwork, we experienced some household heads engaged on cattle raising not willing to talk about cattle ranching related questions, while few others, not engaged in such activity, demonstrated strong opposition for their neighbors involvement with it. In this context, pasture lands often do not surround the house, and can be distributed in ve ry different parts of ones landholding. Those respondents with a greater tendency to expand ranching, knowing the rules limiting the expansion of pasture areas, tend to distribu te pasture land in different parts of the colocao with no continuing pasture areas. This makes it more unlikely for a landholder to be penalized. In fact, this corroborates the argument that individuals respondents ar e indeed conscientious about ranching and land-use regulations w ithin the reserve and the political issues that cattle ranching carry, contrasting with th e lack of significance observed in the model. Some household heads defend more strict enforcement of the current regulation measured by overall size allowed for deforestation, while others defend a more direct regulation about cattle, for instance controlling the number of heads a householder may be allowed to hold. Land use regulat ions directly related to cattle raising (5% of the overall 10% allo wed) are increasing re jected by individual respondents more engaged in cattle ranching acti vities, showing a trend that cattle herds will continue to grow if depe ndent on a individuals will.

PAGE 179

179 There are current contrasting perspectives among individuals. We interviewed an older household heads that has lived in the same rubber estate all his life and is becoming part of a minority still linked to rubber tapping. His connec tion is expressed not only by his significant rubber production, but also by a profound tie to th is activity in a much more symbolic than material relation. His tie to extractivism may explain the absence of cattle ranching in his settlement, and his expressed lack of a plan in regards to raising cattle. It is notorious that new generations of household heads are well aware of the laborious hard work of rubber tapping practices and, more importantl y, its seemingly poor economic return, experienced by their fathers or other family members. As one younger household heads (and son of a rubber tapper) explained, Who wants to get up at four in the morn ing to tap rubber, in th e dark, with all of the wild animals and the snakes? I would rather be ou tside in the sun in my pasture. Cattle herds appear as an attractive option that provides liquefiable assets and perhaps status among community members. Local development approach, as evidenced on the forest government concept of Florestania or forest citizenship, assume that cultural values have priorities over material factors in determining livelihood options for the future, and that the tr aditional perception of rubber tappers, traditionally cons ervationist, guarantees the viabil ity of the extractivist economy. Nevertheless, the traditional values shaping rubber tapper definition are at first rooted in their historical material conditions a nd not the opposite as is enforced in local development agenda which builds on traditiona l rubber tapper definition serving to legitimate the political movement and current governmental agenda as being green. In this sense, cultural values alone cannot be used as a guarantee for conservation in the Extr active Reserve. Even though rubber tappers today still have a generally pro-conser vation mindset, it is idealist to believe that their traditional

PAGE 180

180 culture per se can be translated into conservationist practices. Their pro-conservation culture and values can and do change with changes in materi al condition for subsistence. As this analysis shows, most of their social or ganization values that were very important for rubber tappers definition twenty years ago are no t influencing their perception as a rubber tapper today, because values are not static and individuals are changi ng their values to adapt to new socioeconomic conditions, as well as new wants and needs. It is not to say that th eir definition is shaped in term of solely their economic condition. The apparent tension between a pro-forest gove rnment and the maintenance of cattle as big business, we argue produces a sort of local schizo phrenia, where cattle culture and forest culture can exist side by side, as evidenced by this yea rs Expo Acre (a regional cattle and agricultural exhibition) that featured an exhi bition of genetically superior cattl e next to a replica of the house where Chico Mendes was murdered, in an exhi bition celebrating the history of the rubber tappers movement on the 20th anniversary of Chico Mendes murder. If government and policy makers can envision these land-uses side-by-side, it is no wonder th at rubber tapper themselves find it easy to justify such radical land-use ch anges while still incorporating them into an extractivist culture. Yet, cowboy culture dominates rural Acre today, and thus has a direct influence on the reserve residents. The music on the radio is dominated by country music, the night clubs in the cities portray the cowboy life-style as a much wanted and well-respected social status. Small businesses in rural Acre focus on th e demands of agropastoral activities. And the mere trip to the city from the forest reveals th e extent to which cattle do minates the landscape. In the imagination and within the media messages bombarding the rural population in Acre, the country lifestyle is bigger and bett er than the forest lifestyle.

PAGE 181

181 Conclusions Despite an a ttempt to explain individuals self -definition as rubber ta ppers based on a wide set of land-use, institutional and livelihood vari ables, results show that only a few were significant as seen in table 5-3. Our attempt to me asure individual self-definition is the result of prolonged reflection about rapidly transforming landscapes and livelihoods witnessed by the authors while fieldwork was bei ng conducted. Perhaps, issues such as self-definition cannot be measured quantitatively, as our qualitative data shows the complexity the livelihood issues facing rubber tappers today. Yet, we still find this exercise valuable in understanding the sheer complexity of a historically marginalized and misunderstood group. What it means to be a rubber tapper today is a complex question with a mi x of rationality and subjectivity, involving affirmation with historically constructed trajectories and challenged with contrasting development circumstances as livelihood issues con tinue to emerge. Cattle raising, for example, as a historically contrasting prac tices, should not be seen as a rupt ure from or within traditional rubber tapper definition, as this option gene rally has a positive impact on a households socioeconomic status. Instead of traditional valu e abandonment, it reflects a rational decision based on the current best available op tion for their economic improvement. Socio-economic conditions and social services have improved substantially since the establishment of the CMER. Some indicators not measured in the analyses could also affect a respondent self-definition. Transportation, family education, and road openi ng facilitating market links for the residents are part of these improvements. Today, kids and teens almost all have access to basic schooling and are, for the most part, literate. This is in st ark contrast to their parents, the majority of whom are illiterate. Such changes wi ll no doubt continue to reshape rubber tapper definition in the future.

PAGE 182

182 Most research on land-use change takes issues of individual self-def inition for granted. To our knowledge, the majority of research in th e CMER starts with the assumption that the reserves residents are indeed rubber tappers. In order for futu re conservation and development initiatives to be successful, we need a better understanding of wh at ties residents of the CMER together. We argue that developm ent policy needs to be designed with flexible definitions of rubber tapper in mind. NGOs and government po licy that focuses solely on extractivism development, though important, may be out of touc h with the complexity of current land-uses and livelihood preferences of todays rubber tapper. The rubber tapper seems to have a stereotype that is still very mu ch labeled by other, and there has been little attention to how their cu rrent livelihood systems and socio-economic conditions challenge this traditional view. In the 1980s, the rubber tapper movement and allies mobilized the global community co mmitted to preserving the environment, in order to ensure their own land and livelihood rights. While we are not arguing that their social movement is or was in any way false, by fi ghting a global fight for the forest, rubber tappers may have unwittingly created for themselves a very limited definition based on land-use. Such activities as tapping rubber and collecting brazil nuts, admittedly sustainable or green, are inserted in their values as well as in their basic identificati on: the rubber tapper. Thus, the environmental definition was imposed on or inserted within th e definition of a rubber tapper, without denying, confirming a mutual influence betw een the local and the global. The Extractive Reserve system was an expr ession of a multifaceted mosaic of social players of great diversity whos e rationale can be reconstructed by recognizing that they share similar problems, despite their differences and overall goals. But the problems faced by rubber tappers in the past (land securi ty) are not the same as today. Their needs have changed, yet many

PAGE 183

183 of their historical external supporters remain cen tered on the same historical issue, environmental protection. We argue that rubber tappers are and remain concerned with environmental issues, but their major problem today is centered on th e socioeconomic development that has been lacking within the Extractive Reserves. Thus, a lot of assumptions we make regarding what defines a rubber tapper reflects a historical mo ment that does not necessarily reflect reserve residents and their development perspectives today. The notion of rubber tapper strengthened in the 1980s is still very alive and in several aspects legitimate, but has continued to evolve and today may be unrealistic, transcendi ng local realities. As rubber tappers agenda in these twenty years since Chico Mendes has evolved from one centered on the struggle for land and social justice to sustained socioeconomic improvement, their past image may be called into question and potentially criticized by thei r historical allies. It is ther efore important to understand how rubber tappers define themselves today if we ar e to ensure the success of the Extractive Reserve tomorrow.

PAGE 184

184 CHAPTER 6 CONCLUSIONS: The overarching goal of this dissertation has been to address critical questions, concerns, and criticism s regarding the e xperience of ERs as a conservation and development tool in the Brazilian Amazon, and in so doing to integrate core themes: the e volution of the ER as a peoplebased conservation policy; ch anging livelihood conditions and natural resource management practices; deforestation and landuse patterns; and the impact of historical-cultural livelihood practices on rubber tappers self-d efinition today. These research issues were addressed in the four research papers co mprising this dissertation. The first paper of this dissertation (Chapter 2) provided a region-wide analysis of the evolution of ERs over the past two decades, ex ploring how the ER model has expanded and how political and regional development dynamics have affected ER policy either favoring or restricting the esta blishment of ERs. In the first wave states, both Rondnia and Acre have placed a significant percen tage of their territory under ERs. The status of ERs in Acre appears to be solid, in terms of territory and political support while in Rondni a, it is stagnated in growth and there is a profound lack of government suppor t for ER. Rondnias state reserves may be the most paradigmatic in thei r application of the mode l after twelve years; a typical case of struggle between the economic development and environmen tal protection camps, in which the winner is usually determined by the group that has more polit ical weight. The situation in the second wave states looks promising if recent advances continue in coming years; it is very likely they will. Par and Amazonas, are far from reaching some of the first wave states in percentage of land under ERs, but have demonstrated a consistent process of establ ishing ERs since the early 2000s, representing the trend for cont inuing growth of ERs areas.

PAGE 185

185 ERs as a public policy is widely considered on e of the options to simultaneously decrease deforestation rates in the regi on while responding to social group demands, especially in development frontier areas. The creation of ERs s hould not be seen as static or synonymous of forest protection per se. The success of creating a new ER ca nnot be seen by both social movement and government as an end of an end in itself, as have been the case with several ER implemented. Three elements should be essentia l to all ER efforts: 1) alternatives for improvement of economic and social well-being of ER residents; 2) a strong monitoring process; and 3) an active and continuous social movement coalition. The general focus of both social movement and government institutions has been to create new ERs areas, and pay less attention to socio-economic conditions with in ERs. The adoption of cattle ranching among residents of the Chico Mendes reserve addressed in the following chapters is an example of the lack of public policy supporting sustainable economic practices for ER residents. This perspective need to be reviewed in order to guarantee sustai nable future for ERs in Amazonia. The second paper (Chapter 3) provided a comp arative analysis of cattle ranching adoption among colonist smallholders and forest extrac tivists in the Amazon, exploring why livelihood strategies among a variety of so cial actors in a highly heteroge neous socioeconomic region are converging on a single activity: cattl e ranching. Showing similar tre nds regarding the growth of cattle husbandry in Xapuri (Acre) and Uruar (Par), the results suggest a regional push toward cattle ranching expansion among colonist smallholde rs and forest extractiv ists in the Brazilian Amazon. Cattle ranching among colonist and forest extractivists are driven by market forces, but also have distinct political development cont exts. Smallholder colonists present a more diversified economy with greater market links in which cattle ra nching is consolidated as a economic practice, while forest extractivists, rega rdless of the potential of forest products to

PAGE 186

186 improve local livelihoods, have more limited eco nomies in which cattle represent a option under an uncertain economy, and they have not yet confronted the political and development controversies of cattle raising in Extractive Reserves. Cattle ra nching among forest extractivists in Xapuri are a challenge and may represent an opportunity to promote a broader debate to review forest resources co-management pacts by government institutions and grass-root organizations for Extractive Reserve administra tion in the very place where it was negotiated twenty years ago. The third paper (Chapter 4) expanded the discussion beyond issues of cattle ranching adoption, to consider deforestation issues by re mote sensing analysis to household surveys addressing the question of what are the dom inant land-use and live lihoods practices of extractivists in the CMER and how are they changing. The six seringais examined are undergoing rapid land-use and land-cover changes with cattle and agriculture as the main drivers of such change. As of yet, however, none of these six seringais have surpassed the 10 % limit set on the amount of allowable deforestation. Al though some communities may surpass allowable deforestation limits in the near future, the rate as compared to outside ER areas should still be seen as a measure of environmental success for the ER system. At the time of the creation of the CMER in 1990, residents depended more heavily on traditional non-timber forest product extraction including Brazil-nut co llection and latex production. Trad itional land-use activities are currently undergoing rapid changes in the CMER. Rubber tapping has gone from being historically one of the most im portant land-use activities, to the least important making up less than nine percent of a household s total income on average. Cattle and market agriculture are and will likely remain part of a r ubber tappers land-use system in the CMER, and consequently need to be addressed realistically by local instituti ons and decision-makers. Cattle, however, remains a

PAGE 187

187 dirty word in the CMER. Policy directions for the CMER need to address sustainable pasture management, as well as a serious re-investme nt in NTFP marketing and development. The debate over rubber tappers and ca ttle husbandry has seeped into the press, motivating pointed criticism and questions about th e sustainability of the CMER. These critics have shown clear bias against rubber tappers that break with their traditional identity, presen ting the case of cattle ranching in the CMER as if it is a phenomenon ex clusive to ERs rather than a generalized economic and environmental problem in the Brazilian Amazon. Some critics have even used this debate to qu estion the viability overall of the ER model in Amazonia. The murder of Chico Mendes has been enlarged to equate cattle with evil livelihoods. As a result of such caricaturizing, the mere idea th at the rubber tappers championed by Mendes are now raising cattle themselves is met with little enthusiasm by some proponents and supporters of ER concept.1 It must be remembered that Chico Mendes and the rubber tappers may have been fighting for the forest, but als o, and more importantly, for a right to both land security and improved rubber tappers well-being. The fourth and final paper (Chapter 5) provi ded an analysis of rubber tapper identity, exploring which institutional, cu ltural, and land-use factors currently shap e rubber tapper selfidentification in the face of their historical-cultural trajectori es and current liv elihood practices. Results show that only a few variables were si gnificant in explaining household self-identity. Larger size of land-holding, more knowledge of management rules, and participation in community activities make it more likely to id entify as a rubber tapper, while more recent 1For example, while doing fieldwork in Acre, I was a bit concerned about where and with whom I would start to discuss my research topic due to it politically sensitive nature. I gave presentations for both local and international NGOs branch offices as well as at various seminars at the Federal University of Ac re (UFAC). An international NGO, with a local branch in Rio Branco (Acre), saw a ne ed and opportunity to develop a proposal regarding cattle ranching expansion in the CMER. I was contracted as a c onsultant to address cattle ranching within the Reserve as part of a project to review regulations regarding cattle raising and pasture management initiatives in the reserve. My final report, despite being well received at the local branch, was excluded from the annual report at the international office in Washington, DC, and no funding was allocated to the proposal presented.

PAGE 188

188 occupancy in the CMER and involvement in the official movement have a negative impact on identifying as a rubber tapper. Ca ttle raising should not be seen as a rupture from or within traditional rubber tapper definition, as this option generally has a positive impact on a households socioeconomic status. Instead of iden tity abandonment, it reflects a rational decision based on the current best available op tion for their economic improvement. Most research on livelihoods change takes issues of identity for granted. To our knowledge, the majority of research in the CMER starts with the assumption that the reserves residents are indeed rubber tappers. In order for future conservation and development initiatives to be successful, we need a be tter understanding of what ties residents of the CMER together. Otherwise, development policy may be out of touc h with the complexity of current land-uses and livelihood preferences of todays ER residents. The stereotype of the traditional rubber tapper is still very much labeled by other, and there has been little attention to how thei r current livelihood syst ems and socio-economic conditions challenge the traditional perceptions of rubber tappers. In the 1980s, the rubber tapper movement was seeking new allies and fostered the construction of new images; as a result, they were able to mobilize a global community committe d to preserving the environment, and ensure the protection of their own land and livelihood ri ghts. Although we are not arguing that this definition is or was in anyway false, it now seems that rubber tappers, by making their struggles a global fight for the forest, may have unwittin gly frozen their definition in time. This environmental image was inserted in the defi nition of a rubber tapper, revealing the crossinfluence between the local and the global. R ubber tapper definition is a complex and everchanging idea, participatively cr eated over time. Although the defi nition was formed publicly at

PAGE 189

189 the height of the rubber tappers movement in the 1980s, it is continuously evolving as new landuse and livelihood issues emerge. The Extractive Reserve concept was an expressi on of a diverse mosaic of social players recognizing that they share similar problems, desp ite their political differences and varied goals. Yet, many of the rubber tappers historical ex ternal supporters remain centered on the same historical issue: environmental protection. We argue that rubber tappers are and remain concerned with environmental issues, but th eir major problem today is centered on the socioeconomic development that has been lack ing within the Extractiv e Reserves. Thus, many assumptions we make regarding what defines a rubber tapper reflects a historical moment in time that does not necessarily concord with the deve lopment perspectives of CMER residents today. The definition which strengthened in the 1980s is still very alive a nd in several aspects legitimate, but has continued to evolve in the past twenty years, bringing up new issues that may be called into question and potentia lly criticized by rubber tappers hist orical allies. It is therefore important to understand how rubber tappers define themselves today if we are to ensure the success of the Extractive Reserve tomorrow.

PAGE 190

190 APPENDIX A CORRELATION ANALYSIS Table A-1. Analysis of relations between outco m e and explanatory variables for socioeconomic aspects of households in the Chico Mendes Extractive Reserve 1 Socioeconomic Indicators: Household self-definition (%) A Household background variables No Rubber Tapper Rubber Tapper Total First occupant of the homestead ** No 14.6 85.4 100 Yes 38.5 61.5 100 Years of residence < 10 20.9 79.1 100 10 < 20 21.3 78.7 100 > 20 21.2 78.8 100 Age household head (year) < 30 17.9 82.1 100 30 < 50 21.3 78.7 100 > 50 24.2 75.8 100 Schooling of households head None 16.7 83.3 100 1 < 4 22.0 78.0 100 4 < 8 25.0 75.0 100 Previous residence In the reserve 18.6 81.4 100 Out the reserve 25.0 75.0 100 Relatives living in this seringal No 11.1 88.9 100 Yes 21.9 78.1 100 B Assets variables Household own a tv No 20.6 79.4 100 Yes 20.0 80.0 100 Household own gas No 18.3 81.7 100 Yes 22.7 77.3 100 Household own a refrigerator No 19.7 80.3 100 Yes 50.0 50.0 100 Manioc processor house No 18.9 81.1 100 Yes 23.2 76.8 100 Household own a solar panel No 21.0 79.0 100 Yes 18.2 81.8 100 Household own a chainsaw No 20.6 79.4 100 Yes 20.4 79.6 100 Household has a wood house No 13. 86.5 100 Yes 23.6 76.4 100 Household own a phone + No 22.2 77.8 100 Yes 0.0 100.0 100 + p < .10; p< .05; ** p < .01

PAGE 191

191 Table A-2. Analysis of relations between outcome and explanatory variables for institutional context aspects of households in th e Chico Mendes Extractive Reserve 2 Institutional context indicators Household self-definition (%) Variables: No Rubber Tapper Rubber Tapper Total Assoc. at municipal level No 20.0 80.0 100 Yes 21.3 78.7 100 Assoc. at community level + No 13.0 87.0 100 Yes 26.1 73.9 100 Assoc. w/ producers cooperative* No 24.0 76.0 100 Yes 4.8 95.2 100 Rural workers syndicate No 18.2 81.8 100 Yes 22.8 77.2 100 Participation in empate No 21.9 78.1 100 Yes 22.2 77.8 100 Participation in multiro No 28.1 71.9 100 Yes 20.2 79.8 100 Household received gov. credit No 20.5 79.5 100 Yes 20.8 79.2 100 + p < .10; p< .05; ** p < .01

PAGE 192

192 Table A-3. Analysis of relations between out come and explanatory variables for land-use aspects of households in the Chico Mendes Extractive Reserve 3 Land-use indicators Household self-definition (%) A Forest extractivism variables No Rubber Tapper Rubber Tapper Total Rubber tapping ** Yes 10.0 90.0 100 No 28.6 71.4 100 Total area ** none 58.8 41.2 100 100 < 300 22.0 78.0 100 300 < 600 12.5 87.5 100 > 600 9.7 90.3 100 Total rubber production none 27.7 72.3 100 < 300 14.3 85.7 100 > 300 4.0 96.0 100 Total number of brazil nut trees None 40.0 60.0 100 < 100 19.3 80.7 100 100 < 300 9.7 90.3 100 > 300 12.5 87.5 100 Total brazil nut production + none 29.3 70.7 100 < 1000 23.1 76.9 100 > 1000 12.7 87.3 100 B Annual crops sold variables Rice production sold none 19.0 81.0 100 < 1000 23.6 76.4 100 > 1000 25.0 75.0 100 Corn production sold none 19.8 80.2 100 < 1000 33.3 66.7 100 > 1000 0.0 100.0 100 Bean production sold none 18.6 81.4 100 < 1000 24.1 75.9 100 > 1000 33.3 66.7 100 Manioc production sold none 23.9 76.1 100 < 1000 12.5 87.5 100 > 1000 0.0 100.0 100 C Cattle and pasture variables Raise cattle No 15.8 84.2 100 Yes 21.9 78.1 100 Year household started to raise cattle None 15.8 84.2 100 < 10 16.0 84.0 100 10 < 20 28.2 71.8 100 > 20 33.3 66.7 100

PAGE 193

193 Table A-3. Continued 3 Land-use indicators Household self-definition (%) C Cattle and pasture variables Total number of head of cattle None 15.8 84.2 100 < 20 19.6 80.4 100 > 20 29.0 71.0 100 Total number of cattle head sold None 21.3 78.7 100 < 10 19.2 80.8 100 > 10 37.5 62.5 100 Year household started to make pasture none 20.0 80.0 100 < 10 19.3 80.7 100 10 < 20 24.4 75.6 100 > 20 23.1 76.9 100 Total size of pasture area (ha) none 20.0 80.0 100 < 10 19.2 80.8 100 10 < 20 23.7 76.3 100 > 20 40.0 60.0 100 Plan for # of cattle in 10 years none 0.0 100.0 100 < 50 21.6 78.4 100 50 < 100 18.2 81.8 100 > 100 20.6 79.4 100 + p < .10; p< .05; ** p < .01 Table A-4. Analysis of relations between outco me and explanatory variables for location of households in the Chico Mendes Extractive Reserve 4 Location indicator Household self-definition (%) Variable: No Rubber Tapper Rubber Tapper Total Household at municipalities Xapuri 3.8 96.2 100 Brasilia 17.0 83.0 100 Assis Brasil 30.9 69.1 100 + p < .10; p< .05; ** p < .01

PAGE 194

194 Table A-5. Analysis of relations between outco me and explanatory variables for households knowledge about land-use regulation in the Chico Mendes Extractive Reserve 5 Land use rules indicators Household self-definition (%) Variables: No Rubber Tapper Rubber Tapper Total Knows the utilization plan + Yes 17.9 82.1 100 No 33.3 66.7 100 Know the 10% deforestation rule Yes 21.0 79.0 100 No 22.2 77.8 100 Know the 5% pasture rule Yes 17.0 83.0 100 No 32.5 67.5 100 + p < .10; p< .05; ** p < .01

PAGE 195


PAGE 196


PAGE 197


PAGE 198


PAGE 199


PAGE 200


PAGE 201


PAGE 202


PAGE 203


PAGE 204


PAGE 205


PAGE 206


PAGE 207


PAGE 208


PAGE 209


PAGE 210


PAGE 211


PAGE 212

212 LIST OF REFERENCES Achard, F., Eva, H. D., Stibig, H. J., Mayaux, P., Gallego, J., Richards, T. & Malingreau, J. P. (2002) Determination of deforestation rate s of the worlds humid tropical forests. Science 297: 999-1002. Adams, W. M. 2004 Against Extinction: The Story of Conservation London: Earthscan. Agrawal, A. & Gibson, C. C. (2001) Communities and the environment: ethnicity, gender, and the state in community-based conservation Brunswick, N.J.: Rutgers University Press Alcorn, J. (1992) Indigenous agroforestry strategies meeting farmers needs. In: Alternatives to Deforestation: Steps Toward Sustai nable Use of the Amazon Rain Forest, ed. A. A. Anderson, pp. 141-148. New York, NY: Co lumbia University Press. Alcorn, J. B. & Lynch, O. J. (1992) Remnant Forests. BioScience 42: 397-402. Aldrich, S. P., Walker, R. T., Arima, E. Y., Ca ldas, M. M., Browder, J. O. & Perz, S. (2006) Land-cover and land-use change in the Brazi lian Amazon: Smallholders, ranchers, and frontier stratification. Economic Geography 82: 265-288. Allegretti, M. H. (1989) Reservas extrativistas: uma proposta de desenvolvimento da floresta amaznica. Par Desenvolvimento 25: 2-29. Allegretti, M. H. (1990) Extrac tive reserves: An alte rnative for reconciling development and environmental conservation in Amazonia. In: Alternatives to Deforestation: Steps toward Sustainable Use of the Amazonia Rain Forest ed. A. B. Anderson, pp. 252-264. New York, NY: Columbia University Press. Allegretti, M. H. (1994) Rese rvas extrativistas: Parme tros para uma Poltica de Desenvolvimento Sustentvel na Amaznia. In: O Destino da Floresta: Reservas Extrativistas e desenvolvime nto Sustentvel na Amaznia, ed. R. Arndt, pp. 17-47. Rio de Janeiro, RJ: Dumar. Allegretti, M. H. (1995a) The Amaz on And Extracting Activities. In: Brazilian Perspectives on Sustainable Development of the Amazon Region, ed. M. C. Godt & I. Sachs, pp. 157-174. Paris: UNESCO. Allegretti, M. H. (1995b) Policies for the use of renewable natural resources: The Amazonian region and extractive activities. In: Extractivism in the Brazilian Amazon: Perspectives on Regional Development, ed. M. Clsener-Godt & I. Sachs, pp. 14-33. MAB Digest Nr. 18, Paris: MAB-UNESCO. Allegretti, M. H. (2002) A C onstruo Social de Polticas Ambientais: Chico Mendes e o Movimento dos Seringueiros. Doctoral Dissertation, Centro de Desenvolvimento Sustentvel/Universidade de Braslia, Braslia, Brazil.

PAGE 213

213 Allegretti, M. H. & Schmink, M. forthcomi ng When Social Movement Proposals Become Policy: Experiments in Sustainable De velopment in Brazilian Amazonia. In Rural Social Movements in Latin America: Organizing for Sustainable Livelihoods (ed. C. D. Deere & S. R. Frederick). Gainesville, FL: University Press of Florida. Allegretti, M. H. & Schwartzman, S. (1989) Ex tractive reserves: a sustainable development alternative for Amazonia: Repor t to the WWF, Project US-478. Almeida, A. W. B. (1995) Carajs: a guerra dos mapas Belm, Par: Editora Supercores. Almeida, M. B. d. (2002) The Politics of Amaz onian Conservation: The Struggles of Rubber Tappers. The Journal of Latin American Anthropology 7: 170-219. Almeida, O. (ed.) (1996) A Evoluo da Fronteira Am aznica: Oportunidades para o Desenvolvimento Sustentvel. Porto Alegre: Edies Caravela. Amazon Press. (2001) Ibama quer transformar 10% da Amaznia em reservas: Report [www document]. URL www.amazonpress.com.br Am igos da Terra. (2005) Incendio Ameaa a Rese rva Chico Mendes: Report [www document]. URL www.amazonia.org.br Anderson, A. B., ed (1990) Alternatives to Deforestation: St eps Toward Sustainable Use of the Amazon Rain Forest Ne w York: Columbia University Press. Anderson, A. B. (1992a) Land-use strategies for successful extractive economies in Amazonia. Advances in Economic Botany 9: 67-77. Anderson, A. B. (1992b) Land-Use Strategies fo r Successful Extractive Economies in Amazonia. In: Non-Timber Products From Tropical Fore sts: Evaluation of a Conservation and Development Strategy, ed. D. C. Nepstad & S. Schwartzman, pp. 67-78. New York: The New York Botanical Garden. Anderson, A. B. (1994) Extrativismo Ve getal e Reservas Extrativistas. In: O Destino da Floresta: Reservas Extrativistas e de senvolvimento Sustentvel na Amaznia, ed. R. Arndt, pp. 227-247. Rio de Ja neiro, Brazil: Dumar. Ankersen, T. & Barnes, G. (2004) Inside the polygon: Emerging community tenure systems and forest resource extraction. In: Working Forests in the Neot ropics: Conservation Through Sustainable Management,? ed. D. J. Zarin, J. R. R. Alavalapati, F. E. Putz & M. Schmink, pp. 156-177. New York: Columbia University Press. Arima, E., Barreto, P. & Brito, M. (2006) Cattle Ranching in the Amazon: trends and implications for enviromental conservation Belm: Imazon. Asner, G. P., Knapp, D. E., Broadbent, E. N., Olivei ra, P. J. C., Keller, M. & Silva, J. N. (2005) Selective logging in the Brazilian Amazon. Science 310: 480-482.

PAGE 214

214 Bakx, K. (1988) From Proletarian to Peasant: Rural Transformation in the State of Acre, 18701986. The Journal of Development Studies 24 ; 141-160. Barbosa, L. C. (2000) The Brazilian Amazon Rainforest: Global Ecopolitics, Development, and Democracy New York: University Press of America. Barbosa, R. I. & Fearnside, P. M. (1996) Pa sture burning in Amazonia: Dynamics of residual biomass and the storage and release of aboveground carbon. Journal of Geophysical Research-Atmospheres 101 : 25847-25857. Barnham, B. L. & Coomes, O. T. (1996) Prosperitys Promise:The Amazon Rubber Boom and Distorted Economic Development. Boulder: Westview Press. Barreto, P., Souza, J., Carlos Noguern, R., Anderson, A. & Salamo, R. (2005) Presso humana na floresta Amaznica Brasileira. Belm: WRI-Imazon. Becker, B. K., Miranda, M. & Machado, L. O. (1990) Fronteira amaznica: questes sobre a gesto do territrio Braslia, DF: Editora UnB. Becker, M. (2004) Peasant Identity, Wo rker Identity: Multiple Modes of Rural Consciousness in Highland Ecuador. Estudios Interdisciplinarios de America Latina y el Caribe 15, 1-26. Berkes, F. (2004) Rethinking co mmunity-based conservation. Conservation Biology 18 : 621630. Brandon, K. (1998) Perils to parks: Th e social context of threats. In: In Parks in peril: People, politics, and protected areas, ed. K. Brandon, K. H. Redford & S. E. Sanderson, pp. 415440. Washington, DC: Island Press. Branford, S. & Glock, O. (1985) The Last Frontier: Fighting Over Land in the Amazon London: Zed Books. Brito, M. F. (2005) Parasos perdid os: (o)eco [www document]. URL http://arruda.rits.org.br Brockington, D., Igoe, J. & Schm idt-Soltau, K. (2006) Conservation, hum an rights, and poverty reduction. Conservation Biology 20: 250-252. Brondizio, E. S. & al., e. (2001) The Colonist Footprint: towards a conceptual framework of land use and deforestation trajectories among sm all farmers in Frontier Amazonia. In: Patterns and Processes of Land Use and Forest Change in the Amazon, ed. C. Wood & e. al.. Gainesville: University of Florida Press. Brondizio, E. S., E. Moran, P. Mausel & Wu, a. Y. (1996) Land Cover in the Amazon Estuary: linking of thematic mapper with botanical and historical data. Photogrammatic Engeneering and Remote Sensing 62: 921-929. Brondizio, E. S. & Siqueira, a. A. (1997) From Extractivists to Fore st Farmers: Changing Concepts of Caboclo Agroforestry in the Amazon Estuary. Economic Anthropology 18 : 233-279.

PAGE 215

215 Brondizio, E. S. e. a. (1999) The Colonist Foot print: towards a conceptual framework of land use and deforestation trajectories among small farmers in frontier Amaznia. In: Partterns and Processes of Land Use and Forest Changes in the Amazon ed. C. W. e. al.. Gainesville: Center for Latin American Studies, University of Florida. Brosius, J. P. (2004) Indigenous peoples and protected areas at the World Parks Congress. Conservation Biology 18 : 609-612. Browder, J. O. (1990a) Beyond the limits of extrac tion: tropical forest alternatives to extractive reserves. In: The Rainforest Harvest vol. 40 (ed.?). London, UK: ? Browder, J. O. (1990b) Extractive re serves will not save the tropics. BioScience 40 : 626. Browder, J. O. (1992a) The limits of extractivis m: tropical forest strategies beyond extractive reserves. BioScience 42 : 174-182. Browder, J. O. (1992b) Social and Economic Co ntstraints on the Development of MarketOriented Extractive Reserves in Amazon Rain Forests. In: Non-Timber Products From Tropical Forests: Evaluation of a C onservation and Development Strategy, ed. D. C. Nepstad & S. Schwartzman, pp. 33-43. New York: The New York Botanical Garden. Brown, I. F. (2004) Significant results of LBA-MAP. Presentation. 20 January 2004 Rio Branco, Acre. Brown, K. & Resende, S. (2000a) Environmentalists, Rubber Tappers and Empowerment: The Politics and Economics of Extractive Reserves. Development and Change 31: 201-27. Brown, K. & Resende, S. (2000b) The institutional architechture of Extractive Reserves. The Geographical Journal 166 : 35-48. Bryant, R. L. & Bailey, S. (1997) Third World Political Ecology. London: Routledge. Bunker, D. E., DeClerck, F., Bradford, J. C., Colwell, R. K., Perfecto, I., Phillips, O. L., Sankaran, M. & Naeem, S. (2005) Species Loss and Aboveground Carbon Storage in a Tropical Forest. Science 310: 1029-1031. Bunker, S. G. (1985) Underdeveloping the Amazon: Extraction, unequal exchange, and the failure of the modern state Chicago, IL: University of Illinois Press. Buschbacher, R. (1986) Tropical defore station and pasture development. Bioscience 36: 22-28. Calaa, M. (1993) Violncia e Resistncia: O Movimento dos Seringueiros de Xapuri e a Proposta de Reserva Extrativista. Doctoral Dissertation, Universidade Estadual Paulista, Rio Claro, Brazil. Caldas, M., Walker, R., Arima, E. Y., Perz, S., Aldrich, S. & Simmons C. (2007) Theorizing Land Cover and Land Use Change: The Peasan t Economy of Amazonian Deforestation. Annals of the Associatio n of American Geographers 97: 86.

PAGE 216

216 Campbell, C. (1996) Forest, Field and Factory: Changing Livelihood Strategies in two Extractive Reserves in the Brazilian Amazon. Doctor al Dissertation, Univ ersity of Florida Gainesville, USA. Campos, M. T. & Nepstad, D. C. (2006) Sma llholders, the Amazon's new conservationists. Conservation Biology 20 : 1553-1556. Cardoso, R. C. L. (1992) Popular Movements in the Context of the Consolidation of Democracy in Brazil. In: The Making of Social Movements in La tin America: Identity, Strategy, and Democrac, ed. A. Escobar & S. E. Alvarez, pp. 291-302. Boulder, Co.: Westview Press. Cavalcanti, F. d. S. (2002) A Poltica Ambiental da Amaznia: Um Estudo Sobre as Reservas Extrativistas. Doctoral Dissertation, Univ ersidade Estadual de Campinas, Campinas, Brazil. Cernea, M. M. & Schmidt-Soltau, K. (2006) Povert y risks and national park s: Policy issues in conservation and resettlement. World Development 34: 1808-1830. Chapin, M. (2004) A Challenge to Conservationists. World Watch Institute November/December. Clay, J. (1992) Some General Principles a nd Strategies for Developing Markets in North America and Europe for Non-timber Forest Products: Lessons from Cultural Survival Enterprises, 1989-1990. In: Non-Timber Products From Tropi cal Forests: Evaluation of a Conservation and Development Strategy, ed. D. C. Nepstad & S. Schwartzman, pp. 101-106. New York: The New York Botanical Garden. Clement, C. R., McCann, J. & Smith, N. ( 2003) Agrobiodiversity in Amazonia and its Relationship with Dark Earth. In: Amazonian Dark Earths: Origin, Properties, Management ed. J. Lehmann, D. C. Kern, B. Glaser & W. I. Woods, pp. 159-178. Kluwer: Dordrecht. CNS. (1985) Primeiro Encontro Nacional de Seringueiros da Amazonia. Braslia, DF: Conselho Nacional dos Seringueiros. CNS. (1992) Levantamento Socio-Economico das Reservas do Vale do Acre Rio Branco, AC: Conselho Nacional dos Seringueiros. Colho, E. M. (1982) Acre: O Ciclo da Bo rracha (1903-1945). Doct oral Dissertation, Universidade Federal Fluminense, Rio de Janeiro, Brazil. Colchester, M. (2000) Self-determination or e nvironmental determinism for indigenous peoples in tropical forest conservation. Conservation Biology 14: 1365-1367. Corra, L. M. (1967) A borracha da Amaznia e a II Guerra Mundial Manuas, AM: Governo do Estado do Amazonas.

PAGE 217

217 Cotta, J. N., Kainer, K. A., Wadt, L. H. O. & Staudhammer, C. L. (2008) Shifting cultivation effects on Brazil nut (Berthol letia excelsa) regeneration. Forest Ecology and Management 256, 28-35. Cunha, E. d. (1946) Margem da Histria Porto: Livraria Lello & Irmo Editores. Daly, D. J. (1990) Extractive Reserves: A Great New Hope. Garden 14: 14-32. Dean, W. (1987) Brazil and the struggle for rubber: A study in environmental history Cambridge, MA: Cambridge University Press. Ehringhaus, C. (2005) Post-Victory-Dilemmas: La nd Use, Development, and Social Movement in Amazonian Extractive Reserve. Doctoral Dissertation, Yale University, New Haven, USA. Euler, A., Millikan, B., Brito, E. B., Cardoso, I. B., Leroy, J. P. L., Caminha, L., Hargreaves, M. I., Motta, R. V., Cunha, S., Matias, S., Alve s, V. & Glass, V. (2008) A Devastao das Unidades de Conservao e Terras Indgenas no Esta do de Rondnia. Rondnia: Grupo de Trabalho Amaznico GTA. Faminow, M. D. (1998) Cattle, deforestation, and development in the Amazon: an economic, agronomic, and environmental perspective Wallingford, Oxon, UK: CAB International. Fearnside, P. M. (1986) Human carrying capacity of the Brazilian rain-forest. New York: Columbia University Press. Fearnside, P. M. (1989) Extractiv e reserves in Brazilian Amazonia. BioScience 39 : 387-393. Fearnside, P. M. (1992) Reservas Extrativ istas: uma estratgia de uso sustentado. Cincia Hoje 14: 15-17. Fearnside, P. M. (2001a) Land-tenure issues as f actors in environmental de struction in Brazilian Amazonia: The case of Southern Para. World Development 29: 1361-1372. Fearnside, P. M. (2001b) Soybean cultivation as a threat to the environment in Brazil. Environmental Conservation 28: 23-38. Fearnside, P. M. & Barbosa, R. I. (1998) Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. Forest Ecology and Management 108: 147-166. Feitosa, M. L. (1995) Chico Me ndes Extractive Reserve. In: Extractive Reserves, ed. M. R. Murrieta & R. P. Rueda, pp. 69-76. Cambridge: IUCN. Ferraz, G., Russell, G. J., Stouffer, P. C., Bie rregaard, R. O., Pimm, S. L. & Lovejoy, T. E. (2003) Rates of species loss from Amazonian forest fragments. PNAS 100: 14069-14073.

PAGE 218

218 FUNDASUR (1996) Pre-Seleo das Iniciativas Prom issoras de Manejo Florestal na Amaznia, Uruar, Brasil. Unpublished report. Urua r, PA: Fundao para o Desenvolvimento Sustentvel de Uruar. Futemma, C. & Brondizio, E. S. (2003) Land refo rm and land-use changes in the lower Amazon: implications for agricu ltural intensification. Human Ecology 31: 369-402. FVPP (2000) Projeto Consolidao da Produo Familiar Rural e Conteno dos Desmatamentos na Transamaznica e Baixo Xingu Altamira, PA: Fundao Viver Produzir e Preservar. Gale, T. (1998) The assault co ntinues: State of th e Amazon since Chico Mendes death 10 years ago. International Wildlife 1998 (Nov.) : 1-6. Gash, J. H. C., Tani, M. & Bruijnzeel, L. A. (2005) Controls of evaporation in lowland tropical rainforest. In Forests, Water and People in the Humid Tropics. (ed. M. Bonell & L. A. Bruijnzeel), pp. 287-313. Ca mbridge: Cambridge University Press. Ghimire, K. B. (1994) Parks and People Live lihood Issues in National-Parks Management in Thailand and Madagascar. Development and Change 25: 195-229. Gibson, C. C., McKean, M. A. & Ostrom, E., eds (2000) People and forests: communities, institutions, and governance Cambridge, Mass: MIT Press. Godoy, R., Lubowski, R. & Markandya, A. (1993) A method for the economic valuation of nontimber tropical forest products. Economic Botany 47: 220-233. Godoy, R. A. (2000) Valuation and consumption and sale of forest goods from a Central American rainforest. Nature 406: 62-63. Godoy, R. A., David Wilkie & Franks, a. J. (1997) The Effects of Markets on Neotropical Deforestation: A Comparative Study of Four Amerindian Societies. Current Anthropology 38: 875-878. Godoy, R. A., et al. (1997) Household determin ants of deforestation of deforestation by Amerindians in Honduras. World Development 25: 977-87. Goeschl, T. & Igliori, D. C. (2004) Reconc iling conservation and development: A dynamic hotelling model of ex tractive reserves. Land Economics 80: 340-354. Gomes, C. V. A. (2001) Dynamics of Land Use in an Amazonian Extractive Reserve: The Case of the Chico Mendes Reserve Acre, Brazil. Master Thesis, University of Florida, Gainesville, USA. Gomes, C. V. A. (2004) Cattle Ranching Expa nsion Among Rubber Tapper Communities in the Chico Mendes Extractive Reserve in the Sout hwestern Brazilian Amazonia. Rio Branco, AC: Report for WWF-Brazil. November 2004.

PAGE 219

219 Gomes, C. V. A. (2005) The Rubber Tappers Ra pidly Transitioning Land Use Strategies: Cattle Ranching Expansion in the Chico Mendes Extr active Reserve in Southwestern Brazilian Amazonia. In Global Environmental Change, Globalizat ion and International Security: New Challenges for the 21st Century. The 6th Open Meeting of the Human Dimensions of Global Environmental Change Research Community Bonn: University of Bonn. Gomez-Pompa, A. & Kaus, A. (1992) Traditional management of tropical forests in Mexico. In: Alternatives to Deforestati on: Steps toward a Sustainable Use of the Amazonian Rain Forest, ed. A. A. Anderson, pp. 43-64. New Yor k, NY: Colombia University Press. Government of Acre (2000a) Zoneamento ecolgico-econmico do Acre: Recursos Naturais e Meio Ambiente, Fase I, Volume I Rio Branco, Acre, Brazil: Governo do Acre Secretaria de Estado de Cincia, Tecnologia e Meio Ambiente. Government of Acre (2000b) Zoneamento ecolgico-econmico do Acre: Aspectos Socioeconmicos e Ocupao Territorial, Fase I, Volume II Rio Branco, Acre, Brazil: Governo do Acre Secretaria de Estado de Cincia, Tecnologia e Meio Ambiente. Government of Acre (2005) Acre: The State of Florest ania in the Brazilian Amazon Rio Branco, Acre, Brazil: Governo do Acre. Government of Acre (2006) Zoneamento ecolgico-econmico do Acre: Fase II, 1:250.000 Rio Branco, Acre, Brazil: Governo do Acre Secretaria de Estado de Meio Ambiente e Recuros Naturais. Green, G. M., C.M. Schweik, and M. Hanson. (2002) Radiometric calibration of LANDSAT multi-spectral scanner and thematic mapper images: Guideline for the global change community. CIPEC Working Paper. CPW-02-03 Bloomington: CIPEC. Green, G. M., Schweik, C. M. & Randolph, J. C. (2005) Retrieving land-cover change information from LANDSAT satellite images by minimizing other sources of reflectance variability. In: Seeing the Forest and the Trees, ed. E. F. Moran & E. Ostrom, pp. 131160. Cambridge: MIT Press. Grimes, A., Loomis, S., Jahnige, P., Burnham, M., Onthank, K., Alarcon, R., Cuenca, W. P., Martinez, C. C., Neill, D., Balick, M., Bennett, B. & Mendelsohn, R. (1994) Valuing the Rain-Forest the Economic Value of N ontimber Forest Products in Ecuador. Ambio 23: 405-410. Guimares, R. (1991) Ecopolitics of development in the third world Boulder, CO: L. Reinner Publishers. Hall, A. (1997) Sustaining Amazonia: Grassroots action for productive conservation Manchester: Manchester University Press. Harris, M. (2007) Introduction: ways of knowing. In: Ways of knowing: anthropological approaches to crafting experience and knowledge, ed M. Harris, pp. 1-26. W iltshire: Berghahn Books.

PAGE 220

220 Hecht, S. B. (1985) Environmental, Developm ent and Politics: Capital Accumulation and the Livestock Sector in Eastern Amazonia. World Development 13: 663-684. Hecht, S. B. & Cockburn, A. (1989) The fate of the forest: Developers, destroyers, and defenders of the Amazon London: Verso. Hecht, S. B. & Cockburn, A. (1990) The fate of the forest: Developers, destroyers, and defenders of the Amazon London: Verso. Holt, F. L. (2005) The catch-22 of conservati on: Indigenous peoples, bi ologists, and cultural change. Human Ecology 33: 199-215. Homma, A. K. O. (1989) Reservas extrativista s: uma opo de Desenvolvimento vivel para a Amaznia? Par Desenvolvimento 25: 38-48. Homma, A. K. O. (1992) The Dyna mics of Extraction in Amazonia: A Historical Perspective. In: Non-Timber Products From Tropical Forests: Evaluation of a Conservation and Development Strategy, ed. D. C. Nepstad, pp. 23-33. New York: The New York Botanical Garden. Homma, A. K. O. (1993) Extrativismo Vegetal na Am aznia: limites e oportunidades. Brasilia, DF: Embrapa-SPI. Homma, A. K. O. (1994) Plant extractivism in the Amazon: Limitations and possibilities. In: Extractivism in the Brazilian Amazon: Perspectives on Regional Development, ed. M. Clsener-Godt & I. Sachs, pp. 34-57. MA B Digest Nr. 18, Paris: MAB-UNESCO. Houghton, R. A. (2005) Tropical deforestation as a source of greenhouse gas emissions. In: Tropical Deforestati on and Climate Change, ed. P. Moutinho & S. Schwartzman, pp. 1321. Belm, PA.: Instituto de Pe squisa Ambiental da Amaznia. Humphries, S. S. & Kainer, K. A. (2006) Local perceptions of forest certification for community-based enterprises. Forest Ecology and Management 235: 30-43. Hutton, J., Adams, W. M. & Murombedzi, J. C. (2005) Back to the barriers? Changing narratives in biodiversity conservation. Forum for Development Studies 32: 341-370. IBAMA/CNPT (1999) Projeto RESEX: Um Futuro Sustentvel para a Amaznia Brasilia: Ministrio do Meio Ambiente/Instituto Br asileiro de Meio Am biente/Centro nacinonal para as Populaoes Tradicionais. IBGE (1973) Censo Demogrfico do Brasil 1970 (D emographic Census of Brazil 1970). Rio de Janeiro: Instituto Brasileiro de Geografia e Estatstica. IBGE (1996) Censo Demogrfico do Brasil 1991 (D emographic Census of Brazil 1991). Microdata Sampl e. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatstica.

PAGE 221

221 IBGE (1998) Censo Agropecurio 1995/96 (Agricultural Census of Brazil-1996/96) vol. 2007. Rio de Janeiro: Instituto Brasilei ro de Geografia e Estatstica. IBGE (2007) Censo Agropecurio 2007 (Agricul tural Census of Brazil 1997) vol. 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatstica. IDB (2002) Brazil: Acre Sustainable Development Program, (BR-0313) Loand Proposal, vol. 2007. Washington, DC: Inter-American Development Bank. IDESP (1990) Uruar Belm, PA: Instituto de De senvolvimento do Estado do Par. IPAM (2004) O Desenvolvimento que Queremos: Orde namento Territorial da BR-163, Baixo Amazonas, Transamaznica e Xing Belm, PA: Instituto de Pesquisa da Amaznia. ISA (2004) Relatrio Encontro BR-163 Sustentvel: Desa fios e sustentabilidade Socioambiental ao Longo do Eixo Cuiab-Santarm Sinop, Mato Grosso: Instituto Socioambiental. Kainer, K., Schmink, M., Leite, A. C. & Fadell, M. J. (2003) Experiments in Forest-Based Development in Western Amazoni. Society and Natural Resources 16: 869. Kainer, K. A., Duryea, M. L., de Macedo, N. C. & Williams, K. (1998) Brazil nut seedling establishment and autecology in extractive reserves of acre, Brazil. Ecological Applications 8: 397-410. Kainer, K. A., Wadt, L. H. O. & Staudhammer, C. L. (2007) Explaining variation in Brazil nut fruit production. Forest Ecology and Management 250: 244-255. Kramer, R. A., Van-Schaik, C. P. & Johnson, J., eds (1997) The last stand: Protected areas and the defense of tropical biodiversity New York: Oxford University Press Lambin, E. & Geist, H., eds (2006) Land-Use and Land-Cover Change: Local Processes to Global Impacts. Berlin: Springer. Laurance, W. F., Lovejoy, T. E., Vasconcelos, H. L., Bruna, E. M ., Didham, R. K., Stouffer, P. C., Gascon, C., Bierregaard, R. O., Laurance, S. G. & Sampaio, E. (2002) Ecosystem decay of Amazonian forest frag ments: A 22-year investigation. Conservation Biology 16: 605-618. Laurance, W. F., Vasconcelos, H. L. & Lovejoy, T. E. (2000) Forest loss and fragmentation in the Amazon: implications for wildlife conservation. Oryx 34: 39-45. Laurence, W. F. & Fearnside, P. M. ( 2002) Issues in Amazo nian development. Science 295: 1643-1643. Lima, R. C. A., Miranda, S. H. G. & Galli, F. (2005) Frebe aftosa: impactos sobre as exportaes brasileiras de carnes e o c ontexto mundial das barreiras sanitrias So Paulo: Icone/Cepea.

PAGE 222

222 Ludewigs, T. (2006) Land-Use Decision Making, Uncertainty and Effectiveness of Land Refor in Acre, Brazilian Amazon. Doctoral Disse rtation, Indiana University, Bloomington, USA. Mahar, D. J. (1979) Frontier development policy in Brazil: A study of Amazonia New York: Paeger Publishers. MAPA. (2005) Programa Nacional de Erradicao da Febre Aftosa: Relatrio Anual 2005 Braslia, DF: Governo Brasileiro Ministrio da Agricultura, Pecuria e Abastecimento. Martinello, P. (1988) A "Batalha da Borracha" na segunda guerra mundial e suas conseqncias para o vale amaznico. Cadernos UFAC Srie C : 13-365. Maxwell, K. (2003) Naked Tropics: Essays on Empire and Other Rogues. London: Routledge. Medina, G. (2006) Castanheiros x grileiros: conflito na ltima fronteira florestal do Par. Cincia Hoje 38: 60-63. Mertens, B., Poccard-Chapuis, R., Piketty, M. G. Lacques, A. E. & Venturieri, A. (2002) Crossing spatial analyses a nd livestock economics to unders tand deforestation processes in the Brazilian Amazon: the case of Sao Felix do Xingu in South Para. Agricultural Economics 27: 269-294. Michelotti, F. (2000) Beneficiam ento local da produo extrativ ista e agroflorestal: o caso da Cooperativa Agroextrativista de Xapuri. Novos Cadernos Naea 3: 17-44. MMA (2000) Sistema Nacional de Unidade s de Conservao SNUC Braslia, DF: Ministrio do Meio Ambiente. MMA (2002) Biodiversity and Forests of Brazil Braslia, DF: Ministrio do Meio Ambiente. Moegenburg, S. M. & Levey, D. J. (2002) Prospects for conserving biodiversity in Amazonian extractive reserves. Ecology Letters 5: 320-324. Moran, E. (1981) Developing the Amazon. Bloomington, IN: Indian a University Press. Moran, E. F. (2000) Human Adaptability: An Introducti on to Ecological Anthropology. Boulder: Westview Press. Moran, E. F. & Ostrom, E., eds (2005) Seeing the forest and the trees: human-environment interactions in forest ecosystems. Cambridge, Mass: Mit Press. Murrieta, J. R. & Rueda, R. P., eds (1995) Extractive Reserves Gland, Switzerland: IUCNCNPT-EU. Myers, N., Mittermeier, R. A., Mittermeier, C. G., Fonseca, G. A. B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature 403: 853-858.

PAGE 223

223 Nepstad, D., Carvalho, G., Barros, A. C., Alencar, A., Capobianco, J. P., Bishop, J., Moutinho, P., Lefebvre, P., Silva, U. L. & Prins, E. (20010 Road paving, fire regime feedbacks, and the future of Amazon forests. Forest Ecology and Management 154: 395-407. Nepstad, D., Moreira, A., Verissimo, A., Lefebvre, P., Schlesinger, P., Potter, C., Nobre, C., Setter, A., Krug, T., Barros, A. C., Alencar, A. & Pereira, J. R. (1998) Forest fire prediction and prevention in the Brazilian Amazon. Conservation Biology 12: 951-953. Nepstad, D., Schwartzman, S., Bamberger, B., Sant illi, M., Ray, D., Schlesinger, P., Lefebvre, P., Alencar, A., Prinz, E., Fiske, G. & Rolla, A. (2006a) Inhibition of Amazon deforestation and fire by parks and indigenous lands. Conservation Biology 20: 65-73. Nepstad, D. C., McGrath, D., Alencar, A., Barros, A. C., Carvalho, G., Santilli, M. & Vera Diaz, M. d. C. (2002) Frontier Governance in Amazonia. Science 295 : 629-631. Nepstad, D. C. & Schwartzman, S., eds (1992) Non-Timbert Products From Tropical Forests: Evaluation of a Conservation and Development Strategy Bronx, New York.: The New York Botanical Garden. Nepstad, D. C., Stickler, C. M. & Almeida, O. T. (2006b) Globalization of the Amazon soy and beef industries: Opportu nities for conservation. Conservation Biology 20 : 1595-1603. Nobre, C. A., Sellers, P. J. & Shukla, J. (1991) Amazonian deforestation and regional climate change. Journal of Climate Change 4: 957-988. Nugent, A. (1993) Amazonian caboclo society: an essay on invisibility and peasant economy Oxford: Berg. Nunes, J. R. P. (1991) Modernizao da agricultura pecuarizao e mudanas: o caso do Alto Purus. Rio Branco-Acre: Editora Tico-Tico LTDA. Oates, J. F. (1999) Myth and Reality in the Rain Forest : How Conservation Strategies Are Failing in West Africa. Berkeley: University of California Press. Oliveira, L. A. P. (1985) O Sertanejo, o Brabo e o Posseiro: cem anos de andanas da populao acreana Rio Branco, Acre: Governo do Estado do Acre. Olmos, F. (2006) As Reservas Extratavis tas so Unidades de Conservao?: (o)eco [www document]. URL http://arruda.rits.org.br Ostrom E. (1990) Governing the commons: the evolution of institutions for collective action Cambridge: Cambridge University Press. Pacheco, P. (2005) Populist and Capitalist Fron tiers in the Amazon: Di versity Dynamics of Agrarian and Land-Use Change Doctoral Dissertation, Clar k University, Worcester, USA. Palmer, C. (2002) An overview of environmental ethics. In: Environmental Ethics: An Antholog, ed. A. Light & H. Halston. Malden: Blackwell Publishers.

PAGE 224

224 Paula, E. A. (1991) Seringueiros e Sindicatos: um Povo da Floresta em Busca de Liberdade. Universidade Federal Rural do Rio de Janeiro, Rio de Janeiro, Brazil. Paula, E. A. (2005) (Des) envolvimento insustentvel na Amaznia Ocidental: dos missionrios do progresso aos mercadores da natureza Rio Branco-Acre: EDUFAC. Paula, E. A. & Silva, S. S., eds (2006) Trajetrias da luta camponesa na Amaznia-Acreana Rio Branco-Acre: EDUFAC. Peres, C. A. (2005) Why we need megareserves in Amazonia. Conservation Biology 19 : 728733. Perz, S. G. (2001a) From Sustainable Deve lopment to "Productive Conservation:"Forest Conservation Options and Agricultural In come and Assets in the Brazilian Amazon. Rural Sociology 66: 93-112. Perz, S. G. (2001b) Household demographic factors as life cycle determinan ts of land use in the Amazon. Population Research and Policy Review. 00: 1-28. Perz, S. G. (2002) The Changing Social Contex ts of Deforestation in the Brazilian Amazon. Social Sciences Quarterly 83: 35-52. Perz, S. G. (2003) Social determinants and la nd use correlates of ag ricultural technology adoption in a forest frontier: A case study in the Brazilian Amazon. Human Ecology 31: 133-165. Perz, S. G. (2004) Are agricultural production and forest conservation compatible? Agricultural diversity, agricultural incomes and primary forest cover amo ng small farm colonists in the Amazon. World Development 32: 957-977. Perz, S. G., Caldas, M. M., Arima, E. & Walker, R. J. (2007) Unofficial road building in the Amazon: Socioeconomic and biophysical explanations. Development and Change 38: 529-551. Perz, S. G. & Walker, R. T. (2002) Household life cycles and seconda ry forest cover among small farm colonists in the Amazon. World Development 30: 1009-1027. Perz, S. G., Walker, R. T. & Caldas, M. M. (2006) Beyond population and environment: Household demographic life cycles and la nd use allocation among small farms in the Amazon. Human Ecology 34: 829-849. Peters, C. M., Gentry, A. H. & Mendelsohn, R. O. (1989) Valuation of an Amazonian rainforest. Nature 339: 655-656. Pimm, S. L., Ayres, M., Balmford, A., Branch, G., Brandon, K. & others, a. (2001) Can We Defy Nature's End? Science 293 : 2207-2208. Rancy, C. M. D. (1992) Razes do Acre (1870-1912) Rio Branco: Editora Paim.

PAGE 225

225 Redford, K. H. & Brosius, J. P. (2006) Diversity and homogenization in the endgame. Global Environmental Change-Human and Policy Dimensions 16: 317-319. Redford, K. H. & Painter, M. (2006) Natural al liances between conserva tionists and indigenous peoples. In WCS Working Paper No. 25 Redford, K. H., Robinson, J. G. & Adams, W. M. (2006) Parks as shibboleths. Conservation Biology 20: 1-2. Redford, K. H. & Sanderson, S. E. ( 2000) Extracting humans from nature. Conservation Biology 14: 1362-1364. Rego, J. F. (1999) Amazonia: do Extrativismo ao Neoextrativismo. Ciencia Hoje 25: 147. Rgo, J. F., Costa Filho, O. S. & Braga, R. R. A. (2003) Anlise econmica dos sistemas de produo familiar rural da regio do Vale do Acre, 1996/1997 Rio Branco, Acre, Brazil: UFAC/SEBRAE/Ford Foundation. Revkin, A. (1990) The burning season: the murder of Chico Mendes and the fight for the Amazon rainforest : Houghton Mifflim. Ribeiro, B., Verssimo, A. & Pereira, K. (2005) O Avano do Desmatamento sobre as reas Protegidas de Rondnia. O Estado da Amaznia pp. 4. Belm, PA: IMAZON. Rockwell, C., Kainer, K. A., Marcondes, N. & Baraloto, C. (2007a) Ec ological limitations of reduced-impact logging at the smallholder scale. Forest Ecology and Management 238: 365-374. Rockwell, C. A., Kainer, K. A., Staudhammer, C. L. & Baraloto, C. (2007b) Future crop tree damage in a certified community forest in southwestern Amazonia. Forest Ecology and Management 242: 108-118. Rueda, B. L., Blake, R. W., Nicholson, C. F., Fox, D. G., Tedeschi, L. O., Pell, A. N., Fernandes, E. C. M., Valentim, J. F. & Carneiro, J. C. (2003) Production and economic potentials of cattle in pasture-based systems of the western Amazon region of Brazil. Journal of Animal Science 81: 2923-2937. Ruiz-Perez, M., Almeida, M., Dewi, S., Costa, E. M. L., Pantoja, M. C., Puntodewo, A., Postigo, A. D. & de Andrade, A. G. (2005) C onservation and development in Amazonian extractive reserves: The case of Alto Jurua. Ambio 34: 218-223. Rylands, A. B. & Brandon, K. (2005) Brazilian protected area. Conservation Biology 19: 612618. Salafsky, N., Dugelby, B. L. & Terborgh, J. W. (1993) Can extractive reserves save the rainforest? An ecological comparison of non-t imber forest product extraction systems in el Petn, Guatemala, and West Kalimantan, Indonesia. Conservation Biology 7: 39-52.

PAGE 226

226 Salati, E. & Nobre, C. A. (1991) Possible climate impacts of tropical deforestation. Climate Change 19: 177. Sanderson, S. & Redford, K. (2004) The defence of conservation is not an attack on the poor. Oryx 38: 146-147. Santos, R. (1980) Histria Econmica da Amaznia (1800) So Paulo, SP: T. A. Queiroz. Sassagawa, H. S. (1999) Tcnicas de Sensoriamento Remoto e Sistema de Informaes Geogrficas (SIG) para Estudo de Ocupao do Espao Fsico e dos Tipos Florestais da Reserva Extrativista Chico Mendes, Estado do Ac re. Master Thesis, In stituto Nacional de Pesquisas Espaciais, So Jos dos Campos, Brazil. Schmidt-Soltau, K. (2004) The costs of rain forest conservation: local responses towards integrated conservation and de velopment projects in Cameroon. Journal of Contemporary African Studies 22: 93-117. Schmidt-Soltau, K. & Brockington, D. (2007) Prot ected Areas and Resettlement: What Scope for Voluntary Relocation? World Development 35: 2182. Schmink, M. & Wood, C. (1984) Frontier Expansion in Amazonia Gainesville, FL: University of Florida Press. Schmink, M. & Wood, C. (1992) Contested Frontiers in Amazonia New York: Columbia University Press. Schwartzman, S. (1989) Extractive reserves: the rubber tappers' st rategy for sustainable use of the Amazon rain forest. In: Fragile Lands in Latin America: Strategies for Sustainable Development, ed. J. O. Browder, pp. 150-165. B oulder, CO: Westview Press. Schwartzman, S. (1991) Deforestation and popular re sistance in Acre: from local movement to global network. The Centennial Review 25: 397-422. Schwartzman, S. (1992) Land Distribution and the Social Costs of Frontier Development in Brazil: Social and Historical Context of Extractive Reserves. In: Non-Timber Products From Tropical Forests: Evaluation of a Conservation and Development Strategy, ed. D. C. Nepstad & S. Schwartzman, pp. 51-66. Ne w York: The New York Botanical Garden. Schwartzman, S., Moreira, A. & Nepstad, D. (2000a) Rethinking tropical forest conservation: Perils in parks. Conservation Biology 14: 1351-1357. Schwartzman, S., Nepstad, D. & Moreira, A. ( 2000b) Arguing tropical forest conservation: People versus parks. Conservation Biology 14 : 1370-1374. Schwartzman, S. & Zimmerman, B. (2005) Conserva tion alliance with indi genous peoples of the Amazon. Conservation Biology 19 : 712-727.

PAGE 227

227 Schwartzmann, S., Moreira, A. & Nepstad, D. (2000a ) Rethinking tropical conservation: peril in parks. Conservation Biology 14: 1351-1357. Schwartzmann, S., Nepstad, D. & Moreira, A. (2000b) Arguing tropical forest conservation: people versus parks. Conservation Biology 14: 1370-1374. Serro, E. A. S., Falesi, I. C., Veiga, J. B. & Neto, J. F. T. (1978) Productivity of pastures grown in low fertility soils of the forest areas of the Brazilian humid tropics Belm: Centro de Pesquisa Agropecuria do Tropico mido. Serro, E. A. S. & Homma, K. O. (1993) Brazil. In: Sustainable agriculture and the environment in the humid tropics, ed. N. R. C. (U.S.)., pp. 265-351. Washington, D.C: National Academy Press. Silva, A. F. (1990) Razes da ocupao recente das terras do Acre: movimento de capitais, especulao fundiria e disputa pela terra Rio Branco-Acre: Casa da Amaznia. Silva, M. (2005) The Brazilian Protected Areas Program. Conservation Biology 19: 608-611. Silva, S. S. (2003) Na fronteira agropecuria acreana Rio Branco-Acre: LEUPAGDEGEO/UFAC. Simmons, C. S., Walker, R. T., Arima, E. Y., Al drich, S. P. & Caldas, M. M. (2007) The amazon land war in the south of par. Annals of the Association of American Geographers 97: 567-592. Skole, D. & Tucker, a. C. (1993) Tropical de forestation and habitat fragmentation in the Amazon: satellite data from 1978 to 1988. Science 260: 1905-1910. Smeraldi, R. & May, P. (2008) O Reino do Gado: Uma Nova Fase na Pecuarizao da Amaznia Brasileira. So Paulo: Amigos da Terra-Amaznia Brasileira. Smith, N. (1982) Rainforest Corridors: The Transamazon Colonization Scheme Berkeley, CA: University of California Pres. Smith, N. (20010 Land Use Dynamics in the Amazon Estuary and Implications for Resource Management. Amazonia 16: 517-537. Smith, N. J. H., Serro, E. A. S., Alvim, P. & Falesi, I. C. (1995) Amazonia: Resiliency and dynamism of the land and its people. Tokyo, New York, Paris: United Nations University Press. Sobrinho, P. V. C. (19920 Capital e trabalho na Amaznia Ocid ental: contribuio histria social e das sindicais no Acre So Paulo: Cortez.

PAGE 228

228 Souza, F. K. A. (2006) Effectiveness of Extr active Reserves, Agro-Extractive Reserves and Colonist Settlements in Southwestern Amazonia: An Economic and Land Cover Comparison of Three Land Tenure Types in Acre Brazil. Master Thesis, University of Florida, Gainesville, USA. Steffen, W. L., Sanderson, A. & Tyson, P. D. (2004) Global Change and the Earth System: a Planet Under Pressure. Berlin: Springer. Stokes, C. E. (2000) The Amazon Bubble: World Rubber Monopoly Fort McKavett: Charles Stokes. Stone, S. S. (2003) From Tapping to Cutting Trees: Participation and Agency in two Community-Based Timber Management Projects in Acre, Brazil. Doctoral Dissertation, University of Florida, Gainesville, USA. Terborgh, J. (1999) Requiem for nature Washington, D.C.: Island Press. Terborgh, J. (2000) The fate of tropic al forests: a matter of stewardship. Conservation Biology 14: 1258-1361. Terborgh, J. & Peres, C. A. (2002) Th e problem of people in parks. In: Making Parks Work: Strategies for Preserving Tropical Nature, ed. J. Terborgh, C. Van Schaik, L. Davenport & M. Rao., pp. 307-319. Washington, D.C.: Island Press. Terborgh, J. & Van Schaik, C. (2002) Why the world needs parks. In: Making Parks Work: Strategies for Preserving Tropical Nature, ed. J. Terborgh, C. Van Schaik, L. Davenport & M. Rao., pp. 3-14. Washington, D.C.: Island Press. Tocantins, L. (1979) Formao Histrica do Acre Rio de Janeiro: Ci vilizao Brasileira. Toni, F. (1999) Relations on the Agricultural Frontier: The Struggle for Credit in the Transamaznica Region. Doctoral Dissertation, Un iversity of Florida, Gainesville, USA. Toni, F. (2003) Uruar: Pecuari zao na fronteira agrcola. In: Municpios e gesto florestal na Amaznia, ed. F. Toni & D. Kaimowitz, pp. 175 -217. Natal, Brasil: A.S. Editores. Toni, F. (2007) Dinmica da pecuria na regio de fr onteira agrcola do Alto Rio Acre Braslia, DF: UnB. Toni, F. & Souza, C. A. (2003) Xapuri: Movimentos sociais no poder. In: Municpios e gesto florestal na Amaznia, ed. F. Toni & D. Kaimowitz, pp. 337-372. Natal, Brazil: A.S. Editores. Tourrand, J. F. & Veiga, J. B., eds (2003) Viabilidade de sistemas agropecurios na agricultura familiar da Amaznia Belm: EMBRAPA Amaznia Oriental.

PAGE 229

229 Tucker, C. (2004) Aiming for sustainable commun ity forest management: The experiences of two communities in Mexico and Honduras. In: Working Forests in the NeoTropics: Conservation Through Sustainable Management?, ed. D. J. Zarin, J. R. R. Alavalapati, F. E. Putz & M. Schmink, pp. 178-199. New York: Columbia University Press. Vadjunec, J. M. (2007) The Role of Instituti ons in Land-Use/Cover-Change in the Chico Mendes Extractive Reserve, Acre, Brazil. Doctoral Dissertation, Clark University, Worcester, USA. Vadjunec, J. M., Gomes, C. V. A. & Ludewi gs, T. (nd) Land-use/land-cover change among rubber tappers in the Chico Mendes Extractive Reserve, Acre, Brazil. Valentim, J. F. & Andrade, C. M. S. (2005) Tropical kudzu (Pueraria pha seoloides): successful adoption in sustainable cattle production systems in the western Brazilian Amazon. Tropical Grasslands 39: 221-221. Valentim, J. F., S, C. P., Gomes, F. C. R. & Sa ntos, J. C. (2002) Tendncias da pecuria bovina no Acre entre 1970 e 2000. Boletim de Pesquisa e Desenvolvimento n 38 Rio Branco, Acre, Brazil: Embrapa Acre. Van Schaik, C. (2002) Anarchy and parks: dealing with political instability. In: Making Parks Work: Strategies for Preserving Tropical Nature, ed. J. Terborgh, C. Van Schaik, L. Davenport & M. Rao., pp. 352-363. Washington, D.C.: Island Press. Van Schaik, C. & Rijksen, H. D. (2002) Integr ated conservation and development projects: problems and potential. In: Making Parks Work: Strategies for Preserving Tropical Nature, ed. J. Terborgh, C. Van Schaik, L. Davenport & M. Rao., pp. 15-29. Washington, D.C.: Island Press. Veiga, J. B., Tourrand, J. F., Piketty, M. G., Poccard-Chapuis, R., Alves, A. M. & Thales, M. C. (2004) Expanso e trajetrias da pec uria na Amaznia, Par, Brasil. Braslia: UnB. Viana, J. (2004) Foreword. In: Working Forests in the Neot ropics: Conservation Through Sustainable Management?, ed. D. J. Zarin, J. R. R. Alavalapati, F. E. Putz & M. Schmink. New York: Columbia University Press. Wadt, L. H. O., Kainer, K. A. & Gomes-Silva, D. A. P. 92005) Popul ation structure and nut yield of a Bertholletia excelsa st and in Southwestern Amazonia. Forest Ecology and Management 211: 371-384. Wadt, L. H. O., Kainer, K. A., Staudhammer, C. L. & Serrano, R. O. P. (2008) Sustainable forest use Brazilian extractive rese rves: Natural regeneradon of Brazil nut in exploited populations. Biological Conservation 141: 332-346. Walker, R., Moran, E. & Anselin, L. (2000) Defo restation and cattle ranc hing in the Brazilian Amazon: External capital and household processes. World Development 28 : 683-699.

PAGE 230

230 Wallace, R. (2004) The Effects of Wealth and Markets on Rubber Tapper Use and Knowledge of Forest Resources in Acre, Brazil. Doctor al Dissertation, University of Florida, Gainesville, USA. Watson, R. T., Noble, I. R., Bolin, B., Ravindranath, N. H., Verardo, D. J. & Dokken, D. J., eds (2000) Land Use, Land-Use Change, and Forestry 2000. Cambridge, UK.: Cambridge University Press. WCED (1987) Our Common Future World Commission on Envir onment and Development. Oxford: Oxford University Press. Weinstein, B. (1983) The Amazon Rubber Boom, 1850 -1920 Stanford, CA: Stanford University Press. West, P. & Brockington, D. (2006) An anthropological perspective on some unexpected consequences of protected areas. Conservation Biology 20: 609-616. West, P., Igoe, J. & Brockington, D. (2006) Parks and peoples: The social impact of protected areas. Annual Review of Anthropology 35: 251-277. Wilkie, D. S., Morelli, G. A., Demmer, J., Star key, M., Telfer, P. & Steil, M. (2006) Parks and people: Assessing the human welfare effect s of establishing protected areas for biodiversity conservation. Conservation Biology 20 : 247-249. Wilshusen, P. R., Brechin, S. R., Fortwangler, C. L. & West, P. C. (2002) Reinventing a square wheel: Critique of a resurgent "protection paradigm'' in international biodiversity conservation. Society & Natural Resources 15: 17-40. Wood, C. & Porro, R. (ed.) (2002) Deforestation and Land Use in the Amazon Gainesville, FL: University of Florida Press. Wu, X. B., Thurow, T. L. & Whisenant, S. G. (2000) Fragmentation and changes in hydrologic function of tiger bush lands capes, southwest Niger. Journal of Ecology 88 : 790-800. Zarin, D. J., Alavalapati, J. R. R., Putz, F. E. & Schmink, M., eds (2004) Working Forests in the Neotropics: Conservation Th rough Sustainable Management? New York: Columbia University Press.

PAGE 231

231 BIOGRAPHICAL SKETCH Carlos Valrio A. Gom es was born in Ourm, State of Par, Brazil. As the son of an eastern Amazonian farmer, he did what most teen agers in this region did growing up: worked in the fields planting and harvesting crops, and rode horses and herded cattle while attending high school. In 1989, he moved to Rio Branco, State of Acre in the southwestern Amazon. He received his undergraduate degree in Geography in 1995 from the Federal University of Acre (UFAC). Post-graduation, he worked as an associ ate researcher with th e Zoobotanical Park of UFAC and as a consultant to the Extractive Reserves Project/Pilot Program for the Protection of the Brazilian Rainforests (PPG-7), contracted by the United Nations Development Program (UNDP). A Fulbright Fellowship (Amazon Basin Program) brought him to the University of Florida (UF) where he received a masters degree in Latin American studies with a concentration in tropical conservation and development in 2001. Hi s graduate research with forest dwellers in the southwestern Brazilian Amazon focuses on Extractive Reserves as a people-based conservation model. His work broadly a ddresses changing livelihood strategies among extractivist people, land-use/la nd-cover change, socioeconomic drivers of deforestation, and development policy for sustainabl e livelihoods in the region. In 2007, he worked as a researchertrainer with one of UFs part ner NGO (Pesacre) in Rio Branco-A cre to analyze a payment for environmental services program for smallholders in the Brazilian Amazon, as part of the ALFA (Alliance for Amazon and Atlantic Forest) Cons ortium program, financed by the USAID-Brazil Environment Program. Upon comple tion of his doctoral program in Geography at UF, Valrio intends to return to the Brazilian Amazon to co ntinue his career worki ng at the interface of applied research, sustainable development st rategies for forest dwellers, and regional environmental/development policy formation.