<%BANNER%>

Dynamics of Land Use In an Amazonian Extractive Reserve: The Case of the Chico Mendes Reserve In Acre, Brazil


PAGE 1

DYNAMICS OF LAND USE IN AN AMAZONIAN EXTRACTIVE RESERVE: THE CASE OF THE CHICO MENDES EXTRACTIVE RESERVE IN ACRE, BRAZIL By CARLOS VALRIO AGUIAR GOMES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2001

PAGE 2

Copyright 2001 by Carlos Valrio Aguiar Gomes

PAGE 3

To Leinha for everything

PAGE 4

iv ACKNOWLEDGMENTS This thesis could not have been completed without the help and support of a variety of friends, colleagues and mentors. First I thank Dr. Michael Binford, my supervisor for all his help, and constructive criticism, which helped me to stren gthen my intellectual independence. I am very thankful to the members of my committee: Dr. Marianne Schmink, Dr. Nigel Smith and Dr. Stephen Perz for their valuable comments, suggestions and guidance. I specially wish to express my deepest appreciation to Dr. Schmink for being a constant source of support and encouragement since I first thought about coming to study in the United States. I gratefully acknowledge the support of the Latin American Scholar Program for American Universities (LASPAU) Fulbright Commission for the fellowship that allowed my enrollment in the graduate school at the University of Florida. In LASPAU I want to thank Ms. Sonia Wallemberg for her wonderful assistance and friendship. I am also grateful to Ms. Rebecca Smith Murdock and M r. Amir Abbassi (from the Intensive English Language Institute of the University of North Texas) for the fellowship for English language training. Moreover, I would like to thank the World Wildlife Fund (WWF Brazil), the Tropical Conservation and Developme nt Program (TCD) at the University of Florida, and the Acre Environmental Agency (IMAC) for funding my research in Acre. My heartfelt gratitude goes to my friend and former advisor Dr. Irving Foster Brown at Federal University of Acre (UFAC). I thank him for his outstanding work

PAGE 5

v helping me and many other Amazonian students to make realities out of our dreams. Without his help, mine would have stayed asleep. Also, I want to thank my colleagues and friends who have given me a great deal of help and encourag ement to get to this point in my career: Andrea Alechandre, Silvia Brilhante, Hirome Sassagawa, Atila Cristina, William Flores, Eufran Amaral, Elsa Mendosa, Benedita Esteves and Monica de los Rios at the Zoobotanical Park at the UFAC; Arthur Leite, Juceir and Nuria Merched, at the Acre State Department of Forest and Extractivism; Jorge Rebouas at the Rubber Tapper National Council; Josemar Amorim, Manoel Feitosa and Jos Alberto at the National Center for Traditional Populations; Veronica Passos and Edgard de Deus at the Acre Environmental Agency. I also wish to express my deepest gratitude to Christiane Ehringhaus at Yale University for helping me very much throughout this entire work. I am flattered with the warmth from the rubber tapper families of the C hico Mendes reserve whom have taken me into their homes and shared their knowledge with me. I own them more than I can say. I also wish to thank Francisco C. dos Santos, my field assistant, for his great help. In addition, want to thank my friends at the U niversity of Florida for their support: especially Olendina Cavalcante, Stacey Shankle, Brandon Knox, David Salisbury, Ronaldo Weigand and Richard Wallace. Last but definitely not least, I would like to express my love and thanks to my parents and family for continuously providing me with love and understanding. All of us agree that eleven years ago this degree would have been unthinkable. I specially thank my nephew Sallon Saad and my niece Susy Leia for their love and sweet words, which always energize m e and remind me of what is important in life.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ iv LIST OF TABLES ................................ ................................ ................................ ........... viii LIST OF FIGURES ................................ ................................ ................................ ............ ix ABSTRACT ................................ ................................ ................................ ........................ xi CHAPTERS 1 INTRODUCTION ................................ ................................ ................................ ........... 1 1.1 Problem Statement ................................ ................................ ................................ ... 5 1.2 Central Questions ................................ ................................ ................................ ..... 8 1.3 Structure of the Study ................................ ................................ .............................. 9 2 DEBATE ABOUT EXTRACTIVE RESERVES ................................ .......................... 11 2.1 Brief History of the Amazon Rubber Boom ................................ .......................... 11 2.2 Evolution of the Extractivist Development Model in Amazonia ........................... 12 2.3 Analysis of the Extractive Reserve Literature ................................ ....................... 16 2.3.1 General Patterns and Trends in the Debate ................................ ..................... 17 2.3.2 Social Development Arguments ................................ ................................ ..... 20 2.3.3 Economic Arguments ................................ ................................ ...................... 21 2.3.4 Ecological Arguments ................................ ................................ ..................... 23 2.4 Achievements and Problems with Extractive Reseves ................................ .......... 24 3 STUDY SITES AND RESEARCH METHODOLOGY ................................ ............... 27 3.1 Study Site Selection ................................ ................................ ................................ 27 3.2 Survey Setting and Planning ................................ ................................ ................... 30 3.3 Survey Methodology ................................ ................................ ............................... 31 3.4 Other sources of Data ................................ ................................ .............................. 31 3.5 Data Processing Variables and Analysis ................................ ................................ 32 3.6 Environmental Characterization of the CMER ................................ ....................... 34 3.6.1 Climate ................................ ................................ ................................ ............. 34 3.6.2 Watershed/Hydrology ................................ ................................ ...................... 35 3.6.3 Geology ................................ ................................ ................................ ............ 35

PAGE 7

vii 3.6.4 Pedology (Soils) ................................ ................................ ............................... 37 3.6.5 Vegetation ................................ ................................ ................................ ........ 37 4 DEFORESTATION TRENDS IN THE EXTRACTIVE RESERVE ............................ 41 4.1 Introduction ................................ ................................ ................................ ............ 41 4.2 Deforestation and Land Use Trends in the State of Acre ................................ ...... 42 4.3 Deforestation rates in the CMER ................................ ................................ ........... 44 4.4 Deforestation at the Seringal Level ................................ ................................ ........ 48 5 HOUSEHOLD DETERMINANTS OF FOREST AND NONFOREST ACTIVITIES 54 5.1 Introduction ................................ ................................ ................................ ............ 54 5.2 Characterization of Rubber Tappers Settlements in the CMER .......................... 54 5.3 Seasonality The Rubber Tappers Year ................................ .............................. 55 5.4 Conceptualization of Outcome and Independent Variables ................................ ... 59 5.4.1 Outcome Variables ................................ ................................ .......................... 59 5.4.2 Independent Variables ................................ ................................ ..................... 63 5.5 Correlations Between Outcome and Independent Variables ................................ 66 5.6 Analysis of Multivariable Regression Models ................................ ....................... 69 5.7 Discussion ................................ ................................ ................................ .............. 73 6 DRIVERS OF LAND USE CHANGE IN THE CMER ................................ ................ 79 6.1 Introduction ................................ ................................ ................................ ............ 79 6.2 Population Dynamics in the CMER ................................ ................................ ....... 81 6.3 Population Dynamics in the St udy Seringais ................................ ......................... 86 6.4 Population Density and Deforestation in the Study Seringais ............................... 91 6.5 Settlement Dynamics in the Study Seringais ................................ ......................... 94 6.6 Changes in the Production System: Seringal Paraguau ................................ ....... 98 6.7 Pasture and Cattle Raising Expansion in the Study Seringais ............................. 102 6.8 Social Organization in the Reserve ................................ ................................ ...... 105 7 SUMMARY AND TRANSI TION ................................ ................................ .............. 112 APPENDICES A DATA CENSUS RELIABILITY AND UNCERTAINTY OF THE CMER ............. 121 B QUESTIONNAIRE USED IN CMER SURVEY ................................ ....................... 124 LIST OF REFERENCES ................................ ................................ ................................ 132 BIOGRAPHICAL SKETCH ................................ ................................ ........................... 141

PAGE 8

viii LIST OF TABLES Table Page 4 1 Total area and percentage deforestation in each seringal of the CMER .......................... 49 4 2 Seringais within the CMER with the highest deforestation rates ................................ ... 52 5 1 Variable names, operational definition, and descriptive statistics ................................ ... 60 5 2 Correlations between outcomes and independent variables. ................................ ........... 67 5 3 Multivariable regression models. ................................ ................................ ..................... 70 6 1 Previous residences of the households in the study seringais ................................ .......... 89 6 2 Time of residence of households in the study seringais ................................ .................. 91 6 3 Deforestation levels and rate s in the study seringais ................................ ....................... 92 6 4 Settlements subdivision in the study seringais. ................................ ............................... 97 6 5 Reasons for settlement subdivision in the study seringais. ................................ .............. 98 6 6 Membership in association and rural worker union in the study seringais ...................... 109

PAGE 9

ix LIST OF FIGURES Figure Page 3 1 Maps of the study site showing the seringais in Xapuri and Brasilia area .................... 28 3 2 Maps of the study site showing the seringais and settlements in Assis Brasil area ......... 29 3 3 Map of the mean rivers bordering the CMER ................................ ................................ 36 3 4 Map of forest typology occurring in the CMER ................................ .............................. 38 4 1 Location of the Chico Mendes Extractive Reserv e in Acre ................................ ............ 42 4 2 Deforestation and rubber production in Acre ................................ ................................ .. 45 4 3 Satellite image mosaic of the CMER illustrating deforestation (1986) .......................... 46 4 4 Satellite image of th e CMER illustrating deforestation (1992 1998). ............................. 47 4 5 Percent of deforestation in the CMER ................................ ................................ ............. 48 4 6 Map of the eight seringais with the highest deforestation rates in the CMER ................ 51 5 1 Representation of the settlement in the CMER ................................ ............................... 56 5 2 Seasonality in the rubber tappers year ................................ ................................ ............ 57 6 1 Total population of the CMER in three municipalities ................................ ................... 83 6 2 Total population in the study seringais ................................ ................................ ........... 87 6 3 Relationship of deforestation rates vs. population density in the study seringais. .......... 93 6 4 Total number of settlements in three municipalities of the CMER ................................ 95 6 5 Total numbers of settlements in the study seringais ................................ ........................ 96 6 6 Average production increments per settlement in the seringal Paraguau ...................... 100 6 7 Average agricultural productions per household: Seringal Paraguau ............................ 101 6 8 Average pasture area per settlement in the study seringais ................................ ............. 104

PAGE 10

x 6 9 Total memberships in CMER associations ................................ ................................ ...... 107

PAGE 11

xi Abstract of thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts DYNAMICS OF LAND USE IN AN AMAZONIAN EXTRACTIVE RESERVE: THE CASE OF THE CHICO ME NDES EXTRACTIVE RESERVE IN ACRE, BRAZIL By Carlos Valrio Aguiar Gomes December 2001 Chairman: Dr. Michael Binford Major Department: Latin American Studies Since the Chico Mendes Extractive Reserve (CMER) was established, different levels of land use change have taken place as a response to local and regional economic forces and prevailing social stresses. The following study addresses these changes and their drivers by analyzing key household characteristics and socio economic factors. This study prov ides both a household and a rubber tapper estates ( seringal) level approach of measuring land use changes. Based on remote sensing methods, four seringais with the highest deforestation rates were selected as the focus of analysis. The reserves census da ta on all residents of the four seringais was analyzed for three different periods (1995/1998/2000). In addition, in depth interviews were carried out with 66 households. The household level analysis was based on the data gathered from the 66 families, whe reas at the seringal level the research explores census data. The former approach considers three groups of independent variables: background of household head (migrant status and age), labor

PAGE 12

xii force availability (sons of the household head) and location of the seringais These were then measured in relation to six outcome variables per settlement: rubber and Brazil nut production, size of swidden plot, size of pasture, head of cattle and total area deforested. The serinigal level analysis concentrated on thr ee main factors: population, settlement dynamic, and changes in the rubber tapper production systems. Household level findings shows that the age of the household heads and the locations of the seringais have a strong association with rubber and Brazil nut production, which suggest that older household heads tend to engage in forest activities. The availability of sons of household heads exerts a strong effect on pasture and cattle raising activities, which suggests that non forest activities tend to be car ried out by the young residents. The migrant status accentuates agricultural activities, implying that non migrant households tend to engage in forest activities. Seringal level findings reveal that population density have increased significantly in the fo ur seringais In addition, it shows a crucial process in occupation of space within these seringais is increasing the subdivision of the settlements. Moreover, rubber tapper production system has been increasingly transformed from extractive to agricultura l and pastoral production activities. These increases in population density, number of settlements and change in the production systems accelerate deforestation, which will pose a serious dilemma for land use regulations in the reserve in the near future. This work is a step toward illuminating the prevailing issues on land use changes in the reserve. To understand which factors are driving these changes is a timely endeavor that will hopefully contribute to the strengthening of sustainable land use manage ment strategies in the CMER and in other extractive communities in Amazonia.

PAGE 13

1 CHAPTER 1 INTRODUCTION Amazon deforestation and its social and environmental consequences have been a major concern in the last few decades. The state of Acre in southwestern Brazilian Amazon is the birthplace of the rubber tapper movement, which was the first gras s roots movement in Brazil to advocate conservation of Amazonian forests through the establishment of extractive reserves. The creation of extractive reserves has been promoted as a major strategy for forest conservation while simultaneously providing a su stainable economic return to local peoples. A decade after the establishment of the federal extractive reserves in Amazonia, however, the extractive reserve model is still facing many challenges to put these goals into practice. The present work examines t he social dynamics of smallholder deforestation in the Chico Mendes Extractive Reserve (CMER), and analyzes the socio economic forces behind land use change in the reserve. The impoverishment of tropical forests through deforestation is a world wide phenom enon. In recent decades, few if any environmental issues have received as much global attention as tropical deforestation. Tropical deforestation will undoubtedly continue to be a central international environmental issue in coming years particularly as global issues such as climate change and the related carbon sinks and sources are discussed. Brazil, as owner of both the largest portion of the world's rainforests and the highest absolute deforestation rate, is de facto a leader in both conservation and destruction of these forests. There is a wide consensus that deforestation has environmental consequences, but there is considerable controversy over its social impacts

PAGE 14

2 and extent. In fact, the socio economic dynamics of deforestation in the Brazilian rain forests and its implications continue to be rather poorly understood, if the widely conflicting claims about its causes, impacts and remedies can be taken as evidence. Brazils legal Amazon 1 region includes about 5 million square kilometers, which compris es 58% of the countrys territory. The regional development policies implemented by military governments from the mid 1960s through the 1970s have engaged in programs aimed at opening up the Amazon region to development (Mahar 1978) These policies included massive highways to link these projects to the rest of the country, such as, the Transamazonica, the Belm Braslia, and the Cuiab Porto Velho roads. These development strategies also included economic programs for mineral extraction (e.g., Grande Caraj s), agricultural colonization and cattle ranching (e.g., Poloamaz nia, Polonoroeste), as well as incentives to timber industries. Therefore, in recent decades massive deforestation due to large scale cattle ranching, colonization, loggi ng, and mining projects has significantly reduced rainforests in the Brazilian Amazon and in some cases caused violent conflicts and severe socio economic problems for local communities (Moran 198 1, Schmink and Wood 1984, Smith 1982 ) Deforestation rates in the Brazilian Amazon ranged from 29,059 square kilometers in 1995 to 18,161 square kilometers in 1996; then 13,227 square kilometers in 1997, and 17,383 square kilometers in 1998, according to Brazils National Institute for Space (INPE 2000) From August 1998 to August 1999, the mean annual rate of 1 Two concepts are used to define the boundaries of the Brazilian Amazon region. The first and more commonly used definition is Legal Amaznia, which was established by military government in 1966. For administrative and planning purposes it comprises nine states: Acre, Amap, Amazonas, Mato Grosso, Par, Rondnia, Roraima, as well as part of Tocantins and Maranho. The second definition, Classical Amazon, the area covered by tropical forests, includes about 3.5 mill ion square kilometers (40 per cent of the country). It includes the Brazilian states of Acre, Amap, Amazonas, Par, Rondnia, and Roraima.

PAGE 15

3 deforestation was 17,259 square kilometers. The provisional estimate for the period of 1999/2000 is 19,836 square kilometers, which represen ts an increase of approximately 14.9% (INPE 2000). These data clearly show that deforestation rates have not decreased in Amazonia despite efforts by the Brazilian government to slow them. Yet the Landsat TM satellite images used by INPE do not include def orestation of areas smaller than 6.25 hectares, so not even all deforestation is being counted. The impact of hundreds of thousands of small scale farmers, including rubber tappers as well as logging activities and wild fires are not included. Logging crew s alone severely damage 10,000 to 15,000 square kilometers a year of forest that are not included in deforestation mapping programs (Nepstad et al. 1999) Despite being considered an agricultural frontier, the Amazon has most of its popula tion concentrated in cities. In 1940, the population of Amazonia was only 1.5 million people, of whom 72% were rural dwellers ( Barraclough and Ghimire 1995 ). According to the Demographic Census of 1970, 35.5% of the population was urban, w hich compares with 44.6% in 1980, 55.2% in 1991 and 61.5% in 1996. A number of explanations exist for these demographic patterns in the Amazon. The Amazon became a place of refuge for a segment of southern Brazils landless population that pursued the goal of obtaining its own land and building a resource base. The decade of the 1970s was a period of intense activity in the region. The highway construction projects, the colonization schemes, and the spontaneous migration into the region were among the most salient factors that altered the demographic profile of the area ( Wood and Wilson

PAGE 16

4 1991 ). Currently, almost two thirds of Amazonias population of almost 20 million people lives in urban areas (Browder and Godfrey 1997, Frie nds of the Earth 1998 ). In the early 1980s, deforestation in Amazonia and its associated effects on the livelihoods of local populations became an international issue. In this scenario, the Brazilian government started to consider development alternatives that use Amazonias diverse resources while maintaining forest cover and valuing the needs of traditional populations. Since the 1980s, recognition by Brazilian and international society of the importance of maintaining the biological and cultural diversi ty of Amazon ecosystems has led to the creation of numerous conservation units, including national forests, national parks, biological reserves, as well as indigenous areas and extractive reserves. In this context, extractivism, the ancient practice of ext racting economically valuable non timber forest products (NTFPs), emerged as a possible means of reconciling economic development with forest conservation (Nepstad and Schwartzman 1992) Extractivism captured global attention in the late 1 980s when a group of rubber tappers fought to protect their lands from encroaching cattle ranchers in Acre, Brazil (Allegretti 1989, 1990 and 1994 Schwartzman 1989 and 1991) culminating in the highly pub licized assassination of the rubber tapper leader Francisco Chico Mendes in 1988. These events led to the creation of the first extractive reserves, which are forest areas inhabited by extractive populations granted long term usufruct rights to resource s which they collectively manage (Schwartzman 1989 MMA/IBAMA/CNPT 1995 ). At that time, e xtractive reserves were considered necessary to act as a counter balance to the pressures of deforestation and cattle ranching, fitt ing the relatively new demand for sustainable development and an opportunity to put into practice an explicit

PAGE 17

5 linkage between conservation and development. By restricting the use of forests to a defined set of products and users, extractive reserves are considered a major strategy for promoting forest conservation and local socio economic development in Amazonia ( Allegretti 1990 Anderson 1989 and 1992 ). Compared to other land use strategies, extractive reserves have been successful in maintaining relatively low deforestation rates (Mendes et al. 1996 Nepstad et al. 1992). Nonetheless, extractive reserves are not stand alone solutions, but they complement other land use options. In the last decade, the co ncept of extractive reserves has evolved in the Brazilian Amazon, and extractive communities in other tropical countries have followed the example of Acre ( Salafsky and Dugelby 1991 ). 1.1 Problem Statement The rubber boom at the end of th e nineteenth century and the brief period during World War II indelibly marked the history and cultural characteristics of Acre ( Rancy 1992 Tocantins 1979 ). Since the 1970s, the eastern part of the state has been marked b y clashes between forest dwelling people and settlement projects and cattle farms. Today, changing land use and land cover are the rule in Acre (153,000 km 2 ), the western most state in the Arc of Fire and Deforestation in Brazilian Amazonia. Acre represe nts an interesting threshold site to study the dynamics of deforestation in a context where forest dwelling communities, cattle ranchers and colonists coexist. The initial assumptions of the extractive reserve concept were that an economy based on non ti mber forest products could increase the economic value of forests and rural income, offering many subsistence resources for extractivist communities, while being ecologically sustainable. However, the extractive reserve model is facing

PAGE 18

6 challenges in puttin g these objectives into practice as economic, social, political and ecological constraints have emerged (Anderson 1992 Browder 1992 a, Homma 1993 ). Extractivist communities are faced with a current pattern of low income generation, and the goal of economic growth is a major constraint forcing land use changes in extractive reserves. Land use patterns within the extractive reserves are changing from strict non timber forest products (NTFP) economies to small scale cash crop agricultural and pastoral activities. One of the key questions is how economic development changes land use in extractive reserves and how these changes can be managed. Of particular importance is the management of land use in a rapidly ch anging socio economic context while protecting forest resources according to the conservation objectives of extractive reserves. To evaluate the efficiency of extractive reserves as a conservation and development model in Amazonia, it is necessary to analy ze land use change and deforestation rates within these areas. Deforestation rates have been reported for areas along the Amazonian development frontier with colonization projects and cattle ranches that result in large scale deforestation patterns, parti cularly in the state of Rondnia and Par ( Alves 1999 Fearnside 2001, Moran and Bondizio1994 Nepstad et al. 1999 Rignot et al. 1997 Skole and Tucker 1993 Skole et al. 1994 Wood and Skole 1998 ). Rarely, however, has forest dwelling smallholder deforestation been examined with the exception of a few studies (Brondizio et al. 1994 and in press, Peralta and Mather 2000 Sassagawa 1999 ). Sassagawa (1999) conducted a remote sensing analysis of the deforestation rates in the Chico Mendes Extractive Reserve (CMER) in Acre, detecting land use changes in the reserve during the p eriod of 1986 1998. The analysis revealed that eight seringais, or

PAGE 19

7 rubber estates, with the highest deforestation rates within the reserve are expected to exceed the legal limit for deforestation (10% of each residents area) within 2.5 to 6 years. This th esis uses Sassagawas work as a basis for the deforestation analysis in the CMER. However, this study is limited to remote sensing analysis of land cover changes and did not incorporate the human socio economic factors driving land use change on the CMER. That is, while this study evaluate land use changes top down looking at general patterns of deforestation in the reserve, it does not examine the on the ground socio economic processes that lead to the patterns of land use. In fact, socio economic aspect s of forest dwelling smallholders are still little analyzed for several parts of the Amazon. Most studies with this population have been unable to achieve an integrated perspective between land use cover and socio economic factors inducing land use changes Few studies have examined the way in which patterns of land cover change relate to underlying social and economic driving forces. The gaps in research in socio economic factors on extractive reserves are important aspects to understand the land use chang e process that has taken place in these areas since they were established. The analysis of land use patterns in the CMER was not necessiry at the creation of the reserve, since the economic production at the time was primarily based on rubber and Brazil nu t extraction from standing forests. In the last ten years, land use decisions made by CMER residents have gradually changed toward a mix of extractive and non extractive activities, in an attempt to increase household income and also in response to develop ment incentives from projects and government initiatives. A s the pressure to diversify production and to increase income grows, rubber tappers are increasingly moving toward land use activities that have more destructive impacts on forest cover.

PAGE 20

8 Therefore, today, cash crop agriculture, and small scale cattle ranching are an important component of modern rubber tapper land use strategies. These non extractivist activities play a dual role as the new economic activities: not only do they increase household in come, but they also increase pressure on natural resources and deforestation rates in the reserve. 1.2 Central Questions This thesis addresses three main questions. First, what are the land use patterns and changes occurring in the Chico Mendes extractive reserve? Specifically, this question analyzes the overall land use patterns of the reserve in order to understand the actual land use changes in recent years. Second, what are the socio economic forces behind these land use changes? Several factors influe nce land use change in the reserve, which might include population increase, development policies for extractivist communities, lack of market for non timber forest products and development projects influencing agricultural and pastoral land use activities Third and finally, what factors are contributing to deforestation in concentrated areas of the reserve? The land use patterns in the reserve follow two separate, opposed tendencies. Some areas of the reserve have continuing and high deforestation rates, while in other parts very little deforestation has taken place in the last decade. Factors influencing deforestation in these areas could include population density, road and market access, distance to neighboring towns, as well as distinct land use strate gies bordering with the reserve.

PAGE 21

9 1.3 Structure of the Study Having here explained the importance of studying smallholders land use patterns in the Chico Mendes Extractive Reserve, chapter 2 will provide an overview of the diverse arguments involving the concept of extractive reserves in the Amazon. This chapter is based on the literature produced in the early debate about the reserve model up through recent works based more on practical experience of the model. In Chapter 3, I will describe the procedures used to define the study site, survey setting and planning, types of primary and secondary data gathered as well as data processing, variable operationalization and definitions, and analysis. In chapter 4, I will analyze the context in which deforestatio n has taken place in the reserve. I begin with a discussion of deforestation in Acres context and how it has affected deforestation in the Reserve. I then take a look at how and why seringal level deforestation rates differs in the reserve. In chapter 5, I present a household level approach of measuring some characteristics affecting different land use activities. First I characterize the major land use activities of the rubber tapper settlements and then relate those to seasonality of land use activities throughout the years and differentiated family labor force allocation. Next, I define six major land use outcome variables that define tappers production systems. I then, define a group of explanatory household characteristics and seringal location variab les as major determinants of extractivist and non extractivist land use activities in the study seringais. Chapter 6 I present general trends of key socio economic factors driving land use changes in the reserve as a whole and in the four specific seringai s, based on census data of the reserve in three different periods. At the reserve level, particular attention is given to population and settlement dynamics and social organization of the residents. At a

PAGE 22

10 seringal level analysis, I explore factors inducing land use change, looking at population and settlement dynamics, extractivist and agricultural production, pasture expansion, and social organization. In chapter 7, I summarize some of the mean funding of the study and discuss some of the challenges that th e reserve will face regarding resource regulation and future economic development of the residents. Finally, I suggest continuing steps to be taken in this research as well as in other related topics. In this regard, study in land use change in the reserve still leaves much to be explored.

PAGE 23

11 CHAPTER 2 DEBATE ABOUT EXTRACTIVE RESERVES 2.1 Brief History of the Amazon Rubber Boom Much of the economic history of Latin America has been characterized by several extractive boom and bust cycles. Most of these cycles have been based on the exploitation of plan ts of commercial value. The Amazon rubber boom, which spans the late nineteenth and early twentieth centuries, was one of Latin America's most extreme natural resource export booms. Rubber had been known for a long time before it became an economically imp ortant commodity. The French naturalist Charles Marie De La Condamine, who traveled down the Amazon from Ecuador in 1736, is considered to be the first person to publish a study about the properties of rubber, the milky latex of the Hevea brasiliensis L. t ree ( Dean 1987 ). In 1839, Goodyear invented vulcanization, a process that made rubber durable, turning it into an essential product of the Industrial Age and triggering inventions that would eventually lead to massive demands. Events such as the invention of the inflatable rubber tire by Dunlop in 1888, the bicycle craze in Europe (especially in France) in the late 1800s, and the development of the automobile industry in the early 1900s explain the worlds explosive demand for rubber towar ds the end of the century. With the increasing world demand for rubber, the price of the product grew steadily. Amazonia, the worlds sole supplier of rubber, experienced decades of economic growth in terms of exports. The wealth generated by the black go ld rubber had fantastic proportions. Rubber exports rose sharply in

PAGE 24

12 quantity, and even more dramatically in value. The rubber trade became a mainstay of the Brazilian economy, providing at its height almost 40 percent of its export revenues, nearly equali ng coffee in importance (Dean 1987). During these giddy decades, Brazil monopolized the world rubber trade. After a few decades, however, the Amazon rubber fever was broken. The deterioration of Brazils rubber economy began in 1912 with the collapse of it s monopoly because of competition from Malaysias rubber plantations. Southeast Asian plantation rubber trees grew faster and gave higher yields than the wild rubber of the Amazon, due to the skilled domestication efforts. Consequently, Asian rubber could produce higher quality rubber more cheaply, quickly, and at vastly greater quantities than its wild Amazonian rival. The once wealthy region never fully recovered from the rubber bust, despite attempts to strengthen Amazonian rubber production. The Brazili an and North American governments heavily invested in rubber production in the Amazon during the Second World War, when American access to the Southeast Asian rubber supplies was blocked. However, after the war and with the development of synthetic rubber substitutes, the period of high rubber production in the Amazon was once again over. Despite its failure to establish long term development of the region, the rubber boom era indelibly marked Amazonia and caused substantial social and economic alterations in the region that can be seen and felt until today. 2.2 Evolution of the Extractivist Development Model in Amazonia Amazonia has been indelibly marked by the rubber boom and bust (1890 1912) that brought thousands of people from the arid Northeast of Bra zil to the rainforest as labor for rubber extraction ( Bakx 1988, Rancy 1992 ) Since that time, rubber tappers

PAGE 25

13 have been a disenfranchised category in Brazilian history. Throughout Amazonia they have worked in the semi sla very aviamento system of debt peonage, a type of social relations that persists in many regions of Amazonia until today ( Weinstein 1983 ). However, the wealth generated for the elites by the black gold had fantastic proportions, bringing several decades of unparalleled prosperity and growth to the region, integrating Amazonia to the world economy, establishing extensive trading networks, constructing massive port facilities, and leaving behind large urban centers along Amazonias rivers (Weinstein 1983) With the end of the rubber boom, forest extractivism was generally viewed as a backward and peripheral activity of economic stagnation, and forest resources and traditional peoples considered as development obstacles in Am azonia (Barbosa 2000) This attitude characterized Brazilian government policies for many decades up through the 1980s and brought development projects to the region that included massive highway construction (e.g., Transamaznica, Belm B raslia), dam building (e.g., Balbina, Tucuru), mineral extraction (e.g., Grande Carajs), agricultural colonization and cattle ranching (e.g., Polonoroeste), as well as incentives to timber industries (Hecht and Cockburn 1989, Moran 1983, Oliveira 199 1, Schmink and Wood 1992 ) These projects brought serious social and environmental problems to rural and urban Amazonia, with deforestation reaching unprecedented rates and rural violence against traditional communities rising to alarming numbers. As a response to the rubber boom legacies and the oppressive development projects, recurrent peasant resistance erupted in Amazonia and led to violent clashes along the development frontier. The rubber tapper movemen t emerged in the state of Acre in the late 1970s. Chico Mendes was the major force behind the movement.

PAGE 26

14 Through collective action, Mendes and the tappers fought for the right over the land and not to leave it for the ranchers whose ownership rested more o n coercion than on legality. Slowly, they began to organize against the major cases of deforestation that occurred in the area by the means of empates or standoffs, blocking people and machinery with their bodies (Calaa 1993 ). The rubber tapper movement only received political clout after the creation of the National Council of Rubber Tappers in 1985, when it was decided that the primary demand would be agrarian reform in the form of extractive reserves ( Allegretti 1994 ). In the late 1980s the first powerful alliances between rubber tapper, indigenous organizations (e.g., UNI Union of Indian Nations) and the environmental movement emerged. Common goals such as indigenous land demarcations, creation of extractive reserves a nd opposition to mega projects, shaped joint strategies of civil disobedience and protests 2 ( Albert 1992 Fisher 1994 Hall 1989) The onset of democracy in Brazil also politicized the national environment al and grassroots organizations and boosted the emerging NGO movement that embraced polemic rubber tappers and indigenous causes (Barbosa 2000, Viola 1988) The Amazonian grassroots efforts also gained considerable strengt h when international interests in Amazonia reached high levels. The internationalization of Amazonian ecopolitics occurred twofold. First, beginning in the mid 1980s, awareness of major global environmental problems such as the greenhouse effect, the ozon e hole, and biodiversity loss, sparked an international debate on the state of the global environment (Stern et al. 1992) Overnight, Brazil was in the spotlight, pressured by environmentalists, 2 In particular the Kayap became widely known for civil disobedience and protests in Altamira in 1989 to prote st the construction of several dams along the Xing river ( Fisher 1994).

PAGE 27

15 international organizations, and first world politicians to stop the devastation taking place in the Amazon rainforest (Barbosa 1993 and 2000) Second, international development schemes were heavily attacked, resulting in international publicity around target dam an d road construction projects that have caused international financial supporters to back out. 3 At first, the international attack on Brazilian development strategies provoked strong protest against this environmental imperialism. 4 However, President Col lors administration (1990 1992) changed toward a greener rhetoric. In the pre UNCED (1992) arena the Brazilian government took visible steps to demonstrate to the developing world and to donor countries serious efforts to protect Amazonian forests and its people. Part of this was the creation of extractive reserves (ERs), national parks and the demarcation of indigenous territories. This fight for ERs lasted from 1985 to1990 and encountered many difficulties. The very idea of extractive reserves was th ought incongruent with the legislation of the time, since this defined property in individual terms rather than collective. Decree No. 98.897/90 was the compromise (Murrieta and Rueda 1995) It defined these areas as property of the union or federal government destined to be used for extractive activities (Allegretti 1994) The creation of ERs represented a novel government response to the pressure both of social movements and conservationists, and a form of legally 3 For example, as a result of rubber tapper lobbying the Inter American Development Bank withdrew its financial support for the road construction of BR 364. It was a first in the banks history that it stopped a loan on the basis of environmental concerns. Another example is the withdrawal of the World Bank from the Xing river dam projects in response to the indigenous Kayap protests (Barbosa 2000). 4 In the Sarney a dministration (1985 1990) t he development of Amazonia became a question of nationalism, of national security, and of sovereignty over encroachable lands. The demarcation of indigenous territories and protected areas was seen as a conspiracy of developed co untries, to hamper competitive development in Brazil and to steal land for global (not national) carbon and biodiversity stocks (Barbosa 2000).

PAGE 28

16 recog nizing land tenure regimes that do not comply with the private property paradigm that characterizes most modern economic systems. 5 In 1990, the federal ERs existed on paper, but it took several years for ERs to take shape. 6 Only by 1994 were the four fede ral Amazonian ERs legally and institutionally consolidated, covering an area of over 2 million hectares (ELI 1995 Murrieta and Rueda 1995) Apart from land tenure, resource utilization regulations at the community level w ere a principal concern in the early stages. The newly created Traditional People Department (CNPT) of the Brazilian Environmental Protection Agency (IBAMA) was made the responsible government institution for ERs. This step was followed by the development of utilization plans, jointly created in by rubber tapper associations and government representatives, where federal environmental legislation and traditional rubber tapper resource utilization were reconciled ( MMA/IBAMA/CNPT 1995 ). 2.3 A nalysis of the Extractive Reserve Literature Despite its charisma, the basis of extractive economies is indisputably fragile and the subject of an active and sometimes polemic debate among scholars and conservation and development professionals of diverse academic backgrounds (e.g., Anderson 1992 Fearnside 1989 and 1992 Browder 1992b, Homma 1989 and 1993 ). Publications on ERs strategies peaked in 1992; however, since then very few works have appeared evaluating the ER debate and experience. Consequently, there is a need for a survey of the discussion. In the next section I will focus on the Extractive Reserve experience in the 5 ERs were created by Brazilian Ministry of Agrarian Reform and Development (INCRA), also responsible for extens ive colonization projects in frontier Amazonia. Approval of this form of land tenure signified a radical departure from the way in which regional development in Amazonia had been carried out in the past ( Allegretti 1990). 6 The first fede ral extractive reserve created was the Upper Jurua River in 1989 (Acre), followed by the Chico Mendes (Acre), Cajari River (Amap) and Ouro Preto Riber (Rondnia).

PAGE 29

17 Brazilian Amazon region in order to analyze the leading discussions about ERs, discerning the major trends of debate; and draw conclusions from current ER experiences about the prospects for ER development in Brazilian Amazonia. 2.3.1 General Patterns and Trends in the Debate When extractive reserves were first proposed by the grassroots National Rubber Tapper Council, the cause was picked up and supported by scientists, activists, and conservation and development professionals and was considerably transformed. The late 1980s and ea rly 1990s were characterized by a relatively fresh enthusiasm for environmental and social causes in the tropics (compared to the learned skepticism of today), and the debates around ERs developed into a curious mix of science, ideology and lobby. The so mewhat emotional content of the debate was further enhanced by the uproar over the assassination of Chico Mendes and the violent suppression of indigenous peoples, transforming them into international environmental heroes. At the time, ERs were optimistica lly promoted as the panacea for most of Amazonias environmental problems. Three intermeshed lines of discourse influenced, shaped and transformed the discussion about ERs. First, the international lobby saw ERs as an answer to the urgent demand for rainfo rest conservation as a reaction to high deforestation rates and the concern about global climate change and biodiversity loss. Second, the promotion of traditional resources management as an example of sustainable livelihoods boosted the appeal of ERs as a n example of "inherently environmental" rainforest people as agents of forest conservation. Third, the proposal of ERs coincided with the promotion of non timber forest products use and marketing as a promising economic alternative for (mostly) biodiversit y rich tropical regions. The different demands of these interest groups

PAGE 30

18 caused an inflation of the expectations on ERs. From a human rights and territorial issue that embraced forest conservation as a means of continued livelihood in the forest, the discus sion spawned high economic expectations. The main participant in the beginning debate about ERs was first a Brazilian, anthropologist Mary Allegretti (1989 and 1990) who publicized the rubber tappers cause and later moved into important g overnment positions. The discussion was later joined by mostly North American conservationists such as Anderson (1989, 1992 and 1994 ) and Fearnside (1989 and 1992) followed by development specialists like Schwartzman (1989 and 1992 ); regional planners such as Browder (1990a,b ) as well as the Brazilian economists Alfredo Homma (1989 1992 and 1993) From the onset it was an interdisciplinar y debate, bringing formerly isolated groups together. However, no consensus existed in the vision and objectives about the nascent ERs model. The strongest proponents of the model were Allegretti (1989 and 1990) and Schwartzman (1989 and 1992), supported b y Anderson and Fearnside. The strongest critiques of ERs were produced by Browder (1990 a,b), and Homma (1989 and 1993). The ER literature has been produced in a characteristic boom and bust pattern. In 1989, the year after Chico Mendess a ssassination, articles mushroomed about the topic both from promoters (Allegretti 1989, Fearnside 1989, Schwartzman 1989 Hecht 1989 ) and skeptics ( Homma 1989, Browder 1990 a) in addition to the popular media coverage. From 1990 1993 a many articles were written about the topic, culminating in some edited books that discussed the non timber forest products strategy and ERs (e.g., Nepstad and Schwartzman 1992, Anderson 1990 Redford and Padoch 1992 Plotkin and Famolare 1992 Homma 1993) However, after 1993, hardly anything was written about ERs as a

PAGE 31

19 development and conservation model. After that, most produ ction has gone into gray literature reports and government policy statements. Notable exceptions include a Brazilian edited book that refutes some of the criticism of ERs and a summary of the ER status by the government staff (Murrieta and Rueda 1995) A paper written by visionary Acre Production Secretary Jos do Rego points out that the new ideas of extractivist development in Amazonia, neo extractivism, incorporate technical improvements in cultivation, harvesting and processing, alli ed with a specific social environment of forest cultures people whose lifestyle is intimately tied to the forest ( Rego 1999 ). Also, two other papers about extractive reserves published recently discuss the institutional architecture of e xtractive reserve, and the political and economic empowerment of rubber tappers communities in the State of Rondnia (Brown and Rosendo 2000 a,b). Another characteristic of the ER debate is that most of the critical body of literature was w ritten before ERs were legally and institutionally functioning on the ground (as they were to some extent after 1994, see Murrieta and Rueda 1995) This temporal gap between intellectual battle about ERs and de facto experiences of ERs sug gest that the debate about ERs was a very theoretical and speculative one. This might also explain some of the arrested analysis of the model, as scholars are waiting for ERs to develop on the ground. Several large national and international projects wer e established within ERs, in great part due to the many publications and the publicity about ERs in the early stages. To understand what happened to the ER debate, one has to examine the role of different arguments in the ER literature.

PAGE 32

20 2.3.2 Social Deve lopment Arguments The rubber tappers demand for ERs was primarily a human rights battle linked to an environmental one (Schwartzman 1989) In extractivist communities, the needs of traditional lifestyles translate directly into territorial rights and the need of relatively large expanses of forests as the basis of their livelihoods. Even so, rubber tappers were also fighting to obtain modern production technology for diversification of their production and the creation of education and heal th systems adapted to the needs of the community (Allegretti 1995) which were, and still are, in great need. The establishment of social institutions, such as cooperatives and associations, and physical infrastructure, such as transport sy stems and roads, have also been determined to be crucial for the success of extractivist populations to sell their products ( Schwartzman 1989) The importance of social and political realities within the extractivist region has repeatedly been pointed out as a critical factor that can define the success or failure of ER endeavors. The behavior and incomes of rural people who extract forest products are often determined by social and economic factors over which they have little or no immedia te control (Browder 1992a) 7 An additional concern about the ER model has been that it requires low population densities and therefore large forest expanses for few people (Fearnside 1989, Browder 1990a) 8 It has been argued that it is hard for land planners to justify ERs that require tens of hundreds of ha of forest to support one 7 For instance regional power relations and national development policies can join forces t o support or to undermine ERs. In that context, it is very telling that ERs were only implemented in Amazonian states without an overwhelmingly strong large scale landowner (latifundia) alliance (i.e., Acre and Amap in contrast to Par and Amazonas). 8 Fe arnside (1989) estimated that Brazil nut harvesters need 300 500ha per family, resulting in about 1 1.7 people/km, concluding that the carrying capacity for extractivist economies is low.

PAGE 33

21 household in countries that have extremely high population densities. 9 2.3.3 Economic Arguments The most advertised promise s and the most serious constraints of the ER reserve model have been identified as economic ones. The conservation community was excited about ERs as a model that could preserve biodiversity while simultaneously providing sustainable economic return to loc al peoples and governments ( Allegretti 1990, Schwartzman 1989) This excitement was fueled by an article by Peters et al. ( Peters et al. 1989) proposing that long term financial return from the harvest of non timber forest products found in a hectare of Amazonian rain forest far outweighed the net benefits of timber production or agricultural conversion from the same area of land (see also: Anderson and Jardim 1989, Anderson and Ioris 1992, Gomez Pompa and Kaus 1990 Panayotou and Ashton 1992 ). Several authors, including Peters et al themselves, have attempted to temper this enthusiasm for income generation by pointing out that hypothetical calculations of the income streams to be derived from an average hectare of tropical forest have significant limitations ( Fearnside 1989, Pinedo Vasquez et al. 1990, Godoy and Bawa, 1993, P erez and Byron, 1999) The fragility of extractive economies was spelled out from the onset (Anderson 1989, Fearnside 1989), as few Amazonian extractive products have augmented in production and value in the last decades. This certainly was a very fragil e base for a development strategy that has as a goal the well being of extractivist populations. The premise that NTFP extraction can produce greater financial return compared to timber or agriculture also has been challenged by many authors (e.g., Browder 1990a,b and 1992a, 9 However, although Amazonia's population is rising and currently r eaching 20 million, rural population is decreasing. In addition, many other areas of Amazonia have extremely low population density ( Martine and Camargo 1998).

PAGE 34

22 Homma 1992 and 1993). Particularly, Homma attacked extractive economies as inherently uneconomic, arguing that all extractive products would eventually be substituted by domesticated plantation producti on or synthetic replacement of the extractive good. 10 This was the case for latex, which was the main product of extraction in Amazonia and is now cultivated in plantation within and outside Brazil as well as produced synthetically. He argued that if the p rices of extracted products fall, degradation is likely to follow because extractors must harvest the resources above sustainable thresholds in order to maintain their living standards. He also argued that investments should be made at the development fron tier, where colonists and cattle ranchers pursued more viable economies. However, this position has been challenged by Allegretti ( Allegretti 1994 and 1995) who points out that these land uses are only viable due to the massive incentives and tax benefits that the Brazilian government has used to encourage cattle ranching and colonist agriculture on the Amazonian development frontier. She argues that similar tax and credit incentives need to be established for extractive economies. Also, the inherent assumption within these critiques, that ERs have to compete with other land uses in the market economy, might be quite reductionist, as they ignore the non market value of these areas, such as the forest conservation, subsist ence value and the opportunity value of untapped resources (Allegretti 1994, Anderson and Ioris 1992 Nepstad and Schwartzman 1992 ). 10 Homma delineates three phases in the extractivism of a forest product: (1) e xpansion; (2) stagnation and (3) decline. The decline is attributed to four main factors: (1) inelastic market for the forest products; (2) extraction levels which surpass rates of natural regeneration; (3) domestication of the product; and (4) substitutio n of the forest product for a similar natural or synthetic product. Extractivism is just a phase in the economic development of a region, and if a forest product is that important, it will either be domesticated (generally somewhere else) or be substituted by a cheaper synthetic product.

PAGE 35

23 2.3.4 Ecological Arguments ERs were championed for two main ecological reasons. First, ERs were proposed as a means of preserving tropical forest cover. This has been confirmed in comparison to other land uses strategies (Mendes et al. 1996 ). Secondly, the early NTFP literature assumed that the harvesting of NTFPs occurrs on a sustainable basis since local people have harvested these products for hundreds of years. Particularly anthropologists (e.g., Posey and Bale 1989 Allegretti 1979 ) had the tendency to think so. However studies have shown that commercialization often leads to overexploitation ( Bodmer at al. 1997) In that respect very little has been published about products from ERs in Amazonia (e.g., Pinard 1991 Kaine r 1995) ERs have also been criticized as areas of open access where extractivists were encouraged to respond to perverse incentives of overexploitation ( Homma 1989 Salafsky and Dugelby 1991 ). This notion has been r efuted by May (1989) and the development of Utilization Plans of the reserves demonstrates the conscious attempt at resource use regulation. A final argument in favor of the ERs strategy has been that it supports a traditional lifestyle, increasing the subsistence value of forest resources, preserving knowledge and thus preserving biological diversity (Balick and Cox 1996) Linked to the discourse of preservation of traditional knowledge is the call for the protection of t he opportunity value of protected forests within ERs as repositories of biological and genetic diversity and therefore potential sources or new products for agricultural and pharmaceutical industries (Balick and Mendelsohn 1995, Gottlieb and Kaplan 1990) However, this argument has proven problematic, as it fueled economic expectations that have not been corroborated, and triggered a highly polemic debate about access to genetic and biological diversity in Amaznia.

PAGE 36

24 2.4 Ac hievements and Problems with Extractive Reseves The discussion about ERs has been overly focused on their economic fragility and neglected more basic community issues that are important for the social success of the reserve and also for their eventual econ omic success. A number of important achievements were made. Today, rubber tappers hold communal territorial rights within federal government protected areas that explicitly state their use and management by traditional residents. A great advance is also th e elaboration of Utilization Plans and an ecological monitoring program, although evaluations of ecological practices has not occurred to a great extent (Millikan and Irving 1997) Another important achievement is, perhaps, the building of social institutions such as associations, cooperatives, affiliations with the rural workers union and the fortification of the Rubber Tapper National Council ( Ramalho 1992, Schmink 1992 ). Nonetheless, a continuous effort will be required in order to maintain internal cohesion and common goals within these organizations, and to acquire the administrative and organizational skills necessary to run their commercial enterprises. In terms of formal education and health service s for residents, ERs are still very fragile, as access to many areas is difficult and cooperation with state and municipal agencies has been limited and fractious. In terms of development support, Extractive Reserves have been supplied with international funds. Most significantly, they have been inserted into the international Pilot Program to Conserve the Brazilian Rain Forest (PPG 7) since 1995. 11 The focus of 11 The PPG 7 is one of the largest conservation programs ever experienced in Amazonia. It is the joint effort on the part of the Group of Seven countries, the Government of Brazil and the European Union, with the World Bank as administrator of the financial resources. The budget of international and national support to the four federal extractive reserves in the last four years were approximately ten million dollars (CNPT 1999).

PAGE 37

25 this mega project has been 1) The consolidation of the legal status and the regulations withi n ERs, 2) Social organization, community participation, capacity building and education and health programs, and 3) Extractivist production organization, diversification, transportation and marketing (IBAMA/CNPT 1999) Moreover, the Brazi lian government has offered credit lines to rubber tappers settled in extractive reserves, an innovative governmental policy, as until recently credit was only accessible for agricultural and cattle production in Amaznia. The Program of Support to Extrac tive Development (PRODEX)s goal is to support the development of extractives activities, providing technologies and improvement in production to the communities, and incentives for diversification of production. However, there are many obstacles that pre vent it from having a significant impact on the improvement of extractive production. First, the financial resources are administered by BASA, the Amazon Bank, which has little operational capacity and is bureaucratically too rigid to work with extractive s communities, which often cannot provide the required documentation. Another weakness of the PRODEX is the quality of the technical assistance, as government technicians have little experience in working with extractives communities. Nonetheless, the prog ram demonstrated the federal governments recognition of the needs of extractive communities. Political changes at a federal and state level have contributed to the ER model in Amaznia. At the federal level, a team of top professionals with great commitm ent to Amazonian environmentalism was selected by the Brazilian minister of environment for cutting edge decision making, such as Mary Alegretti, who greatly contributed to the definition of extractive reserve in the 1980s. At the state level, in the Amazo nian states of Amap and Acre, leftist environmentalist governors have been elected (Capiberibe in

PAGE 38

26 Amap and Jorge Viana in Acre). In Acre where the struggle began, the government has explicitly named itself "The Government of the Forest" and sustainable d evelopment on the basis of forest resources is at the center of its mission (Schmink 1999 ). The best possible political conditions currently exist in these regions. The next years will show whether the structural and policy changes that ar e occurring can considerably improve the prospects of ERs. The biggest difficulty at this point challenging ERs is income generating activities (Mendes et al. 1996, Millikan and Irving 1997) An important, although late recognition has bee n that traditional extractive populations do not need to rely solely on extractivist production. Others land uses activities, such as agriculture, agroforestry systems, and small scale cattle ranching are also important (Anderson 1992 and 1994, Rego 1999) In an ideal case these other land use practices would intensify production per area without sacrificing ecological sustainability. However, a s the pressure to diversify production and increase income buil ds up, extractivists tend to move increasingly toward more uns ustainable land use activities and thus increase deforestation in the reserves. Deforestation rate is a major indicator to evaluate the efficiency of ERs and to fulfill its conservation objectiv es. Chapter 4, addresses land use changes in the state of Acre and the context of deforestation rates and trends in the Chico Mendes reserve.

PAGE 39

27 CHAPTER 3 STUDY SITES AND RESEARCH METHODOLOGY 3.1 Study site selection The field research for this project was carried out from May to August 2000 in the Chico Mendes Extractive Reserve, in Acre, Brazil. The reserve is comprised of 970,570 ha (9,705 square kilomet ers), and is divided into the historically important rubber tapper estates ( seringais). For this study, I selected the four seringais with the highest deforestation rates based on the work of Sassagawa (1999). In addition, I selected these seringais accor ding to their location and their extension across different municipalities within the reserve. The four seringais selected were seringal Filipinas (in the municipality of Xapuri), seringais Porongaba and Humaita (in Brasilia) and the seringal Paraguau (i n Assis Brasil). These seringais are relatively near to the closest towns and can be accessed following the highway that leads from the capital Rio Branco to the Brazilian border with Peru. Seringal Filipinas is closer to Xapuri. The seringais Porongaba a nd Humait are located three hours from the closest town, Brasilia. Seringal Paraguacu is located three hours from the municipal seat of Assis Brasil. The study seringais differ in terms of area, population size and some economic activities practiced by t he residents Figures 3 1 and 3 2 are maps of the Chico Mendes Extractive Reserve showing the four seringais where fieldwork was carried out, including the households where one time interviews were conducted. Overall, the total number of households intervi ewed in

PAGE 40

28 Figure 3 1 Maps of the study site showing the seringais and settlements in Xapuri and Brasilia area

PAGE 41

29 Figure 3 2 Maps of the study site showing the seringais and settlements in Assis Brasil area

PAGE 42

30 the four seringais was 66. In the seringal Filipinas, which is the largest, nineteen interviews (28.8% of the total interviews) were carried out representing approximately 40% of the households in the seringal. In the seringal Porongaba, almost all the household heads were interviewe d, totalling sixteen households (24.2 % of the total interviews). In the seringal Humait, a sample of eleven household heads was interviewed (16.7 % of the total interviews). In the seringal Paraguau, twenty interviews were carried out (30.3 % of the tot al interviews), representing approximately 70% of the total households in the seringal. It is important to stress that the interviews were mostly conducted with the male heads of household; however, in three cases women were the head of household. Intervie ws were limited to household heads as these actors make the land use decisions at a household level. The rubber tappers settlements in the seringais are located 30 120 minutes walking distance from each other. The most distant households surveyed were lo cated at about a four hours walking distance, approximately 20 kilometers from the first household surveyed in the entrance of the seringal Households were selected based on their location in the seringal, so that the sample would be evenly distributed th roughout the area. 3.2 Survey setting and planning In previous work I have gained knowledge of the rubber tappers in the Chico Mendes Extractive Reserve, which helped me set up the research in conjunction with the rubber tapper community. Before visiting the households, the research was discussed in meetings with leaders of rubber tapper social organizations in the three municipalities. In these meetings, I explained the goal of the research as well as the methodology that I was

PAGE 43

31 going to use. These discuss ions provided important feedback that helped me to make some changes in the research goal and methodology. The association leaders also helped me by announcing to the communities that I was carrying out the research, through local radio stations and the as sociations amateur radios. This greatly facilitated the research, as the tappers already had an idea of the purpose of my study when I arrived at their homes, affording more time for interviews and participant observation. In addition, the associations in dicated a community representative in each seringal to guide myself and a research assistant in order to safe find our way through forest trails, to select particular households, and to provide us with invaluable historical background of the community. 3.3 Survey methodology Most of the data for this research was collected through a survey. The survey was conducted orally and carried out using structured interviews. Although the survey was applied at the household level, it targeted land use information at the household, community, and reserve levels. The questionnaire covered a broad range of questions regarding land use patterns, income, extractivist and agricultural production, migration dynamics, demography, transportation, social organization, and reser ve wide management efforts. I and a research assistant, an agronomy student from the Federal University of Acre who also was a certified extension worker applied the survey. The survey interviews lasted about 2 3 hours in each household. Between seven and ten days were spent surveying each seringal. 3.4 Other sources of data The Brazilian Environmental Agency (IBAMA) conducted a census with all the residents of the reserve in 1995, 1998, and 2000. The data was collected as part of the

PAGE 44

32 monitoring strategies of the residents land use activities in the reserve. No one had analyzed the information thus far, although the censuses yield very valuable information regarding temporal changes in land use, population and productivity. IBAMA collected the census data o f the four study seringais in a questionnaire format, which was applied orally to the rubber tappers by extensionists and trained residents. 3.5 Data processing variables and analysis The data collected throughout the survey were first entered into an Ac cess database. Then, the data were entered into an excel database and a codebook was created to facilitate statistical analysis with SPSS program. In the codebook, 128 variables were defined. For the purpose of this Masters thesis, after running several t ests, I selected variables that were related strictly to land use patterns. The other variables will be used for further studies on the Chico Mendes Reserve. To accomplish my analysis, I performed frequency distribution tests, cross tabulation tests, and regression analyses. The frequency distribution and cross tabulation tests considered variables regarding length of residence and previous residence of the household heads, settlement sub divisions and the reasons for these sub divisions, as well as reside nts membership in rubber tapper social organizations. These analyses are presented in chapter 5 together with the analysis of census data of the reserve. In addition, I performed regression tests and created an OLS model. I have defined six dependent a nd six independent variables. The dependent variables were divided into two categories: traditional and non traditional land use activities among tappers.

PAGE 45

33 Traditional Rubber production in 1999 Brazil nuts production in 1999 Non traditional Area deforested. Area deforested here is defined as the permanent or temporary clearing of the forest for pasture and agriculture. I combined two variables (household pasture size and agricultural plot size) into a dependent variable called area deforeste d. For a full picture of deforestation at the household level, other land covers need to be considered. Fallow areas in different stages of regeneration constitute an important component of changes in forest cover. This is problematic because I do not hav e data differentiating 3 year old from 20 year old fallow. Total size of swidden plots in hectares per settlement Total size of pasture in hectares per settlement Total number of head of cattle per settlement The independent variables selected point to some of the key socio economic determinants of deforestation in the reserve and were considered as background of the household head, family labor force availability, and location of the seringais. The following independent variables were used in the analysis: Migrant status. The definition of who is a migrant is not straightforward and many factors can be considered to classify migrants within the reserve. For the purposes of this analysis, I classified as migrants anyone who came from outside the reserve or seringal context, be that former city dwellers or residents that came from another land use areas such as colonization projects or farms. Therefore, if the previous residency was outside a

PAGE 46

34 seringal context, the household head is considered as a migrant. In classifying migrants in the reserve as above, I am aware that I am underestimating several other forms of classification of migrants. The classification considered here is an attempt to identify residents that have engaged in different la nd use practices than those traditionally practiced by rubber tappers. Age of the household head. Number of s ons and daughters of the household head. For this analisis, only children who were fifteen years old or older were considered, as this is t he age when they start to play an important role in the family labor force availability in the settlement. Seringal location. I divided the four study seringais in different categories. S eringal Filipinas was considered the reference category for this analysis: as it most typically represented traditional seringal context. Seringais Humaita and Porongaba were analyzed as one unit since I observed that they have the same land use characteristics, thus not affecting the analysis, and also because they b order on each other. The outcome and independent variables will be further discussed in chapter 5, where they are analyzed. 3.6 Environmental characterization of the reserve 3.6.1 Climate The regions climate can be classified into tropical humid AM, acco rding to Keppens system, and is characterized by a mean temperature between 26 e 27 C, with a short dry season a high precipitation levels. The rainy season lasts from November until April, during which precipitation avera ges amount to more than 110 mm/month, while during the rainy season precipitation averages are less than 93.3 mm/month, July being the driest month of the year ( RADAMBRASIL 1976 ). The highest precipitation can be

PAGE 47

35 observed between the month s of December to March, a period referred to as winter ( inverno ) in the region, while the period between June and September is referred to as summer ( vero ). The average monthly temperature generally varies little between 24 a 26C, with a slight decre ase between the months of June and August. During this period a phenomenon regionally referred to as friagem results in acute drops of temperature to 4 to 6C, caused by an advancement of a polar front from the south (IDEAS 1993) 3.6.2 W atershed/Hydrology The CMERs watershed is represented by two main rivers: the Acre River and the Iaco River, both tributaries of the Purus River (Fig. 3 3). The Acre river watershed, cuts through a large part of the reserve territory and runs from west to east. The meandering river touches the reserve near Assis Brasil and runs along the frontier between Brazil and Peru as well as Bolivia. The river crosses the reserve near a community called Itu The upper Acre river basin is characterized by a strong dis section of relief, forming hills located on the most elevated parts with mild slopes from 3 to 8% covered by podzlicos red yellow soils (RADAMBRASIL 1976). The Acre River has a narrow floodplain and the peaks of the flood are observed between February and April, while July to September are characterized by a hydric deficit. 3.6.3 Geology The geological studies of the region resulting in known descriptions of afloramentos litolgicos (rocky outcrops), as well as mineral and fossil occurrences, are always based on finds on the banks of the main rivers due to the access by boat used by researchers. The CMER is totally composed of cenozoic sediments of the Solimes

PAGE 48

36 Formation, lying on a base of crystalline Craton Guapor, for which so far no outcrop has bee n detected in the reserve area. The recent alluvial with origin in the evolution of Figure 3 3 Map of the mean rivers bordering the CMER. Source: CNPT (1997) the rivers of the actual watershed, include tertiary and quaternary sediments (IDEAS 1993). The Solimes Formation is a result of the decomposition of plio pleistocenica, which is characterized by variations of faciolgicas proeminentes. These are typical sediments of the floodplain geology with characteristic sedimentary structures. These a re part of the following geological composition: solid or stratified argillites (clay) with carboniferous and gypseum concrees; calcite and gypsies veins, occasionally with carbonized plant material; varied argillites with laminated structure (CNPT 19 98).

PAGE 49

37 3.6.4 Pedology (Soils) Maps of the Radambrasil Project (1976), indicate that the reserve territory lies basically on eutrophic soils, which suggest good fertility due to the soil type change capacity, characterized by holding more than 50% of exchange able bases (IDEAS 1997). Approximately 55% of the soils have clay texture and 83% have high clay activity. Such conditions are evidence that the reserves environment has special characteristics with ample potential for production. Among the municipalities that hold territory on the reserve, the municipalities of Sena Madureira and Assis Brasil have more than 90% of its soils constituted of red yellow podzolic eutrophic soils. In some other areas on the extreme north of the reserve, modifications of the soi l profile have been detected with laminar and generalized erosion of the superficial horizon of the soils. Moreover, on other areas of the reserve, a process of slow rastreamento of the soil and localized laminar erosions has been identified. 3.6.5 Veg etation The first inventory and description of the vegetation type of the state of Acre were elaborated by the Radambrasil Project (1976). In the area covering the CMER two phytoecological units were identified: the system of Dense (Closed) Tropical Forest and the system of Open Tropical Forest (IDEAS 1993). The reserve holds only 27% of dense forests, while 73% are areas covered with open forest. Figure 3 4 presents these forest distributions in the reserve. The dense tropical forest is characterized by an understory with a dense layer of shrubs and treelets, and by closed arboreal canopy vegetation. The biggest extension of continuous dense forest is located south of the Xapuri River. Dense forest is classified into two forest types, which are differentiat ed according to different geological and

PAGE 50

38 Figure 3 4 Map of forest typology occurring in the CMER, according to RADAMBRASIL (1976). Source: Sassagawa (1999) geomorphologic properties of the sites, such as: dense forest on terras baixas (lowlands) with un dulating relief and dense forests on dissected relief. Dense terra baixa forests on undulated relief (Fdo ) overlay tertiary sediments and usually reach a canopy height over 30m. In the lower areas, bordering igaraps (creeks), occur seringueiras rubber tr ees ( Hevea brasiliensis L. ), the palms aa ( Euterpe precatoria Mart.), jaci ( Attalea butyracea ), and murumuru ( Astrocaryum murumuru ). In this forest type also occur stretches of lowland forest on dissected relief, in which the typical species is the cere jeira ( Amburana cearensis ) that has a high commercial value and is locally abundant.

PAGE 51

39 In the dense forests on dissected reliefs on ridges and hills (Fda), compositions with smaller trees are found with frequency, in which individual trees are relatively un iform in terms of their height within their forest layer and in which the canopy reaches approximately 30m. The understory in the plateau areas is more open than those situated on the dissected relief and is increasingly dense in the direction of the talv egues, where shrub and palm species can be found in higher abundance (IDEAS). These dense forests occur in three areas: In the northern part of the CMER near the headwaters of the creeks ( igaraps ) Mapinguari,; in the western part, at the headwaters of th e Xapuri river, and in the southern part following the course of the Xapuri river until it reaches the Acre river. The main species found in these areas are: aquariquara (Geissospermum spp.), abiorana (Pouteria spp.), angelim (various Legumes), breu (Tetra gastris and Protium), cariperana, castanheira (Bertholettia), cedrorana, copaiba ( Copaifera spp), louro (Aniba, Ocotea), parapara, pau mulato ( Calycophyllum spruceanum ) e ucuba ( Virola spp.). Open Forest is caracterized by a distribution over lowland cov ering sandy tertiary geology and in the alluvial terrasses of quaternary sediments as well as the submontane relatively high areas on pre cambian ground material. In this forest typology three main types of evergreen open forest types with palms and lianas occur. The first evergreen open forest types with palms and lianas ( Fac) is the forest on alluvial terraces on quarternary geology, covering 7% of the reserve. It occurs on the margin of all the major water flows in the region, however in variable widt h along the banks of the rivers. The forests in that environment can have very large trees. The most characteristic species of this environment are the abiorana seca ( Pouteria sp.),

PAGE 52

40 aquariquara ( Geissospermum sp.), andirobarana, mamorana, munguba, seringu eira and ucuba. Common palms also are: aa, bacaba ( Oenocarpus mapora ) and graminoid species such as the bamboo ( Guadua spp.). The second Open Evergreen Forest with Palms and Lianas (Faa), is the forest of lowland and dissected relief. It occupies 22% o f the reserve area and is located in the margin/edges of residual interflvials. This forest is encountered throughout almost all of the western part of the reserve, limited by a band of Open Forest on Alluvial Terrases and the Iaco River. Its understory i s rich in small palms. In dryer areas and on the margins of residual interfluvials lianas dominate what is called liana forests. In the depression and valleys there is a higher abundance of palms. Among the most characteristic species are the palms aa, i naj (Attalea maripa), jarina (Phytelephas macrocarpa), mumbaca, murumuru, patau (Oenocarpus bataua), paxiubo (Iriartea deltoidea) and paxiubinha (Socratea exorrhiza). Others commons species also include aacu (Hura crepitans), amarelo (Aspidosperma par viflorum), fava de espinho, ings (Inga spp.), matamats (Eschweilera spp), muratinga, seringueira, tauari (Couratari sp.) and ucuba (Virola sp.). The third Open Evergreen Forest with Bamboo on the lowland areas with undulated relief (Fao), covers 44% of the reserve, situated mostly in the area of lowland plateaus and variable drainage. This forest is by far more common than the other types covering most of the central eastern part of the reserve. Three forest communities are part of this forest type: 1) O pen forest with lianas 2) Open forest with palms and 3) Open forest with bamboo.

PAGE 53

41 CHAPTER 4 DEFORESTATION TRENDS IN THE EXTRACTIVE RESERVE 4.1 Introduction This chapter introduces land use changes in the Southwestern Amazonian state of Acre and in the Chico Mendes Extractive Reserve. In particular, deforestation rates outside the reserve and wi thin the reserve will be examined and discussed focusing on the state, municipal, reserve, and seringal levels. In addition, the chapter provides the necessary land use change background on which the following chapters 5 and 6 base their discussion of hous ehold determinants of land use activities and keys socio economic drivers of this change in the CMER. The eastern part of Acre, particularly the Acre River Basin, lies at the westernmost limit of the development frontier that experienced massive changes in land use since the 1970s. The Chico Mendes Extractive Reserve lies just at this limit of the frontier (figure 4 1). To understand deforestation patterns within the Extractive Reserves, one also has to look at land use and deforestation rates in the surrou nding areas of the state. First, land conversion and encroachment into rubber tapper areas were one of the reasons for the mobilization of the rubber tapper movement and the creation of the reserves. Second, until this day, land uses surrounding the reserv e influence land uses in the reserve to a certain extent, as there is a lot of social interaction along the boundaries. Third, by comparing deforestation rates and patterns in the reserve with deforestation outside the reserve can one assess the effectiven ess of the extractive reserve in maintaining forest cover.

PAGE 54

42 ACRE CMER BRAZIL w 69 o 30 w 69 o 00 w 68 o 30 w 68 o 00 s 11 o 00 s 10 o 30 s 10 o 05 Xapuri Brasilia Assis Brasil Figure 4 1 Location of the Chico Mendes Extractive Reserve in Acre 4.2 Deforestation and Land Use Trends in the state of Acre Acre was almost totally covered by dense tropical forest before the 1970s when most of its economy was based on rubber and Brazil nut extraction. According to the Landsat satellite data, in 1978 Acre had only 2,500 square kilometers of deforested land. By 1988, after the onset of larger de velopment efforts, deforestation reached a total area of 8,900 square kilometers (INPE 2000). This relatively low figure, compared to other Amazonian states, does not seem to give cause for alarm. What is significant, however, is that the area cleared down between 1978 and 1988 tripled in just ten years.

PAGE 55

43 According to FUNTAC (1996), which analyzed deforestation at the state level, the increase in deforestation in Acre has been greater in the upper and lower Acre basin, the eastern part of the state, which c omprises approximately 86% of the deforestation in the entire state. In fact, half of this deforestation until 1996 occurred in the six eastern municipalities of the state: Rio Branco, Senador Guiomard, Brasilia, Bujari, Placido de Castro, and Xapuri (in decreasing magnitudes), with between 51% and 17% of these municipalities deforested. While this region experienced most of the cattle ranching expansions it also concentrates a large portion of the extractive population. Deforestation from forest conversi on for agriculture and cattle pasture generated sometimes violent social conflicts as the completely diverse social and cultural standards of production and organization clashed. As extractive activities, mostly rubber and Brazil nut extraction, were almos t abandoned, the living conditions of the rubber tappers, who relied more heavily on forest resources, were severely compromised. Therefore, this conversion caused a significant rural exodus by the landless poor, many of them rubber tappers. As a result of the clashes between, particularly, cattle ranchers and resisting seringueiros the rubber tapper movement was born, struggling for land, traditional livelihoods, and socio economic development. Thus today, Eastern Acres extractive reserves, the major ach ievement of this movement, are islands nested within and along the development frontier on otherwise prime development land. The greater rate of deforestation in the upper and lower Acre River region is directly related to the transition from tradition al non timber forest production to ranching and farming. Since the 1970s, deforestation rates have increased at quite high rates, threatening the vegetation cover and forest of the state already compromised by about

PAGE 56

44 15,000 squares kilometers of clearings i n 1999 (See Figure 4 2) 12 Concurrently, Acres rubber production, which was the basis of the states economy, has decreased drastically. This occurred not only as a result of the deforestation in this region, but also as a result of national and regional d evelopment policies that favored cattle ranching and large scale agriculture over extractivism, and the declining wild rubber market. Figure 4 2 demonstrates that there is a direct inverse relationship between deforestation and rubber production in the sta te of Acre, illustrating how extractive activities (rubber) gradually lost importance and large scale land use activities were prioritized. This inverse relationship tends to continue until today, although rubber production has increased a little in the la st two years due to a new local government subsidy intended not only to provide rubber tapper communities with better income, but to decrease deforestation rates as well. 13 The regional and local socio economic factors that have affected land use change in Eastern Acre also have directly and indirectly influenced land use change within extractive reserves. The Chico Mendes Extractive Reserve in particular is rooted in the history of the grassroots movement for forested land but also in the history of land u se change in the region. 4.3 Deforestation rates in the CMER The CMER, with nearly one million hectares, borders the highway BR 317 leading to the Pacific through Peru. Although the CMER has served to impede large scale 12 The deforestation rates between 1979 and 1988 were estimated by averaging due to the lack of data 13 The Lei Chico Mende s or subsdio da borraha was the first effort of the government of the Forest elected in 1998 to strengthen extractivist communities by boosting rubber production. It provided an additional 0,20 US $ for every kg of rubber for every rubber tapper in a production association. In addition to making it worthwhile for many seringueiros to tap rubber again, it was intended to strengthen and encourage the rubber tapper social organization (Governo do Acre 1999).

PAGE 57

45 0 2000 4000 6000 8000 10000 12000 14000 16000 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 0 Years km 2 0 1 2 3 4 5 6 7 8 9 10 11 12 1000 tons Deforestation [km2/year] Rubber production [1000 tons/year] Source:Anurio Estatstico da Borracha (1986,1987,1988,1989 e 1995); IBAMA 1998; SEFE 2000; INPE, 1999. Figur e 4 2 Deforestation and rubber production in Acre (1979 2000) deforestation along this highway, it has experienced accelerating small scale deforestation since it was established (Figure 4 3 and 4 4). It is predicted that within a decade about 12% of the r eserve area will exceed the legal limit for deforestation of 10%, as determined by the Utilization Plan 14 (Brown 2001). In addition, the higher occupation of some parts of the reserve, mostly the southwest part, may cause land clearing to exceed the 10% lim it for those areas in a few years. Nonetheless, despite increasing deforestation within the reserve areas, compared to levels of deforestation outside the 14 The Plan sets forth an upper limit of 10% of deforested area per household (of an estimated 300 400 hectares), including residential clearings, backyards, pastures, agricultural and abandoned fields, as well as agroforestry plots. The plan also includes strict regulations of the extraction of rubber and Brazil nuts and the development of management plans for new forest products. Timber extraction and hunting is restricted to the residents subsistence use. The Plan defines as common use areas, rivers, lakes, main paths, and beaches, although reside nts close to these environments tend to dominate their access. The resident associations have authority to participate in the monitoring process of the plan and a team of resident woodsmen were trained and given the legal status of environmental inspection agents.

PAGE 58

46 Figure 4 3 Satellite image mosaic of the CMER illustrating deforestation in and out of the reserv e. Source: Sassawaga 1999 s 11 o 00 s 10 o 05 w 68 o 0 0 w 69 o 30 0 30 60 Km Assis Brasil BR 317 BR 317 AC 403 Brasilia Xapuri N Border of the Chico Mendes Reserve Desforestation1986 Road Legend: 003/6 002/6

PAGE 59

47 Figure 4 4 Satellite image mosaic of the CMER illustrating deforestation in and out of the reserve. Source: Sassawaga 1999. Border of the Chico Mendes Extractive Reserve Defo restation: a) 1992; b) 1998. Road Legend: s 11 o 00 s 10 o 05 w 68 o 00 w 69 o 30 0 30 60 Km Assis Brasil BR 317 BR 317 AC 403 Brasilia Xapuri N 003/6 002/6 7 s 11 o 00 s 10 o 05 w 68 o 00 w 69 o 30 0 30 60 Km Assis Brasil BR 317 BR 317 AC 403 Bras ilia Xapuri N 003/6 002/6 7 a) b)

PAGE 60

48 reserve these are relatively small, particularly compared to areas along the BR 317 highway, which is do minated by large scale cattle ranching activities. Figure 4 5 shows the percent of overall deforestation in three periods of time that Sassagawa (1999) determined for the CMER. The mean annual deforestation rates within the reserve were 0,137% a year for t he 1986 1992 period and 0,227% a year for the 1992 1998 period. This shows that deforestation rates of the second period nearly doubled compared to the first one. Figure 4 5 Percent of deforestation in the CMER (1986, 1992 and 1998) Source: Sassagawa (19 99) 4.4 Deforestation at the seringal level Within the extremely large CMER Reserve there is great variability in deforestation levels. Some areas have levels as low as 0.02 % (1998) while others are approaching 8% (1998) almost reaching the legal limit o f 10%. Table 4 1 shows the different rubber tapper estates or seringais of the reserve, the area that they cover and the proportion of that area that was deforested in 1986, 1992 and 1998, respectively. 0.72 1.54 2.9 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1986 1992 1998 Years Deforestation (%)

PAGE 61

49 As noted before, there is a great variability within the reserve as can be seen by the very different deforestation rates that the seringais have experienced. Some, mostly more remote areas such as the Seringal Arari and Seringal Petrpolis have extremely low levels of deforestation, whereas Fazenda Carmen has surpassed the legal limit of 10%. Table 4 1 Total area and percentage deforestation in each seringal of the reserve within the three associations of the municipalities Deforestation/Seringal (%) Seringal rea (Km 2 ) in CMER 1986 1992 1998 Assis Bra sil (AMOREAB) Seringal Guanabara 707,69 0,61 1,67 2,75 Seringal Icuri 644,05 0,54 1,38 2,27 Seringal Paraguass 195,11 0,62 1,62 6,70 Seringal Petrpolis 121,68 0,02 0,02 0,02 Seringal So Francisco 299,33 1,12 2,12 3,27 Brasilia (AMOREB) Fazenda Carmem 70,04 2,43 4,31 12,81 Fazenda Porvir 46,80 0,62 0,85 3,63 Seringal Nazar 128,22 0,26 0,92 2,32 Seringal Amap 513,85 0,79 1,54 2,50 Seringal Apodi 191,17 0,47 1,54 2,43 Seringal Canamari 69,70 0,88 0,92 1,25 Seringal Humait 144,87 1,13 2,93 7,95 Seringal Nova Olinda 358,32 0,06 0,14 0,47 Seringal Pacuara 69,59 0,89 2,04 2,56 Seringal Pindamonhangaba 99,05 1,71 2,46 3,63 Seringal Porongaba 89,57 2,27 4,04 7,15 Seringal So Cristvo 165,81 0,55 1,26 2,98 Seringal So Salvador 62,83 1,40 2,34 3,42 Seringal Tabatinga 759,88 0,24 0,49 0,56 Seringal Triunfo 103,17 1,40 3,57 5,08 Seringal Vale Quem Tem 25,14 0,00 2,31 3,50 Seringal Vrzea Alegre 83,43 1,17 3,03 4,44 Xapuri (AMOREX) Fazenda Bonfim 329,40 0,15 0,53 1,38 Fazenda Ana Cludi a 46,83 0,23 0,81 0,96 Fazenda Filipinas 442,21 1,96 3,62 7,55 Seringal Arari 104,67 0,00 0,00 0,06 Seringal Curitiba 120,75 0,03 0,07 0,07 Seringal Fronteira 339,89 0,37 0,64 1,05 Seringal Remanso 54,05 0,02 0,44 0,91 Seringal So Jos 95,65 0,26 1, 08 2,05

PAGE 62

50 Table 4 1 Continued Seringal Sai Cinza 63,38 1,47 2,46 2,86 Seringal Venezuela 115,41 0,55 2,97 4,80 Seringal Albrcia 255,00 0,25 1,04 2,68 Seringal Barra 183,44 0,18 0,39 0,92 Seringal Boa Vista 315,69 0,29 0,86 2,46 Seringal Dois Irmos 6 8,88 0,99 1,09 3,91 Seringal Floresta 70,09 1,63 4,58 7,35 Seringal Independncia 120,43 5,20 6,02 8,1 Seringal Lua Cheia 54,03 1,11 3,09 3,83 Seringal Nazar 273,06 0,71 1,24 2,50 Seringal Palmarizinho 87,29 1,13 2,26 4,83 Seringal S. F co do Irace ma 456,78 0,37 1,09 2,23 Seringal S. Joo do Iracema 41,27 5,38 7,87 7,87 Seringal So Pedro 143,99 0,22 0,76 2,06 Seringal Sibria 227,54 0,50 1,29 2,48 Seringal Vila Nova 248,76 0,82 2,09 4,30 Source: Sassagawa (1999) Generally, the seringais with h igher deforestation rates are situated at the boundaries of the reserve and in relatively close proximity to Xapuri or Brasilia and the highway (See Figure 4 6). Another area of increased deforestation is in the municipality of Assis Brasil, where there i s more agricultural activity and Brazil nut tree density is extremely low, limiting the extractivist potential of the area. In contrast, many seringais toward the middle or back of the reserve are very little impacted. It is important to note that the se numbers represent overall deforestation per seringal and do not always reflect similarities across these areas. Some seringais with very high levels of deforestation may have largely forested land in one portion of the area whereas some seringais can ha ve large deforested areas although the seringal as a whole is still forested. Therefore, a certain caution needs to be exercised when evaluating these numbers. Another important point of consideration is that prior to the creation of the reserve, some seri ngais of the current extractive reserve had already experienced a small

PAGE 63

51 Xapuri Assis Brasil Brasilia R i o I a c o R i o A c r e Rio Acre Rio Xapuri B R 3 1 7 BR 317 1 2 3 4 5 6 7 8 W 69 W 68 S 10 S 11 Figure 4 6 Map of the eight seringais with the highest deforestation rates in the CMER percentage of deforestation, as evidenced by the 1986 rates This is particularly true for the areas close to the highway and towns, such as the Fazeda Carmen Seringal Floresta and Seringal So Joo do Iracema thus setting a different trend for later land use in these areas. Table 4 2 shows only the seringais t hat have reached the highest deforestation rates since 1986, which are the areas of particular interest to this study The four seringais highlighted in yellow are the ones selected as study seringais. All of them are similar in the level of deforestation and by 1998 they were all reaching the legal deforestation limits of 10% per settlement in the reserve. If the same deforestation rates continue, these four seringais will reach their legal limitation of deforestation in the next few years. The Fazenda Car men which has even surpassed this limit, is currently even in discussion Legend: 1 Paraguau, 2 Humait, 3 Porongaba, 4 Carmem, 5 Filipinas, 6 Floresta, 7 Independncia, 8 So Joo do Iracema

PAGE 64

52 with authorities about being turned into a traditional colonization project (personal communication with Josmar A. Caminha, CNPT IBAMA 2001). Table 4 2 Seringais within the CMER wi th the highest deforestation rates Deforestation/Seringal (%) Seringal (with over 6% of deforestation) 1986 Rate in 86 92 1992 Rate in 92 98 1998 Prediction for 10% (Years) Seringal Paraguass 0,62 0,17 1,62 0,85 6,70 4 Fazenda Carmem 2,43 0,31 4 ,34 1,42 12,81 Seringal Humait 1,13 0,30 2,93 0,84 7,95 2,5 Seringal Porongaba 2,27 0,30 4,04 0,52 7,15 5,5 Seringal Filipinas 1,96 0,28 3,62 0,66 7,55 4 Seringal Floresta 1,63 0,49 4,58 0,46 7,35 6 Seringal Independncia 5,20 0,14 6,02 0,35 8, 10 6 Seringal S.Joo Iracema 5,38 0,43 7,87 0,00 7,87 ** Exceeded the 10% allowed; ** Presented a 0% deforestation rate in the 92 to 98 period. Source: Sassagawa (1999). Particularly in recent years, the communities of the seringais with higher defor estation no longer have the historical close relationship with forest resources, and have developed their agricultural and pastoral activities more intensely. The 1992 1998 figures demonstrate this increasing change in attitudes towards land use, mirroring the trends outside the reserve, though on a smaller scale. The increased adoption of these new land uses; however, might become a potential source of conflict between these practices and the Utilization Plan regulations. Apart from deforestation levels, t he four study seringais also have other characteristics in common which might be determining factors in the increased deforestation levels. All of them, excluding Filipinas, are accessible via roads and are

PAGE 65

53 located near cities. In addition, the outer limit s of the reserve in these areas border on or are close to cattle ranches and colonization projects. However, they differ in area and population size, as well as some land use activities. By more closely examining these seringais, this thesis hopes to eluci date some of the factors that have led to this increased deforestation in parts of the reserve. Chapter 6 will consider what factors at the household level might affect land use patterns in the four seringais. These characteristics include head of househol d traits, family composition, and rubber estate location as determinants of land use activities measured in six outcome variables.

PAGE 66

54 CHAPTER 5 HOUSEHOLD DETERMINANTS OF FOREST AND NONFOREST ACTIVITIES 5.1 Introduction Recently, studies of land use processes at the household level have emerged as an important component in understanding social deforestation dynamics in Amaznia ( Brondizio and Siqueira 1997, Perz 2001a,b) This chapter provides a household level approach, measuring some characteristics as determinants of land use activities in four seringais of the Chico Mendes reserve. The central question of t his chapter concerns how the origin and age of the household head, family composition, and location of a household affect land use choices in the four study seringais. If this question is answered we will understand much more about these diversified land u se activities in these seringais and their implication for forest conservation. The results might also delineate future trends of land use in these more impacted areas of the reserve as well as other areas that can follow this trend. Thus, this is importan t for basic policymaking regarding land use regulation and economic development for these rubber tappers communities. 5.2 Characterization of Rubber Tappers Settlements in the CMER In the Chico Mendes reserve there are 46 seringais which are historic ally determined rubber tapper estates. Each seringal is composed of approximately 10 25 or more household units, the colocao 15 or family settlement. Before the creation of the 15 This term comes from the verb colocar meaning to place or to put. The colocao or placement refers to the area where a rubber tapper was placed or colocado in the forest b y the boss during the rubber boom.

PAGE 67

55 reserve, these units were not formally managed for multiple uses, but rubber ta pper families have managed them for rubber and Brazil nut extraction for generations. Today, with the diversification of rubber tapper production systems, it is exactly at this landscape level that most of the new land use decisions are made and carried ou t. Figure 5 1 shows a typical modern settlement in the CMER. The settlements comprise a residential clearing associated with backyards and small open fields; one or more agricultural plots at some distance in the forest; extractive rubber tapping trails ra diating in loops from the residential clearing; a gathering and hunting territory delimited by trails; and a pasture area, mostly resulting from old agricultural areas that were not left to fallow. The size of these settlements ranges from 300 to 800 hecta res and is informally determined by the number of rubber tapping trails that historically defined the productive unit of the settlement. In other words, settlement territories have no physical boundaries that can be mapped precisely. Instead, the attention is concentrated on the natural resources they contain, measured by the rubber trails and the rubber and Brazil nut trees in that area. As noted, rubber tapper settlements cover a variety of land use activities that are distributed seasonally throughout th e years and require differentiated family labor force allocation. The next section will present the allocation of the rubber tapper major land use activities. 5.3 Seasonality The Rubber Tappers Year A good way to understand factors affecting rubber tappers production is to learn about their main land use activities throughout the year. Although production systems shared by rubber tappers are similar, there are differences. How much time a family

PAGE 68

56 spends on rubber tapping, on Brazil nut collecting, agr iculture, livestock raising, and hunting varies greatly between families and by season. In addition, the patterns of Figure 5 1 Representation of the settlement in the CMER work spent in each activity cannot be studie d in isolation. The outcome from one activity can only be analyzed in the context of another activity in order to understand how each operation influences each other. Also, understanding household members multiple roles and responsibilities is crucial to the study of land use activities. Figure 5 2 presents a seasonal calendar of rubber tapper activities throughout the year. The principal seasonal influence on land use activities is that of the rainy and dry seasons. Forest production activities complem ent one another seasonally. Rubber is harvested in the dry season (April to August) while Brazil nuts are collected in the winter (December to March). Since labor is an important limiting factor in the tappers

PAGE 69

57 production system, the fact that rubber and Brazil nuts do not compete for labor is very beneficial and important to tapper families. Activities J F M A M J J A S O N D Hunting * * * * * * Livestock raising * * * * * * Rubber trail clearing * Rubber harve sting * * Brazil nut harvest * * Land clearing * Field burning * Rice planting * Corn planting * Beans planting * Manioc planting * Rice harvest * Pasture planting Corn harvest * Bean harvest * Manioc floor * * * Dry season * * * Wet season * * * Key: = Period of activity/occurr ence Figure 5 2 Seasonality in the rubber tappers year Land clearing for crop production is performed from July to August. Tappers choose an area close to the house that is free of Brazil nut and rubber trees. Trees and understory must be cut on the land to be cleared. This referred to as the broca and derrubada. Broca refers to cutting vines and other small brush. After this is done, the tappers performs the derrubada which is to cut the larger trees with an ax or a chainsaw usually provided by the Association. After that, the vegetation is on the ground, drying for at least three weeks so that it will dry well enough to burn well, which generally is

PAGE 70

58 carried out in September. The land clearing process and rubber tapping need to be done in the dry sea son and are carried out by the adult males. These concomitant high labor activities cause a constraint in the familys labor force availability. Corn and rice are then planted intercropped at the beginning of the rainy season (September and October). T hese crops are harvested in February and March. Beans are planted in April or May, at start of the dry season, and harvested in July and August. Manioc is planted from October to December. Manioc is used to produce flour (farinha ) and is harvested as nece ssary. Following this planting period there is much weeding of these crops to be done in the wet season. Following the second year, the plot will be left in fallow and then used in cultivation again after a few years, will be transformed into pasture or le ft to regenerate into forest. Usually all adult family members plant, weed and harvest. Women help in planting and harvesting. Besides planting and harvesting, young sons and daughters are involved in the weeding. Livestock raising is carried out througho ut the year. This includes chickens, pigs, goats, sheep and cattle. Women usually take care of most of the livestock, excluding cattle, which is often the responsibility of the sons or heads of households. Hunting continues throughout the year for most rub ber tappers. The head of household usually undertakes this activity, but in the busiest times of the year tappers often have no time to hunt and thus send their oldest sons instead. In general, rubber tappers activities are very well distributed througho ut the year. However, the period that requires the most family labor is between June and October. During this period the family members are harvesting rubber, planting and weeding crops, clearing new agricultural land and taking care of livestock. This per iod of labor

PAGE 71

59 stress limits how much a family can dedicate to any one activity, and ultimately decisions need to be made as whether to invest more into rubber tapping, agriculture, or pasture creation. In the following section I will statistically compare t he different forms of land use with some household determinants, an important one of which is labor availability. 5.4 Conceptualization and Operationalization of Outcome and Independent Variables Rubber tapper activities throughout the year involve a seri es of complementary activities. Four main activities can be defined for rubber tapper production systems: rubber tapping, Brazil nut collection, livestock raising, and crop production. Since these activities have different influences on deforestation, the ones with low impact on the forest were conceptualized as traditional forest activities, and the ones with high forest impact were conceptualized as non traditional forest activities. I consider six land use outcome variables and a set of other independent variables that influence rubber tapper land use activities. Table 5 1 presents outcome and independent variables names, their operational definition and their descriptive statistics. 5.4.1 Outcome Variables The first two outcome variables, rubber and Br azil nut production were conceptualized as the basis of the traditional forest activities because they have low forest impacts and represent the traditional forest based economy practiced by rubber tappers. I operationalized rubber production in terms of k ilos and Brazil nut in terms of latas. 16 The average finding for rubber production per household is 268 kilos per year. It represents a very low rubber production per household, but the distribution is skewed, as indicated by 16 Rubber tappers usually measure Brazil nut by latas, or eighteen to twenty liter cans of cooking oil. One lata of Brazil nut is equal to 13 kilos.

PAGE 72

60 Table 5 1 Variable names, operational definition, and descriptive statistics, CMER Variable Name Operational defi nition Mean Standard Deviation Outcome variables Forest and nonforest activities Rubber Total rubber production in 1999 measured in kilos 268.6061 418.3590 Brazil nut Total brazil nut production in 1999 measured in latas 227 .2424 284.9748 Area deforested Total size of pasture and swidden combined in hectares 9.3788 10.9492 Swidden size Total size of swidden plots in hectares per settlement 2.2803 1.5863 Pasture size Total size of pasture in hect ares per settlement 7.0985 10.0641 Heads of cattle Total number of head of cattle per settlement 8.3485 10.3247 Independent variables Household Background Migrant Residents that came from the city or other locations outside t he reserve, such as colonization projects and ranches .3333 .4750 Age Age of the head of household 42.3788 12.9398 Labor force availability Sons Number of sons in the household aged 15 years or older 2.0308 1.9683 Daug hter Number of daughters in the household aged 15 years or older 1.6154 1.6554 Seringal location (reference category seringal Filipinas) Seringal Paraguau .3030 .4631 Seringais Humaita/Porongaba Seringais Humaita and Porongaba w ere tied together since they present the same land use characteristics .4091 .4953 n = 66

PAGE 73

61 the large standard deviation. It reflects that some households do not tap a ny rubber, while other households still have a considerable rubber production. Brazil nut production averages 227 latas per household per year. In general, these traditional forest activities findings show that rubber tapping activities have limited import ance for many households, although they are living in an extractive reserve where rubber is still supposed to be a key part of the economy. The other four outcome variables, area deforested, swidden size, 17 pasture size and number of cattle heads, I classi fied as non traditional forest activities among rubber tappers. All these variables represent important activities increasingly developed by rubber tappers in the study seringais and are important indicators of the impact on forest cover and deforestation in the reserve. In addition, the way these variables behave has important development and political implications for rubber tapper communities settled in extractive reserves. The area deforested variable is an important measurement of household forest imp act, which has several implications in terms of land use regulation in the reserve, such as the one defined in the utilization plan that sets forth an upper limit of 10% of deforested area per household. In addition, the area deforested variable is further implicated in terms of the cultural identity of tappers as an extractive population. I conceptualized area deforested as the permanent or temporary clearing of the forest for pasture and agriculture. 18 I operationalized this variable combining household pa sture 17 Note that swidden agriculture to some extent is also a traditiona l activity for rubber tappers, mostly for family subsistence. 18 I considered these two variables as a mean of measuring area deforested in the settlement. However, they have different ecological impact on forest cover. Swidden might be left to fallow and t hus have less ecological effect on forest cover than pasture, which usually has greater impact, and is more permanent forest clearing.

PAGE 74

62 size and agricultural plot size. The mean household area deforested area was 9 hectares per household. Swidden agriculture is relatively traditional among rubber tapper because it was practiced as a subsistence activity. This has a relatively small f orest impact because small areas of forest are cleared annually. In addition, the impact is low because a rotation system is practiced by tappers whereby the same plots are used for two or three years and then left in fallow to be used again after a few ye ars. This rotation, if conducted properly, may guarantee sustainability. However, I conceptualized swidden agriculture as non traditional forest activities because, more recently, swidden agriculture has increasingly had greater forest impact in the study s seringais. 19 Since agriculture has become an important source of family income, many families have increased the size of swidden agricultural annually, and swidden areas that before were left as fallow are increasingly turned into pasture. I operationaliz ed the swidden agriculture in terms of the size of the swidden plot in hectares per household. The descriptive statistics show that the mean swidden size was 2.2 hectares per household each year. The pasture outcome variable represents the newest land use among rubber tappers. A n accelerated increase in pasture areas has taken place within the study seringais and, thus, p asture represents the land use with the greatest impact on forest cover. A recent perception among rubber tappers in these seringais is t hat the conversion of land into pasture will increase the value of their settlement. This indicates their plans for future land use choices and has a series of implications in terms of the future economic and ecological viability of their reserve as well as in terms of their very identity 19 Traditionally, a tapper family usually has a hectare of agricultural land that is used for family consumption throughou t the years. In the studied seringais, I observed that many families have increased their swidden plot at least with another hectare for cash crop agriculture. Though I did not distinguish these families in the analysis, I recognize that it would have been important to look at the families separately.

PAGE 75

63 as rubber tappers I operationalized the pasture variable in terms of the size of the pasture area in hectares per household. The mean pasture area was 7 hectares per household. Likewise, the cattle outcome variable is an other important indicator of new economic activities for the rubber tappers. It was agreed upon in the utilization plan of the reserve that families could have a small number of cattle, just to improve their daily diets and transport their products. Howeve r, in the study seringais, cattle raising is the activity showing the greatest increase recently. Some households in the study seringais have no cattle, while others have over 30 heads. It seems that most of the tappers profits in the studied seringais ar e invested in cattle raising. These cattle raising expansion phenomenon has been perceived as a savings account, an investment that can be easily changed into cash. Again, the increase of this economic activity among rubber tappers has implications in te rms of the their concept of economic development and the conservation goal of their extractive reserve. I operationalized the cattle outcome variable in terms of the number of cattle heads per household. The mean number of cattle heads per household was 8. 5.4.2 Independent Variables The production system developed and managed by the rubber tappers can be, to a great extent, characterized in socioeconomic (e.g., age, family, labor force availability, price) and biophysical terms (e.g., location of the hous ehold, distance to market, resource availability). The independent variables consider three groups of two variables each that affect the production systems: household head background (migrant status and age), labor force availability (sons and daughters of the household head) and location of the study seringais.

PAGE 76

64 In addition to these characteristics, a series of other variables 20 were considered, but the ones discussed here were noted to be most significant to the model discussed later in this chapter. Moreo ver, the limited number of cases in my sample (66), lowered the likelihood for statistical significance of these other independent variables. The household background involves two indicators: migrant status and age of the household head. I operationalized the migrant variable according to previous place of residence of the household head, which might allow for the assessment of difference in land use practices. Migrant household heads might be more engaged in non traditional forest activities than non migr atory household heads. The age of the household head might also be an important indicator of the economic activities they have engaged in throughout their life cycle. Older household heads might be more concentrated on traditional forest activities (e.g., rubber tapping), while younger household head may be more engaged on non traditional forest activities (e.g., agricultural crops). The descriptive statistics reveal that 33% of the head of the household were migrants. The mean age of household heads was 4 2 years old. The rubber tappers extractive system varies, depending upon how large the family is and upon how dependent the family is on income producing activities. The family itself provides most of the labor on the landholding of a tapper family. Thus, I conceptualized labor force availability in terms of the number of sons and daughters and operationalized it in terms of their age. The sons and daughthers considered for this 20 Among these are: length of the residence; the total number of the residents in the settlement; how old the settlement is and how many tappers have occupied it. Also, I considered household membership in rubber tapper associations, rural workers union and cooperatives. Moreover, I considered household head access to government credit. In terms of location, I considered the distance to market measures and different transportation means (walking time, anima l time and truck time) and types of access to the market (trail, ramal and road). Some of the variables that were theoretically important, but did not show stronger statistical significance were excluded. Others, that were less theoretically important, but presented a much stronger significance, were kept in the model.

PAGE 77

65 analysis are the ones who are fifteen years old or older. I considered that sons/daughters over this age start to make an important contribution to the family labor force and thus influence productive land use actitivities. Besides the household head, the most important factor for labor force is the number of sons of age fifteen and older to tap rubber or to help clear land. This is because the availability of this manpower influences the area of land the family can exploit for rubber or put into pasture or field crops. For example, rubber and Brazil nut activities require less la bor input, while agriculture activities require more labor input. As noted in table 5 1, the household head had on average 2.03 sons of age fifteen years old or older, and 1.62 daughters of age fifteen years old or older. There are structural forces outsid e the household level caracteristics that are important in determining land use activities. The location of the household itself in each seringal might present some degree of influece in land use activities. However, the location of the seringal as a whole is a more poweful structural factor. Therefore, I conceptualized location at a seringal level. The core aspect here is that the location of the seringal influences land use activities since it involves aspects of market distance and transportation costs t hat may reduce some land use activities and stimulate others. A tapper settled in a seringal closer to towns often presents different activities from the tappers who live farther. For example, more distant seringais tend to influence tappers to emphasize r ubber and Brazil nut activities with high value relative to weight and bulk over less easily transported agricultural products. The four study seringais present different characteristics in terms of location. Seringais Paraguau, Humaita and Porongaba are road accessed seringais, while seringal Filipinas is far way from the road and is easier accessed by river in the Xapuri region. It

PAGE 78

66 is important to analyze land use activities in the seringal Filipinas compared to the other three seringais. I classified t he s eringal Filipinas as the reference category, as it represents most closely the scenario of the traditional seringal, being farther from the road, and having the bigger rubber production. In addition, I operationalized Seringais Humaita and Porongaba a s a unit due to the fact that they border each other and I observed that they have the same land use characteristics, thus not affecting the analysis. The descriptive statistics show that of the 66 households in my sample, 30% were located in seringal Para guau, 40% in the seringais Humaita and Porongaba and 30% in the seringal Filipinas. 5.5 Correlations Between Outcome and Independent Variables Table 5 2 presents correlation coefficients between each land use outcome variable and every independent varia ble. Traditional forest activities variables have a strong correlation with each other. That is to say that households that tap rubber also tend to collect Brazil nuts. However, they differ when correlated with nonforest activities. Pasture and cattle have a strong negative relationship with rubber production, but this is less true for Brazil nuts, and these coefficients are not statistically significant. Also, Brazil nuts have a positive association with swidden agriculture and thus households that collect Brazil nuts tend to have swidden plots, but do not necessarily have pastures or cattle. In general, the traditional and non traditional forest activities outcome variables exhibit the expected correlations. As anticipated, the non traditional activities have strong positive correlations with area deforested. It is not surprising that area deforested has a strong correlation with swidden and pastures as the area deforested variable is equal to swidden plus pasture area. Similarly, area deforested is stron gly associated with cattle. Since most of the area

PAGE 79

67 Table 5 2 Correlations between traditional and non traditional outcomes variables and household background, labor force availability and seringal location variables, CMER. Variables name Rubber Brazil nut Area deforested Swidden Pasture Cattle Outcome variables Rubber 1.00 Brazil nut .303* 1.00 Area deforested .190 .131 1.00 Swidden .104 .324** .598** 1.00 Pasture .254* .018 .915** .338** 1.00 Cattle .247* .127 .540** .105 .613** 1.00 Independent variables Household Background Migrant .287* .186 .131 .268* .099 .079 Age .208+ .261* .284* .246* .269* .186 Labor force Sons .026 .053 .345** .147 .355** .339** Daughter .077 .089 .103 .149 .047 .063 Seringal location Humaita/Porongaba .331** .392** .205+ .267* .176 .107 Paraguau .312* .782** .022 .24 2* .121 .234+ + p < .10; p < .05; ** p < .01.

PAGE 80

68 deforested is for pasture making purposes, it has a positive relationship with cattle. Also, swidden agriculture has a strong positive relationship with pasture, but not with cattle. This implies that hous eholds have increasingly transformed swidden plots into pasture, even when they do not have cattle. In general, the non traditional forest activities correlations are all highly significant, except for swidden and cattle. The forest activities variables ( rubber and Brazil nut) have distinct correlations with the three groups of explanatory variables. Rubber production is negatively associated with migrants and also a weaker correlation with age, but still suggests that the older the head of household, the more rubber tapping activities are performed. In terms of location variables, rubber production tends not to be done by households that live in the seringais Humaita, Porongaba and Paraguau. In other words, the household head more engaged in rubber prod uction tends to live in the seringal Filipinas. This reflects Filipinass involvement with traditional forest based economy. With Brazil nut activities, there is little correlation with migrant status, but it shows a stronger association with age, where th e older head of the household tends to collect more Brazil nuts. In terms of location, seringais Humaita and Porongaba have a strong correlation with Brazil nuts, while seringal Paraguau exhibits a negative correlation. That is, households that collect Br azil nuts tend to be located in Seringais Humaita and Porongaba. In general, seringal Filipinas has lots of rubber tapping and less Brazil nut collectors. 21 Seringais Humaita 21 Note that although seringal Filipinas is considered the more traditional seringal, it has less Brazil nuts production than seringais Humaita and Porongaba. It might be related to market d istance factors. Distance to town is an important factor determinant of how many Brazil nut a family will collect. If there is no such nearby market, it is usually uneconomical to transport Brazil nuts because they are heavy and require more transportation structure. Seringal Filipinas is far from the road. Households in seringal Filipinas might be not as attracted by Brazil nut collection as the households in seringais Humaita and Paraguau. Having road access, households in seringais Humaita and Porongaba have more transportation facilities for Brazil nut activities. Also, it would be important to analyze other variables that consider ecological factors such as distribution and density, affecting the availabilities of Brazil nut trees in the reserve.

PAGE 81

69 and Porongaba have less rubber tapping and more Brazil nut collectors. Seringal P araguau, on the other hand, has less of both rubber tapping and Brazil nuts collecting. Area deforested is strongly correlated with age of the household head and number of sons. That is, older households and households with sons tend to have larger area deforested. In terms of location, area deforested has a small effect in the seringais Humaita and Porongaba. This means that households deforest slightly more in these seringais than in seringal Filipinas. In general, area deforested is related to the age of the household head and male labor force availability. Swidden agriculture has a strong association with the two househod background indicators: migrant status and older households tend to have more swidden agriculture. In terms of the location, househo lds located in the seringais Humaita, Porongaba and Paraguau tend to have much more swidden agriculture area than households in Filipinas. Both pasture and cattle have a strong positive correlation with sons in the household, which demonstrates their rol e in terms of family labor force allocation. However, these variables differ in terms of age of household head and location of the seringais. Pasture is positively associated with the age of the household head, with older household heads having more land i nto pasture. Cattle, on the other hand, are correlated with the location of the seringais. Only seringal Paraguau shows a correlation with cattle and thus in this seringal there are relatively more cattle raising activities than in the other seringais. 5 .6 Analysis of Multivariable Regression Models Table 5 3 presents the results of Ordinary Least Squares (OLS) models of rubber, Brazil nut, area deforested, swidden, pasture and cattle regressed on indicators of household background, labor force availabil ity and seringal location. As noted earlier, I

PAGE 82

70 Table 5 3 Multivariable regression models at rubber extraction, Brazil nut extraction, area deforested, swidden size, pasture size and head of cattle, CMER. Model Independent variables Rubber (1) Brazil n ut (2) Area deforested (3) Swidden (4) Pasture (5) Cattle (6) Constant 3.858* 2.146* .563 9.832 .507 1.528 Household Background Migrant .1.439 .795 .203 .294+ .222 .344 Age 6.035+ 6.180** 1.248 1.012+ 1.647 1.289 Lab or force Sons .213 7.598 .138* 5.147 .200* .367* Daughter .196 9.571 6.515 8.238+ 4.163 .110 Seringal location Humaita/Porongaba 5.417** .409 .587* .216 .937** 1.525* Paraguau 6.088** 5.870** .247 .226 678+ 1.676* (Reference Filipinas) Adjusted R 2 .452 .638 .199 .191 .201 .185 F ratio 9.781** 19.775** 3.656** 3.519** 3.690** 3.421** + p < .10; p < .05; ** p < .01.

PAGE 83

71 considered several other explanatory variables. The proble m with the multivariable model is that, due to the small sample size, I was only able to set six independent variables. These six variables were selected out of several potential variables that I introduced into the model. They were chosen because they pro ved to be statistically significant for at least one or two outcome variables and had theoretical importance. Other variables that are considered theoretically important but did not prove significant were not included in this model. Also, I used the same s et of variables for all outcomes to facilitate comparisons among the models. In this way, I tried to develop a single set of variables that represented a stronger model overall. After this procedure, it became clear that I had a small number of explanatory variables that were stronger predictors of all the outcome variables. The primary determinants of rubber production (model 1) involved the age of the household heads and location of the seringais. The model for Brazil nut (model 2) had similar significan t variables, but differences are evident. Brazil nut has a stronger age effect, even bigger than that for rubber. The stronger age effect for Brazil nut is because it not only has a more stable market than rubber, but also requires less intensive labor for ce The location variable is less important for Brazil nut than for rubber production. That is to say that rubber has a stronger negative effect with seringais Humaita, Porongaba and Paraguau, while Brazil nut has a negative effect only with seringal Parag uau. This seringal is located in the Assis Brazil region, where Brazil nut trees have lower density, leaving fewer options for residents to increment their forest production economy. None of the labor force availability variables show significant net effe cts on rubber tapping and Brazil nuts activities, which suggests that regardless of household

PAGE 84

72 background and location of the seringais, young residents are not engaged in traditional forest activities. The area deforested model (model 3) reveals a limited array of significant factors. The determinants of having area deforested emphasize the labor force factor and the location variable. Sons of the household head exert a strong and positive effect on area deforested. Therefore, the availability of sons of ho usehold head is important to having area deforested, a reflection of the high male labor demands in forest clearing, harvesting cash crop and processing land into pasture. The location factor has important effect in seringais Humaita and Porongaba and thus there are more non traditional land use activities (swidden and pasture) driving deforestation in these two seringais. In contrast, the determinants of agricultural production (model 4) accentuate the household background variables. Ages of the household head and migrant status are determinants of agricultural production. This basically indicates that older household heads imply having a bigger family and they have tended to expand their agricultural production for family consumption. In addition, this de monstrates that migrant households have been more exposed to agricultural activities, while non migrant households tend to engage in traditional forest activities. Moreover, model 4 presents two important factors. It is the only model in which daughters sh ow effects on swidden agriculture, and location of the seringais present no significant effect on swidden activities. This demonstrates that daughters of household head give important contributions to swidden agriculture activities such as planting and har vesting crops. It also shows that agricultural production is practiced in all these seringais, which demonstrate its fundamental role as a subsistence activity.

PAGE 85

73 The models for pasture and cattle (model 5 and 6) are stronger and have the same significant v ariables. Sons of the household heads exert a strong and positive effect on pasture size and number of cattle. This demonstrates that the availability of sons for labor is crucial to having pasture and cattle and suggests that expansion of nonforest activi ties will continue in these seringais, which tend to be carried out by the young residents. As conceptually proposed, pasture area and cattle raising activities are greater on the non traditional seringais (seringal Filipinas was considered the closest one to the traditional seringal). Thus, it demonstrates that households settled in seringais with better road access and closer to the towns have been more involved in non traditional forest activities. In general, the land under pasture and cattle raising ac tivities depends on the availability of labor and the location of the seringais. In this model, household background variables were unimportant to pasture and cattle, implying that non migrant households are as involved in pasture expansion and cattle rais ing as migrants. 5.7 Discussion This chapter examines how some of the household characteristics and rubber estate location determine land use activities among rubber tappers in four seringais of the Chico Mendes reserve. The findings from these empirical models that account for household background, labor force availability and seringal location show that these factors exert a significant effect on land use activities. Household background variables have effect on rubber and Brazil nut extraction, as well as swidden agriculture. Labor force availability variables have significant effects on area deforested, pasture and cattle. Finally, location of the seringais has important effects on almost all outcome land use variables, excluding the swidden model.

PAGE 86

74 Th e findings of age of household head and labor availability call attention to the age structure of the rubber tapper families, showcasing differences in labor allocation at the household level. It may be that heads of household are more involved in forest a ctivities (rubber and Brazil nut) and the sons with pasture and cattle. Older household heads might have worked with rubber activities throughout their life cycle, and thus, although rubber production no longer has market stimulation like before, they cont inue tapping rubber because it is the practice of which they have the greatest knowledge and with which they identify themselves. If this interpretation is correct, older household heads will insist on continuing this activity if at all possible, even if t hey are not maximizing their income. However, it must be stated that this effect of older household heads with rubber activities might be because rubber activities requires heavy labor and more specialized skills so that the younger residents are not as ap t to develop this activity. Due to their skills with rubber, older household heads might have more profit from rubber activities than non traditional activities. By contrast, the younger residents (sons of household heads) probably are relying less and le ss on a purely extractive strategy and more on farming systems. If the younger residents continue to go through their life with the same land use practices, their future trajectory as rubber tappers will be very different from that of their parents, and we might see increases in the extent of land under pasture in these seringais. The distribution of different activities among members of the rubber tapper families shows that people have to make choices, which may favor one component of production at the e xpense of the others. The land use strategies chosen reflect the families judgments of the available

PAGE 87

75 options, taking into account immediate necessities, market criteria, food security as well as personal preferences. The location of the seringais is an im portant factor differentiating diverse livelihood activities. I assume that the seringais, with the exception of Filipinas, have better market integration, which might lead to better prices for cash crops and livestock. This in turn should increase forest clearance. Of course, the forest clearance decision is based on the expected increase in family income, although I did not analyze data about prices and income. Market integration in the long run may foster even more forest clearance in these seringais, if such integration does not result in increased levels of land use regulation enforcement. While market integration and increasing household income may increase forest clearance, the rate at which that occurs could dramatically change rubber tappers relati ons with forest resource. Given the extremely large territory of the Chico Mendes reserve, there are contrasted interests among rubber tapper communities living in different part of the reserve. Development programs in the reserve tend to benefit these co mmunities that live in seringais located closer to the towns. It seems that the rubber tappers settled in these seringais not only have more development benefits, but also will not be able to enforce the reserves land use regulation, as shown by the incre asing deforestation rates. In this case, reserve development programs and land use regulation might need to be approached as a way to benefit a particular tapper community without affecting negatively the interests of others that also deserve attention. I suggest two strategies that might indicate a role for development activities and land use regulation.

PAGE 88

76 Development initiatives for rubber tappers should first consider the different zones in the reserve. Rubber tapper communities settled in seringais clos est to the cities should receive development incentives to stimulate traditional forest activities. Tappers focused on agriculture and livestock raising should be stimulated to occupy more distant seringais, with the stimulus of having more land for defore station. If these approaches do not have any repercussions on the deforestation scenario of today, at least deforestation levels would be more dispersed, which would have some ecological benefits, and rubber tapper communities would be more equally distri buted in the reserves territory, resulting in less pressure on forest resources. Both are unrealistic suggestions. The deforestation situation in these seringais will pose a serious dilemma for land use regulation in the reserve. The forest reserve quota legally required on private property in Amaznia currently is 80%, although it is in a huge debate for changes in the Brazilian congress. Due to extractive reserves importance as forest conservation tools, rubber tappers are allowed to deforest only 10% of their settlement. Some seringais are increasingly deforested areas and will reach the limit quota in few years. In this context, rubber tappers who are supposed to be forest conservationists might reach the same level of deforestation as other non tradi tional rural workers on private property in Amaznia, such as the colonists. The difference is that tappers have the privilege of holding hundreds of hectares of land. In this context, it will become increasingly hard to justify the extractive reserve mode l, if rubber tapper communities cannot keep low deforestation rates. If these study seringais exceed their deforestation legal limit, they might have to be taken out of the reserve territory and transformed into another land use strategy, and thus more peo ple would have access to land.

PAGE 89

77 Cattle and pasture expansion in these seringais will pose a difficult problem for rubber tapper communities in the near future, and one that will call for political decisions. The initial concept of the reserve, which advoca tes finding ways to alleviate poverty without degrading the environment, could be a guide to understanding the new decisions that rubber tapper communities are making in terms of land use and to formulate development policies that regard tappers modified socio economic development perspective and still guarantee forest cover. So far, there are no major concerns from the Brazilian Environmental Agency (IBAMA) and rubber tapper organizations about the pasture and cattle expansion in the reserve. Some princ ipal factors that fall outside of this analysis that influence the tappers production system are the improvement of states transportation networks, and the emergence of autonomous tappers who are more socially and politically organized, and less dependen t on the traditional aviamento system. Experiencing less of the inhibiting effects of the greater isolation and the strict traditional marketing system as in past decades, many tappers now have growing opportunities and incentives to maximize their product ion. Additionally, extractive lands now have more pressure from alternative land uses such as ranching and agricultural colonization than ever before. Different combinations of land use patterns and diverse social relations are part of the rubber tapper communities of today. A broad range of social and environmental factors determines their diversified production systems. The household characteristics and seringal location factors discussed in this chapter offer important insight into land use strategies at the family level, representing an integral fraction of these determinants. Socio economic criteria, such as access to markets, development initiatives and policies,

PAGE 90

78 population dynamic, land use regulation, and so on, need to be broken down into smaller units so that a complete but detailed view of the actual dynamic of land use is obtained for these seringais and the reserve as a whole. In chapter 6 I will discuss some key land use changes taking place in these seringais in a long term perspective, and t heir role in deforestation levels in the reserve and in the study seringais

PAGE 91

79 CHAPTER 6 DRIVERS OF LAND USE CHANGE IN THE CMER: Population and Settlement Dynamics, Changes in the Production System, and Social Organization 6.1 Introduction In the previous chapter, small scale changes in land use at the household level were assessed in order to identify household factors that drive these land use changes. In this chapter I will explore primarily socio economic factors that drive land use change at a seringal and reserve level. Socio economic drivers of land use in small forest dwelling commun ities are still little understood for large parts of the Amazon Basin, compared to studies of large scale development projects. In most cases, studies with forest dwelling small holders have been unable to incorporate an integrated perspective of changes i n land use cover and particular socio economic factors. In the case of the C MER, although a study has established deforestation rates and patterns (Sassagawa 1999), the underlying reasons for changes in land use and land use decisions have not been examin ed. As the establishment of Extractive Reserves reaches 10 years, it is important to breach this gap and determine what is driving land use change in extractivist communities at the threshold of the development frontier. This is of particular importance be cause the appeal and an often cited justification of extractive reserves lies in low deforestation levels and the conservation of forest resources. Thus, understanding what drives deforestation in those areas that have experienced higher deforestation is a timely and fundamental endeavor, that will hopefully aid in the prediction and planning of land use in the extractive reserve.

PAGE 92

80 The reasons for increased deforestation rates in concentrated areas of the CMER most likely lie in a variety of factors. A broad range of factors influences land use choices and the rubber tappers quality of life to some extent or another. First, with the demarcation of the extractive reserve, more families may have moved into the area searching for secure land tenure. This secure d land tenure has also led people to invest more into long term forms of land use which include increased agricultural activity and cattle holdings and as well as a stronger investment into structural elements of their settlements, such as fences, houses a nd storage facilities Second, recent development projects and credit programs have encouraged residents to diversify their land use activities toward agroforestry and cash crop agriculture as well as small scale cattle raising, activities that result in a increased cleared areas in the colocaes. Changes in the production systems are also a result of decreased investment into the extraction of forest products. As discussed in the previous chapter, the price for rubber has decreased significantly in the l ast decades and many people have abandoned rubber tapping in favor of non extractivist forms of land use. Brazil Nut extraction, though more profitable than rubber, also has decreased. Third, the influence of the rubber tapper social movement and its organ ization for productive activities needs to be stressed, as many new opportunities and contacts are accessed through the various organizations, which increasingly mobilize production processes and link individual residents to larger development and other go vernment projects. Fourth, more localized variables that can affect land use choices are the relative proximity to towns, means of transportation, access to markets, and exposure to cattle ranches and colonization projects, as is the case for the four stu dy seringais.

PAGE 93

81 The present land use patterns and decisions are also embedded in the particular history of each seringal and its role on the development frontier and the rubber tapper resistance. Seringais at the periphery of the reserve that had part of th eir land transformed into cattle ranches or adjacent to those are more vulnerable to fast land use changes. Overall, deforestation in these areas occurs as a combination of many factors, which influence each other. It is impossible to pin down increased de forestation levels as the result of a single factor. All of these variables cannot be studied simultaneously; therefore, I will concentrate the analysis on three main factors that influence land use changes in these seringais. First, I will analyze the po pulation and settlement dynamics in the reserve as a whole and the study seringais, which may lead to an understanding of why these four seringais not only have the highest population growth but also have experienced the most drastic land use changes. Sec ond, I will discuss the changes in the rubber tapper production systems in order to explore the economic factors that are leading to land use changes. Finally, I will analyze the social organization model in the reserve and how it may influence land use ch ange through different levels of organization of the communities, political influence of the rubber tapper leaders, and the Residents Associations in these seringais This chapter will mostly focus on an analysis of the IBAMA census data of the reserve in three different periods (1995 1998 2000). 6.2 Population Dynamics in the CMER According to data from the census carried out by the Brazilian Environmental Agency (IBAMA) in 1998, the total number of residents in the reserve is approximately seven thous and people, of which 44.8% are female and 55.2% are male. 63.3% of the

PAGE 94

82 population is under twenty one years of age, which demonstrates that the reserve has a very young population. Although the reserve has territory in six municipalities of the state, the majority of the reserve population is located in three municipalities along the BR 317 highway: Xapuri, Brasilia and Assis Brasil. Figure 6 1 presents the population dynamic of the reserve in these three municipalities from 1995 to 1998. Xapuri, which com prises an area of 3,208.91 km 2 in the reserve, concentrates most of the residents of the reserve. In 1995, Xapuri had 3,660 residents (58.5% of the total reserve), decreasing to 2,612 (43.3%) in 1998, which represent a decrease of 28.6%. Despite this decli ne, it still holds most of the population of the reserve. On the other hand, Brasilia, which comprises 1,330.30 square kilometers, had an increment in its reserve residents from 1,338 (21.4%) in 1995 to 2,120 (35.2%) in 1998, representing an increment of 36.8%. Although this data suggests that there are significant changes in population levels in the two municipalities, this is in part an artifact of an administrative change of 2 seringais from Xapuri to Brasilia between 1995 and 1998 (personal communicat ion Josmar C. CNPT/IBAMA 2001), demonstrating the difficulties of conducting an analysis at the municipal level Assis Brasil, comprising 2,158.30 square kilometers of the reserve territory, maintained approximately the same population size of 1,300 (21.5 %) residents in the two periods. Overall, population levels in the reserve have stayed approximately the same in the period 1995 1998. Nonetheless, this very general data does not reveal more subtle changes in immigration and exodus in the reserve, as wel l as internal migration from one seringal into the other or from one colocao to the other. A true understanding of this dynamic

PAGE 95

83 can only be gained by examining the records of all the settlements in the reserve over time, or at least looking at particular seringais, as does this study. 0 500 1000 1500 2000 2500 3000 3500 4000 1995 1998 Years # of Population Xapuri Brasileia Assis Brasil Figure 6 1 Total population of the CMER in three municipalities Traditionally, rubber tappers have always moved around because their patro determined this in the old seringal times or later on b y their own volition. Mobility has always been an important element in rubber tapper existence be it due to the old rubber system or the more recent insecurity in land tenure. Also, in recent decades with the uncertainty of the extractivist economy, many p eople have moved within the region and from forested areas to urban areas. Different fluxes of migration characterize the area: exodus from the seringal into the urban centers, immigration into the CMER from urban areas, and migrations at a local level fro m one seringal to another. Particularly, Xapuri and Brasilia are towns that have increasingly accommodated rubber tappers It seems that the population in the Xapuri area of the reserve has decreased more than Brasilia has increased, indicating some kind of out migration from the area into urban centers. Xapuri might be particularly vulnerable to exodus because it

PAGE 96

84 is closer to the state capital Rio Branco than the other municipalities, and rubber tappers have better transportation access to the area. In a ddition to out migration, many families in the nearby seringais (e.g., Dois Irmos) have close relatives in Xapuri and thus spend time in town, resulting in almost double residency in and outside the reserve. These people often alternate time spent in the city and the seringal While some people have moved out of the area others have moved into or returned to the reserve for lack of quality of life on the urban periphery, and unemployment in the cities. Before the establishment of the reserve many families had to leave their settlements in this region due to encroaching cattle ranching in the boundary of todays reserve. With the establishment of the reserve, many of these families that had moved to the cities gained right to the land and thus some slowly s tarted returning to the seringal, or if they opted for staying in the cities, passed the settlement rights to relatives. The families that returned to the forest realized that living in the reserve areas granted them not only the rights to the land, but al so to several benefits from the government agency, responsible for the reserve management. Assis Brasil, the most remote of the three municipalities, might have maintained its relatively low population size in these two periods due to its further isolatio n. In the seringais of this region, the influence of urban lifestyle on the reserve residents is less prevalent. This is because Assis Brasil is a very small and remote town that does not offer much better social services than those that are offered in the reserve, with the exception of health care. In addition to the distance, the road access to Assis Brasil from Brasilia and Rio Branco, is very precarious and during the rainy season becomes muddy and

PAGE 97

85 practically impassable. However, this road currently i s being paved and should be completed in 2002. In terms of internal migration, Brasilia and Xapuri do not differ significantly to influence households to move to one or another territory in the reserve. Both cities are close to each other, and many tapper s that live in the Brasilia region may interact and travel to Xapuri and vice verse depending the distance inside the reserve and the means of transportation available. While Brasilia is slightly larger, basically, these cities have the same market optio ns. (i.e., cooperatives to sell their production) and also offer the same goods that they need in their household. However, factors inside the reserve might influence a household more to migrate to another settlement, independently of the municipality area One of the factors might have to do with household consumption, such as game and hunting areas. If game resource become scarce a household is likely to move to another settlement. In addition, rubber tapper families tend to live closer to their relatives If the decision of the household head is to leave, the son/daughter that live closer to them might decide to move as well. Thus, as overall population in the reserve has remained approximately the same, migration occurs both into and out of the reserve, and people move within the reserve, it is hard to determine particular migration patterns and population dynamics at a reserve scale and municipality. When considering deforestation in these municipalities, it is also important to look at population densit y per area and the link of population to deforestation rates. Both population density and deforestation rates vary greatly across these three municipalities. In 1998, Assis Brasil had the lowest population density with 0.60 persons per square kilometer, fo llowed by Xapuri with 0.81 persons per kilometer square. Brasilia, on the

PAGE 98

86 contrary, has a population density of 1.60 person per kilometer square, almost double that of Xapuri. According to Sassagawa (1999), the Xapuri area presents the highest deforestati on percentage in 1998, although it has shown a decrease in population, resulting in a higher per capita deforestation level. Thus, even though the population has decreased, or rather a smaller number of seringais belong to this municipality, the deforestat ion rates continued to increase the region. Overall, the relationship between population growth and deforestation rates is low at the scale of municipalities. This is probably due to the large extent of each of these areas and the very distinct population density in different parts of the municipalities. The northernmost and remote areas in the Brasilia, along the Iaco river has few households and immense areas that are uninhabited, while the southern part holds a high population density, with seringais l inked to roads and closer to the city. Similarly, higher levels of deforestation are concentrated in particular zones of the reserve. Deforestation at this level appears to be driven by the increases of population in concentrated areas of the reserve, whic h would require different scale of analysis. Thus, to understand it, we need to look at different scale processes operating deforestation in the reserve. In the next section I will present a seringal level analysis of population grown and deforestation rat es, for those seringais with particularly high deforestation rates in the reserve. 6.3 Population Dynamics in the Study Seringais Figure 6 2 demonstrates the changes in rubber tapper populations in the four study seringais from 1995 to 2000, as obtained f rom the IBAMA census records. In general, the number of residents by seringais has increased constantly from 1995 to 1998 irrespective of the municipalities where they are located. In the seringal Paraguau the

PAGE 99

87 total number of residents was relatively high with 221 people in 1995 rising to 290 residents in 1998, presenting an increase of about 31% in the population growth. Similar population growth has taken place in the seringal Filipinas, the most populated of the four study seringais. In 1995, it had a population of 295 people and in 1998 it reached a total of 378 residents, demonstrating an increase of 28%. Seringal Humait had a population of only 53 residents in 1995, reaching a total of 96 residents in 1998, representing in an impressive increase of about 81% in the population size. 0 50 100 150 200 250 300 350 400 1995 1998 2000 Year # of people Paraguacu Humaita Filipinas Porongaba Figure 6 2 Total population in the four study seringais The highest rate of growth occurred in the seringal Porongaba, which in 1995 had only 23 residents, growing to 126 in 1998, an increme nt of about 447%, although overall population density is lower compared to the other study seringais. It is clear that the period of largest population growth was from 1995 to 1998, with the exception of seringal Humait. Here, population growth continued through 2000, reaching a total of

PAGE 100

88 210 residents, which represents an increment of about 118% compared to 1998 and 296% in relation to 1995. While both Porongaba and Filipinas seringais maintained approximately equal population levels from 1998 to 2000, pop ulation levels decreased significantly in the seringal Paraguau, by about 17%. Various factors contribute to this increase in population in these four seringais. Immigration of former and new residents from urban areas as well as rural areas close by, pl ay a important role in increasing rubber tapper population levels. More significant, the move of CMER residents from remote areas to more accessible areas such as the four study seringais is an important factor. It can be argued that rubber tappers who liv e in more isolated areas may be more likely to migrate to areas near cities, in part because they may have lower income levels. In the these four seringais the increase of population occurs partly because many families left isolated deep forest seringais o f the reserve and settled in more accessible. Table 6 1, which presents data from my own sample in these seringais, shows that the largest proportion of the household heads (39.4%) have migrated from other seringais in the reserve. These household heads ha d frequently migrated from the more isolated, deep forest seringais For instance, the seringais Amap and Guanabara in Brasilia, despite being the biggest seringais in the reserve and the ones with the largest rubber production, experienced severe out m igration as rubber tappers faced great economic difficulties due to the decline of the rubber price. With the falling price of rubber, these families had little to sell. Moreover, there were fewer marreteiros or traveling middlemen from whom they could buy dry goods and sell their production, forcing them to move into other areas with more services closer to the cities, such as the four study seringais.

PAGE 101

89 Table 6 1 Previous residences of the households in four seringais in the CMER Seringais Locations Filipinas Humait Porongaba Paraguau Total N N N N N % Different seringal 7 3 4 12 26 39.4 Other location* 1 6 7 2 16 24.2 Other settlement in the same seringal 7 2 1 4 14 21.2 City 2 -2 2 6 9.1 Always lived in the same settlement 2 -2 -4 6.1 Total 19 11 16 20 66 100 Households have lived in one of the following: colonization projects, ranches, and Bolivia In addition, as showed in table 6 1, a large part of these residents (24.2%) came from other locations outside the res erve, basically from colonization projects, and cattle ranching areas. To some extent, colonists that hold land in the settlement projects along the road also have a settlement in the reserve, in the seringais bordering the settlement projects, which may b e occupied by a family member. This is the case of the seringal Humait, where sons of colonists leaving the settlement project bordering the reserve, have moved into the reserve. Although the residents association has the role of monitoring the entrance of new residents into the reserve, it is very limited in controlling these residents. 22 These new rubber tappers although aware that they have moved into the reserve, continue the same land use patterns as practiced by their parents in the 22 The reserve residents Associations are legally responsible for the entrance of new residents into the reserve, according to the Utilization Plan. However, they are very limited in monitoring it, because the residents who want to leave can sell their settlements to anyone. The entrance of new residents is supposed to be first approved by the communities in each seringal and then be approved by the Association. Rarely is this process followed in the seringais because of political interests of the leaders, family ties, and lack of Association capacity to monitor the reserv e.

PAGE 102

90 colonization pr ojects. Therefore, these new kind of neo seringueiro residents might have influenced land use change in this seringal in particular as well as others in this region. Another important factor influencing population increase in these four seringais is the re turn migration of families that had to leave their settlements in the late 1980s, mostly in the Brasilia region. For example, in the seringal Porongaba, out of about 20 families that lived in this seringal in the late 1980s, only two families refused to l eave their home when they were faced with encroaching ranching in the region. With secured land tenure through the establishment of the reserve many families retuned to their settlements after living in another seringal or in the nearby city. Another aspec t that needs to be taken into consideration is the time of residence of households in the study seringais. Table 6 2 shows that most residents in these four seringais are relatively new in their settlements; 63.6% have lived in their colocao for ten year s or less. This demonstrates the constant population dynamic of rubber tappers families in these seringais and in the reserve as a whole since these data follow the same trends of socio economic data of the entire reserve. For example, a study in 1998 dete rmined that when the reserve was established only 35.4% of the current households were already living the Xapuris area of the reserve, while 40.3% and 50.4% of the households where already settled in Brasilia and Assis Brasil, respectively (ISPN 1998). T he same study determined that approximately 23.9% of the reserve residents had worked or lived in the cities. In Xapuri and Brasilia 27.1% and 26% respectively of the population had done so, while in Assis Brasil only 7.9% had spent time in urban areas.

PAGE 103

91 Table 6 2 Time of residence of households in the study seringais, CMER Seringais Years Filipinas Humait Porongaba Paraguau Total N N N N N % 1 10 13 8 9 12 42 63.6 11 20 3 1 3 6 13 19.7 21 30 1 2 -2 5 7.6 31 40 2 -3 -5 7.6 41 50 --1 -1 1.5 Total 19 11 16 20 66 100 6.4 Population Density and Deforestation in the Study Seringais The four study seringais are similar to each other in terms of population increase and are not representative for general trends within ea ch municipality. Within the reserve and within each municipality they represent seringais with relatively high population density. Some seringais in these municipalities, although being bigger in size than these discussed here, have lower population densit y. Therefore, population density seems to be a matter of location. The four study seringais are located in the southern boundary of the reserve, with relatively easy access to roads and transportation. Generally, the further the seringais are from roads an d urban centers, the lower the population. This trend seems to be strongly related to deforestation levels. Table 6 3 represents deforestation levels and rates in the four study seringais Although their area is quite different, the rates at which they wer e deforested are quite similar. Seringal Paraguau, in the municipality of Assis Brasil, has the lowest

PAGE 104

92 deforestation level of them all, although the rate of deforestation has increased tremendously in the 92 98 periods. The seringal Humait starts with re latively low deforestation levels in 1986, increases in 1992 and surpasses other seringais in terms of deforestation levels in 1998, mirroring the steep increase in population size it experienced. Seringal Porongaba had a relatively high deforestation rate from 1986 to 1992 compared to the other study seringais, but from 1992 to 1998 its deforestation rate was lower than the other seringais. Seringal Filipinas, with the largest population but little increase since 1998, experienced relatively high increases in deforestation rates until 1998, resulting in the second highest deforestation level among the study seringais. Table 6 3 Deforestation levels and rates in the four study seringais Deforestation/Seringal (%) Seringal (over 6% of deforestation) 1986 Ra te in 86 92 1992 Rate in 92 98 1998 Ser. Paragua 0,62 0,17 1,62 0,85 6,70 Ser. Humait 1,13 0,30 2,93 0,84 7,95 Ser. Porongaba 2,27 0,30 4,04 0,52 7,15 Ser. Filipinas 1,96 0,28 3,62 0,66 7,55 Source: Sassagawa, 1999. These data seem to suggest that there is a correlation between rising population levels and deforestation rates in these four seringais. However, this relationship is not a 100% clear one. In figure 6 3 one can detect a positive correlation, although a weak one. Compared to the muni cipal level though, the relationship between deforestation rates and population density is stronger. Possibly, to demonstrate this relationship more clearly the sample size needs to be larger and various types of seringais might need to be compared to be a ble to associate low deforestation rates with low population densities and high

PAGE 105

93 deforestation rates with high population densities. In addition, even at the seringal level a lot of heterogeneity exists between different colocaes Possibly, a comparison o f different areas within seringais or at the colocao level might be more appropriate to irrefutably establish this relationship. 1.24 1.45 1.32 0.86 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.00 0.50 1.00 1.50 2.00 Pop. Density 2000 (km2) Deforestation Rate 1992 98 (%) Figure 6 3 Relationship of deforestation rates vs. population density in the four seringais. Lan d use changes in these seringais, then, is very likely to have been influenced by rising population levels in the area, although in combination with other factors. These seringais are attractive to new residents because they are closer to the cities and ca n be accessed by road, possibilities that do not exist in more remote areas. This proximity allows income generation with products other than rubber. These new forms of income generation are likely to involve those that change forest cover more significant ly. Nonetheless, while the analysis of the population growth presented here suggests that it has affected land use changes in the seringais, other socioeconomic and ecological characteristics of the area need to be taken into account for a more complete u nderstanding of their land use dynamics in recent years.

PAGE 106

94 6.5 Settlement Dynamics in the Study Seringais Not only have population levels changed within the reserve and particular seringais but also the settlement dynamics and distribution have changed in recent years. Today, with the diversification of land use activities, it is exactly at the landscape level of individual settlements that most of the new land use decisions are made and carried out. Therefore, apart from looking at population levels per se it is important to examine increases in settlement numbers within the reserve and the study areas. Figure 6 4 shows the number of settlements in the Reserve. In the 1995 1998 period the total number of the rubber tapper settlements in the Xapuri area was 681 (62% of the total reserve) in 1995 and 507 (45.9%) in 1998, followed by Brasilia, which had a total of 219 (20%) in 1995 and 369 (33.4%) in 1998. However, similarly to the population data, the colocaes in the seringais that were changed from Xapuri to Brasilia between 1995 and 1998, also changed. Thus, the decline and increase for Xapuri and Brasilia respectively are less significant than they seem according to this data. As for Assis Brasil, it had a total of 197 (18%) in 1995 and 228 (20.7%) in 1998, which shows an increase of 13.5% in the number of settlement. As noted early, Assis Brasil had presented the same population size in the period 1995 1998; however, this area has had an increase of 13.5% in the number of settlements within the reserve Settlement expansion follows the same trends with the population when it is looked at a municipalities level. That is to say that in these three municipalities settlement number has a direct relation with the population trends. But a few new questions a rise. Are brand new settlements being opened? That might be the case if new residents are migrating to the reserve. Or are the oldest settlements being subdivided due to family size? In this latter case, settlement dynamic is not related to migrant populat ion in these

PAGE 107

95 0 100 200 300 400 500 600 700 800 1995 1998 Years # of settlement Xapuri Brasileia Assis Brasil Figure 6 4 Total number of settlements in three municipalities of the CMER areas. In the following we will look at settlement expansion at a seringal level and try to examine the implications that these factors c onducing the increase in settlement number may have for deforestation in these seringais Figure 6 5 shows the number of settlements in the four study seringais. Overall, the number of settlements in these seringais has increased continuously, with the ex ception of Paraguau, which decreased from 1998 to 2000, and Porongaba that remained constant from 1998 to 2000. Overall, all four areas experienced the greatest increase in settlement from 1995 to 1998. The seringal Paraguau in 1995 had 33 settlements, i ncreasing to 52 in 1998, which represents an increase of 57.6%. In the same way, seringal Humait went from 10 settlements in 1995 to 23 in 1998, showing an increment of 130%, and seringal Filipinas, with 52 settlements in 1995, reached 63 in 1998 increasi ng by 21.15%. Only seringal Humait and Filipinas continued increasing the number of settlements in the entire period analyzed (1995 2000). The former reached

PAGE 108

96 a total number of 32 settlements in 2000, a total increase of 220%. The latter reached a total of 72 settlements, a total increase of 38.5%. 0 10 20 30 40 50 60 70 80 1995 1998 2000 Years # settlements Paraguacu Humaita Filipinas Porongaba Figure 6 5 Total numbers of settlements in the four study seringais As noted before, 39.4% (Table 6 1) of the residents in these seringais came from different seringais within the reserve. This speaks for the arguments that the increase in the number of settlements in these seringais is strongly due to migration from other areas. In addition to migration, which implies opening new settlements, settlement subdivision is another impor tant factor responsible for settlement increase. Table 6 4 shows the number of settlement subdivisions in these seringais. Overall, it shows that 65.2% of the settlements have been subdivided, while 34.8% of the settlements have not been sub divided in the last ten years. Thus, the great majority of settlements have experienced some type of subdivision. Several factors influence the settlement subdivision in these seringais As presented in table 6 5 the main factors is marriage of the son/daughter of the household

PAGE 109

97 head (50%). This happens with the increase of family size in a settlement or due to the marriage of the son/daughter of the household. In this case, they usually divide a settlement by opening a new clearing for the house where the new family wil l be settled close to their parents. This practice also can have a considerable impact on land use change, since with the establishment of a new family, new shifting cultivation plots and other land use initiatives will take place to provide subsistence g oods and income for the additional household. Table 6 4 Settlements subdivision by seringais in the CMER. Seringais Filipinas Humait Porongaba Paraguau Total Settlements Subdivision N N N N N % Yes 13 9 11 10 43 65.2 No 6 2 5 10 23 34.8 Tot al 19 11 16 20 66 100 Another factor influencing settlement subdivision is the intent of some households to expand the legally deforested area in the settlements. In settlements with an already elevated deforested area and household heads that are cogniz ant of limitations in the Utilization plan, they subdivide the settlements, making a new settlement that might be occupied by a family member, have a new registration of the settlement with the Resident Association and thus have more areas for deforestatio n in two settlements. Another factor that often influence subdivisions is when a son or another relative moves into the same colocao and gets allotted part of the land with some estradas de seringa and land. This relative might be moving from close by or might originate in the city.

PAGE 110

98 Table 6 5 Reasons for settlement subdivision in four seringais, CMER. Seringais Filipinas Humait Porongaba Paraguau Total Reasons for subdivision N N N N N % Marriage of son/daughter 10 7 6 10 33 50 Settlement has not been subdivided 6 2 5 10 23 34.8 To expand area for deforestation 2 -2 -4 6.1 Lease 1 2 3 -6 9.1 Total 19 11 16 20 66 100 This section demonstrates that not only are population levels and population density affecting factors i n land use change, but the particular dynamic of settlement patterns at the local level are of great importance in understanding how deforestation comes about in these areas. This is an area that needs increased research attention given the current poor un derstanding of what a colocao is, how this concept works at a specific geographical scale, and legal issues such as colocao delimitations and deforestation levels. What happens at the colocao level is also of crucial importance in understanding land use change and land use decisions that are made at this level. 6.6 Changes in the Production System: Seringal Paraguau In the four study seringais the extraction of non timber forest products, mainly rubber and Brazil nut, has suffered a drastic decreas e, while pasture creation, and thus, cattle raising as well as agricultural production has increased. In this section I will discuss the changes in the production system between 1995 and 2000, looking at the Seringal Paraguau as a specific case. I selecte d this seringal because this is far from the other three seringais and is located in territory of the more distant city (Assis Brasil). Thus,

PAGE 111

99 changes in the production system are expected to be greater at least on other two seringais (Porongaba and Humaita ) than in seringal Paraguau. According to census data, in 1995, the total rubber production in the seringal Paraguau was 14,180 kilos. In 1998, the total rubber produced was 12,040 kilos, representing a decrease of 15%. By 2000, however, rubber productio n has suffered a drastic decrease, with a total yield of only 2,110 kilos, a decrease of 85% in just five years. Cattle raising, on the other hand, in 1995 had a total number of only 129 heads, followed by an increase of 255 heads in 1998 or 98%. By 2000 c attle raising experienced a very large increase of 344 heads, presenting a total increment of 167% in five years. Figure 6 6 illustrates these changes by presenting the average proportional production increments of rubber, Brazil nut and cattle heads per s ettlements. The average rubber produced per settlement in 1995 was 429 kilos, decreasing to 231 kilos in 1998 and 52 kilos in 2000. Conversely, cattle heads had an average of 3.9 heads per settlement in 1995, going to 4.9 in 1998 and 8.6 in 2000. This figu re illustrates that from 1995 to 2000 the average production of cattle per settlement increased by 120%, while rubber production decreased by 88% and Brazil nut production by 32% in the same period. In addition to increased cattle raising, agricultural pr oduction in that seringal has increased particularly in the second period analyzed (1998 2000). Figure 6 7 shows the average proportional production increments of agricultural production per settlements. It is noted that the two crops with highest increas e per settlement were beans and corn, which are those with higher commercial value in the city, as opposed to manioc and rice, which are often produced for family consumption. The average bean production per settlement was 113.85 kilos in 1995, increasing

PAGE 112

100 0.00 50.00 100.00 150.00 200.00 250.00 1995 1998 2000 Year Production Increment (%) % Rubber % Brazil Nut % Cattle Figure 6 6 Average production increments per settlement in the seringal Paraguau to 200.87 kilos in 1998, an increment of 76.43%. In 2000, the average was 577 kilos, an increase of 477.25%. Likewise, the average corn productio n per household was 244.23 kilos in 1995, increasing to 585.65 kilos in 1998, an increment of 139.8%. In 2000, the average corn production per settlement was 1044 kilos, an increase of 327.9% compared to 1995. Rice and manioc production per settlement expe rienced similar increases during the period 1995 2000. The former had an average production of 619 kilos per settlement in 1995 and reached a total of 1284 kilos in 2000, an increase of 107%. The latter had an average production of 226 kilos in 1995 reachi ng 440 kilos in 2000, an increase of 94%. In general, these data demonstrate that until 1998, fewer households had increased their heads of cattle, but from 1998 to 2000 cattle raising increased in most settlements or the number of cattle in some househol ds is very high, although I have no data on how many households are raising cattle. These data also show that by 2000 very few

PAGE 113

101 0 100 200 300 400 500 600 1995 1998 2000 Year % Manioc Rice Corn Bean Figure 6 7 Average agricultural productions per household: Seringal Paraguau households in that ser ingal were engaged in rubber tapping activities as their main source of income. Instead, rubber tapper families have opted for other economic activities that include cattle raising and agricultural production The low rubber production can be explained by the extremely low prices for rubber, which are exacerbated by the fact that decades of low rubber prices have not motivated younger people to learn the skill of rubber tapping. Sons of the rubber tappers that do not know how to work with rubber, or have ne ver done so, do not see the economic benefit of such a labor intensive activity. Instead, they may engage in activities of agricultural production and livestock that not only require less labor but also are easy to transport and sell. Overall the seringal Paraguau experienced an increase of all agricultural activity both for subsistence and commercial crops. The dramatic increase of cash crop

PAGE 114

102 production denotes the increased market orientation of the rubber tapper agricultural production. This focus on ag ricultural activities in the reserve is much more fixed than the traditional rubber extraction that depended on certain mobility. Also, the willingness to significantly invest in agriculture might also be a reflection of the more secure land tenure in the reserve and the lesser dependence on the marreteiro. The implication of this change in land use choices is that, since agricultural production is increasing and land being increasingly converted into pasture for cattle production, the primary forest in th ese areas will be increasingly cut down for the new cycle of agriculture and then pasture. Often the profits from the sale of these agricultural products, together with others, such as small livestock (chicken), are invested in cattle raising, which for th e residents represents an investment that can be easily changed into cash in times of hardship and new investments into houses, new activities and other assets. 6.7 Pasture and Cattle Raising Expansion in the Study Seringais According to the State Produc tion Secretary, the State of Acre has more than 1.5 million heads of cattle. In addition, data from the Fazenda Secretary of the state (SEFAZ) shows that cattle raising contributed over 50% of the State taxes collected in 1999. According to the 1996 Agro p astoral census (IBGE) of the 23,788 small farmers (landowners or not) in the State in 1996, 81% had land converted for pasture and 55% had cattle. Yet the data from IBGE shows that between 1985 and 1996 there was an increase in 177% of cattle heads in the Acres landholdings with less than 100 hectares. These factors have a direct influence on extractivist communities in Acre. The CMER is located within the region with the most cattle raising and small farmers in

PAGE 115

103 eastern Acre. In fact, some seringais of th e reserve border the biggest farm in the region. In addition, the access to the reserve area in many cases is by roads that lead first to the cattle ranch (e.g., seringal Porongaba) and colonization projects (e.g., seringal Humait). Thus, the development example that is set by the areas bordering the reserve is that of cattle ranching. It is agreed upon in the Utilization Plan that rubber tapper families can have a small number of cattle in order to improve their daily diet and the transportation of their production. However, in the study seringais an accelerated increase of pasture has taken place, compared to other areas of the CMER. The greatest increase is noted in the period 1995 1998. According to the indications in the census data, seringal Paragua u in 1995 has a total of 96 hectares pasture reaching a total of 210 in 1998, an increase of 118% in just two years. Likewise, seringal Filipinas had a total of 145 hectares of land in pasture in 1995, going up to 251 hectares in 1998, an increase of 73%. The other two seringais, Humait and Porongaba, which border each other, had the greatest increase in pasture in that period. The former had a total of 47 hectares of pasture in 1995, reaching a total of 208 hectares in 1998, an increase of 342%. The latte r, had a total of 15 hectares in 1995, reaching a total of 109 hectares in 1998, an increase of 625% in land converted into pasture. Figure 6 8 presents the average pasture area per settlement in the four seringais according to the census data. Seringal Pa raguau and Filipinas present almost the same average increase in the 1995 1998 period, from 2.92 ha and 2.79 ha respectively in 1995 to 1998 and 4.5 ha in 1998. Seringal Porongaba had a constant increase in the period 1995 2000 presenting an average of 3. 75 ha in 1995, 4.76 ha in 1998 and 6.83 ha in 2000.

PAGE 116

104 The greatest increase was detected in the seringal Humait, which in 1995 had an average of 4.7 hectares per settlement, growing to 9.4 in 1998, followed by 9.7 in 2000. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 1995 1998 2000 Year Area (ha) Paraguacu Humaita Filipinas Porongaba Figu re 6 8 Average pasture area per settlement in the four study seringais The perception among rubber tappers in these seringais is that the transformation of land into pasture will increase the value of the settlements, a perception that also is strongly de fended by farmers in colonization projects in Acre. In many cases, to convert land into pasture does not necessarily mean that the rubber tapper will, in fact, raise cattle because it depends on the economic success of other productive activities, but it a t least indicates their plans for future land use choice. However, a confounding element in this study is that in some cases rubber tappers have recently moved to settlements in the reserve that might have pasture fields that were created by former reside nts of these settlements. Even though they have pasture, they may not plan on raising cattle or expanding pasture field. In that case, these residents

PAGE 117

105 have limitations in expanding land use, since the land designed for clearing by the utilization plan (10% of the settlement) is already compromised in pasture, precluding for instance new agricultural activities. In these areas of the reserve the recent productivity of extractivist and non extractivist goods and the deforestation rates seem to mirror the deve lopment outside the reserve, though on a much smaller scale. The increase of pasture area and the declining importance of rubber production in these four seringais, and in the reserve as a whole, will greatly influence the future economic viability and con servation objectives of the reserve as well as with the very identity of rubber tappers as forest dwellers and conservationists. 6.8 Social Organization in the Reserve In the historical rubber estates, the main organizational structure was the vertical hierarchy of patron client relationships starting with the main boss ( seringalista ) and going down through intermediate bosses to the rubber tapper at the bottom. With the disarticulation of the rubber market, the collapse of the rubber market and bankrup tcy of seringalistas who were increasingly replaced by cattle ranching in the region, rubber tapper communities started to slowly build their own social organization, mostly in the form of rural labor unions. In the region where the reserve was established the main social organization was the rural workers union in Brasilia and Xapuri, created in the 1970s. Through these, the rubber tapper community made a major step towards the consolidation of their social organization, creating the Rubber Tapper Nation al Council in 1985, which led to the proposal of the extractive reserve. When the Chico Mendes reserve was created the demand was voiced for a stronger social organization that would work together with the already established one,

PAGE 118

106 but would also have the role of administering and managing of the reserve together with the government agency. In this context, three Reserve Residents Associations were created in 1995, dividing the reserve into three major zones of representation corresponding to the reserve a reas linked to each of three municipalities. These three Associations were the Association of the residents of the CMER area of the Xapuri (AMOREX), the Association of the residents of the CMER area of Brasilia (AMOREB), and the Association of the Residen ts of the CMER area of Assis Brasil (AMOREAB). The main goal of these Associations originally was to administer the Government project at a local level so that government resources could reach the communities in the reserve. The Associations then started t o develop an affiliation process with the residents, which led to an improvement of the community organization with strategies of organization that now had to be approached in the reserve level, not at the seringal level as before. Through these organizati ons, many communities had access to better social services such as schools and health care, and received economic initiatives aimed at generating income such as training and equipment for particular economic activities, as well as transport support. Howeve r, the degree of association with these organizations differs in regions and from seringal to seringal. The levels membership with the associations are important indicators of the involvement of the residents in the rubber tapper social movement, in curre nt projects, in new economic activities as well as credit opportunities. Figure 6 9 presents the total number of residents associated with the AMOREX and AMOREB from 1995 to 2000. It clearly shows that there has been a constant increase in the number of p eople affiliated throughout the last years. For example, from

PAGE 119

107 0 100 200 300 400 500 600 1995 1996 1997 1998 1999 2000 Year # of members Amorex Amoreb Figure 6 9 Total memberships in CMER associations 1996 to 2000 AMOREX grew from a number of 130 to 478 affiliated members, an increase of 267% in just four years. L ikewise, AMOREB, which had an initial number of 139 affiliated in 1996, reached 403 members in 2000, an increase of 190% in the same period. According to the Associations directors, the increased affiliation was due to three main factors. First, women sta rted to get associated due to government benefits for rural expectant mothers. Second, older residents also need to be affiliated to have access to a retirement plan. Affiliation means that they can prove that they live in the reserve,

PAGE 120

108 therefore are eligib le for benefits. Finally, many rural credit opportunities for agricultural, pastoral and extractive activities are accessed through these associations. 23 More recently, the new Acre government issued the Chico Mendes Law in 1999 that subsidizes rubber pr oduction throughout the State. To have access to this benefit and thus sell rubber production at higher prices, rubber tappers must be members of the Associations, which accesses these resources from the government and distributes them to tappers. As one c an note, these reasons behind the expanding resident membership in the reserve are not necessarily directly related to land use change in the reserve. However, the development projects and opportunities that are accessed through the social organization do indeed drive some of the more recent land use choices, such as increased agricultural production and in some cases cattle production (Prorural). Also, once rubber tapper communities have access to different benefits, they also use them for more permanent investments into their settlements, which will likely involve increased agricultural production and pasture fields, whether they were originally intended for this purpose or not. Consequently, these government programs can directly affect the new land use choices practiced by the residents. As to the rubber subsidies, they strengthen a return to higher rubber production in the seringais. However, these subsidies alone will not have great impacts on land use choice, if the residents do not want to tap rubb er as is the case with most young residents, or, like the elderly, see 23 In the Amazon the most important credit line is the FNO Especial ( Fundo Constitucional de Financiamento do Norte ), federal funds that are explicitly intended for small holder credits in the Northern region of the country and are administered by the Amazonian Bank (BASA) in cooperation with extension and agricultural research institutions. For several years the main credit line accessible to rubber tappers was PRORURAL ( Programa de apoio a produo familiar rural organizada ), a general rural credit t hat finances agriculture, ranching and fishing and as well as processing facilities. Since 1997, the PRODEX program ( Programa de apoio ao desenvolvimento extrativista ) has been available explicitly for extractivist activities, such as rubber and Brazil nut extraction and agroforestry projects.

PAGE 121

109 another source of income (e.g., retirement) as an alternative that allows them to invest their time in other activities that requires less intensive labor. Generally, the seringais clo ser to the cities like the ones analyzed here, are also the seringais where communities have a higher level of organization in the reserve. Table 6 6 shows the residents membership in the resident association and the rural workers union in the four study s eringais. Overall, it can be noted that 65.2% of the residents are affiliated to the Association, while 54.5% are affiliated to the rural workers union. These are very high levels of affiliation. Even though the rural workers union has a history of more th an twenty years and the Association exists for only six years, the number of members with the Association is higher. This demonstrates clearly the more practical role of the Association in these seringais, which relates mostly with official government and development activities. Table 6 6 Membership in association and rural worker union in the study seringais Resident Associations Rural worker Union Membership N % N % Yes 43 65.2 36 54.5 No 23 34.8 30 45.5 Total 66 100 66 100 Interes tingly, the board of directors of the Associations, who makes the major decisions about activities and investments within the association, is composed mostly of tapper leaders of these four seringais. It seems that having better organization levels, and ma jor representation in these organizations, led to the modification of the communities economic perspective. These current leaders define the major decisions regarding

PAGE 122

110 investment in the reserve and very often the areas granted with investments are these re presented by the director board. In addition, these decisions do sometimes consider family ties over community interest. For instance, two of the current Association board directors in the Xapuri Association came from the Seringal Dois Irmos, which is the best structured seringal in Xapuri both in terms of social services and economic activities. The goal of improving the level of social organization in the reserve does not exist in a vacuum, and communities have changed their economic and social perspecti ves, which in return have an indirect effect on land use activities. Paradoxically, contrary to the original purpose of the associations, the described land use changes might also very well have been a consequence of increased social organization. By mobil izing reserve residents within the resident Associations, these development incentives might have engendered the very changes they originally had hoped to stifle. The great challenge that poses itself for the associations in the next years is the reconcili ation of the originally championed conservation and forest dweller values, with development incentives and opportunities that truly help increasing income. So far, few of the official programs have born tangible fruits in terms of income generation. Hope fully, these experiences will inform the decisions at the associations as well as the government agencies. Today, associations face chronic problems related to financial dependence on government projects, personnel disputes among leaders and political inte rference from community leaders and local politicians. In some seringais members resist contributing to the maintenance of an Association that is only remotely connected to their daily concerns. On the other hand, in other seringais residents have formed interest groups to influence the Association or to elect officials, competing only among themselves, instead

PAGE 123

111 of cooperating as a whole. For example, since the Association in Xapuri was created in 1995, a group of eight rubber tapper leaders from different communities has administered it. Every two years there is a general assembly to elect new representatives. These eight leaders that have political power in their communities established a rotation system whereby offices in the associations board are essen tially exchanged between the same eight men. By blocking a newcomers access to these positions, they manage to keep the benefits of being community representatives to themselves, which, in turn, assistance the communities they represent. These leaders ar e usually the ones with better reading skills who speak more eloquently to the communities and influential outsiders. These representatives fail to identify younger local leadership, which would be based in the communities. Conversely, today, only two of t hese leaders still live in the seringais that they represent; the others have moved to Xapuri, only keeping their settlement and sometimes a family member in the seringal It is evident that this group has not stimulated capacity building of young resident s and thus it is difficult for new leaders, and less politically networked seringais to compete for political office within the Reserve and strengthen locally based social organization.

PAGE 124

112 CHAPTER 7 SUMMARY AND TRANSITION Rubber tapper communities proposed their own strategy of development in the form of extractive reserves and have given historical evidence of being capable of using forest resources without leaving a deforested landscape in their wa ke. Chapter 2 highlighted the evolution of the extractive reserve concept in the Brazilian Amazon context and analyzed the leading discussions about extractive reserves, discerning the major trends of conceptual debate. Some of the ideas that resulted in t he creation of extractive reserves a decade ago have undergone several changes and that the ER experience thus far has been an extended pilot experiment in its character. Since the creation of the extractive reserves, both government and the local communit ies have learned many lessons. One such lesson is that a few small scale, local initiatives cannot solve regional deforestation problems. Nevertheless, it still is one of the few models in which traditional communities have played a major role. In addition leaving aside barriers of a political and economic nature, the success of extractive reserves will depend on the ability of government agencies to cooperate with the extractive communities and maintain a level of dialogue, that allows the periodic redefi nition of goals and mutual responsibilities. On the part of the communities, a continual effort will be required in order to maintain internal cohesion, common goals, and the ability to adapt to the changing socioeconomic context, which today is calling fo r a diversified tapper production system, causing greater impact on forest cover, as was noted with the increasing deforestation rates in the CMER.

PAGE 125

113 Chapter 4 presents a brief overview of land use in the state of Acre, and the areas surrounding the CMER, an d the reserve itself, by examining deforestation rates in these areas. The deforestation process does not occur uniformly across the reserve, or even across a single seringal In areas that are accessed by roads and are closer to the highway and towns, hig her levels of deforestation are more common. Overall, it can be said that while the CMER is little impacted in many large areas, it is experiencing dramatic increases in deforestation in others. In addition, deforestation also occurs on a different and sma ller scale, the colocao level. If the deforestation rates in more impacted areas are indicative of broader trends for future land use patterns in extractivist communities, there is reason for concern. Therefore, understanding the social dynamics of land use change is of great importance for the future management of the reserve and the definition of conservation and development objectives within extractivist communities. Chapter 5 presented some household determinants of diversified land use activities. T he results show that key household characteristics have influenced land use strategies suggesting that attention needs to be placed on the household level factors affecting different land use activities. General trends of these household factors indicate t hat the changes taking place in land use activities are permanent, and demonstrate that deforestation will continue to increase as a result, a concern that must not be overlooked by household level development initiatives that call for the diversification of land use activities. Chapter 6 attempted to analyze the effect of various socio economic drivers on land use change and deforestation rates within the CMER using as case study four seringais with particularly high deforestation rates. First, the role o f population and

PAGE 126

114 settlement dynamic in the reserve was examined to explain why these four seringais not only have the highest population growth but also have experienced the most drastic land use changes. Second, the significant changes in the rubber tappe r production systems were discussed to explore the economic factors that are leading to land use changes. Finally, the social organization model in the reserve was examined to discuss how it may influence land use change. Although overall population levels throughout the reserve have not increased significantly, they have indeed increased in the four study seringais, resulting in increased population density. Behind the increases in these numbers are hidden a number of processes. Exodus from the reserve int o urban centers, immigration from nearby rural areas, urban centers and from other seringais within the reserve, and higher reproduction rates are just a few processes. The main source of origin of newcomers though in the four study seringais are remote se ringais where the life as pure rubber tappers became too much of a hardship. A crucial process in occupation of space within the seringais and the accommodation of an increasing population is the increasing subdivision of the settlements. This increases po pulation and settlement density as well as facilitates further deforestation rates in an area. Nonetheless, although generally higher population density and deforestation rates seem to bear a relationship, there is no clear cut correlation between these tw o variables. The changes in the rubber tapper production system over the last 5 years are striking. The system has been transformed from extractive production and subsistence agriculture to cash crop agriculture and pastoral activities. The low rubber p rices have discouraged residents from rubber tapping. However, the new government policies have

PAGE 127

115 offered incentives for rubber tapper lifestyles. As a result, many families in the reserve returned to rubber tapping activities. In addition, government data s hows that about eight hundred former rubber tapper families living in the cities returned to the forest. Although these numbers are questionable, about 40% of these families returned to municipalities holding territory on the CMER. In addition, this year m any brasilianos (Brazilian rubber tappers in Bolivia) were kicked out because of logging concessions, and are looking for areas to settle. Many of these would like to settle in the CMER. Possibly, the recent rubber subsidy and other government programs mi ght transform the current scenario somewhat in these more deforested seringais of the CMER A major and still little explored role in this process is that of the social organization in the reserve. While the rubber tapper social movement has articulated m any of the original goals of extractive reserves and have promoted traditionalist values of rubber tapping, the more recent associations (created by the federal government) also act as a vehicle for larger development incentives in the reserve. Some of the se, such as credit lines, can significantly influence and transform the rubber tapper production system, even though implemented with the best of intentions. The rubber tapper associations and the government and non governmental organization working in the area face the great challenge to articulate a new vision and new objective of modern rubber tappers. In addition, the need to devise strategies that remain truthful to the assertion that extractive reserves can maintain forest cover and generate income, r emains as alive as ever. The extractive reserve model is dependent upon financial support from international institutions and the Brazilian government. Extractive reserves will need

PAGE 128

116 subsidies for a long period of time if they are to be consolidated as a mo del of conservation and extractivist development. Brazilian funds are not significant enough to support the extractive reserves demands. Therefore, the reserves become even more dependent upon international resources, which requires a slow and complex neg otiation that guarantees financial support for only a short period of time. Should these international investments be seen only as a payment for the conservation services provided by the reserves residents? The answer for this question is no, and internat ional support will not be permanent. However, how long will it take to reconcile income generation to the reserves inhabitants with conservation practices? Neither the communities nor the international agencies know the answer to this question, resulting in a profound lack of definition for the model; thus, rubber tapper tend to continue incorporating new forms of land use as a mean of income generation and economic development. What, then, are the new options for diversifying land uses the in reserves? Land use regulations and enforcement at the community level are a major concern that will determine the future of extractive reserves. The Utilization Plan challenges the rubber tapper residents of the reserves to abide by the federal legislation while the ir economic development goals are as yet undefined. Due to this, and the experience of some years of implementation of the laws in the reserves, several of the norms of this plan have been questioned. For example, in the last two years, there is already a controversial discussion around the CMER, about a proposal to have sustainable timber extraction within the reserve, which is currently illegal according to the Utilization Plan. The current municipal administration of Xapuri, represented by rubber tap per leaders, has established a noble woods processing plant in Xapuri, in an attempt to create economic

PAGE 129

117 alternatives for the region. This proposal conflicts directly with the norms of the Utilization Plan, as the only timber supplies left in the municipal territory lie within the reserve. Therefore, a strong lobby exists calling for changes in the Utilization Plan that can only take place if 51% of the reserves residents, including those from the other four municipalities, agree with this change. In this case, municipal economic development goals could transform the natural resource regulations of almost all districts in Eastern Acre. According to local political forces and many of the residents within the reserve, timber exploitation will be approved soon As this decision conflicts with international and national environmental funding policies for extractive reserves, two options remain: the loss of multimillion funding guaranteed for the next four years (by PPG 7) or a cancellation of any local decisions about this issue. Thus far, no decision has been made. However, it need to be said that a number of larger international donor organizations are supporting community based forest management (timber) in Amazonia. These efforts are still experimental and be nefits to local communities have yet to be evaluated. Therefore, I suggest that federal extractive reserve refrain from this step until the economic and ecological viability of these initiatives has been demonstrated. Who should have the authority to make these kinds of decisions? Local people with subsistence and economic development needs or international agencies that control the funds are mostly concerned with forest conservation? This dilemma demonstrates that the original idea of reconciling forest c onservation and economic development of forest communities within extractive reserves still has a long way to go before it can become consolidated. The CMER is by far the biggest reserve in Amazonia and has been

PAGE 130

118 considered as an experiment to identify succ essful new income generation strategies and its later application to other reserves. If the current initiative carried out by the Xapuri mayor is a success, and other extractive reserves follow those proposals, the extractive reserve concept will pass thou gh profound changes and has the risk of being shaped only in economic terms, as the traditional model of development of the Amazonia. The timber exploitation discussion shows the fragility of the extractive reserve concept when it comes to economic issues. The vision of the inherently environmental rubber tapper has made the rubber tappers struggle popular within the international community; however, their new economic perspectives and focus can invert this historical vision. The tappers proposal for e xtractive reserves became an international symbol for friends of the forest because they were broadcast as an ecological issue by global international communicators. But this tappers proposal is not ecological except in a tactical sense. The goal of the tappers proposal for creating reserves was not to preserve biodiversity, but to attain a forest version of land reform that would give them permanent rights to use the forests where they have made their frugal living for generations. As Schwartzman ( 1989) wrote, extractive reserves are in essence a proposal for an ecologically sound and socially appropriate land reform in the Amazon. However, if the extractive reserve fails to increase the economic value of standing forests, despite the current nat ional and international support, the criticism of the reserves will be confirmed and the pressure for other land use options will increase in these areas and so will deforestation rates. In addition, several others governmental and non governmental develop ment initiatives for Amazonian extractivism will be less effective and therefore extractivist perspectives of development in the region would suffer a

PAGE 131

119 setback. If they survive, it will take at least another decade before their success can be judged and b y then the concept will have gone through several additional changes. This research is not seen as an end in itself. It is but a step towards illuminating the dominant issues about tappers in the CMER. It can hopefully lead to a better understanding of the needs and problems of tapper communities in relation to the process of socio economic transformation at present influencing their lives and destinies as extractivist communities. Areas for further research can be identified and a number of contributory is sues illuminated. First, as observed in chapter 6, population density varies greatly across the reserve. Thus, population dynamic need to be examined at a seringal and reserve scale to discern the major factors causing population mobility in the reserve. T apper mobility is important to the reserve model insofar as such model depends upon the formation of residents associations to collectively secure the designation of a reserve and to foster the social, political and economic changes necessary for the long term success of the reserves. This mobility of residence also presents obvious difficulties for the creation of social relations among tapper communities over a period of many years. In addition, this mobility is related to economic factors. If, as intend ed, living standards improve within the reserve, then spontaneous immigration should be expected. Increasing living standards within the reserve could also favor out migration, however. One of the most challenging and important research topics in the reser ve deforestation context of today is to integrate remote sensing methods with socio economic analysis at a household level. Its value lies in the fact that it can be used to explore the link between small scale land use change and the socio economic factor s affecting these changes and family land use decision making. From this approach, it

PAGE 132

120 would be possible to characterize the general patterns of deforestation in the reserve and relate these to broad socio economic processes affecting land use changes. Anot her topic worthy of inquiry, although not directly related to this thesis, would be to map the dynamic of interaction and interest of institutions linked to the extractive reserve model in Amazonia. This would need to link the rubber tapper grass roots org anization with national and international organizations and would focus on understanding how different actors and groups view, mediate and negotiate land use changes in the extractive reserve. This research would bring important insights in the different d iscussions of goals and views on reserves. Some of these institutions are primarily concerned with forest conservation and biodiversity and see the preservation of extractivist ways of living as a means to achieve their interest. Rubber tappers, on the oth er hand, aim to strengthen their livelihoods and to improve their well being.

PAGE 133

121 APPENDIX A DATA CENSUS RELIABILITY AND UNCERTAINTY OF THE CMER One important requirement of the Brazilian Environmental Agency (IBAMA) for the establishment of a Federal Extractive Reserve is a census survey prior to approval. As for the CMER the first so cio economic survey carried out was in 1992 by the Rubber Tapper National Council (CNS). After the establishment of the Reserve IBAMA assumed the legal responsibility of monitoring the reserve. The second census survey in the reserve was carried out in 199 5. The reserve was divided into zones for each municipality and the Residents Associations selected a group of residents who carried out the survey in the field with a technician from IBAMA. The data was gathered through structured interviews with mostly the household heads. Each Association had a group coordinator, who usually was an Association director, that was responsible for coordnating the group in the field and submitting the data gathered to IBAMA. The types of data gathered include personal and d emographic data, production data, and agricultural plot, fallow, and pasture size data. The reliability of the data gathered is affected by several factors. The first factor was the distance that the members of the survey group had to travel to cover the e ntire reserve. To reach the settlements, it might have taken between two to fifteen hours by foot, by boat, or by animal depending on the accessibility of each seringal. The second factor is the limited skill of the group members on formal education, in in terviewing and in understanding the information gathered. Third, at times the household head might not be at home when the interviewers came and thus indirect data was collected from

PAGE 134

122 neighbors, workers, or the wife. Despite these limitations, it is still better to work with the residents than to work only with an urban technician, which would present others own challenges. In the 1998 census the same methodology was applied, but with important differences. In the 1998 census each one of the members of th e group was equipped with a GPS machine and trained in getting GPS reading of the settlements. This was introduced mostly for two reasons. First, in the previous census some cases were reported that members of the groups had not gone to the more distant se ttlements but had gotten the information through relatives of the residents of these distant settlements or by filling out the questionnaire using their own knowledge of the community. Also, at the same time, a mapping project of the reserve settlement was being carried out by a group of researchers of the Federal University of Acre, including myself. The opportunity to get GPS readings of the settlement of the entire reserve would not only increase the reliability of the census data, since the interviewers must go to each colocao but also would be of great value for the mapping project. Thus, we gave basic training in GPS, mostly in teaching how to get the geographic coordinates of the settlement and recording it in the questionnaire. The 2000 census sur vey followed the same strategies, with the advantage that Association leaders and members were becoming increasingly experienced in these activities, and thus the reliability of the data tended to improve. Despite the amount of information in the census, it has been explored very little. The use of this information is basically restrictive to keep a database with residents names and their locations in the reserve. It is also meant as a tool to keep track of major trends within the reserve in terms of prod uctivity and deforestation levels.

PAGE 135

123 However, when comparing deforested areas according to the census data (adding up agricultural plots, fallow, and pasture at a seringal level) the total sum is very much smaller than that determined with remote sensing tec hniques (own observation). This might result from two factors. First, what is problematic about the census data is that it contains sensitive information, which will ultimately fall into the hands of government officials (IBAMA) that have a certain authori ty. A rubber tapper that is deforesting more than allowed is likely to understate his pasture and agricultural plot size. Second, a major, but hardly addressed problem is that of unreliable estimates of deforested areas and colocao area. A colocao is g enerally defined as the number of rubber trails X 100ha. Thus, a colocao with 6 rubber trails has approximately 600 ha, of which 10% would be 60 ha. However, nobody really knows the size of a colocao and in real life the size of rubber trail might var y tremendously. Therefore, estimating the legal 10% limit of deforestation per household is extremely difficult. Nonetheless, the census data represents a valuable historical record of changing demographics, production and land use that cannot be reconstru cted otherwise. Also, the census data can represent the rubber tappers estimates of production and land use within their particular cultural, economic and political context. This is very valuable in itself.

PAGE 136

124 APPENDIX B QUESTIONNAIRE USED IN CMER SURVEY Question rio no.________ Data: ___/___/___ Entrevistador:____________________ Coord. Geogrficas: Latitude ______ ______ ______ ______ S Longitude______ _______ _____ ______ W I Identificao da Colocao: 01 Seringal:_______________________________ Colocao:_________________________________ 02 Nome do entrevistado____________________________________________ 03 O seringal esta ligado a que Associao: _____AMOREX (Xapuri) _____AMOREB (Brasilia) _____AMOREAB (Assis Brasil) II Dados da familia A residentes na colocao No Nome Sexo M/F Parentesco Idade Estuda S/N 01 02 03 04 05 06 07 08 0 9 10 11 12 13 14

PAGE 137

125 B No residentes na colocao No Nome Sexo M/F Parentesco Idade #Onde mora? *Porque mudou? Tempo que saiu (anos) 01 02 03 04 05 06 # 1 Cidade mais prxima; 2 Capital; 3 outro seringal; 4 mesmo seringal; *1 Para estudar; 2 casou; 3 para trabalhar; 4 conflito com uso da colocao; 5 outros III Historia pessoal (chefe da colocao) 4 Em que estado voc nasceu? ________________ A Idade? ___________anos B Desde quanto tempo mora aqui? _________anos C Onde morava antes? Colocao_________Seringal___________________ Municpio___________Estado______ 5 Qual a forma de aquisico da sua colocao? ( ) Concesso de uso ( ) Herana ( ) Arrendamento ( ) Compra ( ) Posse ( ) diviso familiar ( ) Outro. Qual? ________________________ 6 Desenvolveu quais atividades, se voc j tiver morado na cidade_________________________ ___________________________________ 7 Voc j trabalhou em fazenda? ( ) Sim ( ) No 8 Voc j trabalhou com agricultura fora da reserva ? ( ) Sim ( ) No 9 Voc sabe se existe colocao sem morador neste seringal ? ( ) Sim No ( ) Quantas________________________________________ 10 Algum na sua comunidade se mudou recentemente? ( ) Sim ( ) No Quando?_______________________________________ 11 No passado, (quando a reserva foi criada ) haviam mais morados neste seringal? ( ) Sim No ( ) 12 Voc o primeiro ocupante desta colocao ( ) Sim No ( ) 13 Voc sabe quando esta colocao foi aberta ( ) Sim ( ) No Quando?___________________ _______________________________________ 14 Voc sabe quantos moradores j ocuparam esta colocao? ( ) Sim ( ) No Quantos?________________________________________

PAGE 138

126 IV Produ o (extrativismo e culturas anuais): A Dados gerais da Produo extrativista Tipos Mo de obra Produo (qde) Consum o (qde) Venda (qde.) Preo de venda (R$) H M 99 00 9 9 00 99 00 99 00 Castanha (latas) Borracha (kg) Mel de abe lha (litros) Copaba (litros) Outros: B Dados gerais da Produo agrcola Tipos Mo de obra Produo Consumo (qde.) Venda (qde.) Preo de venda (R$) H M 99 00 99 00 99 00 99 00 Arroz Kg Milho Kg Feijo Kg Mandioca Kg Outros: C Dados gerais da produo anual de animais Mo de obra Tipos Quantidade H M Consumo (qde.) Venda (qde.) Preo de venda (R$) Galinha Porco Pato Ovelha Gado Outro:

PAGE 139

127 V Uso da terra A Roados na colocao Roado Tamanho Idade Vezes utilizado Razo da escolha* 01 0 2 03 04 1 Capoeira; 2 Transio capoeira/mata; 3 Mata bruta; 4 Mata sem castanha e seringa; 5 perto da casa B Capoeira na colocao Capoeira Tamanho (hec) Idade Vezes utilizada Vegetao original 01 02 03 04 1 Restinga; 2 capoeiro; C Distribuio de uso da terra Tipos Tarefas Hectares rea de capoeira (total) rea de pasto (total) rea de roado (total) 15 Nmero total de estradas de seringa na colocao:_______ ________________________ 16 Quantas estradas voc cortou em: 1998_____ 1999_____ 2000_____ 17 Voc pretende cortar mais seringa ano que vem ( ) Sim ( ) No Porque?___________________________ 18 Est fazendo novo roado este ano? ( ) Sim No ( ) 19 Caso resposta negativa, porque no? ________________________________________________

PAGE 140

128 20 Qual o tamanho da rea de roado para este ano? Capoiera___________Mata bruta_______________ ______ 21 Voc Pretende aumentar a rea de cultivo agrcola: Sim ( ) ( )No Porque?______________________ 22 Quais os produtos agrcolas que voc pretende aumentar a produo? ( ) Arroz ( ) Feijo ( ) Milho ( ) Mandioca ( ) outros_____________________________________________ 23 Quando voc veio morar aqui j haviam reas desmatadas? Sim ( ) No ( ) 24 Caso sim, quais eram os usos: ( ) Capoeira abandonada ( ) Ro ados ( ) Pastagem 25 Algum na casa aposentado Sim ( ) No ( ) 26 Poderia dizer quanto voc gasta por ms, na mdia, com as despesas da casa? __________________________________________________________ 27 Voc j recebeu ou e st recebendo algum tipo de financiamento do governo? Sim ( ) No ( ) 29 Caso sim, quais foram? ( ) Prodex ( ) Outros________________ 30 Qual a finalidade?_________________________________________ 31 Caso resposta positiva Pretende obter outros financiamentos? Sim ( ) No ( ) 32 Quais?__________________________________________________ VI Transporte e comercializao 33 Para quem voc vende a produo extrativista? Produtos extrativista Local de venda Ncleo da Associao Cooperativa Comercio na cidade Marreteiro Castanha Borracha Outros:

PAGE 141

129 34 Para quem voc vende a produo agrcola? Produtos agrcolas Local de venda Ncleo da Associao Cooperativa Comercio na c idade Marreteiro Arroz Kg Milho Kg Feijo Kg Mandioca Kg Outros: 35 Como feito o transporte da produo? Trechos Tipo de acesso Meio de transporte Tempo Rio varadouro ramal rodagem barco animal carro A Colocao B C 36 De quem o transporte? ( ) prprio ( ) Ncleo da Associa o ( ) cooperativa ( ) Prefeitura ( ) fretado ( ) comunidade 37 Quantas vezes vai a cidade? ( ) uma vez por ms ( ) duas vezes por ms ( ) uma vez em dois meses ( ) uma vez em trs meses VII Extrutura Social e Participao 38 Voc Associado a alguma organizao comunitria? Sim ( ) ( )No 39 Caso resposta negativa, porque no participa?_____________ ______________________________________________________ 40 Caso resposta positiva, quais as organizaes ? Associao ( ) H quanto tempo? _________anos Si ndicato ( ) H quanto tempo? _________anos Cooperativa ( ) H quanto tempo? _________anos 41 Participou de alguma reunio este ano Sim ( ) No ( ) 42 O que foi discutido nesta reunio?__ ____________________ 43 Voc participa da execuo de trabalhos em multiro Sim ( ) No ( )

PAGE 142

130 44 Quantas vezes voc j participou de Assemblias geral da Associao? _____________________________________________________________________ 45 Tem algum tcnico extencionista trabalhando na sua comunidade? Sim ( ) No ( ) 46 Atravs do trabalho deste tcnico, que novas idias voc usa no seu dia a dia?______________________________________________________________ 47 Algum membr o de sua famlia trabalha for a da coloca o? Sim ( ) No ( ) 48 Em quais atividades?_____________________________________________ 49 Existem pessoas de fora trabalhando na sua colocao? ( ) No Sim ( ) 50 Em qu ais atividades?_____________________________________________ 51 Esta coloca o j foi dividida alguma vez? Sim ( ) N o ( ) 52 Caso resposta positiva, quando? _________(ano) 53 Quais os motivos para diviso? ( ) Casamento do fi lho (a) ( ) Arrendamento ( ) Para ter mais rea para desmatar ( ) Outros. Quais?_________________________________________ VIII Gesto da reserva 54 Neste seringal os moradores j tiveram problemas com caada por pessoas de f ora da comunidades Sim ( ) No ( ) 55 Pesca de pessoas de fora da comunidade Sim ( ) No ( ) 56 J aconteceu algum problema de posse de colocaes dentro da comunidade Sim ( ) No ( ) 57 Na sua comunidade j houve algum confl ito entre seringueiros e/ou pessoas de fora da reserva (fazendeiros, invasores, madeireiros) depois da criao da reserva? Sim ( ) No ( ) 58 Se a resposta for sim, descreva quais os conflitos:__________ ________________________________________ ______________ 59 Voc conhece o plano de utilizao da reserva Sim ( ) No ( )

PAGE 143

131 60 Voc acha que a reserva precisa de um plano de utilizao? ( ) Sim ( ) No Porque? __________ ___ _____________________________________ 61 Voc ac ha que aps a aprovao do plano de utilizao houveram mudanas no dia a dia dos moradores da comunidade? Sim ( ) No ( ) Quais foram as mudanas?__________ ___ _____________________________________ 62 Voc acha que devem ser feitas mudanas no Plano de Utilizao ? Sim ( ) No ( ) Quais?______________________ __ ____ 63 Voc sabe quanto o plano de utilizao autoriza desmatar? Total da colocao? Certo ( ) Errado ( ) Por ano? Certo ( ) Er rado ( ) 64 J veio algum aqui para fazer fiscalizao nesta colocao? ( ) Sim ( ) No 65 De que organizao ele era? ( ) IBAMA ( ) Associao ( ) Sindicato ( ) Outros 66 Voc estar pensando em mudar p ara outro lugar? Sim ( ) ( ) No. Caso sim, para onde pretende mudar?___ _ Porque?____________________________________________________ 67 Na sua opinio qual o principal problema encontrado na sua comunidade?____________ ____________________________________

PAGE 144

132 LIST OF REFERENCES Albert, B.1992. Indian lands, environmental policy and military geopolitics in the development of the Brazilian Amazon: the case of the Yanomami. Development and Change. 23 (1):35 70. Allegretti, M. H.1979. Os Seringue iros: Estudo de Caso de um Seringal Nativo do Acre. Dissertao de Mestrado, Universidade de Braslia. Braslia. Allegretti, M. H.1989. Reservas extrativistas: uma proposta de desenvolvimento da floresta amaznica. Par Desenvolvimento 25:2 29. Alleg retti, M. H.1990. Extractive reserves: an alternative for reconciling development and environmental conservation in Amazonia. In A. B. Anderson (ed.), Alternatives to Deforestation: Steps toward Sustainable Use of the Amazonia Rain Forest, pp. 252 264. Ne w York: Columbia University Press. Allegretti, M. H.1994. Reservas extrativistas: parmetros para uma poltica de desenvolvimento sustentvel na Amaznia. In R. Arndt (ed.), O Destino da Floresta: Reservas Extrativistas e Desenvolvimento Sustentvel na A maznia, pp. 17 47. Rio de Janeiro: Dumar. Allegretti, M. H. 1995. The Amazon and extracting activities. In M.C. Godt and I. Sachs (eds.), Brazilian Perspectives on Sustainable Development of the Amazon Region pp. 157 174. Paris: UNESCO. Alves, D. (i n press). An analysis of the spatial patterns of deforestation in Brazilian Amazon in the 1991 1996 period. In C. Wood and R. Porro (eds.), Patterns and Process of Land Use and Forest Change in the Amazon Gainesville: Center for Latin American Studies, Un iversity of Florida. Anderson, A. B.1989. Estratgias de uso da terra para reservas extrativistas na Amaznia. Par Desenvolvimento 25 :30 37. Anderson, A. B., ed. 1990. Alternatives to Deforestation: Steps Toward Sustainable Use of the Amazon Rain Fores t New York: Columbia University Press. Anderson, A. B.1992. Land use strategies for successful extractive economies in Amazonia. In D. C. Nepstad and S. Schwartzman (eds.), Non Timber Products From Tropical Forests: Evaluation of a Conservation and Deve lopment Strategy, pp. 67 78. New York: The New York Botanical Garden.

PAGE 145

133 Anderson, A. B. 1994. Extrativismo vegetal e reservas rxtrativistas. In R. Arndt (ed.), O Destino da Floresta: Reservas Extrativistas e Desenvolvimento Sustentvel na Amaznia, pp. 227 247. Rio de Janeiro: Dumar. Anderson, A. B. and E. M. Ioris 1992. Valuing the rain forest: economic strategies by small scale forest extractivists in the Amazon estuary. Human Ecology 20(3):337 368. Anderson, A. B. and M. A. G. Jardim. 1989. Costs and benefits of floodplain forest management by rural inhabitants in the Amazon estuary: a case study of aai production. In J. O. Browder (ed.), Fragile Lands in Latin Amercia: Strategies for Sustainable Development, pp. 114 129. Boulder: Westview Press. Bakx, K. 1988. From proletarian to peasant: rural transformation in the state of Acre, 1870 1986. The Journal of Development Studies 24 :141 160. Balick, M. and P. Cox. 1996. Plants, People and Culture: The Science of Ethnobotany New York: Scientific A merican Library. Balick, M. and R. O. Mendelsohn. 1995. The value of undiscovered pharmaceuticals in tropical forests. Economic Botany 49 (2):223 228. Barbosa, L. C. 1993. The "greening" of ecopolitics of the world system: Amazonia and the change in the ecopolitics of Brazil. Journal of Political and Military Sociology 21 :107 134. Barbosa, L. C. 2000. The Brazilian Amazon Rainforest: Global Ecopolitics, Development, and Democracy New York: University Press of America. Barraclough, S. and K.B. Ghimir e. 1995. Forests and Livelihoods: The Social Dynamics of Deforestation in Developing Countries New York: ST. Martin's Press. Bodmer, R. E. 1997. Linking conservation and local people through sustainable use of natural resources. In C. Freese (ed.), Harv esting Wild Species: Implications for Biodiversity Baltimore: Johns Hopkins University Press. Brondizio, E. S., E. F. Moran, P. Mausel, and Y. Wu. 1994. Land use change in the Amazon estuary: patterns of cabloco settlement and landscape management. Huma n Ecology 22 (3):249 277. Brondizio, E. S. and A. D. Siqueira. 1997. From extractivists to forest farmers: changing concepts of cabloco agroforestry in the Amazon estuary. Research in Economic Anthropology 18 :233 279.

PAGE 146

134 Brondizio, E. S., S.D McCracken, E .F. Moran, A.D.Siqueira, D.R. Nelson, and C.R. Pedraza. (in press). The colonist footprint: towards a conceptual framework of land use and deforestation trajectories among small farmers in frontier Amazonia. In C. Wood and R. Porro (eds.), Patterns and Pro cesses of Land Use and Forest Changes in the Amazon Gainesville: Center for Latin American Studies, University of Florida. Browder, J. O. 1988. The social costs of rainforest destruction. Interciencia 13 (3):115 126. Browder, J. O. 1990a. Extractive r eserves will not save the tropics. BioScience 40 :626. Browder, J. O. 1990b. Beyond the limits of extraction: tropical forest alternatives to extractive reserves. The Rainforest Harvest London. Browder, J. O. 1992a. Social and economic constraints on th e development of market oriented extractive reserves in Amazon rain forests. In D.C. Nepstad and S. Schwartzman (eds.), Non Timber Products From Tropical Forests: Evaluation of a Conservation and Development Strategy pp. 33 43. New York: The New York Bota nical Garden. Browder, J. O. 1992b. The limits of extractivism: tropical forest strategies beyond extractive reserves. BioScience 42 (3):174 182. Browder, J.O and B. J. Godfrey 1997. Rainforest Cities: Urbanization, Development and Globalization of the Brazilian Amazon New York: Columbia University Press. Brown, K. and S. Rosendo. 2000a. Environmentalists, rubber tappers and empowerment: the politics and economics of extractive reserves. Development and Change 31 : 201 227. Brown, K. and S. Rosendo 2000b. The institutional architecture of extractive reserves in Rondnia, Brazil. The Geographical Journal 166 (1):35 48. Calaa, M. 1993. Violncia e Resistncia: O Movimento dos Seringueiros de Xapuri e a Proposta de Reserva Extrativista. Tese de Dout orado em Geografia. Universidade Estadual Paulista. Rio Claro. Dean, W. 1987. Brazil and the Struggle for Rubber Cambridge: Cambridge University Press. Friends of the Earth. 1998. Polticas Pblicas Coerentes para uma Amaznia Sustentvel Braslia.

PAGE 147

135 ELI. 1995. Brazil's Extractive Reserves: Fundamental Aspects of their Implementation. Washington, D.C.: Environmental Law Institute. Fearnside, P. M. 1989. Extractive reserves in Brazilian Amazonia. BioScience 39 :387 393. Fearnside, P. M. 1992. Reserva s extrativistas: uma estratgia de uso sustentado. Cincia Hoje 14 (81):15 17. Fearnside, P. M. 2001. Land tenure issues as factors in environmental destruction in Brazilian Amazonia: the case of southern Par. World Development 29(8):1361 1372. Fisher W. 1994. Megadevelopment, environmentalism, and resistance: the institutional context of Kayap indigenous politics in central Brazil. Human Organization 53 (3):220 232. FUNTAC. 1990. Monitoramento e Cobertura Florestal do Estado do Acre Rio Branco: Go verno do Acre. Godoy, R. A. and K. S. Bawa. 1993. The economic value and sustainable harvest of plants and animals from the tropical forest: assumptions, hypotheses, and methods. Economic Botany 47 (3):215 219. Gomez Pompa, A. and A. Kaus. 1990. Traditi onal management of tropical forests in Mexico. In A. B. Anderson (ed.), Alternatives to Deforestation: Steps Toward a Sustainable Use of the Amazonian Rain Forest pp. 45 64. New York: Colombia University Press. Gottlieb, O. R. and M. A. Kaplan. 1990. Am aznia tesouro quimico a preservar. Cincia Hoje 6 : 44 50. Hall, A. 1989. Developing Amazonia: Deforestation and Social Conflict in Brazil's Carajs Programme Manchester: Manchester University Press. Hecht, S. B. 1989. Murder at the margins of the wor ld. Report on the Americas 23(1) : 36 38. Hecht, S. B. and A. Cockburn. 1989. The Fate of the Forest: Developers, Destroyers, and Defenders of the Amazon. London: Verso. Homma, A. K. O. 1989. Reservas extrativistas: uma opo de desenvolvimento vivel para a Amaznia? Par Desenvolvimento 25: 38 48. Homma, A. K. O. 1992. The Dynamics of extraction in Amazonia: a historical perspective. In D.C. Nepstad and S. Schwartzman (eds.), Non Timber Products

PAGE 148

136 From Tropical Forests: Evaluation of a Conservation an d Development Strategy, pp. 23 33. New York: The New York Botanical Garden. Homma, A. K. O. 1993. Extrativismo Vegetal na Amaznia: Limites e Oportunidades Brasilia: EMBRAPA SPI. IBAMA/CNPT. 1999. Projeto RESEX: Um Futuro Sustentvel para a Amaznia. Braslia: MMA/IBAMA/CNPT. IDEAS. 1993. Levantamento Preliminar da Base de Recursos Naturais da Reserva Chico Mendes. Braslia. INPE. 2000. Monitoring of the Brazilian Amazon Forest by Satellite, 1999 2000. 05/22/01. http://www.inpe.br/Informacoes_Even tos/amz1999_2000/Prodes/index.htm. Kainer, K. 1995. In a nutshell: population ecology of Brazil nut harvest. Doctoral Dissertation. University of Florida. Gainesville. Mahar, D. J. 1978. Desenvolvimento Econmico da Amaznia Uma Anlise das Poltic as Governamentais Rio de Janeiro: IPEA/INPES. Martine, G. and L. Camargo. 1998. Growth and distribution of the Brazilian population: recent trends. Brazilian Journal of Population Studies 1 :59 81. May, P. H. 1989. Direitos de propriedade e a sobreviv ncia das economias extrativistas. Par Desenvolvimento 25 (4):65 70. Mendes, A., A. Salassie, and D. Penha. 1996. Pilot Program to Conserve the Brazilian Rain Forest. Brasilia: Ministry of the Environment, Water Resources and the Legal Amazon. Millikan, B. H. and M. A. Irving. 1997. Programa Piloto Para a Proteco das Florestas Tropicais Brasileiras (Relatrio de consultoria). Banco Mundial/Comisso Europeia/IBAMA. Brasilia. MMA/IBAMA/CNPT. 1995. Plano de Utilizao da Reserva Extrativista Chico Mendes Braslia. Moran, E. 1981. Developing the Amazon Bloomington: Indiana University Press. Moran, E., ed. 1983. The Dilemma of Amazonian Development Westview Press, Inc. Moran, E. 1993. Through Amazonian Eyes: The Human Ecology of Amazonian Populatio ns Iowa: University of Iowa Press.

PAGE 149

137 Moran, E. and E. Brondizio. 1994. Integrating Amazonian vegetation, land use and satellite data. BioScience 44(5): 329 339. Murrieta, J. R. and R. P. Rueda. 1995. Extractive Reserves Gland: IUCN CNPT EU. Nepstad, D C., A.Verissimo, A. Alencar, C. Nobre, E. Lima, P. Lefebvre, P. Schlesinger, C. Potter, P. Moutinho, E. Mendoza, M. Cochrane, and V. Brooks. 1999. Large scale impoverishment of Amazonian forests by logging and fire. Nature 398 :505 508. Nepstad, D. C., I. F. Brown, L. Luz, A. Alechandre and V. Viana. 1992. Biotic impoverishment of Amazonian forests by rubber tappers, loggers, and cattle ranchers. In D. C. Nepstad and S. Schwartzman (eds.), Non Timber Products From Tropical Forests: Evaluation of a Conser vation And Development Strategy New York: The New York Botanical Garden. Nepstad, D. C. and S. Schwartzman, eds. 1992. Non Timbert Products From Tropical Forests: Evaluation of a Conservation and Development Strategy New York: The New York Botanical Ga rden. Oliveira, A. U. 1991. Integrar Para No Entregar: Polticas Pblicas e Amaznia Campinas: Papirus. Panayotou, T. and P. Ashton. 1992. Not by Timber Alone: Economics And Ecology for Sustaining Tropical Forests Washington, D.C.: Island Press. P eralta, P. and P. Mather. 2000. An Analysis of deforestation patterns in the extractive reserves of Acre, Amazonia from satellite imagery: a landscape ecological approach. Remote Sensing 21(13) :2555 2570. Perez, M. R. and N. Byron. 1999. A methodology to analyze divergent case studies of non timber forest products and their development potential. Forest Science 45 :1 14. Perz, S. G. 2001a. From sustainable development to "productive conservation:" forest conservation options and agricultural income and a ssets in the Brazilian Amazon. Rural Sociology 66(1) :93 112. Perz, S. G. 2001b. Household demographic factors as life cycle determinants of land use in the Amazon. Population Research and Policy Review 00 :1 28. Peters, C. M., A. H. Gentry and R.O. Mend elsohn. 1989. Valuation of an Amazonian rainforest. Nature 339 :655 656.

PAGE 150

138 Pinard, M. 1991. Impacts of Stem Harvesting on Populations of Iriartea Deltoida (Palmae) in an Extractive Reserve in Acre, Brazil. Master's Thesis. University of Florida. Gainesville. Pinedo Vasquez, M., D. Zarin, P. Jipp and J. Chota inuma. 1990. Use values of tree species in a communal forest reserve in northeast Peru. Conservation Biology 4(4): 405 416. Plotkin, M. and L. Famolare, eds. 1992. Sustainable Harvest and Marketing of Rain Forest Products Washington, D.C.: Island Press. PMACI I. 1990. Diagnstico Geoambiental e Scio econmico: rea de Influncia da BR 364 Trecho Porto Velho/Rio Branco. Rio de Janeiro: IBGE IPEA. Posey, D. A. and W. Bale, eds. 1989. Resource Manag ement in Amazonia: Indigenous and Folk Strategies Advances in Economic Botany. New York: The New York Botanical Garden. RADAMBRASIL. 1976. Folha SC 19 Rio Branco: geologia, geomorfologia, pedologia, vegetao e uso potencial da terra. Rio de Janeiro: Go verno do Brasil. Ramalho, C. 1992. Cooperativa agro extrativista Chico Mendes. Ecologia e Desenvolvimento 1 (11):34 39. Rancy, C. M. D. 1992. Razes do Acre (1870 1912). Rio Branco: Editora Paim. Redford, K. H. and C. Padoch, eds. 1992. Conservation o f Neotropical Forests New York: Columbia University Press. Rego, J. F. 1999. Amaznia: do extrativismo ao neoextrativismo. Cincia Hoje 25(147):62 65. Rignot, E. 1997. Mapping deforestation and secondary growth in Rondnia, Brazil, using imaging rada r and thematic mapper data. Remote Sensing Environment 59 :167 179. Salafsky, N. B., B.L. Dugelby, J.W. Terborgh. 1993. Can extractive reserve save the rain forest? An ecological and socio economic comparison of nontimber forest product extraction systems in Petn, Guatemala, and west Kalimantan, Indonesia. Conservation Biology 7(3) :39 52. 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 R eserva Extrativista Chico Mendes, Estado do Acre. Dissertao de Mestrado, Instituto Nacional de Pesquisas Espaciais. So Jos dos Campos

PAGE 151

139 Schmink, M. 1992. Building institutions for sustainable development in Acre, Brazil. In K. Redford and C. Padoch (e ds.), Conservation of Neotropical Forests New York: Columbia University Press. Schmink, M. 1999. New hope for environmental policies in Acre, Brazil. Latinamericanist 35 (1):1 2. Schmink, M. and C. H. Wood. 1984. Frontier Expansion in Amazonia Gainesv ille: University of Florida Press. Schmink, M. and C. H. Wood. 1992. Contested Frontier in Amazon New York: Columbia University Press. Schwartzman, S. 1989. Extractive reserves: the rubber tappers' strategy for sustainable use of the Amazon rain fores t. In J. O. Browder (ed.), Fragile Lands in Latin America: Strategies for Sustainable Development pp. 150 165. Boulder: Westview Press. Schwartzman, S. 1991. Deforestation and popular resistance in Acre: from local movement to global network. The Centen nial Review 25 (2):397 422. Schwartzman, S. 1992. Land distribution and the social costs of frontier development in Brazil: social and historical context of extractive reserves. In D. C. Nepstad and S. Schwartzman (eds.), Non Timber Products From Tropical Forests: Evaluation of a Conservation and Development Strategy, pp. 51 66. New York: The New York Botanical Garden. Skole, D. and C. Tucker 1993. Tropical deforestation and habitat fragmentation in the Amazon: satellite data from 1978 to 1988. Science 2 60 :1905 1910. Skole, D. 1994. Physical and human dimensions of deforestation in Amazonia. BioScience 44 :314 322. Smith, N. 1982. Rainforest Corridors: The Transamazon Colonization Scheme Berkeley: University of California Press. Stern, P., O. R. Young and D. Druckman 1992. Global Environmental Change Washington D.C.: National Academy Press. Tocantins, L. 1979. Formao Histrica do Acre Rio de Janeiro: Civilizao Brasileira. Uhl, C. and G. Parker 1986. Is a quarter pounder worth a half ton of r ain forest? Interciencia 11 (5):210.

PAGE 152

140 Viola, E. 1988. The ecological movement in Brazil (1974 1986): from environmentalism to ecopolitics. International Journal of Urban and Regional Research 12 (2):211 228. Weinstein, B. 1983. The Amazon Rubber Boom 1850 1920 Stanford: Stanford University Press. Wood, C. H. and D. Skole 1998. Linking satellite, census, and survey data to study deforestation in the Brazilian Amazon. In D. Liverman, E. Moran, R. Rindfuss, and P. Stern (eds.), People and Pixels: Linking Remote Sensing and Social Science, pp. 70 93. Washington D.C.: National Academy Press. Wood, C. H. and J. Wilson 1991. The magnitude of migration to the Brazilian frontier. In M. Schmink and C. H. Wood (eds.), Frontier Expansion in Amazonia, pp. 142 152. Gainesville: University of Florida Press.

PAGE 153

141 BIOGRAPHICAL SKETCH Carlos Valrio Gomes was born on October 24, 1968, in Ourm, State of Par, Brazil. In 1989, he moved to Rio Branco, State of Acre in southwest Amazonia. There, at the Federal University of Acre (UFAC), he received his undergraduate d egree in Geography in 1995. Upon finishing his study at UFAC, he spent 2 years as both an associate researcher with the Zoobotanical Park of UFAC and as a consultant to the on going Extractive Reserves Project/Pilot Program for the Protection of the Brazil ian Rainforests (PPG 7), under contract with the United Nations Development Program (UNDP). He was involved in the first large scale settlement mapping of the Chico Mendes Reserve, community training programs, socio economic surveys as well as preparation of the reserve wide development plan. In 1999, he was awarded a Fulbright Fellowship (Amazon Basin Program) for Masters level studies in the United States. This thesis is the result of his masters program in Latin American Studies, with a concentration in Tropical Conservation and Development, at the University of Florida. In 2001, he was awarded a scholarship from the Brazilian Science Council (CNPq), of the Ministry for Science and Technology of Brazil, to continue his graduate studies seeking a doctorate degree in Geography at the University of Florida.


Permanent Link: http://ufdc.ufl.edu/UFE0000320/00001

Material Information

Title: Dynamics of Land Use In an Amazonian Extractive Reserve: The Case of the Chico Mendes Reserve In Acre, Brazil
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000320:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000320/00001

Material Information

Title: Dynamics of Land Use In an Amazonian Extractive Reserve: The Case of the Chico Mendes Reserve In Acre, Brazil
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000320:00001


This item has the following downloads:


Full Text











DYNAMICS OF LAND USE IN AN AMAZONIAN EXTRACTIVE RESERVE: THE
CASE OF THE CHICO MENDES EXTRACTIVE RESERVE IN ACRE, BRAZIL

















By

CARLOS VALERIO AGUIAR GOMES


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS

UNIVERSITY OF FLORIDA


2001




























Copyright 2001

by

Carlos Valerio Aguiar Gomes

































To Leinha for everything















ACKNOWLEDGMENTS

This thesis could not have been completed without the help and support of a

variety of friends, colleagues and mentors. First I thank Dr. Michael Binford, my

supervisor for all his help, and constructive criticism, which helped me to strengthen my

intellectual independence. I am very thankful to the members of my committee: Dr.

Marianne Schmink, Dr. Nigel Smith and Dr. Stephen Perz for their valuable comments,

suggestions and guidance. I specially wish to express my deepest appreciation to Dr.

Schmink for being a constant source of support and encouragement since I first thought

about coming to study in the United States.

I gratefully acknowledge the support of the Latin American Scholar Program for

American Universities (LASPAU)-Fulbright Commission for the fellowship that allowed

my enrollment in the graduate school at the University of Florida. In LASPAU I want to

thank Ms. Sonia Wallemberg for her wonderful assistance and friendship. I am also

grateful to Ms. Rebecca Smith-Murdock and Mr. Amir Abbassi (from the Intensive

English Language Institute of the University of North Texas) for the fellowship for

English language training. Moreover, I would like to thank the World Wildlife Fund

(WWF-Brazil), the Tropical Conservation and Development Program (TCD) at the

University of Florida, and the Acre Environmental Agency (IMAC) for funding my

research in Acre.

My heartfelt gratitude goes to my friend and former advisor Dr. Irving Foster

Brown at Federal University of Acre (UFAC). I thank him for his outstanding work









helping me and many other Amazonian students to make realities out of our dreams.

Without his help, mine would have stayed asleep. Also, I want to thank my colleagues

and friends who have given me a great deal of help and encouragement to get to this

point in my career: Andrea Alechandre, Silvia Brilhante, Hirome Sassagawa, Atila

Cristina, William Flores, Eufran Amaral, Elsa Mendosa, Benedita Esteves and Monica de

los Rios at the Zoobotanical Park at the UFAC; Arthur Leite, Juceir and Nuria Merched,

at the Acre State Department of Forest and Extractivism; Jorge Reboucas at the Rubber

Tapper National Council; Josemar Amorim, Manoel Feitosa and Jose Alberto at the

National Center for Traditional Populations; Veronica Passos and Edgard de Deus at the

Acre Environmental Agency. I also wish to express my deepest gratitude to Christiane

Ehringhaus at Yale University for helping me very much throughout this entire work.

I am flattered with the warmth from the rubber tapper families of the Chico

Mendes reserve whom have taken me into their homes and shared their knowledge with

me. I own them more than I can say. I also wish to thank Francisco C. dos Santos, my

field assistant, for his great help. In addition, want to thank my friends at the University

of Florida for their support: especially Olendina Cavalcante, Stacey Shankle, Brandon

Knox, David Salisbury, Ronaldo Weigand and Richard Wallace.

Last but definitely not least, I would like to express my love and thanks to my

parents and family for continuously providing me with love and understanding. All of us

agree that eleven years ago this degree would have been unthinkable. I specially thank

my nephew Sallon Saad and my niece Susy Leia for their love and sweet words, which

always energize me and remind me of what is important in life.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST O F TA B LE S ........................ ........................ .. .. .... ........ ......... viii

LIST O F FIG U RE S .......................... ...................... .. .. ........... ..... .. ....ix

A B STR A C T ................................................... ....... ............ ................. xi

CHAPTERS

1 IN TR O D U C TIO N ....................... ........................... .. ........ ..............

1.1 Problem Statement ................................................................. ........ 5
1.2 Central Questions ........ ...... .................................. ............ 8
1.3 Structure of the Study ....................... .................... .............. 9

2 DEBATE ABOUT EXTRACTIVE RESERVES...............................................11

2.1 Brief History of the Amazon Rubber Boom................................ ................. 11
2.2 Evolution of the Extractivist Development Model in Amazonia........................... 12
2.3 Analysis of the Extractive Reserve Literature ............................................... 16
2.3.1 General Patterns and Trends in the Debate ................................... .............. 17
2.3.2 Social Development Arguments ....................................................... 20
2.3.3 Econom ic A rgum ents ......................................................... .............. 21
2.3.4 E ecological A rgum ents.................. ........................... ............ ............ .. 23
2.4 Achievements and Problems with Extractive Reseves .............. ...................... 24

3 STUDY SITES AND RESEARCH METHODOLOGY .............................................27

3.1 Study Site Selection ......... ................ ............................ .............. 27
3.2 Survey Setting and Planning ....................................................... ........ .... 30
3 .3 Su rv ey M eth odology ....................................................... .................................. 3 1
3.4 Other sources of D ata ...... .. ........ ........................ ........... .... .......... 31
3.5 D ata Processing V ariables and A nalysis............................................. ... ................. 32
3.6 Environmental Characterization of the CMER................................... ........ 34
3 .6 .1 C lim ate ............................ ............... ............... ............... ........... ..... 3 4
3.6.2 W atershed/H ydrology ................................................................................ 35
3 .6 .3 G eo lo g y ................................................................................ 3 5









3.6.4 P edology (Soils)... .............. ............................ ........................ ................ .. 37
3.6.5 V egetation............................................ 37

4 DEFORESTATION TRENDS IN THE EXTRACTIVE RESERVE............................41

4 .1 Introdu action ............................ .... ............................................... 4 1
4.2 Deforestation and Land Use Trends in the State of Acre .................................. 42
4.3 Deforestation rates in the CM ER ..................................................................... 44
4.4 Deforestation at the Seringal Level................................................. 48

5 HOUSEHOLD DETERMINANTS OF FOREST AND NONFOREST ACTIVITIES 54

5 .1 In tro d u ctio n ............................ ..................... ................................ 5 4
5.2 Characterization of Rubber Tappers' Settlements in the CMER...................... 54
5.3 Seasonality The Rubber Tapper's Year .............. ............ ...... ........... ......... 55
5.4 Conceptualization of Outcome and Independent Variables........................... 59
5.4.1 O utcom e V ariables.................................................... .......................... 59
5.4.2 Independent V ariables.................................................................................. 63
5.5 Correlations Between Outcome and Independent Variables ............................. 66
5.6 Analysis of Multivariable Regression Models......................... .............. 69
5 .7 D iscu ssion ...... ..................................................... .......... ..... 73

6 DRIVERS OF LAND USE CHANGE IN THE CMER............... ................79

6.1 Introdu action ................................................ 79
6.2 Population Dynamics in the CMER.................................. .............. 81
6.3 Population Dynamics in the Study Seringais............. ..... ............ 86
6.4 Population Density and Deforestation in the Study Seringais ............................ 91
6.5 Settlement Dynamics in the Study Seringais......................... ........... 94
6.6 Changes in the Production System: Seringal Paraguau ............................... 98
6.7 Pasture and Cattle Raising Expansion in the Study Seringais ............................ 102
6.8 Social Organization in the Reserve...... .............................. ..................... .............. 105

7 SUM M ARY AND TRAN SITION ................................... .............................. ....... 112

APPENDICES

A DATA CENSUS RELIABILITY AND UNCERTAINTY OF THE CMER............. 121

B QUESTIONNAIRE USED IN CMER SURVEY ............................ ..................124

L IST O F R E FE R E N C E S ...................................................................... ..................... 132

BIOGRAPHICAL SKETCH .............. ............. .......... ..................... ............... 141
















LIST OF TABLES


Table Page

4-1 Total area and percentage deforestation in each seringal of the CMER.......................49

4-2 Seringais within the CMER with the highest deforestation rates............................. 52

5-1 Variable names, operational definition, and descriptive statistics...............................60

5-2 Correlations between outcomes and independent variables..................... ..............67

5-3 M ultivariable regression m odels........................................................... ............... 70

6-1 Previous residences of the households in the study seringais .....................................89

6-2 Time of residence of households in the study seringais .................................................91

6-3 Deforestation levels and rates in the study seringais ..................................................92

6-4 Settlements subdivision in the study seringais. .................................... .................97

6-5 Reasons for settlement subdivision in the study seringais .....................................98

6-6 Membership in association and rural worker union in the study seringais....................109















LIST OF FIGURES


Figure Page

3-1 Maps of the study site showing the seringais in Xapuri and Brasileia area....................28

3-2 Maps of the study site showing the seringais and settlements in Assis Brasil area.........29

3-3 Map of the mean rivers bordering the CMER ...... .............................................36

3-4 Map of forest typology occurring in the CMER.....................................................3..8

4-1 Location of the Chico Mendes Extractive Reserve in Acre ................. ............... 42

4-2 Deforestation and rubber production in Acre ...................................... ............... 45

4-3 Satellite image mosaic of the CMER illustrating deforestation (1986)...........................46

4-4 Satellite image of the CMER illustrating deforestation (1992-1998)..............................47

4-5 Percent of deforestation in the CM ER ............... ...................................... ........ ... 48

4-6 Map of the eight seringais with the highest deforestation rates in the CMER ................51

5-1 Representation of the settlement in the CMER ............................. .........56

5-2 Seasonality in the rubber tapper's year................................................. ...............57

6-1 Total population of the CMER in three municipalities ...............................................83

6-2 Total population in the study seringais ........................................ ........ ............... 87

6-3 Relationship of deforestation rates vs. population density in the study seringais. .........93

6-4 Total number of settlements in three municipalities of the CMER..............................95

6-5 Total numbers of settlements in the study seringais.....................................................96

6-6 Average production increments per settlement in the seringal Paraguacu ....... ....100

6-7 Average agricultural productions per household: Seringal Paraguacu..........................101

6-8 Average pasture area per settlement in the study seringais .......................... ..........104











6-9 Total memberships in CMER associations......................... ...............107

































































x















Abstract of thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts

DYNAMICS OF LAND USE IN AN AMAZONIAN EXTRACTIVE RESERVE:
THE CASE OF THE CHICO MENDES EXTRACTIVE RESERVE IN ACRE, BRAZIL

By
Carlos Valerio Aguiar Gomes
December 2001

Chairman: Dr. Michael Binford
Major Department: Latin American Studies

Since the Chico Mendes Extractive Reserve (CMER) was established, different

levels of land use change have taken place as a response to local and regional economic

forces and prevailing social stresses. The following study addresses these changes and

their drivers by analyzing key household characteristics and socio-economic factors. This

study provides both a household and a rubber tapper estates (seringal) level approach of

measuring land-use changes.

Based on remote sensing methods, four seringais with the highest deforestation

rates were selected as the focus of analysis. The reserves' census data on all residents of

the four seringais was analyzed for three different periods (1995/1998/2000). In addition,

in-depth interviews were carried out with 66 households. The household-level analysis

was based on the data gathered from the 66 families, whereas at the seringal-level the

research explores census data. The former approach considers three groups of

independent variables: background of household head (migrant status and age), labor-









force availability (sons of the household head) and location of the seringais. These were

then measured in relation to six outcome variables per settlement: rubber and Brazil nut

production, size of swidden plot, size of pasture, head of cattle and total area deforested.

The serinigal level analysis concentrated on three main factors: population, settlement

dynamic, and changes in the rubber tapper production systems.

Household level findings shows that the age of the household heads and the

locations of the seringais have a strong association with rubber and Brazil nut production,

which suggest that older household heads tend to engage in forest activities. The

availability of sons of household heads exerts a strong effect on pasture and cattle raising

activities, which suggests that non-forest activities tend to be carried out by the young

residents. The migrant status accentuates agricultural activities, implying that non-

migrant households tend to engage in forest activities. Seringal level findings reveal that

population density have increased significantly in the four seringais. In addition, it shows

a crucial process in occupation of space within these seringais is increasing the

subdivision of the settlements. Moreover, rubber tapper production system has been

increasingly transformed from extractive to agricultural and pastoral production

activities. These increases in population density, number of settlements and change in the

production systems accelerate deforestation, which will pose a serious dilemma for land

use regulations in the reserve in the near future.

This work is a step toward illuminating the prevailing issues on land-use changes

in the reserve. To understand which factors are driving these changes is a timely

endeavor that will hopefully contribute to the strengthening of sustainable land-use

management strategies in the CMER and in other extractive communities in Amazonia.














CHAPTER 1
INTRODUCTION

Amazon deforestation and its social and environmental consequences have been a

major concern in the last few decades. The state of Acre in southwestern Brazilian

Amazon is the birthplace of the rubber tapper movement, which was the first grass-roots

movement in Brazil to advocate conservation of Amazonian forests through the

establishment of extractive reserves. The creation of extractive reserves has been

promoted as a major strategy for forest conservation while simultaneously providing a

sustainable economic return to local peoples. A decade after the establishment of the

federal extractive reserves in Amazonia, however, the extractive reserve model is still

facing many challenges to put these goals into practice. The present work examines the

social dynamics of smallholder deforestation in the Chico Mendes Extractive Reserve

(CMER), and analyzes the socio-economic forces behind land-use change in the reserve.

The impoverishment of tropical forests through deforestation is a world-wide

phenomenon. In recent decades, few if any environmental issues have received as much

global attention as "tropical deforestation." Tropical deforestation will undoubtedly

continue to be a central international environmental issue in coming years particularly as

global issues such as climate change and the related carbon sinks and sources are

discussed. Brazil, as owner of both the largest portion of the world's rainforests and the

highest absolute deforestation rate, is defacto a leader in both conservation and

destruction of these forests. There is a wide consensus that deforestation has

environmental consequences, but there is considerable controversy over its social impacts









and extent. In fact, the socio-economic dynamics of deforestation in the Brazilian

rainforests and its implications continue to be rather poorly understood, if the widely

conflicting claims about its causes, impacts and remedies can be taken as evidence.

Brazil's legal Amazon1 region includes about 5 million square kilometers, which

comprises 58% of the country's territory. The regional development policies

implemented by military governments from the mid-1960s through the 1970s have

engaged in programs aimed at opening up the Amazon region to development (Mahar

1978). These policies included massive highways to link these projects to the rest of the

country, such as, the Transamazonica, the Belem-Brasilia, and the Cuiaba-Porto Velho

roads. These "development" strategies also included economic programs for mineral

extraction (e.g., Grande Carajas), agricultural colonization and cattle ranching (e.g.,

Poloamaz6nia, Polonoroeste), as well as incentives to timber industries. Therefore, in

recent decades massive deforestation due to large-scale cattle ranching, colonization,

logging, and mining projects has significantly reduced rainforests in the Brazilian

Amazon and in some cases caused violent conflicts and severe socio-economic problems

for local communities (Moran 1981, Schmink and Wood 1984, Smith 1982).

Deforestation rates in the Brazilian Amazon ranged from 29,059 square

kilometers in 1995 to 18,161 square kilometers in 1996; then 13,227 square kilometers in

1997, and 17,383 square kilometers in 1998, according to Brazil's National Institute for

Space (INPE 2000). From August 1998 to August 1999, the mean annual rate of



1 Two concepts are used to define the boundaries of the Brazilian Amazon region. The first and more
commonly used definition is "Legal Amaz6nia", which was established by military government in 1966.
For administrative and planning purposes it comprises nine states: Acre, Amapa, Amazonas, Mato Grosso,
Pari, Rond6nia, Roraima, as well as part of Tocantins and Maranhio. The second definition, "Classical
Amazon", the area covered by tropical forests, includes about 3.5 million square kilometers (40 per cent of
the country). It includes the Brazilian states of Acre, Amapa, Amazonas, Pari, Rond6nia, and Roraima.









deforestation was 17,259 square kilometers. The provisional estimate for the period of

1999/2000 is 19,836 square kilometers, which represents an increase of approximately

14.9% (INPE 2000).

These data clearly show that deforestation rates have not decreased in Amazonia

despite efforts by the Brazilian government to slow them. Yet the Landsat TM satellite

images used by INPE do not include deforestation of areas smaller than 6.25 hectares, so

not even all deforestation is being counted. The impact of hundreds of thousands of

small-scale farmers, including rubber tappers as well as logging activities and wild fires

are not included. Logging crews alone severely damage 10,000 to 15,000 square

kilometers a year of forest that are not included in deforestation mapping programs

(Nepstad et al. 1999).

Despite being considered an agricultural frontier, the Amazon has most of its

population concentrated in cities. In 1940, the population of Amazonia was only 1.5

million people, of whom 72% were rural dwellers (Barraclough and Ghimire 1995).

According to the Demographic Census of 1970, 35.5% of the population was urban,

which compares with 44.6% in 1980, 55.2% in 1991 and 61.5% in 1996. A number of

explanations exist for these demographic patterns in the Amazon. The Amazon became a

place of refuge for a segment of southern Brazil's landless population that pursued the

goal of obtaining its own land and building a resource base. The decade of the 1970s was

a period of intense activity in the region. The highway construction projects, the

colonization schemes, and the spontaneous migration into the region were among the

most salient factors that altered the demographic profile of the area (Wood and Wilson









1991). Currently, almost two-thirds of Amazonia's population of almost 20 million

people lives in urban areas (Browder and Godfrey 1997, Friends of the Earth 1998).

In the early 1980s, deforestation in Amazonia and its associated effects on the

livelihoods of local populations became an international issue. In this scenario, the

Brazilian government started to consider development alternatives that use Amazonia's

diverse resources while maintaining forest cover and valuing the needs of traditional

populations. Since the 1980s, recognition by Brazilian and international society of the

importance of maintaining the biological and cultural diversity of Amazon ecosystems

has led to the creation of numerous conservation units, including national forests, national

parks, biological reserves, as well as indigenous areas and extractive reserves.

In this context, extractivism, the ancient practice of extracting economically

valuable non-timber forest products (NTFPs), emerged as a possible means of reconciling

economic development with forest conservation (Nepstad and Schwartzman 1992).

Extractivism captured global attention in the late 1980s when a group of rubber tappers

fought to protect their lands from encroaching cattle ranchers in Acre, Brazil (Allegretti

1989, 1990 and 1994, Schwartzman 1989 and 1991), culminating in the highly publicized

assassination of the rubber tapper leader Francisco "Chico" Mendes in 1988. These

events led to the creation of the first extractive reserves, which are "forest areas inhabited

by extractive populations granted long-term usufruct rights to resources which they

collectively manage" (Schwartzman 1989, MMA/IBAMA/CNPT 1995).

At that time, extractive reserves were considered necessary to act as a counter-

balance to the pressures of deforestation and cattle ranching, fitting the relatively new

demand for "sustainable development" and an opportunity to put into practice an explicit









linkage between conservation and development. By restricting the use of forests to a

defined set of products and users, extractive reserves are considered a major strategy for

promoting forest conservation and local socio-economic development in Amazonia

(Allegretti 1990, Anderson 1989 and 1992).

Compared to other land-use strategies, extractive reserves have been successful in

maintaining relatively low deforestation rates (Mendes et al. 1996, Nepstad et al. 1992).

Nonetheless, extractive reserves are not stand-alone solutions, but they complement other

land use options. In the last decade, the concept of extractive reserves has evolved in the

Brazilian Amazon, and extractive communities in other tropical countries have followed

the example of Acre (Salafsky and Dugelby 1991).


1.1 Problem Statement

The rubber boom at the end of the nineteenth century and the brief period during

World War II indelibly marked the history and cultural characteristics of Acre (Rancy

1992, Tocantins 1979). Since the 1970s, the eastern part of the state has been marked by

clashes between forest dwelling people and settlement projects and cattle farms. Today,

changing land-use and land-cover are the rule in Acre (153,000 km2), the western-most

state in the "Arc of Fire and Deforestation" in Brazilian Amazonia. Acre represents an

interesting "threshold" site to study the dynamics of deforestation in a context where

forest dwelling communities, cattle ranchers and colonists coexist.

The initial assumptions of the extractive reserve concept were that an economy

based on non-timber forest products could increase the economic value of forests and

rural income, offering many subsistence resources for extractivist communities, while

being ecologically sustainable. However, the extractive reserve model is facing









challenges in putting these objectives into practice as economic, social, political and

ecological constraints have emerged (Anderson 1992, Browder 1992a, Homma 1993).

Extractivist communities are faced with a current pattern of low-income generation, and

the goal of economic growth is a major constraint forcing land-use changes in extractive

reserves. Land-use patterns within the extractive reserves are changing from strict non-

timber forest products (NTFP) economies to small-scale cash crop agricultural and

pastoral activities. One of the key questions is how economic development changes land-

use in extractive reserves and how these changes can be managed. Of particular

importance is the management of land-use in a rapidly changing socio-economic context

while protecting forest resources according to the conservation objectives of extractive

reserves. To evaluate the efficiency of extractive reserves as a conservation and

development model in Amazonia, it is necessary to analyze land-use change and

deforestation rates within these areas.

Deforestation rates have been reported for areas along the Amazonian

development frontier with colonization projects and cattle ranches that result in large-

scale deforestation patterns, particularly in the state of Rond6nia and Para (Alves 1999,

Fearnside 2001, Moran and Bondiziol994, Nepstad et al. 1999, Rignot et al. 1997, Skole

and Tucker 1993, Skole et al. 1994, Wood and Skole 1998). Rarely, however, has forest

dwelling smallholder deforestation been examined with the exception of a few studies

(Brondizio et al. 1994 and in press, Peralta and Mather 2000, Sassagawa 1999).

Sassagawa (1999) conducted a remote sensing analysis of the deforestation rates

in the Chico Mendes Extractive Reserve (CMER) in Acre, detecting land-use changes in

the reserve during the period of 1986-1998. The analysis revealed that eight seringais, or









rubber estates, with the highest deforestation rates within the reserve are expected to

exceed the legal limit for deforestation (10% of each residents area) within 2.5 to 6 years.

This thesis uses Sassagawa's work as a basis for the deforestation analysis in the CMER.

However, this study is limited to remote sensing analysis of land-cover changes

and did not incorporate the human socio-economic factors driving land-use change on the

CMER. That is, while this study evaluate land-use changes "top-down" looking at

general patterns of deforestation in the reserve, it does not examine the on the ground

socio-economic processes that lead to the patterns of land-use. In fact, socio-economic

aspects of forest dwelling smallholders are still little analyzed for several parts of the

Amazon. Most studies with this population have been unable to achieve an integrated

perspective between land use cover and socio economic factors inducing land-use

changes. Few studies have examined the way in which patterns of land cover change

relate to underlying social and economic driving forces. The gaps in research in socio-

economic factors on extractive reserves are important aspects to understand the land use

change process that has taken place in these areas since they were established.

The analysis of land-use patterns in the CMER was not necessity at the creation

of the reserve, since the economic production at the time was primarily based on rubber

and Brazil nut extraction from standing forests. In the last ten years, land-use decisions

made by CMER residents have gradually changed toward a mix of extractive and non-

extractive activities, in an attempt to increase household income and also in response to

development incentives from projects and government initiatives. As the pressure to

diversify production and to increase income grows, rubber tappers are increasingly

moving toward land use activities that have more destructive impacts on forest cover.









Therefore, today, cash-crop agriculture, and small-scale cattle ranching are an important

component of modern rubber tapper land-use strategies. These non-extractivist activities

play a dual role as the new economic activities: not only do they increase household

income, but they also increase pressure on natural resources and deforestation rates in the

reserve.


1.2 Central Questions

This thesis addresses three main questions. First, what are the land-use patterns

and changes occurring in the Chico Mendes extractive reserve? Specifically, this question

analyzes the overall land-use patterns of the reserve in order to understand the actual

land-use changes in recent years.

Second, what are the socio-economic forces behind these land-use changes?

Several factors influence land-use change in the reserve, which might include population

increase, development policies for extractivist communities, lack of market for non-

timber forest products and development projects influencing agricultural and pastoral

land-use activities.

Third and finally, what factors are contributing to deforestation in concentrated

areas of the reserve? The land-use patterns in the reserve follow two separate, opposed

tendencies. Some areas of the reserve have continuing and high deforestation rates, while

in other parts very little deforestation has taken place in the last decade. Factors

influencing deforestation in these areas could include population density, road and market

access, distance to neighboring towns, as well as distinct land-use strategies bordering

with the reserve.









1.3 Structure of the Study

Having here explained the importance of studying smallholders' land use patterns

in the Chico Mendes Extractive Reserve, chapter 2 will provide an overview of the

diverse arguments involving the concept of extractive reserves in the Amazon. This

chapter is based on the literature produced in the early debate about the reserve model up

through recent works based more on practical experience of the model. In Chapter 3, I

will describe the procedures used to define the study site, survey setting and planning,

types of primary and secondary data gathered as well as data processing, variable

operationalization and definitions, and analysis.

In chapter 4, I will analyze the context in which deforestation has taken place in

the reserve. I begin with a discussion of deforestation in Acre's context and how it has

affected deforestation in the Reserve. I then take a look at how and why seringal level

deforestation rates differs in the reserve. In chapter 5, I present a household level

approach of measuring some characteristics affecting different land use activities. First I

characterize the major land use activities of the rubber tapper settlements and then relate

those to seasonality of land use activities throughout the years and differentiated family

labor force allocation. Next, I define six major land use outcome variables that define

tappers' production systems. I then, define a group of explanatory household

characteristics and seringal location variables as major determinants of extractivist and

non-extractivist land use activities in the study seringais.

Chapter 6 I present general trends of key socio-economic factors driving land use

changes in the reserve as a whole and in the four specific seringais, based on census data

of the reserve in three different periods. At the reserve level, particular attention is given

to population and settlement dynamics and social organization of the residents. At a






10


seringal level analysis, I explore factors inducing land use change, looking at population

and settlement dynamics, extractivist and agricultural production, pasture expansion, and

social organization. In chapter 7, I summarize some of the mean funding of the study and

discuss some of the challenges that the reserve will face regarding resource regulation

and future economic development of the residents. Finally, I suggest continuing steps to

be taken in this research as well as in other related topics. In this regard, study in land use

change in the reserve still leaves much to be explored.














CHAPTER 2
DEBATE ABOUT EXTRACTIVE RESERVES


2.1 Brief History of the Amazon Rubber Boom

Much of the economic history of Latin America has been characterized by several

extractive boom-and-bust cycles. Most of these cycles have been based on the

exploitation of plants of commercial value. The Amazon rubber boom, which spans the

late nineteenth and early twentieth centuries, was one of Latin America's most extreme

natural resource export booms.

Rubber had been known for a long time before it became an economically

important commodity. The French naturalist Charles Marie De La Condamine, who

traveled down the Amazon from Ecuador in 1736, is considered to be the first person to

publish a study about the properties of rubber, the milky latex of the Hevea brasiliensis

L. tree (Dean 1987). In 1839, Goodyear invented vulcanization, a process that made

rubber durable, turning it into an essential product of the Industrial Age and triggering

inventions that would eventually lead to massive demands.

Events such as the invention of the inflatable rubber tire by Dunlop in 1888, the

bicycle craze in Europe (especially in France) in the late 1800s, and the development of

the automobile industry in the early 1900s explain the world's explosive demand for

rubber towards the end of the century. With the increasing world demand for rubber, the

price of the product grew steadily. Amazonia, the world's sole supplier of rubber,

experienced decades of economic growth in terms of exports. The wealth generated by

the "black gold" rubber had fantastic proportions. Rubber exports rose sharply in

11









quantity, and even more dramatically in value. The rubber trade became a mainstay of the

Brazilian economy, providing at its height almost 40 percent of its export revenues,

nearly equaling coffee in importance (Dean 1987). During these giddy decades, Brazil

monopolized the world rubber trade.

After a few decades, however, the Amazon rubber fever was broken. The

deterioration of Brazil's rubber economy began in 1912 with the collapse of its monopoly

because of competition from Malaysia's rubber plantations. Southeast Asian plantation

rubber trees grew faster and gave higher yields than the wild rubber of the Amazon, due

to the skilled domestication efforts. Consequently, Asian rubber could produce higher

quality rubber more cheaply, quickly, and at vastly greater quantities than its wild

Amazonian rival.

The once wealthy region never fully recovered from the rubber bust, despite

attempts to strengthen Amazonian rubber production. The Brazilian and North American

governments heavily invested in rubber production in the Amazon during the Second

World War, when American access to the Southeast Asian rubber supplies was blocked.

However, after the war and with the development of synthetic rubber substitutes, the

period of high rubber production in the Amazon was once again over. Despite its failure

to establish long-term development of the region, the rubber-boom era indelibly marked

Amazonia and caused substantial social and economic alterations in the region that can

be seen and felt until today.


2.2 Evolution of the Extractivist Development Model in Amazonia

Amazonia has been indelibly marked by the rubber boom and bust (1890-1912)

that brought thousands of people from the arid Northeast of Brazil to the rainforest as

labor for rubber extraction ( Bakx 1988, Rancy 1992). Since that time, rubber tappers









have been a disenfranchised category in Brazilian history. Throughout Amazonia they

have worked in the semi -slavery aviamento system of debt peonage, a type of social

relations that persists in many regions of Amazonia until today (Weinstein 1983).

However, the wealth generated for the elites by the "black gold" had fantastic

proportions, bringing several decades of unparalleled prosperity and growth to the region,

integrating Amazonia to the world economy, establishing extensive trading networks,

constructing massive port facilities, and leaving behind large urban centers along

Amazonia's rivers (Weinstein 1983).

With the end of the rubber boom, forest extractivism was generally viewed as a

backward and peripheral activity of economic stagnation, and forest resources and

traditional peoples considered as development obstacles in Amazonia (Barbosa 2000).

This attitude characterized Brazilian government policies for many decades up through

the 1980s and brought development projects to the region that included massive highway

construction (e.g., Transamaz6nica, Belem-Brasilia), dam-building (e.g., Balbina,

Tucurui), mineral extraction (e.g., Grande Carajas), agricultural colonization and cattle

ranching (e.g., Polonoroeste), as well as incentives to timber industries (Hecht and

Cockburn 1989, Moran 1983, Oliveira 1991, Schmink and Wood 1992). These projects

brought serious social and environmental problems to rural and urban Amazonia, with

deforestation reaching unprecedented rates and rural violence against traditional

communities rising to alarming numbers.

As a response to the rubber boom legacies and the oppressive development

projects, recurrent peasant resistance erupted in Amazonia and led to violent clashes

along the development frontier. The rubber tapper movement emerged in the state of

Acre in the late 1970's. Chico Mendes was the major force behind the movement.









Through collective action, Mendes and the tappers fought for the right over the land and

not to leave it for the ranchers whose ownership rested more on coercion than on legality.

Slowly, they began to organize against the major cases of deforestation that occurred in

the area by the means of empates, or standoffs, blocking people and machinery with their

bodies (Calaca 1993). The rubber tapper movement only received political clout after the

creation of the National Council of Rubber Tappers in 1985, when it was decided that the

primary demand would be agrarian reform in the form of extractive reserves (Allegretti

1994).

In the late 1980s the first powerful alliances between rubber tapper, indigenous

organizations (e.g., UNI-Union of Indian Nations) and the environmental movement

emerged. Common goals such as indigenous land demarcations, creation of extractive

reserves and opposition to mega-projects, shaped joint strategies of civil disobedience

and protests2 (Albert 1992, Fisher 1994, Hall 1989). The onset of democracy in Brazil

also politicized the national environmental and grassroots organizations and boosted the

emerging NGO movement that embraced polemic rubber tappers and indigenous causes

(Barbosa 2000, Viola 1988).

The Amazonian grassroots efforts also gained considerable strength when

international interests in Amazonia reached high levels. The internationalization of

Amazonian ecopolitics occurred twofold. First, beginning in the mid-1980s, awareness of

major global environmental problems such as the greenhouse effect, the ozone hole, and

biodiversity loss, sparked an international debate on the state of the global environment

(Stern et al. 1992). Overnight, Brazil was in the spotlight, pressured by environmentalists,



2 In particular the Kayap6 became widely known for civil disobedience and protests in Altamira in 1989 to
protest the construction of several dams along the Xingu river (Fisher 1994).









international organizations, and first-world politicians to stop the devastation taking place

in the Amazon rainforest (Barbosa 1993 and 2000). Second, international development

schemes were heavily attacked, resulting in international publicity around target dam and

road construction projects that have caused international financial supporters to back out.3

At first, the international attack on Brazilian development strategies provoked

strong protest against this "environmental imperialism."4 However, President Collor's

administration (1990 1992) changed toward a "greener" rhetoric. In the pre-UNCED

(1992) arena the Brazilian government took visible steps to demonstrate to the

developing world and to donor countries serious efforts to protect Amazonian forests and

its people. Part of this was the creation of extractive reserves (ERs), national parks and

the demarcation of indigenous territories.

This fight for ERs lasted from 1985 tol990 and encountered many difficulties.

The very idea of extractive reserves was thought incongruent with the legislation of the

time, since this defined property in individual terms rather than collective. Decree No.

98.897/90 was the compromise (Murrieta and Rueda 1995). It defined these areas as

"property of the union or federal government" destined to be used for extractive activities

(Allegretti 1994). The creation of ERs represented a novel government response to the

pressure both of social movements and conservationists, and a form of legally





3 For example, as a result of rubber tapper lobbying the Inter-American Development Bank withdrew its
financial support for the road construction of BR 364. It was a first in the bank's history that it stopped a
loan on the basis of environmental concerns. Another example is the withdrawal of the World Bank from
the Xingu river dam projects in response to the indigenous Kayap6 protests (Barbosa 2000).
4 In the Sarney administration (1985-1990) the development of Amazonia became a question of
nationalism, of national security, and of sovereignty over encroachable lands. The demarcation of
indigenous territories and protected areas was seen as a conspiracy of developed countries, to hamper
competitive development in Brazil and to steal land for global (not national) carbon and biodiversity stocks
(Barbosa 2000).









recognizing land tenure regimes that do not comply with the private property paradigm

that characterizes most modem economic systems.5

In 1990, the federal ERs existed on paper, but it took several years for ERs to take

shape.6 Only by 1994 were the four federal Amazonian ERs legally and institutionally

consolidated, covering an area of over 2 million hectares (ELI 1995, Murrieta and Rueda

1995). Apart from land tenure, resource utilization regulations at the community level

were a principal concern in the early stages. The newly created Traditional People

Department (CNPT) of the Brazilian Environmental Protection Agency (IBAMA) was

made the responsible government institution for ERs. This step was followed by the

development of utilization plans, jointly created in by rubber tapper associations and

government representatives, where federal environmental legislation and traditional

rubber tapper resource utilization were reconciled (MMA/IBAMA/CNPT 1995).


2.3 Analysis of the Extractive Reserve Literature

Despite its charisma, the basis of extractive economies is indisputably fragile and

the subject of an active and sometimes polemic debate among scholars and conservation

and development professionals of diverse academic backgrounds (e.g., Anderson 1992,

Fearnside 1989 and 1992, Browder 1992b, Homma 1989 and 1993). Publications on

ERs strategies peaked in 1992; however, since then very few works have appeared

evaluating the ER debate and experience. Consequently, there is a need for a survey of

the discussion. In the next section I will focus on the Extractive Reserve experience in the


5 ERs were created by Brazilian Ministry of Agrarian Reform and Development (INCRA), also responsible
for extensive colonization projects in frontier Amazonia. Approval of this form of land tenure signified a
radical departure from the way in which regional development in Amazonia had been carried out in the past
(Allegretti 1990).
6 The first federal extractive reserve created was the Upper Jurua River in 1989 (Acre), followed by the
Chico Mendes (Acre), Cajari River (Amapa) and Ouro Preto Riber (Rond6nia).









Brazilian Amazon region in order to analyze the leading discussions about ERs,

discerning the major trends of debate; and draw conclusions from current ER experiences

about the prospects for ER development in Brazilian Amazonia.

2.3.1 General Patterns and Trends in the Debate

When extractive reserves were first proposed by the grassroots National Rubber

Tapper Council, the cause was picked up and supported by scientists, activists, and

conservation and development professionals and was considerably transformed. The late

1980s and early 1990s were characterized by a relatively "fresh" enthusiasm for

environmental and social causes in the tropics (compared to the learned skepticism of

today), and the debates around ERs developed into a curious mix of science, ideology and

lobby. The somewhat emotional content of the debate was further enhanced by the uproar

over the assassination of Chico Mendes and the violent suppression of indigenous

peoples, transforming them into international environmental heroes. At the time, ERs

were optimistically promoted as the panacea for most of Amazonia's environmental

problems.

Three intermeshed lines of discourse influenced, shaped and transformed the

discussion about ERs. First, the international lobby saw ERs as an answer to the urgent

demand for rainforest conservation as a reaction to high deforestation rates and the

concern about global climate change and biodiversity loss. Second, the promotion of

traditional resources management as an example of sustainable livelihoods boosted the

appeal of ERs as an example of "inherently environmental" rainforest people as agents of

forest conservation. Third, the proposal of ERs coincided with the promotion of non

timber forest products use and marketing as a promising economic alternative for

(mostly) biodiversity rich tropical regions. The different demands of these interest groups









caused an inflation of the expectations on ERs. From a human rights and territorial issue

that embraced forest conservation as a means of continued livelihood in the forest, the

discussion spawned high economic expectations.

The main participant in the beginning debate about ERs was first a Brazilian,

anthropologist Mary Allegretti (1989 and 1990) who publicized the rubber tapper's cause

and later moved into important government positions. The discussion was later joined by

mostly North American conservationists such as Anderson (1989, 1992 and 1994) and

Fearnside (1989 and 1992), followed by development specialists like Schwartzman (1989

and 1992); regional planners such as Browder (1990a,b) as well as the Brazilian

economists Alfredo Homma (1989, 1992 and 1993). From the onset it was an

interdisciplinary debate, bringing formerly isolated groups together. However, no

consensus existed in the vision and objectives about the nascent ERs model. The

strongest proponents of the model were Allegretti (1989 and 1990) and Schwartzman

(1989 and 1992), supported by Anderson and Fearnside. The strongest critiques of ERs

were produced by Browder (1990a,b), and Homma (1989 and 1993).

The ER literature has been produced in a characteristic boom and bust pattern. In

1989, the year after Chico Mendes's assassination, articles mushroomed about the topic

both from promoters (Allegretti 1989, Fearnside 1989, Schwartzman 1989, Hecht 1989)

and skeptics (Homma 1989, Browder 1990a), in addition to the popular media coverage.

From 1990-1993 a many articles were written about the topic, culminating in some edited

books that discussed the non-timber forest products strategy and ERs (e.g., Nepstad and

Schwartzman 1992, Anderson 1990, Redford and Padoch 1992, Plotkin and Famolare

1992, Homma 1993). However, after 1993, hardly anything was written about ERs as a









development and conservation model. After that, most production has gone into gray

literature reports and government policy statements.

Notable exceptions include a Brazilian edited book that refutes some of the

criticism of ERs and a summary of the ER status by the government staff (Murrieta and

Rueda 1995). A paper written by visionary Acre Production Secretary Jose do Rego

points out that the new ideas of extractivist development in Amazonia, "neo-

extractivism", incorporate technical improvements in cultivation, harvesting and

processing, allied with a specific social environment of forest cultures people whose

lifestyle is intimately tied to the forest (Rego 1999). Also, two other papers about

extractive reserves published recently discuss the institutional architecture of extractive

reserve, and the political and economic empowerment of rubber tappers communities in

the State of Rond6nia (Brown and Rosendo 2000a,b).

Another characteristic of the ER debate is that most of the critical body of

literature was written before ERs were legally and institutionally functioning on the

ground (as they were to some extent after 1994, see Murrieta and Rueda 1995). This

temporal gap between intellectual battle about ERs and de facto experiences of ERs

suggest that the debate about ERs was a very theoretical and speculative one. This might

also explain some of the arrested analysis of the model, as scholars are "waiting" for ERs

to develop on the ground. Several large national and international projects were

established within ERs, in great part due to the many publications and the publicity about

ERs in the early stages. To understand what happened to the ER debate, one has to

examine the role of different arguments in the ER literature.









2.3.2 Social Development Arguments

The rubber tappers demand for ERs was primarily a human rights battle linked to

an environmental one (Schwartzman 1989). In extractivist communities, the needs of

traditional lifestyles translate directly into territorial rights and the need of relatively large

expanses of forests as the basis of their livelihoods. Even so, rubber tappers were also

fighting to obtain modern production technology for diversification of their production

and the creation of education and health systems adapted to the needs of the community

(Allegretti 1995) which were, and still are, in great need. The establishment of social

institutions, such as cooperatives and associations, and physical infrastructure, such as

transport systems and roads, have also been determined to be crucial for the success of

extractivist populations to sell their products (Schwartzman 1989).

The importance of social and political realities within the extractivist region has

repeatedly been pointed out as a critical factor that can define the success or failure of ER

endeavors. The behavior and incomes of rural people who extract forest products are

often determined by social and economic factors over which they have little or no

immediate control (Browder 1992a).7 An additional concern about the ER model has

been that it requires low population densities and therefore large forest expanses for few

people (Fearnside 1989, Browder 1990a).8 It has been argued that it is hard for land

planners to justify ERs that require tens of hundreds of ha of forest to support one






7 For instance regional power relations and national development policies can join forces to support or to
undermine ERs. In that context, it is very telling that ERs were only implemented in Amazonian states
without an overwhelmingly strong large-scale landowner (latifundia) alliance (i.e., Acre and Amapa in
contrast to Para and Amazonas).
8 Fearnside (1989) estimated that Brazil nut harvesters need 300-500ha per family, resulting in about 1-1.7
people/km, concluding that the carrying capacity for extractivist economies is low.









household in countries that have extremely high population densities.9

2.3.3 Economic Arguments

The most advertised promises and the most serious constraints of the ER reserve

model have been identified as economic ones. The conservation community was excited

about ERs as a model that could preserve biodiversity while simultaneously providing

sustainable economic return to local peoples and governments (Allegretti 1990,

Schwartzman 1989). This excitement was fueled by an article by Peters et al. (Peters et

al. 1989) proposing that long-term financial return from the harvest of non-timber forest

products found in a hectare of Amazonian rain forest far outweighed the net benefits of

timber production or agricultural conversion from the same area of land (see also:

Anderson and Jardim 1989, Anderson and loris 1992, Gomez-Pompa and Kaus 1990,

Panayotou and Ashton 1992). Several authors, including Peters et al themselves, have

attempted to temper this enthusiasm for income generation by pointing out that

hypothetical calculations of the income streams to be derived from an average hectare of

tropical forest have significant limitations (Fearnside 1989, Pinedo-Vasquez et al. 1990,

Godoy and Bawa, 1993, Perez and Byron, 1999).

The fragility of extractive economies was spelled out from the onset (Anderson

1989, Fearnside 1989), as few Amazonian extractive products have augmented in

production and value in the last decades. This certainly was a very fragile base for a

development strategy that has as a goal the well-being of extractivist populations. The

premise that NTFP extraction can produce greater financial return compared to timber or

agriculture also has been challenged by many authors (e.g., Browder 1990a,b and 1992a,


9 However, although Amazonia's population is rising and currently reaching 20 million, rural population is
decreasing. In addition, many other areas of Amazonia have extremely low population density (Martine and
Camargo 1998).









Homma 1992 and 1993). Particularly, Homma attacked extractive economies as

inherently uneconomic, arguing that all extractive products would eventually be

substituted by domesticated plantation production or synthetic replacement of the

extractive good. 10 This was the case for latex, which was the main product of extraction

in Amazonia and is now cultivated in plantation within and outside Brazil as well as

produced synthetically. He argued that if the prices of extracted products fall,

degradation is likely to follow because extractors must harvest the resources above

sustainable thresholds in order to maintain their living standards. He also argued that

investments should be made at the development frontier, where colonists and cattle

ranchers pursued more viable economies.

However, this position has been challenged by Allegretti (Allegretti 1994 and

1995) who points out that these land-uses are only "viable" due to the massive incentives

and tax benefits that the Brazilian government has used to encourage cattle ranching and

colonist agriculture on the Amazonian development frontier. She argues that similar tax

and credit incentives need to be established for extractive economies. Also, the inherent

assumption within these critiques, that ERs have to compete with other land uses in the

market economy, might be quite reductionist, as they ignore the non-market value of

these areas, such as the forest conservation, subsistence value and the opportunity value

of untapped resources (Allegretti 1994, Anderson and loris 1992, Nepstad and

Schwartzman 1992).




10 Homma delineates three phases in the extractivism of a forest product: (1) expansion; (2) stagnation and
(3) decline. The decline is attributed to four main factors: (1) inelastic market for the forest products; (2)
extraction levels which surpass rates of natural regeneration; (3) domestication of the product; and (4)
substitution of the forest product for a similar natural or synthetic product. Extractivism is just a phase in
the economic development of a region, and if a forest product is that important, it will either be
domesticated (generally somewhere else) or be substituted by a cheaper synthetic product.









2.3.4 Ecological Arguments

ERs were championed for two main ecological reasons. First, ERs were proposed

as a means of preserving tropical forest cover. This has been confirmed in comparison to

other land-uses strategies (Mendes et al. 1996). Secondly, the early NTFP literature

assumed that the harvesting ofNTFPs occurrs on a sustainable basis since local people

have harvested these products for hundreds of years. Particularly anthropologists (e.g.,

Posey and Balee 1989, Allegretti 1979) had the tendency to think so. However, studies

have shown that commercialization often leads to overexploitation (Bodmer at al. 1997).

In that respect very little has been published about products from ERs in Amazonia (e.g.,

Pinard 1991, Kainer 1995). ERs have also been criticized as areas of "open access" where

extractivists were encouraged to respond to "perverse incentives of overexploitation" (

Homma 1989, Salafsky and Dugelby 1991). This notion has been refuted by May (1989),

and the development of Utilization Plans of the reserves demonstrates the conscious

attempt at resource use regulation.

A final argument in favor of the ERs strategy has been that it supports a

traditional lifestyle, increasing the subsistence value of forest resources, preserving

knowledge and thus preserving biological diversity (Balick and Cox 1996). Linked to the

discourse of preservation of traditional knowledge is the call for the protection of the

opportunity value of protected forests within ERs as repositories of biological and genetic

diversity and therefore potential sources or new products for agricultural and

pharmaceutical industries (Balick and Mendelsohn 1995, Gottlieb and Kaplan 1990).

However, this argument has proven problematic, as it fueled economic expectations that

have not been corroborated, and triggered a highly polemic debate about access to genetic

and biological diversity in Amaz6nia.









2.4 Achievements and Problems with Extractive Reseves

The discussion about ERs has been overly focused on their economic fragility and

neglected more basic community issues that are important for the social success of the

reserve and also for their eventual economic success. A number of important

achievements were made. Today, rubber tappers hold communal territorial rights within

federal government protected areas that explicitly state their use and management by

traditional residents. A great advance is also the elaboration of Utilization Plans and an

ecological monitoring program, although evaluations of ecological practices has not

occurred to a great extent (Millikan and Irving 1997).

Another important achievement is, perhaps, the building of social institutions such

as associations, cooperatives, affiliations with the rural workers' union and the

fortification of the Rubber Tapper National Council (Ramalho 1992, Schmink 1992).

Nonetheless, a continuous effort will be required in order to maintain internal cohesion

and common goals within these organizations, and to acquire the administrative and

organizational skills necessary to run their commercial enterprises. In terms of formal

education and health services for residents, ERs are still very fragile, as access to many

areas is difficult and cooperation with state and municipal agencies has been limited and

fractious.

In terms of development support, Extractive Reserves have been supplied with

international funds. Most significantly, they have been inserted into the international

Pilot Program to Conserve the Brazilian Rain Forest (PPG-7) since 1995.11 The focus of



1 The PPG-7 is one of the largest conservation programs ever experienced in Amazonia. It is the joint
effort on the part of the Group of Seven countries, the Government of Brazil and the European Union, with
the World Bank as administrator of the financial resources. The budget of international and national
support to the four federal extractive reserves in the last four years were approximately ten million dollars
(CNPT 1999).









this mega-project has been 1) The consolidation of the legal status and the regulations

within ERs, 2) Social organization, community participation, capacity building and

education and health programs, and 3) Extractivist production organization,

diversification, transportation and marketing (IBAMA/CNPT 1999).

Moreover, the Brazilian government has offered credit lines to rubber tappers

settled in extractive reserves, an innovative governmental policy, as until recently credit

was only accessible for agricultural and cattle production in Amaz6nia. The Program of

Support to Extractive Development (PRODEX)'s goal is to support the development of

extractives activities, providing technologies and improvement in production to the

communities, and incentives for diversification of production.

However, there are many obstacles that prevent it from having a significant

impact on the improvement of extractive production. First, the financial resources are

administered by BASA, the Amazon Bank, which has little operational capacity and is

bureaucratically too rigid to work with extractives communities, which often cannot

provide the required documentation. Another weakness of the PRODEX is the quality of

the technical assistance, as government technicians have little experience in working with

extractives communities. Nonetheless, the program demonstrated the federal

government's recognition of the needs of extractive communities.

Political changes at a federal and state level have contributed to the ER model in

Amaz6nia. At the federal level, a team of top professionals with great commitment to

Amazonian environmentalism was selected by the Brazilian minister of environment for

cutting edge decision-making, such as Mary Alegretti, who greatly contributed to the

definition of extractive reserve in the 1980s. At the state level, in the Amazonian states of

Amapa and Acre, leftist-environmentalist governors have been elected (Capiberibe in









Amapa and Jorge Viana in Acre). In Acre where the struggle began, the government has

explicitly named itself "The Government of the Forest" and sustainable development on

the basis of forest resources is at the center of its mission (Schmink 1999). The best

possible political conditions currently exist in these regions. The next years will show

whether the structural and policy changes that are occurring can considerably improve the

prospects of ERs.

The biggest difficulty at this point challenging ERs is income-generating activities

(Mendes et al. 1996, Millikan and Irving 1997). An important, although late recognition

has been that traditional extractive populations do not need to rely solely on extractivist

production. Others land-uses activities, such as agriculture, agroforestry systems, and

small-scale cattle ranching are also important (Anderson 1992 and 1994, Rego 1999). In

an ideal case these other land-use practices would intensify production per area without

sacrificing ecological sustainability. However, as the pressure to diversify production and

increase income builds up, extractivists tend to move increasingly toward more

unsustainable land use activities and thus increase deforestation in the reserves.

Deforestation rate is a major indicator to evaluate the efficiency of ERs and to fulfill its

conservation objectives. Chapter 4, addresses land-use changes in the state of Acre and

the context of deforestation rates and trends in the Chico Mendes reserve.














CHAPTER 3
STUDY SITES AND RESEARCH METHODOLOGY


3.1 Study site selection

The field research for this project was carried out from May to August 2000 in the

Chico Mendes Extractive Reserve, in Acre, Brazil. The reserve is comprised of 970,570

ha (9,705 square kilometers), and is divided into the historically important rubber tapper

estates (seringais). For this study, I selected the four seringais with the highest

deforestation rates based on the work of Sassagawa (1999). In addition, I selected these

seringais according to their location and their extension across different municipalities

within the reserve. The four seringais selected were seringal Filipinas (in the

municipality of Xapuri), seringais Porongaba and Humaita (in Brasileia) and the seringal

Paraguacu (in Assis Brasil). These seringais are relatively near to the closest towns and

can be accessed following the highway that leads from the capital Rio Branco to the

Brazilian border with Peru. Seringal Filipinas is closer to Xapuri. The seringais

Porongaba and Humaita are located three hours from the closest town, Brasileia. Seringal

Paraguacu is located three hours from the municipal seat of Assis Brasil. The study

seringais differ in terms of area, population size and some economic activities practiced

by the residents

Figures 3-1 and 3-2 are maps of the Chico Mendes Extractive Reserve showing

the four seringais where fieldwork was carried out, including the households where one-

time interviews were conducted. Overall, the total number of households interviewed in





































SI I .EMi fc lB
--.. -- .. -




ri
I "-




iL', I = I
r t. -


-----------






Figure 3-1 Maps of the study site showing the seringais and settlements in Xapuri and
Brasileia area









































-? *rhI;




8 S gal P l 9 r (






*'.,,h ,, .,







** 4 *".4.


i c 1 D QO


hjwr1ion
7ii IS wc,-
WJvffhd lhlr.i b

Zate N


2dractw' Re-rve Cckc Mendrs



(5


Figure 3-2 Maps of the study site showing the seringais and settlements in Assis Brasil

area


(Nig MwdP~
MOY aSit
QnbPV7viK
R4#jtrmrd


n
aed

E*









the four seringais was 66. In the seringal Filipinas, which is the largest, nineteen

interviews (28.8% of the total interviews) were carried out representing approximately

40% of the households in the seringal. In the seringal Porongaba, almost all the

household heads were interviewed, totalling sixteen households (24.2 % of the total

interviews). In the seringal Humaita, a sample of eleven household heads was

interviewed (16.7 % of the total interviews). In the seringal Paraguacu, twenty interviews

were carried out (30.3 % of the total interviews), representing approximately 70% of the

total households in the seringal. It is important to stress that the interviews were mostly

conducted with the male heads of household; however, in three cases women were the

head of household. Interviews were limited to household-heads as these actors make the

land-use decisions at a household level.

The rubber tappers' settlements in the seringais are located 30-120 minutes

walking distance from each other. The most distant households surveyed were located at

about a four hours walking distance, approximately 20 kilometers from the first

household surveyed in the entrance of the seringal. Households were selected based on

their location in the seringal, so that the sample would be evenly distributed throughout

the area.


3.2 Survey setting and planning

In previous work I have gained knowledge of the rubber tappers in the Chico

Mendes Extractive Reserve, which helped me set up the research in conjunction with the

rubber tapper community. Before visiting the households, the research was discussed in

meetings with leaders of rubber tapper social organizations in the three municipalities. In

these meetings, I explained the goal of the research as well as the methodology that I was









going to use. These discussions provided important feedback that helped me to make

some changes in the research goal and methodology. The association leaders also helped

me by announcing to the communities that I was carrying out the research, through local

radio stations and the associations' amateur radios. This greatly facilitated the research,

as the tappers already had an idea of the purpose of my study when I arrived at their

homes, affording more time for interviews and participant observation. In addition, the

associations indicated a community representative in each seringal to guide myself and a

research assistant in order to safe find our way through forest trails, to select particular

households, and to provide us with invaluable historical background of the community.


3.3 Survey methodology

Most of the data for this research was collected through a survey. The survey was

conducted orally and carried out using structured interviews. Although the survey was

applied at the household level, it targeted land-use information at the household,

community, and reserve levels. The questionnaire covered a broad range of questions

regarding land-use patterns, income, extractivist and agricultural production, migration

dynamics, demography, transportation, social organization, and reserve wide-

management efforts. I and a research assistant, an agronomy student from the Federal

University of Acre who also was a certified extension worker applied the survey. The

survey interviews lasted about 2-3 hours in each household. Between seven and ten days

were spent surveying each seringal.


3.4 Other sources of data

The Brazilian Environmental Agency (IBAMA) conducted a census with all the

residents of the reserve in 1995, 1998, and 2000. The data was collected as part of the









monitoring strategies of the residents' land-use activities in the reserve. No one had

analyzed the information thus far, although the censuses yield very valuable information

regarding temporal changes in land-use, population and productivity. IBAMA collected

the census data of the four study seringais in a questionnaire format, which was applied

orally to the rubber tappers by extensionists and trained residents.


3.5 Data processing variables and analysis

The data collected throughout the survey were first entered into an Access

database. Then, the data were entered into an excel database and a codebook was created

to facilitate statistical analysis with SPSS program. In the codebook, 128 variables were

defined. For the purpose of this Master's thesis, after running several tests, I selected

variables that were related strictly to land-use patterns. The other variables will be used

for further studies on the Chico Mendes Reserve.

To accomplish my analysis, I performed frequency distribution tests, cross-

tabulation tests, and regression analyses. The frequency distribution and cross tabulation

tests considered variables regarding length of residence and previous residence of the

household-heads, settlement sub-divisions and the reasons for these sub-divisions, as well

as residents' membership in rubber tapper social organizations. These analyses are

presented in chapter 5 together with the analysis of census data of the reserve.

In addition, I performed regression tests and created an OLS model. I have

defined six dependent and six independent variables. The dependent variables were

divided into two categories: traditional and non-traditional land use activities among

tappers.









- Traditional

* Rubber production in 1999

* Brazil nuts production in 1999

- Non-traditional

* Area deforested. Area deforested here is defined as the permanent or temporary clearing

of the forest for pasture and agriculture. I combined two variables (household pasture size

and agricultural plot size) into a dependent variable called "area deforested". For a full

picture of deforestation at the household level, other land covers need to be considered.

Fallow areas in different stages of regeneration constitute an important component of

changes in forest cover. This is problematic because I do not have data differentiating 3

year old from 20-year-old fallow.

* Total size of swidden plots in hectares per settlement

* Total size of pasture in hectares per settlement

* Total number of head of cattle per settlement

The independent variables selected point to some of the key socio-economic

determinants of deforestation in the reserve and were considered as background of the

household head, family labor force availability, and location of the seringais. The

following independent variables were used in the analysis:

* Migrant status. The definition of who is a migrant is not straightforward and many

factors can be considered to classify migrants within the reserve. For the purposes of this

analysis, I classified as migrants anyone who came from outside the reserve or seringal

context, be that former city-dwellers or residents that came from another land-use areas

such as colonization projects or farms. Therefore, if the previous residency was outside a









seringal context, the household head is considered as a migrant. In classifying migrants

in the reserve as above, I am aware that I am underestimating several other forms of

classification of migrants. The classification considered here is an attempt to identify

residents that have engaged in different land use practices than those traditionally

practiced by rubber tappers.

* Age of the household head.

* Number of sons and daughters of the household head. For this analysis, only children

who were fifteen years old or older were considered, as this is the age when they start to

play an important role in the family labor force availability in the settlement.

* Seringal location. I divided the four study seringais in different categories. Seringal

Filipinas was considered the reference category for this analysis: as it most typically

represented traditional seringal context. Seringais Humaita and Porongaba were analyzed

as one unit since I observed that they have the same land use characteristics, thus not

affecting the analysis, and also because they border on each other. The outcome and

independent variables will be further discussed in chapter 5, where they are analyzed.


3.6 Environmental characterization of the reserve

3.6.1 Climate

The region's climate can be classified into tropical humid AM, according to

Koeppen's system, and is characterized by a mean temperature between 26 e 270 C, with

a short dry season a high precipitation levels. The rainy season lasts from November until

April, during which precipitation averages amount to more than 110 mm/month, while

during the rainy season precipitation averages are less than 93.3 mm/month, July being

the driest month of the year (RADAMBRASIL 1976). The highest precipitation can be









observed between the months of December to March, a period referred to as "winter"

(inverno) in the region, while the period between June and September is referred to as

"summer" (verdo).

The average monthly temperature generally varies little between 24 a 26C, with a

slight decrease between the months of June and August. During this period a

phenomenon regionally referred to asfriagem, results in acute drops of temperature to 4

to 60C, caused by an advancement of a polar front from the south (IDEAS 1993)

3.6.2 Watershed/Hydrology

The CMER's watershed is represented by two main rivers: the Acre River and the

laco River, both tributaries of the Purus River (Fig. 3-3). The Acre river watershed, cuts

through a large part of the reserve territory and runs from west to east. The meandering

river touches the reserve near Assis Brasil and runs along the frontier between Brazil and

Peru as well as Bolivia. The river crosses the reserve near a community called Itu. The

upper Acre river basin is characterized by a strong dissection of relief, forming hills

located on the most elevated parts with mild slopes from 3 to 8% covered by podz6licos

red-yellow soils (RADAMBRASIL 1976). The Acre River has a narrow floodplain and

the peaks of the flood are observed between February and April, while July to September

are characterized by a hydric deficit.

3.6.3 Geology

The geological studies of the region resulting in known descriptions of

afloramentos litol6gicos (rocky outcrops), as well as mineral and fossil occurrences, are

always based on finds on the banks of the main rivers due to the access by boat used by

researchers. The CMER is totally composed of cenozoic sediments of the Solim6es









Formation, lying on a base of crystalline "Craton Guapore", for which so far no outcrop

has been detected in the reserve area. The recent alluvial with origin in the evolution of


Figure 3-3 Map of the mean rivers bordering the CMER. Source: CNPT (1997)



the rivers of the actual watershed, include tertiary and quaternary sediments (IDEAS

1993).

The Solim6es Formation is a result of the decomposition of"plio-pleistocenica",

which is characterized by variations of"faciol6gicas proeminentes". These are typical

sediments of the floodplain geology with characteristic sedimentary structures.

These are part of the following geological composition: solid or stratified argillites

(clay), with carboniferous and gypseum "concrec6es"; calcite and gypsies veins,

occasionally with carbonized plant material; varied argillites with laminated structure

(CNPT 1998).


0 70km









3.6.4 Pedology (Soils)

Maps of the Radambrasil Project (1976), indicate that the reserve territory lies

basically on eutrophic soils, which suggest good fertility due to the soil type change

capacity, characterized by holding more than 50% of exchangeable bases (IDEAS 1997).

Approximately 55% of the soils have clay texture and 83% have high clay activity. Such

conditions are evidence that the reserve's environment has special characteristics with

ample potential for production. Among the municipalities that hold territory on the

reserve, the municipalities of Sena Madureira and Assis Brasil have more than 90% of its

soils constituted of red-yellow podzolic eutrophic soils. In some other areas on the

extreme north of the reserve, modifications of the soil profile have been detected with

laminar and generalized erosion of the superficial horizon of the soils. Moreover, on

other areas of the reserve, a process of slow "rastreamento" of the soil and localized

laminar erosions has been identified.

3.6.5 Vegetation

The first inventory and description of the vegetation type of the state of Acre were

elaborated by the Radambrasil Project (1976). In the area covering the CMER two

phytoecological units were identified: the system of Dense (Closed) Tropical Forest and

the system of Open Tropical Forest (IDEAS 1993). The reserve holds only 27% of dense

forests, while 73% are areas covered with open forest. Figure 3-4 presents these forest

distributions in the reserve.

The dense tropical forest is characterized by an understory with a dense layer of

shrubs and treelets, and by closed arboreal canopy vegetation. The biggest extension of

continuous dense forest is located south of the Xapuri River. Dense forest is classified

into two forest types, which are differentiated according to different geological and



























1Legend ____ ____

Flofesta OMlbrWila Aberta cOi Pm Pa eiras M cp6 0 30 0 KM
I FloTrsta Outnbr6til Dens corn Pakruairaf
17 Flofest OmWi-tile Aberta corn lBambu
SFloresta Ombrbfila Densa
*i Floresta Ombr6fi Aberta corn Palmeras


Figure 3-4 Map of forest typology occurring in the CMER, according to
RADAMBRASIL (1976). Source: Sassagawa (1999)



geomorphologic properties of the sites, such as: dense forest on terras baixas (lowlands)

with undulating relief and dense forests on dissected relief.

Dense terra baixa forests on undulated relief (Fdo) overlay tertiary sediments and

usually reach a canopy height over 30m. In the lower areas, bordering igarapes (creeks),

occur seringueiras, rubber trees (Hevea brasiliensis L.), the palms acai (Euterpe

precatoria Mart.), jaci (Attalea butyracea), and murumuru (Astrocaryum murumuru). In

this forest type also occur stretches of lowland forest on dissected relief, in which the

typical species is the cerejeira (Amburana cearensis) that has a high commercial value

and is locally abundant.









In the dense forests on dissected reliefs on ridges and hills (Fda), compositions

with smaller trees are found with frequency, in which individual trees are relatively

uniform in terms of their height within their forest layer and in which the canopy reaches

approximately 30m. The understory in the plateau areas is more open than those situated

on the dissected relief and is increasingly dense in the direction of the "talvegues", where

shrub and palm species can be found in higher abundance (IDEAS).

These dense forests occur in three areas: In the northern part of the CVMER near

the headwaters of the creeks (igarapts) Mapinguari,; in the western part, at the

headwaters of the Xapuri river, and in the southern part following the course of the

Xapuri river until it reaches the Acre river. The main species found in these areas are:

aquariquara (Geissospermum spp.), abiorana (Pouteria spp.), angelim (various Legumes),

breu (Tetragastris and Protium), cariperana, castanheira (Bertholettia), cedrorana,

copaiba (Copaifera spp), louro (Aniba, Ocotea), parapara, pau-mulato (Calycophyllum

spruceanum) e ucuuba (Virola spp.).

Open Forest is characterized by a distribution over lowland covering sandy tertiary

geology and in the alluvial terrasses of quaternary sediments as well as the submontane

relatively high areas on pre-cambian ground material. In this forest typology three main

types of evergreen open forest types with palms and lianas occur.

The first evergreen open forest types with palms and lianas (Fac) is the forest on

alluvial terraces on quarternary geology, covering 7% of the reserve. It occurs on the

margin of all the major water flows in the region, however in variable "width" along the

banks of the rivers. The forests in that environment can have very large trees. The most

characteristic species of this environment are the abiorana seca (Pouteria sp.),









aquariquara (Geissospermum sp.), andirobarana, mamorana, munguba, seringueira and

ucuuba. Common palms also are: aCai, bacaba (Oenocarpus mapora) and graminoid

species such as the bamboo (Guadua spp.).

The second Open Evergreen Foi et i iith Palms and Lianas (Faa), is the forest of

lowland and dissected relief. It occupies 22% of the reserve area and is located in the

margin/edges of residual interfluvials. This forest is encountered throughout almost all of

the western part of the reserve, limited by a band of Open Forest on Alluvial Terrases and

the laco River. Its understory is rich in small palms. In dryer areas and on the margins of

residual interfluvials lianas dominate what is called liana forests. In the depression and

valleys there is a higher abundance of palms. Among the most characteristic species are

the palms aCai, inaja (Attalea maripa), jarina (Phytelephas macrocarpa), mumbaca,

murumuru, pataua (Oenocarpus bataua), paxiubdo (Iriartea deltoidea) and paxiubinha

(Socratea exorrhiza). Others commons species also include acacu (Hura crepitans),

amareldo (Aspidosperma parviflorum), fava de espinho, ingas (Inga spp.), matamatas

(Eschweilera spp), muratinga, seringueira, tauari (Couratari sp.) and ucuuba (Virola sp.).

The third Open Evergreen Foi e\t i i/th Bamboo on the lowland areas with

undulated relief (Fao), covers 44% of the reserve, situated mostly in the area of lowland

plateaus and variable drainage. This forest is by far more common than the other types

covering most of the central-eastern part of the reserve. Three forest communities are part

of this forest type: 1) Open forest iilh lianas 2) Open fi ei i lh palms and 3) Open

forest iilh bamboo.














CHAPTER 4
DEFORESTATION TRENDS IN THE EXTRACTIVE RESERVE


4.1 Introduction

This chapter introduces land use changes in the Southwestern Amazonian state of

Acre and in the Chico Mendes Extractive Reserve. In particular, deforestation rates

outside the reserve and within the reserve will be examined and discussed focusing on the

state, municipal, reserve, and seringal levels. In addition, the chapter provides the

necessary land-use change background on which the following chapters 5 and 6 base

their discussion of household determinants of land use activities and keys socio-economic

drivers of this change in the CMER.

The eastern part of Acre, particularly the Acre River Basin, lies at the

westernmost limit of the development frontier that experienced massive changes in land

use since the 1970s. The Chico Mendes Extractive Reserve lies just at this limit of the

frontier (figure 4-1). To understand deforestation patterns within the Extractive Reserves,

one also has to look at land use and deforestation rates in the surrounding areas of the

state. First, land conversion and encroachment into rubber tapper areas were one of the

reasons for the mobilization of the rubber tapper movement and the creation of the

reserves. Second, until this day, land uses surrounding the reserve influence land uses in

the reserve to a certain extent, as there is a lot of social interaction along the boundaries.

Third, by comparing deforestation rates and patterns in the reserve with deforestation

outside the reserve can one assess the effectiveness of the extractive reserve in

maintaining forest cover.







































Figure 4-1 Location of the Chico Mendes Extractive Reserve in Acre


4.2 Deforestation and Land Use Trends in the state of Acre

Acre was almost totally covered by dense tropical forest before the 1970s when

most of its economy was based on rubber and Brazil nut extraction. According to the

Landsat satellite data, in 1978 Acre had only 2,500 square kilometers of deforested land.

By 1988, after the onset of larger development efforts, deforestation reached a total area

of 8,900 square kilometers (INPE 2000). This relatively low figure, compared to other

Amazonian states, does not seem to give cause for alarm. What is significant, however, is

that the area cleared down between 1978 and 1988 tripled in just ten years.









According to FUNTAC (1996), which analyzed deforestation at the state level,

the increase in deforestation in Acre has been greater in the upper and lower Acre basin,

the eastern part of the state, which comprises approximately 86% of the deforestation in

the entire state. In fact, half of this deforestation until 1996 occurred in the six eastern

municipalities of the state: Rio Branco, Senador Guiomard, Brasileia, Bujari, Placido de

Castro, and Xapuri (in decreasing magnitudes), with between 51% and 17% of these

municipalities deforested.

While this region experienced most of the cattle ranching expansions it also

concentrates a large portion of the extractive population. Deforestation from forest

conversion for agriculture and cattle pasture generated sometimes violent social conflicts

as the completely diverse social and cultural standards of production and organization

clashed. As extractive activities, mostly rubber and Brazil nut extraction, were almost

abandoned, the living conditions of the rubber tappers, who relied more heavily on forest

resources, were severely compromised. Therefore, this conversion caused a significant

rural exodus by the landless poor, many of them rubber tappers. As a result of the clashes

between, particularly, cattle ranchers and resisting seringueiros, the rubber tapper

movement was born, struggling for land, traditional livelihoods, and socio-economic

development. Thus today, Eastern Acre's extractive reserves, the major achievement of

this movement, are "islands" nested within and along the development frontier on

otherwise "prime" development land.

The greater rate of deforestation in the upper and lower Acre River region is

directly related to the transition from traditional non-timber forest production to ranching

and farming. Since the 1970s, deforestation rates have increased at quite high rates,

threatening the vegetation cover and forest of the state already compromised by about









15,000 squares kilometers of clearings in 1999 (See Figure 4-2)12. Concurrently, Acre's

rubber production, which was the basis of the state's economy, has decreased drastically.

This occurred not only as a result of the deforestation in this region, but also as a result of

national and regional development policies that favored cattle ranching and large scale

agriculture over extractivism, and the declining wild rubber market. Figure 4-2

demonstrates that there is a direct inverse relationship between deforestation and rubber

production in the state of Acre, illustrating how extractive activities (rubber) gradually

lost importance and large-scale land-use activities were prioritized. This inverse

relationship tends to continue until today, although rubber production has increased a

little in the last two years due to a new local government subsidy intended not only to

provide rubber tapper communities with better income, but to decrease deforestation rates

as well.13

The regional and local socio-economic factors that have affected land-use change

in Eastern Acre also have directly and indirectly influenced land use change within

extractive reserves. The Chico Mendes Extractive Reserve in particular is rooted in the

history of the grassroots movement for forested land but also in the history of land-use

change in the region.


4.3 Deforestation rates in the CMER

The CMER, with nearly one million hectares, borders the highway BR 317

leading to the Pacific through Peru. Although the CMER has served to impede large-scale



12 The deforestation rates between 1979 and 1988 were estimated by averaging due to the lack of data
13 The "Lei Chico Mende s" or "subsidio da borraha" was the first effort of the "government of the Forest"
elected in 1998 to strengthen extractivist communities by boosting rubber production. It provided an
additional 0,20 US $ for every kg of rubber for every rubber tapper in a production association. In addition
to making it worthwhile for many seringueiros to tap rubber again, it was intended to strengthen and
encourage the rubber tapper social organization (Governo do Acre 1999).














-- Deforestation [km2/year]

16000 Rubber production [1000 tons/year]
16000 12
14000 11
10
12000- 9
10000 8 W
E 8000 -_ 6
5 o
6000 -
4 -
4000 -- 3
2000 2
1
0 0
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 0
Years
Source:Anuario Estatistico da Borracha (1986,1987,1988,1989 e 1995); IBAMA 1998;
SEFE 2000; INPE, 1999.




Figure 4-2 Deforestation and rubber production in Acre (1979-2000)



deforestation along this highway, it has experienced accelerating small-scale

deforestation since it was established (Figure 4-3 and 4-4). It is predicted that within a

decade about 12% of the reserve area will exceed the legal limit for deforestation of 10%,

as determined by the Utilization Plan14 (Brown 2001). In addition, the higher occupation

of some parts of the reserve, mostly the southwest part, may cause land clearing to

exceed the 10% limit for those areas in a few years. Nonetheless, despite increasing

deforestation within the reserve areas, compared to levels of deforestation outside the




14 The Plan sets forth an upper limit of 10% of deforested area per household (of an estimated 300-400
hectares), including residential clearings, backyards, pastures, agricultural and abandoned fields, as well as
agroforestry plots. The plan also includes strict regulations of the extraction of rubber and Brazil nuts and
the development of management plans for new forest products. Timber extraction and hunting is restricted
to the resident's subsistence use. The Plan defines as "common use areas", rivers, lakes, main paths, and
beaches, although residents close to these environments tend to dominate their access. The resident
associations have authority to participate in the monitoring process of the plan and a team of resident
woodsmen were trained and given the legal status of environmental inspection agents.











w 690 30' w 680 00'
s10 05'















s11000'

0 30 60

Legend: Border of the Chico Mendes Reserve
-] Road
W Desforestationl986
Figure 4-3 Satellite image mosaic of the CMER illustrating deforestation in and out of the
reserve. Source: Sassawaga 1999.











w 690 30' w 680 00'
s10o
























s 100 05'
















s11 00'

003/6 002/6
0 30 60 Km

Legend: Border of the Chico Mendes Extractive Reserve
SRoad

SDeforestation: a) 1992; b) 1998.

Figure 4-4 Satellite image mosaic of the CMER illustrating deforestation in and out of the
reserve. Source: Sassawaga 1999.







48


reserve these are relatively small, particularly compared to areas along the BR-317

highway, which is dominated by large-scale cattle ranching activities. Figure 4-5 shows

the percent of overall deforestation in three periods of time that Sassagawa (1999)

determined for the CMER. The mean annual deforestation rates within the reserve were

0,137% a year for the 1986-1992 period and 0,227% a year for the 1992-1998 period.

This shows that deforestation rates of the second period nearly doubled compared to the

first one.





3.5

3.0 2.9

2.5

.2 2.0
1.54
S1.5

1.0 0.72

0.5 -

0.0
1986 1992 1998
Years

Figure 4-5 Percent of deforestation in the CMER (1986, 1992 and 1998) Source:
Sassagawa (1999)




4.4 Deforestation at the seringal level

Within the extremely large CMER Reserve there is great variability in

deforestation levels. Some areas have levels as low as 0.02 % (1998) while others are

approaching 8% (1998) almost reaching the legal limit of 10%. Table 4-1 shows the

different rubber tapper estates or seringais of the reserve, the area that they cover and the

proportion of that area that was deforested in 1986, 1992 and 1998, respectively.









As noted before, there is a great variability within the reserve as can be seen by

the very different deforestation rates that the seringais have experienced. Some, mostly

more remote areas such as the Seringal Arari, and Seringal Petr6polis have extremely low

levels of deforestation, whereas Fazenda Carmen has surpassed the legal limit of 10%.


Table 4-1 Total area and percentage deforestation in each seringal of the reserve within
the three associations of the municipalities
Seringal Area (Km2) in Deforestation/Seringal (%)
CMER 1986 1992 1998

Assis Brasil (AMOREAB)
Seringal Guanabara 707.69 0.61 1.67 2.75
Seringal Icurid 644.05 0.54 1.38 2.27
Seringal Paraguassu 195.11 0.62 1.62 6.70
Seringal Petr6Polis 121.68 0.02 0.02 0.02
Seringal Sao Francisco 299.33 1,12 2.12 3.27
Brasileia (AMOREB)
Fazenda Carmem 70.04 2.43 4.31 12.81
Fazenda Porvir 46.80 0.62 0.85 3.63
Seringal Nazare 128.22 0.26 0.92 2.32
Seringal Amapa 513.85 0.79 1.54 2.50
Seringal Apodi 191.17 0.47 1.54 2.43
Seringal Canamari 69.70 0.88 0.92 1.25
Seringal Humaita 144.87 1,13 2.93 7.95
Seringal Nova Olinda 358.32 0.06 0.14 0.47
Seringal Pacuara 69.59 0.89 2.04 2.56
Seringal Pindamonhangaba 99.05 1.71 2.46 3.63
Seringal Poronaaba 89.57 2.27 4.04 7.15
Seringal Sao Crist6vdo 165.81 0.55 1.26 2.98
Seringal Sao Salvador 62.83 1,40 2.34 3.42
Seringal Tabatinaa 759.88 0.24 0.49 0.56
Seringal Triunfo 103.17 1,40 3.57 5.08
Seringal Vale Ouem Tern 25.14 0.00 2.31 3.50
Seringal Varzea Aleare 83.43 1.17 3.03 4.44
Xapuri (AMOREX)
Fazenda Bonfim 329.40 0.15 0.53 1.38
Fazenda Ana Claudia 46.83 0.23 0.81 0.96
Fazenda Filipinas 442.21 1,96 3.62 7.55
Seringal Arari 104.67 0.00 0.00 0.06
Seringal Curitiba 120.75 0.03 0.07 0.07
Seringal Fronteira 339.89 0.37 0.64 1.05
Seringal Remanso 54.05 0.02 0.44 0.91
Seringal Sao Jose 95.65 0.26 1.08 2.05









Table 4-1 Continued
Seringal Sai Cinza 63.38 1,47 2.46 2,86
Seringal Venezuela 115.41 0.55 2.97 4.80
Seringal Albracia 255.00 0.25 1.04 2.68
Seringal Barra 183.44 0.18 0.39 0.92
Seringal Boa Vista 315.69 0.29 0.86 2.46
Seringal Dois Irmdos 68.88 0.99 1.09 3.91
Seringal Floresta 70.09 1,63 4.58 7.35
Seringal Independ6ncia 120.43 5.20 6.02 8.1
Seringal Lua Cheia 54.03 1.11 3.09 3.83
Seringal Nazare 273.06 0.71 1.24 2.50
Seringal Palmarizinho 87.29 1,13 2.26 4.83
Seringal S. F"o. do Iracema 456.78 0.37 1.09 2.23
Seringal S. Jodo do Iracema 41.27 5,38 7.87 7.87
Seringal Sdo Pedro 143.99 0.22 0.76 2.06
Seringal Siberia 227.54 0.50 1.29 2.48
Seringal Vila Nova 248.76 0.82 2.09 4.30
Source: Sassagawa (1999)

Generally, the seringais with higher deforestation rates are situated at the

boundaries of the reserve and in relatively close proximity to Xapuri or Brasileia and the

highway (See Figure 4-6). Another area of increased deforestation is in the municipality

of Assis Brasil, where there is more agricultural activity and Brazil nut tree density is

extremely low, limiting the extractivist potential of the area. In contrast, many seringais

toward the "middle" or "back" of the reserve are very little impacted.

It is important to note that these numbers represent overall deforestation per

seringal and do not always reflect similarities across these areas. Some seringais with

very high levels of deforestation may have largely forested land in one portion of the area

whereas some seringais can have large deforested areas although the seringal as a whole

is still forested. Therefore, a certain caution needs to be exercised when evaluating these

numbers.

Another important point of consideration is that prior to the creation of the

reserve, some seringais of the current extractive reserve had already experienced a small



















S 10030' 1 Xa uri


2 5



S1130 Ro Acre Assis BrsilB il

Legend: 1 Paraguaqu, 2 Humaita, 3 Porongaba, 4 Carmem, 5 Filipinas, 6 -
Floresta, 7 Independ6ncia, 8 SAo JoAo do Iracema


Figure 4-6 Map of the eight seringais with the highest deforestation rates in the CMER



percentage of deforestation, as evidenced by the 1986 rates. This is particularly true for

the areas close to the highway and towns, such as the Fazeda Carmen, Seringal Floresta

and Seringal Sdo Jodo do Iracema, thus setting a different trend for later land-use in

these areas.

Table 4-2 shows only the seringais that have reached the highest deforestation

rates since 1986, which are the areas of particular interest to this study. The four seringais

highlighted in yellow are the ones selected as study seringais. All of them are similar in

the level of deforestation and by 1998 they were all reaching the legal deforestation limits

of 10% per settlement in the reserve. If the same deforestation rates continue, these four

seringais will reach their legal limitation of deforestation in the next few years. The

Fazenda Carmen, which has even surpassed this limit, is currently even in discussion









with authorities about being turned into a traditional colonization project (personal

communication with Josemar A. Caminha, CNPT-IBAMA 2001).




Table 4-2 Seringais within the CMER with the highest deforestation rates
Seringal (with ov Deforestation/Seringal (%) Prediction
Senngal (with over
6% of deforestation) 1986 Rte in 86- 1992 eaten 1998
92 92 98 (Years)
Seringal Paraguassu 0,62 0,17 1,62 0,85 6,70 4
Fazenda Carmem 2,43 0,31 4,34 1,42 12,81 *
Seringal Humaita 1,13 0,30 2,93 0,84 7,95 2,5
Seringal Porongaba 2,27 0,30 4,04 0,52 7,15 5,5
Seringal Filipinas 1,96 0,28 3,62 0,66 7,55 4
Seringal Floresta 1,63 0,49 4,58 0,46 7,35 6
Seringal 5,20 0,14 6,02 0,35 8,10 6
Independ6ncia
Seringal S.Jodo 5,38 0,43 7,87 0,00 7,87 **
Iracema
Exceeded the 10% allowed;
** Presented a 0% deforestation rate in the 92 to 98 period.
Source: Sassagawa (1999).


Particularly in recent years, the communities of the seringais with higher

deforestation no longer have the historical close relationship with forest resources, and

have developed their agricultural and pastoral activities more intensely. The 1992-1998

figures demonstrate this increasing change in attitudes towards land-use, mirroring the

trends outside the reserve, though on a smaller scale. The increased adoption of these new

land uses; however, might become a potential source of conflict between these practices

and the Utilization Plan regulations.

Apart from deforestation levels, the four study seringais also have other

characteristics in common which might be determining factors in the increased

deforestation levels. All of them, excluding Filipinas, are accessible via roads and are






53


located near cities. In addition, the outer limits of the reserve in these areas border on or

are close to cattle ranches and colonization projects. However, they differ in area and

population size, as well as some land use activities. By more closely examining these

seringais, this thesis hopes to elucidate some of the factors that have led to this increased

deforestation in parts of the reserve. Chapter 6 will consider what factors at the household

level might affect land-use patterns in the four seringais. These characteristics include

head-of-household traits, family composition, and rubber estate location as determinants

of land use activities measured in six outcome variables.














CHAPTER 5
HOUSEHOLD DETERMINANTS OF FOREST AND NONFOREST ACTIVITIES


5.1 Introduction

Recently, studies of land use processes at the household level have emerged as an

important component in understanding social deforestation dynamics in Amaz6nia

(Brondizio and Siqueira 1997, Perz 2001a,b). This chapter provides a household level

approach, measuring some characteristics as determinants of land use activities in four

seringais of the Chico Mendes reserve. The central question of this chapter concerns how

the origin and age of the household head, family composition, and location of a

household affect land-use choices in the four study seringais. If this question is answered

we will understand much more about these diversified land use activities in these

seringais and their implication for forest conservation. The results might also delineate

future trends of land use in these more impacted areas of the reserve as well as other

areas that can follow this trend. Thus, this is important for basic policymaking regarding

land use regulation and economic development for these rubber tapper's communities.


5.2 Characterization of Rubber Tappers' Settlements in the CMER

In the Chico Mendes reserve there are 46 seringais, which are historically

determined rubber tapper estates. Each seringal is composed of approximately 10-25 or

more household units, the colocaqdo15 or family settlement. Before the creation of the



15 This term comes from the verb colocar, meaning to place or to put. The colocacgo or placement refers to
the area where a rubber tapper was placed or colocado in the forest by the boss during the rubber boom.









reserve, these units were not formally managed for multiple uses, but rubber tapper

families have managed them for rubber and Brazil nut extraction for generations. Today,

with the diversification of rubber tapper production systems, it is exactly at this landscape

level that most of the new land-use decisions are made and carried out.

Figure 5-1 shows a typical modern settlement in the CVMER. The settlements

comprise a residential clearing associated with backyards and small open fields; one or

more agricultural plots at some distance in the forest; extractive rubber tapping trails

radiating in loops from the residential clearing; a gathering and hunting territory

delimited by trails; and a pasture area, mostly resulting from old agricultural areas that

were not left to fallow. The size of these settlements ranges from 300 to 800 hectares and

is informally determined by the number of rubber tapping trails that historically defined

the productive unit of the settlement. In other words, settlement territories have no

physical boundaries that can be mapped precisely. Instead, the attention is concentrated

on the natural resources they contain, measured by the rubber trails and the rubber and

Brazil nut trees in that area. As noted, rubber tapper settlements cover a variety of land

use activities that are distributed seasonally throughout the years and require

differentiated family labor force allocation. The next section will present the allocation of

the rubber tapper' major land use activities.


5.3 Seasonality The Rubber Tapper's Year

A good way to understand factors affecting rubber tappers production is to learn

about their main land use activities throughout the year. Although production systems

shared by rubber tappers are similar, there are differences. How much time a family









spends on rubber tapping, on Brazil nut collecting, agriculture, livestock raising, and

hunting varies greatly between families and by season. In addition, the patterns of





S- -














Figure 5-1 Representation of the settlement in the CMER


work spent in each activity cannot be studied in isolation. The outcome from one activity

can only be analyzed in the context of another activity in order to understand how each

operation influences each other. Also, understanding household members' multiple roles

and responsibilities is crucial to the study of land use activities.

Figure 5-2 presents a seasonal calendar of rubber tapper activities throughout the

year. The principal seasonal influence on land use activities is that of the rainy and dry

seasons. Forest production activities complement one another seasonally. Rubber is

harvested in the dry season (April to August) while Brazil nuts are collected in the

"winter" (December to March). Since labor is an important limiting factor in the tapper's
..,;. : ,; ,t -
j .































"winter" (December to March). Since labor is an important limiting factor in the tapper's









production system, the fact that rubber and Brazil nuts do not compete for labor is very

beneficial and important to tapper families.



Activities J F M A M J J A S N D
Hunting
Livestock raising
Rubber trail clearing
Rubber harvesting
Brazil nut harvest
Land clearing
Field burning
Rice planting
Corn planting
Beans planting
Manioc planting
Rice harvest
Pasture planting
Corn harvest
Bean harvest
Manioc floor

Dry season
Wet season

Key: = Period of activity/occurrence

Figure 5-2 Seasonality in the rubber tapper's year



Land clearing for crop production is performed from July to August. Tappers

choose an area close to the house that is free of Brazil nut and rubber trees. Trees and

understory must be cut on the land to be cleared. This referred to as the broca and

derrubada. Broca refers to cutting vines and other small brush. After this is done, the

tappers performs the derrubada which is to cut the larger trees with an ax or a chainsaw

usually provided by the Association. After that, the vegetation is on the ground, drying

for at least three weeks so that it will dry well enough to burn well, which generally is









carried out in September. The land clearing process and rubber tapping need to be done in

the dry season and are carried out by the adult males. These concomitant high labor

activities cause a constraint in the family's labor force availability.

Corn and rice are then planted intercropped at the beginning of the rainy season

(September and October). These crops are harvested in February and March. Beans are

planted in April or May, at start of the dry season, and harvested in July and August.

Manioc is planted from October to December. Manioc is used to produce flour (farinha),

and is harvested as necessary. Following this planting period there is much weeding of

these crops to be done in the wet season. Following the second year, the plot will be left

in fallow and then used in cultivation again after a few years, will be transformed into

pasture or left to regenerate into forest. Usually all adult family members plant, weed and

harvest. Women help in planting and harvesting. Besides planting and harvesting, young

sons and daughters are involved in the weeding.

Livestock raising is carried out throughout the year. This includes chickens, pigs,

goats, sheep and cattle. Women usually take care of most of the livestock, excluding

cattle, which is often the responsibility of the sons or heads of households. Hunting

continues throughout the year for most rubber tappers. The head of household usually

undertakes this activity, but in the busiest times of the year tappers often have no time to

hunt and thus send their oldest sons instead.

In general, rubber tappers' activities are very well distributed throughout the year.

However, the period that requires the most family labor is between June and October.

During this period the family members are harvesting rubber, planting and weeding

crops, clearing new agricultural land and taking care of livestock. This period of labor









stress limits how much a family can dedicate to any one activity, and ultimately decisions

need to be made as whether to invest more into rubber tapping, agriculture, or pasture

creation. In the following section I will statistically compare the different forms of land

use with some household determinants, an important one of which is labor availability.


5.4 Conceptualization and Operationalization of Outcome and Independent Variables

Rubber tapper activities throughout the year involve a series of complementary

activities. Four main activities can be defined for rubber tapper production systems:

rubber tapping, Brazil nut collection, livestock raising, and crop production. Since these

activities have different influences on deforestation, the ones with low impact on the

forest were conceptualized as traditional forest activities, and the ones with high forest

impact were conceptualized as non-traditional forest activities. I consider six land use

outcome variables and a set of other independent variables that influence rubber tapper

land use activities. Table 5-1 presents outcome and independent variables names, their

operational definition and their descriptive statistics.

5.4.1 Outcome Variables

The first two outcome variables, rubber and Brazil nut production were

conceptualized as the basis of the traditional forest activities because they have low forest

impacts and represent the traditional forest based economy practiced by rubber tappers. I

operationalized rubber production in terms of kilos and Brazil nut in terms of latas.16 The

average finding for rubber production per household is 268 kilos per year. It represents a

very low rubber production per household, but the distribution is skewed, as indicated by



16 Rubber tappers usually measure Brazil nut by "latas," or eighteen to twenty-liter cans of cooking oil.
One lata of Brazil nut is equal to 13 kilos.













Table 5-1 Variable names, operational definition, and descriptive statistics, CMER
Variable Name Operational definition Mean Standard
Deviation
Outcome variables


Forest and nonforest activities
Rubber

Brazil nut

Area deforested

Swidden size

Pasture size
Heads of cattle
Independent variables
Household Background
Migrant


Age
Labor force availability


Total rubber production in 1999 measured in
kilos
Total brazil nut production in 1999 measured
in latas
Total size of pasture and swidden combined in
hectares
Total size of swidden plots in hectares per
settlement
Total size of pasture in hectares per settlement
Total number of head of cattle per settlement


Residents that came from the city or other
locations outside the reserve, such as
colonization projects and ranches
Age of the head of household


Sons Number of sons in the household aged 15
years or older
Daughter Number of daughters in the household aged 15
years or older
Seringal location (reference category seringal Filipinas)
Seringal Paraguaqu
Seringais Humaita/Porongaba Seringais Humaita and Porongaba were tied
together since they present the same land use
characteristics
n= 66


268.6061

227.2424

9.3788

2.2803

7.0985
8.3485


.3333


42.3788

2.0308

1.6154


.3030
.4091


418.3590

284.9748

10.9492

1.5863

10.0641
10.3247


.4750


12.9398

1.9683

1.6554


.4631
.4953











the large standard deviation. It reflects that some households do not tap any rubber, while

other households still have a considerable rubber production. Brazil nut production

averages 227 latas per household per year. In general, these traditional forest activities

findings show that rubber tapping activities have limited importance for many

households, although they are living in an extractive reserve where rubber is still

supposed to be a key part of the economy.

The other four outcome variables, area deforested, swidden size,17 pasture size

and number of cattle heads, I classified as non-traditional forest activities among rubber

tappers. All these variables represent important activities increasingly developed by

rubber tappers in the study seringais and are important indicators of the impact on forest

cover and deforestation in the reserve. In addition, the way these variables behave has

important development and political implications for rubber tapper communities settled

in extractive reserves.

The area deforested variable is an important measurement of household forest

impact, which has several implications in terms of land use regulation in the reserve, such

as the one defined in the utilization plan that sets forth an upper limit of 10% of

deforested area per household. In addition, the area-deforested variable is further

implicated in terms of the cultural identity of tappers as an extractive population. I

conceptualized area deforested as the permanent or temporary clearing of the forest for

pasture and agriculture.18 I operationalized this variable combining household pasture


17 Note that swidden agriculture to some extent is also a traditional activity for rubber tappers, mostly for
family subsistence.
1 I considered these two variables as a mean of measuring area deforested in the settlement. However, they
have different ecological impact on forest cover. Swidden might be left to fallow and thus have less
ecological effect on forest cover than pasture, which usually has greater impact, and is more permanent
forest clearing.









size and agricultural plot size. The mean household area deforested area was 9 hectares

per household.

Swidden agriculture is relatively traditional among rubber tapper because it was

practiced as a subsistence activity. This has a relatively small forest impact because small

areas of forest are cleared annually. In addition, the impact is low because a rotation

system is practiced by tappers whereby the same plots are used for two or three years and

then left in fallow to be used again after a few years. This rotation, if conducted properly,

may guarantee sustainability. However, I conceptualized swidden agriculture as non-

traditional forest activities because, more recently, swidden agriculture has increasingly

had greater forest impact in the study's seringais.19 Since agriculture has become an

important source of family income, many families have increased the size of swidden

agricultural annually, and swidden areas that before were left as fallow are increasingly

turned into pasture. I operationalized the swidden agriculture in terms of the size of the

swidden plot in hectares per household. The descriptive statistics show that the mean

swidden size was 2.2 hectares per household each year.

The pasture outcome variable represents the newest land use among rubber

tappers. An accelerated increase in pasture areas has taken place within the study

seringais and, thus, pasture represents the land use with the greatest impact on forest

cover. A recent perception among rubber tappers in these seringais is that the conversion

of land into pasture will increase the value of their settlement. This indicates their plans

for future land use choices and has a series of implications in terms of the future

economic and ecological viability of their reserve as well as in terms of their very identity


19 Traditionally, a tapper family usually has a hectare of agricultural land that is used for family
consumption throughout the years. In the studied seringais, I observed that many families have increased
their swidden plot at least with another hectare for cash crop agriculture. Though I did not distinguish these
families in the analysis, I recognize that it would have been important to look at the families separately.









as rubber tappers. I operationalized the pasture variable in terms of the size of the pasture

area in hectares per household. The mean pasture area was 7 hectares per household.

Likewise, the cattle outcome variable is another important indicator of new

economic activities for the rubber tappers. It was agreed upon in the utilization plan of

the reserve that families could have a small number of cattle, just to improve their daily

diets and transport their products. However, in the study seringais, cattle raising is the

activity showing the greatest increase recently. Some households in the study seringais

have no cattle, while others have over 30 heads. It seems that most of the tappers' profits

in the studied seringais are invested in cattle raising. These cattle raising expansion

phenomenon has been perceived as a "savings account", an investment that can be easily

changed into cash. Again, the increase of this economic activity among rubber tappers

has implications in terms of the their concept of economic development and the

conservation goal of their extractive reserve. I operationalized the cattle outcome variable

in terms of the number of cattle heads per household. The mean number of cattle heads

per household was 8.

5.4.2 Independent Variables

The production system developed and managed by the rubber tappers can be, to a

great extent, characterized in socioeconomic (e.g., age, family, labor force availability,

price) and biophysical terms (e.g., location of the household, distance to market, resource

availability). The independent variables consider three groups of two variables each that

affect the production systems: household head background (migrant status and age), labor

force availability (sons and daughters of the household head) and location of the study

seringais.









In addition to these characteristics, a series of other variables20 were considered,

but the ones discussed here were noted to be most significant to the model discussed later

in this chapter. Moreover, the limited number of cases in my sample (66), lowered the

likelihood for statistical significance of these other independent variables.

The household background involves two indicators: migrant status and age of the

household head. I operationalized the migrant variable according to previous place of

residence of the household head, which might allow for the assessment of difference in

land use practices. Migrant household heads might be more engaged in non-traditional

forest activities than non-migratory household heads. The age of the household-head

might also be an important indicator of the economic activities they have engaged in

throughout their life cycle. Older household-heads might be more concentrated on

traditional forest activities (e.g., rubber tapping), while younger household head may be

more engaged on non-traditional forest activities (e.g., agricultural crops). The

descriptive statistics reveal that 33% of the head of the household were migrants. The

mean age of household heads was 42 years old.

The rubber tappers' extractive system varies, depending upon how large the

family is and upon how dependent the family is on income producing activities. The

family itself provides most of the labor on the landholding of a tapper family. Thus, I

conceptualized labor force availability in terms of the number of sons and daughters, and

operationalized it in terms of their age. The sons and daughters considered for this



20 Among these are: length of the residence; the total number of the residents in the settlement; how old the
settlement is and how many tappers have occupied it. Also, I considered household membership in rubber
tapper associations, rural workers union and cooperatives. Moreover, I considered household head access to
government credit. In terms of location, I considered the distance to market measures and different
transportation means (walking time, animal time and truck time) and types of access to the market (trail,
ramal and road). Some of the variables that were theoretically important, but did not show stronger
statistical significance were excluded. Others, that were less theoretically important, but presented a much
stronger significance, were kept in the model.









analysis are the ones who are fifteen years old or older. I considered that sons/daughters

over this age start to make an important contribution to the family labor force and thus

influence productive land use actitivities. Besides the household-head, the most

important factor for labor force is the number of sons of age fifteen and older to tap

rubber or to help clear land. This is because the availability of this manpower influences

the area of land the family can exploit for rubber or put into pasture or field crops. For

example, rubber and Brazil nut activities require less labor input, while agriculture

activities require more labor input. As noted in table 5-1, the household head had on

average 2.03 sons of age fifteen years old or older, and 1.62 daughters of age fifteen

years old or older.

There are structural forces outside the household level characteristics that are

important in determining land use activities. The location of the household itself in each

seringal might present some degree of influence in land use activities. However, the

location of the seringal as a whole is a more powerful structural factor. Therefore, I

conceptualized location at a seringal level. The core aspect here is that the location of the

seringal influences land use activities since it involves aspects of market-distance and

transportation costs that may reduce some land use activities and stimulate others. A

tapper settled in a seringal closer to towns often presents different activities from the

tappers who live farther. For example, more distant seringais tend to influence tappers to

emphasize rubber and Brazil nut activities with high value relative to weight and bulk

over less easily transported agricultural products.

The four study seringais present different characteristics in terms of location.

Seringais Paraguacu, Humaita and Porongaba are road-accessed seringais, while seringal

Filipinas is far way from the road and is easier accessed by river in the Xapuri region. It









is important to analyze land use activities in the seringal Filipinas compared to the other

three seringais. I classified the seringal Filipinas as the reference category, as it represents

most closely the scenario of the traditional seringal, being farther from the road, and

having the bigger rubber production. In addition, I operationalized Seringais Humaita

and Porongaba as a unit due to the fact that they border each other and I observed that

they have the same land use characteristics, thus not affecting the analysis. The

descriptive statistics show that of the 66 households in my sample, 30% were located in

seringal Paraguacu, 40% in the seringais Humaita and Porongaba and 30% in the seringal

Filipinas.


5.5 Correlations Between Outcome and Independent Variables

Table 5-2 presents correlation coefficients between each land use outcome

variable and every independent variable. Traditional forest activities variables have a

strong correlation with each other. That is to say that households that tap rubber also tend

to collect Brazil nuts. However, they differ when correlated with nonforest activities.

Pasture and cattle have a strong negative relationship with rubber production, but this is

less true for Brazil nuts, and these coefficients are not statistically significant. Also,

Brazil nuts have a positive association with swidden agriculture and thus households that

collect Brazil nuts tend to have swidden plots, but do not necessarily have pastures or

cattle. In general, the traditional and non-traditional forest activities outcome variables

exhibit the expected correlations.

As anticipated, the non-traditional activities have strong positive correlations with

area deforested. It is not surprising that area deforested has a strong correlation with

swidden and pastures as the area-deforested variable is equal to swidden plus pasture

area. Similarly, area deforested is strongly associated with cattle. Since most of the area









Table 5-2 Correlations between traditional and non-traditional outcomes variables and
household background, labor force availability and seringal location variables, CMER.
Variables name

Rubber Brazil nut Area Swidden Pasture Cattle

deforested


Outcome variables
Rubber 1.00

Brazil nut .303*
Area deforested -.190

Swidden .104
Pasture -.254*

Cattle -.247*
Independent variables

Household Background
Migrant -.287*

Age .208+
Labor force

Sons .026
Daughter .077

Seringal location
Humaita/Porongaba -.331**

Paraguacu -.312*
+ p <.10; p <.05; ** p <.01.


1.00
.131

.324**
.018

-.127




.186

.261*


-.053
-.089


.392**

-.782**


1.00

.598**
.915**

.540**




.131

.284*


.345**
.103


.205+

.022


1.00
.338**

.105




.268*

.246*


.147
.149


.267*

-.242*


1.00

.613**




.099

.269*


.355**
.047


1.00




.079

.186


.339**
-.063


.176 .107

.121 .234+









deforested is for pasture making purposes, it has a positive relationship with cattle. Also,

swidden agriculture has a strong positive relationship with pasture, but not with cattle.

This implies that households have increasingly transformed swidden plots into pasture,

even when they do not have cattle. In general, the non-traditional forest activities

correlations are all highly significant, except for swidden and cattle.

The forest activities variables (rubber and Brazil nut) have distinct correlations

with the three groups of explanatory variables. Rubber production is negatively

associated with migrants and also a weaker correlation with age, but still suggests that the

older the head of household, the more rubber tapping activities are performed. In terms

of location variables, rubber production tends not to be done by households that live in

the seringais Humaita, Porongaba and Paraguacu. In other words, the household head

more engaged in rubber production tends to live in the seringal Filipinas. This reflects

Filipinas's involvement with traditional forest-based economy. With Brazil nut activities,

there is little correlation with migrant status, but it shows a stronger association with age,

where the older head of the household tends to collect more Brazil nuts. In terms of

location, seringais Humaita and Porongaba have a strong correlation with Brazil nuts,

while seringal Paraguacu exhibits a negative correlation. That is, households that collect

Brazil nuts tend to be located in Seringais Humaita and Porongaba. In general, seringal

Filipinas has lots of rubber tapping and less Brazil nut collectors.21 Seringais Humaita




21 Note that although seringal Filipinas is considered the more traditional seringal, it has less Brazil nuts
production than seringais Humaita and Porongaba. It might be related to market distance factors. Distance
to town is an important factor determinant of how many Brazil nut a family will collect. If there is no such
nearby market, it is usually uneconomical to transport Brazil nuts because they are heavy and require more
transportation structure. Seringal Filipinas is far from the road. Households in seringal Filipinas might be
not as attracted by Brazil nut collection as the households in seringais Humaita and Paraguaqu. Having road
access, households in seringais Humaita and Porongaba have more transportation facilities for Brazil nut
activities. Also, it would be important to analyze other variables that consider ecological factors such as
distribution and density, affecting the availabilities of Brazil nut trees in the reserve.









and Porongaba have less rubber tapping and more Brazil nut collectors. Seringal

Paraguacu, on the other hand, has less of both rubber tapping and Brazil nuts collecting.

Area deforested is strongly correlated with age of the household head and number

of sons. That is, older households and households with sons tend to have larger area

deforested. In terms of location, area deforested has a small effect in the seringais

Humaita and Porongaba. This means that households deforest slightly more in these

seringais than in seringal Filipinas. In general, area deforested is related to the age of the

household head and male labor force availability.

Swidden agriculture has a strong association with the two-househod background

indicators: migrant status and older households tend to have more swidden agriculture. In

terms of the location, households located in the seringais Humaita, Porongaba and

Paraguacu tend to have much more swidden agriculture area than households in Filipinas.

Both pasture and cattle have a strong positive correlation with sons in the

household, which demonstrates their role in terms of family labor force allocation.

However, these variables differ in terms of age of household-head and location of the

seringais. Pasture is positively associated with the age of the household-head, with older

household-heads having more land into pasture. Cattle, on the other hand, are correlated

with the location of the seringais. Only seringal Paraguacu shows a correlation with cattle

and thus in this seringal there are relatively more cattle raising activities than in the other

seringais.


5.6 Analysis of Multivariable Regression Models

Table 5-3 presents the results of Ordinary Least Squares (OLS) models of rubber,

Brazil nut, area deforested, swidden, pasture and cattle regressed on indicators of

household background, labor force availability and seringal location. As noted earlier, I















Table 5-3 Multivariable regression models at rubber extraction, Brazil nut extraction, area deforested,
head of cattle, CMER.


swidden size, pasture size and


Model
Independent variables Rubber Brazil nut Area deforested Swidden Pasture Cattle

(1) (2) (3) (4) (5) (6)
Constant 3.858* 2.146* .563 -9.832 -.507 -1.528
Household Background
Migrant -.1.439 .795 .203 .294+ .222 .344
Age 6.035+ 6.180** 1.248 1.012+ 1.647 1.289
Labor force
Sons -.213 7.598 .138* 5.147 .200* .367*
Daughter .196 9.571 6.515 8.238+ 4.163 -.110
Seringal location
Humaita/Porongaba -5.417** -.409 .587* .216 .937** 1.525*
Paraguacu -6.088** -5.870** .247 -.226 678+ 1.676*
(Reference Filipinas)
Adjusted R .452 .638 .199 .191 .201 .185
F ratio 9.781** 19.775** 3.656** 3.519** 3.690** 3.421**
+ p < .10; p <.05; ** p <.01.









considered several other explanatory variables. The problem with the multivariable

model is that, due to the small sample size, I was only able to set six independent

variables. These six variables were selected out of several potential variables that I

introduced into the model. They were chosen because they proved to be statistically

significant for at least one or two outcome variables and had theoretical importance.

Other variables that are considered theoretically important but did not prove significant

were not included in this model. Also, I used the same set of variables for all outcomes to

facilitate comparisons among the models. In this way, I tried to develop a single set of

variables that represented a stronger model overall. After this procedure, it became clear

that I had a small number of explanatory variables that were stronger predictors of all the

outcome variables.

The primary determinants of rubber production (model 1) involved the age of the

household heads and location of the seringais. The model for Brazil nut (model 2) had

similar significant variables, but differences are evident. Brazil nut has a stronger age

effect, even bigger than that for rubber. The stronger age effect for Brazil nut is because it

not only has a more stable market than rubber, but also requires less intensive labor force

The location variable is less important for Brazil nut than for rubber production.

That is to say that rubber has a stronger negative effect with seringais Humaita,

Porongaba and Paraguacu, while Brazil nut has a negative effect only with seringal

Paraguacu. This seringal is located in the Assis Brazil region, where Brazil nut trees have

lower density, leaving fewer options for residents to increment their forest production

economy. None of the labor force availability variables show significant net effects on

rubber tapping and Brazil nuts activities, which suggests that regardless of household









background and location of the seringais, young residents are not engaged in traditional

forest activities.

The area-deforested model (model 3) reveals a limited array of significant factors.

The determinants of having area deforested emphasize the labor force factor and the

location variable. Sons of the household head exert a strong and positive effect on area

deforested. Therefore, the availability of sons of household head is important to having

area deforested, a reflection of the high male labor demands in forest clearing, harvesting

cash crop and processing land into pasture. The location factor has important effect in

seringais Humaita and Porongaba and thus there are more non-traditional land use

activities (swidden and pasture) driving deforestation in these two seringais.

In contrast, the determinants of agricultural production (model 4) accentuate the

household background variables. Ages of the household head and migrant status are

determinants of agricultural production. This basically indicates that older household-

heads imply having a bigger family and they have tended to expand their agricultural

production for family consumption. In addition, this demonstrates that migrant

households have been more exposed to agricultural activities, while non-migrant

households tend to engage in traditional forest activities. Moreover, model 4 presents two

important factors. It is the only model in which daughters show effects on swidden

agriculture, and location of the seringais present no significant effect on swidden

activities. This demonstrates that daughters of household head give important

contributions to swidden agriculture activities such as planting and harvesting crops. It

also shows that agricultural production is practiced in all these seringais, which

demonstrate its fundamental role as a subsistence activity.









The models for pasture and cattle (model 5 and 6) are stronger and have the same

significant variables. Sons of the household heads exert a strong and positive effect on

pasture size and number of cattle. This demonstrates that the availability of sons for labor

is crucial to having pasture and cattle and suggests that expansion of nonforest activities

will continue in these seringais, which tend to be carried out by the young residents. As

conceptually proposed, pasture area and cattle raising activities are greater on the non-

traditional seringais (seringal Filipinas was considered the closest one to the traditional

seringal). Thus, it demonstrates that households settled in seringais with better road

access and closer to the towns have been more involved in non-traditional forest

activities. In general, the land under pasture and cattle raising activities depends on the

availability of labor and the location of the seringais. In this model, household

background variables were unimportant to pasture and cattle, implying that non-migrant

households are as involved in pasture expansion and cattle raising as migrants.


5.7 Discussion

This chapter examines how some of the household characteristics and rubber

estate location determine land use activities among rubber tappers in four seringais of the

Chico Mendes reserve. The findings from these empirical models that account for

household background, labor force availability and seringal location show that these

factors exert a significant effect on land use activities. Household background variables

have effect on rubber and Brazil nut extraction, as well as swidden agriculture. Labor

force availability variables have significant effects on area deforested, pasture and cattle.

Finally, location of the seringais has important effects on almost all outcome land use

variables, excluding the swidden model.









The findings of age of household head and labor availability call attention to the

age structure of the rubber tapper families, showcasing differences in labor allocation at

the household level. It may be that heads of household are more involved in forest

activities (rubber and Brazil nut) and the sons with pasture and cattle. Older household

heads might have worked with rubber activities throughout their life cycle, and thus,

although rubber production no longer has market stimulation like before, they continue

tapping rubber because it is the practice of which they have the greatest knowledge and

with which they identify themselves. If this interpretation is correct, older household

heads will insist on continuing this activity if at all possible, even if they are not

maximizing their income. However, it must be stated that this effect of older household

heads with rubber activities might be because rubber activities requires heavy labor and

more specialized skills so that the younger residents are not as apt to develop this activity.

Due to their skills with rubber, older household heads might have more profit from

rubber activities than non-traditional activities.

By contrast, the younger residents (sons of household heads) probably are relying

less and less on a purely extractive strategy and more on farming systems. If the younger

residents continue to go through their life with the same land use practices, their future

trajectory as "rubber tappers" will be very different from that of their parents, and we

might see increases in the extent of land under pasture in these seringais. The distribution

of different activities among members of the rubber tapper families shows that people

have to make choices, which may favor one component of production at the expense of

the others. The land use strategies chosen reflect the families' judgments of the available









options, taking into account immediate necessities, market criteria, food security as well

as personal preferences.

The location of the seringais is an important factor differentiating diverse

livelihood activities. I assume that the seringais, with the exception of Filipinas, have

better market integration, which might lead to better prices for cash crops and livestock.

This in turn should increase forest clearance. Of course, the forest clearance decision is

based on the expected increase in family income, although I did not analyze data about

prices and income. Market integration in the long run may foster even more forest

clearance in these seringais, if such integration does not result in increased levels of land

use regulation enforcement. While market integration and increasing household income

may increase forest clearance, the rate at which that occurs could dramatically change

rubber tappers' relations with forest resource.

Given the extremely large territory of the Chico Mendes reserve, there are

contrasted interests among rubber tapper communities living in different part of the

reserve. Development programs in the reserve tend to benefit these communities that live

in seringais located closer to the towns. It seems that the rubber tappers settled in these

seringais not only have more development benefits, but also will not be able to enforce

the reserve's land use regulation, as shown by the increasing deforestation rates. In this

case, reserve development programs and land use regulation might need to be approached

as a way to benefit a particular tapper community without affecting negatively the

interests of others that also deserve attention. I suggest two strategies that might indicate

a role for development activities and land use regulation.









Development initiatives for rubber tappers should first consider the different

zones in the reserve. Rubber tapper communities settled in seringais closest to the cities

should receive development incentives to stimulate traditional forest activities. Tappers

focused on agriculture and livestock raising should be stimulated to occupy more distant

seringais, with the stimulus of having more land for deforestation. If these approaches do

not have any repercussions on the deforestation scenario of today, at least deforestation

levels would be more dispersed, which would have some ecological benefits, and rubber

tapper communities would be more equally distributed in the reserve's territory, resulting

in less pressure on forest resources. Both are unrealistic suggestions.

The deforestation situation in these seringais will pose a serious dilemma for land

use regulation in the reserve. The forest reserve quota legally required on private property

in Amazonia currently is 80%, although it is in a huge debate for changes in the Brazilian

congress. Due to extractive reserves' importance as forest conservation tools, rubber

tappers are allowed to deforest only 10% of their settlement. Some seringais are

increasingly deforested areas and will reach the limit quota in few years. In this context,

rubber tappers who are supposed to be forest conservationists might reach the same level

of deforestation as other non-traditional rural workers on private property in Amazonia,

such as the colonists. The difference is that tappers have the privilege of holding

hundreds of hectares of land. In this context, it will become increasingly hard to justify

the extractive reserve model, if rubber tapper communities cannot keep low deforestation

rates. If these study seringais exceed their deforestation legal limit, they might have to be

taken out of the reserve territory and transformed into another land use strategy, and thus

more people would have access to land.









Cattle and pasture expansion in these seringais will pose a difficult problem for

rubber tapper communities in the near future, and one that will call for political decisions.

The initial concept of the reserve, which advocates finding ways to alleviate poverty

without degrading the environment, could be a guide to understanding the new decisions

that rubber tapper communities are making in terms of land use and to formulate

development policies that regard tappers' modified socio-economic development

perspective and still guarantee forest cover. So far, there are no major concerns from the

Brazilian Environmental Agency (IBAMA) and rubber tapper organizations about the

pasture and cattle expansion in the reserve.

Some principal factors that fall outside of this analysis that influence the tappers'

production system are the improvement of state's transportation networks, and the

emergence of autonomous tappers who are more socially and politically organized, and

less dependent on the traditional aviamento system. Experiencing less of the inhibiting

effects of the greater isolation and the strict traditional marketing system as in past

decades, many tappers now have growing opportunities and incentives to maximize their

production. Additionally, extractive lands now have more pressure from alternative land-

uses such as ranching and agricultural colonization than ever before.

Different combinations of land use patterns and diverse social relations are part of

the rubber tapper communities of today. A broad range of social and environmental

factors determines their diversified production systems. The household characteristics

and seringal location factors discussed in this chapter offer important insight into land-

use strategies at the family level, representing an integral fraction of these determinants.

Socio-economic criteria, such as access to markets, development initiatives and policies,






78


population dynamic, land use regulation, and so on, need to be broken down into smaller

units so that a complete but detailed view of the actual dynamic of land use is obtained

for these seringais and the reserve as a whole. In chapter 6 I will discuss some key land

use changes taking place in these seringais in a long-term perspective, and their role in

deforestation levels in the reserve and in the study seringais.














CHAPTER 6
DRIVERS OF LAND USE CHANGE IN THE CMER: Population and Settlement
Dynamics, Changes in the Production System, and Social Organization


6.1 Introduction

In the previous chapter, small-scale changes in land use at the household level

were assessed in order to identify household factors that drive these land use changes. In

this chapter I will explore primarily socio-economic factors that drive land use change at

a seringal and reserve level.

Socio-economic drivers of land-use in small forest dwelling communities are still

little understood for large parts of the Amazon Basin, compared to studies of large-scale

development projects. In most cases, studies with forest dwelling small-holders have

been unable to incorporate an integrated perspective of changes in land use cover and

particular socio-economic factors. In the case of the C MER, although a study has

established deforestation rates and patterns (Sassagawa 1999), the underlying reasons for

changes in land-use and land-use decisions have not been examined. As the establishment

of Extractive Reserves reaches 10 years, it is important to breach this gap and determine

what is driving land-use change in extractivist communities at the threshold of the

development frontier. This is of particular importance because the appeal and an often-

cited justification of extractive reserves lies in low deforestation levels and the

conservation of forest resources. Thus, understanding what drives deforestation in those

areas that have experienced higher deforestation is a timely and fundamental endeavor,

that will hopefully aid in the prediction and planning of land-use in the extractive reserve.









The reasons for increased deforestation rates in concentrated areas of the CMER

most likely lie in a variety of factors. A broad range of factors influences land-use

choices and the rubber tappers' quality of life to some extent or another. First, with the

demarcation of the extractive reserve, more families may have moved into the area

searching for secure land tenure. This secured land tenure has also led people to invest

more into long-term forms of land-use which include increased agricultural activity and

cattle holdings and as well as a stronger investment into structural elements of their

settlements, such as fences, houses and storage facilities.

Second, recent development projects and credit programs have encouraged

residents to diversify their land use activities toward agroforestry and cash crop

agriculture as well as small scale cattle raising, activities that result in a increased cleared

areas in the coloca(oes. Changes in the production systems are also a result of decreased

investment into the extraction of forest products. As discussed in the previous chapter,

the price for rubber has decreased significantly in the last decades and many people have

abandoned rubber tapping in favor of non-extractivist forms of land use. Brazil Nut

extraction, though more profitable than rubber, also has decreased. Third, the influence of

the rubber tapper social movement and its organization for productive activities needs to

be stressed, as many new opportunities and contacts are accessed through the various

organizations, which increasingly mobilize production processes and link individual

residents to larger development and other government projects. Fourth, more localized

variables that can affect land-use choices are the relative proximity to towns, means of

transportation, access to markets, and exposure to cattle ranches and colonization

projects, as is the case for the four study seringais.









The present land-use patterns and decisions are also embedded in the particular

history of each seringal and its role on the development frontier and the rubber tapper

resistance. Seringais at the periphery of the reserve that had part of their land transformed

into cattle ranches or adjacent to those are more vulnerable to fast land use changes.

Overall, deforestation in these areas occurs as a combination of many factors, which

influence each other. It is impossible to pin down increased deforestation levels as the

result of a single factor.

All of these variables cannot be studied simultaneously; therefore, I will

concentrate the analysis on three main factors that influence land use changes in these

seringais. First, I will analyze the population and settlement dynamics in the reserve as a

whole and the study seringais, which may lead to an understanding of why these four

seringais not only have the highest population growth but also have experienced the most

drastic land-use changes. Second, I will discuss the changes in the rubber tapper

production systems in order to explore the economic factors that are leading to land use

changes. Finally, I will analyze the social organization model in the reserve and how it

may influence land use change through different levels of organization of the

communities, political influence of the rubber tapper leaders, and the Residents'

Associations in these seringais. This chapter will mostly focus on an analysis of the

IBAMA census data of the reserve in three different periods (1995-1998-2000).


6.2 Population Dynamics in the CMER

According to data from the census carried out by the Brazilian Environmental

Agency (IBAMA) in 1998, the total number of residents in the reserve is approximately

seven thousand people, of which 44.8% are female and 55.2% are male. 63.3% of the









population is under twenty-one years of age, which demonstrates that the reserve has a

very young population. Although the reserve has territory in six municipalities of the

state, the majority of the reserve population is located in three municipalities along the

BR-317 highway: Xapuri, Brasileia and Assis Brasil.

Figure 6-1 presents the population dynamic of the reserve in these three

municipalities from 1995 to 1998. Xapuri, which comprises an area of 3,208.91 km2 in

the reserve, concentrates most of the residents of the reserve. In 1995, Xapuri had 3,660

residents (58.5% of the total reserve), decreasing to 2,612 (43.3%) in 1998, which

represent a decrease of 28.6%. Despite this decline, it still holds most of the population of

the reserve. On the other hand, Brasileia, which comprises 1,330.30 square kilometers,

had an increment in its reserve residents from 1,338 (21.4%) in 1995 to 2,120 (35.2%) in

1998, representing an increment of 36.8%. Although this data suggests that there are

significant changes in population levels in the two municipalities, this is in part an artifact

of an administrative change of 2 seringais from Xapuri to Brasileia between 1995 and

1998 (personal communication Josemar C. CNPT/IBAMA 2001), demonstrating the

difficulties of conducting an analysis at the municipal level. Assis Brasil, comprising

2,158.30 square kilometers of the reserve territory, maintained approximately the same

population size of 1,300 (21.5%) residents in the two periods. Overall, population levels

in the reserve have stayed approximately the same in the period 1995-1998.

Nonetheless, this very general data does not reveal more subtle changes in

immigration and exodus in the reserve, as well as internal migration from one seringal

into the other or from one colocacao to the other. A true understanding of this dynamic










can only be gained by examining the records of all the settlements in the reserve over

time, or at least looking at particular seringais, as does this study.






4000
3500 -
.2 3000 -
2500 -I-- Xapuri
2000 -_- -- Brasileia
I 1500 Assis Brasil
0 1000
500
0
1995 1998
Years

Figure 6-1 Total population of the CMER in three municipalities



Traditionally, rubber tappers have always moved around because theirpatrao

determined this in the old seringal times or later on by their own volition. Mobility has

always been an important element in rubber tapper existence be it due to the old rubber

system or the more recent insecurity in land tenure. Also, in recent decades with the

uncertainty of the extractivist economy, many people have moved within the region and

from forested areas to urban areas. Different fluxes of migration characterize the area:

exodus from the seringal into the urban centers, immigration into the CMER from urban

areas, and migrations at a local level from one seringal to another.

Particularly, Xapuri and Brasileia are towns that have increasingly accommodated

rubber tappers. It seems that the population in the Xapuri area of the reserve has

decreased more than Brasileia has increased, indicating some kind of out-migration from

the area into urban centers. Xapuri might be particularly vulnerable to exodus because it









is closer to the state capital Rio Branco than the other municipalities, and rubber tappers

have better transportation access to the area. In addition to out-migration, many families

in the nearby seringais (e.g., Dois Irmdos) have close relatives in Xapuri and thus spend

time in town, resulting in almost double residency in and outside the reserve. These

people often alternate time spent in the city and the seringal.

While some people have moved out of the area others have moved into or

returned to the reserve for lack of quality of life on the urban periphery, and

unemployment in the cities. Before the establishment of the reserve many families had to

leave their settlements in this region due to encroaching cattle ranching in the boundary

of today's reserve. With the establishment of the reserve, many of these families that had

moved to the cities gained right to the land and thus some slowly started returning to the

seringal, or if they opted for staying in the cities, passed the settlement rights to relatives.

The families that returned to the forest realized that living in the reserve areas granted

them not only the rights to the land, but also to several benefits from the government

agency, responsible for the reserve management.

Assis Brasil, the most remote of the three municipalities, might have maintained

its relatively low population size in these two periods due to its further isolation. In the

seringais of this region, the influence of urban lifestyle on the reserve residents is less

prevalent. This is because Assis Brasil is a very small and remote town that does not offer

much better social services than those that are offered in the reserve, with the exception

of health care. In addition to the distance, the road access to Assis Brasil from Brasileia

and Rio Branco, is very precarious and during the rainy season becomes muddy and









practically impassable. However, this road currently is being paved and should be

completed in 2002.

In terms of internal migration, Brasileia and Xapuri do not differ significantly to

influence households to move to one or another territory in the reserve. Both cities are

close to each other, and many tappers that live in the Brasileia region may interact and

travel to Xapuri and vice verse depending the distance inside the reserve and the means

of transportation available. While Brasileia is slightly larger, basically, these cities have

the same market options. (i.e., cooperatives to sell their production) and also offer the

same goods that they need in their household. However, factors inside the reserve might

influence a household more to migrate to another settlement, independently of the

municipality area. One of the factors might have to do with household consumption, such

as game and hunting areas. If game resource become scarce a household is likely to move

to another settlement. In addition, rubber tapper families tend to live closer to their

relatives. If the decision of the household head is to leave, the son/daughter that live

closer to them might decide to move as well. Thus, as overall population in the reserve

has remained approximately the same, migration occurs both into and out of the reserve,

and people move within the reserve, it is hard to determine particular migration patterns

and population dynamics at a reserve scale and municipality.

When considering deforestation in these municipalities, it is also important to

look at population density per area and the link of population to deforestation rates. Both

population density and deforestation rates vary greatly across these three municipalities.

In 1998, Assis Brasil had the lowest population density with 0.60 persons per square

kilometer, followed by Xapuri with 0.81 persons per kilometer square. Brasileia, on the









contrary, has a population density of 1.60 person per kilometer square, almost double that

of Xapuri. According to Sassagawa (1999), the Xapuri area presents the highest

deforestation percentage in 1998, although it has shown a decrease in population,

resulting in a higher per capital deforestation level. Thus, even though the population has

decreased, or rather a smaller number of seringais belong to this municipality, the

deforestation rates continued to increase the region.

Overall, the relationship between population growth and deforestation rates is low

at the scale of municipalities. This is probably due to the large extent of each of these

areas and the very distinct population density in different parts of the municipalities. The

northernmost and remote areas in the Brasileia, along the laco river has few households

and immense areas that are uninhabited, while the southern part holds a high population

density, with seringais linked to roads and closer to the city. Similarly, higher levels of

deforestation are concentrated in particular zones of the reserve. Deforestation at this

level appears to be driven by the increases of population in concentrated areas of the

reserve, which would require different scale of analysis. Thus, to understand it, we need

to look at different scale processes operating deforestation in the reserve. In the next

section I will present a seringal level analysis of population grown and deforestation

rates, for those seringais with particularly high deforestation rates in the reserve.


6.3 Population Dynamics in the Study Seringais

Figure 6-2 demonstrates the changes in rubber tapper populations in the four

study seringais from 1995 to 2000, as obtained from the IBAMA census records. In

general, the number of residents by seringais has increased constantly from 1995 to 1998

irrespective of the municipalities where they are located. In the seringal Paraguacu the









total number of residents was relatively high, with 221 people in 1995 rising to 290

residents in 1998, presenting an increase of about 31% in the population growth. Similar

population growth has taken place in the seringal Filipinas, the most populated of the four

study seringais. In 1995, it had a population of 295 people and in 1998 it reached a total

of 378 residents, demonstrating an increase of 28%. Seringal Humaita had a population

of only 53 residents in 1995, reaching a total of 96 residents in 1998, representing in an

impressive increase of about 81% in the population size.






400
350
300---
20 3- -- Paraguacu
250 ,
o -0- Humaita
.200 A
SFilipinas
o 150 Porongaba
100
50
0 --
1995 1998 2000
Year

Figure 6-2 Total population in the four study seringais



The highest rate of growth occurred in the seringal Porongaba, which in 1995 had

only 23 residents, growing to 126 in 1998, an increment of about 447%, although overall

population density is lower compared to the other study seringais. It is clear that the

period of largest population growth was from 1995 to 1998, with the exception of

seringal Humaita. Here, population growth continued through 2000, reaching a total of









210 residents, which represents an increment of about 118% compared to 1998 and 296%

in relation to 1995. While both Porongaba and Filipinas seringais maintained

approximately equal population levels from 1998 to 2000, population levels decreased

significantly in the seringal Paraguacu, by about 17%.

Various factors contribute to this increase in population in these four seringais.

Immigration of former and new residents from urban areas as well as rural areas close by,

play a important role in increasing rubber tapper population levels. More significant, the

move of CMER residents from remote areas to more accessible areas such as the four

study seringais is an important factor. It can be argued that rubber tappers who live in

more isolated areas may be more likely to migrate to areas near cities, in part because

they may have lower income levels. In the these four seringais the increase of population

occurs partly because many families left isolated deep forest seringais of the reserve and

settled in more accessible. Table 6-1, which presents data from my own sample in these

seringais, shows that the largest proportion of the household heads (39.4%) have

migrated from other seringais in the reserve. These household heads had frequently

migrated from the more isolated, deep forest seringais. For instance, the seringais

Amapa and Guanabara in Brasileia, despite being the biggest seringais in the reserve and

the ones with the largest rubber production, experienced severe out migration as rubber

tappers faced great economic difficulties due to the decline of the rubber price. With the

falling price of rubber, these families had little to sell. Moreover, there were fewer

marreteiros or traveling middlemen from whom they could buy dry goods and sell their

production, forcing them to move into other areas with more services closer to the cities,

such as the four study seringais.