1 SOCIOECOLOGICAL HETEROGENEITY SHAPES LIVELIHOOD STRATEGIES IN AN AMAZONIAN EXTRACTIVE RESERVE: IMPLICATIONS FOR LONG TERM RESERVE VIABILITY By VIVIAN KARINA ZEIDEMANN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Vivian Karina Zeidemann
3 To my family for the love and suppor t provided along this journey and to Amy Jade who is my inspiration
4 ACKNOWLEDGMENTS Many people contributed to the development of this dissertation and were side by side with me on this long journey. My chair Dr. Karen A Kainer, however, played an important part. She was an exceptional mentor, supporter, and especially a good friend. Her outstanding knowledge, guidance, and feedback regarding the socioecological aspects of natural resource manage ment and Amazon conservat ion were invaluable to the development of this work and my perseverance throughout this journey. All members of my doctoral committee, Dr. Marianne Schmink, Dr. Christina Staudhammer, Dr. Grenville Barnes, and Dr. Christopher McCarty offered their unique contributions to all the important steps that I had to overcome during the process of developing my research. I would like also to thank the Amazon Conservation Leadership Initiative and the School of Natural Resource and Environment for their invaluable f inancial support during my time in Gainesville Thi s doctoral research would not have been possible without the financial support provided for my fieldwork in Riozinho do Anfrsio Extractive Reserve by the PSTC Compton Program, Rufford Small Grant, Botany in Action Program, and USAID ICRAF Linkages Programme. I want to thank all of my family for their great support that allowed me to always believe in my capacity to accomplish this professional achievement. My mother has always offered her emotional and fin ancial support, especially during the final stage of this process. To my brother Eduardo and my sister Patricia, who were always ready to help and support me in so many different ways. To my father, who believed since the beginning of my career that the pl anet Earth would need the hard work of conservationists to overcome the environmental problems that our society is generating.
5 And to my niec e Rebeca and my nephew Samuel who have provided me with their happiness and beauti ful smiles when I most needed the m Support from my friends has been invaluable, especially while being in a foreign country far from my loved ones and having to overcome cultural and language barriers. I am a believer that life without friends is not worth it! The fulfillment that my friends provided is not easy to express in words especially in English words. Since they all know me very well, however, they can understand the difficulties that I have to express my feelings and thank them. First of all, I want to thank Ane Alencar, wh o m I consider by far to be my best friend. I only came to Gainesville and UF through her incentive and support. During the almost 5 years that we spent in UF as colleagues and friends, she was by my side helping me in all the tasks of a PhD student. She wa s always ready to discuss my doctoral research and ideas, and gave me feedback on many important aspects related to Amaz on conservation, and also made me beautiful and colorful maps. To my gator roommates, Marco Lentini, Lucimar Souza, Iran Rodrigues, Boni Ramey, and especially Thaissa Sobreiro thanks for the provision of free housing when I could not afford to pay rent. My roommates provided me the best environment needed to concentrate and work by preparing me tasty dishes, just being at home, and giving me their smiles when I was discouraged. Thanks to the Lab gals, Amy Duchelle, Tita Alvira, Marlene Soriano, Shoana Humphries, Marina Londres, Christie Klimas, Joanna Tucker, Rosa Cossio, Jennife r Arnold, Cara Rockwell, Jamie Cotta, Con stanza Rios, and to Antonio Huelsz for the nice meetings, discussion and moments that we s hared during our academic life.
6 Thanks also to my friends and colleagues of the Library, Paula Pinheiro, Ana Carolina Barbosa de Lima, Carol Felix de Oliveira, and especially to Flavia Leite. I want to thanks also all the other friends that participated of my journey in one way or another: my closer friends Arika Virapongse and Simone Athayde, Denyse Mello, Ricardo Mello, Lucima r Souza Valerio Gomes, Eric Carvalho Raissa Guerra, Bruna Brando Cavalcanti, Rosana Resende, Sam Schramski, Lucas Fortini, Geraldo Silva, Ana Alice Eleutrio, Antonio de la Pea, Matthew Palumbo, Vanessa R. da Silva, Ingrid Marisa Tohver, Patricia Delam onica Sampaio, Ludmila Ribeiro, Leonardo Pacheco, Paulo Brando, Alexander Shenkin, Hillary del Campo, Patricia Mupeta, Bothepha Mosetlhi and Ari Martinez. Thank you very much Mason Mathews for the help with the social network analysis and the productive d iscussions, and Wendy Lyn Bartels for the English corrections and the nice feedback provided on C hapter 4 This research was developed in formal partnership with two Brazilian organizations: Instituto de Pesquisa Ambiental da Amaznia IPAM and Fundao V iver Produzir e Preservar FVPP. I want to thank them for all the support they provided that made my days in Brazil during my fieldwork so much more productive and easy. I also want to thank Galvanda Queiroz Ga lvo, Flvio Barros Bezerra, Os ias Costa San tos, and Marcio Luiz da Silva Souza for their nice company and help during my fieldwork in Riozinho do Anfrsio Extractive Reserve. I am grateful to Ana Paula Souza Paulinha, Antnia Melo, Marcos Rocha and all the staff of FVPP and IPAM based in Altamira for their support. I must also thank Dr. Bruce Forsberg of INPA, Dr. Barbara Zimmerman of The International Conservation Fund of Canada Dr. Stephan Schwartzman of the Environmental Defense Fund and Dr. Jos Maria Cardoso da Silva
7 of the Conservao Inte rnacional for the recommendation letters they provided, which allowed me to obtain enough funding to carry out my fieldwork in a remote place like Riozinho do Anfrsio Extractive Reserve. Finally, I want to thank immensely all of Riozinho do Anfrsio Extractive Reserve residents for their smiles, their help, the good food and shelter that they provided me. I have wonderful memories of the amazing moments that I spent with them. I learned so much with them that it i s difficult to explain with just few words. Riozinho do Anfrsio Extractive Reserve is a place that I will never forget and all the people that I met there will be in m y thoughts and my heart forever!
8 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 12 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 INTRODUCTI ON ................................ ................................ ................................ .... 16 What is the Problem? ................................ ................................ .............................. 18 The Place and Community Involved ................................ ................................ ....... 20 Expected Broader Impacts of this Study ................................ ................................ 22 2 INTERNAL SOCIOECOLOGICAL HETEROGENEITY SHAPES BRAZIL NUT CONTRIBUTION TO HOUSEHOLD INCOME IN AN AMAZONIAN EXTRACTIVE RESERVE ................................ ................................ ....................... 24 Non Timber Forest Products and Conservation ................................ ...................... 24 Riozinho do Anfrsio Extractive Reserve ................................ ................................ 28 Methods ................................ ................................ ................................ .................. 30 Livelihood Survey ................................ ................................ ............................. 30 Structured Interviews ................................ ................................ ........................ 31 Forest Inventories ................................ ................................ ............................. 32 Data Analysis ................................ ................................ ................................ ... 34 Brazil nut stand quality ................................ ................................ ............... 35 Brazil nut tree access and management, household characteristics and income ................................ ................................ ................................ .... 36 Results ................................ ................................ ................................ .................... 36 Brazil Nut Stand Quality ................................ ................................ ................... 36 Dens ity and dbh distribution ................................ ................................ ....... 36 Tree characteristics ................................ ................................ .................... 37 Reported and counted fruit production ................................ ....................... 37 Brazil Nut Tree Access and Management Practices ................................ ........ 38 Households Characteristics and Income ................................ .......................... 39 Discussion ................................ ................................ ................................ .............. 40 Brazil Nut Socioecological Heterogeneity ................................ ......................... 41 Brazil nut density and distribution ................................ .............................. 41 Access to Brazil nut trees varied across the reserve ................................ 44
9 Fruit production discrepancies explained by ecological factors, management, and household size ................................ .......................... 46 Regional Variation Shapes Brazil Nut Income ................................ .................. 51 Heterogeneity Inside the Polygon ................................ ................................ ..... 53 3 LINKING SOCIAL NETWORKS AND INCOME IN A SUSTAINABLE USE RESERVE: IMPLICATIONS FOR CONSERVATION INTEGRITY ......................... 64 Social Networks and Natural Resources Management ................................ ........... 64 Riozinho do Anfrsio Extractive Reserve Context ................................ ................... 68 Methods ................................ ................................ ................................ .................. 70 Data Gathering ................................ ................................ ................................ 70 Round one: int ernal RDAER social network structures .............................. 70 Round two: ties with institutions and other external actors ........................ 71 Livelihood survey ................................ ................................ ....................... 72 Data Analysis ................................ ................................ ................................ ... 74 Results ................................ ................................ ................................ .................... 75 Social Networks Within Reserve Polygon ................................ ......................... 75 Social Networks Linked to External Actors ................................ ....................... 77 Contact of RDAER residents with external actors ................................ ...... 77 Assistance received from external actors ................................ .................. 78 Trust as ranked by reserve residents ................................ ......................... 79 Interactions Outside the Reserve and Household Income ............................... 81 Discussion ................................ ................................ ................................ .............. 83 The Primacy of Geo graphic Location ................................ ............................... 83 Internal social networks explained by household settlement patterns ........ 83 Access to RDAER regions determined social networks linked to external actors ................................ ................................ ........................ 85 Formal Social Positions Mattered ................................ ................................ ..... 87 Trust as Explained by Social Interactions and Formal Social Positions ........... 88 Linking Social Networks and Trust with Household Inc ome ............................. 90 Utility of Social Network Analysis Applied to Sustainable Use Reserves ......... 93 4 HETEROGENEITY INSIDE THE EXTRACTIVE RESERVE POLYGON: IGNORING SOCIOECOLOGICAL VARIATION AND LIVELIHOOD OUTCOMES AT THE PERIL OF LONG TERM RESERVE VIABILITY ................ 104 Sustainable Use Reserves and Socioecological Heterogeneity ............................ 104 The Extractive Reserve Model ................................ ................................ .............. 105 How the Model was Conceptualized ................................ .............................. 105 What Socioeconomic and Ecological Factors Influence Livelihood Portfolios? ................................ ................................ ................................ ... 108 The Changing Livelihood Process Occurring in Extractive Reserves ............. 110 Heterogeneity Inside the Polygon: The Case of Riozinho do Anfrsio Extractive Reserve ................................ ................................ ................................ ............. 112 Implications for Sustainable Use Reserves ................................ ........................... 120 Local Specific Approaches ................................ ................................ ............. 120
10 In stitutional Mechanisms for the Governance of Heterogeneous Polygons ... 124 The utilization plan ................................ ................................ ................... 126 Local agencies ................................ ................................ ......................... 128 Challenges and Benefits of Taking into Account this Socioecological Heterogeneity ................................ ................................ .............................. 131 5 CONCLUSION ................................ ................................ ................................ ...... 138 APPENDIX A LANDHOLDINGS SELECTED FOR FOREST INVENTORIES (RED) AND BRAZIL NUT TREES MAPPED IN TRAILS AND TRASECTS IN 2009 (GREEN) AND 2010 (LIGHT GREEN) (PORTUGUESE) ................................ ..................... 143 B BRAZIL NUT QUE STIONNAIRE (PORTUGUESE) ................................ .............. 144 C SOCIAL NETWORK QUESTIONNAIRE (PORTUGUESE) ................................ .. 152 D GOVERNMENT AND NON GOVERNMENTAL ORGANIZATIONS, AND PRIVATE SECTOR ACTORS ................................ ................................ ............... 158 LIST OF REFERENCES ................................ ................................ ............................. 159 BIOGRAPH ICAL SKETCH ................................ ................................ .......................... 174
11 LIST OF TABLES Table page 2 1 Descriptive statistics of Brazil nut density, dbh, counted and reported fruit production in the three RDAER regions and at the Reserve level. ..................... 61 2 2 Best fit models th at expla in Brazil nut quality in RDAER ................................ .... 62 2 3 Descriptive and inferential statistics of household characteristics forest based in come, income from Brazil nuts, and total household income in RDAER. ................................ ................................ ................................ .............. 63 3 1 Households characteristics, representation on AMORA, participation in meetings and courses, frequently visited cities, degree centrality, and household income. ................................ ................................ ........................... 102 3 2 Descriptive st atistics of average household characteristics, forest based and n on forest based income, and total household income in RDAER ................... 103 4 1 Contribution of different sources of income to household total income in the three RDAER regions. ................................ ................................ ...................... 136 4 2 RDAER internal socioecological heterogeneity. ................................ ............... 137
12 LIST OF FIGURES Figure page 2 1 Map showing the distribution of fam ily households and Brazil nut stands along the three RDAER regions. ................................ ................................ ........ 56 2 2 Diameter class distribution of Brazi l nut trees from transects in RDAER regions. ................................ ................................ ................................ ............... 57 2 3 Percentage of trees by crown form category within each of three RDAER regions. ................................ ................................ ................................ ............... 57 2 4 Distances to Brazil nut stands harvested by region ................................ ............ 58 2 5 Percentage of harvesters performing Brazil nut management practices within each of three RDAER regions ................................ ................................ ............ 59 2 6 Number of return visits to harvest trees within the same Brazil nut season by region. ................................ ................................ ................................ ................ 60 3 1 Social network within the reserve featuring subgroups for borrowing canoe s .... 97 3 2 Social network within the reserve featuring subgroups for borrowing motorboats ................................ ................................ ................................ .......... 97 3 3 RDAER household network of number of visits received from representatives of external organizations ................................ ................................ .................... 98 3 4 RDAER network of households that received assistance from at least two external actors ................................ ................................ ................................ .... 99 3 5 RDAER networks of household s that received assistance from different groups of external actors ................................ ................................ .................. 100 3 6 RDAER networks revealing degr ee of trust that household have with external actors ................................ ................................ ................................ ................ 101
13 LIST OF ABBREVIATION S AMORA Associao dos Moradores do Riozinho do Anfrsio CIFOR Center for International Forestry Research FAO Food and Agriculture Organization FVPP Fundao Viver, Produzir e Preservar GPS Global Position ing System ICMBio Chico Mendes Institute for Biodiversity Conservation (Instituto Chico Mendes de Conservao da Biodiversida de) ISA Instituto Socioambiental IUCN International Union for Conservation of Nature MMA Environmental Ministry (Ministrio do Meio Ambiente) NGO Non Governmental Organization NTFP Non Timber Forest Products PEN Poverty and Environment Network RDAER Riozinho do Anfrsio Extractive Reserve RESEX Extractive Reserve (Reserva Extrativista)
14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doc tor of Philosophy SOCIOECOLOGICAL HETEROGENEITY SHAPES LIVELIHOOD STRATEGIES IN AN AMAZONIAN EXTRACTIVE RESERVE: IMPLICATIONS FOR LONG TERM RESERVE VIABILITY By Vivian Karina Zeidemann May 2012 Chair: Karen A. Kainer Major: Interdisciplinary Ecology Sustainable use reserves have been promoted as a way to reconcile conservation and development, based on the premise that non timber forest products (NTFPs) provide subsistence and cash benefits to reserve residents, while protecting forests. These reserve polygons are highly dynamic and heterogeneous, and the livelihood strategies adopted by reserve residents are diverse and affected by several socioec onomic and ecological factors. This internal ecological socioeconomic and cultural heterogeneity, a s well as the regional context in which these reserves are embedded, play important roles in how reserve residents use and manage their natural resources, and on the livelihood portfolios they adopt Using the Riozinho do Anfrsio Extractive Reserve (RDAER ) in the Brazilian Amazon as an example of a sustainable use reserve, the main aim of this study was to evaluate the internal socioecological variation to determine what factors affect forest based and non forest based income in that reserve. To assess thi s internal socioecological heterogeneity, this study used Brazil nut ( Bertholletia excelsa ) as a model NTFP species, and adopted a perspective that takes into account potential variation in Brazil nut stand quality, access to trees,
15 and ultimately, income generation across RDAER. It also evaluated the internal heterogeneity related to social networks within the reserve and with external actors, and how those social networks link to household income. To assess these variables, a livelihood survey, structured interviews, a social network questionnaire, and forest inventories were employed from 2008 to 2010. Results showed that Brazil nut makes an important contribution to forest based income in RDAER, and that its contribution varies within the reserve mainly because of the internal socioecological heterogeneity (some household characteristics, Brazil nut stand quality, management and tree access). In addition, social networks within the reserve, but especially with external actors, were linked to household inc ome as d etermined mainly by geographic location, formal social positions, and the degree of trust and interactions among reserve residents and external actors. These findings indicate that if Brazil nut and other NTFPs are used as a key focal point to reco ncile conservation and development in RDAER and other sustainable use reserves, then policies and initiatives that aim to regulate and promote NTFPs use need to take into account the ecological and socioeconomic heterogeneity inherent in these forest produ cts. Furthermore, the use of social network analysis proved to be extremely useful to determine other social factors affecting household income and the existence of different groups of stakeholders and their social relations, which are extremely important for the c ollaborative management of inhabited reserves. We conclude that the long term viability of sustainable use reserves depends on the integration of t heir internal socioecological heterogeneity and on the adoption of participatory bottom up approache s that directly address the resultant diverse livelihoods
16 CHAPTER 1 INTRODUCTION This PhD dissertation examines the role of the internal socioecological heterogeneity of sustainable use reserves i n natural resourc e use and management, and in the livelihood strategies adopted in those reserves. By focusing on the socioeconomic and ecological aspects of the use of an important non timber forest product ( Brazil nut, Bertholletia excelsa ) and on the social relations that reserve residents of t he Riozinho do Anfrsio Extractive Reserve ( RDAER) have within the reserve and with external actors, this study aims to identif y and obtain a better understanding of how these internal socioecological dynamics influence livelihood strategies that are inten ded to suit the sustainable use reserve model. This dissertation is presented in three separate manuscripts prepared for publication in academic journals. Thus, each of these manuscripts has its unique aspects and can be read independently. The first manus cript, entitle d socioecological heterogeneity shapes Brazil nut contribution to household income in an its contribution to household income in RDAER. Ecological ( density, distribution, fruit production, and tree characteristics) and socioeconomic ( access to trees, management p ractices adopted, and household characteristics) variables were evaluated to identify which factors influenced forest based and total househo ld income in this extractive reserve. The findings show tha t Brazil nut make s an important contribution to forest based income in this reserve, and that its contribution varies among RDAER regions due to internal socioecological heterogeneity ( mainly Brazi l nut stand quality, management practices and trees access, and some household characteristics).
17 aims to provide insights for understanding the social relations that reserve re sidents have within this polygon, but also with external actors, and how these relations relate to conservation and development initiatives taking place in those reserves. This manuscript demonstrates the potential use of social network analysis to underst and social network structures of sustainable use reserves, and to identify the main actors that could be useful to foster participation and collective action for reserve management. It also proved to be extremely useful in linking household income, mai nly sourced from forest and non forest based activities, with the social relations that reserve resid ents have with external actors. Finally, the third Ignoring socioecological variation and live lihood outcomes at the peril of long term discuss es the importance of socioeconomic and ecological heterogeneity existing in sustainable use reserves, and the livelihood portfolios influenced by this heterogeneity. By using the example o f Riozinho do Anfrsio Extractive Reserve, and the internal socioecological heterogeneity found and discussed in the previous chapters, this study provides insights into the importance of this internal heterogeneity for sustainable use reserve management a nd the long term viabili ty of these conservation units. Each of these manuscripts is a result of the contribution of colleagues from University of Florida as well as from partners of the Eastern Amazon, where I carried out my fieldwork. Each submitted pape r will recognize key contributions of these
18 individuals through co authorship; hence the use of third person plural versus first person in these chapters within the dissertation. What is the Problem ? The high rates of deforestation and biodiversity deterio ration in tropical forests have resulted in a search for forest uses that provide economic incentives for their long term retention (Anderson 1990). Because a pproximately 80% of the population living on forest areas in developing countries is strongly depe ndent on forest resources to pursue their livelihoods based mainly on subsi stence activities, t rade in tropical forest products and particularly non timber forest products (hereafter, NTFPs), has been pursued as a strategy that may allow forest conservation while improving local well being. D evolution of forest rights to rural communities is also expanding due to the struggle of indigenous and traditional people to defend their lands and livelihoods. In parall el, the poor conservation outcomes that followed decades of top down resource management strategies and planned development (Allegretti 1990) have forced policy makers and developers to reconsider the role of forest based communit ies and recognize the impo rtance of customary laws for res ource use and conservation. As a consequence participation of local residents is seen as a fundamental part of sustainable use and management of natural resources (Colfer & Byron 2001) and sustainable use reserves emerged a s a strong conservation strategy to reconcile conservation and development. Sustainable use reserves are government owned conservation units designated for sustainable use and conservation of common pool resources but also individually designated resource s, by traditional communities (Allegretti 1990). The initial assumption behind these types of conservation units is that non timber forest products (NTFP) would sustain a forest economy that will provide better income to extractive
19 communities while ensuri ng forest conservation. However, these reserve polygons present a range of socioeconomic and ecological characteristics and are highly internally dynamic (Ankersen & Barnes 2004). Indeed, t hey are embedded in different regional contexts, which also impl y that they are subject to a variety of external forces. The livelih ood strategies adopted by resident communities are influenced by a diversity of different factors, such as natural resources access and quality, cultural, historical and institutional aspect s, labor availability, education level s, and market integration (Peralta & Kainer 2008; Pyhl et al. 2006; Shone & Caviglia Harris 2006 ), but also historical, socioeconomic and ecological contexts in which these communities are embedded. Indeed, recent st udies have shown a tendency of transformation of the se traditional livelihood strategies ( Campbell 1996; Castelo 1999; Ehringhaus 2 006; Gomes 2001; Stone 2003; Vadjunec 2009; Wallace 2004) which are in part a response to those external forces. Even though these studies have show n the occurrence of livelihood changes in extractive reserves only a few have analyzed the socioecological internal dynamics of these reserves, and how they influence the adoption of livelihood portfolios that can contribute to the conservation and development of these polygons (Brown & Rosendo 2000; Gomes 2 009; Salisbury & Schmink 2007). In this context it is essential to incorporate the internal socioecological dynamic s of sustainable use reserves and to understand how this internal heterogeneity influences the divers e livelihood portfolios. Forest based and non forest based household income is determined by those factors and the lack of integration of this knowledge wh en implementing and managing these reserves can compromise their final goals, which are to conserve the natural resource uses while protecting the
20 traditional way of life of the communities dependent on those resources and improving their livelihood s in a way that is compatible with these goals. The Place and Community Involved The research underlying this dissertation was carried out in Riozinho do Anfrsio Extractive Reserve, located in Par state, in the Eastern Amazon of Brazil This Reserve was create d by a presidential decree in 2004, mainly as a result of a very well articulated alliance between social and environmental movements in response to a process of land speculation and violence in the region (Campos & Nepstad 2006). This extractive reserve p rotects 736,430 ha of mature, old growth forest and hosts around 26 hous eholds (ISA 2006). These forest dwelling families are dispersed, isolated, and without access to health care or education. They live by subsistence agriculture, hunting, fishing and c ollection of forest products (Rocha et al. 2005). To sell their production, RDAER families depend on itinerant traders, or regates who often manipulate prices and foster dependency, but also being historically important as providers of information and ma rket access. Contemporary members of these families are descendants of migrants from the Brazilian Northeast who arrived in the region to tap rubber in the late nineteenth and early twentieth centuries. At that time, approximately 500,000 people migrated t o the Amazon region as a result of a rapid process of occupation of this region in response to increased demand for rubber by industrialized nations (Allegretti 1990). The collection and processing of rubber were carried out by these migrant families, who were inserted from the start in a highly regressive system of debt peonage. Despite numerous financial crises, rubber tapping continued in the region until the middle of twentieth century, when this activity was replaced by jaguar skin commercialization (1 975 1997).
21 Even though RDAER families viewed jaguar skin commercialization as one of the more profitable activities, it still was carried out in the same debt peonage system that dominated the rubber tapping activity (Rocha et al. 2005). In addition to the predominance of these two activities, Brazil nut, andiroba ( Carapa guianensis ) and copaiba ( Copaifera langsdorffii ) harvests were conducted and continue to be important natural resources for the RDAER family outcomes in this extractive reserve (Rocha et al. 2005). RDAER can be divided into three ecological and socioeconomic regions, which are self recognized by RDAER residents: the Upper, Middle and Lower Riozinho regions. These three regions present different historical socioeconomic dynamics, mainly as a result of differences in seasonal river access; during the summer season, the Upper region and some of the Middle region residents become completely isolated. In the past, these families received an advanced payment (food and supplies) from rubber barons at the beginning of the summer season to survive during this hard time. D uring the winter season these residents used their Brazil nut production to pay for the supplies received in advance. Recently, this RDAER internal division was highlighted by a co nflict generated when residents of the Upper region opposed the Reserve creation while Lower region residents supported it The conflict was generated mainly because of the different perspectives of residents and their contact with external actors. The Up per residents, at the time of the Reserve creation, were in frequent contact with ranchers, loggers and land grabbers that entered the Reserve, coming mainly from the cities of Trairo and So Flix do Xingu through the Western borders of the Reserve (Roch a et al. 2005). On
22 the other hand, during the same period residents of the Lower region were contacted by non governmental organizations and government agencies that offered the possibility of extractive reserve creation in an attempt to stop the deforest ation promoted by these land grabbers, loggers, and ranchers, provid ing land security to Riozinho residents (Rocha et al. 2005). My preliminary research clearly revealed that Brazil nut is currently the most important source of income of RDAER residents, m ainly in the Middle and Lower region of the Reserve. These families have collected Brazil nuts traditionally for decades and have tremendous knowledge regarding this natural resource. Less important NTFPs, but also contributing to household income in RDAER are copaiba, andiroba and fish. Recently, manioc fl our and other agriculture products have become an important complement for famil y income s especially in the Upper region where families are less engaged with Brazil nut harvest s Expected Broader Impac ts of this Study This study responds to the need for a comprehensive analysis of th e overall understanding of the socioecological factors that influence the diverse livelihood portfolios found within sustainable use reserves It was argued that the social relations that reserve residents have within the reserve, and with external actors, need to be integrated with the ph ysical, biological, geographic and social variation taking place within these polygons. This heterogeneity has an important role i n the adoption of livelihood strategies that could potentially meet conservation and development goals. In the global context, sustainable use reserves are touted as an ideal model of forest conservation for t ropical regions worldwide Thus, this research will be relevant to biodiversity conservation and community based developm ent initiatives currently
23 under way across the tropics. Information presented herein will also contribute to a wide range of regional and national needs, and natural resources use and mana gement programs that aim to foster sustainable forest livelihoods and biodiversity conservation. At a community level, this research will provide the baseline information needed to build a management plan for RDAER, and also specifically for Brazil nut, wh ich is an important resource supporting long term sustainable use in this reserve The results of this research should be equally informative for grassroots organizations and government institutions responsible for forest conservation and development initi atives as well as other related approaches in Amazon region.
24 CHAPTER 2 INTERNAL SOCIOECOLOGICAL HETEROGENEITY SHAPES BRAZIL NUT CONTRIBUTION TO HOUSEHOLD INCOME IN AN AMAZONIAN EXTRACTIVE RESERVE Non Timber Fores t Products and Conservation Approxim ately 80% of the population from developing countries depends primarily on forest products, mainly non timber forest products (NTFPs), for income, food, and medicine (FAO 1995 ; Shackleton & Shackleton 20 04 ; Ahenkan & Boon 2008) Many scholars have advocated NTFP extraction and trade as an important tool for supporting traditional forest live lihoods and sustainable forest management (Belcher et al. 2005 ; Ahenkan & Boon 2008 and 2010). Th e widely held assumption that NTFP extraction is typically less ecologically destructive than timber extraction helped to strengthen awareness of the environmental benefits of this forest use strategy (Neumann & Hirsch 2000) Nonetheless, there remain conflicting conclusions about whether NTFPs promote conservation as well as improve livelihoo ds of forest peoples (Salafsky & Wollenberg 2000 ; Arnold & Ruiz Prez 2001 ; Kusters et al. 2006) This ambiguity stems from the ecological, socioeconomic and cultural complexity of NTFP use. NTFP uses vary widely, even at a small scale, because of the socioeconomic, demographic and geographic variability that exists within communities (Coomes & Burt 2001 ; Mahapatra et al. 2005 ; Belcher et al. 2005; Newton 2011) This community heterogeneity is matched by that of the NTFPs available for harvest, including types of NTFPs exploited, intensity and frequency of harvest, and relative contribution of those products to use r welfare (Cavendish 2000 ; Janse & Ottitsch 2005) Furthermore, Ahenkan and Boon (2010) argued that NTFP policy failures can often be traced back to a lack of adequate and detailed species and product information, and suggested that more detailed
25 information on any given NTFPs is needed if it is to be used as an effective tool to reconcile conservation and livelihood improvements. However, a few studies have integrated resource heterogeneity, such as qu ality and availability, with the socioeconomic aspects related to them (although see Takasaki et al. 2010 ; Newton 2011). In the Brazilian Amazon the promise of NTFPs to reconcile conservation and sustainable development is exemplified in the extractive reserve model. Extractive reserves are government owned conservation units designated for sustainable use and conservation o f common pool resources by traditional communities (Allegretti 1990) The central assumption behind extractive reserves is that the traditional management r egimes of reserve residents are sustainable and thus the standing forests on which they depend for their livelihoods will also persist in the long term (Allegretti 1990; Hall 1997) The policy goal of the extractive reserve model that poverty alleviation can be reconciled with resource conservation is attractive, and has resonated glo bally, even precipitating a new World Conservation Union (IUCN) protected area designation of Managed Resource Protected Areas. Three decades later in the Brazilian Amazon alone, there are 44 federal Extractive Reserves (RESEX), 19 Sustainable Development Reserves, conceptually akin to RESEX, and 26 state Extractive Reserves, comprising approximately 25 million ha of forests (ICMBio 2011). L ike other conservation areas, extractive reserves comprise social ly and ecological ly complex systems where a large num ber of relevant variables and their interactions influence how residents operate at multiple levels. While federal and/or state designation imparts clear definition of the external reserve boundary, internally
26 these spaces present a range of ecological and socioeconomic characteristics that are highly dynamic (Ankersen & Barnes 2004) As reserves mature over time, variation in quality, access and management of natural resources, and differences in household characteristics inside the reserve polygon, determine whether residents continue to embrace livelihood strategies that foment conserva tion or whether they pursue less sustainable economic opportunities. Because of its basin wide distribution, its significance in global markets, and its potential for sustainable use and forest conservation, Brazil nut ( Bertholletia excelsa ) has played a k ey role in the designation of many Brazilian extractive reserves (Campbell 1996 ; Assies 1997) Additionally, Brazil nut is one of the few economically important NTFPs under the control of poor people, and has become the principal resource that prevents marginalized community members from becoming even more impoverished (Collinson et al. 2000 ; Escobal & Aldana 2003 ; Stoian 2005) However, even with the massive international and domestic investments applied to these conservation u nits and to the promotion of forest products such as Brazil nut, resident communities still face poverty and recent studies have shown a tendency for these traditional livelihood strategies to transform to alternatives that can compromise conservation goa ls (Wallace 2004 ; Salisbury & Schmink 2007 ; Gomes 2 009) Several factors can be enumerated as drivers of these changes, but certainly one is the lack of integration of the well adapted and traditional management strategies with policies and initiatives that aim to promote the use and trade of forest pro ducts (Schmink & Wood 1987 ; Coomes & Burt 2001). In addition, there is a tendency to invest solely in what is considered the most important forest products, without taking into account the likely
27 variation in distribution and quality of these species across the reserve or the accompanying variation in how reserve residents might invest and depend on these products. In this study we aim to understand the role of the internal socioecological dynamics of Brazil nut identified as a key forest product to achieve conservation and development goals in Riozinho do Anfrsio Extractive Reserve (RDAER) in Par, Brazil (Rocha et al. 2005) The known existence of Brazil nut across the reserve coupled with the joint conservation and livelihood promise of this NTFP generated great enthusiasm, and was one of the main aspects that determined selection of the extractive reserve model for this particular geographic area, ultimately resulting in the creation of RDAER in 2004. Thus, our main question is to what extent do the benefits of this particular NTFP extend to all residents within this reserve? More specifically, we ask: How does Brazil nut stand quality [densities, distribution by diameter at breast height (dbh), tree characteristics, counted and reported fruit production] vary within RDAER? How do Brazil nut tree access and management practices vary within RDAER? and, How does this Brazil nut variation and differences in household characteristics affect forest based household income derived from Brazil n uts? The strong emphasis on Brazil nut as a potential forest product that could provide enough income to reserve residents while meeting reserve conservation goals has not been adequately evaluated to date. Our study seeks to fill that gap by adopting a pe rspective that takes into account potential variation in NTFP quality, access and ultimately income generation across the reserve. Our work also provides insights into the importance of integrating internal socioecological heterogeneity within conservation
28 units, where external boundaries are well defined, but internal spaces are not as well understood. Riozinho do Anfrsio Extractive Reserve RDAER is located in the municipality of Altamira, Par state, in the Eastern 3.98"S) (Figure 2 1). It was created by presidential decree in 2004, mainly as a result of a very well articulated alliance between social and environmental movements in response to a process of land speculation and violence in the region (Campos & Nepstad 2006) The reserve encompasses 736,430 ha of m ature old growth forest and hosts around 26 households or approximately 300 people (ISA 2006) The RDAER region is characteriz ed by a hot and humid climate, with an annual mean temperature of 27C, and annual mean rainfall of 1885 mm (Silva 2007) The forest dwelling families are dispersed, isolated, and without access to health care or education. They live by subsistence agriculture, hunting, fishing and collection of forest products (Rocha et al. 2005) To sell their production, RDAER families depend on iti nerant traders, or regates who often manipulate prices and foster dependency, but are also historically important providers of information and market access. Contemporary members of these families are descendants of migrants from the Brazilian Northeast who arrived in the region to tap rubber in the late nineteenth and early twentieth centuries. The collection and processing of rubber were carried out by these migrant families, who were inserted from the start in a highly regressive system of debt peonag e. Even though Brazil nut was economically less important than rubber for decades, resident families have traditionally harvested it since their arrival, and with the decline of rubber production in the middle of twentieth century, Brazil nut became the
29 mo st important forest product to RDAER families. Other less important NTFPs that may also contribute to household income inclu de copai ba ( Copaifera spp .), andiroba ( Carapa guianensis ), fish, and game. While some NTFPs have a broad distribution across the ent ire reserve, others are more concentrated in specific regions. Nowadays RDAER families still rely on itinerant traders to sell their Brazil nuts and other forest products mainly because of the remoteness of RDAER and thus high costs of transportation. The Brazil nut value chain in that region is controlled by one family from Belm, the capital city of Par, which buys the entire Brazil nut production from Altamira sellers and also subsidizes the itinerant traders operating in RDAER. Thus differences of Bra zil nut income among RDAER families are not a consequence of price variation, but rather of Brazil nut harvest intensity. RDAER is commonly divided into three ecological and socioeconomic regions, which are self recognized by RDAER residents: the Upper, Mi ddle and Lower Riozinho regions (Figure 2 1). Each of these three regions presents unique dynamics due to distribution of natural resources, differences in seasonal river access, and access to the surrounding cities. During the dry season, the residents in the Upper and parts of the Middle region become completely isolated, without access to Altamira because of the lower level of water in the Riozinho do Anfrsio river (Figure 2 1). Even though Upper river residents have some access to Itaituba (the closest city of RDAER Figure 2 1), the use of this access is extremely limited since they have to walk for several days along an old logging road to reach the city. This isolation was even recognized during the rubber era, when these families received an advanc e of food and supplies from rubber barons to survive the dry season.
30 Methods To evaluate the Brazil nut socioeconomic and ecological variation within RDAER we employed a livelihood survey, structured interviews, and forest inventories from 2008 to 2010. T he livelihood survey, applied to 2 3 of 26 total households (85%), provided information on household socioeconomic characteristics, total household income, and the Brazil nut contribution to total income. Structured interviews of 24 members of these same households were used to determine Bra zil nut management practices and household access to Brazil nut trees. In addition, a subset of 6 households was randomly selected for forest inventories to assess Brazil nut stand quality (i.e., densities, distribution by diameter at breast height (dbh), tree characteristics, counted and reported fruit production). Livelihood Survey A livelihood survey adapted from the Center for International Forestry Research Poverty and Environment Network ( CIFOR PEN; http://www.cifor.org/pen/research tools/the pen prototype questionnaire.html ) was administered to the head (typically male) of 23 of the total 26 households in the entire Reserve, and when possible, discussions were conducted with participation of other household members. We recorded general household characteristics such as time of residence, household size, age and schooling of the head of household, and calculated a household dependency ratio ( household consumers: household producers ). Total household income was quantified as all forest and agricultural products that were sold or traded plus wage labor, social benefits from government, and other minor sources of external economic support (e.g. mining). Forest produ cts included both unprocessed forest products (mainly Brazil nuts, copaba, andiroba, rubber, honey, and fish) and processed ones
31 (baskets, brooms, canoes, and paddles). Data on the amount of game consumed was not collect ed since most of the interviewe e s c ould not provide accurate information on it. Hunting is an activity that is performed by all RDAER residents and there are no restrictions on game consumption in RDAER (no published data). Agricultural products quantified were mainly manioc, but also chick ens. With this detailed information we calculated total household income and the proportion of it sourced from Brazil nuts. Income is presented in Brazilian Reais (1 U.S. dollar 1.5 Brazilian reais) and refers to the year of 2009. Structured Interviews W ithi n the 23 households, structured interviews were applied to 24 adult members of these household s who harvest Brazil nuts to gather information on access to Brazil nut trees and practices used to manage individual trees and stands. Tree access was define d as the walking time (minutes) to reach the closest and the second closest Brazil nut stands that were most frequently harvested. To understand stand and tree management, we selected eight Brazil nut management practices frequently adopted by other Amazon ian communities (Kainer et al. 2007; Duchelle 2009) : (1) clearing of Brazil nut trails, (2) use of fire underneath individ ual trees to facilitate harvest, (3) Brazil nut enrichment plantings, (4) clearing around seedlings and saplings, (5) purposefully protecting seedlings a nd saplings, (6) liana cutting, (7) slashing the outer bark of mature exuded when slashed), and (8) washing nuts after harvest. Using this list, we asked each househol d member if they practiced these activities in their stands. Finally, we
32 also asked harvesters the number of times they returned to their trees within the same Brazil nut season. Forest Inventories Forest inventories were conducted on landholdings of a subset of six RDAER households, using an initial regional stratification to ensure representation by households across the length of the Riozinho do Anfrsio river (two in the Lower, three in the Middl e, and one in the Upper) ( Appendix A ) (diameter at breast height) in the m ost frequently harvested stands by each of these households were geo referenced ( latitude, longitude, and altitude) u sing a Garmin GPSmap 60CSx unit Dbh was measured using a steel diameter tape, and fruit production as reported by the household head (average amount of Brazil nuts harvested per year in terms of number of boxes: 1 box 22 kilos) was also collected for each tree mapped. To collect more detailed data on Brazil nut population structure, tree characteristics, and fruit production by household, we sampled t he active stands described above where Brazil nuts were currently collected as well as nearby forest areas where harvesting was not cond ucted. To sample active stands, we randomly selected one Brazil nut trail previously mapped within each of the six households, and cutoff at which Brazil nut trees are as sumed to be reproductively mature (Kainer et al. 2007) In addition, from each of these stands, one transect (2500m x 40m = 10 ha) was done to sample a typically unharvested forest ( Appendix A ) Start point and direction of these unharvested transects were randomly selected by obtaining GPS points every 50
33 m when walking the active Brazil nut trail, randomly selecting one of these GPS points, and then randomly selecting one of eight possible geographical directions (N, NE, E, SE, S, SW, W, NW, respectively) for the transect. After start point and direction were defined, we checked to ensure there was no overlap with active Brazil nut trails. If overlap did exist, then the opposite direction was selected, and if the opposite direction still overlapped an active stand, then a c ompletely new direction was randomly selected. measured. Brazil nut tree characteristics (liana loads, crown form and crown position) and fruit production data were collected transects, following the same methods described in Kainer et al. (2007). Thus, liana 75% crown covered, and (4) > 75% crow n covered. Crown form categories were defined using a method adapted from Synnot 1979 : (1) complete or irregular circle (perfect or good), (2) half crown (tolerable), (3) less than half crown (poor), and (4) one or a few branches (very poor). Finally, crown position also was scored using four categories: (1) dominant (full overhead and side light), (2) co dominant (only full overhead light), (3) intermediate (some overhead or side light), and (4) suppressed (not receiving direct light). The opinion of at least 3 people went into determination of these characteristics: the main resear cher, a field assistant, and the local Brazil nut harvester. unharvested transects (10 ha transects sampled from each of six stands). This resulted in a total sampled area wit hin the Reserve of 60 ha.
34 Finally, fruit production was measured in two ways: counted fruit production and reported fruit production. For the former, number of fruits found below the tree crown area was counted for each mature Brazil nut tree sampled in th e trails and transect. Fruit counting was performed at the end of February, after fruit fall and before the Brazil nut harvest began. During the period of fruit counting a few previously selected Brazil nut trees were replaced by other trees located on the same trail since some fruit fall areas overlapped. Data Analysis To understand the Brazil nut internal socioecological heterogeneity among RDAER regions, we initially generated descriptive statistics for (1) Brazil nut stand quality, (2) access to Brazil nut stands and management practices performed, and (3) household characteristics, income from Brazil nuts, forest based income, and total household income. Brazil nut stand quality in this study was defined using measured Brazil nut biological data (densit y, tree dbh, liana loads, crown form, crown position and counted fruit production) and reported production. for all pairs of measured variables and performed several statistical analyses to evaluate how different variables explained Brazil nut stand quality, access to Brazil nut trees, management practices, household characteristics, and household income derived from Brazil nut heterogeneity within RDAER. Considering t hat the fore st inventories were performed on a subset of households, and that structured interviews and livelihood surveys were conducted on approximately 85% of RDAER households, we estimated different statistical models to explain Brazil nut stand qualit y, tree access and management as well as household characteristics and income variation. These diverse
35 models were tested using various independent (predictor) variables and combinations thereof, and residual analyses were used to check model validity. All data analyses were performed with SAS statistical software, version 9.2 (SAS 2004) with significance defined at p <0.05. Brazil nut stand quality Models of Brazil nut stand quality (density, tree dbh, liana loads, crown form, crown position, counted and reported fruit production) were estimated from the subset of households in which we collected forest inventory data, and models were fitted to test RDAER internal variability of Brazil nut stand quality with three predictor variables. First, instead of using the originally designed regional stratification of RDAER, we used distance from the river mouth to each inventoried Brazil nut tree encountered in both trails and transects along the length of the Riozinho do Anfrsio river. Thus, a relative to the river mouth. The other predictor variables incorporated in our Brazil nut was lo cated on an existing collection trail versus a random transect, and tree altitude (m above sea level). We estimated General Linear Mixed models to predict Brazil nut density, dbh, liana load, crown form, crown position, counted and reported fruit productio n with the SAS procedure PROC MIXED. We used the General Linear Mixed model to account for possible correlation among trees sampled within the same trails. Reported and counted Brazil nut fruit production variables were also square root transformed to ensu re that statistical assumptions were met.
36 Brazil nut tree access and management, household characteristics and income In contrast, data for Brazil nut tree access, management practices, household characteristics, income from Brazil nuts, forest based incom e, and the total household income were collected in 2 3 of 26 households along the entire length of the Reserve. For these variables we considered the original regional stratification of RDAER and ated General Linear models, variation regarding Brazil nut management practices we performed a Chi Square test using the SAS procedure PROC FREQ and determined if the distri bution of households that frequently performed the management practices described above differed by region Results Brazil Nut Stand Quality Density and dbh distribution Ninety in ventoried, resulting in a Reserve level density of 1.58 individuals per hectare Brazil nut d ensity increased steadily from the Lower to the Upper region (Table 2 1). Distribution of transect trees by dbh categories showed some variation by region (Figure 2 2); however, when considering both transect and trail trees (1276 trees total), dbh was not explained by tree distance (p = 0.2064, Table 2 2), meaning that dbh was not significantly related to distance of any given Brazil nut tree to the mouth of the Ri ozinho do Anfrsio river. The mixed model for d bh as a function of X, Y, and Z indicated that dbh var ied significantly by tree location type (p < 0.0001, Table 2 2); trees located on trails had greater diameters than those found in transects (Table 2 1).
37 Tree characteristics Two hundred and forty trails and transects, and used to study tree characteristics in more detail. Liana loads were significantly different by tree location type (p < 0.01) (Table 2 2), such that trees on trails, frequented by residents who seasonally collect Brazil nuts, had fewer lianas than those located on rarely visited transects. The model also detected an interaction effect between tree altitude and location (p < 0.01) (Table 2 2). Brazil nut trees on trails tended to average slightly lower altitudes ( x = 201 m 3 m ) than those on transects ( x = 222m 3 m ), a nd trail trees at the lowest altitudes tended to present crowns with limited liana coverage whil e crowns of transect trees at the lowest altitudes tended to present greater liana coverage. Crown form was significantly different by tree altitude (p < 0.05) and tree distance (p < 0.01) (Table 2 2), such that the further trees were located from the mout h of the Riozinho do Anfrsio river, the poorer their crown forms. Translating this finding to regions, trees in the Lower and Middle regions tended to have better crown forms (97% categories combined respectively) th an those in the Upper region (76%) (Figure 2 3). Finally, more than 95% of the Brazil nut trees inventoried had dominant crown positions, and therefore, no statistical differences were observed for this variable. Reported and counted fruit production Fruit production was reported by harvesters (in number of boxes) for 956 trees out of the 1276 mapped Brazil nut trees. Descriptive statistics of the 2009 harvest showed that households from the Lower and Middle regions harvested roughly 2 to 3 times
38 more Bra zil nuts (662 and 731 boxes per region respectively ) than households from the Upper region (242 boxes) (Table 2 1). Our best fit model, however, detect a clear relationship between reported production and distance of tree from the mouth of the Riozinho do Anfrsio river (p = 0.1112) (Table 2 2). Counted fruit production, based on a smaller sample of 128 reproductively mature trees in 6 trails and 6 transects, showed a slight tendency to be explained by tre e distance (p = 0.0952) (Table 2 2). Descriptive statistics again showed that Upper region trees produced fewer fruits (Table 2 1); however, this value was similar to counted fruits of the Middle region, both of which were over three times lower than total counted fruits in the Lower region (Table 2 1). A comparison of the reported and counted fruit production reveals the inconsistency that even though counted fruit production was more than 3 times less in the Middle region, those residents reported harvest ing a slightly larger number of boxes than the Lower region and 3 times more Brazil nut boxes than residents of the Upper region (Table 2 1). Finally, tree location type was a significant predictor of counted fruit production in RDAER (p < 0.05) (Table 2 2 ) with mean counted fruits higher on trails than in transects across all regions (Table 2 1). Brazil Nut Tree Access and Management Practices Although not statistically significant, households in the Lower and Middle regions may have better access to Brazi l nut trees. On average, the closest Brazil nut stands for residents in these regions were less than one and two hours away respectively, while those in the Upper region walked an average of almost five hours to reach their closest Brazil nut stands (Figur e 2 4 A and Table 2 3 ). Indeed, Upper region residents also lived
39 almost 3 times further from their second closest Brazil nut stands than residents in the other two regions (Figure 2 4 B and Table 2 3 ) Among the Brazil nut management practices studied, only liana cutting was significantly different by region (p > 0.01) (Figure 2 5 ). While almost half in the Lower region and six of seven Middle region residents performed this task, none of the Brazil nut harvesters in the Upper region cut lianas (Figure 2 5 ). Overall, the most frequently performed management practices by RDAER residents were clearing of Brazil nut trails and washing of harvested nuts, the latter required by local buyers. Harvesters s, they did not perform this practice. Finally, the number of return visits to harvest Brazil nuts in the same season was statistically significant by region (p < 0.05) (Figure 2 6). While households from the Middle region returned to harvest their trees an average of 4.4 times, those in the Lower region returned an average of 2.8 times. In contrast, households from the Upper region returned to harvested trees only 0.8 times on average within the same Brazil nut season ( Table 2 3). Households Characteristi cs and Income Of all characteristics studied to understand household variation by region, only household size was significantly different by region (p < 0.0001); results showed a decrease in average household size from the Lower to the Upper region (Table 2 3). Although not significantly different, there was a tendency for heads of households in the Upper region to be younger, have attained higher levels of schooling, and have a higher household dependency ratio, meaning a higher number of consumers to prod ucers in the household (Table 2 3). Finally, average residence time in the Reserve was similar
40 among regions ( x = 38 13 years); however, when examining how long families had lived at their current landholding, those in the Upper region had sp ent far fewer years in their current residence than families of the Lower and Middle regions (Table 2 3). Total household income was not statistically d ifferent by region (p = 0.8623); however some tendencies in NTFP income were detected, with a trend for families to increasingly invest more in processed forest products the closer they were to the mouth of the Riozinho do Anfrsio river (Table 2 3). Middle region households tended to have a higher proportion of their income from forest resources than those in the Lower and Upper regions (p= 0.1064), and though not statistically different (p = 0.2271), the contribution of Brazil nut to household income in this same region was almost three times higher (23%) than in the Upper region (9%) (Table 2 3). Finally, when considering only forest based income, Brazil nut contribution to forest based income was signif icantly different by region (p = 0.05 24 ); decreasing as one gets further away from the mouth of the Riozinho do Anfrsio river (Table 2 3). Descriptive stat istics showed that the contribution of processed forest products to total household income in RDAER was only 7 to 16% across the three regions. Discussion Our findings confirm that Brazil nut makes an important contribution to forest based income in Riozin ho do Anfrsio Extractive Reserve in the Brazilian Amazon, and that its contribution varies among RDAER regions due to internal socioecological heterogeneity. This information can serve to guide initiatives that aim to improve Brazil nut household income, and to construct a Brazil nut managemen t plan that guarantees the long term sustainability of this important species in RDAER.
41 More broadly, our findings bring to light the importance of carefully analyzing the ecological and socioeconomic factors that det ermine use and management of natural resources by extractive reserve residents, especially the key NTFPs resident families rely on for their livelihoods. The extractive reserve model has been created to conserve a wide range of natural ecosystems, from the Amazon rainforest to mangroves and coastal areas. Even though reserve residents are apparently similar in the way that they use and manage their natural resources across these areas, their economic portfolio varies significantly according to their socioec ological context. The internal heterogeneity of extractive reserves defined at a fine scale, such as the spatial distribution of natural resources and household structure, has a strong influence on resident engagement with different economic activities. To be a successful conservation unit, reserve implementation and management must take into account this internal socioecological heterogeneity to incorporate the locally specific and often diverse economic repertoire at hand, and thus, guarantee reserve viab ility in the long run. Brazil Nut Socioecological Heterogeneity Brazil nut density and distribution Using random transects we found a density of Brazil nut trees in RDAER of 1.58 trees ha 1 which is similar to densities found in other Brazilian Amazon reg ions (Campbell et al. 1986 ; Peres & Baider 1997 ; Wadt et al. 2005) Densities among RDAER regions varied from 0.9 to 2.7 trees ha 1 (Table 2 1) and can be attributed to methodological issues, such as the difference in the number of transects sampled in each RDAER region, or simply loca l variation of tree densities in this region. Peres and Baider (2003) also found local variation in Brazil nut densities in their study area at the Kayap Indigenous Land, which is located in the same general region as RDAER, and
42 attributed it to the prese nce of Brazil nut groves. Even though we did not analyze the spatial distribution of Brazil nut to assess a random or clustered distribution, many ls). Results of dbh distribution at the reserve level (Figure 2 2) are similar to distributions found in other Amazon regions, with the highest presence of trees with 100 (Salomo 1991 and 2009) ; Zuidema & Boot 2002 ; Peres et al. 2003) However, average dbh in RDAER ( x = 120.7 55.6 cm, Table 2 1) was higher than average dia meters found in other extractive reserves or indigenous lands (Peres & Baider 1997; Wadt et al. 2005), but similar to some populations in Peres et al. (2003). These authors argued that persistently harvested populations present an average dbh > 100 cm, and that Brazil nut populations with a higher average dbh and a low frequency of juvenile trees, such as in RDAER are a result of overexploitation, and serve as an alert for the potential failure of recruitment in these pop ulations. We do not believe, howeve r, that contemporary overexploitation necessarily explains our dbh distributions, but rather a possible combination of past land use patterns and natural disturbances. Natural Brazil nut regeneration depends on canopy openings such as forest gaps and large scale blowdowns (Baider & Peres 1997); shifting cultivation also has been documented to play an important role on Brazil nut recruitment (Viana et al. 1998 ; Cotta et al. 2008) Dbh distributions, such as in RDAER, with a prevalence of large diameter trees, can be a consequence of an intense past human occupation followed by a drastic depop ulation, resulting in an absence of shifting cultivation activity (Scoles & Gribel 2011) and/or to peaks of seedling recruitment that
43 follow individual treefalls and large scale blowdowns (Baider & Peres 1997). Thus, past land use patterns and natural disturbances can foment Brazil nut re generation, and under the absence of such events, Brazil nut populations may present dbh distributions that reflect a paucity or absence of young trees. The human population of RDAER region was much higher in the past than today, with elderly residents est imating that more than 350 rubber tappers and families lived in this area during the rubber boom. [Orbignya speciosa (Mart.) Barb. Rodr.] both indicators of the prese nce of pre Colombian societies (Pires & Prance 1985 ; MMA 2003) Finally, RDAER r esidents reported that the occurrence of blowdowns is not a rare event, with some stating that they had witnessed this type of event in their landholdings. Past occurrence of seedling peaks, theoretically from blowdowns or past shifting cultivation activit y, could also explain variation in Brazil nut densities found among RDAER regions and presence of necessary to test these explanatory hypotheses of population dynamics A gradual decline in Brazil nut production was an expressed concern by harvesters in RDAER. In this extractive reserve as with most others, Brazil nut stands are inherited and resident families have continued to harvest Brazil nuts almost exclusively in those trails worked by their ancestors. As a result, it is possible that current residents are concentrating their harvests on trees that are relatively old and tending toward senescence, rather than peak production. In a Western Amazonian population, Kai ner et al. (2007) reported that Brazil nut fruit production increases with diameter until it peaks somewhere between 100 and 150 cm dbh, subsequently
44 declining with increased diameters in a quadratic manner. In our study, trees in harvested areas (trails = x = 154.9 41.6 cm dbh) presented statistically greater diameters than trees in unharvested areas (transects = x = 120.7 55.6 cm dbh) (Table 2 1). We suggest investments by residents and possibly incentives by government and NGO s to search for new reproductively mature trees to include in routine annual harvests, coupled with the opening of new trails that have gone unharvested in recent years. Access to Brazil nut trees varied across the reserve Access to natural resources is an important factor driving resource use and management. Forest dwellers that have better access to natural resources tend to invest and rely more on those resources (Pattanayak & Sills 2001 ; Duchelle et al. 2011) Mahapatra et al. (2005) studying NTFP sales on rural households in India found that the proximity of available non timber products in nearby forests was o ne of the most important factors determining NTFP income. In addition, results from a second study in a Brazil ian extractive reserve showed that even though all households ha d access to the overall natural resources to sustain their livelihood s access to economically important NTFPs was an important component driving resource harvest for ca sh income (Newton 2011). Our results showed that residents of the Upper region are much farther from their Brazil nut trees than residents of the other 2 regions (Figure 2 4 ) which could explain in part the smaller amount of Brazil nuts harvested in 2009 and the income secured from it by those households (Table 3 3) Upper residents have to travel by canoe about 5 hours to get to their Brazil nut stands T o harvest their Brazil nut trees they temporarily move from their landholdings to camp near their Brazil nut stands leav ing their crops and
45 small domestic animals behind. Thus they only spend minimal time in these stands, returning as quickly as possible to their land holdings to avoid crop and other losses. In contrast residents of the Lower and Middle regions reach their Brazil nut stands in less than 2 hours. Our results also indicate that access to Brazil nut trees has a positive impact on the number of return visits to harvest Brazil nut trees within the same Brazil nut season, and subsequently on Brazil nut income. Residents with good access, particularly those in the Middle region, returned to their trees much more tha n residents of other regions (Figure 2 6), invested more in management of those trees (largely through liana cutting) (Figure 2 5 ), and tended to have a greater proportion of their income from Brazil nuts (Table 3 3). However, even though our results showe d a significant regional variation in access to Brazil nut trees and number of return visits we d id not find a statistically significant correlation between these two variables, indicating that other factors also influence number of return visits to harve st trees. Indeed, Pattanayak and Sills ( 2001) using number of trips as an indicator of effort invested on forest products showed that older household heads and those with a longer time of residence generally took more trips to the forest, which they argued was related to the accumulated knowledge that these older household heads have about their forests. In our study we d id not f ind statistically significant differences between age of household head nor time of residence across region but household head s of the Upper region tended to be relatively younger and had lived less years in their current landholdings, which could be one additional explanation as to why these residents tended to invest less and receiv e lower returns from Brazil nuts
46 Fruit production discrepancies explained by ecological factors, management, and household size Brazil nut fruit production is the ecological parameter of greatest importance for thousands of people that depend on this spec ies for their subsistence and cash income. Understanding the heterogeneity of Brazil nut fruit production in the context of extractive reserves is essential to understanding the potential of this NTFP to support families as well as understand where investm ents could be made to improve household welfare. Because of its importance, we measured fruit production by two means: (1) counted fruit production by the number of fruits encountered below each tree and (2) reported fruit production, or the average numbe r of boxes collected by families in each tree in past seasons. Within RDAER, heterogeneity of the more ecologically controlled variable (counted fruit production) exists, and can be explained by individual Brazil nut tree characteristics, as well as management practices. Brazil nut counted fruit production differences between trails and transects in our study (Table 2 1) clearly showed the negative influence of two linked tree level characteristics, liana loads and crown forms (Table 2 2). Brazil nut trees located at trails were less encumbered by lianas, which can be explained by the cutting of lianas on those trees, a practice that is easily carried out since trail trees are visited and harvested annually. In contrast, transect trees were, by definition, not systematically visited by residents, and thus had more liana s which were not cut, perhaps explaining in part, the reduced Brazil nut frui t production of transect trees. Regional differences (as measured in tree distance from the mouth of the Riozinho do Anfrsio river) in counted fruit production were not as clear (p = 0.0952; Table 2 2),
47 although descriptive statistics revealed that trees in the Lower region seemed to produce three times more fruits than those in the other two regions (Table 2 1). Tree diameter, a variable that explained over half of Brazil nut pro duction variation in a Western Amazonian study (Kainer et al. 2007), did not vary by region; however crown form did (Table 2 2), such that crown form was poorer as distance from the mouth of Riozinho do Anfrsio river increased (Figure 2 3). Of tree charac teristics that explain Brazil nut fruit production, crown form and liana loads are linked, such that trees with greater liana loads are much more likely to have poor crown forms (Kainer et al. 2006) and correspondingly, liana cutting gradually improves crown form (Kainer et al. 2007). While our Brazil nut stand quality model did not d etect significant differences in liana load by distance (p = 0.1579) (Table 2 2), liana cutting was the only management practice we investigated that did vary statistically by region (Figure 2 5 ). Upper residents did not invest in liana cutting, while over 80% of residents in the Middle region and over 40% in the Lower cut lianas. To our knowledge no other study has analyzed the influence of altitude on Brazil nut fruit production or other tree level parameters. Our results showed that trees on trails were at lower altitudes than those on transects an expected result since Brazil nut harvesters favor trees on locations of easy access. The RDAER region has a relatively rugged relief, including many hills and ravines that harvesters tend to avoid. Consequent ly, when collecting Brazil nut data on trees located in our 2.5 km transect we inventoried several trees located at hill peaks at higher altitudes. Because transect trees were not as productive as trail trees, we found a highly significant negative corre lation between tree altitude and reported fruit production (p < 0.0001) and also a
48 less significant correlation with counted fruit production at the 90% level of confidence (p < 0.0797). Descriptive statistics indicated that trees in the Lower region produ ced over three times more counted fruits than the other two regions (Table 2 1), although our Brazil nut stand quality model only revealed a tendency for tree distance from the mouth of the Riozinho do Anfrsio river to explain counted fruit production (p = 0.0952) (Table 2 2). This p value could be explained by the high degree of variation existing within regions, which made the statistical f test less sensitive across regions Our counted fruit production data were not normal ly distributed, and therefore, did not present a common variance required by the f test. However, when these results were presented and discussed with reserve residents in a community meeting in 2010, they agreed that Lower region indeed produces more Brazil nut fruit s than the other R DAER regions Still, did the seemingly greater numbers of counted fruits in the Lower region translate to a greater n umber of nuts reportedly sold? According to our descriptive statistics (Table 2 1), Middle region residents sold over three times more boxes of nuts than those in the Upper (and Lower) regions (Table 2 1), and thus a much greater percentage of Brazil nuts available on their existing trails. This seemingly contradictory result is proba bly a consequence of the types of data that these two variables represent. While Brazil n ut counted fruit production is more biologically controlled (number of fruits counted per tree), reported fruit production better reflects not only the biological cond ition of Brazil nut stands, but also the social aspects that influence Brazil nut harvest. For example, Brazil nut reported production (number of boxes harvested per tree) depends on the effort invested by harvesters (i.e., liana cutting and number of time s returned to each
49 knowledge about each of their Brazil nut trees and the accuracy of respondents to recall the average number of boxes harvested in each tree durin g the past seasons. For behavior is inaccurate ( Bernard et al. 1984) increases if they have to recall information that happened within six months of when they were interviewed (Bernard 2000) Reported fruit production could also reflect limited cash income options, market access, and market deman d (Shackleton & Shackleton 2004, Mahapatra et al. 2005). In RDAER, resident families sell most of their Brazil nut production, even when the production is small, relying exclusively on itinerant traders who buy mostly Brazil nuts and fish. Some itinerant t raders exclusively trade Brazil nuts, visiting RDAER solely during Brazil nut season. Following this rationality, one can expect that if RDAER families want to increase their cash income, they are constrained by market demand as defined by itinerant trader s. Finally, household characteristics also influence engagement and reliance on forest products by resident communities (Coomes & Burt 2001, Pattanayak & Sills 2001, Newt on 2011). Most of the household characteristics in RDAER were relatively even across the 3 regions studied; however household size was significantly larger in the Lower and Middle regions (Table 2 3), where number of Brazil nut boxes harvested was approximately three times larger than i n the Upper region (Table 2 1). A study in another extractive reserve in Brazil found a positive correlation between household size and the amount of fruit palm extracted, and attributed it to the higher labor input on this
50 activity (Newton 2011). Brazil n ut extraction is a labor consuming activity since harvesters need to walk to distant areas to extract this NTFP and then carry the heavy nut laden baskets to their homes. Therefore, the number of economically active household members directly affects the a mount of Brazil nut harvested and, consequently, the income sourced from this forest product. Even though we did not find a statistically significant variation of dependency ratio among regions, descriptive statistics suggested that Upper region residents had the highest dependency ratio (Table 2 3), indicating that possibly these families face labor limitations in concert with a lack of access to this important NTFP. In contrast, in the Middle region, household size is higher and the dependency ratio tends to be the lowest of RDAER, suggesting fewer to no labor restrictions. The access that these Middle region households have to Brazil nut stands also is much better than households of the Upper region (Figure 2 4 ). Finally, again contrasting Middle (most Br azil nut engaged) and Upper region (least Brazil nut engaged) residents, our descriptive statistics suggest that household heads in the Upper region are comparatively younger, have lived less time in their landholding, and have more formal education (Table 2 3). These findings support a similar NTFP engagement pattern identified in a different Brazilian extractive reserve in that younger heads of household had less accumulated knowledge regarding their forests a shorter time of residence, and invested less on NTFPs (Pattanayak & Sills 2001). Additionally, (Shone & Caviglia Harris 2006; te Velde et al. 2006) found that male heads of NTFP collecting households tend ed to have fewer years of schooling compared to non NTFP households. Jointly, these variables suggest that internal variation of household
51 characteristics also influence Brazil nut use, management and contribution to household income in RDAER. Regional Variation Shapes Brazil Nut Income The importance of NTFPs for income and livelihood improvements of forest peoples is well documented (Shackleton & Shackleton 2004; Vedeld et al. 2007 ; Maske et al. 2011) Studies have also shown that a large degree of heterogeneity exists within communities in relation to the typ es and frequency of engagement i n activities for income generation (Coomes & Burt 2001, M ahapatra et al. 2005, Belcher et al. 2005, Newton 2011); and that the wide variation observed is a consequence of the socioeconomic, demographic and geographic heterogeneity that exists within communities (Svarrer & Olsen 2005 ; McElwee 2008 ; Newton 2011). Our study supports these fin dings. We found no household total income heterogeneity in RDAER ; however, contr ibution of Brazil nut to forest based income varied by region (Table 2 3) indicating that Brazil nut socioecological heterogeneity shapes Brazil nut income in RDAER. R esidents of the Middle region, which had the lowest total income o f RDAER, had the highest forest based income with Brazil nut contribution playing a central role: 58 % of total income was from forest products, and Brazil nut alone contributed 23%. Some scholars h ave argued that the poorest forest residents rely more heavily on NTFPs than wealthier ones (Cavendish 2000 ; Belcher 2005 ; Mahapatra & Tewari 2005) while other studies have not found a direct relationship between poverty and NTFP income (Pattanayak & Sills 2001, McElwee 2008). W e did not aim to a nalyze the relationship of forest income and poverty, but our results seem to support the assumption that the poorest forest users rely more on NTFPs
52 Finally, and perhaps the most important factor in the context of RDAER, are the differences of household access to other sources of income as a result of the geographic location of each RDAER region. Studies have demonstrated that if forest people lack access to different sources of income they tend to rely more on NTFPs (Coomes et al. 2004 ; McElwee 2008; Duchelle et al. 2011). McElwee (2008) found that those with less access to income from government jobs or wage labor were more dependent on forest products. In a comparison between Brazil nut h arvesters from Pando, Bolivia, and an extractive reserve in Acre, Brazil, Duchelle et al. (2011) found that even though the Pando communities presented more constraints to Brazil nut investments than communities in Acre, Pando residents invested more on Br azil nut harvest mainly because of the lack of other sources of income. Comparing average Brazil nut and forest based income distribution within RDAER we observed that Middle region households presented the highest income sourced from Brazil nuts and tha t their reliance on forest products was higher than households of the other two regions (Table 2 3) While households of the Lower region have a good connection with Altamira during the entire year and the Upper region is located near Itaituba, which can be reached in 2 3 days during the summer season when they lack their river connection with Altamira, the Middle region residents stay almost completely isolated from these two sources of external income during the Summer season (Figure 2 1) Thus, in RDAER the lack of access to other sources of income as a consequence of a geographic location could, among other factors, be an important factor determining a
53 Heterogeneity Inside the Polygon The future viability of the extractive reserve model depends on both socioeconomic and ecological sustainability. Most extractive reserves present a highly complex socioecological system and a bundle of shared rights that drive the use of natural resource s, and consequently determine livelihood strategies of reserve residents (Ankersen & Barnes 2004, Cronkleton et al. 2011) Thus, understanding the variation of natural resource use and availability inside the reserve polygon is critical, especially if reserve residents rely on only a few NTFPs to meet their needs. This is the case in Riozinho do Anfrsio Extractive Reserve, where Brazil nut plays a critical role as a main contributor to household income. Findings from this st udy support the initial assumption that Brazil nut ecological and socioeconomic heterogeneity exists in Riozinho do Anfrsio Extractive Reserve, and that this internal heterogeneity determines the use and management of this important forest product. Resour ce quality and access are important aspects driving Brazil nut use and management, and lastly in shaping its contribution to household income. The socioeconomic differences found among RDAER residents, and their impacts on Brazil nut use, management and income, confirm the importance of taking into account household socioeconomic variables when analyzing the contribution of forest resources to household income. Furthermore, our results showed that even though hous ehold income is relatively even within RDAER, with a more detailed analysis of the forest products and Brazil nut contribution to household income, coupled with the geographical, ecological and socioeconomic aspects that affect household income in that res erve, a much more complex scenario of the economic context of RDAER emerged. This complex economic scenario is a result of the internal RDAER
54 socioecological heterogeneity, which is rarely incorporated into management plans and supporting conservation and development initiatives and policies in these reserves. Extractive reserves have been sold as a successful model to reconcile conservation and development goals that take into account the sociocultural diversity of Amazonian societies, and their traditiona l and well adapted management practices. However, when implementation and management plans are formulated for these conservation units, reserve residents have their management rights restricted and external agents make major decisions regarding natural res ource use (Cronkleton et al. 2011). Furthermore, conservation and development projects tend to focus either on a specific forest product, or on a specific group of forest users, and within group differences are typically neglected and do not take into acco unt the internal socioecological heterogeneity of these conservation units. We demonstrated the need to analyze internal ecological heterogeneity of one of the most important NTFPs in the context of extractive reserves in Brazil, and the relevance of takin g into account socioeconomic aspects when identifying the main factors affecting reserve residents income. If Brazil nut as well as the few other important cash generating NTFPs are to be used as a tool to guarantee the sustainability of RDAER and other Br azilian extractive reserves, policies and projects that aim to reconcile conservation and development goals need to take into account the ecological and socioeconomic heterogeneity surrounding these forest products in these complex multi use polygons. And the only way to do so is by putting more effort on site specific and highly detailed assessments of the forest products used by the different groups of reserve residents,
55 with the subsequent formulation of a management plan that encompasses relevant hetero geneity dimensions in these conservation units.
56 Figure 2 1 Map showing the distribution of family households (black triangles) and Brazil nut stands along the three RDAER regions (Lower, Middle, and Upper) (Source: Adapted from FVPP map in Rocha et al 2005).
57 Figure 2 2 Diameter class distribution of Brazil nut trees from transects in RDAER regions. Figure 2 3 Percentage of trees by crown form category within each of three RDAER regions.
58 Figure 2 4 Distances to Brazil nut stands harvested by region (1 = Lower, 2 = Middle, and 3 = Upper). A) Distance (minutes) to closest Brazil nut stands. B ) Distance (minutes) to second closest Brazil nut stands
59 Figure 2 5 Percentage of harvesters performing Brazil nut man agement practices within each of three RDAER regions. NS = not significant at p 0.05
60 Figure 2 6 Number of return visits to harvest trees within the same Brazil nut season by region (1 = Lower, 2 = Middle, and 3 = Upper).
61 Table 2 1. Descri ptive statistics of Brazil nut den sity, dbh, counted and reported fruit production in the three RDAER regions and at the Reserve level.
62 Table 2 2. Best fit models that explain Brazil nut quality in RDAER. A general linear mixed model ANOVA was used to explain tree dbh, liana loads, crown form, counted and reported fruit production.
63 Table 2 3. Descriptive statistics of household characteristics, forest based income, income sourced from Brazil nuts and the total household income in RDAER ), and p values denoting the significance of regional differences.
64 CHAPTER 3 LINKING SOCIAL NETWO RKS AND INCOME IN A SUSTAINABLE USE RESE RVE: IMPLICATIONS FOR CON SERVATION INTEGRITY Social Networks and Natural Resources Management Sustainable use reserves represent the fastest growing type of protected area worldwide. According to the World Conservation Union (IUCN) protected area designation, sustainable use reserves are classified as categories V and VI (Protected Landscape/Seascape and Protected Area with Sustainable Use of Natural Resources, respectively), and their primary goal is to protect natural ecosystems and promote the sustainable use of natural resources, while fostering economic and soc ial benefits to local communities (Dudley 2008). The emergence of this type of protected area resulted from grassroots social movements (Borrini Feyerabend et al. 2004) merging with conservationist defined needs to integrate local people into long term man agement of these areas (Dudley 2008). Successful implementation requires crucial understanding of the dynamics of social interactions taking place in these reserves, and the role they play in natural resource use and management. Collaborative management am ong stakeholders implies a certain level of trust, which is formed through social interactions among these groups (Coleman 1988, Fukuyama 1995). Many studies highlight the role of social networks i n trust emergence and maintenance within groups (Granovette r 1985, Powell 1990 Putman 1993), and emphasize that trust can facilitate cooperation and coordination for mutual benefits (Putman 1993). The internal relations that take place w ithin these protected areas, and also interactions and relationships with ext ernal actors, shape behaviors and institutions. These social interactions, and their emergent patterns, determine the flow of resources through the social system (Weimann 1982, Abrahamson & Rosenkopf 1997,
65 Reagans & McEvily 2 003). Thus, cooperation and pro vision of assistance to manage natural resources in sustainable use reserves is embedded in social networks (Granovetter 1985). Scholars and practitioners have progressively appreciated that resource users do not behave as isolated individuals, but rather, they form part of a concrete and dynamic system of social relations. Interpersonal influences play an important role in formulating actor opinions and attitudes, and thus, form the foundation of actor decisions and behavior (Friedkin 1998). Indeed, psycho log ical theory has shown that individuals are more influenced by those with whom they interact regularly and often (Festinger et al. 1950). These usually are individuals with whom they share the same beliefs, views, and perceptions; that is they are part o f the same subgroup (Cooley 1909, Homans 1950, Kadushin 1966). Reserve resident communities are not static and homogeneous units, but rather are composed of different subgroups with different interests, resources, and perceptions, as well as different l ev els of influence (Carlsson & Berkes 2005, Nygren 2005). Understanding these different subgroups and how they relate to each other is crucial to build consensus and cooperation in light of natural resources management. Social networks are a way to operatio nalize existing social theories and to understand the views and behaviors of natural resource users, such as those living in sustainable use reserves. Indeed, social networks can improve or constrain user capacity to tackle both internal and external chall enges to the intertwined socioecological system in which they are embedded ( Batterbury 2003 Crona & Bodin 2006). While systematic research has been conducted to understand how social
66 network structure influences natural resource management and community d evelopment ( Freudenburg 1986, Tindall 2002 Crona & Bodin 2006), to our knowledge no study has attempted to link social network structure with household income in the context of a sustainable use reserve. By applying methods from the field of social networ k analysis, we assess the structural aspects of social networks of households within sustainable use reserves, and explore how these networks link to family reliance on different s ources of income, mostly forest based vs. non forest based income. We use th e Amazonian extractive reserve of Riozinho do Anfrsio (RDAER) in Brazil as a model, examining the following specific questions: To what extent do household social networks within RDAER vary? What types of residents play a central role within the reserve polygon and external to it? To what extent do ties of resident households with external actors vary within RDAER? Do households from different RDAER regions rank external actors differently in terms of trust? Are resident social networks in each region linked to differe nt income (forest based vs. non forest based) sources? To determine the social position of actors within the reserve and to understand internal sociological dynamics, we use the concept of centrality. Centrality is an actor level concept t hat generally reflects ideas of relative power, prestige, or popularity (Bodi n & Prell 2011). Centrality measures the social position of each actor ( household s in our case) within the entire network structure. Social network analysis provides different way s to measure centrality, but here we will rely on degree centrality, which is simply the number of direct connections that one actor has in a network ( Freeman 1979 ). Actors with high degree centrality are connected to many other actors and thus
67 have more s ocial capital giving them more opportunit ies to exchange resources and ideas directly. We will also use the concept of centrality to identify residents who play an important role to connect other reserve residents to external actors, or that can serve as i mportant sources to disseminate ideas and interests linked to these external actors. We assume that organizations involved in reserve implementation and management, or which are part of the socioeconomic context of this reserve, play an important role i n c onstraining and guiding reserve resident attitudes and behaviors, as articulated by Bandaragoda (2000). Thus, we analyzed how reserve residents link to those external actors, using a two mode network approach ( Bodin & Prell 2011 ). In contrast to a one mode network where the relational data are measured between actors with all other actors, a two mode approach measures the interaction of actors with organizations or events. By examining the pattern of co use of organizations or events we can imply a type of relationship between actors. For example, the position of natural resource users in a network based on participation in certain events, such as membership in cooperatives and associations, or residence in a community, can elucidate many important connectio ns related to the formation of social groups and their roles to disseminate knowledge, and to influence other me mbers of these groups (Bodin & Prell 2011). Social network analysis can also identify differe nt groups of stakeholders and provide insights into the role that they play to foster cooperation for natural resource use and management. Thus, by assuming that RDAER is composed of groups of actors with different values, perceptions and opinions, we also use social network analysis to
68 identify different subgroups and relate those subgroups with natural resources use in this extractive reserve. Riozinho do Anfrsio Extractive Reserve Context Riozinho do Anfrsio Extractive Reserve (RDAER) was created in 2004 by presidential decree, and hosts approximately 26 households. RDAER encompasses an area of 736,430 ha of the municipality of Altamira; it is commonly divided into three geographic regions : Upper, Middle and Lower (Figure 2 1). Resident households are isolated from one another and dispersed along the Riozinho do Anfrsio River, a settlement pattern that constrains daily contact between resident families. These families rely on subsistence a ctivities such as hunting, fishing, slash and burn agriculture, and collection of forest products (Rocha et al. 2005). Itinerant traders who operate in the region provide the few goods that are not produced locally. All RDAER residents have access to the c ity of Altamira by river, but the Upper residents frequently visit Itaituba, which is the closest city to the reserve. Itaituba can be reached by walking approximately 2 3 days from the penultimate landholding of RDAER to arrive at a small road frequently travelled by ranchers and loggers that live in the surrounding area. No public transportation is available on this unpaved road, and thus for reserve residents, access to Itaituba is solely via opportunistic rides with ranchers and/or loggers, and only dur ing the 4 months of the summer season when rainfall is low. Lower region residents have greater access to Altamira, which is located approximately 10 15 days by river. Although it clearly takes Middle region residents longer to reach Altamira than their Lo wer region neighbors, they always use this river route when ne eding to access city resources.
69 Alta mira is the main center for non forest products trade, and all governmental and non governmental organizations (NGOs) that operate in RDAER are based t here. S ervices provided to reserve residents by these organizations include health and education, and technical assistance for natural resource management and reserve management in general. Altamira is also the base of several conservation organizations, and thus plays a central role in the conservation agenda for the region. Historically, Altamira had a socioeconomic character more oriented to agricultural, ranching, and logging activities a description that currently applies to Itaituba. This change started in the late 1 99 0s when a social movement of smallholders of the Transamazon Highway proposed to the Brazilian government an agenda that would reconcile conservation with e conomic rural development. This culminated in the creation of many conservation units i n 2004, including Riozinho do Anfrsio Extractive Reserve (Campos & Nepstad 2006). As with other extractive reserves in the Brazilian Amazon, management and composed of seve ral groups: the Federal Environmental Agency (ICMBio) whose role is to coordinate activities related to reserve management, representatives of other local governmental organizations and NGOs, plus reserve resident representatives. Once a year this Manageme nt Council assembles to make decisions and monitor reserve activities. The RDAER Management Council was created officially by decree in 2008; however, many activities started on the ground in 2006, and those activities were mainly carried out by NGOs that were involved in RDAER creation. Simultaneously, the local
70 RDAER association (AMORA Associao dos Moradores da Reserve Extrativista do Riozinho do Anfrsio) was also created to provide reserve residents with services related to citizenship, health, edu cation, and trade of forest products. While all RDAER residents are members of this associatio n, only 12 reserve residents are officially elected to be representatives of AMORA for a two year appointment Methods Data G athering From a total of 26 RDAER hou seholds, 25 household heads ( all male ) were interviewed to determine the social network structure among reserve residents. Social network interviews were applied in two rounds, February 2009 and August 2010. The first round of interviews sought to understand the resident network structure within RDAER. The second round identified ties that reserve residents had with external actors (the two mode approach) and how resident households ranked actors in terms of their trust, and usefulness to support resident livelihoods. Simultaneously, a livelihood survey was applied to determine household characteristics, to tal household i ncome, and forest based vs. non forest based income as a way to identify livelihood strategies adopted by each RDAER household. Round one: internal RDAER social network structures This first series of questions explored relationships intervi ewees had with each other. R espondents were asked to provide names of other Reserve residents from items (canoes, motorboats, and manioc flour) that are typically used by hou seholds and exchanged within RDAER. The main means of transportation in RDAER are canoes and motorboats, b ut not all residents own canoes or motorboats. RDAER households
71 that lack these items have to rely on others to access services such as health assista nce, or to communicate with other reserve residents since families live apart from each other and radio communication is scarce. Manioc flour is the key staple food consumed in RDAER and is extremely important for reserve residents. By selecting these thr ee topics to elicit the social network structure within RDAER, we assume that reserve residents will rely mostly on residents with whom they have an established and good relationship, or on those that they believe have the ability to provide these supplies and equipment. Round t wo: t ies with institutions and other external actors To determine external household ties, we first developed a list of possible actors by using different sources of information: reports and other documents that attested to the inter action of RDAER residents with those actors, previous informal conversations with RDAER residents as well governmental and non governmental representatives, and participation in meetings and workshops. T his list included external actors present in RDAER at the time of the survey as well as past relationships among RDAER residents and external actors. A total of 25 governmental organizations, 8 NGOs, and 8 commercial enterprises and cooperatives related to the private sector were used to elicit ties that res erve residents ha d with external actors ( Appendix D) Most of the governmental organizations operate d at the local level and we re responsible for providing different services to RDAER residents, such as health and education services, agrarian and environmental technical assistance, citizenship documentation, and information related to reserve management. The NGOs considered in this study were local, national, and international organizations mainly concern ed with socio environmental issues. Finally,
72 the private sector wa worker union Flash cards were subsequently prepared containing the names of each external actor that was part of this list, and differen t questions were asked related to six main topics: (1) recognition, (2) frequency of contact, (3) visit of external actors at households, (4), assistance, and (5) trust. We first asked respondents if they recognized each external actor on the card. Actors not recognized were discarded for the rest of the interview For those that were recognized, we asked respondents to select one of the following contact frequencies: n o contact, once a year, twice a year, three times a year, eve ry two months, once a month, or once a week. Then we asked questions related to number of visits received, and number of times that each actor assisted their particular household. The following ranks were coded for these groups of questions: never, on ce or twice, three to f ive times, and more than five times since reserve creation in 2004 Visiting and assistance delivered were used to determine more practical and concrete interactions between resident families and external actors. Assistance was also used as an indicator of benefits distribution among RDAER residents since many external actors have the responsibility to implement, manage, and provide socioeconomic and environmental services to residents. Finally, t o determine the degree of trust that respondents have with e ach actor, we asked respondents to select one of the f ollowing trust categories: no trust, neutral, trust, or u nknown. Livelihood survey To determine general household characteristics, such as time of residence, household size, age, schooling and affiliati on of head of household to RDAER
73 association and external organizations, and household income, we applied a livelihood survey adapted from the Center for International Forestry Research Poverty and Environment Network (CIFOR PEN; http://www.cifor.org/pen/research tools/the pen prototype questionnaire.html ) to the head of household among the 2 3 (of a total of 26) available households in the entire Reserve. To determine total household income, we quantified all forest and agricultural products that were traded and sold, and the economic benefits from government, wage labor, and other sources of e xternal income earned during the year of 2009. Processed (baskets, brooms, canoes, and paddles) and unprocessed (mainly Brazil nuts, andiroba, copaba, honey, rubber, and fish) subsistence forest products were also quantified to determine total household f orest based income. We excluded the amount of game consumed due to lack of confidence in our ability to measure this accurately No game restrictions exist in RDAER and all RDAER residents hunt. Manioc, which is the main RDAER agricultural product, and chi cken, were also measured Based on this detailed information, we calculated total household income and the contribution of for est based and non forest based income to total household income. Income is re presented in the Brazilian currency, reais (1 U.S. do llar 1.5 reais). Finally, we sought a better understanding of the institutional and social context in which each RDAER household was embedded. We asked questions related to the frequency of visits to the closest cities (Altamira and Itaituba), affiliati on to RDAER association or other external institutions, and participation in meetings and courses organized by external actors. Participation in meetings within and outside the reserve is
74 important since most resident input in decision making related to re serve management occurs during those meetings. Capacity building courses to develop skills for reserve management spanned from educational and legal aspects of reserve management to non timber forest product information to basic reading and writing skills Data A nalysis To determine the RDAER internal social network structure and degree of connectivity that reserve residents have with external actors we used UCINET software, version 6 (Borgatti et al. 2002), which provided visual and statistical network an alyses. With the exception of data regarding the frequency of c ontact that RDAER residents had with external actors, we transformed our two mode data (social networks linked to external actors) using dich otomization. Thus, data that were ordinal were reduc ed to a binary matrix (presence or absence of social ties ). We used a procedure to create one mode household by household matrices from the two mode data where the tie linking every pair of households was the number of external actors they shared in common Girvan Newm an subgroup analysis (Girvan & Newman 2002) was used to identify different groups in RDAER as well as to test differences among RDAER regions in terms of ties at the local and regional level. The Girvan Newman procedure identifies subgroups by identifying the most bridging ties in the network, simulating their deletion and identifying subgroups that become disconnected (Girvan & Newman 2002). In an attempt to identify potential overlap of reserve resident social networks (within reserve and wit h external actors) with the three RDAER regions we forced the software to search for three clusters. Finally, we created a set of matrices using household head attributes ( region, age, schooling, time of residence within the reserve, time of residence at p articular landholding, frequency of visits to closest cities, representation on RDAER
75 association, and household income ) and r a n a Quadratic Assignment Procedure ( QAP ) correlation analysis to test if the structure based on these household head attributes was correlated to RDAER social networks based on interaction with external actors. Descriptive statistics were used to determine the contribution of different sources of income (forest based vs. non forest based) to total household income in each RDAER reg ion. To understand internal RDAER variation regarding household characteristics, their institutional and social context, forest based and non forest based income, and the total household income, we estimated General Linear models, with ictor variable, and also calculated Pearson correlation coefficients for all pairs of measured variables, using SAS statistical software, version 9.2 (SAS 2004) Results Social N etworks Within Reserve P olygon Results from the analyses of internal rese rve dynamics showed that household networks were strongly related to geographic location and proximity. In most cases, households relied on nearest neighbors to borrow supplies and equipment (Figure s 3 1 and 3 2 ). The network related to borrowing canoes pr esented three subgroups that partially related to the spatia l arrangement of RDAER (Figure 3 1 ), such that the three clusters observed in the diagram almost matched with the three RDAER regions commonly recognized by both internal and external actors. Simi larly, when asked from whom they would borrow a canoe, most responded that they would ask their neighbors or other households of the same region. O nly households of the Middle region that are located closer to the Lower region or closer to the Upper region interact more with residents of these two regions since those are their neighbors.
76 The network for borrowing motorboats also tended to match the geographical internal division of RDAER ( Figure 3 2 ). The main component (a set of nodes tied directly or ind irectly) was composed of three subgroups, each of them composed predominantly of households from those regions. Middle region residents were again present in these three main subgroups, demonstrating the importance of location. The only exception to this i s a small subgroup centered around one Lower region resident (Figure 3 2 household # 4), who are also neighbors, and therefore, borrow motorboats from each other because of their closer location. The network for borrowing manioc flour revealed a similar p attern (not shown) In RDAER, we observed that the number of households with a central position (or more connections) within the internal social network decreased from the Lower to the Upper region, when focusing on provision of canoes and motorboats (Figures 3 1 and 3 2 ). In the Upper region, most residents relied almost completely on one household for borrowing canoes and motorboats In the Middle region we found two main actors that provided canoes and motorboats, while in the Lower region centrality did not appear to be a s strong many resident households served as lenders (Figure s 3 1 and 3 2 ). No households with a central position, regardless of region, were observed in the case of manioc flour since that network was not cohesive, but presented only small sparse groups of neighbors that borrowed manioc from each other. QAP analyses showed a positive statistical correlation between provision of motorboats (mirrored in degree centrality) and total income ( p < 0.005), and between provision of motorboats and representation o n AMORA ( p < 0.005). This means that the structure of motorboat lending was by those with higher incomes and those people
77 tended to be AMORA representatives. No statistical centrality related correlations were found between provision of canoes or manioc fl our and the diverse household variables studied here. S ocial N etworks Linked to External A ctors Contact of RDAER residents with external actors S ubgroups found when analyzing the data related to recognition, frequency of contact, or number of visits by ext ernal actors did not align with the regions However, when considering only those households that received more than 3 visits from external actors, two subgroups emerged (Figure 3 3 ). The first was a small Upper Region subgroup composed of only two residen ts: a municipal health agent who received visits from many external actors related to his job, and a second resident who not only received visits from governmental and non governmental organizations (NGOs), but also had good connections with private secto r actors. The second subgroup was composed of a diversity of residents predominantly of the Lower region the region most visited by external actors. Nonetheless, it was a household in the Middle region (represented by a large square) that received the hi ghest number of visits from different actors, demonstrating a higher degree centrality and a central position in the network ( Figure 3 3 ). Furthermore, QAP correlations corroborated these visual results, demonstrating that the number and diversity of visit s by external actors were statistically different by region ( p = 0.027). Similarly, AMORA representatives tended to receive a higher number of visits than non representatives ( p = 0.063) No other correlations were detected regarding other household charac teristics and visits from external actors.
78 Assistance received from external actors When analyzing the number of external actors that provided assistance to any given household, no subgroups were identified; however, our results showed that most external assistance was concentrated on the Lower region ( p = 0.007) since the number of residents benefiting from external relations was much higher in comparison to the other two RDAER regions (Figure 3 4 ). Assistance received from external actors was not correla ted to other household characteristics tested. Although no clear correlation was found between assistance received from external actors and representation on AMORA ( p = 0.110), we observed that most elected AMORA representatives tended to receive assistanc e from a higher num ber of external actors (Figure 3 5 and Table 3 1). Indeed, these households were the ones that received external assistance from more actors both at the scale of the entire reserve and when analyzing each region separately. The only exce ption was the health agent of the Upper region (who was not an AMORA representative) who was well connected with several governmental and non governmental organizations because of his job. Despite this exception, it was clear that Upper region residents benefitted the least from their external assistance since five out of the seven regional households received no assistance from external organizations, including those who were AMORA representat ives (Figure 3 4 and Table 3 1). On the contrary, in the Lower region, we observed that even household heads that were not AMORA representatives (four of eight) received assistance from many external actors. Finally, in the Middle region, of those variables studied, only representatives of AMORA benefited from their connections with external actors.
79 Analyses run separately by group of external organizations (government, NGO or private) showed that most assistance came from governmental institutions and NGOs the two groups officially responsible for implementation an d management of t his extractive reserve (Figure 3 5 ). Subgroups were identified only in the case of assistance received from the private sector, which included entities linked to the logging and ranching sectors, and unions from the main municip alities of this region (Figure 3 5 ). Overall, government and NGO assistance mostly benefitted Lower and Middle region residents as well as re presentatives of AMORA (Figure 3 5 and Table 3 1). Assistance from governmental institutions was more concentrated on a few h ouseholds, mainly from the Middle and Lower regions. In contrast to this limited government reach, NGOs provided assistance to a higher numb er of RDAER households (Figure 3 5 ). In the case of private sector assistance, two groups emerged: one formed by thr ee Lower region households, and the other formed by two households from the Upper region (Figure 3 5 ). Furthermore, Lower residents received assistance from more actors of the private sector than Upper residents, and all three Lower region households assis ted by the private sector were AMORA representatives (Figure 3 5 and Table 3 1). Trust as ranked by reserve residents No subgroups emerged when examining the degree of trust that RDAER households had with external organizations. However, when excluding the households that trust less than three external organizations by increasing the number of relations in the Girvan Newman analysis, results showed that residents of the Upper region presented the lowest degree of trust in comparison to residents of the o the r RDAER regions (Figure 3 6 ). Results of the QAP correlations confirmed that household trust
80 with external actors was statistically different by region (p = 0.027), and AMORA representatives were the most trusting of external actors (p < 0.01). When restricting our analysis to governmental institutions and only those households that trust more than three of such institutions, two subgroups emerged: one composed only of Upper region residents and another network of both Lower and Middle region resident s (Figure 3 6 ). These two networks present different scores of degree centrality as represented by node size, suggesting that overall, residents of the Lower and Middle regions trust more governmental institutions compared to Upper region residents. Furthe rmore, most of these residents identified, were members of AMORA, regardless of region. In the case of trust that households have with NGOs, no subgroups were clearly identified, even when increasing the number of relations that households have in common w ith these organizations. Nonetheless, most Upper residents were located at the network periphery, indicating that these households potentially trusted a lower number of NGOs in comparison to residents of the other two regions. As we increase the number of ties in the Girvan Newman subgroups analysis, these peripheral households were excluded from the network and only one resident of the Upper region, a very active mem ber of AMORA, remained (Figure 3 6 ). Finally, no distinct subgroups were found regarding th e number of private sector actors that RDAER household s trusted when all households w ere included in the analysis. Even when we excluded households that trust less than three private sector actors, no subgroups emerged. However, residents of the Lower and Middle regions presented a higher degree centrality related to trust with external private actors,
81 indicating that overall, they tend to trust a higher number of private sector entities than residents of the Upper region which again, were located on the ne eriphery (black nodes) (Figure 3 6 ). Interactions Outside the Reserve and Household I ncome No household heads interviewed were members or representatives of external institutions. Within the reserve polygon, although all RDAER households were lega l members of AMORA, only the 11 elected representatives recognized and articulated their membership (Table 3 1). Data related to cities most frequented by households of the three RDAER regions showed that residents of the Lower and Middle regions only visi ted Altamira (Table 3 1). On the contrary, in the Upper region, we found a more diverse scenario in terms of cities visited: two households exclusively visited Altamira, two exclusively visited Itaituba, and two visited both. In addition, results of partic ipation in meetings showed no correlation with representation in AMORA. However, we found a significant correlation with AMORA representation and participation in capacity building courses to improve reserve management (p = 0.0457)(Table 3 1). Total househ old income in RDAER ranged from R$ 596 to R$ 21,871 in 2009, and in all three regions, we observed that at least two households showed outstanding total household income (Table 3 1). We found no significant correlations between total household income and a ge of head of household, schooling, nor time of residence at that particular location. However, representation on AMORA was strongly correlated representatives of AMORA earned a mu ch higher total income in comparison to non representative residents. This is the case in the Lower and Middle regions, where the two highest earning households were members of AMORA. In the Upper region, the
82 household with the highest total income is an A MORA representative, but the next highest earning household is not. Finally, we also found a significant positive correlation between total household income and time of residence within the reserve (p = 0.0381), with long term residents earning a higher to tal income. Non forest based income was also highly correlated with AMORA representation (p = 0.0112). Results showed that in the Lower and Upper region most AMORA representatives earned the majorit y of their 2009 income from non forest based sources. In t he Lower region, more than 70% of total income of AMORA repre sentatives was sourced from non forest based products. We observed this same trend in the Upper region. Middle region AMORA representatives diverged from this pattern in that only one AMORA repre sentative had a higher portion of total income sourced from non forest based products. In addition, age of household head seemed also to have a positive effect on total non forest based income (p = 0.0775). No variation in total household income was eviden t when comparing earnings among residents from the three RDAER regions (Table 3 2). Nonetheless, families from the Middle region tended to earn a higher proportion of their total income from forest products (Table 3 2, p = 0.1064), in opposition to the Low er and Upper regions where households had most o f their income sourced from non forest based activities. Contributions of non forest based activities were not significantly different between regions (Table 3 2). Nonetheless, residents of the Lower and Upper regions receive d twice as much income from non forest activities as those in the Midd le region, but the types of non forest activities they focused on were different. Upper region residents tended to have a greater proportion of their income from agr icultural activities,
83 gold minin g and contract wages (64% vs. 46 % in the Lower region). In contrast, Lower region residents earned a larger portion of their income from government social benefits (mostly retirement, maternity, and school payments) (p = 0.2 064). In that region, the contribution of governmental social benefits to total household income was almost 3 times higher than in the other RDAER regions. Discussion W e set out to explore social network structures within sustainable use reserves and how t hey link to different sources o f income, mainly forest and non forest based income. Below, we highlight three key factors that we hypothesized would explain social network structures and their link to sources and amount of household income: geographic loca tion of households, formally recognized social positions of residents, and trust that residents have with external actors. We also explored the importance of trust with external actors for building collaboration for reserve management. Finally, we provide insights on the utility of social network analysis as applied to sustainable use reserves. The Primacy of Geographic L ocation Internal s ocial networks explained by household settlement patterns Our f indings showed that geographic location of reserve households was the primary factor defining social networks among households in the reserve Indeed, geographic location is considered one of the most basic sources of influence of social connections because it is easier and less costly to connect people who are in close proximity (McPherson et al. 2001). In RDAER, where residents had no access to technologies that could imp rove communication, geographic space played an important role on the intensity and frequency of contact with other reserve residents (McPherson
84 et al. 2001). Accordingly, the Upper region was the most isolated area given that the river flows in a northeasterly direction and navigability during part of the year is restricted; residents of this region had much less acces s to residents of the other two regions. In contrast, closer to the mouth of Riozinho do Anfrsio river in the Middle and Lower regions, navigability and access to other reserve residents increased, allowing more interactions among reserve residents of tho se two regions. Still, RDAER social interactions were mostly restricted to those households that shared the same region as defined b y household settlement pattern. Based on these results, we can assume that these three RDAER regions represent not only the attachment that reserve residents have with their physical environment, but also the social relations that they have with residents that share the same place. When individuals share a socially constructed setting, they learn how to live with others, and develop practices of co habitation (Cheng et al. 2003; Kemmis 1990). This seems to be the case of RDAER, such that the geographic regions represent three socially cons tructed settings where people influence each other (Cooley 1909, Festinger et al. 1950, Homans 1950, Kadushin 1966). Geographic location seemed also to influence the number of households that had a central position as providers of supplies and equipment in RDAER given that the number of centralized hous eholds decreased as geographic isolation increased. In RDAER, more isolated households had less access to residents of other regions, forcing them to depend on few households for supplies and equipment. Inde ed, scholars have argued that isolation, and consequently, the limited communication that isolated areas have with the outside world, strengthen s the dependence of residents on
85 just a few key individuals ( Da Cunha 2003 ). This is clearly the case of the soc ial network structure related to the provision of canoes, where the number of households that lent this equipment increased the closer one lived to the mouth of Riozinho do Anfrsio river Even though this same structure was not so apparent regarding provi sion of motorboats, we still observed that the Lower region, the most accessible region in RDAER, presented a higher number of households that served as providers of motorboats in comparison to the other more isolated RDAER regions Access to RDAER regions determined social networks linked to external actors When analyzing linkages to external actors, geographic location also appeared as an important determinant of RDAER social network structures. Overall, households of the Lower and Middle regions were muc h more connected with external actors than Upper region residents who tended to occupy a more peripheral location in all social networks studied. External actors from Altamira had easier and more consistent access via river traffic to the Lower and Middle regions. In contrast, river access to the Upper region from Altamira, but also overland from Itaituba, was extremely difficult a nd restricted most of the year. The emergence of two subgroups related to external visits also seemed to be a consequence of ge ographic location in this case, city access. External actors located in Altamira had better access to the Lower and Middle regions and thus visited more residents in those regions, while Upper region residents received more visits from exter nal actors lo cated in Itaituba. Geographic location and access also seemed to influence comparative levels and types of assistance received by RDAER residents. First, more Lower region residents d assistance
86 from external actors than their resident counterparts in the two other RDAER regions. Secondly, NGOs and governmental organizations, all located in Altamira, predominantly assisted households located in the more easily accessed Lower and Middl e regions, while Upper residents remained on the periphery of both the governmental and NGO networks This uneven coverage is significant since governmental organizations and NGOs are officially responsible for provision of most socio environmental service s to the reserve. On the other hand, private sector actors have no official authorization to operate within the reserve, which could explain why only 5 out of 24 households received assistance from private sector act ors. An alternative geographically linke d explanation could be that private sector actors, almost entirely located in Itaituba, could not provide much assistance because they had almost no physical access to the Lower and Middle regions and only restricted access to the Upper region. Still, the social network structure related to private sector assistance revealed two subgroups: one composed by Upper region residents and one by L ower region residents These results indicate that households from the Upper region have connections with a different s et of private sector actors than Lower region residents, which was agai n determined by the geographic location of RDAER households and external actor access from Itaituba and Altamira, re spectively. Geographic location can also explain the fact that no Mid dle region residents received assistance from private external actors. Private sector actors diffe r not only on their geographic location, but also their activities. The Altamira based private actors are worker union organizations that provided citizenship and organizational services at the inception of reserve creation, and are thus highly linked to governmental and NGO activities within the reserve. Itaituba based
87 private actors are mainly ranchers, loggers and land grabbers, which in most cases, operate illega lly within the reserve context. Formal Social Positions Mattered Formal social positions also emerged as an important factor linked to RDAER social network structures. Residents who were elected representatives of the local RDAER association (AMORA) were particularly important actors both within the reserve and with external actors Because they were required to participate in external meetings related to reserve management, AMORA representatives had greater access to external resources and therefore serve as important sources of information and goods accessed through their social connections ( Burt 2000, 2004 ) Indeed, in a remote region such as RDAER, this greater access through social relations made other reserve residents perceive these representatives to be potential providers of supplies and equipment. With few exceptions, they were indeed key lenders of can oes and motorboats (Table 3 1, Columns 5, 10 and 11). Likely, their possession of motorboats, a very expensive item that most RDAER residents could not afford on their own, was a direct result of their formal position. Most reserve motorboats were donated by NGOs that operate in the reserve, and their allocation to AMORA representative households makes institutional sense. However, while households in which a resident had a formal job with external actors also presented more interactions with external actor s as did AMORA representatives, they did not emerge as central actors within the reserve. Perhaps because they were not elected representatives, reserve residents did not expect them to be benefits providers as were AMORA representative households. Our fin dings also showed that visits received by external actors were concentrated on households that had AMORA representatives (Table 3 1 Columns 5 and 12 ) and on
88 a few households that had a job linked to external actors, partially explaining why only 9 out of 24 RDAER households received visits from more than three external actors. Similarly, external assistance was concentrated on these same households. This was the case of one Upper region household head with a formal job in one of those organizations (Figure 3 3 and Table 3 1; household 19). While it was expected that AMORA representatives would receive more visits from external actors, particularly from government organizations and NGOs operating within the reserve since its creation in 2004, we had not expe cted that AMORA representatives would also receive assistance from more external actors than non AMORA representatives We expected government organizations and NGOs responsible for reserve management and implementation to have a n egalitarian approach and thus provide benefits to all reserve residents Trust as Explained by Social Interactions and Formal Social Positions Social scientists have argued that it is through social interactions that trust is formed (Coleman 1988, Fukuyama 1995, Warren 1999); and simultaneous exchange of goods fosters reciprocity, which then increases trust (Putman 1993, Pretty 2003). Thus, it was expected that households that received more visits, but especially more assistance from external actors since from the perspective of reserve residents assistance could be understood as an act of reciprocity would have more trust with those external actors. Indeed, the households that trusted most of the external actors were households from the Lower and Middle regions that received m ore visits and assistance from external actors. On the contrary, residents of the Upper region, which had a more peripheral position within the social networks related to visits and assistance, had also a peripheral position in the social network related t o trust,
89 indicating that those households trusted only few external actors. Correlations between trust to external actors and location related to the three RDAER regions were statistically significant (p = 0.027), which confirmed the variation found in thi s study in relation to the degree of trust that households from different RDAER regions have to external actors. More importantly, the strong statistically significant correlations found between AMORA representation and degree of trust (p < 0.01) may indic ate that trust in RDAER is formed through formal social positions, or perhaps that social position is a consequence of trust. In all three RDAER regions the residents that trusted most of the external actors had a formal social position, mainly as represen tatives of the local RDAER association (AMORA). Our findings also showed that few households had trust with most governmental organizations, which were the primary organizations responsible for reserve management. Furthermore, those households formed two s ubgroups, indicating that those reserve residents trusted different sets of governmental organizations The social network structure related to trust that reserve residents had with NGOs showed no subgroups, but Upper region residents, had again a peripher al position. Furthermore, most of the households that trusted most of the governmental organizations and NGOs were AMORA representatives (Figure 3 6 see households # 2, 3, 7, 9, 12, 13, 15, 16, 20, and 22). Historically, Upper residents have been perceive d by government, and also non government organizations, as opponents to reserve creation because of their past connections with external actors, mainly linked to the rancher, logger, and land grabber sectors located in Itaituba (information sourced from in terviews). Indeed, conflicts
90 between Upper and Lower region residents, which were the most connected with governmental and non governmental organizations, were frequent during reserve creation. A lmost 8 years later, some governmental and NGO representative s are not allowed to enter landholdings of some Upper region residents. While these conflicts have diminished in recent years, our findings suggest that Upper residents still do not trust the majority of those organizations. Trust building in RDAER seems t o be predominantly achieved through ongoing social interactions linked to formal social positions a mechanism that has reached only few households in RDAER Linking Social Networks and Trust with Household Income Our findings showed that greater inte ract ions favored by geographic location and formal social position seemed to promote formation of trust among reserve residents and external actors, which may affect total household income, and income sourced externally. Scholars have also pointed out the dire ct effect of membership (in clubs, associations, and other social organizations) and contract work among managers on income (Corral & Reardon 2001; Boxman et al. 1991). Thus, the links among AMORA representativeness, their high connectedness with external actors, and the total household income found in this study (Table 3 1), plus the statistically significant correlation o f AMORA representation with non forest bas ed income (p = 0.0112), seems to support the findings of those scholars. We have argued above that the higher degree of trust that households with a formal social position had with external actors could be explained by the higher interaction that those hou seholds had with those external actors, as required by their formal social position (Coleman 1988; Fukuyama 1995). Using the same rationale, we also argue that the higher degree of trust that households with a formal social position had with external actor s, which determined their social
91 networks with external actors, could also favor their access to external reso urces provided by those actors. Analysis of the different sources of income, in light of the connections that reserve residents from different reg ions have with different groups of external actors, make the links among social networks, trust and income more apparent. Thus, contribution of government benefits to total household income is much higher in the Lower region (31%) than in the other RDAER r egions (15% and 13% in the Middle and Upper, respectively) (Table 3 2), which seems to be a consequence of the higher number of Lower region residents that had good connections and trusted most NGOs and governmental organizations. In fact, it is through go vernmental organizations and NGOs that government benefits can be accessed in this reserve, since those organizations are responsible to assist reserve residents to access benefits such as retirement salary. Thus, it would be expected that households with a better relationship with those external actors would also have better access to resources accessed through those actors. Lower region residents also sourced part of their total income from formal government jobs or contract wages in temporary positions w ith fishing or as boat drivers for itinerant traders. This is also a consequence of the greater and more seasonally consistent connections that Lower region residents have with fishers and itinerant traders not afforded to the other two RDAER regions where river navigation is e xtremely restricted year round. An almost opposite scenario was observed on the Upper region. Here, most households e arned the majority of their non forest based income from contract wages linked to ranching, logging and mining activi ties, and only one Upper resident had a
92 formal job in a governmental organization. These reserve residents clearly presented less connection and had less trust with governmental organizations and NGOs than residents from the two other regions. However, the y had more access to the rancher, miner, and logger sectors than residents of the other two regions, which could explain the higher income from activities linked to those external actors and lower income from government benefits. Finally, Middle region res i dents had the lowest non forest based income in RDAER, even though they had more connections and more trust with external actors than residents of the Upper region. They were well connected with NGOs and governmental organizations and indeed had a higher income from government benefits than Upper residents. However, the lack of social interactions among Middle region residents and private sector actors, and the fact that only three Middle residents trusted actors from those sectors seemed to limit their ac cess to these other sources of income, which contributes to the fact that those residents rely more on forest products (Coomes et al. 2004, McElwee 2008, Duchelle 2009, Duchelle et al. 2011). Several other ecological, social and economic factors affec t hou sehold income in RDAER ( Chapter 2), but social networks and degree of trus t, as determined by geographic location and formal social positions, seemed to also influence total household income in this reserve since it allows reserve residents to access diffe rent external sources of income. Studies have demonstrate d the importance of geographic location on the engagement of rural people in non farm activities because it influences the access to financial and human resources (Berdegu et al. 2001; Lanjouw 2001; Reardon et al. 2001; Ruben & Van den Berg 2001); and empirical evidence of the
93 importance of personal relationships on access to jobs and market opportunities was et al. (2008) tested different types of d eterminants of business income in rural households in Nicaragua, and found that trust, particularly generalized trust (Knack & Keefer 1997), was an important factor determining household income. Accordingly, our findings indicat e the importance of geograph ic location as a factor affecting the social interactions that RDAER residents have with different groups of external actors, and therefore, their access to different sources of income. This study also highlight s the importance of the r ole of formal social positions i n RDAER social networks, and consequently the importance of those social networks and the degree of trust that reserve reside nts have with external actors for household income sourced externally. Utility of Social Network Analysis Applied to Su stainable Use Reserves Sustainable use reserves aim to protect natural ecosystems and promote the sustainable use of natural resources, while fostering economic and social benefits to local communities (Dudley 2008). Success of these reserves is highly dep endent on cooperation among different actors collaboratively involved in reserve management and how residents use reserve resources to make their living. Social network analysis can be a powerful tool to help reserve managers identify: (1) the different s ocial groups operating within the reserve, (2) the social position of reserve residents, (3) interactions between those inside the reserve polygon and external actors, (4) factors that explain the structure of extant social networks, and (5) how those soci al interactions are linked to household sources of income. Sustainable use reserves like extractive reserves, are commonly demarcated in huge geographic areas that comprise multiple subgroups of individuals communities
94 within, what an outsider perspecti ve might define, as one large reserve community. Although these subgroups have their own patterns of social interactions (including informal arrangements of assistance and degrees of trust as observed in the three RDAER geographic regions), they tend to b e lumped together under the same reserve level governance framework. Distinct needs, geographic and/or socioeconomic orientations, as well as the potential conflictive and cooperative elements that may exist between those subgroups are potentially ignored (Cronkleton et al. 2011). Furthermore, ronkleton et al. 2011, p. 467). Within these different socially constructed groups, identification of central households can facilitate reserve management. Highly connected residents disseminate ideas, practices and resources, and can serve as agents to foster collective action. Central individuals have the potential t o influence members of the same social group, but also individuals that pertain to other places and compose other groups (Cheng et al. 2010). T heir social standing can elevate their position such that they can influence individuals outside their particular subgroup (Cheng et al. 2010); they have potentially disproportionate effects on attitudes and opinions of others (Friedkin 1998). Thus, they have the potential to influence natural resource management plans a nd initiatives within reserves. RDAER management, as in most sustainable use reserves globally, involves reserve residents as well as government and non governmental organizations under a collaborative governance structure. Success in these collaborative management
95 settings is contingent upon providing equal opportunities to participate in the decision making process related to rules and norms regarding the use of natural resources (Ostrom 1990), and also access to adequate mechanisms that allow interactions at local scales of governance (Cronkleton et al. 2011). Our network analysis revealed discrepancies in levels of engagement between external actors and reserve residents. Most visits and assistance by external organizations were concentrated on households with easy access because of th eir privileged geographic location and their formal social position related to reserve management. Additionally, the positive correlation between participation of households in courses and representation in AMORA indicates that most households were exclude d from capacity building initiatives t hat took place in this reserve. It is important to highlight that not all forms of social interactions produce the best outcomes for everyone. Associations can serve household residents in many different ways, but they can also perpetuate inequity and encourage conformity, which can constrain the emergence of sustainability (Portes & Landolt 1996). In our case, while the prevalence of social interactions that AMORA representatives in particular had with external actors could be viewed as a problem success of this strategy is highly dependent on the existence of appropriate mechanisms that guarantee the transfer of the knowledge gained and access to external resources to non representatives, thus spreading more effective ly within the reserve Still, the fact that non representatives of AMORA did not even recognize their membership in this association suggests a fundamental lack of awareness of the governance structure created to manage this
96 reserve. It symbolizes perhaps the unequal levels of participation by reserve residents that can ultimately compromise reserve governance. Finally, social network analysis proved to be extremely useful in linking household income, mai nly sourced from forest and non forest based activities, with the social relations that reserve residents have with external actors. Indeed, research and theory emphasize the role of formal structures i Thus, by sharing values developed through organization al culture, formal structures constrain views and behaviors of its members (Deal & Kennedy 1982, Hill & Jones 2000, Hill 2009). Our results showed that reserve residents that are well connected to external actors tended to earn part of their income from ac tivities linked to those external actors. However, how these linkages and income sources may foster or compromise res erve integrity remains unclear. In conclusion, social network analysis can be a powerful tool for managing sustainable use reserves. It she ds light on internal and external social interactions in a management setting highly dependent on collaboration among different actors to guarantee long term viability of these reserves. Furthermore, it is a useful tool that can be integrated with other an alytical tools to capture the socioeconomic and ecological complexity of those sustainable use reserves.
97 Figure 3 1. Social network within the reserve featuring three subgroups represented by the different colors and degree centrality (number of conne ctions an actor has with other actors) for borrowing a canoe, and shaped by regional classification (circle nodes = Lower, square nodes = Middle, and triangle nodes = Upper region). Bigger nodes represent higher degree centrality Figure 3 2. Social network within the reserve featuring three subgroups represented by the different colors and sized by degree centrality (number of connections an actor has with other actors) for borrowing motorboats, and shaped by regional classification (circle nodes = Lower, square nodes = Middle, and triangle nodes = Upper region).
98 Figure 3 3. RDAER household network of number of visits received from representatives of external organizations using Girvan Newman sub group analysis. The number of relations in this netw ork was defined as greater than 3, which means that households that received visits from less than 3 different representatives of external organizations were excluded from the analysis ( these isolates are seen in the column on the left hand side of the fig ure) Degree centrality is represented by node size and measures the number of external actors that visits each household (biggest nodes = high degree centrality); and region by node shape (circle nodes = Lower, square nodes = Middle, and triangle nodes = Upper region)
99 Figure 3 4. RDAER network of households that received assistance from at least two external actor s using Girvan Newman sub group analysis. Degree centrality is represented by node size and measures the number of external actors that provided assistance to each household (biggest nodes = high degree centrality); and region by node shape (circle nodes = Lower, square nodes = Middle, and triang le nodes = Upper region)
100 Figure 3 5. RDAER networks of households that received assistance from external actors. Diagrams are presented separately for each of the three different groups of external actors: governmental institutions, non go vernmental organizations, and private sector. Degree centrality is represented by node size and measures the number of external actors that provided assistances to each household (biggest nodes = high degree centrality); and region by node shape (circle no des = Lower, square nodes = Middle, and triangle nodes = Upper region)
101 Figure 3 6 RDAER networks revealing degree of trust that household have with external actors using G irvan Newman sub group analysis. Diagrams represent degree of trust that househol ds have with the three different groups of external actors: governmental institutions, non governmental organizations, and private sector. Diagrams of governmental and non governmental organizations show households that trust more than two organizations, w hile private sector diagram represents all RDAER households, even the ones that trust only one external actor. Degree centrality is represented by node size (biggest nodes = high degree centrality) and measures the number of external actors that each house hold trusts; and region is represented by node shape (circle nodes = Lower, square nodes = Middle, and triangle nodes = Upper region)
102 Table 3 1. Households and courses, most frequently visited cities, degree centrality of borrowing canoes and motorboats, and number of visits and assistance received by external actor s, and household income.
103 Table 3 2. Descriptive statistics of average household characteristics, forest based and non forest based income, and the total household income in each RDAER region (Lower, Middle and Upper) and p values denoting the signi ficance of regional differences.
104 CHAPTER 4 HETEROGEN E ITY INSIDE T HE EXTRACTIVE RESERVE POLYGON: IGNORING SOCIOECOLOGICAL VARIATION AND LIVELIHOOD OUTCOMES AT THE PERIL OF LONG TERM RESERVE VIABILITY Sustainable Use Reserves and Socioecolo gical Heterogeneity Extractive Reserves have been seen, and also sold, as a successful model that can reconcile human well being, sustainable resource use, and forest conservation. In general, the implementation and management of extractive reserves, and other types of sustainable use reserves, are based on the premise that these areas are relatively homogeneous, which in consequence, promotes the implementation of development and conservation initiatives that are also homogeneous can compromise the long term viability of these conservation units because it neglects the socioeconomic and ecological variation that drives distinct livelihoods that are more (and less) conservation friendly and economically sustainable. In this s tudy, we aim to bring heterogeneity to the fore, with a detailed focus on one Brazilian Extractive Reserve (Riozinho do Anfrsio Extractive Reserve, hereafter RDAER). We also discuss the impli cations of this heterogeneity for resident livelihoods and ulti mately, the viability of the extractive reserve model. Reserve residents have different degrees of access to natural resources, and the resources themselves vary biologically and spatially in these conservation units. S ocial aspects, such as education, lab or availability age of household members time of residence, and consequently, the household knowledge about available natural resources also vary internally. Driving forces also include variations in access to sources of income, different household interests, and the related likelihood to engage in different activities, which are in part related to household social networks ( Chapter 3)
105 and the interactions that reserve residents have with external agents This complex mix of factors determine s reside nt livelihoods and must be integrated into the implementation and management plans of these conservation units for long term reserve viability. L ivelihood choice s are crucial for the success of the extractive reserve model and its dual goals to conserve t he natural resources and biodiversity within these units and to sustain reserve residents. I n this article we aim to: Briefly review the conception of the extractive reserve model. We commend this novel conservation strategy that formally acknowledges that humans and their forest based livelihoods can and should be integrated into conservation agendas. This model largely assumes that all resident livelihoods a re based on non timber harvests tha t could conserve the forest and economically sustain reserve residents Demo livelihoods are diverse. Not all residents rely solely (or even mostly) on NTFPs for income generation. Provide insights into the importance of r ecognizing and integrating this heterogeneity into reserve implementation for long term success. The Extractive Reserve Model How the Model was C onceptualized The creation of the extractive reserve model in the 1980s represents the main achievement of a social movement by Western Amazonian rubber tappers who were struggling to guarantee their rights to their land, natural resources, and traditional way of life (Allegretti 1994; Allegretti & Schmink 2009; Almeida 2002; Hall 2004) The great uniqueness of the extractive reserve model is that it allows the reconciliation of development, environmental conservation and social equity because it defines a territory, a space, and a way to secure livelihood systems of traditional communities over large tracts of w ell conserved forest areas (Allegretti 1989; Keck 1995) The rubber
106 tapper livelihood system, which is based on the extraction of forest products dispersed through hundreds of hectares used by each rubber tapper family, was seen as highly compatible with f orest conservation (Gradwohl & Greenberg 1988; Schwartzman 1989 ). First, this form of land use stands in stark contrast to the dominant Amazonian development model at the time that stimulated conversion of inhabited forests, albeit at low human densities, to pastures and agricultural areas, which resulted in land concentration in the hands of few and violent conflicts. On the conservation end of the spectrum, this emerging model provided a dramatic break from the dominant paradigm that dictated that protec ted areas need be devoid of humans for effective conservation. The concept of extractive reserves integrated conservation and development like never before; they were viewed as a promising solution to secure land rights of traditional Amazonian communitie s and their forest based livelihoods while conserving large tracts of forests ( Fearnside 1989 ; Keck 1995 ; Schwartzman 1989 ) This promising model of reconciling conservation and development rapidly spread throughout the Amazon region, attaining even global implementation. In the Brazilian Amazon alone, approximately 25 million ha of forests (ICMBio 2011) are protected under the principles of this conceptual model. It has been extended to a diversity of traditional populations that encompass rubber tappers a nd other NTFPs harvesters, lowland (vrzea) riverine communities, fishers, and small scale ranchers It also includes social groups that engage in subsistence production, but also embrace commercial activities linked to forests ( Castro 2009). The success o f the extractive reserve model had such an impressive impact on the international conservation community that it precipitated the creation of a new category in the World Conservation
107 Union (IUCN) Protected Area Categories system, the category of Managed Re source Protected Areas (Category VI). Furthermore, IUCN recently changed the name of that category to Protected Area with Sustainable Use of Natural Resources (Dudley 2008), fortifying the idea that traditional management regimes of forest communities can facilitate long term conservation. Scholars have long recognized the biological and sociocultural diversity within sustainable use reserves (Ankersen & Barnes 2004), even in the earlier years of model creation (Allegretti 1 99 4; Hall 1 99 7). This internal bi ological and sociocultural heterogeneity, as well as the regional context in which these reserves are embedded, play important roles in how reserve residents use and manage their natural resources, and on the livelihood portfolios adopted by these resident s. Indeed, scholars have pointed out the need to integrate extractive reserves into the regional economy, and have recognized that agriculture and small scale livestock activities are also part of the traditional livelihood portfolio of reserve residents ( Allegretti 1994; do Rgo 1999). One question debated since the creation of the extractive reserve model has Are th e se reserves socioeconomic ally viable considering their focus on NTFP extraction the livelihood portfolios adopted by reserve residents, since those livelihoods need to link conservation goals with improvement of reserve resident well being. Overall, traditional livelihood portfolios, which include those of sustainable use reserve resi dents, are a combination of resource use activities, small scale livestock, and agriculture (Castro 2009; Coomes et al. 2004). Those activities can have a subsistence or commercial character, and are influenced by several socioeconomic, cultural, and biolo gical factors. Therefore, the complexity of
108 livelihood portfolios found in reserves and the different socioeconomic and ecological factors that influence them are central to the debate of sustainable use and management of those sustainable use reserves. What Socioeconomic and E cological F actors Influence Livelihood P ortfolios? One of the primary factors that directly influence livelihood portfolios in the context of forest conservation is providing secure rights to forest communities, so they have guarant eed access to natural resources (Ostrom 1990; Scoones 1998). This key condition is successfully achieved in the case of sustainable use reserves since the designation of those reserves, and the clear definition of the external boundaries, guarantee secure rights to resident families (Brown & Rosendo 2000; Cronkleton et al. 2011). However, the rights to use natural resources are not the only factor that determines whether traditional residents continue to engage solely or even mainly on forest based activiti es. Access to and quality of natural resources are important factors determining resource use and management (Castro 2009; Duchelle 2009; Newton 2011, Pattanayak & Sills 2001) and therefore, they influence directly the adoption of a more forest based livel ihood portfolio. In addition, other ecological factors, such as habitat type, spatial distribution of natural resources, and seasonality of production are also crucial for the contribution of forest resources to livelihoods in those reserves (Brown & Rosen do 2005; Newton 2011, Duchelle 2009). Key factors are also related to the socioeconomic diversity of reserve residents. The ability of reserve residents to benefit from different opportunities and to overcome inevitable constraints is directly linked to th eir ability to exploit those opportunities. The household structure directly influences the ability of reserve residents to adopt a diversified livelihood portfolio or to incorporate a specific economic activity (Futemma et
109 al. 2002; Perz & Walker 2002), a nd informs its reliance on forest products versus non forest products (Coomes & Burt 2001; Newton 2011; Pattanayak & Sills 2001). Castro households with a larger number of adult members tend to add more activities to t heir economic repertoire In addition, characteristics such as household headship, sex and age of forest users, schooling, and time of residence also influence engagement in forest based activities. Ultimately, they determine whether household livelihood portfolios meet conservation goals (Cavendish, 2000; Pattanayak & Sills 2001; Shone & Caviglia Harris 2006). Access to different livelihood resources is also of particular importance in the context of sustainable use reserves. Differential access that rese rve residents have to financial and economic resources, through organizational issues, institutional arrangements, politics, and power (Castro 2009; Scoones 1998), can contribute or not to the adoption of portfolios that suit sustainable use reserves. Mult iple exogenous socioeconomic factors at larger scales (i.e., regional, national and global) are also key factors that influence livelihood portfolios adopted by reserve residents (Castro 2009; Coomes and Barham 1997; Homma 1992; Mahapatra et al. 2005; Shac kleton & Shackleton 2004). Across the tropics, drivers of deforestation have shifted from rural farmers to major industries and economic globalization (Butler and Laurence 2008). In the Amazon, soy and beef industries are the primary drivers, responding to worldwide economic signals from such seemingly distant factors such as and avian flu ( Nepstad et al. 2006 ). Other factors external to the reserves, such as technological innovation, market access
110 and demand, and policy measures all ultimately impact livelihood adoption within the reserves. Finally, a wide range of risks (such as injury and illness, disability, death, drought and other crop failures, theft, livestock disease and lo s ses, and many others) is direct ly linked to livelihood portfolios, given that households have to respond to those risks ( Cavendish 2003). Livelihood portfolios are also outcomes of trade offs faced by forest households, where differential access to different resources may have negative or positive implications for the adoption of livelihood portfolios that meet sustainable use reserves goals (Scoones 1998). Therefore, livelihood portfolios are constantly being created, negotiated, and adapted through the economic, social and environmenta l changes taking place within and outside reserve polygons. The Changing Livelihood Process Occurring in Extractive Reserves Scholars emphasize the need to take into account the diversity and dynamism of extractive reserve resident livelihoods (Campbell 19 96; Salisbury & Schmink 2007), and the importance of household level factors as drivers of the different livelihood portfolios adopted (Gomes 2001). Recent studies also show that livelihoods of reserve residents are a combination of subsistence activities with market production and wage labor (Campbell 1996; Ehringhaus 2005), and in some cases, this income diversification has indeed improved the livelihoods of reserve residents (Ruiz Prez et al. 2005). ustainable use reserves, resident livelihoods are in a continuous process of change (Ehringhaus 2005; Gomes 2009; Souza 2006; Wallace 2004). The creation of extractive reserves itself, with the provision of services such as basic health, schooling and the emergence of entrepreneurial within reserve cantinas, coupled with new employment opportunities
111 that arose with the insertion of infrastructure and social services, are contributors to economic diversification within reserves (Ruiz Prez et al. 2005). Cons traints related to forest product extraction may be another reason for changing livelihoods. Reserve residents rely on few key NTFPs, dictated partially by the ecological variability and extraction seasonality of any given NTFP, the lack of access to marke ts, and the poor prices of those forest products (Brown & Rosendo 2000; Ruiz Prez et al. 2005). Even though extraction of forest products is still an important activity performed by reserve residents, agriculture is becoming an increasingly important sour ce of cash income (Castelo 1999; Guerra 2004; Ruiz Prez et al. 2005; Salisbury & Schmink 2007). In addition, these same studies have shown that cattle ranching and pensions have gradually expanded in some reserves (Ruiz Prez et al. 2005; Salisbury and Sc hmink 2007; Vadjunec et al. 2009), showing a changing process from forest based to non forest based livelihoods. This transition threatens to compromise extractive reserve goals. These changes are also a clear result of the external economic forces influen cing reserve residents, mainly driven by market demand and prices (Vadjunec et al. 2009). Indeed, research results from one of the oldest and most studied extractive reserve in Brazil, the Chico Mendes Extractive Reserve, showed that it was prisingly rapid land use and land cover changes with cattle and et al. 2009, p. 269). Extractive reserves and other sustainable use reserves are surely an important conservation strategy since these conservation units serve as efficient barriers to deforestation in tropical areas (Nepstad et al. 2006; Vadjunec et al. 2009). However, the livelihood change processes taking place in these reserves suggest that greater focus
112 on the internal socioeconomic and ecological dynamics is needed to address long term reserve viability. H eterogeneity Inside the Polygon: The Case of Riozinho do Anfrsio Extractive Reserve Riozinho do Anfrsio Extractive Reserve (RDAER) was created in 2004 by the Brazilian federal government in response to a violent process of land speculation and deforestation (Campos & Nepstad 2006). The need to secure rights of resident families, coupled with the existence of Brazil nut across this region and the high reliance of the local population on this NTFP, generated great enthusiasm among the social and environmental movements to create this reserve. Located in the municipality of Altamira, Par st population of approximately 300 people in 26 households sparsely distributed along the Riozinho do Anfrsio River ( 2 1) Thus, RDAER is on the low population end when considering the population of other s ustainable use reserves in Brazil It encompasses approximately 736,430 ha of old growth forest, and is commonly divided into three distinct socioeconomic and ecological regions: the Lower, Middle, and Upper regions. This geographic division is mainly dete rmined by biophysical difference s and the different geographic locations along the Riozinho do Anfrsio River, which determine proximity and access to the main surrounding cities (Altamira and Itaituba) (Figure 2 1). Household residents traditionally live on subsistence hunting, fishing, agriculture small scale poultry, and the harvest of forest products ( Rocha et al. 2005). NTFPs and some agricultural products, mainly manioc flour, are sold for cash income and usually to itinerant traders ( regates ).
113 We s et out to evaluate the socioecological heterogeneity within RDAER and how it is linked to the livelihood portfolios adopted by reserve residents. To do this, we: (1) used Brazil nut as a model NTFP species to analyze the ecological and social factors that determine the contribution of this important forest product to RDAER household income, and (2) identified how the connections between RDAER households and different groups of external actors (government institutions, non government organizations, and comme rcial enterprises and cooperatives related to the private sector) influence household incomes. To determine the different sources of income and total household income, we applied a livelihood survey adapted from the Center for International Forestry Research Poverty and Environment Network (CIFOR PEN; http://www.cifor.org/pen/research tools/the pen prototype questionnaire. html ) to 2 3 of the 26 RDAER households. In addition, we used structured interviews and forest inventories to evaluate internal socioecological variation related to Brazil nut use and management. Finally, we used social network questionnaires to identify ho usehold social networks with external actors. For details on the methodology used see Chapters 2 and 3. Even though we found no internal variation in total household income, we did find variation in the contribution of forest vs. non forest based income to total income (Table 4 1 ) This variation was mainly influenced by the spatial distribution and quality of Brazil nut stands, which were linked to access to Brazil nut trees, management practices, and househ old size. Household geographic location was also an important factor determining income in RDAER since it influenced access to external sources of income
114 through the social relations that reserve residents had with acto rs located outside the reserve. Results from the evaluation of Brazil nut internal soc ioecological heterogeneity showed that Brazil nut stand quality, as indicated by tree characteristics (diameter, crown form, crown position, and liana loads), counted fruit production, and reported production, varied among the three reserve regions (Table 4 2 ). In addition, reserve residents from different regions had different access to Brazil nut trees, which influenced their investment on this product as evidenced by management practices performed and number of return trips made to individual trees durin g the same Brazil nut season. All these socioecological factors affected income sourced from Brazil nuts in the reserve and played an important role on reliance of RDAER househo lds on this NTFP. Overall, the further tree stands were geographically located from the mouth of Riozinho do Anfrsio River, the lower their stand quality, which in turn decreased overall productivity. Access to Brazil nut trees was also different among RDAER regions, with residents from the Upper region having the most difficult acc ess to Brazil nut trees in comparison to resid ents of the other two regions. Simply, Brazil nut stands in the Upper region were located further from their homes, making it more time consuming and physically challenging to reach productive trees. The lack of Brazil nut access and the lower quality of Brazil nut stands in the headwaters of Riozinho do Anfrsio River seemed to be the main factors that resulted in minimal investment of Upper region residents on Brazil nut management and correspondingly, limite d use of this NTF P for household income (Table 4 1 ). In addition, some household characteristics, mainly
115 household size and dependency ratio (household consumers: household producers), also influenced resident reliance on Brazil nut in this reserve. Brazil nut harvest is a labor consuming activity and the number of economically active household members directly affected the amount of Brazil nuts harvested. Finally, descriptive statistics suggest that household heads in the Upper region were comparatively yo unger and had lived for less time on their landholdings. Furthermore, they had more formal education, which could also affect their NTFP reliance since forest users with less accumulated forest knowledge, shorter residence times, and more years of schoolin g tend to invest less in NTFPs (Pattanayak & Sills 2001; Shone & Caviglia Harris 2006; te Velde et al. 2006). Interestingly, although stand quality in the Lower region was superior to the similarly lower quality stands in both Upper and Middle regions, it was Middle region residents who reported higher Brazil nut income than either of their regional counterparts. Key household factors did not explain why Lower region residents did not earn more from Brazil nuts than their Middle region neighbors: household size, dependency ratio, age, time of residence, and time lived in landholding were comparable for these two regions. Middle region residents even had poorer access to their Brazil nut trees than Lower region residents. Why did Middle region residents, then earn more from Brazil nuts? Our data suggest that because of their more distant geographic location from the mouth of the Riozinho do Anfrsio River, and consequently lack of access to external sources of income, they invested much more effort in this N TFP than residents of the other two regions. Middle region residents invested significantly in Brazil nut management practices, specifically cutting lianas, a recognized
116 practice that increases Brazil nut fruit production ( Kainer et al. 2007) yet barely pr acticed by Lower and Upper region residents. They also pursued a larger number of return trips to individual Brazil nut trees. The amount of time and effort that they invested resulted in a higher number of Brazil nuts harvested, and consequently, a higher income sourced from this NTFP (Table 4 1 ). One of the main assumptions of the extractive reserve model is that the traditional livelihoods of forest populations meet forest conservation goals since these populations rely mostly on the extraction of forest products or on activities of low environmental impact, such as small scale mostly subsistence agriculture and livestock (Allegretti the livelihood portfolios of RDAER re sidents traditional and complementary to the among the Our results showed that overall RDAER residents indeed mostly rely on forest products and small scale agr iculture to source their income (Table 4 1 ), which are traditional activities recognized as complementary to forest conservation and in the context of the extractive reserve model (Allegretti 1994; do Rgo 1999; Hall 1997). However, the contribution of inc ome sourced externally in this reserve is considerably high er especially considering that RDAER is located in a very remote area, without roads or other infrastructure that provides easy access to the surrounding cities and external income sources. Furthe rmore, RDAER residents were practically isolated from external actors and external sources of income until the 19 80s, when small mining areas opened in RDAER and some reserve residents engaged in this activity (Rocha et
117 al. 2005). Only more recently, land grabbers and loggers reached RDAER and used reserve resident labor to open logging roads and build runways (Rocha et al. 2005). Following the creation of the reserve in 2004, residents began to access government jobs and social benefits. Therefore, the inc orporation of external sources of income, mainly contract wages, formal jobs, and government social benefits to household livelihood portfolios in RDAER, is a relatively recent process. When comparing the livelihood portfolios adopted in the different RDAE R regions, our results showed that Middle region residents presented the most traditional livelihood portfolio in the context of an extractive reserve, since most of their income was sourced from forest products (58%), with a relevant contribution of Brazi l nuts (23%) and from small scale agriculture and chicken husbandry (16%) (Table 4 1 ). On the contrary, forest based income in the Lower and Upper regions was approximately half of that observed in the Middle region (24% and 23%, respectively vs. 58%), whi le the contribution of agriculture and chickens in the Lower and Upper regions was twice as much as in the Middle region (27% and 33%, respectively vs. 16%), mainly in the Upper region w h ere agriculture and chicken were the main source of income to those r esidents. In the Low er region, households sourced 50 % of their income externally, mai nly from government benefits (30 %), but also from formal jobs and tem porary labor wage (20%) (Table 4 1 io in RDAER. The Upper region presented an intermediary livelihood portfolio in comparison to the other two RDAER regions. In this region, most household income was traditionally sourced, from agriculture (33%) and forest product extraction (23%).
118 However, formal jobs and wage labor from temporary jobs and gold mining income (totaling 31%) indicate a high regional reliance on income sourced externally. Indeed, results from our social network analysis with external actors ( Chapter 3) sh owed that household geographic location influenced access that reserve residents had to different sources of income. Households from different RDAER regions had varying degrees of access to the most proximate cities of Altamira and Itaituba, which influenced their social netw orks wit h urban external actors (Table 4 2 ), and thus, their access to external resources. While residents from the Upper region had better access to Itaituba, and Lower region residents were more connected to Altamira, Middle region residents had no acces s to Itaituba, and less access to Altamira. Consequently, this geographic constraint limited their access to external sources of income, as indicated by the negligible contribution of government social benefits and the other sources of income to Middle reg ion residents (Table 4 1 ). Similarly, Upper region residents were also constrained in accessing government social benefits becaus e of their geographic location. Although well connected to Itaituba, they were further away from Altamira, the city that provid es government social benefits. As a result, Upper region residents had poorer social relations with the Altamira based government and non governmental organizations. However, Upper residents had better connections with external actors located in Itaituba, who were mostly linked to the private sector, and consequently, Upper region residents earned a higher income from wage labor and temporary jobs linked to those private sector actors. Among all RDAER regions, Lower region residents had the best access to A ltamira during the entire year. This better access allowed them to cultivate better
119 connections with most external actors studied, particularly those linked to government and non government organizations. As a result, Lower region residents earned the high est income from government social benefits, and also considerable income sourced from formal jobs linked to government organizations located in Altamira. Social network analysis results also showed that formal social position was linked to the social relat ions that RDAER households had with external actors. Residents who were elected representatives of the local RDAER association, AMORA, or had formal positions with external organizations such as a health care worker, had greater access to external actors, and overall received more visits and assistance from those external actors ( Table 4 2 ). Indeed, by the nature of holding a formal social position, these employed residents had to have more frequent access to external actors. These households also displayed a higher degree of trust toward external actors than households without formal social positions. Furthermore, the degree of trust that households with formal social positions showed toward external actors was linked to non forest based income ( p < 0.01). These findings indicate that formal social position could favor access to external resources provided by external actors with whom residents are formally linked. Findings from this study demonstrate that key factors that influence household income and henc e livelihood portfolios adopted by reserve residents include, Brazil nut socioecological heterogeneity, and variation of household social networks with external actors, as well as the determinants of these social networks. The long term viability of RDAER, and indeed other extractive reserves, depends to a high degree on the
120 livelihood portfolios adopted within those polygons, and how those portfolios meet conservation goals while improving resident well being. Implications for Sustainable Use Reserves Sust ainable use reserves tend to encompass larger territories and combine multiple communities that use and manage different forest resources. The livelihood strategies adopted in those conservation units, even in a reserve such as RDAER with only 26 household s, are diverse since they are the result of resident transformation of the available resources into goods, services and income. This transformation is highly dynamic and can be achieved through a variety of strategies. Therefore, the success of sustainable use reserves as a conservation strategy is related to its capacity to accommodate this sociocultural and ecological diversity. Results from our study in RDAER confirmed that livelihood portfolios adopted by reserve residents were linked to the socioecolog ical heterogeneity that this reserve encompasses. We showed that the context of each of RDAER regions is a key determinant of the availability of resources that reserve residents can use to pursue their livelihoods, and these contexts influence the livelih ood portfolios adopted. Reserve managers and conservationists that aim to foster livelihood portfolios that are suitable to the context of sustainable use reserves such as RDAER, need to understand the internal socioecological contexts and how they affect the adoption of diverse livelihood portfolios. Furthermore, this internal heterogeneity needs to be integrated into the reserve manageme nt plan and its implementation. Local Specific A pproaches We argue that a primary aspect related to interventions that a im to foster suitable livelihood portfolios in sustainable use reserves is the understanding of the local specific
121 contexts of those reserves. Although conservation and development initiatives taking place in sustainable use reserves frequently build upon traditional natural resource uses and management, addressing the internal complexity and heterogeneity of these polygons in the design and implementation of these interventions is critical to their long run viability. Detailed assessment of the socioeconom ic, ecological, and cultural factors influencing livelihoods is necessary, including how the diverse livelihood systems within these reserves function or fail to function. Qual itative and quantitative data are required to assess the variety of livelihood p ortfolios adopted and the outcomes that they produce, most of which can be assessed through secondary data sources. However, the specific contexts of different reserves also demand an investment in local field research (Ahenkan & Boon 2010; Cavendish 2003; Meikle et al. 2001 ; Sallu et al. 2009 ). This assessment will help managers and conservationists to gain a better understanding of the socioeconomic, ecological and cultural factors that contribute to or constrain the adoption of different livelihood portf olios. The RDAER study presented in this article, and many other studies ( Boxman et al. 1991 ; Cavendish 2000 ; Gomes et al. 2012; Janse & Ottitsch 2005 ; Newton 2011; Shone & Caviglia Harris 2006 ), provides insights into some of the socioecological factors that affect livelihood adoption. Because of the socioecological diversity that sustainable use reserves comprise, we emphasize the need for local specific research to guide local specific interven tions. The scale at which these assessments take place is also of great relevance when analyzing the adoption of different livelihood strategies, since even within the same household, different individuals may adopt different livelihood strategies to pursu e their needs. The degree of household engagement in different livelihood strategies is related
122 to their skills, knowledge, physical capability, social networks, and the financial and economic resources available. This same rationale can be applied at the household, community, and regional scales; and because each of these scales presents a unique dynamic that produces differentiated outcomes, these different scales need to be considered within the context of sustainable use reserves and the livelihood port folios adopted (Scoones 1998). Following an assessment of local specific information, managers and conservationists involved in reserve implementation and management, would be able to identify the strengths and weaknesses of the different portfolios adopte d within households, but also within different reserve regions. Or in other words, it would help to identify the variety of factors operating at different scales (individual, households, or communities), and which livelihood portfolios meet sustainable use reserve goals at differentiated levels. This information can thus guide the adoption of local specific interventions that integrate the internal local heterogeneity of these polygons into their implementation and management plan. Furthermore, such an appr oach would help managers to address the constraints that limit adoption of livelihood portfolios that meet conservation and development goals. Another important aspect to be taken into account is that livelihood strategies are a process and not a simple en d outcome (Meikle et al. 2001). Because livelihood portfolios are a result of the social and ecological relations taking place on those polygons, they are continually changing and adapting to the specific conditions that reserve residents face. Livelihood strategies have the primary aim to address reserve residents needs and interests, and thus, assessment and intervention needs to be a
123 consultative and participatory process. Interventions also require a level of flexibility and adaptability that provide r oom to accommodate the changing process that takes place on these reserves. Scholars have also emphasized the importance of recognizing the socioeconomic aspects that promote changes in forest user identities (Vadjunec et al. 2012), and the different cultu ral and value systems associated with the adoption of different livelihood strategies ( Berkes & Davidson Hunt 2006 ). The expansion of urban areas and infrastructure around sustainable use reserves increases the connection between reserves and the surroundi ng cities. Furthermore, policies that favor and promote the expansion of activities that are not compatible with sustainable forest uses, influence the cultural values and identities of reserve residents, and thus affect their livelihood portfolios ( Vadjun ec et al. 2012). Taking into account those important aspects, and also the findings of the RDAER study, we advocate that conservation and development interventions within sustainable use reserves acknowledge and incorporate: The variety of livelihood portf olios adopted by reserve residents; The complexity and local specificity of these livelihood portfolios; The different scales at which the socioecological factors operate (individu al, households, and community); The social relations related to livelihood portfo lios at these different scales; The cultural values related to the adoption of livelihood portfolios and the conflicts that may be generated by divergent value systems; The continued and dynamic process of change in livelihood portfolios adopted; The adoption of interventions that allow flexibility and the ability to respond to these changes;
124 Interventions that aim to remove the constraints that limit adoption of livelihood portfolios that meet conservation and development goals; and Investments to se ek compatible and alternative future livelihood portfolios if current one s do not meet reserve goals. Institutional Mechanisms for the Governance of Heterogeneous Polygons Many scholars have emphasized the importance of robust institutions for community ba sed natural resource management and aspects related to the governance of protected areas, such as sustainable use reserves (Agrawal 2007; Agrawal & Gibbson 1999 ; Cox et al. 2010 ; Ostrom 1999). It is not the aim of this article to provide a detailed discuss ion on institutions and principles underlying community based natural resources management, but instead to focus on some important aspects related to the institutional framework adopted for extractive reserve governance, and implications for the livelihood strategies of residents and the long run viability of these reserves. The institutional framework set up for extractive reserves is a crucial mechanism for the long run viability of these reserves (Hall 1997) since it influences the socioeconomic activity and behavior of residents ( Agrawal 2001 ; Ostrom 1990), and lastly, affects the economic repertoire that they adopt (North 1999). The formal regulations and procedures, organizational issues, politics, and power established by the institutions operating on these reserves are important determinants of how reserve residents can use and manage their natural resources, and also access financial and social resources to meet their needs ( Agrawal 2005 ; Gautam & Shivakoti 2005; Scoones 1998). In extractive reserves as in other sustainable use reserves, the institutional framework was designed to guarantee participation of local communities in reserve decision making processes (Mannigel 2008). Therefore, it is based on a co
125 management system where the local communit ies, formally represented by their local associations, the federal environmental agency (Chico Mendes Institute of Biodiversity Conservation ICMBio), and other stakeholders, are jointly responsible for reserve implementation and management (Brasil MMA 20 00 ). Even though the aim of this co management system is to assure participation of local communities in reserve governance, scholars have pointed out that the institutionalized rules that were created reduce the autonomy of reserve residents because they limit their management rights (Cronkleton et al. 2011; Cunha 2010; Pacheco 2010). Furthermore, current structures focus on local organizations (associations formally recognized by the reserve model) which lack autonomy and external agents tend to make the major decisions regarding natural resources use (Cronkleton et al. 2011; Cunha 2010; Pacheco 2010 ). The institutionalization of the extractive reserve model surely represents an important achievement for Amazonian fore st users since it secures the rights of land and natural resources that those populations depend on for their livelihoods. However, it was conceived under the assumption that reserve residents were relatively homogeneous communities, and thus would engage cooperatively on natural resource management to maintain their forest based livelihoods ( Cardoso 2002 ). Indeed, and as stated by Turner (1999 immutable group of people jointly managing a deli mited common resource through As illustrated in our RDAER study findings, these reserve polygons cannot be of heterogeneous systems, wit h different social groups operating at different scales and
126 using a variety of resources and livelihood strategies independently (Cardoso 2002; Cronkleton et al. 2011). However, the institutional framework adopted on those reserves tends to fortify the id ea of homogeneity as attested by at least two formal mechanisms required on management and implementation, and the way that they have local agencies to represent rese rve residents. The utilization plan The Utilization Plan is a formal mechanism that regulates the use, management, and monitoring of natural resources (See Cardoso 2002 for a detailed description of the legislation and the formal procedures of extractive r eserves). Local agencies must exist even before reserve creation, because local communities can only request extractive reserve denomination through organizations that formally represent them. These are also the institutional mechanisms whereby reserve res idents participate in reserve governance (Brown and Rosendo 2000 ). Even though these two mechanisms secure the rights of reserve residents over land and natural resources and regulate the use of those resources that they depend on, in most cases, they do n ot provide enough room for diversity since they do not take into account the socioeconomic and ecological heterogeneity of those polygons. Indeed, Cardoso (2002) questioned the effectiveness of these mechanisms in her analysis of the robustness of common p roperty institutions in Chico Mendes Extractive Reserve, in Acre. She showed that the institutional framework and the mechanisms that govern the use of natural resources in that extractive reserve had a differential impact on independent rubber estates (or seringais ) mainly because they neglected the differences among those estates. This impact may be negative or positive for natural
127 resource management (Cardoso 2002), and therefore, it could also have negative or positive impacts on reserve resident liveli hoods. As also exemplified by the RDAER results presented here, it is crucial that the Utilization Plans of extractive reserves take into account the local and internal socioeconomic and ecological diversity when establishing rules and norms that regulate natural resource use (Agrawal 2002; Cox et al. 2010; Ostrom 1992). In this sense, we suggest the disaggregation of the Utilization Plan into different and perhaps more flexible sets of rules and norms that address the internal heterogeneity of those polygo ns. These different sets of rules should be based on the local biophysical, ecological, and socioeconomic conditions of the different areas that comprise these large conservation areas ( Agrawal & Chhatre 2006; Tucker et al. 2007; Schlager et al. 1994 ). Pla ns also need to take into account the different forest uses of the social groups that inhabit those reserves, and engage them directly in making the rules that apply to reserve (or their region of the reserve) governance. Participation by residents in int ernal rulemaking has been found to increase the probability of obtaining positive forest conservation and livelihood results (Persha et al. 2011). Finally, managers need to consider the social relations that residents have with external agents and how thos e relations affect their engagement in different economic activities that compose their livelihood portfolios (Sick 2008; Ostrom 1990). Thus, ideally a zoning strategy approach should be used to identify the different socioecological regions that exist wit hin those polygons as applied in the Ja National Park in the Brazilian Amazon (Oliveira & Anderson 1999), and subsequently different sets of rules could be used to regulate the use of natural resources in those different zones by the social groups that op erate in those areas.
128 Local agencies Local agencies are undoubtedly an important mechanism for reserve residents since they provide a spa ce for residents to discuss their own needs and priorities, and serve as a channel to communicate those needs, prioriti es, and views on issues that affect them. In this way, they generate information that can be useful in the decision making process related to livelihood strategy adoption. They are also an important mechanism to disseminate useful and important information related to reserve management regulation, relevant within the context of natural resources management and livelihoods. Hall (1997) argues that local agencies are a key component for the success of extractive reserves since they ensure that reserve residen ts exercise effective control over natural resources. However, if these local agencies do not acknowledge the heterogeneity related to the social groups, power inequalities, and different interests operating within reserve, they may not provide egalitarian benefits or serve residents positively (Cox et al. 2010; Korten 1980; Scott 1998 ). Furthermore, if those local agencies are not self organized and lack the autonomy required to ensure that reserve residents maintain control over natural, financial and soc ial resources, then their effectiveness can be compromised. In our RDAER study we found important relationships between household income, degree of trust, and formal social positions linked to the local RDAER association (AMORA). These relationships result in unbalanced access to external resources among AMORA and non AMORA representatives. In this way, they highlight the importance of evaluating the impact of institutional mechanisms on the adoption of livelihood portfolios and the distribution of services and benefits among reserve residents. Within the context of sustainable use reserves, primarily because of their
129 internal complexity and heterogeneity as well as the large areas that they comprise, local agencies and mechanisms that regulate reserve imple mentation and management need to be designed to address the different problems and demands existing in those reserves. In addition, scholars pointed out that trust, transparency, and legitimacy are important conditions for the functionality and effectivene ss of institutions related to natural resource management (Cox et al. 2010; Harkes 2006 ) mainly because they favor engagement in natural resources management and monitoring activities. Thus, the adoption of several small local agencies to represent reserve residents might be preferable than the adoption of only one central local agency that represents the entire reserve (Cardoso 2002), since they could address the socioecological heterogeneity of those polygons. In RDAER we could envision at least two local agencies, if not three, to compose the institutional framework of that reserve: one to represent Upper region residents, and another for the Lower and Middle regions residents. We recognize that RDAER has only a small number of households and it may be costly to have more than one local agency to represent this limited population. However, these households are distributed in a huge area, and indeed are located in different socioe cological zones and represent different social groups operating in this reserve. Therefore, we believe that the adoption of an institutional framework with this type of configuration could provide many advantages for the extractive reserve governance and t heir residents and overcome the costs to create and maintain these local agencies. Some of the advantages envisioned are the following: Association representatives may be more efficient in their tasks since they will be more aware of the problems and dema nds of their associates and the local conditions related to them, which would contribute to their ability to respond to
130 those demands and issues, and not have to spend time with issues in which they are not knowledgeable or do not recognize as important; T he dissemination of information among members of those local associations would be facilitated since residents that share the same place are usually well connected with each other because of their physical proximity; This configuration would favor the occu rrence of meetings and other spaces of discussion because associates share the same place and would be abl e to participate more actively; A more equitable representati on would be promoted since groups that share the same place and face the same livelihood conditions would have more trust in local residents to represent their needs; Consequently, it would reduce the number of conflicts generated by different interests, opinions, and representativeness and power inequalities, which emerge in local agencies th at represent innumerous different social groups; Local agencies that represent a well connected group of residents who share the same physical space may also deal with fewer and less complex issues. Considering the need that reserve residents have to adapt to such new institutional mechanisms, and also the time that it requires, local agencies that are small and function in a simple way, would also contribute to the learning process on adapting to new institutional mechanisms; and Management efficiency and accountability may also be favored since local agencies would coordinate and search for financial support for specific activities that respond to a small, similar group of residents. It may prove to be an advantageous strategy since access to small and spe cific funds may be easier through different mechan isms and from different donors. We recognize that these insights relate to only two institutional mechanisms that shape the governance of extractive reserves. Many other institutional aspects are important in the context of sustainable use reserves. However, these two mechanisms are a crucial part of reserve governance because they directly affect the livelihood strategies adopted by reserve residents by: (1) regulating the use and management of natural reso urces, and (2) facilitating greater participation and sharing of information within the reserve an d with other stakeholders. These are two recognized conditions for
131 the long term reserve viability (Brown & Rosendo 2000; Gibson 2001; Mannigel 2008; Potetee & Ost rom 2004). Challenges and Benefits of T aking into A ccount this S ocioecological H eterogeneity Sustainable use reserves are a considerably new conservation strategy that has grown immensely in the last years. Concurrently, we see an increase in the poli tical and financial investments made on these reserves. However two decades after the creation of this model and following massive international and domestic investments applied to these conservation units, the development im provements in these reserves ha ve been relatively slow Furthermore, the limited evidence that traditional livelihood strategies can transform into alternatives that complement conservation goals puts in question the viability of these units. This scenario may indicate a certain degree of failure of the conservation and development interventions taking place in these reserves, as a consequence of limited integration of multi level socioeconomic and ecological heterogeneity that encompass these reserves. Indeed, there is an emerging consensus that adoption of interventions and institutional mechanisms that integrate the internal heterogeneity of those areas is central to achieving positive conservation and development outcomes In addition, such mechanisms must promote an egalitarian distribution of benefits and empower local communities through their participation in the decision making and governance processes of these conservation units ( Campbell et al. 2010; Cronklenton et al. 2011 ). The governance of sustainable use reserve is not an easy task considering the remoteness and large areas that these polygons comprise, the socioecological complexity involved in natural resources management, and the interactive effects of
132 numerous variables that operate within and across multiple socioe conomic and environmental scales in those reserves. Thus, we recognize that the adoption of an integrative approach as suggested in this article is extremely challenging. Furthermore, we emphasize that the adoption of this integrative approach must complem ent rather than replace conventional conservation and development approaches taking place in those areas. An integrative approach in the context of sustainable use reserves will require tha t practitioners and managers be willing to invest a significant amo unt of time and resources within these conservation units to bui ld the trusting relationships and long term commitment s The investment of managers in building trustworthy relationships with local communities can allow for a more nuanced understanding of t he socioecological interactions taking place in those reserves. Also, such relationships can empower and promote the participation and engagement of the local communities in reserve governance. A more consistent, physical presence within these reserves wou ld also facilitate a process of cooperation among managers and local residents to find joint solutions that address felt needs. Thus, one of the main challenges for the application of these integrative approaches is short handedness of qualified staff in g overnment and non governmental agencies engaged in reserve governance. The problem here is not a lack of qualified professionals, but the lack of financial resources to hire these professionals. Currently, most of available funds for conservation and devel opment interventions are directed to investment in infrastructure and short term initiatives. Governance of sustainable use reserves is a long term adaptive process, which implies
133 a necessary long term commitment and inve stment of stakeholders and professi onals in conservation and development interventions. Another important challenge is the lack of skills among lo cal people related to engaging i n these integrative approaches and interventions. Resident communities of these reserves usually are illiterate, have cultural and language barriers that limit their ability to express their perspectives, and lack a history of cooperation to solve common problems. These conditions apply to many Amazonian extractive reserves. Within this context, natural resources man agement is based on i ndividual families rather than o n joint activities that involve the populations within those reserves. Because sustainable use reserve governance implies a learning and adaptive long term process, it also requires that managers and pra ctitioners understand and acknowledge the constant socioeconomic and ecological changes taking place in these reserves across temporal and spatial scales. Such a dynamic context demands flexibility to adequately respond to these changes. Many conservation and development initiatives taking place on sustainable use reserves are highly dependent on earmarked funds, which mean that they are unable to spend their financial resources on activities that were not previously specified and approved by donor institut ions. This certainly represents an important challenge for most of organizations involved in these initiatives and on reserve governance. We recognize that with creativity and persistence this has been successfully achieved in some reserves in the Amazon ( e.g. in the Mamirau sustainable use reserve, which has managed to have long term goals and the corresponding funding that goes with it). However, this is not the case in many sustainable use reserves. Thus, it is important that those organizations start a dialog
134 with donor institutions to guarantee that conservation and development funding allow for more flexible and ad aptable approaches to spending. Considering the lack of information about how the systems of sustainable use reserves function, a higher in vestment in social and ecological research is also required for the sustainable viability of these conservation units. Nowadays, most of the investment and effort expended within these reserves is on conservation and development initiatives that are rarely based on the local socioecological conditions of these reserves because of the lack of this specific information. Indeed, there is not a formal or mandatory requirement for local specific socioecological research for the formulation of an extractive reser ve Utilization Plan. In addition, most organizations (governmental and non governmental) that compose the Management Council of sustainable use reserves do not have a strong research tradition. These circumstances constrain research activities in reserves, because they must rely on third party institutions or even hired individual researchers to carry out such local specific research. In the case of Brazil, a formal collaboration among government agencies and NGO members of a reserve Management Council, spe cifically the ICMBio, with federal and state universities could be a good alternative to overcome this challenge. Masters and PhD students could carry out systematic research in these conservation units that could serve the development and conservation in terventions taking place in those units. This type of arrangement would guarantee that the local specific research would target the main needs and problems faced by reserve residents and organizations involved in the implementation and management of these conservation units. Collaboration would also facilitate the acquisition of official authorization required in order to carry out rese arch in
135 these protected areas. Authorizations could be provided through the federal and state universities collaborating wi th ICMBio. Finally, such agreements would also ensure that ICMBio maintained control over the research carried out and the data generated in these reserves. We recognize that these collaborations are opportunistically occurring in some extractive reserve s in Brazil; however, this is not a common practice. We have argued in this article that the success of sustainable use reserves depends on the integration of the socioeconomic and ecological local specific conditions and the multiple scales implied in the s ustainable use reserve governance. This integration would ensure that reserve residents are the main players and beneficiaries of the governance processes taking place in the reserves. It would guarantee the efficient investment of scarce human and financi al resources in interventions that truly address the specific conservation and development needs of these reserves. And finally, it would facilitate the creation of robust institutional mechanisms that respond to the specific conservation and development c ontexts and needs of those conservation units. We also recognize that major challenges exist to the integration of complexity and heterogeneity within reserve implementation and management. However, we believe that long term reserve viability would be bett er served if stakeholders involved in reserve governance would invest in efforts to (1) acknowledge and integrate socioecological heterogeneity and (2) develop participatory bottom up approaches that take into account the multiple scales necessary to gover n these reserves.
136 Table 4 1. Contribution of different sources of income to household total income in the three RDAER regions and p values denoting the significance of regional differences
137 Table 4 2. RDAER internal socioecological heterogeneity (positive signs indicate intensity)
13 8 CHAPTER 5 CONCLUSION Sustainable use reserves have been touted as a land tenure model that reconciles development and conservation. Th e central assumption behind these reserves is that the traditional management regimes of reserve residents are sustainable, and thus the standing forests on which they depend for their livelihoods will also persist in the long term. Therefore, adoption of livelihood strategie s that meet conservation and development goals is a crucial factor for the long term reserve viability. The three manuscripts th at form this dissertation explore s the socioecological factors that determine household income in RDAER, and how the socioecolog ical heterogeneity existing in this extractive reserve links to different livelihood portfolios adopted by reserve residents. Findings from the first manuscript (Chapter 2) support the initial assumption that Brazil nut ecological and socioeconomic heterogeneity exists in Riozinho do Anfrsio Extractive Reserve, and that this internal heterogeneity determines the use and management of this important forest product. Resource quality and access are important aspects driving Brazil nut use and management, and lastly in shaping its contribution to household income. The socioeconomic differences found among RDAER residents, and its impacts on Brazil nut use, management and income, confirm the importance of taking into account household socioeconomic variables when analyzing the contribution of forest resources to household income. Furthermore, our results showed that even though household income is relatively even across all region s of RDAER, a more detailed analysis of the contribution of forest products and Brazil nut to household incom e, coupled with the geographic ecological and socioeconomic aspects
139 that affect household income in that reserve, revealed a more complex economic s cenario. This complexity resulted from internal RDAER socioecological heterogeneity, which is rarely incorporated into management plans and supporting conservation and development initiatives and policies in these reserves. In the second manuscript (Chap ter 3) findings showed that social network analysis proved to be a powerful tool to identify: (1) the different social gro ups and their interactions within RDAER (2) the social position of reserve residents, (3) interactions between those inside the reser ve polygon and external actors, (4) factors that explain the structure of extant social networks, and (5) how those social interactions are linked to household sources of income. All these aspects are extremely important in the context of sustainable use r eserves since the success of those conservation units is highly dependent on cooperation among different a ctors collaboratively involved i n reserve management. The social network structures within the reserve, and with external actors, showed that geograph ic location was the primary factor defining social networks in RDAER. Within the reserve, geographic location constrained interactions among reserve residents of different RDAER regions, and also determined the number of households that had a central posit ion as providers of supplies and equipme nt. As geographic isolation increased, the number of centralized households decreased Geographic location and access to surrounding cities seemed also to influence comparative levels and types of assistance received by RDAER residents, which lastly influenced livelihood portfolios adopted in the three RDAER regions. Social networks of reserve residents with external actors were also linked to the formal social positions that reserve residents had with external organi zations or with the local RDAER association
140 (AMORA). Residents who were elected representatives of AMORA were particularly important actors both within the reserve and with external actors. Our findings also showed that visits and assistance received by ex ternal actors were concentrated on households that were AMORA representatives and on the few households that had a permanent job linked to external actors. These same hous eholds also showed more trust in external actors than households that had no formal s ocial positions. Finally, social network analysis also proved to be extremely useful in linking household income, mainly source d from forest based and non forest based activities, with the social relations that reserve residents had with external actors. T hese results showed that reserve residents that are well connected to external actors tended to earn part of their income from activities linked to those external actors, such as government pensions However, further studies are needed to determine how the se linkages and income sources may foster or compromise reserve integrity. Finally, the third manuscript showed that livelihood portfolios adopted by reserve residents were linked to the socioecological heterogeneity and the social relations that reserve r esidents had with external actors. Therefore, reserve managers and conservationists that aim to foster livelihood portfolios that are suitable to the context of sustainable use reserves such as RDAER, need to understand the internal socioecological context s and how they affect the adoption of diverse livelihood portfolios. This internal heterogeneity needs to be integrated into the reserve management plan and its implementation. We conclude that a detailed assessment of the socioeconomic, ecological, and cu ltural factors influencing livelihoods is necessary, including how the diverse livelihood systems within these reserves function or fail to
141 function. This assessment will help managers and conservationists to gain a better understanding of the socioeconomi c, ecological and cultural factors that contribute to or constrain the adoption of different livelihood portfolios, and thus guide the adoption of local specific interventions that foster livelihood portfolios that produce positive outcomes for these conse rvation units. We also argue that some aspects of the institutional framework adopted for the governance of sustainable use reserves need to provide room to accommodate the socioecological diversity, which define these polygons. By analyzing two institutio nal mechanisms of the extractive reserve model ( the Utilization Plan and local agencies) we provided insights of how those mechanisms can be adapted to accommodate socioecological heterogeneity. Thus, we suggested the disaggregation of the Utilization Plan in different sets of rules and norms to address the local biophysical, ecological, and socioeconomic conditions of the different areas that comprise those large conservation areas. In addition, we suggested that the adoption of several small local agencie s to represent reserve residents might be preferable to adoption of only one local agency that represents the entire reserve. We recognize that sustainable us e reserves governance by itself is not an easy task considering the remoteness and large areas of these polygons, the socio ecological complexity involved i n natural resources management, and the interactive effects of numerous variables that operate within and across multiple socioeconomic and environmental scales in these reserves. Adoption of a kind of integrative approach as suggested in this disserta tion is extremely challenging. It implies proactive involvement of the different stakeholders engaged in reserve governance, and a great er investment to identify and integrate socioecological
142 heterogene ity as well as a participatory bottom up approach that takes into account the multiple scales needed to govern these reserves. However, adoption of this integrative approach would allow reserve resi dents to rise to the top as key players and beneficiaries of the governance processes taking place in these reserves. It would guarantee more efficient investment of the scarce human and financial resources on interventions that truly address the specific conservation and development needs of these reserves. Fina lly, it would facilitate the creation of robust institutional mechanisms that respond to the specific conservation and development needs of these novel integrated conservation and development units.
143 APPENDIX A LA NDHOLDINGS SELECTED FOR FOREST INVENTORIES (RED) AND BRAZIL NUT TREES MAPPED IN TRAILS AND TRASECTS IN 200 9 (GREEN) AND 2010 ( LIGHT GREEN ) (PORTUGUESE)
144 APPENDIX B BRAZIL NUT QUESTIONN AIRE (PORTUGUESE) Data: Nome da Localidade: Nome do Entrevistador: Nome do Informante: 1. Se respondeu que no nasceu na Reserva perguntar quando chegou? Voc sempre viveu nesta colocao? (0=no, 1=sim) Se no, em que outros locais da Reserva viveu antes de chegar aqui? (1 0) 2. Qual o tamanho aproximado da sua colocao? (metros, linhas, braas, etc.) 3. Como voc se tornou dono desta colocao? (1=herdada, 2=comprada, 3=trocada, 4=outros) De quem? (1 4) 4. Desde quando voc e sua famlia exploram castanha nos castanhais de sua colocao? 5. Quem explorava estes castanhais antes de voc e sua famlia? (1=pai, av, etc., 2=desconhecido, 3=outros) (1 3) 6. Alm de sua famlia quem mais tem o direito de coletar castanha nos castanhais localizados em sua colocao? 7. Como voc, sua famlia ou a comunidade decidem quem pode coletar castanha nos castanhais da Reserva, ou seja, nas reas comunitrias e nas colocaes de cada famlia? 8. Como foram definidos os limites atuais das colocaes de cada famlia e das reas comunitrias e quem fez esta delimitao? 9. As famlias e todos os membros da comunidade reconhecem os limites destas reas? (0=no, 1=sim) (1 0) 10. Voc coleta a castanha em outras reas da Reserva e que no sejam dentro da sua prpria colocao? (0=no, 1=sim) (1 0)
145 Se sim, em que reas? (1=reas comunitrias, 2=colocaes de outras famlias, 3=outras) Se no, passe para a questo # 17 (1 3) 11. Quantas reas comunitrias para a coleta da castanha existem na Reserva? Qual o nome destas reas? E qual o tamanho destas reas? Em quais destas reas comunitrias voc e sua famlia coletam castanha regularmente? # reas 12. Como a sua famlia e as outras famlias se organizam para fazer a coleta da castanha nestas reas comunitrias e nas colocaes de outras famlias? 13. Como e por quem feita a coleta da castanha nestas reas? (1=famlia, 2=coletiva, 3=gnero, 4=faixa etria) (4 1) 14. Como feita a partilha/diviso da castanha coletada na sua colocacao, nos castanhais comunitrios e nas colocaes de outras famlias? 15. Existe alguma diviso de castanhais ou de reas de coleta definidas dentro destas reas em que cada famlia ou membro comunitrio deve coletar? (0=no, 1=sim) (1 0) 16. Ou todos podem coletar castanha em todos os castanhais localizados nestas reas? (0=no, 1=sim) Se sim na #13 como isto decidido e quem decide que famlia ou membro da comunidade pode coletar em cada um destes castanhais? (1 0) 17. Existe algum limite mximo na quantidade de castanha que pode ser coletada nestas reas (comunitrias ou outras colocaes) por safra por famlia ou por membro da comunidade? (0=no, 1=sim) Se sim, qual esta quantidade? E como e quem decide esta quanti dade mxima? (1 0)
146 18. Existem atividades especficas e obrigatrias que as famlias tm que fazer quando coletam a castanha nas reas comunitrias ou em colocaes de outras famlias? (0=no, 1=sim) Quais so estas atividades? (1 0) 19. Existem reas comunitrias para outros produtos da floresta? (0=no, 1=sim ) Quais so estas reas e para que produtos? (1 0) 20. Voc usa a sua colocao para outras atividades produtivas e para coletar outros produtos da floresta alm da castanha? (0=no 1=sim) (1 0) 21. Qual a rea que roou no ano passado (2009)? 22. A que distncia est a roa da sua casa? 23. O que plantou na roa (alm de mandioca)? 24. Quantas pessoas trabalharam na roa? So todos membros da famlia? (0=no, 1=sim) # homens e mulheres; (1 0) 25. Qual o trabalho dos homens na roa? 26. Qual o trabalho das mulheres na roa? 27. Que instrumentos utiliza pra fazer a roa? 28. Quantas pessoas trabalharam na fabricao da farinha? So todos membros da famlia? (0=no, 1=sim) # de homens e mulheres (1 0) 29. Que instrumentos utiliza pra fazer a farinha? 30. Que dificuldades voc tem para trabalhar na roa? (qualidade do solo, falta de gente pra ajudar, no tem tempo porque est envolvido em outras atividades, falta de apoio tcnico, falta de sementes, outras)
147 31. Quantas pessoas trabalham na pesca. So tod os da membros da famlia? (0=no, 1=sim) # de homens e de mulheres (1 0) 32. Que instrumentos usa pra pescar? 33. Qual o produto da floresta que voc considera como o mais importante pra sua famlia? (a importncia pode ser em termos de renda, de gosto, de facilidade de trabalhar, etc.) Coleta e manejo da castanha 34. Voc saberia dizer quantas castanheiras existem dentro de sua colocao? Deste total quantas castanheiras voc explora em mdia por safra? # rvores # rvores 35. Existem castanheiras da sua colocao que nunca foram exploradas? (0=no, 1= sim) Se sim, quantas e por qu? # rvores 36. Voc saberia dizer qual a mdia de frutos (ourios) produzidos por cada castanheira da sua colocao por safra? # Frutos 37. Voc coleta castanha nas mesmas castanheiras ou castanhais mais de uma vez em uma nica safra de castanha (em um nico ano)? (0=no, 1=sim) (1 0) 38. Quantas vezes voc retorna a estas castanheiras e castanhais? Por qu? # Retornos 39. Quando comeou a coleta da castanha na safra deste ano (2010)? 40. Qual o papel dos homens na coleta da castanha? 41. Qual o papel das mulheres na coleta da castanha? 42. Qual o papel das crianas e jovens na coleta da castanha?
148 43. Que instrumentos utiliza na coleta da castanha? 44. Quantos dias em mdia por semana voc trabalha na coleta da castanha durante a safra? # Dias 45. Quantas horas por dia em mdia voc trabalha na coleta da castanha? # Horas 46. Tem algum dia da semana ou data especfica em que voc no vai coletar a castanha? Quando? Por qu? 47. Quais so as atividades associadas com a coleta da castanha que voc e sua famlia fazem? (1 0) 48. Como e por quem feita a coleta da castanha na sua colocao? (1=famlia, 2=coletiva, 3=gnero, 4=faixa etria) (1 4) 49. Quais so os membros da sua famlia que trabalham na coleta da castanha e participam das atividades relacionadas coleta da castanha? ( Preencher tabela abaixo ) Membro da Famlia Quantos dias por semana/Quantas horas por dia Voc paga esta pessoa? (1=sim; 0=no) Como feito o pagame nto e qual o valor mdio deste pagame nto? 50. Que dificuldades voc tem para coletar a castanha? (distncia dos castanhais, falta de gente pra ajudar, no tem tempo porque est envolvido em outras atividades, falta de apoio tcnico, outras) 53. Existe algum apoio ou financiamento para a coleta da castanha? (0=no, 1=sim) Se sim, qual(is)? (1 0) 54. Voc j recebeu este financiamento ou apoio alguma vez? (0=no, 1=sim) Que tipo de ajuda ele fornece? (dinheiro, material para a coleta, rancho, assistncia tcnica, apoio com transporte, outros) (1 0)
149 55. Como voc usa a castanha em sua casa (alimento, cura de doenas, rituais, outros)? 56. O que voc faz com a castanha coletada? Vende, troca ou consome? Se vende ou troca, desde quando voc e sua famlia fazem isto? Para quem voc vendeu a castanha coletada nos ltimos anos (2008, 2009 e 2010)? Nome Tipo de Comprador (1=regates, 2=empresa, 3=associao; 4=outros compradores) Quantas latas ou caixas foram vendidos? Preo da lata ou caixa da castanh a vendida ? 2008 2009 2010 57. Voc saberia dizer onde esta castanha vendida depois que sai aqui do Riozinho? (Supermercado, feira, loja, restaurantes, sai da cidade, vai pra outras cidades do Brasil ou do mundo, outros) Atividades de preparao da coleta da castanha 58. Voc limpa os castanhais antes da coleta da castanha? (0=no, 1=sim) Se sim, quais reas especficas dos castanhais voc limpa? (1=caminhos at os castanhais, 2=rea ao redor das rvores, 3=outras reas) Por qu? Voc faz esta limpeza na sua colocao e nos castanhais comunitrios? (0=no, 1=sim) (1 0) (1 3) (1 0) 59. Voc usa fogo para fazer a limpeza dos castanhais de sua colocao e nos castanhais comunitrios? (0=no, 1=sim) Por qu? (1 0) 60. Quem trabalha na limpeza dos piques e castanhais de sua colocao? So todos membros da famlia? (0=no, 1=sim) Se no, que outras pessoas ajudam nesta limpeza e por qu? (1 0) 61. Quem trabalha na limpeza dos piques e castanhais comunitrios? So todos membros da famlia? (0=no, 1=sim) (1 0)
150 Se no, que outras pessoas ajudam nesta limpeza e por qu? 62. Voc planta castanheiras? (0=no, 1=sim) Se sim, quantas castanheiras voc j plantou ou quantas castanheiras voc planta por ms, ou ano, ou safra da castanha? Onde? (1=roa; 2=roa abandonada/capoeira, 3=floresta, 4=outras reas) (Para cada um destes perguntar onde exatamente, por exemplo, onde na floresta se em sua colocao, outra colocao se em outras reas, quais reas de uso comum, outr os). Por qu? (1 0) # rvores (1 4) 63. Voc limpa as reas ao redor das plntulas e juvenis de castanha na floresta ou plantadas por voc ou sua famlia? (0=no, 1=sim) Se sim, por que voc faz esta limpeza? (1 0) 64. Voc protege as plntulas e juvenis da castanha da floresta ou plantadas por voc do fogo ou da predao de animais? (0=no, 1=sim) (1 0) 65. Voc corta os cips das castanheiras do sua colocao? (0=no, 1=sim) Se sim, por qu? Faz isso tambm nos castanhais comunitrios onde voc coleta? (0=no, 1=sim) (1 0) (1 0) 66. Voc faz algum corte no tronco da castanheira ou machuca o tronco da castanheira? (0=no, 1=sim) Se sim, por qu? (1 0) 67. Voc quebra o fruto (ourio) para retirada das amndoas da castanha no mesmo dia da coleta? (0=no, 1=sim) Se sim, por qu? (1 0) 68. Voc transporta a castanha coletada para a sua casa no mesmo dia da coleta? (0=no, 1=sim) Se sim, por qu? (1 0)
151 69. Por quanto tempo voc deixa os frutos de castanha coletados na floresta antes de transport los para sua casa? Por qu voc deixa esta tempo ou no deixa mais tempo? # Dias 70. Como feito o transporte da castanha coletada para a sua casa? (barco, caminhando, outros) e (paneiro, saco, outros) 71. Voc lava e seca a castanha coletada antes de vend la? (0=no, 1=sim) Se sim, como voc lava e seca a castanha e onde? (1 0) 72. Voc caa durante a coleta da castanha? (0=no, 1=sim) Se sim, voc caa mais durante a coleta da castanha do que durante a coleta de outros produtos ou outras atividades produtivas? (1=mais, 2=menos, 3=igual) (1 0) 73. Que animais que voc caa durante a coleta da castanha e quais destes animais voc caa com maior frequncia? 74. Voc caa e consome cotia? (0=no, 1=sim) Quantas cotias em mdia voc caa ou consome por semana ou por ms? (1 0) # Cotias
152 APPENDIX C SOCIAL NETWORK QUESTIONNAIRE (PORTUGUESE) Data: Nome da Localidade: Nome do Informante: COLETA DA CASTANHA 1. Se o seu castanhal no produz suficientemente no ano voc pede permisso pra coletar castanha em outra colocaco? (Sim/No) Pra quem? _________________________________________ _____________________________ a. Voc daria alguma coisa em troca? O qu? __ ______________ Se no, porque no daria nada em troca?_______________________________________________________ _________ b. ___________ seu/sua pare nte, amigo, ou vizinho? ________________________ c. Ele / ela tambm coleta castanha? Quais as outras ati vidades que ele / ela faze na colocaco? ______________________________________________________________ ________ d. Teria mais algum que voc escolheria se ele / ela dissesse no? (Sim/No) Quem? ________________________________________ ________________ ______________ 2. Voc ajuda a limpar os piques de castanha de outros moradores do Riozinho? (Sim/No) De quem? ______________________________________________________ ________________ a. Ele / ela te do alguma coisa em troca? (Sim/No) O qu? ___________ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Ele / ela tambm coleta castanha? Quais as outras atividades que ele / ela faz na colocaco? _________________________________________ _____________________________ d. Teria mais algum que voc ajudaria se ele / ela te pedisse? (Sim/No) Quem? ________________________________________ ______________________________ 3. Quando voc coleta castanha nos castanhais comunitrios voc convida outros da comunidade pra irem junto com voc? (Sim/No) ______________________________________________________________________
153 Quem? _________________________________________________________________ a. ___________ seu/sua parente, amigo, ou vizi nho? ________________________ b. Por que voc convida ele / ela (especificamente)? ________________________________________ ______________________________ c. Teria mais algum que voc convidaria pra ir coletar nos castanhais comunitrios junto com voc (Sim/No) Quem? ________________________________________ ______________________________ d. Voc d a ele / ela alguma coisa em troca? O qu? _____________________________ S e no, porque no daria nada em troca?_______ EMPRSTIMOS 6. Se voc precisasse pe dir emprestado uma canoa, para quem voc pediria emprestado? ___________________________________________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?___________________________________ ____________________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? ________________________________________ ______________________________ 7. Se voc precisasse pedir emprestado uma rabeta, para quem voc pediria emprestado? _____________________________________________________________ _________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troc a?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou vizinho? __________ ______________ c. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? _______________________________________ ______________________________ 8. Se voc precisasse pedir emprestado o rancho para a coleta de castanha, para quem voc pediria emprestado? ________________________________________ ______________________________
154 a. V oc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Ele / ela tambm coleta castanha? Quais as outras atividades que ele / ela faz na colocaco? ______________________________________________________________ ________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? _______ _________________________________ ______________________________ 9. Se voc precisasse pedir emprestado 1 lata (caixa) de castanha pra saldar sua dvida (ou pegar mercadorias) com o regato, para quem voc pediria emprestado? ______________________________ __________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou v iz inho? ________________________ c. Ele / ela tambm coleta castanha? Quais as outras atividades que ele / ela faz na colocaco? ______________________________________________________________ ________ d. Teria mais algum que voc escolheria se ele disse sse no? (Sim/No) Quem? ________________________________________ ______________________________ 10. Se voc precisasse pedir emprestado 10 latas (caixas) de castanha pra saldar sua dvida (ou pegar mercadorias) com o regato, para quem voc pediria emprestado? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_____________________________________________________ __ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Ele / ela tambm coleta castanha? Quais as outras atividades que ele / ela faz na colocaco? ______________________________________________________________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? ________________________________________ ______________________________
155 COMPRAS 11. Se voc precisasse comprar uma canoa, de quem voc compraria? ______________________________ __________ ______________________________ a. Por que voc compraria dele / dela? ________________________________________ ______________________________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Como voc pagaria esta canoa? Daria algo em troca? O qu? ________________________________________ ______________________________ d. Teria mais algum de quem voc compraria uma canoa? (Sim/No) Quem? ____________________________________________________________ __________ 12. Se voc precisasse comprar um kilo de farinha, de quem voc compraria? ________________________________________ ______________________________ a. Por que voc compraria dele / dela? _____________________________________________________ _________________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Como voc pagaria este um kilo de farinha? Daria algo em troca? O qu? ________________________________________ ______________________________ d. Teria m ais algum de quem voc compraria este um kilo de farinha? (Sim/No) Quem? _________________________________________________________________ 13. Se voc precisasse comprar 15 kilos de farinha, de quem voc compraria? ____________________________________ ____ ______________________________ a. Por que voc compraria dele / dela? ________________________________________ ______________________________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. Como voc pagaria estes 1 5 kilos de farinha? Daria algo em troca? O qu? ________________________________________ ______________________________ d. Teria mais algum de quem voc compraria estes 15 kilos de farinha? (Sim/No) Quem? _______________________________________________ __________________ _____
156 ASSISTNCIA 14. Se voc precisasse de ajuda para arrumar os documentos para receber os benefcios de aposentadoria, bolsa famlia, etc., de quem voc pediria ajuda? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou vizinho? __ _____________________ c. O que que ele / ela faz? _________________________________________________ _____________________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? ________________________________________ _________ _____________________ 15. Se voc precisasse de ajuda para resolver uma briga ou problema com algum da comunidade, de quem voc pediria ajuda? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_______________________________________________________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. O que que ele / ela faz? ___________________________ ______________________ _____________________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? _________________________________________ _____________________________ 16. Se voc precisasse de informao sobre assuntos da Res erva e do IBAMA com quem voc falaria? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?____________________________________________ ___________ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________
157 c. O que que ele / ela faz? _________________________________________________ _____________________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? __________________________________________________ ____________________ 17. Se voc precisasse de informao sobre os preos dos produtos que voc vende e compra, com quem voc falaria? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em t roca?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. O que que ele / ela faz? _____________________________ __________________ d. Teria mais algum qu e voc escolheria se ele dissesse no? (Sim/No) Quem? ________________________________________ ______________________________ 18. Se voc precisasse de ajuda pra levar algum da sua famlia para Altamira para tratar de algum problema de sade, com quem voc falaria? ________________________________________ ______________________________ a. Voc daria alguma coisa em troca? O qu? ________________ Se no, porque no daria nada em troca?_______________________________________________________ __________ b. ___________ seu/sua parente, amigo, ou viz inho? ________________________ c. O que que ele / ela faz? _________________ ______________________________ d. Teria mais algum que voc escolheria se ele dissesse no? (Sim/No) Quem? ________________________________________ ______________________________
158 APPENDIX D GOVERNMENT AND NON GOVERNMENTAL ORGANIZATIONS, AND PRIVATE SECTOR ACTORS
159 LIST OF REFERENCES Agrawal, A. (2001) Common property institutions and sustainable governance of resources. World Development 29 : 1649 1672. Agrawal, A. (2005) Environmentality: Technologies of government and the making of subjects. : Duke University Press. Agrawal, A. (2007) Forests, governance, and sustainability: common property theory and its contributions. International Journal of the Commons 1 (1): 111 136. Agrawal, A. & Chhatre, A. (2006) Explaining success in the commons: community forest governance in the Indian Himalaya. World Development 34 (1): 149 166. Agrawal, A. & Gibson, C. C. (1999) Enchantment and disenchantment: the role of community in natural resource conservation. World Development 27 (4): 629 649. Abrahamson, E. & Rosenkopf, L. (1997) Socia l networks effect on the extent of innovation diffusion: a computer simulation. Organization Science 8 (3): 289 309. Ahenkan, A. & Boon, E. (2008) Enhanced food security, poverty reduction and sustainable forest management in Ghana through non timber forest products farming: a case study of Sefwi Wiawso District. In: http://www.grin.com/en/e book/86653/enhancing food security and pov erty reduction in ghana through non timber : GRIN. Ahenkan, A. & Boon, E. (2010) Assessing the impact of forest policies and strategies on promoting the development of non timber forest products in Ghana. J Biodiversity 1 (2): 85 102. Allegretti, M. H. (1989 ) Reservas Extrativistas: um proposta de desenvolvimento da floresta amaznica. Par Desenvolvimento 25 : 3 29. Allegretti, M. H. (1990) Extractive reserves: an alternative for reconciling development and environmental conservation in Amazonia. In: Alternatives to Deforestation: Steps toward Sustainable Use of the Amazonian Rain Forest. ed. A. B. Anderson, pp. 253 264. New York: Columbia University Press. Allegretti, M. H. (1994) Reservas extrativistas: parmetros para uma poltica de desenvolviment o sustentvel na Amaznia. In: O destino da Floresta. Reservas extrativistas e desenvolvimento sustentvel na Amaznia. ed. R. Arnt, pp. 17 47. Curitiba, PR, Brazil: Instituto de Estudos Amaznicos e Ambientais, Fundao Konrad Adenauer. Allegretti, M. H. & Schwartzman, S. (1986) Extractive reserves: a sustainable development alternative for Amaznia. In: Report to the World Wildlife Foundation US (Project US 478).
160 Allegretti, M. H. & Schmink, M. (2009) When social movement proposals become policy. In: Rur al social movements in Latin America. Organizing for sustainable livelihoods eds. C. D. Deere & F. S. Royce, pp. 196 213. Gainesville, FL: T he University Press of Florida. Almeida, M. B. (2002) The politics of amazonian conservation: the struggles of rubb er tappers. The Journal of Latin American Anthropology 1 : 170 219. Anderson, A. B. (1990) Alternatives to deforestation New York : Columbia University Press. Ankersen, T. & Barnes, G. (2004) Inside the Polygon: Emerging community tenure systems and forest resource extraction. In: Working Forests in the Tropics: conservation through sustainable management? ed D. Zarin, F. E. Putz, M. Schmink & J. R. R. Alavalapati, pp. 253 264. New Y ork: Columbia University Press. Arnold, J. E. M. & Ruiz Prez, M. (2001) Can non timber forest products match tropical conservation and development objectives? Ecological Economics 39 (3): 437 447. Assies, W. (1997) Going nuts for the rainforest. Non timber forest products, forest conservation and sustainability in Amazonia. Amsterdam. Bandaragoda, D. J. (2000) A framework for institutional analysis for water resources management in a river base context. In: ed. I. W. M. Institute, Colombo, Sri Lan ka. Belcher, B. (2005) Forest product markets, forests and poverty reduction. International Forestry Review 7 (2): 8 2 89. Belcher, B., Ruz Prez, M. & Achdiawan, R. (2005) Global patterns and trends in the use and management of commercial NTFPs: Implications for livelihoods and conservation. World Development 33 (9). Berdegu, J., Ramirez, E., Reardon, T. & Escobar, G. (2001) Rural non farm employment an d incomes in Chile. World Development 29 : 411 425. Berkes, F. & Davidson Hunt, I. J. (2006) Biodiversity, traditional management systems, and cultural landscapes: examples from the boreal forest of Canada. In: UNESCO, Blackwe ll Publishing Ltd., Oxford, UK. Bernard, H. R. (2000) Social research methods: Qualitative and quantitative approaches. Thousand Oaks, California: Sage Publications, Inc. Bernard, H. R., Killworth, P. D., Sailer, L. & Kronenfeld, D. (1984) The problem of informant accuracy: The validity of retrospective data. Annual Review of Anthropology 13 : 495 517.
161 Bodin, O. & Crona, B. (2011) Barriers and opportunities in transforming to sustainable governance: the role of key indi viduals. In: Social networks and natural resource management: uncovering the fabric of environmental governance ed. O. Bodin & C. Prell, pp. 75 94. Cambridge, UK: Cambridge University Press. Bodin, O. & Prell, C. (2011) Social networks and natural resourc e management: uncovering the fabric of environmental governance Cambridge: Cambridge University Press. Borgatti, S. P., Everett, M. G. & Freeman, L. C. (2002) Ucinet for Windows: software for social network analysis. In: Analytic Technologies Harvard, Ma ssachusetts. Borrini Feyerabend, G., Kothari, A. & Oviedo, G. (2004) Indigenous land and local communities and protected areas: towards equity and enhanced conservation. In: Best practice protected area guidelines No. 11 Gland and Cambridge: IUCN. Boxman, A. W., De Graaf, P. M. & Flap, H. D. (1991) The impact of social and human capital on the income attainment of Dutch managers. Social Networks 13 : 51 73. Brasil, MMA (2000) Sistema Nacional de Unidades de Conservao da Natureza. Lei no. 9985 de 18 de Julho de 2000. http://www.planalto.gov.br/ccivil_03/Leis/L9985.htm Brown, K. & Rosendo S. (2001) Environmentalist, rubber tappers and empowerment: the politics and ec onomics of extractive reserves. Development and Change 31 : 201 227. Burt, R. S. (2000) The network structure of social capital. Research in Organizational Behaviour 22 : 345 423. Burt, R. S. (2004) Structural holes and good ideas. American Journal of Soci ology 110 (2): 349 399. Butler, R. A. & Laurance, W. F. (2008) New strategies for conserving tropical forests. Trends in Ecology and Evolution 23 (9): 469 472. Butterbury, S. (2003) Environmental activism and social networks: campaigning for bycicles and alternative transport in West London. The Annals of the American Academy of Political and Social Science 590 : 150 169. Campbell, C. E. (1996) Forest, Field and F actory: Changing livelihood strategies in two extractive reserves in the Brazilian Amazon. In: Center for Latin American Studies p. 208. Gain esville: University of Florida. Campbell, D. G., Daly, D. C., Prance, G. T. & Maciel, U. N. (1986) Quantitative ec ological inventory of terra firme and vrzea tropical forest on the Rio Xingu, Brazilian Amazon. Brittonia 38 (4): 369 393.
162 Campbell, B. M., Sayer, J. A. & Walker, B. (2010) Navigating trade offs: working for conservation and development outcomes. Ecology a nd Society 15 (2): 16 [online] URL: http//www.ecologyand society.org/vol15/iss12/art16/. Campos, M. T. & Nepstad, D. C. (2006) Smalholders, the Amazon's new conservationists. Conservation Biology 20 (5): 1553 1556. Cantrill, J. G. (1998) The environmental self and a sense of place: communication foundations for regional ecosystem management. Journal of Applied Communication Research 26 : 301 318. Cardoso, C. A. S. (2002) Extractive reserves in Brazilian Amazonia: loca l resources management and the global political economy. Burlington, VT, U SA: Ashgate Publishing Company. Carlsson, L. & Berkes, F. (2005) Co management: concepts and methodological implications. Journal of Environmental Management 75 : 65 76. Castelo, C. E F. (1999) Avaliao econmica da produo familiar na Reserva Extrativista Chico Mendes. In: Porto Velho Rondnia, Brazil: UNIR, UFSC. Castro, F. (2009) Patterns of resource use by caboclo communities in the middle lower Amazon. In: Amazon peasant socie ties in a changing environment. ed. C. Adams et al. pp. 157 177. Amsterdam, The Netherlands: S pringer Science Business Media. Cavendish, W. (2000) Empirical regularities in the poverty environment relationship of rural households: evidence from Zimbabwe. World Development 28 (11): 1979 2003. Cavendish, W. (2003) How do forest support, insure and improve the livelihoods of the rural poor? A research note. In: Bogor, Indonesia: Center for I nternational Forestry Research. Cheng, A. S., Kruger, L. E. & Daniels, S. E. (2003) "Place" as an integrating concept in natural resource politics: propositions for a social science research agenda. Society and Natural Resources 16 : 87 104. Chomitz, K. M. (2007). At loggerheads? : agricultural expansion, poverty red uction, and environment in the tropical forests. In: ed. K. M. Chomitz Washington D.C.: The International Bank for Reconstruction an d Development / The World Bank. Coleman, J. S. (1988) Social capital in the creation of human capital. American Journal of Sociology 94 : 95 120. Colfer, C. J. P. & Byron, Y. (2001) People managing forests: the link between human well being and sustainability Washington, D.C.: Center for International Forestry Research.
163 Collinson, C., Burnett, D. & Agreda, V. (2000) Economic viability of Brazil nut trading in Peru. In: Natural Resources and Ethical Trade ed. N. R. Institute, p. 62. Kent: University of Greenwich. Cooley, C. H. (1909) On self and social organization. Chicago, Illinois: The University of Chicago Press. Coomes, O. T. & Barham, B. L. (1997) Rain forest extraction and conservation in Amazonia. The Geographical Journal 163 (2): 180 188. Coomes, O. T., Barham, B. L. & Takasaki, Y. (2004) Targeting conservation development initiatives in trop ical forests: insights from analyses of rain forest use and economic reliance among Amazonian peasants. Ecological Economics 51 : 47 64. Coomes, O. T. & Burt, J. G. (2001) Peasant charcoal production in the Peruvian Amazon: rainforest use and economic relia nce. Forest Ecology and Management 140 : 39 50. Corral, L. & Reardon, T. (2001) Rural non farm incomes in Nicaragua. World Development 29 : 427 442. Cotta, J. N., Kainer, K. A., Wadt, L. H. O. & Staudhammer, C. L. (2008) Shifting cultivation effects on Brazil nut ( Bertholletia excelsa ) regeneration. Forest Ecology and Management 256 : 28 35. Cox, M., Arnold, G. & Toms, S. V. (2010) A review of design principles for community based natural resources management. Ecology and Society 15 (4): 38 [online] URL: http://www.ecologyand society.org/vol15/iss34/art38/. Crona, B. & Bodin, O. (2006) What you know is who you know? Communication patterns among resources users as a prerequisite for comanagement. Ecology and Society 11 (2): 7. Cronkleton, P., Bray, D. B. & Me dina, G. (2011) Community forest management and the emergence of multi scale governance institutions: lessons for REED+ development from Mexico, Brazil and Bolivia. Forests 2 : 451 473. Cunha, C. C. (2010) Reservas extrativistas: institutionalizao e implementao no estado brasileiro dos anos 1990. In: p. 308. Rio de Janeiro: Rio de Janeiro Federal University. Da Cunha, E. (2003) Amaznia: um paraso perdido. In: ed. G. d. E. d. Amazonas, Manaus: Editora da Universidade Federal do Amazonas. Deal, T. & Kennedy, A. (1982) Corporate cultures. In: Addison Wesley, Reading Massachusetts, USA.
164 D'haese, M., De Ruijter De Wildt, M. & Ruben, R. (2012) Business incomes in rural Nicaragua: the role of household resources, location, experience and trust. Entrepren eurship and Regional Develpment 20 (July): 345 366. Do Rgo, J. F. (1999) Amaznia: do extrativismo ao neoextrativismo. Cincia Hoje 25 (146): 62 65. Duchelle, A. E. (2009) Conservation and livelihood development in Brazil nut producing communities in a tri national Amazonian frontier. In: p. 214. Gain esville: University of Florida. Duchelle, A. E., Cronkleton, P., Kainer, K. A., Guanacoma, G. & Gezan, S. (2011) Resource theft in tropical forest communities: implications for non timber management, livelih oods, and conservation. Ecology and Society 16 (1): 4. Dudley, N. (2008) Guidelines for applying area management categories. Gland, Switzerland: IUCN. Ehringhaus, C. (2005) Post Victory Dilemmas: Land Use, Development, and Social Movement in Amazonian Extra ctive Reserves. In: School of Forestry and Environmental Studies p. 430. Yale University. Escobal, J. & Aldana, U. (2003) Are nontimber forest products the antidote to rainforest degradation? Brazil nut extraction in Madre de Dios, Peru. World Development 31 (11): 1873 1887. Fafchamps, M. (2004) Market institutions in Sub Saharan Africa: theory and evidence. Cambridge, Mass.: MIT Press. FAO (1995) Non wood forest products for rural income and sustainable forestry. Non wood Forest Products 7 : 1 2. FAO (2001) Forest resource assessment 2000. In: p. 479. Rome, Italy: FAO Fearnside, P. M. (1989) Extractive reserves in Brazilian Amazonia: An opportunity to maintain tropical rain forest under sustainable use. BioScience 39 : 387 393. Festinger, L., Schachter, S. & Back, K. (1950) Social pressures in informal groups: a study of human factors in housing. New York, NY: Harper and Brothers. Freeman, L. C. (1979) Centrality in social networks. Conceptual clarifications. Social Networks 1 : 215 239. Freudenburg, W. R. (1986) The density of acquantainceship: an overlooked variable in community research. American Journal of Sociology 97 : 27 63. Friedkin, N. E. (1998) A structural theory of social influence New York, NY: Cambridge University Press.
165 Fuku yama, F. (1995) Trust: the social virtues and the creation of prosperity London: Hamish Hamilton. Futemma, C., Castro, F., Silva Forsberg, M. C. & Ostrom, E. (2002) The emergence and the outcomes of collective actions: an institutional and ecosystem appro ach. Society and Natural Resources 15 (6): 503 522. Gautam, A. P. & Shivakoti, G. P. (2005) Conditions for successful local collective action in forestry: some evidence from the hills of Nepal. Society and Natural Resources 18 (2): 153 171. Gibson, C. C. (2001) Forest resources: institutions for local governance in Guatemala. In: Protecting the commons: a framework for resource management in the Americas. eds. J. Burger, E. Ostrom, R. B. Noorgaard, D. Policansky & B. D. Goldstein, pp. 71 89. Washi ngton, D.C., USA: Island Press. Girvan, M. & Newman, M. E. J. (2002) Community structure in social and biological networks. PNAS 99 (12): 7821 7826. Goeschl, T. & Igliori D. C. (2006) Property rights for biodiversity conservation and development: ext ractive re serves in the Brazilian Amazon. Development and Change 37 (2): 427 451. Gomes, C. V. A. (2001) Dynamics of land use in an amazonian extractive reserve: the case of the Chico Mendes Extractive Reserve in Acre, Brazil. In: Geography Department p. 1 53. Gain esville: University of Florida. Gomes, C. V. A. (2009) Twenty years after Chico Mendes: Extractive Reserve's expansion, cattle adoption and evolving self definition among rubber tappers in the Brazilian Amazon. In: p. 231. Gain esville: University o f Florida. Gomes, C. V. A., Vadjunec, J. M. & Perz, S. G. (2012) Rubber tapper identities: political economic dynamics, livelihood shifts, and environmental implications in a changing Amazon. Geoforum 43 : 260 271. Gradwohl, J. & Greenberg, R. (1988) Saving the tropical forests. Washington, D.C.: Longwood Publisher Group. Granovetter, M. (1985) Economic action, social structure, and embeddedness. American Journal of Sociology 91 (481 510). Guerra, R. M. N. (2004) Verificando a viabilidade do PDS So Salvador no estado do Acre. Ambiente & Sociedade 7 (1): 157 167. Hall, A. (1997) Sustaining Amazonia: grassroots action for productive conservation. England.
166 Hall, A. (2004) Extractive reserves: b uilding natural assets in the Brazilian Amazon. In: Working Paper Series ed. U. o. M. Amherst, pp. 1 27. Politi cal Economy Research Institute. Harkes, I. H. T. (2006) Fisheries co management, the role of local institutions and decentralisation in Southeas t Asia: with specific reference to marine sasi in Central Maluku, Indonesia. In: Institute of Environmental Sciences Leiden, The Netherlands: Leiden University. Hill, C. W. L. (2009) Stragegic management theory: an integrated approach. Mason, Ohio, USA: C engage Learning. Hill, C. W. L. & Jones, G. R. (2000) Cases in strategic management. Boston, Massachusetts, USA: Houghton Mifflin Company. Homans, G. C. (1950) The human group. London, UK: Routledge and Kegan Paul. Homma, A. K. O. (1992) The dynamics of extraction in Amazonia: A historical perspective. In: Non timber products from tropical forests: Evaluation of a conservation and development strategy. ed. D. C. N. a. S. Schwartzman, pp. 23 31. New Yor k: Advances in Economic Botany. ICMBio (2011) Instituto Chico Mendes de Conservao da Biodiversidade, Ministrio do Meio Ambiente, Braslia, Brasil. In: http://www.icmb io.gov.br/comunicacao/downloads ISA, Instituto Socioambienta. (2006) Relatrio do Seminrio "Perspect ivas da Terra do Meio". In: p. 44. Altamira, Brazil: Instituto Socioam biental. Janse, G. & Ottitsch, A. (2005) Factors influencing the role of non wood forest products and services. Forest Policy and Economics 7 : 309 319. Kadushin, C. (1966) The friends an d supporters of psichotherapy: on social circles in urban life. American Sociological Review 31 (786 802). Kainer, K. A., Wadt, L. H. O., Gomes Silva, D. A. P. & Capanu, M. (2006) Liana loads and their association with Bertholletia excelsa fruit and nut production, diameter growth and crown attributes. Journal of Tropical Ecology 22 : 147 154. Kainer, K. A., Wadt, L. H. O. & Staudhammer, C. L. (2007) Explaining variation in Brazil nut production. Forest Ecology and Management 250 : 244 255. K eck, M. E. (1995) Social Equity and environmental politics in Brazil: lessons from the rubber tappers of Acre. Comparative Politics 27 (4): 409 424. Kemmis, D. (1990) Community and the politics of place Norman: University of Oklahoma Press.
167 Knack, S. & Keefer, P. (1997) Does social capital have an economic payoff? A cross country investigation. The Quaterly Journal of Economics 112 (November): 1251 1288. Korten, D. C. (1980) Community organization and rural development: a learning process approach. Publi c Administration Review 40 (5): 480 511. Kusters, K., Achdiawan, R., Belcher, B. & Prez, M. R. (2006) Balancing development and conservation? An assessment of livelihood and environmental outcomes of nontimber forest products trade in Asia, Africa, and Lat in America. In Ecology and Society 11 (2): 20 41. Lanjouw, P. & Feder, G. (2001) Rural and non farm activities and rural development: from experience towards strategy. In: Rural Develoment Strategy Background Paper Washington, D.C.: World Bank. Mahapatra, A. K. & Tewari, D. D. (2005) Importance of non timber forest products in the economic valuati on of dry deciduous forests of I ndia. Forest Policy and Economics 7 : 455 467. Mahapatra, A. K., Albers, H. J. & Robinson, E. J. Z. (2005) The impact of NTFP sales on rural households' cash income in India's fry decidous forest. Environmental Management 35 (3): 258 265. Mannigel, E. (2008) Integrating parks and people: how does participation work in protected area management? Society and Natural Resources 2 1 : 498 511. Maske, M., Mungole, A., Kamble, R., Chaturvedi, A. & Chaturvedi, A. (2011) Impact of non timber forest produces (NTFP's) on rural tribes economy in Gondia District of Maharashtra, India. Archives of Applied Science Research 3 (3): 109 114. McElwee, P. D. (2008) Forest environmental income in Vietnam: household socioeconomic factors influencing forest use. Environmental Conservation 35 (2): 147 159. McKean, M. A. (2000) Common property: what is it, what is it good for, and what makes it work? In: People and forests : communities, institutions, and governance ed C. C. Gibson, M. A. McKean & E. Ostrom, pp. 27 55. Cambridge, Mass.: MIT Press. McPherson, M., Smith Lovin, L. & Cook, J. M. (2001) Birds of a feather: homophily in social networks. Ann ual Review of Sociology 27 (2001): 415 444. Meikle, S., Ramasut, T. & Walker, J. (2001) Sustainable urban livelihoods: concepts and implications for policy. In: Working paper The Development Planning Unit London: University College London.
168 MMA (2003) Projeto realizao de estudos preliminares e formulao de uma proposta tcnica para a implantao de um mosaico de unidades de conservao no Mdio Xingu. In: Instituto Socioambiental and Instituto de Pesquisa Ambiental da Amaznia, p. 237. Nep stad, D., Schwartzman, S., Bamberger, B., Santilli, M., Ray, D., Schlesinger, P., LeFebvre, P., Alencar, A., Prinz, E., Fiske, G. & Rolla, A. (2006) Inhibition of Amazon deforestation and fire by parks and indigenous lands. Conservation Biology 20 : 65 73. Neumann, R. P. & Hirsch, E. (2000) Commercialisation of non timber forest products: review and analysis of research. Bogor, Indonesia: Center for I nternational Forestry Research. Newton, P. (2011) Opportunities for conservation and livelihoods in Amazonian extractive reserves. In: School of Environmental Sciences p. 170. East Anglia: University of East An glia. North, D. C. (1999) Institutions, institutional change, and economic performance. Cambridge, UK: Cambridge University Press. Nygren, A. (2005) Commu nity based forest management within the context of institutional descentralization in Honduras. World Development 33 (4): 639 655. Oliveira, R. & Anderson, E. S. (1999) Gender, conservation, and community participation: the case of the Ja National Park, Br azil. MERGE, Managing Ecosystems and Resources with Gender Emphasis 2 : 1 13. Ostrom, E. (1990) Governing the commons: the evolution of institutions for collective action Cambridge, New Yo rk: Cambridge University Press. Pacheco, L. M. (2010) Arising from the trees: achievements, changes, and challenges of the rubber tappers movement in the Brazilian Amazon. In: Gainesvi lle, FL: University of Florida. Pattanayak, S. K. & Sills, E. O. (2001) Do tropical forests provide natural insurance? The microecono mics of non timber forest product collection in the Brazilian Amazon. Land Economics 77 (4): 595 612. Peralta, P. A. & Kainer, K. A. (2008) Market integration and livelihood systems: a comparative case of three Ashninka villages of the Peruvian Amazon. Jou rnal of Sustainable Forestry 27 (1 2): 145 171. Peres, C. A. & Baider, C. (1997) Seed dispersal, spatial distribution and population structure of Brazil nut trees ( Bertholletia excelsa ) in Southeastern Amazonia. Journal of Tropical Ecology 13 (4): 595 616.
169 P eres, C. A., Baider, C., Zuidema, P. A., Wadt, L. H. O., Kainer, K. A., Gomes Silva, D. A. P., Salomo, R. P., Simes, L. L., Franciosi, E. R. N., Valverde, F. C. G., R., Shepard Jr., G. H., Kanashiro, M., Coventry, P., Yu, D. W., Watkinson, A. R. & Freckl eton, R. P. (2003) Demographic threats to the sustainability of Brazil nut exploitation. Science 302 : 2112 2114. Persha, L. A., Agrawal, A. & Chhatre A. (2011) Social and ecological synergy: Local rulemaking forest livelihoods and biodiversity conservation. Science 331 (1606 1608). Perz, S. G. & Walker, R. T. (2002) Household life cycles and secondary forest cover among small farm colonists in the Amaz on. World Development 30 (6): 1009 1027. Pires, J. M. & Prance, G. T. (1985) The vegetation types of the Braz ilian Amazon. New York: Pergamon Press. Portes, A. & Landholt, P. (1996) The downside of social capital. The American Prospect 26 (May June): 18 21. Potetee, A. & Ostrom, E. (2004) Heterogeneity, group size, and collective action: the role of institutions on forest management. Development and Change 35 (3): 435 461. Powell, W. (1990) Neither market nor hierarchy: network forms of organisation. Research of Organisational Behaviour 12 : 295 336. Pretty, J. (2003) Social capital and collective management of resources. Science 302 (5952): 1912 1914. Putman, R. D. (1993) Making democracy work. Civic traditions in modern Italy. Princeton, NJ: Princeton University Press. Pyhl, A., Brown K. & Adger W. N. (2006) Implications of livelihood dependence on non timber products in Peruvian Amazonia Ecosystems 9 : 1328 1341. Reagans, R. & McEvily, B. (2003) Network structure and knowl edge transfer: the effects of cohesion and range. Administrative Science Quarterly 48 (2): 240 267. Reardon, T., Berdegu, J. & Escobar, G. (2001) Rural non farm employment and incomes in Latin America: overview and policy implications. World Development 29 : 395 409. Rocha, C. G. S., da Silva, P. A., Carvalho, S. A. & Salgado, I. (2005) Diagnstico scio econmico da Reserva Extrativista Riozinho do Anfrsio. In: p. 62. Altamira, Brazil : Universidade Federal do Par. Rocheleau, D. & Edmunds D. (1997) Women, men and trees: Gender, power and property in forest and agrarian landscapes. World Development 25 (8): 1351 1371.
170 Ruben, R. & Van den Berg, M. (2001) Nonfarm employment and poverty alleviation of rural farm households in Honduras. World Development 2 9 (3): 549 560. Ruiz Prez, M., Almeida, M., Dewi, S., Costa, E. M. L., Pantoja, M. C., Puntodewo, A., Postigo, A. A. & Andrade, A. G. (2005) Conservation and development in amazonian extractive reserves: the case of Alto Juru. AMBIO 34 (3): 218 223. Salafs ky, N. & Wollenberg, E. (2000) Linking livelihoods and conservation: A conceptual framework and scale for assessing the integration of human needs and biodiversity. World Development 28 (8): 1421 1438. Salisbury, D. S. & Schmink M. (2007) Cows versus rubber: changing livelihoods among Amazonian extractivists. Geoforum 38 : 1233 1249. Sallu, S. M., Twyman, C. & Thomas, D. S. G. (2009) The multidimensional nature of biodiversity and social dynamics and implications for contemporary rural livelihoods in remote Kalahari settlements, Botswana. African Journal of Ecology 47 (1): 110 118. Salomo, R. P. (1991) Estrutura e densidade de Bertholletia excelsa H. and B. ("Castanheira") nas regies de Carajs e Marab, Estado do Par. Boletim d o Museu Paraense Emilio Goeldi. Srie Botnica 7 : 47 68. Salomo, R. P. (2009) Densidade, estrutura e distribuio espacial da castanheira do Brasil ( Bertholletia excelsa H. & B.) em dois plats de floresta ombrfila densa na Ama znia setentrional brasilei ra. Boletim do Museu Paraense Emilio Goeldi. Cincias Naturais 4 (1): 11 25. SAS Institute, Inc. (2004) S AS, Version 9.2, Cary, NC, USA. Schlager, E., Blomquist, W. & Tang, S. Y. (1994) Mobile flows, storage, and self organized institutions for governing common pool resources. Land Economics 70 (3): 294 317. Schmink, M. & Wood, C. H. (1987) The "political ecology" of Amazonia. In: Lands at risk in the Third World: local level perspective ed. P. D. Little, M M. Horowitz & A. E. Nyerges, pp. 38 57. London: Westview Press. Schmink, M. & Wood C. H. (1992) Contested Frontiers in Amazonia New York: Columbia University Press Schwartzman, S. (1989) Extractive reserves: the rubber tappers' strategy for sustainable use of the Amazon rainforest. In: Fragile Lands of Latin America. Strategies for Sustainable Development. ed. J. O. Browder, pp. 150 163. San Francisco: Westview Press.
171 Scoles, R. & Gribel, R. (2011) Population structure of Brazil nut ( Bertholletia excelsa Lecythidaceae ) stands in two areas with different occupation histories in the Brazilian Amazon Human Ecology 39 (4): 455 464. Scoones, I. (1998) Sustainable rural livelihoods: a framework for analysis. In: IDS Working Paper pp. 1 22. In stitute of Development Studies. S cott, J. C. (1998) Seeing like a state: how certain schemes to improve the human condition have failed. New Haven, Connectic ut, USA: Yale University Press. Shackleton, S. & Shackleton, C. (2004) The importance of non timber forest products in rural livelih ood security and as safety nets: a review of evidence from South Africa. South African Journal of Science 100 : 658 664. Shone, B. M. & Caviglia Harris, J. L. (2006) Quantifying and comparing the value of non timber forest products in the Amazon. Ecological Economics 58 : 249 267. Sick, D. (2008) Social contexts and consequences of institutional change in common pool resource management. Society and Natural Resources 21 (2): 94 105. Silva, P. A. (2007) Contextualizao socioambiental da Reserva Extrativista Ri ozinho do Anfrsio: um estudo de caso Altamira, PA. In: p. 231. Manaus: Universidade Federal do Amazonas. Souza, F. K. A. (2006) Effectiveness of extractive reserves, agro extractive settlements, and colonist settlements in southwestern Amazonia: an econ omic and land cover comparison of three tenure types in Acre, Brazil. In: Latin American Studies p. 174. Gain esville: University of Florida. Stoian, D. (2005) Making the best of two worlds: rural and peri urban livelihood options sustained by nontimber forest products from the Bolivian Amazon. World Development 33 (9): 1473 1490. Stone, S. (2003) From tapping to cutting trees: participation and agency in two community based timber management projects in Acre, Brazil. In: Gainesville, FL, USA: University of Florida. Svarrer, K. & Olsen, C. S. (2005) The economic va l ue of non timber forest products A case study from Malaysia. Journal of Sustainable Forestry 20 (1): 17 41. Synnot, T. J. (1979) A manual of permanent sample plot procedures for tropical rainfo rest. In: Tropical Forestry Papers No. 14 : Commonwealth Forestry Insti tute, University of Oxford, UK. Takasaki, Y., Barham, B. L. & Coomes, O. T. (2010) Smoothing income against crop flood losses in Amazonia: rain forest or rivers as a safety net? Review o f Development Economics 14 (1): 48 63.
172 te Velde, D. W., Rushton, J., Schreckenberg, K., Marshall, E., Edouard, F., Newton, A. & Arancibia, E. (2006) Entrepreneurship in value chains of non timber forest products. Forest Policy and Economics 8 : 725 741. Tin dall, D. B. (2002) Social networks, identification and participation in an environmental movement: low medium cost activism within the British Columbia widerness preservation movement. Canadian Review of Sociology 39 : 413 452. Tucker, C. M., Randolph, J. C & Castellanos, E. J. (2007) Institutions, biophysical factors and history: an integrative analysis of private and common property forests in Guatemala and Honduras. Human Ecology 35 (3): 259 274. Turner, M. D. (1999) Conflict, environmental change, and social institutions in dryland Africa: limitations of the community resource management approach. Society and Natural Resources 12 (7): 643 657. Vadjunec, J. M., Gomes, C. V. A. & Ludewigs, T. (2009) Land use/land cover change among rubber tappers in the Chico Mendes Extractive Reserve, Acre, Brazil. Journal of Land Use Science 4 (4): 249 274. Vadjunec, J. M., Schmink, M. & Gomes, C. V. A. (2012) Rubber tapper citizens: emerging places, policies, and sh ifting rural urban identities in Acre, Brazil. Journal of Cultural Geography 28 (1): 73 98. Vedeld, P., Angelsen, A., Boj, J., Sjaastad, E. & Berg, G. K. (2007) Forest environmental incomes and the rural poor. Forest Policy and Economics 9 : 869 879. Viana, V. M., Mello, R. A., Moraes, L. M. & Mendes, N. T. (1998) Ecologia e manejo da castanha do Par em reservas extrativistas de Xapuri, Acre. In: Floresta amaznica: dinmica, regeneraco e manejo. ed C. Gascon & P. Mou tinho, pp. 277 292. Manaus, Bras il: M inistrio de Cincia e Tecnologia Instituto Nacional de Pesquisas da Amaznia. Wadt, L. H. O., Kainer, K. A. & Gomes Silva, D. A. P. (2005) Population structure and nut yield of a Bertholletia excelsa stand in Southwestern Amazonia. Forest Ecology and Management 211 : 371 384. Wallace, R. H. (2004) The effects of wealth and markets on rubber tapper use and knowledge of forest resources in Acre, Brazil. In: p. 303. Gainesville: University of Florida. Warren, M. (1999) Democratic theory and trust. In: Democrary and Trust ed. M. Warren, pp. 310 345. Cambridge: Cambridge University Press. Weimann, G. (1982) On the importance of marginality: one more step into the two step flow of communication. American Sociological Review 47 (6): 764 773.
173 Williams, D. R. & Stewart, S. I. (1998) Sense of place: An elusive concept that is finding a home in ecosystem management. Journal of Forestry 96 : 18 23. Zuidema, P. A. & Boot, R. G. A. (2002) Demography of the Brazil nut tree ( Bertholletia excelsa ) in the Bolivian Amazon: impact of seed extraction on recruitment and population dynamics. Journal of Tropical Ecology 18 : 1 31.
174 BIOGRAPHICAL SKETCH Vivian Zeidemann has had diversif ied work experience s that range from environmental education, to marine conservation, soil geochemistry, human health, Third Sector administration, and community based natural resources management. She received a T eaching degree in b iology from the Federal University of Santa Cata rina, Florianpolis, Brazil, a m aster i n natural resources and tropical b iology from the National Institute of Amazonian Research, Man aus, Brazil, and completed her doctorate in Philosophy in the School of Natural Resources and Environment at the University of Florida. F or the past 10 years Vivian has been working with several aspects of conservation and development with different traditional communities in Brazi l. Just after she received her b iology undergraduate degree she started to work in a turtle marine c onservation project with environmental education for tourists and fisherman communities, lecturing about turtle m arine ecology and conservation as well as managing turtle marine nests. For her thesis in Manaus, Brazil, she developed a research pro ject on soil contamination by mercury in the Rio Negro basin where she was able to interact wi th several traditional riverine communities affected by mercury contamination. She also had the opportunity to spent five months in 1 997 at the Universit du Qu bec Montral ( UQAM), where she received training i n soil analysis techniques. After her master she was hired by Conservation International Brazil to work in a malaria control and prevention project in a Kayap indigenous community located in the Br azilian Amazon region. This project was carried out in partnership with the Brazilian National Health Foundation a nd had the main goal to educate the Kayap community members on how to control and prevent malaria, but also teach them how to
175 diagnose and treat the disease, and to collect the malaria mosquitos for the National Health Foundation researchers. village for almost a year, when she had the chance to be exposed to the Kayap culture and their li fe style. This health project grew in objectives and expanded to other four Kayap communities demanding the creation of an institution to manage the funding for these projects and initiatives. Thus, the Protected Forest Association was created in 2002, a nd Vivian assumed the position of project manager In that position she was responsible for administering the resources and managing activities related to the protection of the Kayap territory and to promote sustainable economic initiativ es in these four Kayap communities She also developed a program of institutional capacity building for the Kayap members of the Association board. After the experiences with these different groups of traditional communities she realized the importance to integrate conse rvation and development for the improvement livelihood s while pursuing conservation goals In spite of all her experience with traditional communities in the Amazon region and other parts of Brazil, she felt she needed more acad emic training and decided to apply for the interdisciplinary ecology graduate program at University of Florida. During Vivian Ph.D. fieldwork she had the opportunity to interact with extractive reserve residents, and many governmental and non governmenta l organizations responsible for reserve management and implementation This interaction had an and development initiatives, but also related to natural resource manageme nt and its elihoods. Because she believes i n the
176 importance of academic research to contribute to the success of conservation and development initiatives taking place at the ground, she made a commitment to give ba ck her Ph.D. research results to reserve residents and other reserve managers. Vivian intends to continue to work with conservation and development projects that have a focus on traditional communities and the integration of the socioecological aspects needed to foster sustai nable livelihood strategies