1 INSTITUTIONAL ARRANGEMENTS FOR FIRE MANAGEMENT IN THE BRAZILIAN AMAZON By MARIA LUCIMAR DE LIMA SOUZA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2009
2 2009 Maria Lucimar de Lima Souza
3 To my family, friends, professors and to all those who love the Amazon, and understand that we, humans, are just part of nature
4 ACKNOWLEDGMENTS It was a ch allenge for me to come all the way from the Brazilian Amazon to the United States of America to attend a graduate school without almost any English skills. There were so many people and institutions that helped me. The lis t is almost infinite, and I hope not to forget anybody. First of all, I would like to say thanks to Professor Dr. Marianne Schmink. It is the most special thank you here. I Thank her for he r support and vision since the first time we met each other in Acre. I thank her for accepting me as a student even in her sabbatical year. My sincere thank for all the conversation we had in Gainesville, for each suggestion she gave me during my three years living here and for being so supportive. I th ank her for all the revisions on my thesis, for being so careful to the point of really reading everythi ng, including footnotes of my work. I thank her for being the best advi ser ever. I thank Professor Charles Wood and Professor Karen Kainer for their participati on in my committee, and for being such great professors and source of inspirat ion to my future professional life. I thank Professor Carmen Deere for opening the doors of Latin America thr ough the course she offers in the Center for Latin American Studies. I also wa nt to say thanks to Professor John Dain for being one of the nicest human being I have ever met and for the extreme respect he shows for all his students. I also want to thanks to professor Antony Oliver-S mith for the passion he shows for his subjects of study and for the great travel in which he conducted me thought anthropological thinking. Finally, thanks to professor Huge Popenoe for showing me the importance of loving your job in life. Thank to all my professors for the know ledge shared, the patien ce they had with my difficulties, and for the good example they were to me. My view of the academic world is completely different now, and all of you have a piece of responsibility in this change. I am grateful to the people of Piquiatuba, Ch ib, Marai, Paraso e Prainha II communities, for the receptivity, and also for the patience to answ er all my questions. I am also grateful to all
5 the families living within Tapajs National Forest for the four years of work together, for all the meetings we carried out together, all the surveys answered, all the visits to their fields. I cannot forget the first communities where I started working as practitioner. Thanks to all the families of Del Rei and Rio Bonito communities. I cannot expr ess how much I learned working there. I am grateful to the Moore Foundation for the financial support during the years of my Masters. I also thank IBAMA and ICMBio employees for the suppor t while I was carrying out my research. My special thank to Manuella de Souza for all the support and information sh e sent to me. I also thank IPAM for the first opportunity to work as an environmentalist in the Amazon. I am especially thankful to Ricardo Mello, Rosana, Leuza, Marcos and Edivan for being such a nice group to work with. You guys rock! Thank you to Claudia Azevedo-Ramos for the opportunity to be a trainee in the program and to Daniel Ne pstad for believing that I could work with rural communities. Thanks to Ana Cristina Barros for pu shing me to work hard and thanks to David McGrath for all the conversation during the ye ars of work and for the great letter of recommendation he wrote for me. Thanks to Oriana and Oswaldo for their friendship at IPAM and for their great sense of humor that made our days easy. Thanks to Jaqueline for her constant smile at work and for the tenderness she used to have with all of us. Thanks also to Gina and Liana for helping so much during my application process with all the Eng lish difficulties and so forth. I thank my parents Antonio and Zimar for ev erything I am today, and because at the end they gave me the freedom I needed to take my own decisions in life, to grow financially and intellectually independent. Thanks to my sisters Elizete and Lucild a for all the love, all the talks and for being there for me always and always. Th anks to my brothers for the respect and support to my decisions even when they thought I was go ing too far. Thanks to all my nieces, nephews,
6 sisters and brothers-in-law, and finally to my grandnieces for helping me every day of this journey. Without my family nothing of that would have been possible for me. I also thank all my paidegua friends. Evelyn, Wendell, Tamar, Meriet e, Adriana, Monica, Daniela, Alex, Darlin, Allyne, Marcia, Adilton, Carlinhos, and Nuccia are the people that were with me everyday helping me with all the problems I had, and sayi ng all the time they were there to support me. Special thanks to Evelyn for all the messages ex changed during these three years, and for sharing your life with me at such a deep level. Thanks to my good friends Salete e Dilene for helping and making my life easier and funnier during the years of Par Stude nt House, and for our lively friendship nowadays. Thanks to Leandro for suppor ting my decision to come to the US, and for being with me in the first years of this adventur e. Finally, I thank my new friends of Gainesville. The time in Gainesville was an opportunity to m eet so many nice people. Thanks to Valerio for helping me to settle in Gainesville, for each conversation and for answering all my calls asking for help. Thank you Iran for being with me all the years, and for being so patient to review many papers, for driving me so many places, and help ing to understand the mean ing of modernization theory, dependency theory and othe r class topics. Thanks to Ane and Vivian for just being there for me in many different situati ons. My special thanks to Ane for her friendship, for being here with me during this time supporting me, and also for all her contributions to my thesis. They made my thesis much easier to read and more in teresting. Florzinha, I am eternally grateful to you. Thanks to Simone and Morena for the free friendship they offered to me, for all the conversations where we could talk about everyt hing, and for the fun we had together. Thank to Ludmila for making my days in Gainesville be tter, for being such a nice friend since the beginning, for all the talks, and also for helping me in my research and data analysis. Thanks to Evelina e Ani for helping to overcome the fear of speaking in English. Thanks to Evelina for still
7 being such a wonderful friend. Thanks to Dave fo r all the opportunities to talk and for being my American friend. Thanks also to Hilary for the wonderful friendship, and for being one more of my American friends. Thank to Pedro for all the tenderness when I needed it, and for introducing Ana to our group. Thanks to Ana for being so nice and friendly, for all the conversations we had, and for sharing her happiness and calm with me every day. Thanks to Ricardo Mello and Denyse for making me feel safe when they were around. Special thanks to Ricardo for reviewing my chapters and helping with ideas. Thanks to Bruna and Karina for the friendship and for being so fun and giving us all the good vibrations all the time. Thanks to Michael Sepulveda for the first invitation to me to come to the USA, fo r supporting me during this adventure, and for the free friendship. Thanks to Everton for helping to overcome my first fears of learning English and for being so fun in class. Thanks to Armando Marc os for being such a nice friend, and for all the help. Finally thanks to Christine Housel for not just being so lovely and kind, for being my friend, even though we did not meet often, and al so for accepting the challenge of reviewing my English in this thesis.
8 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES .........................................................................................................................10LIST OF FIGURES .......................................................................................................................11ABSTRACT ...................................................................................................................... .............12 CHAP TER 1 INTRODUCTION .................................................................................................................. 14Research Objectives ........................................................................................................... .....14Research Site ..........................................................................................................................15Research Methodology ...........................................................................................................15Organization of the Thesis ......................................................................................................152 THE USE OF FIRE IN THE AMAZ ON: IMPORTANCE, P ROBLEMS AND ATTEMPTS TO DECREASE ACCIDENTAL FIRES ......................................................... 18Introduction .................................................................................................................. ...........18The Importance and Problems Associated with the Use of Fire in Amazon .......................... 20Increasing Risk of Forest Fires ........................................................................................22Ecological Losses ............................................................................................................ 24Economic and Social Losses ...........................................................................................25Looking for Solutions: the Attempts to Decr ease Accidental Fires in the Brazilian Amazon ........................................................................................................................ .......26Legal Attempts to Decrease Fire ..................................................................................... 27Governmental Programs and Other Initiatives to Decrease Fire ..................................... 30Regional level ...........................................................................................................30From regional to local level ..................................................................................... 32Local level ................................................................................................................34Some Thoughts on Fire in the Amazon and the Experiences Carried Out ............................. 35Conclusions .............................................................................................................................363 INSTITUTIONAL ARRANGEMENTS FOR FI RE MANAGEMENT: THE C ASE OF TAPAJS NATIONAL FOREST PAR BRAZIL......................................................... 38Introduction .................................................................................................................. ...........38Institutional Arrangements to Avoid the Tragedy of the Commons ......................................39Building Institutional Arrangements for Fire Management: Tapajs National Forest ........... 44History of the Project .......................................................................................................45The Tapajs National Forest ...........................................................................................46The Fire Management Project at Flona-Tapajs ............................................................. 49
9 Changes in Fire Management Practices at Flona-Tapajs .............................................. 56Institutional Arrangements for Fire Manage ment Avoiding the Tragedy of Commons ...... 61Conclusions .............................................................................................................................674 PARTNERSHIP IN AN ACTION RESE AR CH PROJECT: BUILDING AND ANALYZING PARTICIPATION .......................................................................................... 70Introduction .................................................................................................................. ...........70Background .................................................................................................................... .........73Attempts to Build a Participatory Action Research at Tapajs National Forest .................... 80The Fire Management Project at Flona-Tapajs ............................................................. 82Reflections on the Process of Partne ring for Action Research Project ........................... 93Do Numbers of People Count? Particip ation and Compliance with Rules ............................ 95Communities Studied ...................................................................................................... 99Piquiatuba community ..............................................................................................99Chib community ...................................................................................................100Paraso community ................................................................................................. 101Prainha II community .............................................................................................102Results ...........................................................................................................................104Discussion .................................................................................................................... ..107Conclusions ...........................................................................................................................1095 CONCLUSIONS .................................................................................................................. 111LIST OF REFERENCES .............................................................................................................115BIOGRAPHICAL SKETCH .......................................................................................................122
10 LIST OF TABLES Table page 2-1 Positive and negative impacts of the use of fire in the Amazon in relation to the scale of influence and tim e. ........................................................................................................ 223-1 Design principles illustrated by long-e nduring common-pool resources institutions ....... 413-2 Concerns expressed by some families during the fire management project and the steps or arguments used to overcome those concerns. ....................................................... 553-3 Percentage of the families using techni ques and recommendations for the years of 2001 and 2007. ...................................................................................................................583-4 Analysis of Flona-Tapajs experience using the principles designed by Ostrom (1990) ........................................................................................................................ .........644-1 Average number of adults participating in the meetings as a percent of the adult population of each community for the years of the project. .............................................. 974-2 Fire management techniques in the comm unities agreements, and rules in common .... 1054-3 Number of rules in each community agreement, and minimum and maximum of rules followed...................................................................................................................1064-4 Frequency of people using the 7 rules in common in the four communities. .................. 1064-5 Percentage of the population by community following the rules .................................... 108
11 LIST OF FIGURES Figure page 2-1 Distribution of fire hot pi xels in 2008 in relation to the Am azon Arc of Deforestation and Fire. .............................................................................................................................192-2 Representation of main losses caused by accidental fires in the Amazon ......................... 233-1 Location of the communities and muni cipalities of Flona-T apajs, communities participating in the fire management project since 2001, and the communities interviewed in 2007............................................................................................................ 473-2 Fire management main activities carried out within the communities. ............................. 513-3 Average of families compliance with techniques and recommendations based on a labor classification for the year of 2007. ...........................................................................604-1 Range of participatory developmen t initiatives. Source: Stone 2003:36 ........................... 764-2 Partners at the beginning of the project their position, their in terests, their power regarding to the project and th eir level of interaction direct with IPAM. Straight lines and their denseness indicate the level of power for each partner in the project. The dashed lines and their denseness indicate th e relationship of IPAM with each partner .... 844-3 Partners after the project started, their position, their inte rests, their pow er regarding to the project and their level of interaction direct with IPAM. .......................................... 904-4 Partners in the third and fourth year, their interests and their power of decisionmaking in the project. ........................................................................................................ 91
12 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts INSTITUTIONAL ARRANGEMENTS FOR FIRE MANAGEMENT IN THE BRAZILIAN AMAZON By Maria Lucimar de Lima Souza August 2009 Chair: Marianne Schmink Major: Latin American Studies Fire is a very important element in the agricultural production system s used all over the world. It is through the use of fire that large areas of tropical forest s are burned each year, opening space for the establishment of new past ure and agricultural fields. In the Brazilian Amazon, more than 15,000 km2 per year of forest biomass have been burned and converted into crop fields and pastures in the last decade. In spite of its importance as an accessible landscape transformation tool, sometimes farmers lose c ontrol of the fire, causi ng losses, which have ecological, economic and social consequences. To avoid the losses associated with accidental fires, government and civil society have been carrying out different actions such as new legislation, as well as regi onal and local programs. In this study I analyzed some of those expe riences developed in th e Amazon that tried to decrease the numbers of accidental fires within rural regions. Specifically, the research focused on the contributions of institutional arrangements for fire management based on a project carried out at Flona-Tapajs. The analyses indicate that institutional arrangements can contribute to cope with the problem of accidental fires by in creasing the numbers of techniques and recommendations farmers apply to avoid acciden tal fires, and by increasing the number of farmers applying each technique. However, the process of building institutional arrangements
13 needs to be both participatory, re quiring strong involvement of the us ers of fire, and adaptive, so that farmers can try to improve the use of techniques from year to year, and they can modify the rules of the arrangements when they understand it to be necessary. The process of building institutional change needs to em power local people so that they can decide year-to-year what techniques and recommendations ar e important to be applied, cons idering the changes that might occur within the communities.
14 CHAPTER 1 1 INTRODUCTION Fire is the m ost reliable and least expensive means of converting forest biomass to agriculture and pasture fields. Because of that, fire is a universal tool in rural Amazonia, extensively used by small, medium and la rge farmers throughout the region. Although its benefits are very important for the land uses that are part of the region al economy, much research has called attention to the problems associated with this practice. One of these problems is escaped fire. In Brazil, every year, cattle ranchi ng and swidden agriculture are sources of ignition to forest fires (Ray, Nepstad and Moutinho 2005 ) that generate economic, environmental and social damages. After the accidental fire that occurred in the Brazilian state of Roraima in 1998, significant actions were implemented in the region focusi ng on the problem and risks of large accidental fires. These actions were carried out by th e federal and state governments and by other organizations of civil society. At the federal government level, one important action was the regulation of fire use for agri cultural activitie s through decree No 2661. Be sides new legislation, other projects and programs were carried out in Amazon with the goal of decreasing accidental fires such as PROTEGER, PDA/PADEQ and Roa Sem Queimar project. According to Nepstad et al. (2001), long-term reduction of Amazon fire, and thus its s ubstantial costs to society, can happen through investments and policies that stimulate permanent agriculture and forest production. However, they are needed in rural regi ons at the local level, in communities of small farmers, but also for medium and larger producers who still use fire to manage their properties. Research Objectives This thes is discusses the contributions of ins titutional arrangements for fire management in the Amazon through an analysis of an action resear ch project for fire management developed in
15 at Tapajs National Forest (Flona-Tapajs), a protected area located in Pa r, Brazil. This project contributed to the establishment of fire manageme nt agreements to avoid accidental fires. This thesis had two main objectives: to understand the contributions of institutional arrangements for fire management in the Amazon, based on a analysis of the project ca rried out within Fl ona-Tapajs; and to discuss the importance of local participation and partnerships to build an action research project, as well as the effectiveness of the institutional arrangements for fire management. Research Site This research was carried out at the Tapajs National Fore st, a 600,000 hectares protected area located in the State of Pa r, Brazil, where a project was developed through a partnership between an NGO and rural comm unities of Flona-T apajs with the ultimate goal of decreasing accidental fires within the protected area. Research Methodology In 2007, a survey was carried out within th e protected area to an alyze the level of com pliance with the rules established in the fire agreements, and how the communities were using the agreements. To do this analysis, two sources of data were used: (a) an assessment carried out in 2001 within 11 communities of Fl ona-Tapajs, where 71 families were interviewed at the beginning of the project; and (b) survey s carried out in 2007 in 4 communities, where 53 families were interviewed. The communities interviewed were Prainha II, Paraso, Chib and Piquiatuba. I also used my experience as part of the NGOs staff that carried out the project within Flona-Tapajs, discuss the partnerships an d local participation in the research project. Organization of the thesis The thesis h as four main chapters. Chapter Two reviews and discusses th e use of fire in the Amazon together with the problems associated with this practice. In addition, it presents attempts by governments and civil society to decrease the number of accidental fires in the Amazon. At
16 the federal government level, one important action was the establishment of fire legislation. It was established in 1998, after the big accidental fires in Roraima. Although the legislation meant that the federal government recognized accide ntal fires as a problem in the Amazon, it recommended procedures very difficult to be pu t into practice by farmers. The need to obtain authorization to burn is one example. Othe r regional programs car ried out by NGOs were effective in increasing the concerns of rural l eaders for environmental is sues, and in promoting the exchange of information among farmers in the Amazon, but there is no clear information about the contribution of those programs to an actual decr ease in accidental fires. Chapter Three discusses the contributions of instituti onal arrangements for fire management in the Brazilian Amazon by analyzing the rules established by the families, and the farmers compliance with those rules, in a fire management program. The case of Flona-Tapajs shows that institutional arrangements can positiv ely contribute to addressing the problems of accidental fires. However, the process of building institutional arrangements needs to be adaptive so that farmers can improve the use of techniques from year to year, and change the rules when they deem it necessary. Chapter Four discusses the importance of local participation and partnerships in building an action research project and ultimately, the eff ectiveness of the institutional arrangements that were developed during the fire management project at Tapajs National Forest. The case of Tapajs National Forest shows that the engage ment of many different stakeholders present several challenges. It takes time, resources, continuous de bate of interests and goals and constant adaptation, but the openness to participation by different stakeholders ultimately resulted not only in the execution of the action research projec t, but more importantly, in the achievement of the main goal of the partnership itself: the reduction of accidental fires at Flona-Tapajs.
17 Secondly, the chapter analy zes the relationship between the numb er of people par ticipating in the process of formulating local agreements at Tapa js National Forest and the level of compliance with the fire management agreem ent rules. The results indicate that there is no significant difference in the average number of rules applied by the families of communities with high level of attendance in comparison to the communities that had lower levels of attendance at the meetings to formulate fire agreements. However, while at the community level, numbers did not seem to be decisive for the success of the initiativ e, the engagement of different stakeholders and their participation in different moments, seems to have contributed to legitimating the project and thus, contributing to the achievement of the part nerships main goal, which was the reduction of accidental fires. Finally, Chapter Five presents the conclusions of this research. It indicates that the attempts to build the participatory action resear ch not only contributed to the execution of the project, but were also fundamental to the effect iveness of the institutional arrangements for fire management that were developed during the project at Tapajs National Fo rest. Thus, the case of Tapajs National Forest shows that a partners hip between an outside r organization and the families increased the concern of the people regardi ng fire, and that local participation is not just simply desirable for fire management, but rather is a requirement. Without involving the users of fire in the processes of looking for solutions to accidental fires, desi rable results cannot be achieved, because the farmers are the ones that ul timately are going to use fire, deciding which techniques and recommendations to use to avoid accidental fires. Just as important, the results of this research indicate that it is possible to develop partnerships that involve both outsiders and locals and thus to contribute to conservation initiatives in ways that do not ignore the needs of the local people.
18 CHAPTER 2 1 THE USE OF FIRE IN THE AMAZON: IM PORTANCE, P ROBLEMS AND ATTEMPTS TO DECREASE ACCIDENTAL FIRES Introduction Fire is a very im portant element in the ag ricultural production syst em used all over the world. It is through the use of fire that large areas of tropical forest s are burned each year opening space for the establishment of new pasture and agricultural fields (Conklin 1954; Nepstad, Moreira and Alencar1999; Ickowitz 2006; Vayda 2006). In the Brazilian Amazon, which holds the largest tropical fo rest of the world, about 18,000 km 2 per year of forest biomass have been burned and converted into crop fields and pastures in the last decade (INPE 2008). The distribution of fire in the region follows th e spatial pattern of the roads and deforestation, defining the area where the agriculture frontie r is being established (Wood 2001). This area known as the Arc of deforestation or Arc of fi re (Figure 2-1) has been a major focus of government and non-government intervention in relation to fire prevention and control. The high occurrence and frequenc y of fire hot pixels in th e Brazilian Amazon indicate that it is a tool being widely used by small, medium and large farmers and producers in the region. The popularity of fire in the Amazon production system is associated with the relative low cost of burning and short terms benefits associated with it. In spite of its popu larity and benefits to rural producers, there are several risks associated with the use of fire in the Amazon. Fire can easily spread and escape from its intended ar ea, affecting neighbori ng areas and land uses. Accidental fires can promote damage and loss of infrastructure, agricultural systems, pasture fields and forests resources, which can stimulate more extensive and unsustainable land use practices (Nepstad, More ira, and Alencar 1999).
19 Figure 2-1. Distribution of fire hot pixels in 2008 in relation to the Amazon Arc of Deforestation and Fire. The risks of escaped or accidental fires are likely to increase in a future of warmer climate and population growth. These two elements are e xpected to affect the flammability of the vegetation and increase the sources of ignition, respectively, causing more accidental and recurrent fires (Cochrane 2003). In fact, there is already an indicat ion that dense forest is losing its resistance to fire due to the disturbances promoted by human activities in forest areas (Alencar, Nepstad and Diaz 2006; Nepstad et al. 1999). These dist urbances associated with extreme drought events have caused large-scale bu rns (Alencar et al. 2004; Arago et al. 2008). In the last two decades alone th e Amazon suffered large-scale accide ntal fires such as the one in 1998 in the state of Roraima and the one in Ac re in 2005 (Arago et al. 2008; Barbosa and
20 Fearnside 1999). Both events represented a benchm ark that called the atte ntion to the issue of fire in the Amazon. Both government entities and other organizations started several reactions and initiatives to decrease fire use, and to prevent and control accidental fires in the region. In this chapter I discuss the use of fire in the Amazon together with the problems associated with this practice. In addition, I present some of the attempts by governments and civil society to decrease the num ber of accidental fires in Amazon. The chapter is organized in two parts. In the first I focu s on the importance of fire for agriculture, and the problems associated with this practice. In the second part I discuss some attempts by the government and other organizations, at different scales, to decr ease accidental fires within the region. The most important Brazilian legislation abou t fire was established in 1998, af ter the big accidental fires in Roraima. Although the legislation meant that the federal government recognized accidental fires as a problem in the Amazon, it recommended procedures that are difficult to be put into practice by farmers. The need to obtain authorization to burn is one example. Other regional programs carried out by NGOs are effective in increasing the concerns of rural leaders for environmental issues, and in promoting exchange of inform ation among farmers in Amazon, but there is no clear information about the effect of those program s to decreasing incidence of accidental fires. The Importance and Problems Associated w ith the Use of Fire in Amazon Fire is part of the system of production of four million rural people in Amazonia (Nepstad et al. 2001). One of the main economic actors in the region that uses fire as an important agricultural tool is the small farm er. According to data from the Instituto Brasileiro de Geografia e Estatstica IBGE (Brazilian Geographical and Statistical Institute) there are more than 600,000 small farmers in the Amazon (Almeida, S ouza, and Rodrigues 2006:155). These farmers own land parcels smaller than 100 hectares, a nd have production systems based on the family labor force. The average size of their fields is around 1 to 3 hectares per year (Walker and
21 Homma 1996). They represent 70 % of the total rura l population of Amazon, and are responsible for 36% of the agricultural Gross Domestic Pr oduct of the region (IBGE 1998; Vosti, Witcover, and Carpentier 2002). These families use fire as an agricultural tool to prepare and manage fields for crops and grazing. In general, farmers consid er fire practical, chea p, and quick. In fact, almost none of them have the economic means a nd/or knowledge to use alternative techniques or to implement alternative agricultural systems that do not need fire for maintenance. Besides, most of these farmers live in areas of poor soils, wh ere forest is still abundant and land is usually cheap. The labor available is from their family and they have little capital to start alternative systems of production. In such conditions, fire represents the most viable alternative to provide nutrients to the soil through the combustion of forest biomass in ash-fertilized fields, being the perfect substitute for fertilizers, pesticides and machinery (Nepstad et al. 2001: 395). Following this logic of fire use in the Amazon, there are se veral benefits and costs associated with this practice that can dir ectly or indirectly affect farmers. Thes e benefits include: c onversion of forest biomass to nutrients, decrease in the costs of pr oduction, control of diseases and weeds, and use of available technology and knowledge. The costs, known as impacts, include: soil impoverishment, CO2 emissions, air pollution through smoke, and the risk of accidental fires (Table 2-1). These impacts vary in the scale and in the way they affect users of fire, as well as the time frame in which they feel the consequen ces of this activity. In general, most of the positive impacts directly affect the fire user, in a very short time period (Table 2-1). On the other hand, most of the negative impacts or costs associated with fire do not nece ssarily affect the fire user, and when they do, the effects of the practice of using fire are felt in the medium to long term. These relationships help explain why fire is such an important tool in the rural Amazon context, and why it is very difficult to exclude it from the farming system.
22 Table 2-1. Positive and negative impacts of the use of fire in the Amazon in relation to the scale of influence and time. Impacts of fire Type of benefits and costs Scale of influence Time to detect Positive1 Negative2 Direc t Indirect Short term Medium term Long term Conversion of biomass to nutrients x x x Decrease in costs of production x x x Control of diseases and weeds x x x Use of available technology x x x Soil impoverishment x x x x CO2 emissions x x x Smoke x x x x Risk of accidental fires x x x x x 1 Positive impacts represent the benefits of using fire for a rural producer 2 Negative impacts represent the costs or prob lems of using fire for a rural producer From this list of benefits a nd costs, the one of greatest policy concern is the risk of accidental fires. In many cases, when farmers burn their fields, they lose control of fires and inadvertently burn forest, agricultural systems, pasture, and other kinds of vegetation, such as secondary forest. Accidental fires are responsible for 50% of the area burned every year in the Amazon (Nepstad, Moreira, and Alencar 1999) and can cause many losses (Figure 2-2). The magnitude of unwanted fires in the Amazon has generated government and societal concern about the risk of large-scale forest fire ev ents and has transformed fire into the main environmental problem in the region. Increasing Risk of Forest Fires Forest fires in the Am azon dense forest used to be rare, and associated with Mega-El Nino events in pre-Colombian times (Meggers 1994, Sa ndford et al. 1985). This infrequent natural fire regime has been disrupted by the expans ion of anthropogenic activ ities such as logging,
23 Figure 2-2. Representation of main losses caused by accidental fires in the Amazon Fires Controlled for production Accidental fires Forest Opened areas Loss of livestoc k Loss of infrastructure Change of b ehavio r Loss of means of p roduction Climate change Increase in flammability Loss of timber and NTFPs
24 cattle ranching and large and small-scale agricult ure and road infrastructure (Alencar et al. 2004; Cochrane 2003; Nepstad et al. 2001; Uhl and Kauffmann 1990). Th ese activities are responsible for increasing forest flammability by promoting changes in the forest microclimate, through the increase of edge effect caused by fragmenta tion and deforestation, and the increase of canopy openness promoted by logging (Cochrane 2001; Ferre ira. and Laurence 1997; Holdsworth and Uhl 1997; Ray, Nepstad and Moutinho 2005; Uhl and Kauffman 1990). The large number of families using fire every year combined with the increased flammability of Amazon forests during El Nio-Sout hern Oscillation (ENSO) events, results in increased probability of accidental fires (Alencar et al. 2004) such as those that occurred in the Brazilian states of Roraima in 1998 or Acre in 2005 (Barbosa and Fearnside 1999; Arago et al. 2008). The total area burned in Roraima was 38,144 40,678 km2, where about 55% of the fire affected primary forest (Barbosa and Fearnside 1999). The area of forest fires in the Amazon is estimated to vary from 200,000 to 1.600,000 ha in years of non-ENSO and ENSO, respectively (Alencar, Nepstad and Diaz 2006:10). Besides accidental forest fires, the area burned after deforestation to establishment to agricultural fields has averaged 1.800,000 ha in the last two decades (INPE 2008). Ecological Losses Accidental fires can provoke eco logical, economic and social dam ages. The effects of fires in the Amazon region have global effects on the Ea rths system because they alter the chemical composition of the atmosphere and the Earth's reflectance (Artaxo et al. 2005). Among them, the most important changes are altera tions in the dynamics of nutrie nts and biomass, changes in the hydrological cycle, and reductions in the number of local plants and animals (Barlow, Haugaasen and Peres 2002; Gerwing 2002; Haugaasen, Barl ow and Peres 2003; Holdsworth and Uhl 1997; Nepstad et al. 1999; Ray, Nepstad and Moutinho 2005).
25 Fires can damage the structure and composition of the forest (Alencar et al. 2005; Barbosa and Fearnside 1999; Cochrane and Schulze 1999; Gerwing 2002; Haugaasen, Barlow and Peres 2003). Cochrane and Schulze (1999:2) found that about 50% of trees di e after a forest fire. This number increases and varies depending on the inte nsity and frequency of the fire. Thus forest structure is expected to change dramatically after fire since large trees have an inverse relationship with fire intensity and frequency, while pioneers tend to increase with burn intensity, generating an area dominated by pioneers (Cochrane and Schulze 1999). Further, once a forest burns, the probability of another accidental fire increases. Fires increases the flammability of the Amazon lands cape and initiates a positive feedback under which tropical forests are gradually substituted by species more prone to fire (Nepstad, Moreira and Alencar 1999). Once a forest is burned, trees ar e killed and start to sh ed their leaves, opening the canopy, increasing the amount of solar radiation and the amount of fuel material on the forest floor. The insulation changes the internal air hum idity and temperature of the forest, drying the leaves and dead woody biomass on the forest floor. Economic and Social Losses In addition to the eco logical damage provoked by accidental forest fires, there are also economic and social impacts of fire in th e Amazon region. Every year farmers lose wood, pasture, fences, other investments and productive systems that are worth millions of dollars. Economic damages of having pasture fields and fences burned alone vary between US$12 to $97 million per year (Diaz et al. 2002). The annual economic losses associated with burned timber inside the forest vary between US$ 1 and 13 m illion. Scaling up, these losses represent 0.1% and 2% of the GDP of the Amazon, and 0.2% and 1.6% of GDP of the agricultu ral production of the Amazon region.
26 Within rural communities families accumulate damage when they lose pastures and crops by accidental fires. In 1998, a survey carried out in Del Rei, a community in northeastern Par, Brazil, showed that 98 percent of the families had one or more accidental fires on their properties during the six-year period covered in the study. In most cases, the study s howed that accidental fires started within the community where th e damage occurred, where in 1997, ninety-eight percent of the accidental fires began in the community fields, where families lost timber, animals, productive agricultural systems and pastures. Because of the pasture losses, some farmers had to rent pasture from their neighbors. Also in some cases, the accidental fires burned fields that were not prepared to burn. Because of that some families could not produce adequately, since the competition between plants increases when fields do not burn well. There are also social damages associated with accidental fires. They include conflict between farmers, and an increase in poverty levels due to the constant risk of accidental fires that stimulate extensive forms of land use. Farmers are not always inclined to invest in more sustainable forms of land use (Nepstad et al. 2001: 399) when the constant risk of accidental fires can destroy years of work and investments. Respiratory problems are also among the negati ve impacts of fire. Every year the local hospitals in the rural areas of the Amazon receive hundreds of pe ople with respiratory problems due to the constant effect of smoke inhalation promoted by burnings. These costs represent around 0.2% of the regions GDP during the 1996 -1999 period (Diaz et al. 2002; Mendona et al. 2004:89) Looking for Solutions: the Attempts to Decreas e Accidental Fires in the Brazilian Amazon In spite of the importance of fire in the Amazon, only in the last decades did government and society start to pay attention to the problem that this practice may cause. The debate arose in the beginning of the1990s about increasing defore station, with the releas e of the first annual
27 deforestation statistics for the Amazon by the Instituto Nacional de Pesquisas Espaciais (INPE). The issue of deforestation and fire use gained political force during the ECO 92 Summit in Rio and became one of the main foci in the agenda of the Pilot Program to Conserve the Brazilian Rain Forest (PPG7). In 1996 the World Bank, worried about increasing of de forestation and fire in the Amazon, hired a research team to unders tand and quantify fire in the region. This study generated a book that brought public a ttention to the fact that about half of the area had burned in the region had burned accidentally (Nepstad, Moreira and Alencar 1999). The Amazon Environmental Res earch Institute (IPAM), an NGO based in Belm, Par, conducted this study. At that time this NGO wa s the main protagonist in promoting the discussion of fire. After the study was concl uded, IPAM organized two workshops, one in Braslia and another in Belm, to present to th e government the concerns of many scholars about the probability of widespread ac cidental fires in the Amazon forest. The increasing concern about fire issues gained force when much of the state of Roraima burned in 1998. After the publication of the study results a nd the accidental fires of Roraim a, many activities took place in the region, and different actors started looking for solutions. Legal Attempts to Decrease Fire The federal governm ent has been creating a nd changing laws and programs to approach fire use in Amazon, and to create a legal framewor k to regulate this prac tice. Brazilian attempts to regulate fire use are not new. In the XVII Century there was a law in Brazil that forbade the use of fire in areas where pau brazil ( Caesalpinia echinata) could be found, through the regulation of December 12th 1605 (Ramos 1998:4). In 1934, decree No 23.793 established the Brazilian Forest Code; it forbad e the use of fire without previous permission of the forestry authorities (Ramos 1998). In 1965 the federal government created Law No 004771 that regulated the use and protection of forests in Brazil. In that law fire appear s in four articles (11, 25, 26 and
28 27) in very general terms. This law was the first that made the setting of fire in forests or other vegetation types without precautions a crime, and clearly forbade the use of fire. However, in its unique paragraph, it says that if local and regional aspects justify the us e of fire, then, the government can permit its use with norms of precaution which were left undefined. From 1965 to 1979 there were other federal act ions regarding the environment, but none specifically addressed the issue of fire use. However, some of these actions were important to forest conservation and control of deforestation, such as the creation of the Instituto Brasileiro de Desenvolvimento Florestal (IBDF), and the creation of the Brazilian Environmental Secretariat. In 1979, through decree No 84.017, the federal gove rnment regulated national parks, and forbade any activity that could provoke fire within national parks. In 1981, the government created a set of institutions to deal with environmental issues, among them the National Environmental Policy (PNMA), the National Environmental System (SISNAMA), and the National Environmental Council (CONAMA) and instituted the Environmental Defense Register. In 1983, the government instituted a na tional campaign to prevent agricultural and forest fires (regulation No 000326 of 1983), probably in respon se to the ENSO of 1981-1982. The Brazilian Constitution of 1988 brought in an enti re section about environment, but without any consideration of a specific problem or pr actice. In 1988, the IDBF through regulation No 000231, required authorization for the use of controlled burns, as was establ ished in the Brazilian Forest Code of 1934. Still in 1988, the IBDF, created the National Commission of Prevention and Control of Forest Fires (CONACIF), with the objective of coordina ting and regulating the use and control of fire through regula tion No. 000254. In 1989, decree No 97.635 regulated article 27 of the Forest Code, and created the National System of Prevention and Control of Forest Fires (PREVFOGO). In 1991, through cooper ative agreements, the government created an
29 emergency program to control deforestation and fire in Legal Amazon, establishing agreements with the Amazon State Environmental Agenci es (OEMAs). In 1993, a regulation of the Environmental Ministry (MMA) promoted a campaign called Liv e Amazon to control illegal practices of burning and deforestation in the region. During 1996 and 1997, a multilateral agreement between MMA, the German Agency for Technical Cooperation (GTZ), the United Kingdom Department for International Developm ent (DFID), and the World Bank created the Pilot Program for the Protection of Tropical Fore sts in Brazil (PPG-7), which later would support many other different programs regarding to fire monitoring systems and fire prevention. In 1998 decree No 2661 redefined the attr ibutes of PREVFOGO, and esta blished specific rules for the use of fire. Still in 1998, decree No 2662 created the Program for Prevention and Control of Fires in the Brazilian Amazon Forest (PROARCO) as a response to the Roraima fires and international pressure. The law No 9605, of 1998, established the Law of Environmental Crimes. Its article 41 makes a crime to set fire in forests. In 1998, the Brazilian Institute of Environment and Renewable Natural Resources (IBAMA) published a regulation regarding controlled burning and its procedures, and also esta blished the practice of solid arity burning in which rural communities could organize themselves to get one collective authorization to burn their fields, instead of many different authorizations for many different families within a community. In 2000, decree No 3420 created the National Forest Program (PNF), in which the FLORESCER Program was specifically established to work with prevention and combat of deforestation, burning and forest fires. In 2003, the federal government created th e Plan of Action to Prevent and Control Deforestation, Burning and Illegal Logging in the Brazilian Amazon. The review of fire legislati on shows that the use of fire a ppeared in some laws before 1998, such as the Brazilian Forest Code of 1934, and the decree No 97.635 of 1989 that created
30 the PREVFOGO. However, although fire was noted, it did not appe ar in specific regulations before 1998. It seems that the federal government only responded to the problem after Roraimas accidental fires in 1998. This event had national and international repercussions, and in that year, the federal government created decree No 2661 redefining the attributions of PREVFOGO, and establishing specific rules for the us e of fire. This was the first regulation that went to the level of defining procedures to be used by fa rmers during burning of their fields. Governmental Programs and Other Initiatives to Decrease Fire Regional level Besides laws and regulations there are two program s of federal government that require more attention. The first, called PROARCO, deve loped emergency actions in partnership with many state and local governments. The second on e called Demonstrative Projects (PDA) had a very practical configuration to decrease accidental fires. PROARCO was created in the context of the dry season of 1998, and the consequent risk of repetition of accidental forest fires such th e one that happened in Roraima state at the beginning of that year. The project was carried out by IBAMA with financial support from the World Bank, and started its activit ies in 2000 (Sauer et al. 2004). The main goal of the project was to prevent and/or control large-scale wildfi res in the Brazilian Am azon during the dry season of 1998, and generate lessons to ensure conditions were created to prevent large fires in the midterm (MMA 2009). According to the Environmenta l Ministry (2009), there were five benefits expected to be achieved with the project. They were: a) Reduce occurrence of large-scale fires a nd potential economic a nd social losses to residents of the Amazon; b) Enhance protection of the Amazon rain forest an d the environmental services it provides; c) Improve knowledge of how to preven t escaped fire in this region;
31 d) Foster sustainable partnerships among federal, state and civil society organizations in an emergency setting; and e) Develop a system of rapid response to fires which could be used in future emergencies, including other recurr ent natural disaster s and accidents. The program had 4 main components: a) Risk assessment and monito ring of critical areas; b) Forest fire prevention; c) Forest fire suppression; and d) Project coordination, monitoring and eval uation, and studies (World Bank 2005). The project aimed to work w ith rural populations living along of the Deforestation Arc, urban populations of the Amazon, state and regi onal electric power generation and distribution companies and their users, indigenous people, and society in general. Established in 1995, the PDA was developed by the Environmental Ministry (MMA), and started its activities in 1996. This program, a result of the intern ational cooperation that created the PPG-7, was a sub-program of the PPG-7. Its main goals were (a) to support local organizations to develop innovati ve experiences to promote sustainable development; and (b) to promote the registration, diffusion and incorpor ation of those experi ences to other rural communities, organizations and governmental ins titutions. In its first period, the PDA supported 147 projects within th e Amazon region (MMA 2009). In 200 4, the PDA started a new period with three new components. One of these, the A lternatives to Deforestat ion and Burning Project (PADEQ), promoted the sustainability of rural prope rties, reducing or eliminating the use of fire through the implementation of alternative systems of production. PADEQ has supported 49 projects within the Amazon (MMA 2009). The state governments of the Amazon also have been looking for solutions to the problem of accidental fires. One example of state gover nment action is the case of the government of Roraima to the accidental fire in 1998. In response that fire, the state government, in partnership
32 with IBAMA, created the State Committee for Prev ention and Control of Burning and Combat of Forest Fires. This committee inst ituted groups of fire fi ghters within some municipalities, trained to combat fires, and also to support local rural communities in carrying out controlled burnings. The state committee established partnerships with state government secretaries, IBAMA, Military Police, National Institute for Coloni zation and Agrarian Reform (INCRA), National Institute of Amazonian Research (INPA) and Ci vil Defense. In general this committee had the goal of coordinating actions to avoid accidental fires within Roraima (Sousa Jr. 2006). However, although the Roraima government was involved in th e committee, at the lo cal level the structure created to implement actions had no financial support, nor adequate tr aining and materials. IBAMA acts mostly in emergency situations, and there is no institutional plan to avoid and control accidental fires for Roraima. Likewise, the other organizations involved in the committee did not carry out structured and continued actions to prevent and combat accidental fires. The state government did not support the committee to consolidate itself and act, and also did not have a plan to change the agricultural system in order to decrease the need for fire use as an agricultural tool (Souza Jr. 2006). From regional to local level There were also reg ional level experiences trying to decrease accidental fires number. The Mobilization and Training of Family Farmers fo r the Prevention of Accidental Fires in the Brazilian Amazon Project (PROTEGER) is one exam ple of a regional-scale project which started in 1998. It was a result of a partnership betw een Grupo de Trabalho Amaznico (GTA), with Amazonian farmers unions through PPG7 and PROARCO. The project had the financial support of USAID (U.S. Agency for Internati onal Development), and was developed in two distinct periods. The first pe riod, in 1998 and 1999, had an emergency character (Sauer 2005), and tried to develop activities to sensitize Amazonian farmers to the use of methods that could
33 control fires. PROTEGER also stimulated the organization of community groups of people called brigadas comunitrias, to work locally within their own communities to decrease accidental fires. The Project collected informa tion about systems of production that could be alternatives to fire use. At the end of two years, PROTEGER had worked in 322 municipalities in the nine Amazon states. It organized 400 training courses and workshops in which 200 local leaders participated. Those local leaders, called monitores and monitoras, were responsible to disseminate the information within their communities. Between 2001 and 2004, based on the learning experience of PROTEGER in its first period, the GTA coordinated PROTEGER II. In its second period, the general goals of the project were to contribute to the preservation of the Amazonian ecosystem and to the improvement of life quality of the local population, and to prevent forest fires set by family farmers in Amazon. The specific goals were: a) increase the level of burning control by family farmers; b) promote forms of sustainable pr oduction without fire use; c) stimulate recovery of deforested areas; d) strengthen local organizations fo r environmental management; and e) influence public policies that could contribute to the reduction of accidental forest fires in the Amazon (Lins 2005). A group of consultants analyzed the many re sults of PROTEGER (Sauer et al. 2004). Some of those results were (a) the establishment of a channel to think about, debate and increase the concerns of rural union leaders with envi ronmental problems within the Amazon; (b) the exchange of information about alternative syst ems of production, between farmers of different regions; and (c) the formation of local groups to mobilize families within communities. Consultants concluded that the challenge of decr easing accidental fires was difficult to verify
34 because there were no data availabl e about numbers of fires and their origins, but they noted that at least there was more engagement am ong farmers to avoid accidental fires. Local level Many local projects in the Amaz on have tried to decrease accidental fires and/or diminish fire use within rural properti es. An example of such local level action is case of the Roa Sem Queimar Project. This project was ca rried out in 11 municipalities of the Transamazon region, state of Par. With support from the Secretaria de Coordenao da Amaznia (SCA) and MMA, it was carried out by Fundao Viver, Produzir e Preservar (FVPP), a local non-governmental organization. The goal of the project was to e xperiment with crop production without using fire to prepare the fields. The projec t was carried out in partnership with 150 participating farmers, and with rural labor unions of the municipalities. Each farmer got a fund of US 500.00 per year to experiment with producing on one hectare wi thout using fire. In addition the farmers got training and technical support (Sau er et al. 2004). The farmers i nvolved in the project could not use fire on that one hectare, nor could they use ex ternal inputs to manage the system. They had to follow agro-ecological principles such as: a) conservation of biodiversity by keep ing trees with economic value; b) maintenance of soil cover by planting species to produce biomass and organic matter for the soil; and c) planting diversified plants in th e same area (Sauer et al. 2004). In an analysis of the project in 2004, one of the principal problems cited was the fact that it was necessary to wait one year between the in itial preparation of the field and the planting process, because it takes time for the organic ma tter to decompose (Sauer et al. 2004). Another problem was that some farmers found it difficult to believe in agro-ecological principles. Another difficulty was the pres ence of many branches and trunks in the field, which impede
35 farmers movements while working. Despite these difficulties, there were positive points, such as the fact that the farmers were directly involved in the planning and actions of the project. According to Sauer et al. (2004), farmers experiences were valu ed, and they learned from the training opportunities, but more importantly fr om trying to produce wit hout the use of fire. Although more research is necessa ry, the project demonstrated efficiency in producing crops such as cocoa, a very important crop in that region, as well as banana s (Sauer et al. 2004). Some Thoughts on Fire in the Amaz on and the Experiences Carried Out In 1998, the scientific comm unity knew that the Am azon forest could suffer large accidental fires, and had tried to call the atten tion of the government to that, however, only after the case of Roraima, in 1998, did significant actions take place in the region focused on the problem and risks of large accidental fires. Th ese actions were carried out by the federal and state governments and by other organizations of ci vil society. At the federal government level, one important action was the regulation of fire uses for agricultural ac tivities through decree No 2661. It was important, first, because it meant recognition by the government of the problem. Also the laws gave accidental fires the character of a crime and specified who could use fire and under what conditions. However, the structure of the government in Amazon is too weak to guarantee the enforcement of the laws, and secondly fire laws created by the federal government are distant from those who actually use fire to pr oduce. In my experience working in more than 30 rural communities in different municipalities of different stat es in the Amazon, I never found one farmer who was aware of federal legisla tion about fire. For those reasons, laws and regulations alone are not enough to decrease the number of accidental fires in Amazon. Fire is still going to be part of the agricultu ral system in the Amazon for the next decades. There are two basic important things that need to be done to eff ectively decrease accidental fire in Amazon. First, educational program such as PROTEGER are important, because as pointed
36 out by Sauer et al. (2004), educational activit ies promote local responsibility by people for environmental issues. Second, a massive investme nt is needed in alternative systems of production that can help to decrease the need to burn every year, and/or to exclude fire completely from the system of production, such as the PDA/PADEQ and Roa Sem Queimar projects. According to Nepsta d et al (2001), long-term reduc tion of Amazon fire, and its substantial costs to society, can happen th rough investments and po licies that stimulate permanent agriculture and forest production. However, they are needed in rural regions at the local level, in communities of sm all farmers, but also for medium and larger producers who still use fire to manage their properties. As point ed out by Souto (2003) in an analysis of the government program PREVFOGO, those programs need to work directly within rural communities. Conclusions In this chapter I reviewed and discussed the use of fire in the Am azon together with the problems associated with this practice, and pr esented some attempts by government and civil society to decrease the number of accidental fi res in the region. The an alysis indicates that changes in the use of fire in Amazon are not going to happen from one year to another. Fire is so important in the agricultural system of the region that important changes need to be carried out through initiatives that engage direct the users of fire locally, and that are established through adaptive process where people can experience new models, evaluate it, and adapt according to emerging conditions. Although the fire legislatio n of 1998 is important as federal government recognition of accidental fires as a problem in the Amazon, it recommends procedures that are difficult to be put into practice by the farmers. Th ere are some examples of the difficulties faced by the farmers to follow the law: the bureaucratic procedures that a farmer has to go through to obtain authorization to burn, and also the size of fire break requi red by law, that according to the
37 farmers is too large and requires lots of labor Other regional programs carried out by NGOs are effective in increasing the concerns of people ab out environmental issues in general, but not specifically about fire management. In the next chapter I discuss a project deve loped at the local level in eighteen rural communities of a protected area in the north region of Brazil. Th rough a partnership between an Amazonian NGO, IBAMA, the Rural Labor Union of Belterra municipality, and rural community organizations, the project was carried out with the clear goal of reducing accidental fires. The experience was based on the assumption th at effective fire management requires direct involvement of fire users, and a continuous process to promote changes. Institutional arrangements for fire management were established and they were: a) elaborated by the users of fires; b) adequate to their reality; c) experimented and adapted by them through the years, and d) required a strong level of partnership and pa rticipation among the people and organizations involved.
38 CHAPTER 3 1 INSTITUTIONAL ARRANGEMENTS FOR FI RE MANAGEMENT: THE C ASE OF TAPAJS NATIONAL FOREST PAR BRAZIL Introduction There are many important factors contribute to the occurrences of accidental fires in the Amazon. Climatic conditions such as extreme drought events, the characteristics of the vegetation surrounding the area to be burned, and the fragmentation status of this vegetation are some of the elements that can contribute to th e occurrence of accidental fires (Alencar et al. 2004). However, there are techniques used by the farm ers to prepare their fields for the burn that are very important to avoid accidental fires. The decision of using one or more of these techniques is taken by the farmer at the indivi dual level. Even though the decision is individual, the consequences of an escaped or accidental fire can be shared, and affect every family in one community or even in neighboring communities. B ecause of this risk, some communities in the Amazon have been trying to reduce the risk of accidental fires through institutional arrangements where rules of how to use fire were established by the families living within each community. In this chapter I discuss the contributions of institutional arrangements for fire management in the Brazilian Amazon by analyzing the rules established by the families, and the farmers compliance with those rules, in a fire management project. The case of Flona-Tapajs shows that institutional arrangements are effectiv e in the short and medium terms in coping with the problem of accidental fires. However, th e process of building institutional arrangements needs to be adaptive so that farmers themselves can try to improve the use of techniques from year to year, and change it when necessary. The chapter is organized in three parts. In the first part I present the literature about governance of the commons, and how institutional arrangements have been used by different groups in different parts of the world as a way to avoid destruction of th eir natural resources and
39 of their source of livelihood. In th e second part, I present the expe rience of fire management in the Tapajs National Forest where institutional arrangements were developed as a solution for accidental fires. Finally, I draw upon the common property principl es designed by Elinor Ostrom to analyze institutional arrangements developed at the Tapajs National Forest. Institutional Arrangements to Avoid the Tragedy of the Commons In 1968 Garret Hardin published an articl e called the Tragedy of the Commons. The tragedy of the commons refers to peoples tende ncy to destroy natural re sources when they are not privately held. According to Hardin, indivi duals will always exploit public resources to maximize profits. He gave an example of a comm on pasture where every herdsman can place as many cattle as possible because as a rational being, each herdsman seeks to maximize his gain (Hardin 1968: 1244). The result is tragedy, b ecause overgrazing will ruin the resource and everybody will have problems. To Hardin, freedom in commons systems brings ruin to all. Another important aspect of the commons is the tendency to share losses and damages. Here is not a question of taki ng something out of the commons, but of putting something in (Hardin 1986: 1245), but the rationality is simila r to the cases of taking something out, according to the author. Using an example of the use of wa ter as a common resource, Hardin explained that the rational man finds that his sh are of the cost of the wastes he discharges in the commons is less than the cost of purifying his waster before releasing them (1986: 1245). The result, again, is the tragedy by f ouling our own nest so long as we beha ve only as independent, rational, freeenterprisers (1245). To avoid th e tragedy of the commons and the destruction of the common resources, Hardin proposed privatization or state control, as the two possible alternatives. In his view, the appeal to conscience or to peoples sense of responsibi lity would not work without a system of sanctions and coercion mutu ally defined by those people affected.
40 Other scholars have challenged Hardins theor y. The principal argument is that Hardin did not recognize the capacity of i ndividuals involved in situations where resources are common properties to use their knowledge to create rules and to ch ange problems they face (Gautam 2005). Many recent scholars have written about the use of common-pool resources and the ways communities and other users have organized themselves to create institutional arrangements that avoid the tragedy of commons. To Agrawal ( 2001:1649), scholars working with the commons have shown that resource users organize themselv es to establish systems and regimes where they allocate benefits of resource us e equitably, for long periods of tim e. Ostrom (1990) discusses the best ways to limit the use of natural resources so they will not be completely exploited and can be continuously used by people over time. The principal point she makes is that individuals develop strategies to govern and manage th eir common resources a nd avoid the tragedy. Common-pool resour ces (CRPs) refer to a natural or man-made resource system that is sufficiently large as to make it costly (but not impossible) to exclude potential beneficiaries from obtaining benefits from its use (Ostrom 1990: 30). To study cases in which people developed stra tegies to avoid environmental destruction, Ostrom limited her research to situations wh ere renewable resources were scarce, and where users could harm one another. After analyzing many cases of long-enduring, self-organized, and self-governed CRPs, Ostrom (1990) presented a list of principles (Table 21) that are indicators of a successful collective process of management. Based on this list, principle 1 refers to the im portance of clear defined boundaries, so that it is clearly established who has ri ghts to withdraw and use the re sources in question, and also to define the limits of the resource. According to Ostrom (1990), the definition of boundaries should be the first step in organizing for colle ctive action. Just the defining of boundaries, users
41 and not authorized users, can give a certain level of guarantee that the efforts of those involved in collective action will be worth it. Table 3-1. Design principles illustrated by l ong-enduring common-pool re sources institutions Principles Characteristics 1. Clearly defined boundaries: Individuals or households who have rights to withdraw resources units from the CRP must be clearly defined, as must the boundaries of the CRP itself 2. Congruence between appropriation and provision rules and local conditions : Appropriation rules restricting time, place, technology, and; or quantity of resources units are related to local conditions and to provision rules requiring labor, material, and; or money. 3. Collective-choice arrangements Most individuals affected by the operationa l rules can participate in modifying the operational rules 4. Monitoring Monitors, who actively audit CPR conditions and appropr iator behavior, are accountable to the appr opriators or are the appropriators. 5. Graduated sanctions Appropriators who violate operational rules are likely to be asse ssed graduated sanction s (depending on the seriousness and context of the offense) by other appropriators, by officials accountable to these appropriators, or by both. 6. Conflict-resolution mechanisms Appropriators and their officials have rapid access to low-cost local arenas to resolve conflicts among appropr iators or between appropriators and officials. 7. Minimal recognition of rights to organize The rights of appropriators to devise their own institutions are not challenged by external governmental authorities. 8. Nested enterprises (For CRPs that are part of larger systems) Appropriation, provision, monitoring, enforcement, conflict resolution, and governance activities are organized in multiple layers of nested enterprises. Source: Adapted from Ostrom 1990 page 90. Principle 2 refers to the need for congruence between appropriation ru les, provision rules, and local conditions. Local partic ularities can make it necessary to have different rules for
42 different places even when the ru les are set up to govern the same kind of resource in a way that consider proportional equiva lence between costs and be nefits (Ostrom 1990; 2005). Principle 3 refers to the possibility that in dividuals directly affected by the institutional arrangements and its rules can modify them wh en necessary. When there are clear boundaries, there is congruence between a ppropriation, provision and rules, and when the users can contribute to the creation of the rules and change them, there is a greater likelihood that the rules will be appropriate to the case to which they. Ho wever, it does not necessarily mean that the users will follow the rules. It is easier to agree with rules, than to follow them after they are elaborated (Ostrom 1990). Principle 4 refers to the need for a monitori ng system. Trust and reciprocity among users to avoid breaking rules are charac teristics of only a few long-su rviving regimes (Ostrom 2005). A monitoring system is therefore needed that ca n identify those who are and are not following the rules. Principle 5 refers to the need for a sanction sy stem for those who viol ate established rules. In the analyses of successful cases, Ostrom ( 1990) found that it was local people, instead of external authorities, carried out monitoring and sanctions, and that initial sanctions were very low considering the expected benefit-cost ratio of breaking the rules. It seems that the initial sanctions were more to inform the one breaki ng the rules. According to Ostrom, the sanction system is more like a quasi-voluntary cooperation system than coerced cooperation. The monitoring and sanctioning system has to be understood together wi th the other initial principles, because the level of commitment of participants is not related to the possibility that someone is monitoring them or that they might be punished. The level of commitment was also related to the fact that people were able to choo se their own rules that were appropriate to their
43 reality. Success was also enhanced when it was cl ear who could participate, and when there were incentives for people to do so (Ostrom 1990). Principle 6 refers to the importance of rapi d, low-cost, local spaces to solve conflicts. There is always a possibility that someone will un derstand a rule in a different way from others, or break a rule. However, in long term regimes pa rticipants need a place to debate their problems and find solutions. Principle 7 refers to the freedom of partic ipants to devise their own rules, and the importance of governmental recogn ition of those rules. Local user s ability to develop more effective regimes over time is influenced by th e national or local government recognition that those people can, themselves, deci de their own rules. Official recognition of lo cal rules gives more power to the locals in the monitoring process. Principle 8 refers to the importance of having an organization based on different scales, for more complex and larger systems. Nested enterp rises can help to overcome the weakness of one scale or another. According to Ostrom (2005), in a polycentric system the users of a commonpool resource have some authority to make some rules; however, they can count on other authority levels to cope with problems such as local tyrannies and inappropriate discrimination. Polycentric system are ones in which people ar e able to organize not just one, but multiple governing authorities at different scales (Ostrom 2005, 283). Although the design principles can predict th e level of success of a common property regime, there are also other factors that can th reaten its success. A ccording to Ostrom (2005, 272) these threats are: 1. rapid exogenous change; 2. transmission failures from one generation to th e next of the operational principle on which community governance is based;
44 3. programs relying on blueprint thinking and easy access to external funds; 4. corruption and other forms of opportunistic behavior; 5. lack of large-scale institutional arrangements related to reliable information collection, aggregation, and dissemination; 6. fair and low cost conflict -resolution mechanisms; and 7. educational and extension facilities; and facili ties for helping when natural disasters or other major problems occu r at the local level. However, according to the author there ar e ways to address these threats. The establishment of community-gov ernance entities to provide continuous support to the communities, besides research on self-governing systems, and the development of better curricula on local governance, are actions that ma y cope with threats to sustainable systems of governing common resources (Ostrom 2005). Building Institutional Arrangements for Fi r e Management: Tapajs National Forest In this section I analyze the case study of a fi re management project within a protected area in the Brazilian Amazon carried out by a region al NGO, with direct participation of local communities, local organizations, and the Brazili an Environmental Agency (IBAMA Instituto Brasileiro de Meio Ambiente e dos Recursos Naturais Renovveis). At Flona-Tapajs, IPAM stimulated and supported the develo pment of institutional arrangements as an attempt to increase the number of techniques used by the families during the burning process of their fields, and consequently decrease the number of accidental fires. The arrangements were composed of a list of techniques and recommendations that each fam ily within the communities should apply when burning. The process was developed in an adaptive way in which families established rules, tried to apply them, evaluated the results, changed the rules, tried to apply them and evaluated them again.
45 The general research question of this section is: do institutional arrangements contribute to fire management? Specifica lly, the questions are: 1. Are institutional arrangements effective in decreasing accidental fires? 2. What are the basic principles for building in stitutional arrangements for fire management? History of the Project In the beginning of the 1990s a rural community in the northwest part of Par state called Del Rei asked its Rural L abor Union to help them solve the problems of ac cidental fires. During the drought of 1991 and 1992, many families suffered damages caused by accidental fires. From that time to 1996, the community in partnership w ith IPAM (Amazon Institute of Environmental Research) a regional NGO based in Belm, Par carried out research and organized many meetings to discuss issues related to accidental fires and how the families could decrease their losses caused by them. These meetings were seen as an opportunity for knowledge sharing about the use of techniques that could he lp keep fire within the borders of their agricultural fields. Even though several people attended the meetings, the community continued to have problems in the following years. It was in 1996 that they decided, with the help of IPAM, to create a local law with clear rules about how to use fire within the community. The experience within Del Rei community indicated that instituti onal arrangements could contribute to fire management in the Amazon. However, more research was necessary re garding this subject, and IPAM had interest in carrying out this research. In 2000, IBAMA and a group of local volunteer environmental agents working with in national forest located in another region of the same state, carried out a meeting to evaluate the environmental problems of the pr otected area. During this meeting, accidental fires were pointed out as an important problem, and because of that IPAM was invited to develop a project within the communities to establish local agreements fo r fire management to decrease accidental fires.
46 The Tapajs National Forest The Flona-T apajs is located in the north of Brazil, northwest region of Par State (Figure 3-1) in the municipalities of Belterra, Aveiro, Rurpolis and Placas. Created by the federal government in 1974 through decree No 73.684, the Fl ona-Tapajs is a protected area for sustainable use (SNUC 2000). This means that such areas should integrate conservation of natural resources with the sustai nable use of parcel of its natu ral resources. At the time of creation, there were many conflicts between the government and the families living there because there were some families who were not originally from the region (IBAMA/ProManejo 2005), and because at that time people were not allowed to live within National Forests (FLONAS). The Tapajs National Forest (Flona-Tapajs ) is a 600,000 hectare protected area nowadays under management responsibility of the Instituto Chico Mendes de Conservao da Biodiversidade (ICMBio). When the fire ma nagement project started, IBAMA (Instituto Brasileiro de Meio Ambiente e dos Recursos Na turais Renovveis) was the agency responsible for the management of the area. The area has more than 7,000 families living in an urban area and 30 different rural communities. The rural communities have more than 2,000 families that have their livelihood based mainly on agricultura l production. The use of fire is one of the traditional agricultural tools for all these families. The Flona-Tapajs has families of traditional people, indigenous and colonos. These families have limited land use rights that include the use of forest resources to produce crops, extract timber and non-timber products, although they do not officially own the land. The use of fire is one of the traditional agricultural tools for all these families.
47 Figure 3-1. Location of the communities and municipalities of Flona-Tapajs, communities participating in the fire management project since 2001, and the communities interviewed in 2007.
48 The Flona-Tapajs communities in 2001 were exte rnally represented by local leaders, in some cases chosen through a formal election process, but most of the communities did not have a formal association. Besides the local leader s of each community they had four intercommunitarian associations that were formal or ganizations, each one representing part of the communities. The Rural Labor Union of Belterra (STR-Bel) was another grass-root organization that could represent most of the communities, but also not all of them because the Flona-Tapajs has part of its territory in other municipa lities such as Aveiro and Placas. In 2001, the communities of Flona-Tapajs were involved in many different projects with different goals. There were projects to support the inter-community a ssociations. There were projects supporting alternative systems of production such as the us e of rubber to produce ha ndcrafts, the production of oils, the establishment of agroforestry system s, the raising of small animals, and so forth. Most of them had financial support from IB AMA/ProManejo. Currently, 12 communities are involved in a timber project coordinated by the Cooperativa Mista Flona Tapajs Verde (COOMFLONA). The COOMFLONA was created to manage Amb Project, a demand of the local people of Flona-Tapajs, after the federal government, through IBAMA, started a process of timber exploitation in partnership with an international timber company. The families questioned the government, on the grounds that they should be the ones to exploit the timber and get the benefits. It was a long process of negotia tion until the families started the Amb Project (COOMFLONA 2009). Regarding fire management, there were other ac tions carried out within Flona-Tapajs that had the goal of decreasing problems with fire, directly and indirectl y. In 2002 the PREVFOGO, an IBAMA program, started a partnershi p with the communities to form the Brigada de Fogo. The Brigada is a temporary group of local people trained by IBAMA/PREVFOGO to control
49 escaped fires within the protected area. The comm unities, through an election process, select the local people who make up the Brigada. After that they are tr ained by IBAMA/PREVFOGO and the Fire Fighters, and have to take an exam that selects the ones who will compose the Brigada for a period of six months of th e burning season. From 2002 to 2008, the Brigada had to fight two accidental fires within Flona-Tapajs. Howeve r, other accidental fires were registered by IBAMA in the years of 2004, 2006 and 2007. Anot her important acti on of IBAMA that contributed to the process of fi re management within the protected area was an initiative to control deforestation. In 2002, IBAMA invited the leaders of all the comm unities to establish a process to control deforestation. During those meetings, it was decided that each family could cut down the vegetation, burn and produce in a maximu m two hectares per year. The decision of producing in forest or secondary forest was to be negotiated with IBAMA by each community. Those areas would also be monitored by IB AMA year to year through a random process involving all the communities. The Fire Management Project at Flona-Tapajs The fire m anagement project was carried ou t at Flona-Tapajs. It started in 2001 and finished in 2004. The project was carried out by IPAM in partnership with the local people from within the communities, the Rural Labor Uni on of Belterra and IBAMA/ProManejo. The main goal of the project was to decrease accidental fires within the pr otected area. To IPAM, the main goal of the project was test the e fficiency of institutional arrangem ents for fire management, as a way to decrease the risks of accidental fires. Before the project started, the idea of formul ating local institutional arrangements within each community of Flona-Tapajs was presented and evaluated by the leaders and other local organization representatives in two different mee tings of three days each. In those meetings the leaders of all the communities helped IPAM to define activities, a cal endar and indicated the
50 communities with more problems of accidental fi res, where the project should start working. After that, the project was submitted to IBAM A/ProManejo, and once approved the activities started within the communities. The process within communities to formulate the fire agreements followed many steps (Figure 3-2). First, each co mmunity was invited to discuss their fire use practices and later to formulate fire agreements. Each community had the opportunity to debate internally and then to respond to IPAM if they wanted to particip ate in the projec t or not. Sixteen communities were pointed out by the leaders as the ones where the project should be carried out. Those were invited to participate in 2001, but four declined the invitation, including some with a history of accidental fires. Th e communities that declined the i nvitation, during the meetings said that they were not interested in regulations about fire. In their vi ew, fire is a practice that they know how to use, and they did not want to ch ange anything regarding that point. So, IPAM started working only within the communities that accepted the invitation. In the following years, other communities invited IPAM to work with them: In 2002, four communities started participating, and in 2003 another two communities. The second activity of the project was a fi re use assessment with in each participant communities. The assessment was carried out through interviews with families of the communities. The interviews were carried out at the household level, and all the questions were related to the system of producti on, fire practices, cases of accide ntal fires, and losses caused by accidental fires. The information was organized by IPAM and was used in many ways: first to create a base line to see if there was change in the families practices during future years. Second, in each meeting within the communities, th e information was used to debate past cases of accidental fires, and good practices identifi ed by the families. Third, annual information was
51 organized in a way that enabled participants to evaluate the effectiveness of the management system, and to make changes accordingly. Figure 3-2. Fire management main activiti es carried out within the communities. The third activity was the organization of one to three-day meetings within each community, in which families and IPAM discussed practices of fire use; the benefits of fire management for communities; the problems associat ed with fire use for the local, regional, national and international environm ent; and ways to avoid accident al fires. Alternative systems of production were also discussed, as well as the responsibilities of Amazonian farmers for responsible fire use. The communities then had meetings to discuss specific practices and methods used to prepare fields to burn. IPAM suggested techniques they could start using or changes that could more effectively control accide ntal fires. In these meetings the communities formulated temporary agreements that they woul d try to implement in the next burning season. The meetings were organized in two different mome nts. The meetings to debate practices and to formulate the agreements were carried out just be fore the families began to prepare their fields for the year, generally around June and July. Meetings to debate fire use Specific meeting to design fire agreements Fire season Collective Evaluation of fire season Assessment of fire use Meetings with local leaders Invitation to families
52 After the burning seasons in which families were to try to apply the rules of the temporary agreements, there was one meeting in which the families evaluated the burning season, the rules they tried to use, and whether they needed to ch ange the agreements to ma ke them more feasible. These evaluations were carried out in two ways. First IPAM techni cians would visit a sample of fields to interview the owners of the fields re garding the use of the techniques. Then a meeting with all the farmers was carried out, and each farm er had the opportunity to talk about his or her attempt to use the rules, the difficulties they encountered, and their opini on about each rule, and the changes the thought were necessary. Based on information produced by the families, and also by IPAM, all the communities reformulated the agreements every year. The idea was to develop realistic rules that they could follow, and at the same time, that coul d result in control of accidental fires. By the end of the project the communities established definitive fire agreements. The meetings to evaluate the use of the rules esta blished in agreements we re carried out after the burning season, during the months of January and February. An average of four meetings was carried out within each community each year. These meetings were open to all the families within each community. In general there were more men than women participating, and there were also participation by teenagers. Besides the process within the communities the project develope d three other main activities. The first was the training of two pe ople chosen by the families from within each community, who were responsible for keeping up the debate about fire management in other opportunities besides the activit ies organized with the pres ence of IPAM. Throughout the project, these individuals, called fire management facilitators were engaged in planning, giving inputs to the project and helping to carry out the activities with in their communities. This group was voluntarily involved in the pr oject. Most of the volunteers we re young people, not traditional
53 leaders of the communities. The training proces s therefore focused not only on the debate over fire and alternative systems of production for fire management, but the group also demanded training in techniques of how to present ideas, how to speak in public, and how to organize and write proposals. The facilitators faced some challenges. The most important was that some families attributed to them responsibilities that they had not in itially accepted, such as being present and assisting at every burning event. The agreement in the meeting at which they were selected established that the facilitators were there to talk with people, not to work for them. The second activity was a pro cess involving teachers and st udents of some schools at Flona-Tapajs in an attempt to introduce fire us e in the curriculum of the schools. This process was difficult because any change in the schools had to enlist the interest of the local government. In the case of Belterra, the local government was not interested in fire control. The third activity was the creation of a group of lo cal leaders and organizations (F orum for Fire Management of Flona-Tapajs) to discuss fire use, to ensure the continuity of the pr oject, and other possible projects for Flona-Tapajs that could help to decrease the use of fire. Although the Forum started with the participation of many local organizations, by the end of the project, only two local grass-roots organizations remained interested in collaborating with IPAM and the communities. When IPAM began to work within Flona-Tapajs, the idea was that, at the end, the communities would formulate and abide by definitive agreemen ts. However, it did not take long for them to say that continued support was necessary to keep the discussion alive, and also to promote the introduction of alternative systems of production th at could, in the long term diminish the use of fire. Those ideas were discussed within each comm unity and their local lead ers, and also within the Forum. Within the communities, the leaders gua ranteed that they would address the issue in their meetings with the support of the facilitators, and would remember and discuss the rules they
54 established. Also by the end IPAM helped two lo cal grass-roots organizations (Sindicato dos Trabalhadores Rurais de Belterra and Federao das Comunidades Tradicionais da Floresta Nacional do Tapajs) wrote proposals for funds to keep the work going within the communities already involved in the process. Funds were also requested to in corporate other communities that asked for the project in the following years, and to work within the communities living along the borders of the protected area. During the process, some people living within Flona-Tapajs had di fferent concerns and reactions to the project, especia lly in the first year (Table 3-2). During the first meetings the participants were afraid that at the end of the process they were going to be forbidden to use fire in agriculture. In fact this was never the goal for IPAM, for the leaders involved in the planning of the project, or for IBAMA, since everybody was aware that the fam ilies had no alternative technologies. IPAM repeatedly affirmed that the prohibition of fire use was not the goal of the project. Some people would ask why they were de bating an old practice that they learned how to use from their ancestors. The responses of IP AM were that the idea was not going of teaching anybody how to burn, but instead help ing with the improvement of the practices where they were not working, to avoid more accidents in the fu ture. Some families often answered this point by reminding each other of the accidenta l fires they had before, and the losses they had in the past. During the meetings, in which the debate was ab out their local practices and the impacts of fire use at different scales, the communities in general were very open an d curious to participate in the debate about climate change and the connec tions between what happens in their fields and people all around the globe. In the specific meetings to formulate fire agreements, some families expressed that fire could not be controlled, that only agricultur al mechanization could prevent accidental fires. IPAM tried to highlight the ex perience of those families who never started
55 Table 3-2. Concerns expressed by some families during the fire management project and the steps or arguments used to overcome those concerns. Concerns Arguments from IPAM, and other families to overcome the concerns 1 Concern about a possible prohibition of fire use Fire would not be prohib ited in the end of the process but only controlled 2 Fire was an old known practice. There was nothing new to learn The process was not about teaching new methods, but of how to improve the use of known techniques, and to guarantee that everybody used the known techniques. 3 Curiosity about global relationship of environmental issues More exchange of information. 4 Need for better technology (mechanization) Valorize the experiences and techniques used by families that did not have accidental fires. Discussion of agricultural mechanization benefits and cost, alternative systems of production, sources of financial support, organization necessary for them to access those resources. accidental fires within the communities, and trie d to emphasize the techniques they used, and explore the possibility of other families follo wing those examples. Families that had never caused accidental fires within the communities ofte n responded to this point saying that if people would use the techniques they knew, they were not going to have problems in the future. Because in many situations, the topic of mechani zation as the best solution to accidental fires would be brought to the meetings, IPAM did exercises to analyze economic and social aspects of mechanization, its benefits and costs, and the deba te of how they could us e it. Mechanization of agriculture was pointed out as the best solution for fire, because at the time the project started, a strong process of mechanization for the pr oduction of soybeans was going on outside the protected area through the movement of agribusin ess to the region. Beside s this, the discussion of alternative systems of production, and sources of financial support that the families could get to start those systems was always part of the deba te in the meetings. In later years, the debate
56 was more related with their experiences in tryi ng to use the techniques they had selected, and changes that needed to be made to their agreements. The initiative at Flona-Tapajs was ba sed on dialogue between families of each community and researchers, join tly trying to identify good technique s to avoid accidental fires. During those years, each community involved in the process formulated agreements in which they tried to regulate the use of agricultural fi res to avoid accidents. The agreements were composed of rules to be followed by each farm du ring the preparation of the productive field. All of the rules were techniques and procedures to control fire and to avoid physical damage to the farmers. Changes in Fire Management Practices at Flona-Tapajs The project carried out by IPAM finished in 2004. The project that was written for the Federation to continue to work within the co mmunities of Flona-Tapajs was not funded. The Federation was not a form al orga nization at the time and this made it impossible for them to apply for other additional support. The project presented by STR-Bel was approved, though, and they are carrying out activities to promote fire management and alternative systems of production within some communities on the border of Flona-Tapajs. In 2007, I carried out research wi thin the protected area to an alyze the level of compliance with the rules established in the agreemen ts, and how the communities were using the agreements. To do this analysis, I used two sources of data: (a) an assessment carried out in 2001 within 11 communities of Flona-Tapajs, where 71 families were interviewed at the beginning of the project; and (b) surveys carried out in 2007 in 4 communities, where 53 families where interviewed. To select the four communities I built an index of participation that considered the number of people participating from 2001 to 2004 in the meetings were the agreements were formulated within the communities; in chapte r 3 the construction of this index is better
57 explained. The communities interviewed were Prainha II and Paraso with the highest level of participation and Chib and Piquiatuba with the lo west level of participation (Table 3-1). The surveys were conducted in July of 2007, and in most cases the couples were interviewed together. The results from the early assessment indicated that, before the project started, the use of techniques to avoid accidental fires by families living in the protected area was minimal (Table 3-3). According to Assuno (2001), most of the families used on average four techniques. Carvalho et al. (2007) listed 18 techniques or recommendations that can be used to avoid accidental fires. The list is composed by tec hniques and recommendation that farmers need to apply months before the burning day, such as the definition of the place to start a new field, or the plan to work in partnership with neighbor s. There are also techniques and recommendations specific to the day of burning such as time, number of people or starting th e fire against the wind. The list is a result of action research carried out in more than 20 rural communities of Par. The knowledge of the rural families was tested to an alyze the effectiveness of each technique in helping to control accidental fires. In 2001, at Flona-Tapajs, less than 20 percent of the families used to build a firebreak around the field before burning; almost fifty percent of the farmers burned their fields at noon (the hottest hour); and less than 20 percent of them remained to watch the fire until the burning process finished (Assuno 2001). The result of th ese practices was a history of accidental fires throughout the years, but especially in 1997/1998. The 11 communities involved in the assessment carried out in 2001 had accidental fire in their areas, resultin g in losses of forest, pasture and productive systems (Assuno 2001). During the years of th e project, all the
58 communities involved described accid ental fires in their areas, although they were not precise about the years they occurred. Table 3-3. Percentage of the families using techniques and recommendations for the years of 2001 and 2007. Techniques or recommendations Years 2001 n=71 2007 n=53** 1 Distance from rivers 100.0 2 Bring tools to the field 100.0 3 Wear adequate clothes 100.0 4 Authorization from IBAMA 98.1 5 Watch the field until the fire finishes*** 13.0 96.2 6 Month of burn 95.3 7 Burn some days after raining 38 92.3 8 Distance from roads 88.4 9 Plan the burning day 87.1 10 Prepare source of ignition before 83.3 11 Time to burn*** 6.0 83.0 12 Cut the trees towards the field 70.8 13 Start the fire against the wind*** 31.0 69.2 14 Lower the fuel load 66.0 15 Inform neighbors of the day to burn 63.6 16 Clean the fire break before burning 62.5 17 Water*** 20.0 56.6 18 Build fire break around the field*** 18.0 52.8 19 Cut dead trees 8.0 52.3 20 Number of people*** 19.0 47.2 21 Talk to neighbors to work together 14.2 Information not available in the database of 2001. ** Percentage calculated based on valid responses, because there are rules that do not apply for all the communities and it was not asked in the survey. *** Techniques or recommendations indicated by the families as the important ones. Techniques or recommendations indicated by the technicians as the important ones. The results from 2007 indicated that the numbe r of techniques used by each family, and also the number of families using each technique, had increased. Most of the families living at Flona-Tapajs in 2007 used at l east 11 techniques. For the techni cians involved in the project, the five most important techniques are: build a firebreak around the field, time to burn (avoiding
59 the hottest time of the day), number of people participating during the fire, starting the fire against the wind, and watching the field until the fire finishes. For the families interviewed in 2007, the most important techniques were the same ones pointed out by the technicians, plus bringing water to the field. Build ing a firebreak increased from around 20 percent of use to 52 percent. Preferred time to burn increased from 6 percent of the families, to 83 percent. In 2001 31 percent of the families used to start the fire against the wind, and in 2007 this recommendation was used by 69 percent of the families. Watching the field until it finished burning increased from less than 20 percent to mo re than 95% (table 3-3). The five techniques pointed out by the technicians were the same ones pointed out by the families interviewed as the most important to avoid accidental fires. But th e families also considered that it was very important to have water available to help with small accidental fires. In 2001, 20 percent of the families said they used to bring water to th e field during the burning day. In 2007, 56.6 percent of the families asserted they were following this recommendation. The process of requiring official authorization to burn star ted during the years of the project. While IPAM talked with the communitie s about the importance of keeping their fields legal, IBAMA developed a system in which they would visit each community to give families the authorization for cutting and burning their fields every year. This explains the high level of compliance with this rule (98 percent). Family labor force is considered an importa nt element that guides the decisions made by small farmers in Amazon regarding their land uses strategies (Perz and Walker 2002; Perz 2004). During the elaboration of the fire agreements, technicians and families were concerned not to introduce solutions to accidental fires that would increase labor demands, and consequently the costs of production for the farmers. Considering th at the requirement of labor can be one factor
60 influencing the decision of families to use a specif ic technique or not, I classified the techniques used by the families in 2007 in 4 categories (Figure 3-3). There are techniques that: a) depend on family labor; b) depend on family labor and technology (chainsaw); c) depend on family and neighbors interest and labor; d) no family labor required. An average of the compliance with the techniques and recommendations using these categories indicated that families followed more the techniques and recommendations that do not require hard labor (81.2 percent), and applied less the ones that required partnership with neighbors (49.5 percent). Of the six most important techniques pointed out by the families (watch the field until the fire finishes, start the fire against the wind, water, build fire break around the field, time to burn, number of peopl e), three that depend more on planning than on hard family labor (watch the field until the fire finishes, start the fire against the wind, and time to burn) were used by more than sixty percent of the families. Figure 3-3. Average of families compliance with techniques and recommendations based on a labor classification for the year of 2007.
61 In 2007, almost 95 percent of the families knew that the community had fire agreement. Those that did not know were new couples, families that were not living in the area and old people that did not participate in the process during the years of the project. Most of the people thought that the agreements still needed changes. More than 50 percent said that they kept discussing fire use and the techni ques; however, almost 50 percent said that they only do it when IBAMA comes to the community to give the authorizations to burn. The analyzes of the techniques applied in 2001, before the project star ted, and in 2007, two years after the project ended, show that not only did people use more techniques, but also the number of people using each of them increased for all the tec hniques. Although the changes were considerable, 83 percent of the families interviewe d recognized that there was still a risk of accidental fires. They also recognized that the risks were lower in 2007 than in the past (97 percent). During the interviews, some families asserted that in 2005 they had a severe drought in the region and that no big accidental fires occurred within the protected area because families are more worried and more careful now, than they were in the past. In 2007, fires escaped in the field of 5.6 percent of the families interviewed, but none of them spread to huge areas. Institutional Arrangements for Fire Management Avoiding the Tragedy of Commons Fire is an ag ricultural tool us ed today by almost all small, medium and big farmers in the Amazon. The practice and problems of using fire as an agricultural tool have three aspects of the commons: 1. Accidental fires can cause damages to every farm in a community and sometimes in neighboring communities. 2. The decision of decreasing the risks of accidental fires needs to be taken by the community, and not just by one farmer. Even if one individual manages fire in a way to avoid accidents, his property or productive system still has the risk of being burned by an accidental fire provoked by his neighbor, or by someone living in another community. This means that the more people involved in the process of controlling accidental fires, the greater the chances that the effort will be successful; and
62 3. Some techniques of controlled burns require or are easier to put in practice through collective work. Hardin (1968), presented situ ations in which the conseque nces of one practice could provoke problems for many other individuals, and t hose situations could also be identified as tragedies. He cites pollution as one example. In the case of agricultural fire, it does not matter if the system is private property or communal; accid ental fire is an element that will not respect boundaries, and can provoke indisc riminate damages in all the properties of a community. The solution to this problem also re quires a certain level of collective organization. During a meeting involving many farmers in the Del Rei communit y, many years ago, after they tried different strategies to avoid accidental fire without too much success, one farmer said that to control fire effectively they should create a local law. He unders tood that together they needed to create local rules, and follow them, in orde r to control agricultural fire s. The community of Del Rei established the first fire agreement of the regi on. This event showed an understanding that the strategy of fire control necessar ily had to have components of co llective action. Just one farmer (A) isolated in his property usi ng the correct techniques to control fire in his field is not enough. His neighbor (B) can provoke a fire accident and this fire can burn forests and systems of production of farmer A. To have an efficient system, they need to work together, and use as many techniques as they can, as well as comm unicating effectively, even forging collective management strategies. The experience developed at Flona-Tap ajs was thought out based upon those considerations, and instit utional arrangements were develope d in 18 communities, as an attempt to increase the number of techniques used by th e families during the burning process of their fields, and consequently decrease the number of accidental fires. The arrangements, in most of the cases were composed of a list of techniques and recommendations that each family within the
63 communities should apply when burning. The proce ss was developed in an adaptive way in which families established rules, tried to apply them, evaluated the results, changed the rules, tried to apply them and evaluate them again. The analysis carried out in this research indicated that this process contributed to an increase in the number of techniques used by the families, and also in the percentage of the families using each technique. A later analysis of that e xperience drawing upon the principl es designed by Ostrom (1990) (table 3-4) shows that fire management agreem ents at Flona-Tapajs had clear boundaries, the rules were thought out considering cost/benefits analyses, the process was adaptive and families could modify the agreements anytime, and external agencies recognized the rights of the communities to take their decisions. However dur ing the years of the project, communities did not establish a monitoring and sanction system, mo st of the communities did not set up arenas to solve conflicts, and the intera ction among other organizations to support fire agreements was only indirect. Regarding the definition of boundaries, it wa s easy in the case of the Flona-Tapajs agreements due to two factors. First, each community already had their own boundaries, and clear definitions were already established regarding of which families were part of the community and which ones were not. Also, the use of fire within each community was only for agricultural purposes, and the right of farmers to burn for ag ricultural purposes was guaranteed in law. During the meetings the families just emph asized that the rules created were applied to everybody who would burn a field. Regarding the congruence between benefits a nd costs of following the agreements, there was a concern not just among the technicians, bu t also among the families not to propose rules that would increase the costs of production for the families. This concern was based on the fact
64 that the use of techniques and recommendations for fire management does not increase the profits of a farmer. It can decrease damages, but accidental fires are a possibility. Burning fields does not always result in accidental fire. And not all accidental fires mean damages in the view of farmers. It is therefore complicated to incr ease labor when there is no consequent increase in benefits. The results of this study show that the families were more likely to follow the rules that required less labor. Table 3-4. Analysis of Flona-Tapajs experience using the principles designed by Ostrom (1990) Design principles illustrated by longenduring commonpool resources institutions Analyses for Flona-Tapajs Clearly defined boundaries All the agreements were clear regard ing the rights to use fire, and for what purpose Congruence between appropriation and provision rules and local conditions The rules set up were based on the farmers practices, and there was an effort to think of rules that were not going to increase the labor required, because the use of techniques for fi re management does not increase the profits of a farmer. It can decrease damages, but accidental fires are a possibility. Not all the time a farmer burns his field is there is an accidental fire. And not all accidental fires mean damages in the view of farmers. Collective-choice arrangements The agreements can be modified any time by the families, however, although they have suggestions for ch anging it, they are not bringing the ideas to the communities meetings Monitoring There is no formal system of monitoring who is doing what within any of the communities. Graduated sanctions From the 8 communities i nvolved in the project from the beginning to the end, four have defined that people would be responsible for damages caused, but no sanction have been applied Conflict-resolution mechanisms Three communities decided that comm unity meetings were their space to solve conflicts, In one community they defined if they could not solve it, IBAMA was going to be invited to intervene. Minimal recognition of rights to organize Communities were free to decide thei r own rules, and even now IBAMA is not interfering with fire agreements. Nested enterprises (For CRPs that are part of larger systems) There is not any other organiza tion responsible for helping the communities, despite attempts. ICMbio is currently the organization that can support communities regarding fire management.
65 Regarding the possibility of the participants to change the rules, during the years of the project there were special meetings designed just to change the rules, and the communities did change their agreements many times. Nowadays, the agreements can be modified any time by the families; however, although they have some suggestions for change they do not always bring them to the communities meetings. While the agreements were being developed, many times the communities were asked to think of a system of monitoring, and sanctions for those who didnt follow the rules. Although some communities did create procedures, most of them did not want to do it. When I went back to the field in 2007, the son of one family caused an accidental fire that burned the productive system of his parent s. The parents did not want to discuss the case even with the leaders of the community, as was established in their agreements; the main reason was because the person that had caused the accident al fire was their own son, and they were the only family that had damages in the case. The str ong kinship level within the communities in that region may be one explanation for why families did not want to establish a system of punishment and sanction. However ICMBio has a strong pr esence within the comm unities and they can always report problems to that governmental organization, although there is no evidence of families doing that. Also, IBAMA monitors fields w ithin each community every year as part of a process established in agreement with families of Flona-Tapajs to control deforestation. This monitoring activity carried out by IBAMA is the only action of its kind, and maybe it plays the role of monitoring the system. The families interv iewed in 2007 said that they only know what their neighbors are doing because they sometimes work together or they go hunting, and they go through neighbors fields and can see what they are doing.
66 Regarding the solution of conflicts, some of the communities decided that the community meetings were their space to resolve conflicts, bu t there were no cases after the agreements were established in which the community had to interfere in a conflict situation. While the project was going on, IBAMA left the communities free to participate and decide their own rules, and even now they are not interfering w ith fire agreements. However, they helped the communities many times by prov iding information about each technique and the importance of using them, leaving the decision about including the tech niques or not in the agreement to the farmers participating in the meetings. Because the project had an e nding date, there were many attempts to build a network of organizations that could keep th e debate about fire management alive better described in the chapter four. However, it was not possible due to three factors: a) most of the organizations were not able to work within all the communities of FlonaTapajs; b) some organizations were not interested in working with fire management; and c) the only organization that was interested in k eeping the project was not able to apply for financial support because it was a new orga nization that had no formal status. Every year just before the burning season, IBAMA goes to the communities to authorize the deforestation and burning of the fields. The fa milies indicated this visit as the moment when the fire agreements were remembered. This coul d be a great opportunity to review the rules and re-enforce the process, but this is not occurring. The lack of continuous discussion of the fire agreements is one of the biggest threats to their continuing to function. In 2006 there was one case of accidental fire that was provoked by a young couple that did not participat e in the project activities. Al so, there were a few families who moved to the community afte r 2004 who do not know that there was a fire agreement in the community. If the community or the IBAMA/ICMBio or the local organizations do not continue
67 to promote the discussion of fire management there is a chance that those communities will decrease the number of techni ques and recommendations being us ed, and consequently increase the of accidental fire in the future. Conclusions The case of Flona-Tapajs shows that institu tional arrangem ents can contribute to coping with the problem of accidental fires, even when th e process is stimulated by an external agent. At Flona-Tapajs four years of open discussion ab out techniques and recommendations available to avoid accidental fires, and a process of prom oting dialogue among families where farmers could express their interests and opinions and by themselves decide which techniques and recommendations to use to control accidental fire s, evaluate them and change them according to their opinions and debates, resulted in a increase in the number of techniques used, and also in the number of people using them. Thus, the risk of accidental fires decreased, as pointed out by all the families. What are the basic principles to build instit utional arrangements for fire management? The experience of Tapajs National Forest indicates th at the effectiveness of institutions for fire management depends on some important principl es such as the ones pointed out: congruence between costs and benefits to put the rules in practice, or clear definition of boundaries (Ostrom 1990). However, based in the case of Flona-Tapajs it is possible to say the effectiveness of the initiative also relied on the facts that: 1. The initiative took place at the very local le vel, with strong participation and engagement of the farmers using fire; 2. To build institutional arrangements, the tech niques and recommendations to be used were chosen by the farmers who would use them. Is does not matter what ex ternal agents think, because in the end, the farmers decisions will be implemented during the burning of their fields;
68 3. The decision power in the hands of the familie s was to be clearly stated. All communities and families need to knew that they could decide to participate or not, to stop the process anytime they wanted, and the decisions regardin g the techniques were truly in the hand of the families. In the case of Flona-Tapajs some communities were never involved in the project and there was no penalty for that. Ma ybe because of it, during the following years the number of participating communities increased. 4. The process was adaptive so that all those i nvolved could learn from successful activities and also from mistakes, improving the understa nding of the process, and changing course of actions when necessary (Mulder and Coppo lillo 2005). In the case of fire agreements, evaluations were carried out year from year to analyze the attempts to use the techniques and recommendations, and to decide what was necessary to change. 5. The process was focused and continuous, with different kind of activities, allowing longterm debates on fire. Otherwise, it is not possi ble to guarantee any ch ange in the behavior of people through a lecture on good techniques a nd recommendations of how to burn fields for agriculture. When the project for fire management starte d within Flona-Tapajs technicians thought that after some years of adaptive changes, the families could define a permanent set of rules that could be applied for long time. However, it did not take long for us to understand that institutional arrangements can always change, and th at the families need to control the process of organizing and changing it from year to year if it is necessary. A change in the weather can make a difference in the set of rules necessary to control accidental fires, and the communities internally need to be able to work out the necessary changes. In the case of Flona-Tapajs, families keep taking the decision at the individua l level, and although the number of techniques and the number of families using them increased, for fire management, it would be better if the decisions for using the techniques or recommenda tions were debated among all the families year to year, as a way to stimulate constant concern and also to avoid the failure of transmitting this concern to new generations. The case of Flona-Tapajs also indicates that through the process of building institutional arrangements for fire management, farmers star t thinking and debating ab out alternative systems of production to fire, and in this way, the proc ess of discussing fire management creates an
69 environment for the introduction of more important changes in their agricultural systems, that could ultimately diminish or even excl ude fire from the system of production
70 CHAPTER 4 1 PARTNERSHIP IN AN ACTION RESEARCH PROJEC T: BUILDING AND ANALYZING PARTICIPATION Introduction Participatory approaches started to gain acc eptance in developm ent projects and also within the environmental field due to the failu re of many top-down initi atives in achieving the goals established, and when externally imposed m odels often proved to be ineffective (Chambers 1992; Guijt and Shah 1998). After decades of de bate, the importance of having local people participating actively in decision-making pro cesses is now a consensus among scholars and practitioners working in devel opment initiatives and environment. Agencies and foundations started requiring strategies and ac tivities that would guarantee lo cal participation in proposals supported by them, as a way to empower local pe ople to analyze their own situation and take action and improve their livelihoods (Guijt and Sh ah 1998). However, there is still the challenge of putting into practice the cons ensus about importance of part icipatory process. The term participation has often been used to describe very rudimentary levels of participation where agencies staff just consults community me mbers (Guijt and Shah 1998). The analysis of participatory processes have not yet resulted in a theory about local participation, but ranges and frameworks to analyze the subject have been proposed to identify if a process is locally participatory or not (Adams and Hulm e 2001; Cohen and Uphoff 1980; Stone 2003). The discussion about local participation is not specific to development and environmental projects. Participation can also be discussed in the context of research projects. According to Ferreyra (2006:577) in the last few decades, there has been an emerging debate in participatory action research (PAR) in the context of environmental governance involving multiple stakeholders.
71 In this study, I discuss the impor tance of local participation a nd partnerships in building an action research project, and ultimately, to the effectiveness of the institutional arrangements that were developed during the fire management proj ect at Tapajs National Forest. The project was developed through a partnership between an NGO, rural communities of a protected area and many other stakeholders. The project aimed to build institutional arrangements for fire management through a participatory process, as a way to decrease accidental fires within the protected area. I used my experience as part of the NGOs staff, to discuss the attempts to make it a participatory process during the years. This an alysis suggests that it is complex to define what is local in a process, in which more than one community and more than one organization is involved. The engagement of many different stakeholders is not easy. It takes time, human and financial resources, continuous deba te of interests and goals and c onstant adaption. In the case of Flona-Tapajs, the openness to participation by diffe rent stakeholders cont ributed not only to the execution of the action research project, but more importantly, to the achievement of the main goal of the partnership itself: the reducti on of accidental fires at Flona-Tapajs. Secondly in this study, I examine how numb ers of people participating during the formulation of local agreements at Tapajs Nation al Forest are related with levels of compliance with the fire management agreements rule s the families follow nowadays. I decided to investigate this question because during the ye ars of the project, th e group of technicians carrying out the activities and the people coming to the meetings were always worried about the number of families present in the meetings. The concern was motivated by many reasons: a) part of the community was going to take d ecisions in the name of other community members that were not participating in the meetings; b) those who were not coming to the meetings might not follow the decisions made. Farmers used to bring up the subject of the importa nce to guaranteeing that the neighbors were going to make the same effort to avoid acc idental fires as thes e people attending the meetings. Otherwise, their own effort would not avail, because a single accidental fire
72 could damage all the farmers. This point wa s debated in all the meetings within the communities. c) Finally, the number of people involved in the activities we were developing was one evaluation criterion used by the funding agencies. These were the reasons for the concern about the number of families in each community participating in the meetings. However, during fo ur years working at Tapajs National Forest, I observed that the number of people that were pa rticipating in the meeting varied by community. Some communities had higher levels of partic ipation with many people coming to each meeting, while others had less people par ticipating in the meetings. To understand how different levels of particip ation during the process of local agreements formulation at Flona-Tapajs were related with levels of complianc e with the rules, I carried out a survey in four communities of Flona-Tapajs. Two of those had the highest levels of adult participation in the meetings carried out there, while the other two had the lowest levels of participation. In July of 2007 I carried out 53 interviews with families of the four communities. I calculated frequencies of the degree to which peopl e used the agreements rules to compare the communities. My findings indicate that at FlonaT apajs, there is no significant difference in the average number of rules applied by the families of communities with higher level of attendance in comparison with the communities that had lower levels of attendance at the meetings to formulate fire agreements. Because there is no important distinc tion between the two communities, the findings suggest that in co mmunities in which families have long-term interactions and where the level of kinship is al so strong, the number of people participating in the meetings is not a determinant of the actions the families will take. This chapter is organized in four parts. In the first, I discuss the importance of local participation to manage natura l resources, and the importance of local participation in the specific case of building institut ional arrangements to avoid natu ral resources depletion. Then, I
73 discuss the experience of fire management in th e Tapajs National Forest within the context of participation. The third part brings an analys is of the relationship between number of people participating in meetings and compliance with rules they established during those meetings. Finally, I draw some conclusions regard ing the two goals of this research. Background Local governance and local participation in e nvironm ental issues are widely supported by many scholars as a way to guarantee sustainability and development. Despite the current interest in participatory methods, the involvement of lo cal people in decision-making processes is not new. In 1930 there were initiative s in India trying to stress part icipation as a way to empower and promote collective local acti ons (Guijt and Shah 1998:3). However, in the past, participatory approaches sometimes were actually discouraged in development projects and programs (Stone 2003). During the 1960s within the growing envir onmental movement, participatory approaches did not have the global space and importance that they get nowadays. As an example, at that period and before, many protected areas were created in many different places without any consideration of local in terests, and local people, and primar ily by decisions of outsiders (Mulder and Coppolillo 2005). Only during the 1970s, after the failure of many t op-down the initiatives and when ineffective of externally imposed mode ls became clear, did participatory approaches start to gain more serious consideration (Chambers 1992; Guijt and Shah 1998). Bottom up development projects, started to gain more accep tance when conservationists embraced the idea, as a way to avoid the large-scale and expens ive failures to top-dow n models (Mulder and Coppolillo 2005). In the early 1990s participation started to be used as synonym for good or sustainable development (Guijt and Shah 1998:4). Participati on requirements by agencies and foundations, were based on the assu mption that participatory appr oaches empower local people with the skills and confidence to analyze their situa tion, reach consensus, make decisions and
74 take action, so as to improve their circumstan ces, and promote more equitable and sustainable development (Guijt and Shah 1998:1). In that period participatory development emphasized not only the need for local input in projects but also the decentrali zation of and access to decision-making processes (Stone 2003:28). Nowadays, even in the governmental sphere the importance of participatory processes is gain ing space. There is a shift going on in state environment governance program to promote mo re participatory approach involving local stakeholders and the private sector. Thus, many actors, decision makers, and partnerships have come to play increasingly important roles in what happens to the en vironment (Agrawal and Lemos 2007:38). The current advocates of partic ipatory approaches emphasize not only voice but agency, the capacity and power for local people to define development according to their aspirations and to negotiate vis--vis other actor s, the conditions of their participation in all aspects of a development project (Stone 2003:32). Thus, the term co-participation appears to define a process of collaborative participation where local communities and other actors form partnerships to define and implement developmen t initiatives, as a more recent approach (Stone 2003). But what does local participation mean or what is a participatory approach? Participation has often been used to describe very rudimentar y levels of consultation between agency staff and community members and sometimes it has been used to manipulate people under a cloak of social palatability but it se ems that scholars are accepting that genuine participation should embody some form of empowermen t of the population, especially participation in decision making processes where people should be involved in the whole project or program cycle (Guijt and Shah 1998:9). Mere consultation, and limited involvement in activities, should not be considered to be a genuine participation pro cess (Schneider and Libercier 1995:10). For Cohen
75 and Uphoff (1980:214), participation is a single phenomenon that by general definition is the involvement of a significant number of persons in situations or actions wh ich enhance their wellbeing. To some, participation is an end in itself: a human right. In conservation and development initiatives, community supporters and activists ar gue that local people have the right to participate in any decision process that could influence their lives (Weigand Jr. 2003:1). Although it seems that nowadays there is ag reement among academics and practitioners that engagement of local people, groups and communities foster more effective environmental governance, there is also criticis m of participatory approaches a nd the way outsiders work with rural communities and groups rega rding to the level of particip ation local people have in the projects and programs in which they are involv ed. Adams and Hulme (2001) say that there are two polar forms of community conservation. On one side there is centralized or top-down protected area management that does not make efforts to include local opinions and views, and on the other side, there are initiatives that aim to promote a complete devolution of resource management to local authorities. Stone ( 2003) also proposes a range of participatory development initiatives that go from little participation, in which projects and programs are defined externally and local people are just providers of resources such as labor and information, to full participation in which projects are locally driven since their conceptualization, and the involvement of outsiders is minimal (Figure 4-1). Cohen and Uphoff (1980:214) designed a framew ork to analyze participation. The authors say that it is important to think of participation in three dimensions: what kind of participation is under consideration, how participation is occurri ng, and who participates at which level (Cohen and Uphoff 1980; White 1996). Besides these three dimensions, the context of a participatory
76 process needs close consideration too, because the situations in which participatory actions take place may vary widely (Cohen and Uphoff 1980). Figure 4-1. Range of partic ipatory development initiatives. Source: Stone 2003:36 Regarding the kind of participation possi ble, the authors list four types: a) participation in decision-making processes; b) participation in implementation; c) participation in benefits; and d) participation in evaluation. LITTLE PARTICIPATION FULL PARTICIPATION Externally defined and administrated project Significant involvement of outsiders (foresters, extensionists, etc.) Local peoples participation largely confined to providing resources (labor, information, etc.) Little to no participation of local people in decision-making processes Little capacity of local people to change or redefine project according to own needs, problems, etc. Dependent on outside ass i sta n ce, r esou r ces, etc. Project externally defined but with some input from local people Local peoples participation: providing Local people have some power to change or redefine project resources and involved in some decision-making and administrative processes Outside professionals provide assistance and act as facilitators Gradual independence from outside assistance, resources, etc. Co-participation Multiple actors (local and external) define project and share decision-making authority and responsibility Participation of each varies (according to skills, interest, and needs) but deemed equal or appropriate by all participants involved Interdependency among actors involved Locally-driven in all aspects of the project (conceptualization, planning, implementation administration, accountability, etc). Minimal involvement of outside actors (only upon request of local participants) Auto-dependent
77 Decision-making participation refers to th e generation of ideas, formulation and assessment of options, and making choices about th em, as well as the formulation of plans for putting selected options into effect (Cohen a nd Uphoff 1980:220). Three different moments are involved: initial decision s related to the identification of local needs; ongoi ng decisions taken during the execution of the project or program; and operational decisions that mean specific arrangements or partnerships established by th e project as an effort to involve people. Implementation participation refers to people s participation in the implementation aspects of a project. Cohen and Uphoff ( 1980) conceptualized it in three ways: resource contribution can means provision of labor, cash, material goods and information; administration and co-ordination refers to the possibility of peopl es participation as employees or as members of project advisory or decision-making boards, which helps increase se lf-reliance. The last form of participation regarding implementation is enlistme nt in projects or programs. Benefits of participation refer to the material social or personal gains someone can obtain for participating in a project. Mate rial benefits refer to private goods, and social benefits refer to public goods such as schools or water systems. Pe rsonal benefits refer to the possibility of a group member acquiring political or social pow er through the execution of the project or program, as well as self-esteem and sense of effi cacy. Besides the study of what are the benefits of participation in a project a nd who acquires the benefits, for Cohen and Uphoff (1980:221) it is also important to understand possible harmfu l consequences of participation, and who participates in adverse outcomes. Evaluation participation refers to the ways people participate in ev aluating a project or program. In the project context, evaluations can happen through formal review processes or informal consultation, and it is im portant to know who participates in it, how continuously and
78 what power people have to, based on the evaluation, change the cour se of the actions within the project. Regarding who participates, the authors state that this depends on the characteristics of a project or program, and the distin ction of the participants. In rural development programs it is possible to distinguish possible pa rticipants in four groups, but w ithin each group other kinds of distinction are possible. The gr oups are: local people, local l eaders, government personnel and foreign personnel. How participation occurs is anot her important point, and it adds a qualitative aspect to the analysis of participatory pro cesses. According to Cohen a nd Uphoff (1980:224) the important points to consider are: the ini tiative of participation comes fr om above or from below; the inducements for participation are more volunteer or coercive; the structur e of participation and the channels of participation (i ndividual/collect ive, direct/indirect) ove r time; the duration, and the scope of participation; and finally the em powerment of people to obtain results through participation. To the authors, the processes wh ich are initiated from below and are voluntary, organized, direct, continuous, broa d in scope and empowered are considered to be the most participatory (Cohen and Uphoff 1980:225). The context of participati on, according to Cohen and Uphoff (1980) also deserves important consideration. The author s distinguish context in two form s: (1) the characteristics of the project that refers to t echnological complexity, resource requirements, tangibility, probability, immediacy and divisibility of benefits, program linkages and flexibility, administrative accessibility and coverage; and (2) th e task environment that refers to physical and biological, economic, political, social, cultural and hist orical factors.
79 Finally, it is important to cons ider the purposes of particip ation. According to Cohen and Uphoff (1980), in one project or program the pur poses can be seen different by depending on whose perspective is taken into consideration. However, there are two important things to consider: whether the purpose of one group of pa rticipants differs from the purpose of other groups, and who gets the benefits of participation. Within communities there are many different as pects that contribute to the process of participation in a given project. However, no single theory is able to explain all those aspects, and each author seems to make separate contri butions. Guijt and Shah talk about differences within communities that interfere in the partic ipation of people. There is a nave view of community as a unit that is an harmonious and in ternally equitable collective (Guijt and Shah 1998:1). However it is important to understand the complexity of community differences in participatory processes. Age, economic, religious caste, ethnic and in particular, gender are some of these differences that need consider ation (Guijt and Shah 1998:1). Besides internal differences, participation is also influenced by other aspects such as dissatisfaction with a project. In one case of a participatory rural appraisal in Bahia, Brazil, some people refused to participate in the process as a silent protest against the gov ernment for the creation of an extractive reserve in the area, and as a rejection of the group that was conducting the process (Weigand Jr. 2003). This example shows that fam ilies within communities consider a variety of issues in participating or not in any process. Issues related to local participation are also relevant to participatory research action projects. According to Ferreyra ( 2006:577) in the last few decade s, there has been an emerging debate in participatory research action (PAR) in the context of envir onmental governance that involves multiple stakeholders. The assumption of participatory action research (PAR),
80 according to Wadsworth (1998) represent a new paradigm of science that considers the importance of social and collective processes in obtaining conclusions about a specific question or subject. Most of the part icipatory action research involv es those who share concerns, experiences and interests: those th at have a problematic situation; those trying to assist to change it and those who provides material resources ne cessary to change the situation (Wadsworth 1998). In this study, I first discuss th e importance of local participat ion and partnerships to build an action research project, and the effectiven ess of the institutional arrangements for fire management at Tapajs National Fo rest, considering the frameworks of local participation in the context of a participatory acti on research project. Secondly, I look to how numbers of people participating during the process of local agreemen ts formulation at Tapajs National Forest are related to levels of compliance with the fire management agreements rules the families follow. This was an attempt to measure local participation, and test its relationship to management results. According to Guijt and Shah 1998:10 feasibility of 100 percent of participation is a myth. The local participation context will strongly influence the degree and form of participation. But if 100 percent is a myth, wh at proportion is necessary? What percentage of people participating in processes within a commun ity is sufficient to guarantee that results will be achieved? This is the question that this study attempts to answ er. For that, I test the hypothesis that the greater the number of co mmunity participants, the greater the degree of compliance with the rules of fire management agreements. Attempts to Build a Particip atory Action Research at Tapajs Nation al Forest The fire management project was develope d within Tapajs National Forest (FlonaTapajs), a 600,000 hectares protected area located in the State of Par, north part of Brazil. The area is celebrating 35 years of creation. Flona-Tapajs was esta blished in 1974 through federal
81 decree No 73.684, and its creation was part of a federal government strategy to integrate the region to the other parts of the country (IBAMA-ProManejo 2005). The many communities living within the area just learned of the protect ed areas creation in 1975, when the Instituto Brasileiro de Desenvolvimento Florestal (IBDF Brazilian Institute for Forest Development) started doing an assessment in the area to identify the number of families that were going to be driven out. IBDF was the government agency respons ible for the management of the area at that time. From the year of creation until the end of the 1990s, conflicts marked the relationship between communities, other groups living wi thin the Flona-Tapajs and IBDF (IBAMAProManejo 2005). The main points of conflict we re: (a) IBDF was extremely authoritarian towards the families (Faria 2003; IBAMA-ProManejo 2005); and (b) the limits established by the federal government included areas in which fa milies were living, and the legislation of that time did not allow people to live within nationa l forests. Because of that, a long period of conflict and resistance by the families, supporte d by local organizations, took place in the region (IBAMA-ProManejo 2005). In 1994, through decree No 1.298, the government recognized the presence of the families living within the area, an d their rights to be there. However, among the families there was a strong debate about the possibility of having their areas excluded from the protected area limits. In 1996, 62 per cent of the families voted to have their areas excluded from the Flona-Tapajs. In 2003, a new voting process was carried out and 95 percent of the families decided the opposite (IBAMA-ProManejo 2005). Th rough development projects and a more democratic model of relationship between the government agency and communities, the relationship between government and communitie s has changed in the last decade. Many partnerships were established for the management of the area, and also to promote alternative systems of production for the families living within Flona-Tapajs. Projeto de Apoio ao Manejo
82 Florestal Sustentvel na Amaznia (ProManejo ) is a project funded by PPG-7, and managed by IBAMA. This project at Flona -Tapajs supported activities regarding community management of natural resources, control a nd monitoring of the protected ar ea, ecotourism and environmental education (IBAMA-ProManejo 2005). The Fire Management Project at Flona-Tapajs IBAMA had developed a partnership with comm unities within Flona-Tapajs and established a group of volunteer environmental agen ts composed by local people trained to work in their communities regarding environmental issues. In 2000, IBAMA/ProManejo promoted a meeting with the participation of local leaders representatives of inter-communitarian associations of Flona-Tapajs, the Rural Labor Union of Belterra, the volunteer environmental agents (AVVs), research institutions and other local organizations, to discuss environmental problems of Flona-Tapajs. IPAM was invited to participate and present the experience working with fire management in other regions of Par st ate. Because of that experience, and because one of the problems debated in the meeting was the nu mber of accidental fires that occurred at FlonaTapajs during 1997 and 1998, IPAM was invited to write a project to work at Flona-Tapajs. The previous experiences of IPAM group work ing with fire management had shown some important lessons that helped to guide the process at Tapaj s National Forest. One important lesson was that fire management could not be improved without a strong engagement of local people. Because of that and after deciding to propose a project, the fi rst action of the IPAM group was to organize two meetings with repres entatives from all the communities within FlonaTapajs, and some other local organizations to discuss if the idea was important to the region or not, the activities to be carried out, the communities to be involve d, the time of the project, and the necessary partnerships. At that time, mo st of the communities did not have formal
83 organizations to represent them. Their representati on at the meetings happen ed in three different levels: a) the presidents of the communities (leaders choose by the families to coordinate the communities), b) the presidents of the inter-communitarian a ssociations (formal orga nizations representing communities, and also able to propose projects; Flona-Tapajs had four intercommunitarian associations and each repr esented part of the communities), and c) the Rural Labor Union of Belterra (the organi zation that at the time represented all the communities of Flona-Tapajs located within Belterra municipality). There are some communities of Flona-Tapajs that are located in the municipality of Aveiro, but the Rural Labor Union of this city neve r participated in the activities of the project, although they were invited many times. Based on the debate carried out in those m eetings, a one-year project was proposed to IBAMA/ProManejo. The common goal of the project was to decrease accidental fires at FlonaTapajs (Figure 4-2). The partners at the beginn ing of the project were IPAM as the research proponent, the communities as the research part icipants, IBAMA/ProMan ejo as the financial support and also as the federal government agency responsible for the management of the area, and the Rural Labor Union of Belterra (STR-Bel) as local representative of the communities, and a reference group. Each pa rtner had its own goal. In the case of IPAM, a research institute, we had the goal of studyi ng the effectiveness of local agreements in decreasing the number of acci dental fires within protected areas. This goal was clearly expressed all the time of the partnership, and it brought a constraint in the sense that the communities involved in the process by defi nition were those ones th at wanted to try to formulate agreements and valuate them. Because of that, during the conception of the project, the decision-making power was strong for IPAM th at was the proponent of the project and the manager of the financial resources, followed of IBAMA/ProManejo, the Rural Labor Union and
84 Figure 4-2.Partners at the beginning of the project, their position, their interests, their power rega rding to the project and their level of interaction direct with IPAM. Straight lin es and their denseness indicate the level of power for each partner in the project. The dashed lines and their denseness indicate the relationship of IPAM with each partner Main goal of the project: Decrease accidental fires at Flona-Tapajs IPAM Proponent of the research Interests: verify the efficiency of institutional arrangements for fire management for protected areas, and verify its efficiency in decreasing accidental fires IBAMA/ProManejo financial support and manager of the area Interests: Avoid environmental problems within the protected area and decrease accidental fires. Communities research participants Interests: decrease accidental fires within the communities, get to know the new actor, and guarantee that no decisions regarding to fire was going to happen without them. Rural Labor Union of Belterra local representative of the communities. Interests: increase their political influence within Flona-Tapajs, and support the idea of decreasing accidental fires within the communities Local leaders Local representatives of the communities Interests: Guarantee their inputs in the new project and decrease accidental fires within the communities.
85 the representatives of the communities that par ticipated in the meeting where the ideas for the project formulation were discussed. In the Figur e 4-2 the straight lines and their denseness indicate the level of power for each partner in the project. The dashed lines and their denseness indicate the relationship of IPAM with each partne r. The direct power of the families within the communities only started to exist after the firs t activity carried out within each community. Through this partnership the project was concei ved, and the local leaders and IPAM were the ones that contributed most with definition of acti vities to be developed within the communities, and defining the areas of priority to the project to start. All thos e partners had other interests and other partnerships not re lated to the fire management project, and not represented here in the diagram. After the approval of the project, the fi rst activity was a visit to each community indicated by the local leaders of Flona-Tapajs as the ones with more problems with accidental fires. The trip to the communities had the goal of explaining the project, its objectives, and activities, and to invite the communities to be part of the process. The families in the communities alone could decide to participate in it or not. Sixteen communities were invited based on the decision of the leaders of Flona-Tapajs; 12 accepted, at first, others said that the project was not interesting for them, and some co mmunities invited IPAM to start the project in future years. The acceptance by the communities was the basic condition for the project to happen, considering the IPAM, IBAMA or the Labor Union were not users of fire. It was not the first communication within the communities. The leaders that participated in the formulation of the project had already discussed it internally. However, the ma in goal of the partnership, the construction of it, and also the goal of it to each partner were topics care fully discussed within the communities. It was important to clarify goals of each one, because IPAM was interested in
86 testing a specific instrument that was the formulation of fire agreements, hence, the communities that accepted the project knew preciously that it wa s a proposal to discuss their fire practices, and they were going to try to formulate rules to burn fields. The idea of bringing out the debate about rules to burn to within the comm unities was an attempt to let the families to establish their own rules of how to burn. That was based in two basi c assumptions. First, the federal fire law had no influence in the practice of local families. Although the law was speci fic at the level of establishing procedures of what to do during a burning, some of th e procedures needed adaption, considering farmers reality. Sec ond, there was the assumption that simply discussing techniques of how to burn with the families, without an instrument for organi zing them, would not be effective. This assumption was based on the experience in communities where IPAM worked before. The process was a discussion about fa milies experiences and practices, technical knowledge introduced by IPAM and other agencies and the current Brazili an legislation about fire. It was a process of talking about the families practices, the benefits and problems with them and the possible local solutions. It was not a teaching process about the fire federal law, and what it says farmers should do. It was a process of c onsidering the federal law, considering technical knowledge, and discussing what was possible to be carried out or not. Based on that, year-to-year the families were setting up a set of rules, trying to apply them, and them evaluations carried out by families and IPAM would indicate possible changes in the rules, in an adaptive process. This process was repeated until the commun ities were able to elaborate a li st of rules that was feasible to be applied, and efficient to control accidental fires. The first meetings with the communities gave us, the practitioners, the first lesson regarding participation and participatory proces ses. Because the project was conceived based on a demand from the volunteer environmental agen ts, and because we had discussed its details
87 with most of the leaders of the communities, the IPAM group supposed it was an idea accepted by most of the people. However, there were a number of communities, and families within the communities, that were not interested in the disc ussion of fire management. It is important here to question: which level is local? For us, as practitioners or outsiders, the local leaders and the AVVs were local, and because we were working with them, the project was based in a local demand. However, the practice when we were within the communities showed that even the decisions made by a group of local leaders, peop le from within the communities, did not always represent the demand of the communities, and even when a community agreed with the project there were still the family level to be considere d, and not the families agreed with it. To design the project, we did not have financial support to go to each community and have the same interaction we had with the leader s, but because of that it was im portant to start the process all over within each community, to expl ain all the ideas, the objectives, how it was conceived, to be open to new inputs, criticisms, and to change ideas, activities and plans based on that. During the first year of the project and the beginning of the second, the main activities were the meetings within the communities to di scuss fire practices and techniques implemented by the families to avoid accidental fires, the experimental testing of the efficiency of the techniques available to control fi re, and the training of the facilitators. All these activities were carried out within the communities that accepte d to participate of the project. For each community the actions were: a) an assessment of the fire practices for agriculture; b) experimentation with the efficiency of the t echniques (demonstrative burnings where all the techniques indicated by the families were a pplied, and analyzed after burning); and c) the selection of two people who were trained to carry out some activit ies locally called the facilitators of fire management.
88 The communities indicated them, and to promot e the participation of women in the fire management debate, the communities were asked by IPAM to indicate one man and one woman to represent the communities. At the end, from the 18 communities involved in the project during the years, six indicated women. The responsibili ties of the facilitators were to mobilize the communities for the meetings, to discuss and plan the activities with IPAM, and to help IPAM with the facilitation of the m eetings within their communities. With the involvement of the facilitators, the project had a different configur ation of partnership beca use this group of local people was dialoging with IPAM, planning the next steps of the project and evaluating it (Figure 4-3). The direct interaction of IPAM with community leaders changed once the facilitators started getting involved in the project. The decision of involving f acilitators in the project was based on the fact that the lead ers had many other responsibilities in representing the communities in other initiatives going on in the region, so we discussed together the possibility of new people participating in the fire project. Most of the communities indicated new people, but for some of them, the facilitators were the traditional leaders. Besides, the families were the other group contributing, participating, and evaluating. IBAMA and STR-Bel had less influence. In the case of IBAMA, it was thought that because the goa l was to leave the communities free to take decisions, and although IBAMA was in a process of building a more democratic relationship with families, IBAMA was the federal government agency responsible for the management of the area. In the case of STR-Bel, at that time it was going through a political transition, and there were no representatives directly involved with the project actions. When the new board was elected, IPAM had to start all the negotiations with STR-Bel again, and the new leaders showed more interest in the project.
89 By the end of the second year of the project, IPAM decided to invite some organizations that were involved in activities or projects at Flona-Tapajs to create a discussion group about fire use at the protected area. The main interests of IPAM with this activity were: a) to open the project to be discusse d formally by other organizations; b) to identify common interests of working within the protected area; and c) to guarantee the continuity of the fire management actions within the protected area after the end of the project. The Forum for Good Fire Management was composed by IBAMA/ProManejo, representatives of the communities through the inter-communitarian a ssociations and Rural Labor Union of Belterra, IPAM, Federal Univers ity of Par (UFPa), Projeto Sade e Alegria (PSA), CEFTBAM-Proteger Project, GCI, and la ter the Federation (Federao das Comunidades Tradicionais da Floresta Naciona l do Tapajs) (Figure 4-4). In th is Forum activity plans, agendas and budget were discussed among the organizations that were already working at the protected area, but also new ideas were discussed and so me of the organizations participating had not worked at Flona-Tapajs, but had inte rests in starting to work there. Within the Forum, IPAM started discussing a bout the continuation of the activities within the protected area. The most important questi on was: which organizatio n could take over the project? There were some difficulties to transf er the project to a loca l organization. The first attempt was to find an organization interested in working with fire management. Second, that organization had to be able to work within a ll the communities involved in the project, but also be able to start it in other communities that had showed interest. It was a problem for the intercommunitarian associations b ecause each of them was repres entative of just a number of communities, and political issues would not allow them to work in all of them. Third, after identifying that organization it was necessary to train that organization to use the adaptive
90 Figure 4-3.Partners after the proj ect started, their position, their interests, their power regard ing to the project and their level of interaction direct with IPAM. Main goal of the partnership: Decrease accidental fires at FlonaTapajs IPAM research proponent Interests: verify the efficiency of institutional arrangements for fire management in protected areas, and verify its efficiency in decreasing accidental fires IBAMA/ProManejo financial support and manager of the area Interests: Avoid environmental problems within the protected area and decrease accidental fires. Communities research participants Interests: decrease accidental fires within the communities and guarantee that no decisions regarding to fire was going to happens without them Rural Labor Union of Belterra local representative of the communities. Interests: increase their political influence within Flona-Tapajs, and support the idea of decreasing accidental fires within the communities Facilitators representatives of the communities Interests: contribute to the planning of the project, get training and grow as leaders within the community.
91 Figure 4-4.Partners in the third and fourth year, their interests and their powe r of decision-making in the project. Main goal of the partnership: Create a common background about fire management within Flona-Tapajs and identify possibilities of partnership between different organizations workin g in the area. Rural Labor Union of Belterra local representative of the communities. Interests: continue the activities regarding to fire management and consolidate its representation within the communities. Federation New organization with representation in all the communities of Flona-Tapajs. Interests: continue the activities regarding to fire management and consolidate its representation within the communities. PSA Working within almost all the communities in project related to health and alternative systems of production. Interests: include fire management in their projects about alternative systems of p roductio n IPAM Proponent Interests: Reflect its project within Flona-Tapajs and guarantee the continuity of fire mana g ement p ro j ect. IBAMA/ProManejo financial support and manager of the area Interests: Avoid environmental problems within the protected area Inter-communitarian representatives of the communities Interests: guarantee demands of families in the debates. GCI organization working with indigenous communities of FlonaTapajs. Interests: adapt the experience to those communities. UFPa Research institution Interests: was implementing a research project within the protected area CEFTBAM (Projeto proteger) Developing another project with focus on fire management. Interests: contribuite with tier experience to the project within Flona-Tapajs
92 methodology that was initiated within the co mmunities, and finally to find financial support for that organization to keep the activit ies going. While the fire management project was going on, a commission invol ving all the communities of Flona-T apajs was formed to discuss with IBAMA the agrarian situation of the prot ected area. This commission later formed the Federation (Federao das Comunidades Tradicionais da Floresta Nacional do Tapajs), and the leaders were interested in IPAM activities w ithin the area. A partne rship was established between IPAM, the Federation and the Rural Labor Union of Belterra. IPAM was going to train these two organizations and support them to write projects to work with in, and on the border of Flona-Tapajs. With the Forum, the number of partners increas ed, but after some mee tings just some of them kept interested in discus sing fire and planning activities together. Thus, by the end those still more involved were the communities, the faci litators, IPAM, the STRBel, after the political transition, and the Federation. Before the projec t was finished, the Federation and the STR-Bel were the organizations trained and with projects submitted to funders to de velop activities within and along the border of Flona-Tapajs, to work with fire management and also alternative systems of production that could decrease not only th e accidental fires, but also the need to burn every year. One of the projects was approved, and the STR/Bel is implementing it in the border of Flona-Tapajs. The project of the Fede ration was not approved, mainly because the organization was not formal at that time, a nd all the funders available asked for formal organizations with the legal requirements to submit proposals. IPAM, IBAMA and the Federation are still discussing a way to keep the work with fire management within the communities of Flona-Tapajs.
93 Reflections on the Process of Partn ering for A ction Research Project Based on Stones schematic range of participa tion, the fire management project could be classified as a co-participation initiative, because the project was coordinated by IPAM, but had many other actors involved such as local families, local representatives (fac ilitators and leaders), other local organizations such as associat ions, labor unions, NGOs and some government organizations/institutions such IBAMA/ProManej o, and the university. At different moments, and at different levels, all those organizations contributed to proj ect and had power to change its activities and plans. The projec t was a local demand, if we c onsider leaders as legitimate representatives of local people. It may not be located at the end of the line, as a full participatory process, because the financial administration of the project was done by IPAM, it was not a family level demand that brought the necessity of a project to discuss fire management, and it was not a community project, but a regional action involving many di fferent communities. Besides, at least among the main partners (IPA M, facilitators and co mmunities), there was a strong level of dependency. In all the conversations it was stated th e project would continue only if the communities were still in terested in discussing their fire management practices. During the four years of the project three communities opted for stopping the activities. In two of them, one that was on the boarder of the pr otected area, and the second one th at had half of its territory inside the protected are, and th e other half outside, many families were selling their lands to the soybean producers that were moving from cen tral Brazil to the north region. This process resulted in a low level of interests of the remaining families regarding the fire management project. The third community had internal po litical problems, and af ter a community board transition, the new leaders were not interested in organizing the meetings for fire management, mainly because it was an activity that started in partnership with the other group of leaders. From
94 2001 to 2004, five new communities started involved in the activities, and in the last year, other invited IPAM to visit them, but by that time the project was being concluded. Regarding Cohen and Uphoffs (1980) contribut ions to the analyses of participatory processes, the fire management pr oject could not be characterized st rictly in any of the kinds of participation described by the authors. There were situations during the process that most of the actors involved participated in decision-making processes. At th e beginning of the project, its conceptualization, all the actor s involved (IPAM, IBAMA, local leaders, and STR-Bel) had important role in defining the project, its goa ls and activities. After the conceptualization moment, when the project went to the communiti es, the families had the opportunity to make decisions regarding first: to participate or not in the initiative, the activities of the project, the rules in the agreements, and whether to use th e rules or not. The moments of implementation were basically carried out by IPAM and facilitators, organizing the trainings and the meetings within the communities, and by the families, who were the ones to try to apply the rules of their agreements. The evaluation processe s happened in many different levels First, a sample of fields was selected and visited by IPAM technicians to analyze the use of the rules. Besides that, a meeting with all the families was carried out w ithin each community, every year, to discuss the use of the rules, to analyze their efficiency and difficulties, and based on that evaluation, to change the rules for the next burning period. Evaluation of each activity was also carried out in each community, and that was the opportunity for families to express their concerns, propostions, and ideas. During the trainings, the facilitator also had the opportunity to evaluate each activity and plan future actions. The participation was voluntary for all the partners, and the communities that decided not to participate had not suffered any kind of pressure for that decision. The invitation was always open for them to reconsider. Considering the goals of the
95 project to reduce accidental fires, but also consid ering the interests of eac h partner, the benefits of the project were many. The numbers of accident al fire decreased the establishment of fire agreements through an adaptive process showed to be effective for protected areas and it was a benefit to all the partners. One important point regarding the project was that IPAM made clear to all the communities, that despite the goal of decreasing accidental fires, for IPAM it was a research project. That clarification of goals created a demand within the communities to understand the usefulness of research projects in general, no t only this one. There were other projects being developed at Flona-Tapajs, thus when IPAM st arted the action research, we faced a need to explain to the communities what is research, why is that important, for whom, and how it is done. That clarification was necessary because th e communities had complaint about research. In their words, researchers were pe ople who would come to then, as k many questions, and leave. In the case of fire management, in each evaluation, we would bring the information produced in each year to be used in the m eeting to support the decisions. The attempts to build participatory action research at Flona-Tapajs show that the engagement of many different stakeholders is not an easy task. It takes time, human and financial resources, continuous debate of interests and go als and constant adaptation, but the openness to participation of different stakehol ders contributed not only to the execution of the action research project, but more importantly to the achievement of the main goal of the partnership itself: the reduction of accidental fires at Flona-Tapajs. Do Numbers of People Count? Participation and Compliance with Rules The m ain goal of this part of the study is to understand if different numbers of people participating during the process of fire agreements formulation would result in different levels of compliance with the rules established by the fa milies for their fire management agreements.
96 The concept of community used in this study if that community is a geographic collective, a group of people who live (full and/or part time ) in a locality and ar e connected by a web of emotional, economic and/or relational bonds and a culture, and share a set of values, norms, and meanings (Guijt and Shah 1998:268). The hypothesis I tested in this study was: Communities with higher level of participation during the process of formulation of the agreements follow the rules more than communities with lower levels of participation. The first step of this research was the iden tification, for each commun ity, of the level of communities adult participation in the meetings where the fire management agreements were elaborated or reviewed. Because communities started the project in different years, I decided to work only with the communities that were pa rt of the project from 2001 to 2004. From 18 communities involved in the project, 8 were partic ipating all the four y ears since the beginning. To analyze the level of particip ation I used the attendance lists for each of the 8 communities to see what percentage of the population above 16 years old living in each community was participating in the meetings car ried out during the years of the project. The age of 16 was taken based on the fact that the teenagers around that ag e used to come to the meetings sometimes, in most of the communities. During the project, th e communities held many meetings related with fire management, but some of those meetings were not directly related with the formulation or reformulation of the agreements. Therefore, I used only the attendance lists from the meetings in which the families formulated and reviewed the agreements, as there were very important moments in the process of formulation and adap tation of the agreements. The list of attendance had only the name of the person participating in the meeting, and I was interested in knowing the proportion of the adult population those people in the lists repr esented to each community. To
97 find this information I first listed the number of families in each community in each year of the project. Then I used the average number of adult population using census data from IBAMA to estimate what percentage of that population was present in the meetings for the four years of the project. Communities were ranked according to the le vel of participation in the meetings carried out during the period between 2001 and 2004 (table 4-1). Table 4-1.Average number of adults participatin g in the meetings as a percent of the adult population of each community for the years of the project. Communities Average for all the years (%) 1 Prainha II 64 2 Paraso 49 3 Itapuama 37 4 Itapaiuna 34 5 Prainha I 30 6 Nazare 27 7 Chib 25 8 Piquiatuba 15 Using this ranking I selected four communities in which to interview families. The first community selected was Prainha II, which had the highest percen tage of participating in the meetings. It means that for the meetings I analyz ed, in the community of Prainha II, an average of 64% of the adult population came to the meet ings. However, at the time the project was carried out, Prainha II was not yet recognized by IBAMA as a comm unity. In order to avoid bias, I also selected the community of Paraso, which had the second highest leve l of participation in the meetings. Finally, I selected Chib and Pi quiatuba, which had the two lowest level of participation in the meetings duri ng the four years of the project. After this selection I traveled to each community to intervie w the families. My goal was to interview all the families of each community, but in the end, because of time constraints, I interviewed as many families as possible for each community. In Prainha II, I interviewed twelve
98 of nineteen families. In Paraso, I interviewed ten of the eleven families. In Chib I interviewed twelve of 43 families, and in Piquiatuba I interviewed nineteen of fifty-six families living within the community at the time I carried out the surveys. The families were selected randomly at first. However, beyond the families selected randomly, I did a second selection to interview only families who had burnt fields during the summer of 2006 or 2005, and families in which the interviewed people were actually in the field on the day it was burnt. These criteria were necessary because I was in terested in the use of fire manage ment techniques, and only those who had burned their fields could gi ve me that information. The in terviews were conducted at the families homes with male and female heads present in almost all the interviews. There were a few cases in which the woman was not at home at the time of the interview, and some cases of a single member family. The interviews were conducted during June of 2007. I used a structured survey to interview all the families. The survey was organized in different sections. The first section had questions related to the identi fication of the household heads (male and female) such as their names, ti me living within the community, and if they had lived in other regions before co ming to the area. The second secti on had questions related to their last crop, and how they burned th eir fields, which techniques they used, and if they had accidental fires. In this section the people interv iewed were asked to draw their plots and show where they built break fires, in which position they started the fire, the vegetation surrounding the field, and the direction of the wind. This info rmation was necessary to evaluate if they had followed some techniques previously established in their fire agreements. The third section was related with their fire management agreements. In this section I aske d questions about their opinions regarding their fire agreements, if they knew the community had a fire agreement, if they participated during the process of fire agre ements formulation, if they remembered the rules
99 of the agreements, their opinion regarding the rules, and the current process of discussion of fire management within the communities. The last part of the survey had questions related with their participation in community organizations such as associations, religi ous organizations, labor unions, and so on. For this study, I had one unit of analysis which was the community, one dependent variable and one independent variable. To test the hypothesis the de pendent variable was compliance with the rules, and the independent variable was the level of participation. Communities Studied Piquiatuba community There is no clear information on when Pi quiatuba community was founded. However, accord ing to Faria (2003) in 1912 there were already some families in the area. By 1956 there were 14 families living in the area. Today, Piquiatuba has 70 families living in three different clusters of houses. Even though this community ha s strong family ties, nowadays there are many different family nuclei. The community has a total population of 370 people. Piquiatuba has many different forms of organi zation. One is the Associao dos Moradores e Produtores Rurais e Extrativistas da Comunidade de Piquiatuba that re presents the community externally, and organizes internal activities such as monthly meetings, parties, games and other collective activities. The community members elec t the association council for a two-year term. Piquiatuba is also represented by ASMIPRUT (Associao Intercomunitria de Mini e Pequenos Produtores Rurais e Extrativistas da Margem Di reita do Rio Tapajs de Piquiatuba a Revolta). The ASMIPRUT is an intercommunity associati on that represents eight neighbor communities. The families of Piquiatuba are associated to the Rural Labor Union of Belterra. There is also one youth group responsible for a collective radio, an d for organizing the you ngest people of the community. Piquiatuba has a womens group, and four soccer teams. Although most of the
100 families are Catholics, there is also a Protestant church within the community. Piquiatuba has a school that offers classes from the first grade to high school. For that reason students from other communities attended there. The families living at Piquiatuba practice subsistence agriculture, and have an economy based on crop production, extractivism and governme nt benefits. The main products are cassava flour and cattle. Also families collect some non-tim ber forest products such as aa and Brazil nuts. There are more than 50 people in the comm unity that work for the government, private companies, or receive money through governme nt programs and retirement. Fishing is an important source of protein to this comm unity, but is not a commercial activity. Piquiatuba has been involved in many differe nt projects. The Projeto Sade e Alegria (Health and Happiness Project), a local NGO located in Santarem city, carried out different initiatives involving communities at Flona-Tap ajs. Alternative systems of production, organizational strengthening and acti vities regarding to health care were some of the projects in which Piquiatuba community was involved. The co mmunity is currently involved in the Amb Project, an initiative ca rried out by the Cooperativa Mista Flona Tapajs Verde (COOMFLONA). The project exploits forest products to the benefit of the families living within the communities. Regarding the fire management project Piquiatuba was involved since the first activities in 2001. During the four ye ars of the fire management proj ect activities, Piquiatuba was the community with the lowe st level of participation. Chib community The fa milies that nowadays are the community of Chib were part of the neighbor community of Tauari. Political problems and also the distance between the clusters of families resulted in the separation of the one group of people that in 2000 started the community of
101 Chib. Chib has 43 families and approximate ly 180 people. Although there are extended families living in Chib, this community has a variety of different families. Chib is organized in a formal associa tion called the Associao Comunitria de Produtores Rurais e Extrativis tas da rea do Planalto do Chib (ASCPREC). The association represents the community extern ally, and organizes internal ac tivities. The community members elect the association council for a two-year term. Chib is also represented by AITA (Associao Intercomunitria do Tapajs de Itapuama a Marituba). The families of Chib are associated to the Rural Labor Union of Belterra. Chib has Catholic and Protestant members, but they only have a Protestant church. The lo cal school offers classes from the first to the fourth grade. The families living at Chib practice subsiste nce agriculture, and have an economy based on crop production, extractivism a nd government benefits. The main products are cassava flour and cattle. Some years ago through the local association, around 15 families of the community got involved in a project to plant cumaru ( Dipteryx odorata ) and curau (Ananas erectifolius ). The plan is to produce and commercialize the cumaru seeds and curau fiber. The project had support of state and federal government through th e local technical assi stance agency and banks Besides the cumaru-curau project, Chib has been involved in the activities carried out by Sade e Alegria Project. Regarding to fire ma nagement project Chib was involved since the first activities in 2001. During the f our years of the fire management project activities, Chib had a low to medium level of participation. Paraso community The first fam ilies living in Paraso were in the area since the 1930s (Faria 2003). At that time they were three brothers and sisters w ho moved to the area and started the community. Strong kinship is a very important characteristic of Paraso. Almost all th e families are relatives.
102 According to Faria (2003) only in 1991, did the fa milies living in the area recognize themselves as a community. The community has a total population of 75 people divided in 11 families. Paraso does not have a formal organization. The community has an informal council that represents them and organizes in ternal activities. The community members elect this council for a two-year term. Formally, Paraso is repr esented by AITA (Associao Intercomunitria do Tapajs). The AITA is an intercommunity associ ation that represents the neighbor communities. The families of Paraso are associated to the Rural Labor Union of Aveiro. Most of the families are Catholics, but the co mmunity does not have a church. Paraso has a school that offers classes from the first grade to the fourth grade. Some t eenagers of the community have to go everyday to a neighbor community that offers higher levels of education. The families living at Paraso practice subsistence agriculture. The economy of the families living at Paraso is based on crop production, cattle and government benefits such as retirement, and the Bolsa Famlia Program. There are two people who work in the local school, and are government employees. Paraso was involved with the activities de veloped by the Projeto Sade e Alegria. Currently the community is invol ved in the Amb Project. Regarding to the fire managemnt project, Paraso was involved since the first activities in 2001. During the four years of the project activities, Paraso had always around fifty percent of the families participating in the meetings. Prainha II community Prainha II was founded in 1998. The fa milies livi ng in Prainha II previously were part of the neighbor community of Prainha. Because of political issues, one part of the community moved to another area and started Prainha II. Siblings mostly formed the group that moved in 1998, and their father was an important leader in Prainha, who moved with his sons and
103 daughters and created the new community. Prainha II has 15 families and approximately 112 people. The community has a formal organization ca lled ASNUTA (Associao Intercomunitria Nova Unio do Tapajs) that organizes intern al activities and represents the community externally. The community members elect the boa rd of the ASNUTA. The families of Prainha II are associated to the Rural Labor Union of Belterra. Most of the families are Protestants, and they have a church that is al so a training base for the regi on. The children and teenagers of Prainha II attend the school at Pr ainha, the neighbor community that offers classes from the first grade to high school. The families living at Prainha II practice s ubsistence agriculture. The economy of the families living at Prainha II also is based on crop production, cattle and government benefits such as retirement, and Bolsa Famlia Program. Most of the families received credit finance to produce urucum ( Bixa orellana L.) and curau (Ananas erectifolius ). The families sell the seeds of urucum and fiber of curau. Cattle is also a important activity in Prainha II, but most of the families get income through the production of cassava flour, urucum and curau. Prainha II has been involved in many different projects. The Projeto Sade e Alegria is one of the NGOs that worked within this community carrying out hea lth activities su ch as vaccines and preventive programs. The community memb ers through ASNUTA got involved in a project to produce urucum ( Bixa orellana L.) and curau ( Ananas erectifolius). The community is currently involved in th e Amb Project. Regarding the fire management project Prainha II was involved since the first activities in 2001. During th e four years of the fire management project activities, Prainha II always had the highest level of familys participation.
104 Results This part of the study I aim ed to understand if different numbers of people participating during the process of fire agreements formulation would result in different levels of compliance with the rules established by the families for th eir fire management agreements. For the four communities I analyzed, there are seven fire ma nagement rules common in all the agreements. These include: Authorization from IBAMA, cu t dead trees, build fire break around the field, time to burn, number of people, start the fire agai nst the wind and watch the field until the fire finish (Table 4-3). However this is not the to tal number of rules that each community has. Piquiatuba has a total of 12 rules; Chib and Para so have each 15 rules, while Prainha II has 16 rules in its agreement. Thus the communities with higher levels of participation in rule-setting meeting also had higher number of t echniques included in the agreements. Using the data from the four communities a bout the techniques each family used to burn their field during the summer of 2006 or 2005, a nd using the list of t echniques previously defined in each fire agreement of the four communities, I anal yzed the frequency of people in each community following the rules in their agre ements. Piquiatuba has 12 rules. The minimum of rules people use is 5, and the maximum is 12 rules. The majority of them (68.4 percent) follow 7 rules or more. Chib has 15 rules in its agreement. The minimum of rules people follow is 7 and the maximum is 15. The majority (91 percent) follows 10 rules or more. Paraso agreement has 15 rules. People follow the minimu m of 5 rules and the maximum of 10 rules. People follow from five to 10 rules. The majority (80 percent) follow 7 rules or more. Prainha II agreement has 16 rules. The minimum rules peop le use is 8 rules and the maximum is 15. The majority (91 percent) follows 10 rules (Table 4-3).
105 I found that from those 7 rules, most of the families (77.3 percent) follow four or more seven rules (Table 4-4). Twenty-two percent of the families use only two or three of the seven common rules. Table 4-2.Fire management techniques in the communities agreements, and rules in common Column A Column B Column C Column D Column E Communities Rules Piquiatuba Chib Paraso Prainha II Rules in common Authorization from IBAMA X X X X X Cut the trees towards the field X X X Lower the fuel load X X Cut dead trees X X X X X Talk to neighbors to work together X X X Build fire break around the field X X X X X Clean the fire break before burning X X Time to burn X X X X X Number of people X X X X X Plan the burn day X *X X Start the fire against the wind X X X X X Watch the field until the fire finish X X X X X Wear adequate clothes X Prepare source of ignition before X *X Month of burn X X X Inform neighbors of the day to burn X X Water X X X Bring tools to the field X X Distance from rivers X X Distance from roads X X Total rules in each community 12 15 15 16 7
106 Table 4-3. Number of rules in each community agreement, and minimum and maximum of rules followed. Communities # of rules in the agreement Minimum of rules followed Maximum of rules followed Piquiatuba 12 5 12 Chib 15 7 15 Paraso 15 5 10 Prainha II 16 8 15 Table 4-4. Frequency of people using the 7 rules in common in the four communities. Frequency Percent Valid Percent Cumulative Percent 2,00 5 9,4 9,4 9,4 3,00 7 13,2 13,2 22,6 4,00 12 22,6 22,6 45,3 5,00 11 20,8 20,8 66,0 6,00 13 24,5 24,5 90,6 7,00 5 9,4 9,4 100,0 Total 53 100,0 100,0 Analyzing the communities separately considering each category of participation (higher and lower) and the seven rules all of them have in common, I found that in Prainha II and Paraso (higher participation) 81 percent of the families use four to seven of those common rules. In Piquiatuba and Chib, 74.1 percent of the familie s use four or more of the common rules. Separating the analyses of common rules for the community with hi gh and low level of participation, I found that Prai nha II and Paraso (high participation) have 11 rules in common, and 81 percent of the families follow six rules or more. The communities with low participation (Piquiatuba II and Chib) have eight common rules, and 51.5 percent of the families follow six rules or more. In Prainha 87.5 percent of th e rules of their agreements are followed by more than 50 percent of the families. In Paraso 60 percent of the rules are followed by more than 50 percent
107 of the families. In Chib, 100 percent of the rules are followed by more than 50 percent of the families, and in Piquiatuba 67 percent of the rules are followed by 50 percent of the families. All the analyses suggest that different number of people participating in the meetings is not an indicator of how people wi ll proceed with rules complian ce. Although a comparison between Prainha II (higher level) and Piquiatuba (lower le vel) indicate differences in the number of rules they set up in the their agreements (Table 3-4), and also in the percentage of the rules followed by most of the people, this is not the case for the comparison between Paraso (higher level), and Chib (lower level) where the number of rules are the same and the percentage of people following the rules is higher in Chib. Looking at the common rules (Table 4-5) there are rules that Prainha and Paraso follow more than Piquiat uba and Chib, but there are also rules that Piquiatuba or Chib follow more. Discussion Regarding to the analyses of the difference in number of people par ticipating in the meetings and possible differences in levels of co mpliance with rules, in this study, in the four communities where I carried out the research, the analyses suggest that there are no differences in the behavior of the families regarding compliance with the rules. One possible explanation for that is the fact that at Flona-Tapajs the co mmunities are stable, the families living there are known by each other for a very long time, in most of the cases since they were born. The fact that there were not new families moving to the communities, and that they are usually relatives, can result in a strong le vel of communication between them th at might be more important than their presence in the meetings. Participation, according to Ostrom (1990) is ju st one part of the process of having strong common rules to manage natural resources. Ther e are other important elements such as a monitoring system and sanction system for thos e that do not follow the rules. At Tapajs
108 National Forest these two important elements are weak. This could explain why there are not large differences between the comm unities with high or low levels of participation in meetings, and in the percentage of families in each commun ity following the rules of the fire agreements. Table 4-5.Percentage of the population by community following the rules Column AColumn BColumn CColumn D Column E Rules PiquiatubaChib Paraso Prainha II Rules in common Authorization from IBAMA 94.7 100.0 100.0 100.0 X Cut the trees towards the field 78.9 60.0 58.3 Lower the fuel load 50.0 91.7 Cut dead trees 10.5 50.0 10.0 16.7 X Talk to neighbors to work together 21.1 20.0 0.0 Build fire break around the field 42.1 75.0 50.0 50.0 X Clean the fire break before burning 50.0 75.0 Time to burn 84.2 100.0 70.0 75.0 X Number of people 31.6 75.0 30.0 58.3 X Plan the burn day 84.2 91.7 Start the fire against the wind 52.6 58.3 80.0 91.7 X Watch the field until th e fire finish 100.0 91.7 100.0 91.7 X Wear adequate clothes 100.0 Prepare source of ignition before 66.7 Month of burn 89.5 100.0 100.0 Inform neighbors of the day to burn 66.7 60.0 Water 75.0 30.0 58.3 Bring tools to the field 100.0 90.0 Distance from rivers 83.3 100.0 Distance from roads 63.2 91.7 Total rules 12 15 15 16 7 There are other points to consider in the disc ussion of the results. In the case of FlonaTapajs, the number of families living within the communities suffers a year-to-year variation. Some families leave the communities to work in temporary jobs. Those families continue to be counted as families living within the communitie s and this fact can have influenced the construction of the index of participation. If one community had only 50 families at the
109 community when the project was carried out, but the leaders said they were 60 families counting the absent, it would meant that inaccurate information was used to build the index. More research is necessary to define the variables that explain why communities have a very similar pattern regarding to the application of the rules established in their agreements. One possibility is that the number of adult people, that was the information I used to create the index of participation, is not the best measure to qua ntify participation. Maybe the numbers of families represented in each meeting could give a better idea of which household was represented or not in the meetings. Nevertheless, in the case of this research this information was not available. Quality of participation is also another kind of information that could be used to compare communities, but again there were not system atized information about the quality of participation for the communitie s involved in the project. Conclusions The construction of partnerships to guarant ee the involvement of local people and local organizations in a research project is a comple x process, and for that reason it is difficult to capture all the aspects that contribute to it or not. However, the descriptions of experiences are helpful to bring an understanding of how it occurs in the field. The case of Tapajs National Forest presented here shows some importa nt lessons and reflections pointed out. 1. In a process involving many different co mmunities and different organization, the definition of local is complex and has different levels. If we consid er as local, only the families within each community, then all the strategies of projects and programs that consider local leaders as representative of th e communities need to change. The case of the fire management project at Flona-Tapajs s howed that locals have many different levels, such as local leaders, the community level, fa mily level, but also ot her organizations that work in the region and are considered to be local organizations. In the case of the fire management project at Flona-Tapajs, an e xperience involving more than one community, to say that the process was participatory a nd included local people, it is necessary to consider all these levels. 2. Even when the leaders of a region are involved in the construction of a project, it does not mean that their ideas are repr esentative of all th e communities and families within a region.
110 However, even a process that did not start w ith the acknowledgement of all the families can change during its execution in a way to guarantee more effective participation of locals in different scales. In the case of Flona-Tapajs, it reflected in a challenge for IPAM during the first year of the project. Families were conc erned with the real intentions of the project, and it took time and dialog until the families understood that no decisions regarding fire management were going to be made without their involvement, and felt more comfortable to participate and contribute in the debate of fire management strategies. Considering that, practitioners have to be aware of the importa nce, once within the communities, of dialoging with families, and that most of the time it is going to be necessary to start the process all over again, re-discuss everything, including the go als of the project a nd activities planned. It also requires a level of flexibility of the practitioners, to change plans as needed. 3. An initiative that was not conceived or dire ctly administrated by the communities can still have a high level of sharing power, if outsiders are willing and able to change initial plans and to see local people inputs as contributions and not as criticisms The agencies funding the projects also have to be open to cha nges in what was previously determined 4. Just because the work is being develope d closely involving representatives of the communities, it does not mean that the information is going to be distributed to most of the families. Again, sometimes topics needed to be discussed again and again. 5. Finally, in the case of Flona-Tapajs, the fact that IPAM had been working in the region with other projects before, and that there were a local technician res ponsible for the project helped to build the necessary alliances to guarantee more involvement of local organizations and even of the families within the communities. In the case of fire management at Flona-T apajs, although at the community level this study showed that numbers of people participati ng in the meetings was not decisive to their compliance with the rules, the qualitative analys is showed that only with clear dialog between the research organization, the communities and many other partners working in the region could the action research process be carried out. The atte mpts to build different levels of partnership meant more time, resources, cont inuous debate of interests and goals and many changes during the process, but it was fundamental to the legi timacy of the project, and ultimately, to the effectiveness of the institutional arrangements for fire management that were developed during the fire management project, and that was main goal of the partnership its elf at Tapajs National Forest.
111 CHAPTER 5 1 CONCLUSIONS Fire is a very im portant element in the agricultural production syst em used all over the world. In the Brazilian Amazon, wh ich holds the largest tropical forest of the world, about 18,000 km2 per year of forest biomass has been bu rned and converted into crop fields and pastures in the last decade (INPE 2008). Althoug h its benefits are very important for the land uses that are part of the regi onal economy, much research has called attention to the problems associated with this practice. Fire can easily spread and escape from its intended burning area, affecting neighboring areas and productive systems. In spite of its problems, fire will remain a part of the agricultural system in the Amazon in the coming decades. The analyses of the use of fire in the Amaz on along with the problems associated with this practice, and the attempts by government and civil society to decrease the number of accidental fires in the region all indicate that changes in th e use of fire in Amazon are not likely to take place from one year to the next. Fire is so importa nt in the agricultural system of the region that important changes need to be carri ed out through initiatives that dir ectly engage the users of fire locally, and that are establishe d through an adaptive process where people can experience new models, evaluate and adapt them according to th eir conditions. Although the fire legislation of 1998 is an important gain marking federal governm ent recognition of accidental fires as problem in the Amazon, it recommends procedures that are very difficult for famers to implement. The need of authorization to burn, a nd the size of fire breaks are exam ple of those procedures. Other regional programs carried out by NGOs are effectiv e in increasing the concerns of people about environmental issues in general, but not specifically about fire management. This study analyzed the contributions of ins titutional arrangements for fire management based on the project carried out at Flona-Tapaj s. The analyses indicate that institutional
112 arrangements can contribute to cope with the pr oblem of accidental fires, even when the process is stimulated by an external agent. At Flona-Tapajs, four years of participatory process to establish fire management rules resulted in a increase in the number of technique s used, and also in the number of people using them. Thus, the risk of accidental fires decreased as pointed out by all the families. However, developing successful institutional arrangements for fire management depends on several basic principles. Examples include the congruence between costs and benefits to implement the rules, and clear definition of boundaries (Ostrom 1990). The case of Flona -Tapajs showed that it is also important that the initiative take place at the local level, with strong participation and engagement of the farmers who use fire and al so chose the techniques and recommendations to be applied. It also showed that the decision power that is in the hands of families needs to be clearly stated. All communities and families need to know that they can choose to participate or not, and can stop the process at a nytime. The decisions regarding th e techniques also have to be truly in the hand of the families. The process needs to be adaptive so that all those involved learn from success and mistakes, and adapt accordi ngly (Mulder and Coppolillo 2005). It needs to happen through a focused and a continuous process with different kinds of ac tivities. Finally, it is necessary to design ways, at the community level, for families to debate fire management agreements year to year. This stimulates a con tinued concern and awarene ss of the topic, which transmits this concern to subsequent generations and creates an envir onment to the introduction of other changes in the agricultural systems, wh ich could ultimately diminish or even exclude fire from the system of production. Finally, this study aimed to discuss the fi re management project developed through a partnership between an NGO and rural communitie s of FlonaTapajs. The study explored the
113 relantionship between the number of community participants and the level of compliance with the fire management agreements rules that famili es follow. In the four communities I carried out the research, the analyses suggest that there were no significant diffe rences in the behavior of the families regarding compliance with the rules. One po ssible explanation for that is the fact that, at Flona-Tapajs, the communities are stable and the families living there know each other for a very long time, in most cases since they were born. The fact that there are not new families moving to the communities, and th at often they are relatives ma y result in a strong level of communication between them that can be more im portant that their presence in the meetings. The results of this research sugge st that, at the community level, perhaps there is no need to have full community participation in their meetings. However, in the case of Flona-Tapajs, higher rates of participation generated more debate duri ng the meetings, and greater confidence that the decisions were representa tive of the community. The construction of partnerships that guarant ee the involvement of local people and local organization in a project is a comp lex process. It is therefore difficult to capture all the aspects that contribute to the success of it. Although community level results s uggested that the number of people participating in the meetings was not d ecisive to their compliance with the rules, the qualitative analysis showed that the action research process depended on a dialog between the research organization, the communities and other local stakeholders. The process involved more time, resources, continuous debate of interest s and goals, as well as, many changes during the process, but these investments were fundamental to the legitimacy of th e project. They were similarly important to the effectiveness of the institutional arrangements for fire management that were developed during the fire manageme nt project, which was the main goal of the partnership itself at Tapajs Na tional Forest. Thus, the case of Ta pajs National Forest indicates
114 that a partnership between an outsider organizatio n and the families increased the concern of the people regarding fire, and that local participation is required if su ccessful fire management is to be achieved. Without the users of fire invol ved in the process no re sults can be achieved. Ultimately, farmers are the ones who are going to use fire, and decide they are the ones who which techniques and recommendations they will use to avoid accidental fires.
115 LIST OF REFERENCES Adam s, William and David Hulme 2001 Conservation and Community : Changing Narratives, Policies and Practices in African Conservation. In African W ildlife and Livelihoods: the Promise and Performance of Community Conservation. David Hulme and Marshall Murphree, eds. Pp. 9-23. Oxford: James Currey. Agrawal, Arun and Maria C. Lemos 2007 A Greener Revolution in the Maki ng? Environment Governance in the 21st Century. Environment 49 (5):36-45. Agrawal, Arun. 2001 Common Property Institutions and Sustainable Governance of Resources. World Development 29: 1649-1672. Albuquerque Jos H. B., James J. B. Gomes, Carlos B. da Costa, Charles da S. Santos and Irving F. Brown 2007 Viso da Defesa Civil do Estado do Acre na Aplicao das Ferramentas de Sensoriamento Remoto para o Controle e Combate s Queimadas do ano de 2005. Anais XIII Simpsio Brasileiro de Sensoriamento Remoto INPE. 4413-4420. Alencar Ane, Daniel Nepstad and Maria del Carmen Diaz 2006 Forest Understory Fire in the Brazilian Amazon in ENSO and Non-ENSO Years: area Burned and Committed Carbon Emissions. Earth Interactions 10 (6): 1-17. Alencar, Ane., Daniel Nepstad and Paulo Moutinho. 2005. Carbon Emissions Associated with Forest Fires in Brazil. Tropical Deforestation and Climate Change. P. Moutinho and S. Schwar tzman. Belm, PA, Instituto de Pesquisa ambiental da Amaznia and Environmental Defense: 23-33. Alencar, A., L. Solorzano and D. Nepstad. 2004. Modeling Forest Understory Fires in an Eastern Amazonian Landscape. Ecological Application 14(4): S139-S149. Almeida, Oriana., Lucimar Souza, and Liana Rodrigues 2006 Conservation of Biological Diversity and Fi re Control on Small Amazonian Properties. In What on Earth is Biodiversity? Sa o Paulo: Editora Pe irpolis: 151-154. Arago Luiz E. O. C., Yadvinder Malhi, Ni colas Barbier, Andre Lima, Yosio Shimabukuro, Liana Anderson and Sassan Saatchi 2008 Interactions between rainfall, deforest ation and fires during recent years in the Brazilian Amazonia. Philosophical Transa ctions of the Royal Society B 363: 1779.
116 Artaxo Paulo, Luciana V. Gatti, Ana M. C. Leal, Karla M. Longo, Saulo R. Freitas, Luciene L. Lara, Theotonio M. Pauliquevis, Alin e S. Procpio and Luciana V. Rizzo 2005 Atmospheric Chemistry in Amazonia: The forest and the biomass burning emissions controlling the composition of the Amazonian atmosphere. Acta Amazonica 35(2): 185 196. Barbosa Reinaldo I. and Philip M. Fearnside 1999 Incndios na Amaznia Brasileira: Estima tiva da Emisso de Gases do Efeito Estufa Pela Queima de Diferentes Ecossistemas de Roraima na Passagem do Evento El Nio (1997/98). Acta Amazonica 29 (4): 513-534. Barlow Jos, Tor Haugaasen and Carlos A. Peres. 2002 Effects of Ground Fires on Understory Bird Assemblages in Amazonian Forests. Biological Conservation 105(1):157-169. Barrow, Edmund and Marshall Murphree 2001 Conservation and Community: Changing Narra tives, Policies and Pr actices in African Conservation. In African W ildlife and Livelihoods: the Promise and Performance of Community Conservation. David Hulme and Marshall Murphree, eds. Pp. 24-37. Oxford: James Currey. Brown, Irving F. 2006 Monitoring Fires in Southwestern Am azonia Rain Forests. Eos 87 (26): 253-264. Carvalho, Edivan, Ricardo Mello, Maria L. Souza and Leuzabeth Silva 2007 Tcnicas de Preveno de Fogo Aciden tal: Mtodo Bom Manejo de Fogo para reas de Agricultura Familiar. Belem: Ipam. Cochrane, Mark A. 2003 Fire Science for Rain forests. Science 412: 913-919. Cochrane, Mark A. 2001 In the Line of Fire: Understanding the Im pacts of Tropical Forest Fires. Environment 43 (8): 28-38. Cochrane, Mark. A. 2001 Synergistic Interactions be tween Habitat Fragmentation a nd Fire in Evergreen Tropical Forests. Conservation Biology 15(6): 1515-1521. Cochrane Mark A. and Mark D. Schulze 1999 Fire as a Recurrent Event in Tropical Fore sts of the Eastern Amazon: Effects on Forest Structure, Biomass, and Species Composition. Biotropica 31 (1): 2. Cochrane Mark. A. and Carlos. M. Souza Jr. 1998 Linear Mixture Model Classi fication of Burned Forests in the Eastern Amazon. Int. J. Remote Sensing 9 (17): 3433-3440.
117 Conklin Harold C. 1954 An Ethnoecological Approach to Shifti ng Agriculture. The New York Academy of Sciences 17 (2): 133-142. Cohen John M. and Norman T. Uphoff 1980 Participations Place in Rural Development: Seeking Clarity throug h Specificity. World Development 8:213-235. COOMFLONA 2009 O Projeto Amb URL: http://www.coomflona.com.br/ popup.php?pages=projetoambe&mais=mais (March 14, 2009). De Souza Jr, Afonso F. 2006 Dislexia Institucional: Fogo, Fumaa e Cinzas em Roraima. PhD dissertantion, Universidade de Braslia/Centro de Desenvolvimento Sustentvel, Braslia. Diaz, Maria del C., Daniel Nepstad, Mendona M.J ., Motta R.S. da, Ane A. Alencar, Gomes J., Ortiz R. 2002 O Preo Oculto do Fogo na Amaznia: Cust os Econmicos associados s Queimadas e Incndios Florestais. IPAM / WHRC / IPEA. Faria, Andreia A. da C 2003 Uma Viso da Floresta Nacional do Tapaj s: Informaes sobre o Diagnstico Rural Participativo (DRP) realizado durante os me ses de junho e setembro de 2003. Santarm: IBAMA/ProManejo. Ferreira Leandro V. and William F. Laurence 1997 Effects of Forest Fragmentantion on Mort ality and Damage of Selected Trees in Central Amazon. Conserva tion Biology 11 (3): 797-801. Ferreyra, Cecilia 2006 Practicality, Positionality, and Emancipa tion: Reflections on Participatory Action Research with a Watershed Partners hip. Syst Pract Act Res 19:577-598. Gautam, Ambika P. and Ganesh P. Shivakoti 2005 Conditions for Successful Local Collectiv e Action in Forestry: Some Evidence From the Hills of Nepal. Society and Natural Resources 18:153-171. Gerwing, Jeffrey. J. 2002 Degradation of Forests through Logging a nd Fire in the Eastern Brazilian Amazon. Forest Ecology and Management 157:131-141.
118 Guijt, Irene and Meera K Shah 1998 The Myth of Community: Gender Issues in Participatory Development. London: Intermediate Publications. Haugaasen Tor, Jos Barlow and Carlos A. Peres. 2003 Surface Wildfires in Central Amazonia: S hort-Term Impact in Forest Structure and Carbon Loss. Forest Ecology and Management 179: 321-331. Holdsworth Andrew R. and Christopher Uhl 1997 Fire in Amazonian Selectively Logged Rain forest and the Potential for Fire Reduction. Ecological Appli cations 7 (2): 713-725. Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renovveis (IBAMA) and Projeto de Apoio ao Manejo Florestal na Amaznia (ProManejo) 2005 A Transformao para Conservar Est em Nossas Mos: Plano de manejo Floresta Nacional do Tapajs. Santarm: IBAMA ProManejo. Ickowitz Amy 2006 Shifting Cultivation and Deforestation in Tropical Africa: Cr itical Reflexions. Development and change 37 (3): 599-626. Instituto Nacional de Pe squisas Espacais (INPE) 2008 Monitoramento da Cobertura Vegetal da Amaznia por Satlit e: Sistema PRODES, DETER, DEGRAD e queimadas 2007-2008. URL: http://www.obt.inpe.br/prodes/index.html (April 01, 2009). Kauffman, John B. 1991 Survival by Sprouting Following Fire in Tropical Forests of the Eastern Amazon. Biotropica 23 (3): 219-224. Lei federal N 9.985 2000 Sistema Nacional de Unidades de Conservao da Natureza (SNUC). Lei federal N 9.985 de 18 de julho de 2000 Lins, Alexandre C. 2005 Sistema de Monitoramento e Avaliao do Proteger. AMA/PPG7 Ministrio do Meio Ambiente (MMA) 2009 Subprograma Projetos Demonstrativos PDA. URL: http://www.mma.gov.br/sitio/index.php ?ido=conteudo.monta&idEstrutura=51 (February 04, 2009).
119 Mendona Mrio J. C., Maria del. C. V. Diaz, Daniel Nepstad, Ronaldo Seroa da M., Ane Alencar, Joo C. Gome s and Ramon A. Ortiz. 2004 The economic cost of the use of fire in the Amazon. Ecological Economics 49: 89105. Meggers, Betty J. 1994 Archeological Evidence For the Impact of Mega-Nilqo Events on Amazonia During the Past Two Millennia. Climatic Change 28: 321-338. Mulder Monique B and Petter Coppolillo 2005 Conservation: linking ecology, economic s, and culture. Princeton: Princeton University Press. Nepstad Daniel, Georgia Carvalho, Ana Cristina Barros, Ane Al encar, Joo Paulo Capobianco, Josh Bishop, Paulo Moutinho, Paul Lefebvre, Urbano Lopes Silva Jr. And Elaine Prins 2001 Road Paving, Fire Regime Feedbacks, a nd the Future of Amazon Forests. Forest Ecology and Management 154:395-407. Nepstad Daniel, Moreira, Adriana. G. and Alencar, Ane A. 1999 A Floresta em Chamas: Origens, Im pactos e Preveno de Fogo na Amaznia. Programa Piloto para a Proteo das Florestas Tropicais no Brasil. Braslia. Nepstad Daniel.C., Adalberto Verssimo, Ane Al encar, Carlos Nobre, Eirivelthon Lima, Paul Lefebvre, Peter Schlesinger, Christopher Potter, Paulo Moutinho, Elsa Mendonza, Mark Cochrane and Vanessa Brooks. 1999a. Large Scale impoverishment of Am azonian Forests by Logging and Fire. Nature 398: 505-508. Ostrom, Elinor 2005 Understanding Institutional Diversity. Princeton: Princeton University Press. 1990 Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press. Penner Joyce E., Robert E. Dickinson and Christine A. O'Neill 1992 Effects of Aerosol from Biomass Burning on the Global Radiation Budget. Science, New Series 256 (5062): 1432-1434. Perz, Stephen G. 2004 Are Agricultural Production and Forest Conservation Compatible? Agricultural Diversity,Agricultural Incomes and Primar y ForestCover Among Small Farm Colonists in the Amazon. World Development 32 (6): 957. Perz, Stephen G. and Robert T. Walker 2001 Household Life Cycles and Secondary Forest Cover Among Small Farm Colonists in the Amazon. World Development 30 (6): 1009.
120 Ramos, Paulo C. M. 1998 Polticas Pblicas que Afectan a los Ince ndios Forestales en el Brasil. Reunin Sobre Polticas Pblicas que Afectan a los Incendios Forestales FAO. URL: http://www.fao.org/docrep/ 003/X2095S/x2095s0x.htm#fn B38 (February 02, 2009). Ray David, Daniel Nepstad and Paulo Moutinho. 2005 Micrometeorological and Canopy Controls of Fire Susceptibility in Forested Amazon Landscape. Ecological Applications 15(5): 1664-1678. Sanford Robert L., Juan Saldarriaga, Kathleen E. Clark, Christopher Uhl and Rafael Herrera 1985 Amazon Rain-Forest Fire s. Science 227 (4682):53-55. Sauer Srgio., Luciana Miranda Costa, Soraya Fernandes Martins and Arthur Oscar Guimares 2004 Relatrio de Consultoria: Anlise de Experincias Inovadoras na Ireveno e Controle de Queimadas na Amazni a Brasileira. Braslia: PPG7/AMA Sauer, Srgio 2005 Preveno de Incndios Florestais na Amaznia: Lies Aprendidas no Projeto Proteger. Braslia: Srie Estudos 1/MMA. Secretaria de Estado de Cincia, Tecnologia e Meio Ambien te (SECTMA) AND Instituto de Meio Ambiente do Acre (IMAC) 2006 Meio Ambiente no Governo da Floresta: Relatrio de Atividades da SECTMA e do IMAC 1999-2002. SECTMA E IMAC. Silva, Leuzabeth A. 2001 O Fogo na Floresta Nacional do Tapajs Panorama Atual e Perspectivas Futuras. Belem: Ipam. Stone, Samantha S. 2003 From Tapping to Cutting Trees: Particip ation and Agency in Two Community-Based Timber Management Projects in Acre, Brazil. Schneider, Hartmut and Marie-Hlne Libercier 1995 Participatory Development: from Advo cacy to Action. Paris: Organisation for Economic Co-operation and Development Uhl, Christopher 1998 Perspectives on Wildfire in the Humid Tropics. Conservati on Biology 12 (5): 942-943. Uhl Christopher and John B. Kauffman 1990 Deforestation, Fire Susceptibility, and Pote ntial Tree Responses to Fire in the Eastern Amazon. Ecology 71 (2):437-449.
121 Vayda, Andrew P. 2006 Causal Explanation of Indonesian Forest Fires: Concepts, Applications and Research Priorities. Human Ecology 34: 615-635. Vosti Stephen A., Witcover and Chantal L. Carpentier 2002 Agricultural Intensification by Smallholders in the Western Brazilian Amazon. International Food Policy Research Institute. Wadsworth, Y. 1998 What is Participatory Action Research? Action Research International, Paper 2. Available on-line: http://www.scu.edu.au/schools/gcm/ar/ari/p-ywadsworth98.html Walker Robert. T. and Alfredo Homma 1996 Land Use and Land Cover Dynamics in the Brazilian Amazon: An Overview. Ecological Economics 18:67-80. Weigand Jr., Ronaldo 2003 The Social Context of Pa rticipation: Participatory Ru ral Appraisal (PRA) and the Creation of a Marine Protected Area in Bahi a, Brazil. PhD dissertation, University of Florida, Gainesville. White Sarah C. 1996 Depoliticising Development: The Uses and Abuses of Participation. Development in Practice 6 (1): 6-15. Wood, Charles H. 2002 Introduction: Land Use and Deforestation in the Amazon. In Deforestation and Land Use in the Amazon. Charles H. Wood and Roberto Porro, eds. Pp. 1-38. Florida: Gainesville. World Bank 2005 Implementation Completion and Results Report Emergency Fire Prevention and Control Project. URL: http://www.wds.worldbank.org/ex ternal/default/WDS ContentServer/ WDSP/IB/2005/07/27/000012009_20050727091223/Rendered/PDF/327870rev0.pdf (March 27, 2009).
122 BIOGRAPHICAL SKETCH Maria Lucim ar de Lima Souza was born in 1976, in Altamira, state of Par, Brazil. Daughter of farmers that moved from Northeast of Brazil to the Amazon during the colonization period, she always had an interest in the human aspects of environmental field. She attended the Federal University of Par where she graduate d as a Psychologist in 2003. While at Federal University of Par, she started working as a trainee at Amazon Institute for Environmental Research (IPAM) where she worked for 8 years w ith rural communities from different regions in the Amazon in projects related to environmental e ducation and partnerships for fire management. After 8 years, she decided to come back to school and rethink her experience as practitioner. This major decision on her life brought her to the Univ ersity of Florida where she was part of the Amazon Leadership Initiative Program, and where sh e pursued her Master of Arts in the Center for Latin American Studies. She pursued a minor in tropical conservation and cevelopment. She did her field work in 2007, in Santarm region, wi thin a protected area ca lled Tapajs National Forest. Her research was related to the effectiveness of institutional arrangements for fire management in the Amazon region.