The Bolsa Floresta1 program encompasses a set of integrated interventions aimed at conserving forests and improving the welfare of residents in selected sustainable development reserves (SDRs) of the state of Amazonas in the Brazilian Amazon. Only 1 of the 15 conservation units where Bolsa Floresta is working is a certified REDD+ initiative (Juma), while the other 14 are in the REDD+ readiness phase.

The program was started in 2007 and contains a financial compensation program, where a small economic incentive of BRL 50 per month (USD 30)2 is paid to households for their commitment to zero net deforestation. Moreover, the program includes a set of integrated conservation and development components implemented at the community or association level and aimed at improving livelihood opportunities and thereby preparing communities to meet oncoming deforestation pressures.

3.1 Basic facts: Where, who, why and when

3.1.1 Geography

The state of Amazonas is the largest of 10 federal states in the Brazilian Legal Amazon and still has 96% of its forests intact (INPE/PRODES 2014). In recent decades, pressures have expanded from the ‘Arc of Deforestation’ to new frontiers, including the southeastern border of Amazonas state (May et al. 2011). However, with drastic reductions in Brazilian deforestation since 2004, deforestation pressures in Amazonas have also diminished.

The Bolsa Floresta program currently works in 15 state conservation units – or protected areas – throughout Amazonas state and covers over 10 million ha (FAS 2014). Protected areas have been shown to be effective barriers to deforestation in the Amazon. This includes human-occupied protected areas, such as indigenous territories, extractive reserves and SDRs (Nepstad et al. 2006, 2014; Viana et al. 2009). Thirty-seven percent of the reduction in the Brazilian Amazon’s total deforestation between 2004 and 2006 has been attributed to the expansion of protected areas (Soares-Filho et al. 2010). The addition of Bolsa Floresta has been questioned because the program has been implemented in protected areas. For our study, we selected two SDRs that face high deforestation risk as projected in deforestation scenarios, in order to assess conditions where there is potential for good outcomes from REDD+.

In this chapter, we focus on the implementation of Bolsa Floresta in two SDRs in Amazonas: Juma (589,612 ha) and Uatumã (424,430 ha; see Figure 3.1), both of which face relatively high internal and external deforestation pressures compared with other SDRs, as shown in a recent study of Bolsa Floresta by Börner et al. (2013). SDR Juma was created in 2006 and is located in the Municipality of Novo Aripuanã in the southeastern region of Amazonas. It lies approximately 200 km south of Manaus, the capital of Amazonas, and 100 km north of the regional market town of Apuí. The western boundary of the reserve is defined by the Mariepauá River, which forms the border with Manicoré Municipality. The southern boundary is defined by federal land (100 km north of the Transamazon Highway, BR-230). The Acari and Madeira Rivers limit the reserve in the east and north, respectively (SDS 2007).

Juma can be accessed by the Aripuanã, Mariepauá and Madeira Rivers, and by road from Novo Aripuanã (AM-174). According to early simulation models, plans to pave AM-174 and BR-319 further south would threaten to escalate deforestation, with up to 62% of forest cover potentially lost by 2050 (based on the SimAmazonia I model by Soares-Filho et al. [2006]), although other models that consider different inputs have indicated a more modest forest-cover loss (19% by 2050) (Yanai et al. 2012).

The SDR Uatumã, created in 2004, is 337 km east of Manaus by road (SDS et al. 2009). Projections simulate that 70% of the reserve’s forest cover could be lost by 2050 (Soares-Filho et al. 2006; SDS et al. 2009). The northern shore of the Uatumã River and upland is part of the Municipality of São Sebastião do Uatumã, where the southern side belongs to Itapiranga Municipality. The reservoir for the Balbina hydroelectric plant is located on the western side of the reserve.

Industrial logging has taken place historically in forest concessions on the southern side of the reserve on lots both inside and outside the Uatumã reserve, which has led to the illegal timber exploitation of commercial hardwoods. These illegal practices are currently being converted to sustainable forest management operations, with support from a local NGO, The Institute for Conservation and Sustainable Development of the Amazon (IDESAM).

Current access to Uatumã reserve is by road or river from Manaus. There is a regular boat service from Manaus to Itapiranga or São Sebastião do Uatumã. From the north, there is also a boat service, and from the south, the state roads AM-010 and AM-363 connect the reserve to Itapiranga. There are better connections to the west via the federal road BR-174, which connects to Balbina and then continues via AM-240 from Balbina to Ramal da Morena. These transport connections west of the reserve have allowed better access to markets and thus for the development of more commercially oriented agriculture and production of perishable yet valuable agricultural products (such as watermelons). Table 3.1 gives a description of the biophysical characteristics of the study areas.

SDRs were created to conserve nature and biodiversity, while ensuring sustainable resource use and the development of local communities’ livelihoods (Medeiros 2006). Both Juma and Uatumã have been divided into strict preservation, intensive-use and extensive-use zones, which increase in the degree of allowed human intervention and land use. Main livelihood activities inside the reserves include smallholder agriculture (cassava as the mainstay, and banana and watermelon), fishing and harvest of forest products including fruits, nuts, bushmeat and timber – mainly for subsistence, but sometimes also for trade. Small livestock (chickens and pigs) are common, but only a few households have cattle (SDS et al. 2009, 2010).

According to SDR rules, households are allowed to deforest up to 20% of their land but land sales are strictly prohibited. Management plans specific to each reserve stipulate the exact area of forest that resident households may clear each year. In Juma and Uatumã, forest clearing is limited to 3–4 ha/year, preferably of secondary forest (with some exceptions for clearing primary forest) (SDS et al. 2009, 2010). The state is the de jure land owner, but concessão de direito real de uso (CDRUs or concessions of real right of use)3 are issued individually to households or to community associations (Carvalheiro et al. 2010). However, over half the resident families in the Uatumã reserve held no CDRU documentation (SDS et al. 2009). Table 3.2 compares the rules of the Bolsa Floresta contract with the specific SDR management rules for Juma and Uatumã. Bolsa Floresta imposes conservation rules going slightly beyond the reserve management plans in terms of stabilizing agricultural production areas and including social requirements (children’s education, association membership).

Both reserves house relatively small populations. We sampled approximately 122 households inside Uatumã SDR, and 70 households inside Juma SDR. An additional 47 households were sampled outside of Juma SDR, as they also received benefits from Bolsa Floresta. However, for the purposes of comparing benefits at these sites to REDD+ benefits in Juma, only the households inside the carbon accounting area were included. Both reserves have had low deforestation, and households survive mainly on subsistence agriculture of cassava, and production of a cassava derivative, farinha, for later sale (Table 3.3).

3.1.2 Stakeholders and funding

Fundação Amazonas Sustentável (Amazon Sustainable Foundation; FAS) is a private Brazilian nongovernmental foundation based in Manaus, Amazonas. In a partnership between the Amazonas State Government and Bradesco Bank, FAS was created in 2007 to implement environmental conservation activities through sustainable livelihood development programs in state conservation units.

Bolsa Floresta is the flagship program of FAS, and is implemented with assistance from affiliated NGOs, government bodies and other private actors. IDESAM supports Bolsa Floresta in aspects of monitoring, natural resource management and community organization in SDR Uatumã through its program on supporting protected area management. It also contributed to the technical coordination of baseline methodology and monitoring in Juma SDR and development of the PDD (FAS et al. 2008). Various state agencies are also tied to the management of state protected areas. The Amazonas State Secretariat of the Environment and Sustainable Development (SDS) coordinates the State Center for Protected Areas, which is responsible for the administration and management of protected areas in Amazonas. The State Center on Climate Change and the State Secretariat of Planning and Economic Development provide assistance in reserve management. Amazonas State Institute for Environmental Protection carries out monitoring and law enforcement in the protected areas. The National Institute for Amazon Research is also responsible for carbon monitoring. Several private organizations offer pro bono assistance in initiative management and auditing, including PricewaterhouseCoopers and Bain & Company, among others.

The Bolsa Floresta program is funded mainly by Bradesco Bank and the Amazon Fund (The Brazilian National Development Bank BNDES/Government of Norway). Almost 80% of FAS funding is from private sources, including Coca-Cola, Samsung, Abril Media Group and Marriott International, among others. In addition, the Juma REDD+ initiative has been co-funded by Marriott International and Abril Media Group (personal communication from V Salviati, 8 October 2014).

3.1.3 Motivation

Brazil has had the highest deforestation in the world with a 10-year average of 19,500 km2/year between 1996 and 2005. Annual forest loss then gradually reduced from the peak in 2004 until 2012 by 70% in the Legal Amazon, although deforestation increased slightly from 2011 in the state of Amazonas. This large overall decrease in deforestation was mainly through changes in national protection and enforcement policies, as well as supply-chain interventions that aimed at some of the major drivers of historic deforestation (Nepstad et al. 2014).

The Bolsa Floresta program was born out of a series of initiatives begun in 2003 by SDS, entitled Zona Franca Verde. These initiatives aim to promote the sustainable use of natural resources in the state with a view to increasing the value of the environmental benefits provided by forests (Viana 2008).

Protected sustainable-use areas became the target sites for the Bolsa Floresta interventions to equip local forest dwellers with sustainable alternatives in the face of oncoming deforestation pressures. Following dire deforestation forecasts, the Juma SDR was established in 2006 and became the site of the first test REDD+ initiative in the Amazon and the first to be implemented under the new State Policy on Climate Change Law (PEMC; Law 3.135/2007) and the State System of Protected Areas (SEUC; 53/2007), which both provided the legal framework necessary to implement REDD+ initiatives in Amazonas state (Viana et al. 2008).

3.1.4 Timeline

The start date for the Bolsa Floresta program is essentially the time when sustainable development policies were being created at the state level to give value to the environment and its services. From that point on, the enabling policy environment assisted the creation and implementation of the Bolsa Floresta program and initiation of the first REDD+ initiative in Juma SDR (Figure 3.2).

Juma was also independently validated to ‘gold status’ according to CCBA Standards in 2008 and became the first certified REDD+ program in the Amazon. Over its first crediting period in 2006–2016, the program is expected to prevent an estimated 3.6 million tCO2e, and 190 million tCO2e by the initiative end in 2050 (Viana et al. 2009). To access a wider voluntary carbon market, plans are already underway to generate verifiable emissions reductions (VERs) using the VCS, which started with the approval of a carbon accounting methodology in 2010 (WWF 2009; FAS 2012).

3.2 Strategy for the initiative

The major objectives of the Bolsa Floresta program are forest conservation and improvements of local people’s livelihoods. FAS sees these two goals as inherently interrelated. The underlying logic closely resembles that of ICDPs: change the productive logic locally so as to increase sustainability, add value to the forest and divert production factors from environmental degradation. At the same time, productive investments and income transfers are seen as tools to make smallholder SDR residents into potential conservation allies against predominantly external deforestation threats, thereby reducing needs-based deforestation. Avoiding outmigration, and thus maintaining high smallholder populations, is seen as a plus in this respect. Nevertheless, beyond the ICDP logic, Bolsa Floresta has also integrated a minor (in terms of cost), but much debated PES component (see Section 3.5).

Participation in the Bolsa Floresta program is voluntary. Nevertheless, some components, e.g. education and health facilities, are delivered at the community or reserve level, and therefore equally benefit nonparticipants. To avoid ‘magnet effects,’ a prerequisite for receiving PES is residence in the reserves for at least two years. In practice, almost all households that are entitled to participate in Bolsa Floresta do.4

The program is comprised of four interrelated components:

1. Bolsa Floresta Renda (income component)

Communities in SDRs receive annual equivalent productive investments of BRL 140,0005 (USD 84,916) per SDR, to support income-generating activities that are in line with the protected area’s management plan. Examples are on-farm processing activities for value-added of existing products, NTFP value chains, or for alternative income sources such as ecotourism, aquaculture, small-livestock breeding and natural honey production.

2. Bolsa Floresta Social (social component)

An additional total of BRL 140,000 (USD 84,916) per SDR is allocated to improve education, sanitation and health services, as well as communication and transportation infrastructure in the reserves. This is done in collaboration with the respective public sector institutions.

3. Bolsa Floresta Associação (association component)

Targeting reserve-level collective action, this component supports the associations of reserve dwellers. It corresponds to 10% of all the family forest allowances granted (see item 4. Bolsa Floresta Familiar), and associations may decide freely on how to allocate these funds for the benefit of their members.

4. Bolsa Floresta Familiar (family component)

This component is a monthly transfer of BRL 50 (USD 30) to the female spouse at the household level, subject to a signed agreement that commits the household to good forest management practices, including zero net deforestation (see Table 3.2). In itself, it is not designed to be a source of income for families, but rather a reward for forest conservation (Viana et al. 2008).

3.3 Smallholders in the initiative

Table 3.4 summarizes the main socioeconomic characteristics of the reserves. Average household total income is similar for Uatumã and Juma, as is average household forest income. However, Uatumã residents have a higher average household value of transport assets and higher average household value of livestock in this reserve, reflecting a small concentration of nonrecipients of Bolsa Floresta engaging in livestock farming, as well as a generally better market integration as a result of proximity to Manaus.

Generally, Uatumã communities have more access to piped water or artesian wells (31%) and have better sanitation (97%) than do communities in Juma. Nevertheless, Juma households have more access to electricity (89%) and control, on average, larger areas of land.

This difference in socioeconomic characteristics between the two reserves is best reflected in the distribution of income sources per household (Figure 3.3). While forest shares are quite similar, Uatumã households have a greater share from wage labor, which may also reflect the better connection to regional markets around Uatumã. Juma households have a greater share from agricultural sources, where 74% of household members 16 years or older report agriculture as their primary or secondary occupation (compared with 56% in Uatumã; Table 3.5).

In cash terms, the average household income in the two study areas was found by Börner et al. (2013) to be close to Brazil’s ‘extreme poverty’ line (monthly BRL 70 or USD 42 per capita). However, households are shown to be markedly better off when subsistence sources are included (as seen in Figure 3.3, which gives average household income). High subsistence incomes come from swidden cassava cropping, and forest and fish extraction. Indeed, forest and fish extraction, making up 29% of income in both reserves, was generally found to be as important an income source as agriculture, even across income quintiles (Börner et al. 2013). In the women’s meetings, participants reported that men go to the forest weekly, if not daily (81–100% frequency), throughout the year to hunt and collect wild products. Throughout the year, women occasionally go to the forest (0–20%) to collect wild products or traditional medicines, but go daily during the dry season to make charcoal and hunt alongside the men. One community reported daily visits only during the wet season (21–40%) for wild food and fuelwood collection. The sparse population in the SDRs benefits from protected access to land and natural resources of high quality and value, which clearly makes the difference that lifts reserve dwellers above the levels of extreme rural poverty.

Smallholders in the reserves are not the primary threat to forest conservation, as clearing in the forest buffer has been found to be several times higher (Reimer et al. 2012). Current livelihood strategies are largely focused on forest extraction and small-scale clearing for agriculture and are therefore still quite harmonious with SDR and Bolsa Floresta regulations (Table 3.5) (Börner et al. 2013). Indeed, several households in the reserves were reported to have left up to 30 ha fallow over the last three years, which could be a result of conservation efforts under Bolsa Floresta. Also, the availability of forestland for clearing has decreased for almost two-thirds of respondents from Uatumã and Juma. Sixty percent of respondents attributed this decrease to the creation of the reserve or related reasons (e.g. sanctions).

3.4 Challenges facing the initiative

The vastness, expanse and seasonality in these areas make it difficult for the proponent to maintain a technical and institutional presence at a reasonable cost. According to the proponent, transaction costs including start-up fees were also initially a huge challenge for the organization, reducing their cost efficiency; they have been limited to an average of 25% of the overall budget for the program in all conservation units (personal communication from J Tezza, 1 May 2013).

The large geographical scale of the initiative also poses a challenge for monitoring and enforcement. Remote sensing cannot detect the relatively small changes in land cover that are characteristic of the small-scale forest clearing in this region (Börner et al. 2013). In BBC coverage of the Juma REDD+ initiative in 2009, Amazonas State Governor Eduardo Braga was reported as saying that although policing infrastructure should be improved, surveillance needs to be delegated to the people in the region. In fact, Bolsa Floresta’s implementers have placed much emphasis on this – namely, building the capacity of local communities and reserve associations to assist in monitoring and to take responsibility for the program, as well as developing a monitoring methodology that can detect small-scale clearing of 0.25 ha. More than 180 families to date have been excluded from the program for reasons including relocation or violating the rules of the contract, and the current process of enforcement and sanctioning relies on an external FAS monitor to evaluate the infringement (personal communication from V Salviati, 8 October 2014). The remoteness of communities has implications for compliance, detection and enforcement, both now and in the future, making it critical to invest in the capacity of the communities to ensure effective reductions in deforestation and degradation.

Another challenge has been the focus on Bolsa Floresta’s family component, which was the first direct cash payment for avoided deforestation to be implemented in the Amazon (see also Section 3.5). Currently, the financial reward (family component) is very minimal, and according to the proponent, it is not meant to be permanent, nor necessarily intended as compensation for the opportunity cost of keeping forest standing. Rather, it represents a conditional reward in return for good forest stewardship. However, when asked about their hopes for the program, the majority of respondents mentioned the need to increase the value of this payment (62% and 77% for Uatumã and Juma, respectively). Similarly, half of the main worries (49% and 50% in Uatumã and Juma, respectively) were that Bolsa Floresta would lead to a decrease in household income through changes in livelihood practices, or that it would not properly compensate the household for the loss in forest income. However, the proponent argues that increasing the value of the payment to match opportunity costs would only compensate for current income, which is very low relative to the national average (personal communication from J Tezza, 1 May 2013). Thus, such a complementary mix of integrated development components targeted at health, education and sustainable livelihoods could do more than cash payments to improve lives in the communities.

3.5 Lessons from the initiative

The provision of a small economic reward, or PES, makes this initiative a pioneer in direct cash transfers for avoided deforestation in the Brazilian Amazon.6 Past studies have shown that conditional cash transfer programs, defined as any program that requires a specific course of action to receive a benefit (cash or in-kind; after Das et al. 2005), can improve household well-being through impacts on health, purchasing power, household productivity and resource allocation, asset consolidation, and reduction of inequality (Hanlon et al. 2010; Arnold et al. 2011; Barrientos 2012). Regular and reliable transfers can enhance investment into productive assets (e.g. livestock and agricultural production in Paraguay; Soares et al. 2010), facilitate coping with shocks and improve labor market access (Arnold et al. 2011). Only a few studies (e.g. Alix-Garcia et al. 2012, 2013), however, have dealt with potential environmental and socioeconomic spillover effects from PES, i.e. effects that go beyond the implied conditionality rules on how recipients spend their additional money, and what feedback effect this has on development paths and the environment.

To analyze these effects, Bakkegaard et al. (2013) looked at the effect of cash transfers on conservation behavior, or so-called ‘conservation spillovers,’ in the sample group, since governmental transfers were the main source of cash income for these households. Results showed that transfers did not have negative spillovers on land use, but rather led to (positive) marginal reductions in the area of land under crop and early fallow. Using a Gini decomposition, they found (unsurprisingly) that transfers were income equalizing, while BAU strategies (i.e. agriculture) were unequalizing. Thus, conditionalities attached to conservation transfers could potentially curb these unequalizing effects. Thus, for the time being, these transfers have positive conservation and equality effects and provide a welcome cash injection, in addition to the in-kind benefits from other Bolsa Floresta components. Figure 3.4 shows that the cash payments mainly supporting current consumption, such as the purchase of food, clothes and other basic goods, are similar to how cash transfers are used elsewhere (Hanlon et al. 2010; Arnold et al. 2011).

The use of cash transfers to promote forest stewardship could be considered as culturally appropriate for Brazil. Eighteen million households in Brazil already benefit from cash transfers, such as the retirement pension and Bolsa Família (family grant) – one of the largest conditional cash transfer programs in the developing world that provides transfers based on educational and health conditionalities to poor families (Medeiros et al. 2008; Bunting 2010). Combined, Brazil’s cash transfers have helped in reducing the number of people in poverty from 28% in the decade preceding their introduction in 2001, to 17% in 2008 (Ferreira et al. 2009; Hanlon et al. 2010).

Bolsa Floresta’s success is largely dependent on contextual factors, including: low deforestation pressure in target areas; homogeneous, subsistence-oriented resident populations; and program rules that are very similar to preexisting conservation rules in the protected areas (Börner et al. 2013). The replication of low-value, uniform per-household cash transfers such as Bolsa Floresta’s family allowance may be challenging in different settings. That said, the Bolsa Floresta approach is an innovative model and offers an important example of implementing REDD+ on the ground.

3.6 Acknowledgments

We are first and foremost grateful to the several hundred people in and around SDR Juma and SDR Uatumã, who let our team of dedicated enumerators and assistants into their homes with our seemingly endless questionnaires. We thank the enumerator team Denise Reis dos Nascimento, Simone Santos, Mario Acuna, Ivan Zarros and Patricia Gallo, research assistant Kaline Rossi do Nascimento, and boat crews. We also thank Fundacão Amazonas Sustentável (FAS) and The Institute for the Conservation and Sustainable Development of Amazonas (IDESAM) for their field support.

We would like to acknowledge the financial support that facilitated this research, including that by CIFOR-GCS, NEWFOREX, Skovridder Mindefond, Torben and Alice Frimodts Fond, WWF REDD initiative, LIFE Copenhagen University, Stiftelsen Løvstrupgaard, Jordbrugs Akademiker (JA), Landbrugets Studiefond, Agricultural Development Grant, DLH, Oticon, PLAN DK and Danida.