INTRODUCTION
Despite the inadequacy in the implementation of democratic decentralization in the area of natural resources governance, the efficient management and sustainable use of environmental resources has become a question of survival in most parts of Africa in the face of mounting population pressures. Those who stand to benefit directly from the sustainable management of environmental components like forest resources can best manage the environment (Nelson & Agrawal Reference Nelson and Agrawal2008). It is virtually impossible to manage the environment from the centre without the active support and participation of local communities, and without assuring them of direct benefits and costs from sustainable management of environmental resources. In fact, good practices are already emerging in the local management of the environment and local resources, such as community-based natural resource management (CBNRM) initiatives throughout southern Africa (Godana & Naimhwaka Reference Godana and Naimhwaka2002; Mukamuri et al. Reference Mukamuri, Manjengwa and Anstey2009). Thus, democratic decentralization is a promising means of institutionalizing and scaling up the popular participation that makes participatory management and CBNRM effective. CBNRM increases environmental management efficiency and improves equity and justice for local people who stand to lose or gain from the mismanagement or optimal use of natural resources (Ribot Reference Ribot, German, Karsenty and Tiani2009). Emerging experience from some African countries, suggests that an approach that puts communities at the centre of the development process is more likely to succeed than the centralized bureaucratic approach that has characterized previous rural development initiatives (Parker Reference Parker1997; Nelson & Agrawal Reference Nelson and Agrawal2008; German et al. Reference German, Karsenty and Tiani2009).
Almost all developing countries are undertaking decentralization reforms (Ribot Reference Ribot2004; Hobley Reference Hobley2007) and a growing body of literature focuses on African experiences with CBNRM (Nelson & Agrawal Reference Nelson and Agrawal2008; German et al. Reference German, Karsenty and Tiani2009; Mukamuri et al. Reference Mukamuri, Manjengwa and Anstey2009). Motives for decentralization vary greatly, although most donors and governments justify decentralization as a means for increasing access, use, management and voicing of claims and concerns about natural resources among communities. Decentralization reforms change the institutional arrangements for local natural resource management and may provide an institutional basis for more popular and participatory management and use of natural and other public resources (Ribot Reference Ribot2002; Reference Ribot, German, Karsenty and Tiani2009). Ribot (Reference Ribot, German, Karsenty and Tiani2009) argued that there was limited scholarship on the relationship between decentralization in forestry and the establishment of democracy at the local level. Nonetheless the overall macro-political contexts for each country are given to situate the institutional arrangements for forest management at the local case level. This paper provides evidence of what has and what has not taken place in forestry decentralization and the impediments to it within some countries.
Five issues are at the intersection between this article and broader discussions around decentralized forest management. Firstly, forest tenure is at the centre of discussion on property rights and relations. All access to forests and forest products is dependent on the rights and mechanisms that communities are allocated by the state, in turn dependent on the history of land alienation (Ribot & Peluso Reference Ribot and Peluso2003). Within common property theory, the argument most consistently made is that securing rights results in improved management by resource users. Forests in Africa are often marked by absence of secure rights for communities, such that decentralization is touted as offering an avenue improving tenure over forests and forest products (German Reference German, Karsenty and Tiani2009). Secondly, organizational structures for resource management at the local level may vary depending on the level of state decentralization of forest management. Without transfer of management authority to suitable local institutions, central governments' transfer of responsibilities may remain inadequate to generate management success (Ribot Reference Ribot, German, Karsenty and Tiani2009).
Thirdly, one of the principal barriers to effective resource management at the local level is the lack of downwardly accountable institutions (Shackleton et al. Reference Shackleton, Campbell, Wollenberg and Edmunds2002; Ribot Reference Ribot, German, Karsenty and Tiani2009). While there are links to the broader political-administrative processes in which power is transferred to institutions or leaders that are accountable to their constituents for decentralization to be effective and democratic, in forestry decentralization, the major setback has been the transfer of authority to institutions that are largely accountable to the state or central government. To whom elected officials are accountable is central to effective decentralization (Manor Reference Manor1999; Ribot Reference Ribot2002). Fourthly, linked to accountability, the extent to which decision-making power is transferred to communities or lower levels to enable forest or resource management. The various forms of decentralization that Manor (Reference Manor1999, p. 6) defined were in the realm of ‘democratic decentralization’. This form of decentralization, also referred to as ‘devolution’, may involve the ‘transfer of resources, power and tasks to lower level authorities that are largely or wholly independent of higher levels and democratic in some respects (Manor Reference Manor1999). Such devolution has a higher probability of failure if lower level authorities lack both financial and administrative ‘powers and resources’ (Manor Reference Manor1999, p. 7). Yet another feature of decentralization involves the level to which power and resources are transferred, and it is important to distinguish between intermediate (region or district) and local levels. Each of the levels at which power is transferred has implications for resource governance in so far as rights enjoyed by local resource users are concerned. Democratization processes are a common feature of the Southern African region. These processes are still ongoing, although they have attained different features across the region (Lalá Reference Lalá2004). This framework of democratic processes has affected the management of natural resources in the developing world through the content and scope of policies that shape their management and use.
Finally, livelihoods are linked to decentralization, in so far as they provide benefits that accrue to communities, households or groups at community level. Benefits may be in the form of incomes from stumpage fees or from forest products, or in terms of the safety net function of forests (namely that forest goods and services assist poor and vulnerable households in times of shocks and stresses by reducing risk and improving their livelihood strategies in the form of food, medicines, materials for shelter and income sources). In this paper, owing to limitations in available data, we were unable to provide a more nuanced analysis of livelihoods to follow through the emerging scholarship on decentralization and impact on livelihoods (Shackleton et al. Reference Shackleton, Campbell, Wollenberg and Edmunds2002; Hobley Reference Hobley2007; Tacconi Reference Tacconi2007).
We here consider what outcomes decentralization and experiments in CBNRM have for improving livelihoods of rural people living within and around forests in Southern Africa. We review experiences in giving authority and control over resources to communities in relation to their contributions to rural livelihoods in three case studies, one each from Zimbabwe, Mozambique and South Africa. We identify common features in how each of the cases’ practices over decentralization have progressed or failed to change state-centred management and improve communities' livelihoods.
METHODS
We reviewed secondary information from our earlier studies, complemented by other relevant work. We selected the Zimbabwe case on the basis of a long history of joint management between the state and different groups of local people through various phases of experimentation. Between 1994 and 1999, the Forestry Commission of Zimbabwe was supported by external funding bodies to undertake a pilot resource-sharing scheme that would provide lessons for other state forests on devolved resource management. From 2000 to 2003, the Centre for International Forestry Research (CIFOR) initiated the Adaptive and Collaborative Forest Management Project to enhance the Zimbabwean Forestry Commission's learning-by-doing experiments involving resource user groups at a lower administrative level than the resource management committees (RMCs). RMCs were formed to represent communities and take responsibility for routine management of allotted spaces within the protected forest.
In Mozambique, we selected the Chipanje Chetu case study for reasons identified by Anstey (Reference Anstey, Dzingirai and Breen2005, p. 141). The majority of CBNRM projects in Mozambique were aimed at specifically empowering local communities, but, as well as specific objectives of land transfers, the Chipanje Chetu project also incorporated natural resource tenure and promotion of local democracy (Salomão & Matose Reference Salomão, Matose, Dewees and Campbell2008). The Chipange Chetu project was created in 1998, in Matchedje and Macaloge (North Sanga District, Niassa Province, Mozambique) occupying an area of 3500–4000 km2 and with the aim of eliminating furtive hunting, wood smuggling to Tanzania and the use of poison to catch fish. The project also aimed to ensure the transfer of authority and power over natural resources to the local community level in order to promote local economic development.
Finally, we also selected the Tsitsikamma Indigenous Forest case study as the most recent experience on the impact of participatory forest management policies in South Africa. Most experiences of decentralization in South Africa tend to be centred on forest plantations. This case study developed against a backdrop of policy changes within the forestry sector to facilitate participatory forest management (PFM) after the democratization of political governance in South Africa post-1994. Previously, the Tsitsikamma Forest was under the management of the Department of Water Affairs and Forestry (DWAF), and then South African National Parks (SANParks), a parastatal body responsible for nature conservation and management. Scotney Watts collected data for the Tsitsikamma case from secondary materials collected during several studies between 2003 and 2006, as well as a review of available government literature.
DECENTRALIZATION CONTEXTS
The CBNRM concept popularized in the 1970s has strong connections with Zimbabwe, through the development of the Communal Areas Management Programme for Indigenous Resources (CAMPFIRE). The perceived success of the CAMPFIRE initiative led to the development of similar programmes in other eastern and southern African countries (Nelson & Agrawal Reference Nelson and Agrawal2008; DeGeorges & Reilly Reference Degeorges and Reilly2009; Mukamuri et al. Reference Mukamuri, Manjengwa and Anstey2009). Local resource-user committees grew in the process and the CBNRM initiative enshrined in the CAMPIRE programme for devolving the management and revenue from safari hunting and ecotourism became internationally renowned and generously funded. The same concept was expanded to the water sector, whereby the Water Act of 1998 introduced a catchment-based management system premised on the concept of integrated water resource management and devolved responsibility for management decision-making. This is in contrast to the largely commercial farming-dominated River Boards that represented only the water-user elites. However, the emergence of the Movement for Democratic Change (MDC) as a credible challenger to the Zimbabwe African National Union Patriotic Front (ZANU-PF)'s authority has led to the politicization of decentralized institutions and activities for them to function or survive (Sustainable Livelihoods in Southern Africa [SLSA] Team 2008). The ongoing land reform initiatives occasioned by chaotic farm invasions reduced gross earnings (previously US$ 20 million per annum) from devolved natural resources management programmes in the country after 2000 (DeGeorges & Reilly Reference Degeorges and Reilly2009).
Mozambique has an evolving policy framework to enable effective participation of local communities in the management of natural resources. This has created opportunities for rural communities to manage and benefit from forest and wildlife resources (Anstey Reference Anstey, Dzingirai and Breen2005). The Mozambican policy for wildlife and forest management requires the involvement of people who are dependent on forest and wildlife resources in the planning and sustainable use of resources (Salomão Reference Salomão2004). This is amplified by the Forest and Wildlife Law that recommends integrated management of natural resources and effective participation of local communities, associations and the private sector (Governo de Moçambique Reference de Moçambique2004; Serra Reference Serra2004). However, the requirement for participatory democracy in the management of natural resources does not derive from the country's supreme law or constitution that informs all sectoral policies and legislation. It is on this basis that decentralization of natural resources governance in Mozambique has been categorized as sector-specific and hence problematic. The lack of integration of natural resources management policies with the broader debates about democratic local governance has increasingly made local and central state authorities suspicious of decentralized management of natural resources. A further disadvantage of sector-specific decentralization is that it hardly fits the reality of rural livelihoods, which are not confined to distinct institutional arrangements dealing with different and specific resources (SLSA Team 2008).
Consequently, the role of local communities in forest resource use and management in Mozambique is considered passive without appropriate institutional arrangements for benefit-sharing. The harvest of forest produce by private concessionaires without the active involvement of respective communities, has been observed as a major challenge for CBNRM in Mozambique. Permits for logging community forests are issued by the provincial authorities where revenues from forest exploitation are retained (Gilmour Reference Gilmour2000). Community participation in the management of natural resources in Mozambique is treated cursorily if at all. Long-term participation of rural Mozambicans as stakeholders in natural resource conservation is still viewed in the context of creation of low-paid employment and other superficial benefits that do not reflect genuine participation or CBNRM (Tanner Reference Tanner2001).
In South Africa, the Constitution is the supreme law in the land, covering all economic decisions and activities, including access to environmental resources. The Bill of Rights in the second chapter of the Constitution promulgates rights to equality and property, and protects a healthy environment against unfair discrimination, amongst other things. Section 24 of the Constitution grants rights to environmental security for every person, including people's well-being and rights to participate and enjoy the benefits of a healthy and well-protected environment. This section requires the protection of the environment through reasonable legislation and other measures (Government of South Africa 1996). This constitutional directive has been implemented through the National Environmental Management Act (NEMA) of 1998. This law covers the different aspects of the environment and provides mechanisms for people to participate in decisions and activities that safeguard a healthy environment upon which the economy depends (Government of South Africa 1998a).
Section 2 of NEMA outlines the principles of environmental management and provides the legal basis for community involvement in conservation (Government of South Africa 1998a). The White Paper for Sustainable Forestry Development in South Africa (DWAF 1997a) emphasizes the participation of forest and woodland-dependent communities in the management of state forests and woodlands. The need for inclusion of local communities is explicitly articulated in the National Forestry Action Programme (NFAP) and the National Forests Act (NFA) No 84 of 1998 (DWAF 1997b; Government of South Africa 1998b). The NFA requires forests to be developed and managed in ways that conserve biodiversity, ecosystems and habitats, while simultaneously sustaining the supply of socioeconomic and environmental benefits. These benefits should be realized through collaborative management between the DWAF and local people (Government of South Africa 1998b). The NFA authorizes active community participation in the management of forests, by people who live within or at the margins of natural forests. It is unlikely that equitable distribution of benefits with local people can occur if they do not influence the institutions that distribute benefits among stakeholders. Similarly, marginalized people cannot emerge out of the vicious cycle of poverty when they do not actively partake in negotiating for their rights and benefits with other stakeholders, especially the state forestry department.
The concerns expressed in the White Paper for Sustainable Forestry Development, NFAP and NFA for active involvement of local people in forest and woodland management caused DWAF to develop a participatory forest management (PFM) strategy (DWAF 2004) This resulted in the further development of principles for PFM in state forests. These principles require that indigenous forest management should be people-centred, participatory and holistic, transparent and result in equitable allocation of benefits, capacity building and use of indigenous knowledge. The principles also emphasize the need to establish interdependent partnerships among forest user groups and reinforce participatory management as a management style in state forests (Watts Reference Watts2003). However, local communities in South Africa are not always involved in the hands-on management of forest resources, especially in the Southern Cape Forests, including Tsitsikamma (Holmes-Watts & Watts Reference Holmes-Watts and Watts2008).
CASE STUDIES
Mafungautsi (Zimbabwe)
A resource-sharing project around Mafungautsi Forest (a state forest, reserved for protection) initiated in 1993–1994 by the Forestry Commission of Zimbabwe (FC) progressed rapidly in the first year, but progress slowed thereafter. Progressive development in the initial stages led to the formation of RMCs and training them in their roles and responsibilities. In practice, memoranda of agreement were entered into between the different RMCs and the FC to facilitate collaboration in the management of the forest. The agreements moved beyond the provisions of the Forest Act by giving authority to the RMCs as community ‘representatives’. Under the agreements, each RMC is responsible for an area close to its village, managing and controlling resource access. The agreements cover minor forest products such as deadwood, mushrooms, fruits, edible caterpillars, thatch and broom grass. The RMCs operate under constitutions that authenticate them as recognized bodies/organizations within local governance structures. However, these agreements clearly state that the ownership of the forest remains with the State and the FC still continues to be the legal custodian of the forest (Matose Reference Matose2006). Therefore ownership and use rights remain the prerogative of the FC and could be revoked if it felt there was a general decline in the status of protected forest resource (Matose Reference Matose2006). The memoranda have aspects that pose challenges to the development of meaningful partnership arrangements between the state and local people. The challenges arise from the limitations of the ‘partnership’, whereby local people have little participation in the resource-sharing scheme, resulting in their continuing to pursue their access regimes as if the scheme did not exist.
The creation of the RMCs was the FC's way of implementing the concepts of ownership and participation towards CBNRM. However, the resource-sharing model falls short of securing tenure for communities around the forest. The State retains ownership of the forest resources and communities are only allowed access to non-timber forest products (NTFPs) under the sharing arrangement. Importance has been placed on harvesting grasses for thatching and brooms. However, most communities were not clear about the role of the RMCs and saw them as an extended arm of the FC. The RMCs were accountable only to the FC and did not respect existing traditional leadership structures, leading to conflicts between the two structures.
The RMCs opened bank accounts to deposit money raised by issuing permits to resource harvesters in the forest. Although RMC's said these accounts belonged to the communities, a FC officer was a signatory on the account and money could not be withdrawn without the FC's approval. This demonstrated that RMCs were upwardly accountable to the FC and not to the communities they represented (Manor Reference Manor1999). The FC remained reluctant to grant local communities the power to decide how to use their money and resources. Decisions about harvesting remained with the FC, indicating that power still lay with the state and had not been transferred to local organizations that represented communities (Mapedza Reference Mapedza, German, Karsenty and Tiani2009).
Conflicts continued to plague the project and were inadequately addressed (Matose Reference Matose2006) because no mechanisms existed to resolve them with the introduction of the resource-sharing project. The FC believed good governance centred on the introduction of the resource-sharing model of devolved management. The introduction of RMCs to the institutional landscape only aggravated existing tensions; what was needed instead was a meaningful transfer of power and responsibility to communities through them. Consequently, the chaotic relationship between the state and local people failed to improve during the resource-sharing project, and the forest continued to deteriorate as communities accessed resources beyond what was sustainable. It took a third party to break the impasse. In 1999, when foreign assistance ended, CIFOR's adaptive and collaborative management (ACM) approach, based on improved social learning, was introduced to the FC's resource-sharing project, with the objective of improving and learning from collaboration with selected forest user groups from communities.
ACM is a value-adding approach whereby individuals or groups who use or manage a forest agree to act together and draw up plans for their forests (Ruitenbeek & Cartier Reference Ruitenbeek and Cartier2001; Colfer Reference Colfer2005). These plans are then implemented, with the recognition that they may not fulfil intended objectives. In this process, stakeholders learn collaboratively from implementation the negotiate changes in the plans for forest use. Lack of progress in the resource sharing project was the entry point for introducing CIFOR's ACM research project. The ACM approach was introduced to add value to the joint forest management scheme that was in place around Mafungautsi Forest between the FC (on behalf of the state) and local communities. The scheme, as explained above, had not improved collaboration between the state and local communities, as intended at the start in 1994. Thus, the focus of the CIFOR initiative was on developing approaches that would enhance learning and improve the existing management systems.
CIFOR's project was taken over by the FC through officers who died respectively in 2005 and 2007. Although learning and collaboration activities stopped when the first officer died, one RMC continued organizing meetings with local resource user groups and ordinary members of the community to plan and decide how to manage their forest resources on their own. The local community members decided to use money (US$ 188.16) raised in July 2004 to build a small house for a teacher at a local primary school. The construction of the teacher's house was successful, and community members commended the committee for being accountable to them in use of funds raised from forestry activities.
While the RMCs had become more accountable and transparent to the communities they represented, they became less operational (in terms of activities on the ground) after 2006, as problems and conflicts intensified amongst their membership (Mutimukuru-Maravanyika et al. Reference Mutimukuru-Maravanyika, Prabhu, Matose, Nyirenda, Kozanayi, Mandondo, Prabhu and Matose2008). However, community members in some areas continued to use the knowledge and networks they had gained during the ACM project. For instance, some community members had met their member of parliament, who had not previously visited their area, and later made several requests for services that they eventually received (Mutimukuru-Maravanyika et al. Reference Mutimukuru-Maravanyika, Prabhu, Matose, Nyirenda, Kozanayi, Mandondo, Prabhu and Matose2008). The downward accountability that had been initiated by the decentralization of forest resources had thus promoted broader political representation, an unintended outcome, albeit positive.
In terms of livelihoods, participants were able to invest their income from forestry in livestock, which in turn provided much of the financial capital that was gained by households participating in the ACM project. Livestock acted as a buffer in the event of crop failure. Of the 70% of households owning cattle, 31% were involved in collaborative management activities. Before the collaboration, 61% of the households owned cattle, of which only 4% were involved in serious collection of forest products (Mutimukuru-Maravanyika et al. Reference Mutimukuru-Maravanyika, Prabhu, Matose, Nyirenda, Kozanayi, Mandondo, Prabhu and Matose2008). Some incomes raised by two RMCs as part of the livelihoods were changed by the ACM project (Table 1).
Table 1 Net income raised by RMCs after harvesting permits. In 2005, all money raised through the issuing of permits (together with the funds generated in 2004) was used to purchase 15 bags of cement as well as pay the builder who built a two-roomed house for one of the teachers at Gababe Primary school.
Chipange Chetu (Mozambique)
Before the creation of the Project, traditional authorities controlled the use of natural resources, but weak enforcement capacity resulted in uncontrolled access to resources. To address this problem, a local committee was created and its members trained in forest and fauna legislation, land management legislation, leadership and community regulation. The process was participatory and democratic, and resulted in the election, by local communities, of the president, vice- president, secretary and treasurer. Traditional leaders were appointed as committee advisors to avoid unnecessary accumulation of powers and ensure accountability. A complementary group integrated government and non-governmental institutions and incorporated provincial forestry and wildlife services, the District Agricultural Directorate, District Administration of Sanga and non-governmental organizations (NGOs) with the mandate of promoting community training, research and other activities.
Central to the Chipange Chetu project was security of tenure that was provided to the community through a title to the land (DUAT) under Mozambique's land law of 1997 (Salomão & Matose Reference Salomão, Matose, Dewees and Campbell2008). The DUAT, granted in 2003, represented the largest community title of its kind in Mozambique (Tanner et al. Reference Tanner, Baleira, Norfolk, Cau and Assulai2006). Effectively the community had official title to the land, such that, in theory, they could benefit from natural resources on that land. In practice, however, the provisions of the Wildlife and Forestry Law (1999), and decrees of 2002, completely undermined the security of tenure provided under the ownership of land, through cumbersome bureaucratic procedures that took away benefits from concessions from communities to the state (Salomão & Matose Reference Salomão, Matose, Dewees and Campbell2008). Nonetheless, through the intervention of the governor of the province and the provincial head of wildlife and forestry, communities were able to receive 80% of hunting fees, with 57% being direct cash dividend and 20% going to the district authorities. This special arrangement was permitted for three years after the signing of the DUAT. The special granting of benefits to communities strengthened interaction between communities and district authorities, where the relations were previously tense due to wildlife hunting (Anstey & Rihoy Reference Anstey, Rihoy, Mukamuri, Manjengwa and Anstey2009).
Overlapping authority between a traditional queen and the local committee resulted in differing treatment of external investors. Internal conflicts within the community led to a government decision to reorganize. This reorganization is ongoing and supported by the Irish International Development Cooperation. Community consultation is seen as central to determining continuation of project activities. Recommendations by a team that made a field visit to the project, defended the transformation of the project into a coutada (community hunting area) together with tourism, where communities could invest as entrepreneurs. However, this suggestion was not approved by the Ministry of Wildlife and Forestry Services (SPFFB), which preferred its transformation into a fazendas do bravio (wildlife farm), a legal mechanism seen as a more open and accessible model for community participation.
As a result of the training provided by NGOs in the area, local communities acquired a good level of legislative knowledge. Current benefits of the project implementation consist of production and sale of honey and sawn wood, and construction of access roads and health clinics. The roads facilitate communication with areas that were once inaccessible, access to markets has thus been opened and local products can be exchanged both within and outwith the district. The community received payments resulting from the different activities of the project to the value of US$ 6180 in 2001 and US$ 31 000 in 2004.
Economic benefits from sport fishing and game shooting were also generated. At the end of each year of activities, each community involved presents a report indicating the expenditure of the allocated revenues. The Project has undertaken a zoning of the areas with abundant resources. The contents of the management plan were defined in a participatory way involving the community and the officers from SPFFB. The decisions for participation in resources management are entrusted to the management committee facilitated by SPFFB.
Tsitsikamma Forest (South Africa)
Access to information is central for communities becoming knowledgeable about participatory forest management (PFM) for socioeconomic development and conservation of natural resources. Such information creates community awareness about their rights, privileges and responsibilities. This need is expressed in the White Paper for Sustainable Forestry Development in South Africa, the National Forestry Action Programme and the National Forests Act of 1998, which collectively seek to achieve sustainable forest management in South Africa (DWAF 1997a, b; Government of South Africa 1998b).
Local people around Tsitsikamma, particularly the forest-margin dwelling community of Covie, are unaware of the devolved forest management policy. DWAF officials had not put the PFM policy for the Covie community into practice. Eighty per cent of households around Covie were aware of DWAF, but were poorly informed about PFM, indicating that the department had not involved local people in the management of Tsitsikamma Indigenous Forest (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006).
Lack of active participation by the Covie people in the management of the natural resources in their vicinity contradicts the conservation laws that encourage their involvement in biodiversity conservation. There are two reasons for the mismatch between policy and practice. Firstly, forestry officials responsible for PFM lacked understanding of what was required by devolved management practice. This was exacerbated by the lack of leadership from head office to implement the policy. Consequently, forest managers at the local level pursued PFM independently without the needed knowledge and training. On the ground, managers tended to focus on the formation of PFM forums that generated few tangible results from the resolutions made. Despite a series of such meetings held over two years, forest-based projects proposed by local people failed to achieve approval, resulting in disillusionment and waning interest in PFM (Watts Reference Watts2006). Consequently, forest managers were unwilling to call more PFM forum meetings because proposed projects submitted to the regional office in Knysna for approval were not being addressed.
Secondly, community representatives who attended PFM forum meetings were not mandated by their respective communities and were not accountable to the communities that they purported to represent in such meetings. As a result, local people were not uninformed about the state's intentions and activities. Conversely, forest officials were ignorant about what the community representatives reported back to their constituencies. By contrast, local communities perceived that officials interacted more willingly with business people that relocated to Tsitsikamma, who consequently benefited more from the management of protected areas in Tsitsikamma. A further weakness was also observed in relation to community representation on PFM forums. Communities that have 20 or 150 households each have only one representative on the PFM forum and with the exception of the Thornham Community in Tsitsikamma lacked structures and organization for effective internal or external communication (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006).
More than half of the people in Covie indicated in 2006 that the management of the indigenous forest in Tsitsikamma had deteriorated with the introduction of PFM (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006). They pointed out that previously when they had not been prohibited from entering the forest in their backyard to gather wood. However, they were being denied access to such resources through increased restrictions through the issuance of permits. Communities around Tsitsikamma perceived permits as unnecessary restrictions, and many believed that they had to pay to be issued forest use permits, although permits for certain forest products and services were issued without fees to resident communities (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006).
Most households in Covie stressed that they had not benefited from the management of the indigenous forest (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006). Natural forest products used by the community were extracted illegally and were not considered to be benefits. Formal and casual employment was the most important source of income in Covie. Very few households worked for the conserved Tstitsikamma Forest. The nearby forest was a convenient source of fuelwood for all the householders in the Covie Community. However, local people would not readily supply this information for fear of retribution or increased policing by forest authorities. In fact, few locals regarded forest products such as wood and medicinal plants as benefits because they had been exploiting these resources without formal access prior to the introduction of PFM. As a result, there was an increased expectation after SANParks took over the practical management of the forest for SANParks to do more in and around Covie to improve benefits to local communities (Watts & Holmes-Watts Reference Watts and Holmes-Watts2006).
The illegal use of wood for energy by local communities had no discernible impact on the forest structure because they collected only dead, dried and fallen twigs and branches at the periphery of the forest. They also collected tree debris generated during the felling of diseased and dead trees.
DISCUSSION
Review of forest management reforms across Southern Africa has revealed relationships differed between forest governance reforms and their impacts upon local livelihoods (Table 2).
Table 2 Summary of findings across the three cases. ACM = adaptive and collaborative management, RMC = resource management committee.
Five salient features for policy emerge from the three experiences spanning the three countries (Table 1). Firstly, there has been no tenure reform in Zimbabwe and South Africa, where communities neither have land nor forest resource rights, however in Chipange Chetu (Mozambique) the state transferred land rights to communities through the DUAT, although forest and wildlife tenure remained with the state. Governments need to give communities rights to forests and resources (German et al. Reference German, Karsenty and Tiani2009). Even in Chipange Chetu, where land tenure is secure, forest tenure remains with the state in a manner similar to Mafungautsi and Tsitikamma.
Secondly, organizational structures for resource management appear to be problematic. Chipange Chetu presents a mixture of informal customary organizations and local elected committees under district councils working in tandem (Matakala & Mushove Reference Matakala and Mushove2001). Conversely, the Mafungautsi and Tsitsikamma examples promote greater participation through user groups/committees (sensu Manor Reference Manor, Ribot and Larson2005), or as indicated previously through scaling down. Depending on the scale of the forest, the presence of user group committees might facilitate increased participation, although there are also concerns about elite capture (Iversen et al. Reference Iversen, Chhetry, Francis, Gurung, Kafle, Pain and Seeley2006). Thirdly, local communities are not involved in decision-making processes and accountability to the local community is either lacking or weak. Yet, local participation in decision making may create a sense of ‘ownership’ and involvement in decisions (Larson & Ribot Reference Larson and Ribot2004). The participation of local communities needs to be significant for them to obtain benefits from CBNRM initiatives (Matakala Reference Matakala, Nhantumbo, Foloma and Puná2004). Effective participation requires that communities have abilities and capacities, meaning that community capacity building should be one of the forestry sector's priorities for effective CBNRM. However, in all three cases, decision making over key forest resources still remained with the state, including timber and wildlife concessions, although in Mozambique, negotiations were conducted into attempts to scale down to communities. In Mafungautsi, RMCs were solely upwardly accountable until the ACM project initiated plans to democratize these structures. State officials responsible for the Tsitsikamma Forest were accountable only to the state, and required only limited consultation with local communities.
Fourthly, transfer of power to lower levels has not occurred uniformly throughout the case studies, confirming earlier views on decentralization (Larson & Ribot Reference Larson and Ribot2004).
In Mafungautsi and Chipange Chetu, some adaptive management systems were initiated in order to provide lessons for devolved management for other forest or protected areas by the respective governments. In Mozambique, at least some financial resources from sport hunting and timber harvesting were transferred to communities, but decision-making power over the same resources was retained by the state, as in Zimbabwe. Conversely, in Tsitsikamma, power remained with SANParks as the statutory body. Governments need devolve power to communities to manage resources and equitably share the revenues derived from forests.
Fifthly, in relation to livelihood outcomes, our case studies reveal differing results and show limited progress, with the exception of Mozambique an increase of up to 40% in household incomes was recorded (Anstey Reference Anstey, Dzingirai and Breen2005). In Mozambique, Article 102 of the Forest and Wildlife Regulation states that 20% of the revenue from forest and wildlife exploitation be returned to the local communities living in the area from where the resources are extracted (Sitoe & Tchaúque Reference Sitoe and Tchaúque2008). Article 112 of the same regulation establishes that 50% of the fines collected from transgressors of the legislation are given to forest patrol personnel and community members who participate in law enforcement activities or report infringements of forest and wildlife law (Sitoe & Tchaúque Reference Sitoe and Tchaúque2008). However, the procedure for returning the money accrued from forest resources to local communities is still somewhat unclear (FAO [Food and Agriculture Organization of the United Nations] 2008). For example, the interministerial decree that established the revenue-sharing system requires that local communities are represented by a legal entity, which must have a bank account. Although this may appear to be a minor requirement, most rural communities in Africa need assistance from the state or a community advocacy group to establish a committee and receive basic legal and management training. At the same time, few rural districts are served by financial institutions. The express aim of these two articles in Mozambique is to strengthen the participation of local communities in forest management, ensure benefits to communities and ultimately contribute to poverty alleviation (Sitoe & Tchaúque Reference Sitoe and Tchaúque2008; Ribot Reference Ribot, German, Karsenty and Tiani2009). This is a means of compensating communities that have land-use rights and both articles are in line with the sustainable forest management principles of providing social benefits to local communities and ensuring the monitoring of forest operations (ITTO [International Tropical Timber Organization] 2003; Colfer Reference Colfer2005).
In the Zimbabwe and South African cases, small revenues were generated from minor activities in which communities engage under the joint management schemes, but these did not contribute substantially to community or household livelihoods. As a result, there is a need to make benefits from forests more tangible to the communities involved, because forests offer a wide range of both tangible and intangible benefits that are not adequately recognized in current forestry enterprises (Rana et al. Reference Rana, Shrestha and Silwal2008; German et al. Reference German, Karsenty and Tiani2009).
Experiments in devolved forest management have thus been initiated in the region, but much more needs to be done in practice. In fact, the scale of transferring power and resources is an issue that still requires attention in order for communities to enjoy rights and accrue meaningful benefits. The Mozambican case appears to offer some limited hope, as it combines local customary institutions with local elected officials and also allocates defined percentages of revenue from forest resources to local communities.
CONCLUSIONS
The country case studies provide very mixed experiences of decentralization outcomes, particularly as they relate to livelihoods of communities. Where there are high-value forest resources, the state exhibits reluctance to transfer power over resource management (see Shackleton & Campbell Reference Shackleton and Campbell2001; Nelson & Agrawal Reference Nelson and Agrawal2008) as witnessed in Chipange Chetu and Tsitsikama. The Mafungautsi case was more complex in relation to the value of resources argument, and determining what had inhibited the state from transferring power given the low value of resources. Perhaps the decade-long political impasse has stalled the evolution of decentralization as even the limited revenue derived from such forests as Mafungautsi has gone to the state (Mapedza Reference Mapedza, German, Karsenty and Tiani2009).
Regardless of the limited experiences of the cases reviewed, overall decentralization leads to the transfer of forestry power to some local levels, provides for more secure access to forest resources, which in turn increases benefits for local communities. While decentralization may not result in pro-poor forestry (Hobley Reference Hobley2007), the resultant changes in improving property and access rights, as well as accountability and limited transfer of forestry powers, facilitates benefits being derived by local communities across the three cases reviewed. The challenge is whether first-generation experiments in decentralization will be followed up to lead to more enduring outcomes, as the ACM project demonstrated. The cases also show that decentralization has to be situated in the broader political economy context which in turn, provides challenges for activities in the forestry sector and at the local level.
