Green grabbing and the dynamics of local-level engagement with neoliberalization in Tanzania's wildlife management areas

This paper analyzes the politics and struggles ongoing within wildlife management areas (WMAs) in Tanzania to discuss the dynamics of neoliberalization of the wildlife sector. We discuss neoliberalization as a new political-economic context within which the ongoing politics of natural resource management are played out, and focus on green grabbing as an expression of these politics. We discuss how local-level actors are engaged in these processes, often in strategic ways, to negotiate their roles within WMAs and address green grabbing by the state. Secondly, we discuss an example of the politics of land control and local-level actors’ enactment of accumulation by dispossession within a WMA.


Introduction
Debate about the nature of neoliberal conservation and the complex processes of neoliberalization in the environmental sector (e.g. Harvey 2005;Igoe and Brockington 2007;Kelly 2011;Peluso and Lund 2011) has included discussion of the wide range of actors involved and the injustices of associated processes of 'green grabbing' (Castree 2010;Fairhead, Leach, and Scoones 2012). There is a growing literature that considers the way local actors engage with the sweeping changes being brought about by neoliberalization, respond to this changing context and resist green grabbing (e.g. Benjaminsen et al. 2013;Gardner 2012). In this paper, we contribute to this literature by examining the dynamics of neoliberalization of the wildlife sector of Tanzania, focusing on the local level and the emergence of new forms of engagement with processes of green grabbing and accumulation by dispossession.
We focus on community-based natural resource management (CBNRM), as implemented in Tanzania in a new wildlife policy in 1998 through the creation of wildlife management areas (WMAs). Tanzania's adoption of CBNRM in the late 1990s followed policy change across eastern and southern Africa in search of models to address both rural development and wildlife conservation (Hulme and Murphree 2001;Nelson 2010;Nelson and Agrawal 2008). Income from game viewing tourism and (except in Kenya) tourist hunting has been the common focus of CBNRM policies across the region, but it has taken a variety of forms in different countries. In some there has been a degree of devolution of resource rights and control (Namibia, Botswana, Zimbabwe), while in others there has been little (Zambia, Mozambique, Kenya and Tanzania, Nelson and Agrawal 2008). In all cases, donor support has been considerable, but in Tanzania (and these latter countries), strongly centralized governance and lack of transparency have limited donor pressure to decentralize the control of resources (Nelson and Agrawal 2008).
Research on neoliberalization and CBNRM has tended to focus on the nature and impacts of green grabs (e.g. Benjaminsen and Bryceson 2012;Gardner 2012), the impacts of such processes on local people (e.g. Büscher and Dressler 2012) and the creation of new, neoliberal environmental subjects (e.g. Fletcher 2010). Here, we aim to extend the understanding of the role local-level actors play in neoliberalization by considering the way they are engaged (cognitively and politically) in the processes of neoliberalization. We also examine the extent to which these dynamics and processes should be seen as a product of neoliberalization. Here, we draw on the longstanding and wide-ranging literatures on the politics of natural resource management and participatory development to ask what has changed under neoliberalization. In their discussion of green grabbing, Fairhead, Leach, and Scoones (2012) argue that what distinguishes neoliberalization from previous discussions of the politics of natural resource management is the number and variety of different actors involved, and the focus on novel forms of valuation and the commodification of nature. Here we investigate how these changes have altered such politics of natural resource management.
We begin by outlining neoliberalization in nature conservation, and green grabbing in the context of CBNRM. Next, we frame this within a discussion of the politics of natural resource management and the struggles taking place around this. Using examples from Tanzania, we then build on previous discussions of green grabbing in WMAs to frame the wider dynamics of neoliberalization. We emphasize, firstly, the role of local actors in these processes and, secondly, the interactions between these processes and longstanding struggles over natural resource management that span the local, national and international levels.
The paper is based on fieldwork by the first author in Tanzania from March-August 2010 and January-June 2011. Data collection included qualitative interviews with actors across all levels of WMA governance, including ministry officials, regional and district authorities and Authorized Association Consortium members (see below). The findings also draw on interviews and a focus group carried out at the local level with Authorized Association representatives and leaders, village authorities and local people during an indepth investigation of one WMA in southeastern Tanzania.

Neoliberal conservation, green grabbing and community-based natural resource management
Neoliberalization needs to be understood as a multifaceted and dynamic process that involves a wide range of both state and non-state actors across different levels (Castree 2008(Castree , 2010Harvey 2005). A growing literature explores the implications of neoliberalization (an ongoing and unfinished process involved in the testing, application and revision of neoliberal philosophy; Castree 2010) for biodiversity conservation, and analyses the resulting 'neoliberal conservation' (Igoe and Brockington 2007;Kelly 2011;Sullivan 2013).
Neoliberal conservation is market-driven, and consists of the reregulation of nature through commodification, the spread of free markets to manage these commodities and the development of a 'green economy' (Fairhead, Leach, and Scoones 2012;Igoe and Brockington 2007). It involves the opening up of areas of land for capital accumulation under the paradoxical assumption that the tense relationship between capitalism and the environment can be solved through the latter's complete integration into the capitalist system, resulting in growth, improved efficiency, development, democracy and sustainability Castree 2010;Kelly 2011). A critical aspect of neoliberal conservation is the reduced role of the state in the management and conservation of nature, and a larger role and increased collaboration with other organizations, including non-governmental organizations (NGOs) and private enterprises (Castree 2010;Igoe and Brockington 2007). These new collaborations and alliances are described as the creation of new, hybrid forms of environmental governance, in which responsibility for the conservation and management of the natural world is no longer assumed to be the responsibility of the state, but instead involves both state and non-state actors (Igoe and Brockington 2007;Peck and Tickell 2002).
CBNRM is the devolution of authority to manage natural resources, often involving the granting of tenure rights, to autonomous local/community institutions, based on three core principles of participation, empowerment and decentralization (Dressler et al. 2010;Hulme and Murphree 2001). The connections between neoliberalization and CBNRM are complex; the enrolment of communities in official systems of natural resource management reflects neoliberalism's mobilization of non-state actors, but also opens the possibility of solidarity-based resistance to processes of privatization and market extension (see also Li 2007). However, CBNRM is compatible with neoliberalism's philosophy of a reduced role for the state in natural resource management (or a reduced role for national governments with the devolution of powers to local governments), and often involves the development of close relationships with the private sector (e.g. tourism companies and NGOs).
Moreover, the expansion of capitalism involves natural resources being commodified and traded in markets (Kelly 2011), and such markets are a common feature of CBNRM (e.g. for wild products or tourism). Such markets need not be novel: Tanzania, for example, has operated a national and international market for wildlife hunting since the colonial era, pre-dating neoliberalization by decades (see Leader-Williams, Kayera, and Overton 1996;Siege 2001). However, as in other African countries, the neoliberalization of the wildlife sector has developed these markets, and brought local communities into engagement with them (Nelson 2010). In Tanzania, WMAs now allow local communities to hold formal rights to manage wildlife resources and to benefit financially from tourist hunting and game viewing tourism facilities.
CBNRM therefore represents a hybrid form of environmental governance that merges neoliberal ideas of private markets (Dressler et al. 2010;Peck and Tickell 2002) with populist arguments for participatory development (Mohan and Stokke 2000). The adoption of CBNRM in Africa from the 1980s reflected a global 'turn' in conservation policy towards saving nature through 'win-win' strategies that involved nature 'paying its way' (Adams 2004;Dressler and Büscher 2008;Dressler et al. 2010;Goldman 2011). CBNRM seemed to promise 'a world in which it is possible to eat one's conservation cake and get a development dessert too' (Grandia 2007, 480) by handing power to local communities and giving them access to the markets generating revenue from wildlife resources. However, the proposed 'win-win' outcomes from CBNRM are not easily achieved and not uniformly distributed (Büscher and Dressler 2012;Dressler and Büscher 2008;Dressler et al. 2010;Frost and Bond 2008;Selfa and Endter-Wada 2008;Sikor and Nguyen 2007). Across eastern and southern Africa, patterns of elite patronage and rent-seeking, allied with a lack of transparency in the allocation of hunting contracts and consequent benefits, have encouraged a high degree of centralized control of wildlife resources, except in one or two cases (notably Namibia, Nelson and Agrawal 2008).
Green grabbing is a term that denotes the privatization or appropriation of land and the exclusion of local people from natural resources on the basis of 'green' credentials (Fairhead, Leach, and Scoones 2012). Green grabbing arises from neoliberalization's drive for privatization, resulting in what Marx described as primitive accumulation, 1 and Harvey (2005) discusses as 'accumulation by dispossession'. Accumulation by dispossession involves the 'commodification and privatization of land and the forceful expulsion of peasant populations' (Harvey 2005, 159), bringing about capital accumulation for some at the expense of others. Primitive accumulation and accumulation by dispossession apply not only to physical enclosure of land and physical removal of people, but also to changes in access, rights, institutions and rules as a result of commodification, as well as the restructuring of authority and human-ecological relationships (Fairhead, Leach, and Scoones 2012;Kelly 2011). It is in this respect that we investigate accumulation by dispossession, drawing on Kelly's (2011, 686) description of how local communities engaged in CBNRM are entrained in the neoliberal processes of engaging natural resources with the market and 'at the same time relieving local people of their land/resources while … [they] become dependent on an industry (often ecotourism) over which they have no control and from which they earn very little money'.
Previous research has argued that green grabbing has taken place in Tanzanian WMAs through three processes of dispossession of local actors (Benjaminsen and Bryceson 2012): firstly, the continual restriction of grazing land (see also Gardner 2012; Patinkin 2013); secondly, a loss of control over crops due to an increased level of wildlife-related damage; and, thirdly, through the emergence and institutionalization of markets for wildlife resources over which local communities have little control. In this paper, we develop this analysis, focusing on the third of these processes of green grabbing.
The politics of nature-state-territory in natural resource management Struggles over land, authority and natural resources are not new in Tanzania. Issues of access and property regarding natural resources are intimately tied to the exercise of and struggle over power and authority (see Gardner 2012;Sikor and Lund 2009). The enclosure of land for conservation purposes has previously been discussed as a critical component of statecraft (Neumann 2004), and the devolution of rights to manage natural resources has been widely discussed in terms of the recentralization of power that may emerge (Ribot, Agrawal, and Larson 2006). The politics of devolved natural resource management are therefore well known (see Nelson 2010), and, arguably, the institutional blending involved in such public-private-community arrangements as CBNRM allows the state to extend its capacity to achieve public conservation policy outcomes (Hodge and Adams 2012).
Land and authority over natural resources have long been central to both state-society and nature-society relationships in the politics of what Neumann (2004) terms 'nature-stateterritory', Lund (2011) term 'land control' andVandergeest andPeluso (1995) discussed as 'internal territorialization'. Such processes of struggle, however they are termed, involve multiple actors, not just the state, in negotiations and contestations that set the rules which allow and disallow certain forms of land use and access (Sikor and Lund 2009). We follow Neumann (2004) in using the term 'nature-state-territory' to describe a complex set of relationships including state-society and nature-society relationships with respect to territory and enclosure.
In Tanzania, these politics have a long and complex history, indicating important shifts in state-society and nature-society relations over the twentieth century. This has been the subject of extensive discussion (e.g. Brockington 2002;Goldman 2003Goldman , 2011Nelson 2007) and such politics are clearly not confined to Tanzania (see Nelson and Agrawal 2008), but we focus here on this case to contextualize our own findings.
The politics of nature-state-territory in Tanzania were first documented in relation to the processes of enclosure and displacement under colonial policies of social control and spatial segregation, in which protected-area enclosures fixed boundaries between nature and culture, and ultimately between society and the state (Neumann 2001(Neumann , 2004. Centralized control over natural resources has been the norm in Tanzania since the first game reserve was gazetted in 1896 (Nelson, Nshala, and Rodgers 2007;Siege 2001) and, to this day, the state retains ownership of all wildlife resources, even if communities now have the right to manage them in WMAs (Government of Tanzania (GoT) 2009). The history of evictions and colonial appropriation of land have created tense and untrusting relationships between the state and local populations, particularly amongst pastoral groups displaced by protected areas (Brockington 2002;Goldman 2003) and blamed for environmental degradation (Walsh 2012). Access to rural land and resources was further reshaped, and service delivery to rural areas was reduced by the policy of Ujamaa or villagization, in which over 75 percent of the population was relocated into planned settlements between 1967 and 1975 (Coulsen 1982). Ujamaa, however, was also the first step in securing the legal rights of the village, rather than central government, to be responsible for land and natural resources (Nshala 2002).
The management of wildlife resources, and the responsibilities and relationships between the state and local communities, changed significantly in the 1980s, a product of both the introduction of neoliberal reforms and a growing political agenda in Tanzania for multi-party democracy, human rights and good governance (Mniwasa and Shauri 2001). In 1985, Tanzania's acceptance of the International Monetary Fund's Structural Adjustment Programme for induced privatization reduced the role of the state in service delivery and brought a huge influx of foreign capital and donor-funded projects in the wildlife sector (see Neumann 1998). Around the same time, important shifts in the policy environment 2 and a 1984 amendment to the constitutions formalized institutions and rights of local government, established village land as a legal category and authorized village authorities to manage natural resources on that land (Mniwasa and Shauri 2001;Nshala 2002).
These factors were important in the shift towards community conservation in Tanzania (see Igoe and Croucher 2007;Nelson 2007), first seen in the Serengeti Regional Conservation Strategy in 1985. However, the politics and struggles surrounding devolved natural resource management continued (Mustalahti 2007;Nelson 2010). In WMAs, these political struggles centred on: (1) the devolution of power to local communities, and how this is recentralized, controlled and restricted by the state, particularly through the application procedure to create a WMA and a lack of autonomy of the local level in managing a WMA; (2) the complex revenue-sharing arrangements in place and inequalities in benefit distribution; and (3) the often disappointing outcomes for community development (Humphries 2013;Nelson 2007;Nelson, Nshala, and Rodgers 2007).
The processes of green grabbing that have drawn so much attention in recent years should be seen in the context of these ongoing politics of nature-state-territory in Tanzania. Two key points need to be made about green grabs: first, they are not homogeneous or single events, but are processual, involving 'a series of changing contexts, emergent processes, forces and contestations', and are under constant negotiation regarding access (Peluso and Lund 2011, 669); second, the role of local actors is central to these struggles. Green grabbing does not happen to local communities as an undifferentiated whole, but is continually negotiated among local actors. Gardner's (2012) discussion of struggles between local groups, private investors and the state in the context of tourism in Tanzania is a good example of these political interactions. Local groups have seized on the neoliberalization of Tanzania's wildlife sector as an opportunity to regain authority over land and the benefits that can be derived from wildlife resources in an active challenge to state authority, and to address previous injustices (Gardner 2012). Here, we provide evidence of the shifting parameters of these struggles and the politics of nature-state-territory as local actors get involved with the processes of neoliberalization, and try to find ways to benefit from these changes and to challenge the inequalities of green grabbing taking place within WMAs.

Tanzania's wildlife management areas
WMAs were created across Tanzania from 1998, under the Wildlife policy for Tanzania (GoT 1998). A WMA is defined as 'an area set aside by a village government for the purpose of biological and natural resource conservation' (GoT 1998, 35). The 1998 wildlife policy introduced CBNRM in the wildlife sector in Tanzania, providing a legal mechanism 'to allow rural communities and private land holders to manage wildlife on their land for their own benefit' (GoT 1998, 14), with the aim that participating communities would have the 'full mandate of managing and benefitting from their conservation efforts' (GoT 1998, Statement 34). Day-to-day management of a WMA is conducted by an elected group of local actors (who form the Authorized Association) in accordance with a management plan.
WMAs reflect the neoliberalization of the wildlife sector through the reduced role of the central state in wildlife management, and the central idea that communities receive economic benefits through investment agreements for the consumptive and non-consumptive use of natural resources (most commonly through tourist hunting and photographic tourism enterprises). This was intended to make wildlife management a competitive form of land use, and thereby incentivize the sustainable management of wild animal populations (Nelson 2007). WMAs depend on the commodification of wild animals and the marketing of the resulting products in order to generate revenue that can be targeted at the local level. In WMAs, local communities become partners in business ventures, wildlife resources are commodified and markets for these are expanded (Igoe and Croucher 2007).
WMAs are involved in markets for wildlife commodities in two ways: firstly, through tourist hunting taking place on areas of land gazetted as hunting blocks; secondly, through investment agreements between Authorized Associations and private investors for non-consumptive uses of wildlife (such as photographic tourism). Figure 1 shows the revenue system that was in place at the initiation of WMAs in 1998, and subsequent changes. These arrangements correspond to what might be called 'green grabbing' in WMAs set out by Benjaminsen and Bryceson (2012) in both streams, as described below.
Commercial tourist hunting in Tanzania has taken place since the 1960s, with a short hiatus between 1973and 1978(Leader-Williams, Kayera, and Overton 1996. The Wildlife Division, which has overseen the tourist hunting market in Tanzania since 1984, has been widely criticized for its opaque system of hunting block allocation, opportunities for corruption and rent-seeking and for the expansion of state control and influence over investments in the wildlife sector (Baldus and Cauldwell 2004;Benjaminsen and Svarstad 2010;Gardner 2012;Igoe and Croucher 2007;Leader-Williams, Baldus, and Smith 2009;Nelson et al. 2009). Within this political economy of wildlife management, it is unsurprising that central government officials have been reluctant to devolve power or revenue collection arrangements to the local level (see Nelson and Blomley 2010). A former pilot project member of staff described how colleagues found the initiation of WMAs ideologically challenging and uncomfortable because it involved a shift in attitude from one of 'enforcers controlling populations' to one of 'working with them as colleagues'. This caused serious doubts and resistance to WMAs within the Wildlife Division, which a member of staff explained led to a situation where 'whilst they officially praised it, [they] tried to sabotage the whole operation' (interview, 2011). Another former pilot project member of staff recalled that during participatory discussions with the Wildlife Division around policy development, certain areas were off limits to discussion, particularly regarding authority for tourist hunting (decision-making and revenue arrangements), over which the central authorities were unwilling to relinquish control (interview, 2010).
As a result of these politics, tourist hunting in WMAs has excluded local communities from both decision-making and revenue collection, and simultaneously created the exclusion of local people from village lands. The responsibility for allocating hunting blocks, issuing hunting concessions and collection of tourist hunting revenue remained with the Wildlife Division, despite promises that WMAs would phase out this state control and initiate locally managed tourist hunting (Benjaminsen et al. 2013). One of the main beneficiaries of the commodification of wildlife resources and the development of markets for wildlife tourism has been the Tanzanian state, and individuals within government organizations (Benjaminsen and Bryceson 2012). Commercial tourist hunting is the Tanzanian Wildlife Department's chief source of revenue (a recent article in the New York Times, published 17 March 2013, by the Director of Wildlife in Tanzania exemplifies this importance), but also a major element in patterns of patronage and rentseeking (Leader-Williams, Baldus, and Smith 2009;Nelson and Agrawal 2008). Whilst this hunting took place on village land within hunting blocks inside WMAs, villagers had no role in the management of this system for tourist hunting. Between 1998 and 2012, just 25 percent of revenue generated from tourist hunting was channelled back to the district level, and then shared with the participating villages in the WMA (although with little opportunity for Authorized Associations to hold the District Authorities or Ministry to account on this score; see Baldus and Cauldwell 2004;Gardner 2012;Sachedina and Nelson 2010). The second major beneficiary of the system for tourist hunting in WMAs has been the private safari hunting industry, often internationally owned. Gardner (2012) argues that the granting of hunting rights by the state to private business investors amounts to green grabbing through the 'foreignization of space' (Zoomers 2010): the exclusion of people from village lands in the interests of the conservation of game animals, which international operators' paying clients have the sole right to shoot. This territorialization is direct and enforced: as one informant commented: 'when we see villagers in our area, we chase them and, if we catch them, we beat them, we beat them hard. They know now not to come' (interview with tourist hunting operator staff member, 2010).
Importantly, arrangements for allocating hunting revenue underwent a significant shift in 2012 with the publication of updated WMA Regulations (URT 2012, see Figure 1). These stipulated that Authorized Associations would receive 75 percent of the hunting block fee, 5 percent of the game fee, conservation fee and observers fee and 15 percent of the permit fee paid by tourist hunting operations, URT 2012). WMAs with a designated hunting block now publicize opportunities and negotiate directly with tourist hunting operators (although a representative of the Director of Wildlife and state authorities at the District level must be involved in scrutinizing the tendered investments, and the Director of Wildlife within the Ministry of Natural Resources and Tourism must sanction the agreement before the Authorized Association signs; URT 2012, Article 51).
The second source of revenues in WMAs comes from investment agreements for nonconsumptive wildlife tourism (such as tourist lodges). These kinds of arrangements had existed on village land prior to the initiation of WMAs, and had been an important source of income for some villages, with the most lucrative agreements generating up to USD 40,000 annually (Benjaminsen et al. 2013). However, the central state (and its functionaries) had no say in these investments and was unable to benefit from this revenue, prompting a change in the rules (Igoe and Croucher 2007). In 2000, the Ministry of Natural Resources and Tourism published an amendment to the Wildlife conservation (tourist hunting) regulations which declared that all investment agreements within any wildlife Protected Area (including hunting blocks and WMAs) required written approval by the Director of Wildlife (MNRT 2000). This change was significant because it declared all previous such investment agreements to be illegal, extending the authority of the state over these lands (despite their legal status as village lands), dispossessing local communities of the rights to autonomously set up such agreements and dispossessing them of this means of generating revenue from wildlife resources. However, it should also be noted that this change does bring a benefit of protection for local communities from unfair investments from which they would not benefit adequately, and also serves to protect the environment from damaging investments (A. Songorwa, 3 personal communication, 2011).
This move by the state was undoubtedly designed to encourage investments through WMAs instead of individual villages, but Figure 1 highlights that such non-consumptive investment agreements within WMAs, through the Authorized Associations, also underwent a similar process, again comprising green grabbing. Non-consumptive investments in a WMA are subject to a tender system and are assessed by a Tender Evaluation Committee made up of members of the Authorized Association, the district authorities and a representative of the Director of Wildlife (URT 2012, Articles 49 and 50). In the original revenue arrangements, Authorized Associations received 100 percent of the revenue generated from these agreements. However, the Non-consumptive wildlife utilization regulations were revised in 2008 (GoT 2008), to reduce the share of revenue kept by the Authorized Association from 100 percent to 65 percent: the remaining 35 percent went to the central state.
As a result of these changes, local actors can only enter investment agreements to manage or benefit from wildlife resources with approval from the Director of Wildlife in the Ministry. The central government has thereby extended its authority over lands within WMAs and restricted the role of local actors in decision-making over these lands and, furthermore, these changes have dispossessed local actors of a significant proportion of the revenues from such agreements (see also Benjaminsen and Bryceson 2012;Gardner 2012).
Changing state-society relationships and response to green grabbing As communities respond to the changing opportunities in WMAs, and the processes of neoliberalization they unleash, both nature-society and state-society relations change. We address this by considering the changing relationships and power patterns around the governance of WMAs.
The neoliberalization of Tanzania's wildlife sector and the creation of WMAs have brought about various changes in natural resource governance and in relationships between state and society. First and foremost, they set out a reduced role for the central state and transfer of authority for wildlife resource management to local communities. In reality, this shift in state-society relations has been only partial and contested, as the governance arrangements in WMAs and a large literature on the politics of power devolution in WMAs have clearly shown. The value of wildlife resources, the long history of centralized control in the wildlife sector and the substantial opportunities for rent-seeking surrounding wildlife resources have all contributed to delayed or restricted devolution, or have recentralized power (see Igoe and Croucher 2007;Nelson et al. 2009). State-society relationships have remained tense and contested over the issue of authority to manage wildlife resources.
Secondly, the initiation of WMAs created opportunities for new alliances between local communities and NGOs, as each WMA is assigned an NGO facilitator to support the WMA and its Authorized Association. The NGO facilitator is involved in the application for and gazettement of the WMA, and the production of a management plan; the NGO provides technical advice to the Authorized Association, is charged by the Wildlife Division with conflict arbitration for the WMA and participates in an advisory board for the WMA that is involved in decisions to appoint non-consumptive investment agreements (URT 2012, Article 28). NGO facilitators may also act as a liaison between central ministries, district authorities and local communities (URT 2012). These new collaborations are an example of the hybrid forms of environmental governance that are an integral part of neoliberalization as a broader range of actors become involved (see Peck and Tickell 2002), and they represent important shifts in the relationships between state and non-state actors.
Local actors and members of Tanzania's Authorized Associations recently began to use these collaborations to position themselves to bring about change in the governance system for WMAs and to address the green grabbing that has been taking place. In 2010, a new organization was formed, named the Authorized Association Consortium (GoT et al. 2011), made up of representatives of WMAs across the country and facilitated by the international conservation NGO Worldwide Fund for Nature, with funding provided by the United States Agency for International Development. The official objective of the consortium was declared to be 'liaison with government, non-governmental organizations and/or private sectors in all matter [sic] that affect the consortium positively or negatively. These included development policies, legislations, guidelines, community development and distribution of land and natural resources' (GoT et al. 2011). The purpose of the consortium was to act as a forum for the unification of processes and systems in WMAs across the country, and also to provide a platform for Authorized Associations to present a coherent voice of WMAs to central government, increasing their influence at the national level and their role in the governance system for WMAs (A. Songorwa, personal communication, 2011). The chairman of one Authorized Association suggested that the consortium could be seen as 'the equivalent of the National Park Authority for WMAs in Tanzania … which has the power to speak to government and the ministry' (Interview, 2011). Through the Authorized Association Consortium, local actors have a channel through which to claim a legitimate voice in national-level decision-making, and members of the Consortium have begun to use this opportunity to challenge the institutional arrangements and access arrangements for wildlife revenues in WMAs. These claims to authority are evident in the consortium's ambition, expressed in a poster produced by the Authorized Association Consortium about its work, to achieve 'secured markets for different tourism products in the WMAs … representing Authorized Association members in decision making during wildlife quota allocation … [and] having favourable environment [sic] for investment in WMAs' (GoT et al. 2011).
It is interesting that the consortium's official objective particularly mentions engaging in issues of land and natural resource distribution, whilst the expected benefits of the consortium refer specifically to gaining Authorized Association members' representation in decision-making bodies over wildlife quota allocation for tourist hunting (see also Authorized Association Consortium 2013). Both of these issues have been central to the struggles taking place within the politics of nature-state-territory for decades. It is not surprising that local actors are attempting to effect change on these very issues, but it is important that neoliberalization, by opening up opportunities for collaboration with NGOs and shifting the state-society relations for wildlife management, has provided the context to enable these attempts to be made.
The Authorized Association Consortium allows local actors to engage not only with the national state regarding wildlife resources, but also with the system of power and authority within WMA governance. The Consortium is a vehicle by which local actors can position themselves to legitimately challenge the green grabbing that has been enacted by the state in the form of centralized control of tourist hunting and non-consumptive wildlife management in WMAs. Local actors were engaged in attempts to rearrange the governance system, redefine their authority and redistribute access to decision-making and benefit distribution in their own interests. Neoliberalization (by facilitating community management of natural resources, bringing these actors into national and international markets for wildlife resources, and through the facilitation of close relationships with other non-state actors) has played a key role in enabling this engagement, and it is also the framing within which actors are operating.
It has long been recognized that actors will deliberately adopt different discursive frames to suit their political motivations and personal interests (Brockington 2004). In the wildlife sector of Tanzania, Gardner (2012) argues that neoliberalization is being embraced by local actors and perceived as an opportunity to challenge the state with respect to ongoing political struggles over land. In WMAs, neoliberalization has significantly altered the context in which these struggles over land and state-society relationships are taking place, and the engagement of local actors within the consortium with the state relates to both of these processes. Firstly, it is evident from the discursive frame employed by the consortium in this literature, which shows that this engagement with the state and resistance to green grabbing does not consist of a rejection of neoliberalization in the wildlife sector. The discourse put forward by the consortium draws on neoliberalism's core concept of building market mechanisms for effective wildlife management and focuses on the creation of investments to achieve this. The consortium should not, therefore, be understood as a challenge to processes of neoliberalization, but as an attempt to utilize the opportunities presented by neoliberalization to reposition the role of local actors within WMA governance, enabling them to challenge centralized control in WMAs and the processes of green grabbing by the central state, and address the politics of naturestate-territory that have historically disadvantaged local people. Thus local actors are engaged in attempts to utilize and shape neoliberalization for their benefit, drawing on the discourse of neoliberal wildlife conservation in the politics of this struggle. The extent to which this engagement with neoliberalization is effective at bringing about change will take time to become clear. The role of the Authorized Association Consortium and actors linked to it, for example, in the recent changes in decision-making and revenue arrangements for tourist hunting, is a matter of current debate.

Changing nature-society relations and local struggles over land and authority: intra-village accumulation by dispossession
To explore the dynamics of these processes in greater depth, we investigate an act of political struggle over land and wildlife resources at the local level. We analyse a WMA named MBOMIPA, a Kiswahili acronym for Matumizi Bora Maliasili Pawaga na Idodi, meaning 'sustainable use of natural resources in Pawaga and Idodi' (the two administrative divisions in which the WMA is located). MBOMIPA represents 21 villages (with a total population of over 60,000) within the district of Iringa Rural, in the region of Iringa, southeastern Tanzania (see Figure 2). The MBOMIPA WMA is located along the eastern edge of Tanzania's largest national park, Ruaha, and was formerly designated the Lunda-Mkwambi Game Control Area (Hartley 1997 andsee Murphree 2000;Nelson et al. 2009;Walsh 1995Walsh , 1998Walsh , 2003. We discuss two examples from MBOMIPA that show how green grabbing and accumulation by dispossession have been enacted by local actors. Conflict over land and natural resource ownership has emerged among villages participating in the MBOMIPA WMA as a result of the commodification of wildlife resources and the development of markets. Conflict intensified following the low levels of community benefits paid to villages: of the USD 110,000 generated by investment agreements in the years , only USD 1590 was given to each village. These funds were invested at the community level in services such as health clinics and education resources. These are an important source of revenue for local communities, and indicate that the WMA system can channel benefits to local communities. However, disappointment and conflict stems mostly from the small amounts received (if the funds made available for community services are translated to a per capita financial benefit within the local communities, the WMA generated just USD 0.55 and USD 0.20 in these years), and the lack of adequate compensation mechanisms for crop damage by wildlife crossing into agricultural areas bordering the WMA.
When the WMA began its application for official gazettement in 2002, all of the 21 participating villages were made equal partners in the governance structure, had representatives within the Authorized Association, and received an equal share of WMA revenue (MBOMIPA 2002). However, only 14 villages actually held land inside the WMA, which was created from the southern portion of the former Lunda-Mkwambi Game Control Area (Figure 2). The other seven had no registered village land within the WMA. The WMA was set up in this way because it was feared that, if left out, the other seven villages might resent the WMA and serve as gateways for hunting and illegal activities within both the WMA and the adjoining National Park (M. Walsh, 5 personal communication, 2010; interviews with Authorized Association Representatives, 2011). Including the 4 Conflicts with investors and within the WMA were responsible for the reduction in both the total amount of revenue collected by the Authorized Association and the amount shared with the participating villages. 5 Martin Walsh was a member of staff on the donor-funded project that initiated community-based wildlife management in this location prior to the gazettement of MBOMIPA. additional villages in the WMA not only provided their inhabitants with opportunities to benefit financially, but also created a buffer of protection against illegal users of wildlife. This argument, and a vision of participatory conservation and community cooperation, was stressed continually by members of the villages that had not contributed land to the WMA: 'The primary objective of the project at the start was to stop hunting' (interview with village chairman, 2010); 'Those who want to change the system are only interested in money, not in conservation' (interview with village authorities, 2010); 'The idea was working together for conservation, participating together' (interview with Authorized Association representative, 2011); 'MBOMIPA belongs to society, not to one village' (interview with Authorized Association representative, 2011).
However, some of the villages with legal title to land within the WMA argued that, as the wildlife resources that underpinned revenue occurred on their land, they should receive a greater proportion of the revenue. This was further justified by arguments that these villages were generally closest to Ruaha National Park, and so were subject to higher levels of crop damage and human-wildlife conflict. During fieldwork (2010-2011), a group of representatives within the Authorized Association, supported by members of some village councils, were attempting to bring about a change in the revenue-sharing arrangements so that the seven villages without land inside the WMA would receive only a 10 percent share of the total revenue. 6 The 14 villages with land inside the WMA would then share the remaining 90 percent. A villager from a 'landed' village claimed that this was necessary because 'landless' villages were 'eating for free ' (interview, 2011), and an Authorized Association representative from a landed village described his support for altering the revenue distribution system because 'here [landed villages] we are hurt by conservation, but they [landless villages] only know the importance of conservation, but we are all paid the same' (focus group, 2011).
Some villages with land that was richer in wildlife claimed it should be treated as still belonging to them, rather than as a collective or common area shared between all the villages, and threatened to leave the WMA if changes were not made. Others challenged this commodified view of wildlife resources, arguing that 'wild animals don't belong to us; they are the wealth of the state. There isn't a single person that can say they belong to a specific village' (interview with Authorized Association representative, 2011).
These struggles were played out discursively at the local level as different groups employed contrasting justifications to gain or retain control. Those who wanted to change the revenue system territorialized wildlife as assets on their land, and drew on commodified language in their arguments; their opponents challenged these meanings. They described wildlife as a resource in ways that emphasized the stewardship and protection of nature, arguing that it was not all about money, and that revenue should be shared equally because: Those who have little are doing their work just as those that have lots of land. Will you withhold from them so that they are given little? … We share together because we all have one job to protect our land so that it can be sustainable, and even those who don't have [land], protect it. It is something that all of us should receive equally, therefore. (interview with Authorized Association representative, 2011) One village, Nyamahana (see Figure 2), which had not contributed land to the WMA, was a particular point of fierce debate amongst the villages. Members of the village authorities in Nyamahana contested any change to the revenue sharing, and fought for the vital role their village played in the success of the WMA: 'They know the importance of our village for conservation. The village is a gateway for hunting: MBOMIPA cannot un-invite us because poachers would come to the village' (focus group with Village Council and Authorized Association representatives, 2010). Their discursive framing of the WMA emphasized cooperation and participation and described the role of landless villages as vital to the concept of 'good neighbourliness'. 7 One informant rejected thinking about wildlife primarily in terms of the revenue it can generate, arguing that we have to learn this is not the answer because, if we go down this path, even those villages with land are different; some of them have large areas of land but no attractions for tourists; some of them have no business investments. If we keep thinking of it just in these terms, it will disrupt us a lot. (interview with Authorized Association representative, 2011) As actors from some villages attempted to remove others' rights to land, or access to benefits, and claim an increased proportion of revenue, they enacted accumulation by dispossession at the local level. This was facilitated by the commodification of wildlife resources within the WMA, which resulted in a shift in the conceptualization of the relationships between nature and society, as people viewed and valued these resources differently. Supporters of a change to the revenue distribution arrangements argued from a very commodified and financialized point of view, saying that it was their land that provided the conservation value of the WMA because it hosted the wildlife, and thus enabled the WMA to survive because these resources then generate the revenue. The commodified framing of wildlife management was well summarized by two informants, who explained that: 'education was needed in the communities because people didn't understand the meaning of the animals they were hunting. People began to learn about the money that would come from these, and that future generations would benefit also' (interview with Authorized Association representative, 2011); and that 'we were educated to know that wildlife are an opportunity for business, that can bring benefits long into the future' (interview with Authorized Association representative, 2011).
The commodified value of wildlife resources amongst local people was evident in common descriptions of animals as 'wealth' and a 'crop', and comparisons between WMA management of wildlife and the management of livestock. Actors across all levels indicated this shift by drawing on financial and market terms to discuss wildlife resources. A District Natural Resources Officer described the excitement surrounding WMAs, saying 'wildlife are "hot cakes" (Interview, 2010). The money from tourist hunting activities is very large'. Similarly, discussions with the Chairman of the MBOMIPA Authorized Association emphasized the commodification and financialization of wildlife resources in WMAs, and the dependence of the WMA and its participating villages on the development of markets and securing of investment agreements. When asked what he saw as the most important objectives for the future of the WMA and what was necessary to achieve these, he answered that MBOMIPA needs to build two luxury camps and an airport to get lots of money. After five years, we could get TzSh 500 million [USD 344,000] each year … There are lots of tourist activities we want to start … We are fortunate we have these animals (Interview, 2011). This commodification of wildlife resources is clear in a statement from a villager (quoted in Walsh 2003, 25): 'in the past, wildlife was used for bushmeat for home consumption, whereas cattle were cash assets. Now wildlife is also a cash asset, so we see it differently'.
The discourse of commodified wildlife resources among local-level actors, and the emphasis on market-based forms of exchange, indicates a shift in nature-society relations brought about by neoliberalization. The conflict is essentially a struggle amongst local actors to access the benefits provided by the WMA. We do not claim that these struggles are unique to wildlife management under neoliberalization; instead, we believe that these struggles are an inherent component of the politics of natural resource management which, in this case, have emerged within the framing provided by neoliberalization and the changes in nature-society relations this involves.
Increasingly, villages participating in MBOMIPA have begun to look for opportunities to negotiate investments with private companies for tourist accommodation. Tourists can undertake game viewing inside the WMA and national park, but not tourist hunting, which is restricted to the designated hunting block within the WMA. Some participating villages (particularly Tungamalenga; see Figure 2) had tourist infrastructure that predated the WMA, but several expressed their desire to enter into similar agreements. Thus, in 2011, one village council, with no prior tourist investment agreements, was negotiating with five potential investors to build tourist camps on village land along the boundary of the WMA.
These negotiations were driven by perceptions of wildlife as a commodity, but also by dissatisfaction with WMA revenues. These villages saw the opportunity to use the wildlife resources on WMA land to secure investment and revenue for their village, rather than having to share it with other villages through MBOMIPA. These investment negotiations fuelled the conflicts with other MBOMIPA villages. One informant said, 'people are trying to benefit privately by setting up tourist enterprises, but village investors cause much conflict because people disagree over whether the money should go to the village or to MBOMIPA' (interview with Authorized Association representative, 2011). Members of the village council in the village that was negotiating these investments justified their actions as good for both conservation and development, because the revenue that the village received would be invested in the village, which would help to maintain support for wildlife conservation and the continuation of the WMA.
These examples from MBOMIPA illustrate two important points: (1) accumulation by dispossession is something carried out not only by states, NGOs and private companies, as has previously been identified in green grabbing in WMAs, but also by local actors at the inter-village level (in this case, through some villages trying to dispossess others); (2) these processes are a product of the commodification of wildlife resources and a shift in naturesociety relations, but are also part of the struggles that local actors engage in to access benefits from wildlife resources. As Peluso and Lund (2011) argue, a 'grab' needs to be seen not as a separate process, but as the result of a series of changing contexts that facilitate shifts in land control, including exclusion, territorialization and enclosure. WMAs have altered this context, bringing areas of wildlife land into a capitalist network of marketbased exchange, thereby creating opportunities for development, and, quite literally, 'beggar thy neighbour' competition. This changing context has been drawn in to the terrain of struggle over land, power and access at the local level as different groups try to find ways to position themselves to benefit from the opportunities presented by neoliberalization.
These examples of accumulation by dispossession at the local level, and the different discourses employed by local-level actors within these conflicts, highlight the tensions within the neoliberalization of wildlife management. This is central to the changing politics of nature-state-territory at this level. As actors struggle to find and maintain ways to access and benefit from wildlife management, they exemplify the strategic adoption of discourse within natural resource politics. The result is competing discourses pitched between opposing groups at the local level that are working to secure control over land and revenue from wildlife resources.

Discussion
The neoliberalization of the wildlife sector in Tanzania has opened up new spaces to markets for tourist hunting and wildlife-related investment agreements within WMAs, and has resulted in green grabbing by the state and accumulation by dispossession with detrimental effects for some local people. However, the development of WMAs, the neoliberalization of wildlife management, accumulation by dispossession and green grabbing are not things that can be understood as simply happening externally to or being inflicted upon local communities. Instead, they are ongoing processes in which local actors are actively engaged.
Discussions of neoliberalization need to consider the space for contestation and negotiation that this process involves, and the ways actors across all levels are engaged in the reshaping of configurations of power, authority and access to land and natural resources. Here, we have referred to the politics surrounding these issues and the state-society and nature-society relationships that shape them as the politics of nature-state-territory (see Neumann 2004). Our view conceptualizes green grabbing not as a final and discrete event or act, but instead as part of the politics of the nature-state-territory relationships being played out as different actors try to position themselves to benefit from the opportunities presented by neoliberalization (Peluso and Lund 2011;Sikor and Lund 2009).
As neoliberalization progresses, new institutional forms and new nature-society relationships are produced (Büscher and Dressler 2012). In Tanzania, neoliberalization has facilitated collaborations between local actors and NGOs in WMAs in the Authorized Association Consortium, now engaged in political struggle around revenue and decisionmaking arrangements and over central state control of wildlife resources. Important features of the politics of nature-state-territory are the way local-level actors, firstly, negotiate their role in revenue sharing and decision-making arrangements and, secondly, engage in processes of accumulation by dispossession.
The strategic adoption of neoliberal discourse (see Brockington 2004) by the actors involved in both the Authorized Association Consortium and the conflicts taking place within MBOMIPA suggests complex local understandings of the values of nature. These ideas are relevant to the creation of environmental subjects through the internalization of values that define the way people care about the environment (environmentalities; Agrawal 2005). Fletcher (2010) argues that neoliberal environmentalities are produced in response to the creation of an incentive structure by the state, which then shapes people's decision-making and priorities in line with neoliberalization. Our examples suggest, however, that there is a strategic element to these neoliberal environmentalities, that actors are not always unaware of these changes and that they will try to use them to their advantage.
We agree with Gardner (2012) that for many local actors, neoliberalization creates an opportunity to generate real financial benefits from wildlife management. Whilst inequalities and injustices are apparent, for local actors, the growth of markets based on wildlife resources (e.g. photo tourism) is an opportunity that they do not want to miss. Non-consumptive tourism in WMAs substantially increased revenues received by the Authorized Associations (MBOMIPA revenues were USD 110,000 in 2011). This brought funding for community development projects, although the level of these benefits is generally disappointing. Certain local actors also benefit substantially from their involvement with the Authorized Association, particularly through allowances paid to its members. The elite capture this involves (Humphries 2013) is a central feature of CBNRM in Tanzania, as in other eastern and southern African countries with weak institutional environments (Nelson and Agrawal 2008). The commodification of wildlife resources, and the expansion of markets for these commodities, generates patterns of winners and losers at the local as well as the national level. Villages located close to the WMAs have seized the opportunity to develop investment agreements for wildlife tourism activities on their own land, in competition for revenue with the Authorized Association management of official WMA investments.
The engagement of local actors with neoliberalization is complex. These examples from Tanzania show local actors' strategic adoption of neoliberal discourse and use of the changing contexts of neoliberalization to negotiate roles within wildlife management, to position themselves to accumulate capital from wildlife resources, and to address state-society relations with respect to wildlife management. Clearly, in the examples discussed here, local actors cannot be described as simply resistant to neoliberalization and green grabbing by the state and the private sector within WMAs: local actors do not attempt to challenge the neoliberal regime of governance. Nor do local actors unconsciously take on neoliberal environmentalities. Instead, neoliberalization is simultaneously adopted and challenged. The dynamics of neoliberalization and the engagement of local actors in these processes have material consequences, both through addressing the inequalities witnessed, and in generating new sources of inequality. Tanzania is not the only country undergoing neoliberalization in CBNRM, and these insights point to the need to extend the consideration of the politics of nature-state-territory that permeate these processes and their outcomes across southern and eastern Africa, particularly in the context of the current 'crisis' of CBNRM (see Dressler et al. 2010).
It is important to consider the role of local communities not simply as passive observers of green grabbing and bearers of the inequalities facilitated by neoliberalization. Nor are local actors representative of an awaiting uprising against state control and green grabbing. Instead, we understand local communities as important agents within the processes of green grabbing and accumulation by dispossession, and engaged in the politics of natural resource management, a game for which neoliberalization is shifting the rules, and local communities are learning to play.
What we see in green grabbing and responses to it is not a new process, although as Fairhead, Leach, and Scoones (2012) point out, it can be distinguished by the number and variety of different actors involved, and the focus on novel forms of valuation and the commodification of nature. We argue here that discussion of neoliberalization and green grabbing needs to recognize the politics of nature-state-territory of which they are an expression. These politics have been shaped at different times in the past century by colonial appropriation, devolved environmental management and neoliberalization. We have witnessed these politics shift in discursive terms several times, incorporating emphases on displacement, development, participation, land tenure and, most recently, neoliberalization. Whilst the focus is currently on green grabbing, it is important to note that this is a current expression of a fundamental concern and struggle over natural resource governance. As politico-economic philosophies change over time and new modes of environmental governance emerge, the terminology we use to describe these politics will undoubtedly change again in the future, and the context in which they are played out will continue to shift, but these struggles and their fundamental concerns over state-society and nature-society relations with respect to land and power will remain.