Power Struggles in the Management of Wildlife Resources The Case of Burunge Wildlife Management Area, Tanzania

Through a cross-sectional research design, this study examined power struggles in Burunge Wildlife Management Area (WMA) in Tanzania. Four out of ten villages com-prising the WMA were purposively selected, and data were collected via focus group discussions, key informant interviews, questionnaires to household heads, and a litera-ture review. Results showed that the central government, investors and non-government organisations held institutional and strategic powers, while the democratically elected Village Councils held structural powers and lost most of their pre-WMA institutional powers to a legally required new institution, the Authorised Association. Therefore, Village Councils lost influence on strategic, institutional and management decisions per tinent to the WMA and their constituencies’ livelihoods. Accordingly, Burunge WMA de-democratised wildlife management by eroding the relevance of Village Councils to their constituencies. The study also found power struggles over revenues, land management and access to resources among the stakeholders, mainly due to a divergence of interests. However, there was no conflict management mechanism in place. Hence, we recommend that the institutional powers to establish, govern and dissolve WMAs should go back to Village Councils. The purpose is to establish economic incentive structures that promote (i) wildlife conservation, (ii) an equitable distribution of associated costs and benefits between Village Councils forming WMAs and (iii) an equitable distribution of costs and benefits between WMAs and higher levels of government as well as international con servation NGOs.


Introduction
'Power' in the context of natural resource management has no single universally accepted definition. For instance, some scholars characterise power as the relationship between people [1], while others relate it to coercion, constraint, consent, social relationship and interaction [2]. However, power occurs in various forms including structural, strategic and institutional [3][4][5]. Structural power, or domination, is asymmetrical, hierarchical, hard to challenge and difficult to reverse [3]. The subordinate actors in structural power arrangements have little room for manoeuvre because their margin of liberty is extremely limited [5]. Strategic power involves structuring possible fields of action of others, and it is expressed through daily interactions between individuals and groups [3]. Accordingly, strategic power has many expressions, including economic ability, rational and ideological argumentation or manipulation [4]. Officialised institutional power is systematised, regulated and involves mandates that are legally defined [3,5].
Furthermore, power is often exercised through multiple, intertwined institutions 1 following what is socially possible or acceptable [5]. Power analysis must involve observation of actors in decision-making [2,3,5,6]. Therefore, stakeholders' power can be understood as the extent to which they can persuade or coerce others to accept decisions and follow certain courses of action [7,8]. This may also involve actors resisting the management actions over resources. Counterpower such as foot-dragging occurs in situations where the less powerful try to express their concerns. For example, Scott [9] revealed that in developing countries, communities rarely choose open confrontations against onerous new laws but instead resort to more subtle actions of everyday resistance including foot-dragging and hidden non-compliance, that is, by using the weapons of the weak. When this happens, it is usually a sign of great desperation [9].
In Tanzania, power struggles among actors over decision-making related to wildlife management have lasted for decades [10]. For example, the lucrative nature of wildlife resources has created strong incentives and vested interests for the central government and the associated bureaucracy to maintain control over the resource [11]. As a result, institutional reforms or legal adjustments officially intended to enhance local communities' rights to benefit from wildlife have often been circumscribed to reassert the state's ultimate power in matters of wildlife governance [8]. Over time, these reforms and legislative adjustments have produced asymmetrical power relations in the way decisions are made about wildlife resources and how benefits from wildlife are shared among the actors (ibid.).
The United Republic of Tanzania (URT) has decentralised wildlife management through the introduction of Wildlife Management Areas (WMAs) [12]. The Wildlife Policy of Tanzania defines WMAs as a new category of protected areas for community-based wildlife management.
WMAs can be established when a group of Village Councils agree to set aside adjacent parts of their respective village lands to form a WMA [12]. However, the process that leads to a fully functioning WMA is long and bureaucratic. It begins with the formation of a community-based organisation among the involved villages, which among others must prepare a draft so-named Management Zone Plan for the WMA area. When the Director of the Wildlife Division endorses the Management Zone Plan, the community-based organisation is granted wildlife user rights and becomes a legally recognised institution, called an Authorised Association 2 , which is authorised to manage its WMA [13]. With decentralisation by devolution as envisioned in the 1998 Wildlife Policy [12], the Authorised Association is expected to be accountable to the local communities. However, empirical evidence has revealed that Authorised Associations are more accountable to the central government ( [14,15], see also below). This contradicts Ribot's definition of devolution, which 'involves transfers of power to elected local authorities and enables local people to make decisions for themselves through their representative local authorities' [9,16].
Before the introduction of WMAs, other functionally protected areas included National Parks and Game Reserves where no human settlement is allowed and Game Controlled Areas as well as the Ngorongoro Conservation Area-where settlement is allowed albeit under tight regulation [17]. National parks are managed by the Tanzania National Park Authority (TANAPA); game reserves and game controlled areas are managed by the Tanzania Wildlife Management Authority (TAWA); and WMAs are managed by communities through Authorised Associations, which, except for the secretary and trustee of each Authorised Association, are democratically elected community-based organisations. Therefore, WMAs are a form of common property regimes (CPR), whereby communities are expected to sustainably manage and benefit from wildlife [18,19]. However, decentralisation usually threatens some actors' powers and interests. As a result, official power transfers from the central government to local actors are often resisted overtly and covertly by different actors [20].
Currently, Tanzania has 38 WMAs at different stages of development [21]. Burunge WMA was among the first nine pilot WMAs in Tanzania. It was launched in 2003 and granted official status in 2006 [22,23]. Burunge WMA has great significance to Tanzania's protected area network given its function as a wildlife corridor linking Tarangire National Park, Lake Manyara National Park and Ngorongoro Conservation Area [22,23]. The importance of Burunge WMA has attracted numerous actors to participate in its management. Divergent roles, interests and expectations of these actors are, however, major causes of power struggles, which unfortunately is not unique. For example, Shilereyo [24] reported a situation where local people had no power to influence management decisions in a WMA, while Manyika et al. [25] reported power struggles over forest governance. Also, Bluwstein et al. [14] showed how the process of rule-making and rule-changing in WMAs generated power struggles. However, analyses addressing issues of interests, power disparity and power struggles among the actors in community-based conservation are relatively few. Thus, our understanding of the outcomes including the management of such power struggles is inadequate.
This study examines power struggles among the key actors in Burunge WMA, specifically by attempting to answer the following questions: who are the actors and what are their roles and interests in the management of Burunge WMA? What powers (structural, strategic and institutional) do these actors possess and how do they affect the management of the WMA?
How do these powers determine the outcome of struggles among the actors? And how are these power struggle conflicts managed?

Theoretical framework
Drawing on Raik et al. [2], Nuijten [3], Kajembe et al. [5] and Manyika et al. [25], this study focuses on identifying actors including their roles, interests and powers to understand power struggles and their outcomes in natural resource governance. The emergence of conflicts is an important indicator of power struggles among actors [2,25]. In this study, power means the relationship among the actors, while power struggle denotes a negative situation in the course of decision-making where certain actors at the expense of other legitimate actors push forward their interests in the management of WMAs. Also, the study draws on common property theory dealing with the governance of common pool resources. The theory is expounded by the principles of subtractability/rivalry (that one person's resource use diminishes what is available for others) and excludability (that people are divided into 'insiders' who have rights to a defined resource and 'outsiders' who do not have rights) [19,26].

Description of the study area
The study was carried out in Burunge WMA located in Babati District in Northern Tanzania (Figure 1). The H-shaped WMA was established in 2003 and covers about 283 km 2 . It is formed among 10 villages, namely Mwada, Sangaiwe, Ngoley, Vilima Vitatu, Kakoi, Olasit, Manyara, Magara, Maweni and Minjingu. The residents in these villages practice livestock keeping and small-scale farming as their main economic activities. Their main food crops are maize, beans, bananas, paddy rice, potatoes and millets. Sesame is the major cash crop. The area experiences a bimodal rainfall: short rains from May to June and long rains between November and January. The annual rainfall ranges between 400 and 500 mm, while the temperature ranges between 18 and 33°C [27,28].

Research design and data collection methods
The research was carried out between 2014 and 2016 using a cross-sectional design. Four villages, Mwada, Vilima Vitatu, Minjingu and Kakoi, were purposely selected for this study. The main criterion for selection of the study villages was residents' level of acceptance of Burunge WMA and associated arrangements of access to resources (i.e., satisfied or dissatisfied). For example, preliminary survey results showed that Mwada villagers were fairly satisfied with the WMA, while villagers in Minjingu, Vilima Vitatu and Kakoi (a relatively newly formed village separated from Minjingu) were dissatisfied and wanted to withdraw from the WMA.
For triangulation purposes, the study applied several methods of data collection including focus group discussions, key informant interviews and a questionnaire. Focus group discussions were conducted with Village Council members, with youths, women and senior men from each village. Selection for the focus group discussions was conducted so that each group had different interests, positions and roles in the management of natural resources [29]. Key informant interviews were conducted up to the point at which no new information was obtained, that is, the data saturation point [30]. Informants involved were respondents who had specific knowledge about Burunge WMA including the village chairpersons, senior men and women, the Burunge Authorised Association and Babati District and Wildlife Division officials.
Focus group discussions and key informant interviews sought to identify powers held by the key actors and uncover emerging struggles originating from these power differences. The powers of different actors in the WMA were identified through a stakeholder analysis that uncovered stakeholders with vested interests in the current situation and stakeholders who were likely to be affected, either positively or negatively, in case conditions change or remain the same. The questionnaire survey involved respondents drawn randomly from the village registers (n = 140 households in total) [31]. The questionnaire had both closed-ended and openended questions. Along with key informant interviews and focus group discussions, the questionnaire survey was used to assess the existing resource use conflicts. Secondary data, mainly law texts that related to actors, powers of actors, power struggles and conflicts were reviewed.

Data analysis
Content analysis was used to analyse data related to the institutional and different categories of power possessed by different actors. In this case, all qualitative data from focus group discussions, the questionnaire and key informant interviews were categorised into related themes [32]. Underlying themes in the categories of words were identified, explained, clarified and interpreted [33].

Key actors in the management of the WMA
The study revealed various actors with different roles and interests in the WMA ( Table 1).
The interests were both overlapping and divergent. However, all actors supported the objective of wildlife conservation. The local communities viewed conservation as an opportunity to improve their livelihoods, while the central government, through the Wildlife Division, enjoyed the expansion of conservation territory as well as improved revenue collection Actor

Roles Interests
Village Assembly Collective decision-making such as approving the size of land to be allocated for WMA establishment.
Conservation of wildlife; accruing benefits from the WMA and improvement of livelihoods.

Representing villagers in all matters
about the villagers' well-being.
Protecting their fellow villagers' rights and promoting their interests.

Authorised Association
Managing the WMA on behalf of the villagers.
Act as a power broker between the villagers and the central government agencies and local government.
Ensuring the conservation of wildlife and that associated benefits are realised at WMA level.

District Council
Oversee management of the WMA. Making sure wildlife is conserved. Generating benefits from the WMA.

Wildlife Division
Making rules for WMA management.
Increasing the area of protected landscapes including wildlife corridors where most WMAs are located.
Collecting revenues from tourism investments in the WMA.

Non-Governmental Organisations (World Wildlife Fund for Nature, African Wildlife Foundation)
Facilitating WMA establishment. Conserving wildlife and ensuring the increase of protected landscapes including wildlife corridors between otherwise protected areas.

Investors (Maramboi and Burunge tented camps)
Entering into contracts with the Authorised Association to undertake business ventures in the gazetted

WMA.
Generating profits through wildlife conservation.

Conserving Wildlife
Ensuring that Burunge WMA acts as a buffer zone and corridor.
Source: field data. The Wildlife Division has three major roles: policy formulation, regulation and coordination.
Until recently when Tanzania Wildlife Management Authority was established by government order, the Wildlife Division was responsible for overseeing the management of wildlife outside national parks and the Ngorongoro Conservation Area. However, since the formation of WMAs was based on the promise of enabling people to manage wildlife on their lands and benefit from this resource, the role of revenue collection assumed by the Wildlife Division contradicted expectations of the local communities, whom government officials during the sensitization process verbally promised full control over the WMA and a say over the revenues and other benefits. By assuming the role of revenue collection, our respondents felt that the Wildlife Division had grabbed an intended power of the Authorised Association-defined in the WMA Regulations as 'a community-based organisation, whose primary objective is to conserve wildlife resources for the benefit of local community members ordinarily residing in that particular area' [13].
Besides its conservation role, the Authorised Association is intended to act as a power broker between the local communities and the central/local government agencies. However, the Authorised Association was found mostly to represent the interests of the central government. These findings are similar to another study in the area by Kicheleri et al. [15] where the Authorised Association was found to be accountable to the central government, quite in contradiction to the WMA Regulations, which require Authorised Associations to be accountable to the local communities. Accordingly, the interests of the central government take precedence over those of the local communities. The Tanzania National Park Authority's (TANAPA's) interests are that Burunge WMA serves as a buffer zone to and a corridor between 'its' national parks.
Kajembe et al. [5] contend that in the management of natural resources, triangles of actors including state agencies, NGOs (Non-Governmental Organisations) and local communities are created where power struggles take place. This happened in this case as well where procedural flaws have clouded the Burunge WMA establishment process during which some actors pushed forward their interests without due regard to other actors' interests or formal democratic procedures. For instance, the conflicting interests and power differences among communities, donors, the central government and the African Wildlife Foundation (AWF) negatively affected the AWF's work with communities in Burunge WMA [15,34]. Instead of acting as a power broker between the central government and local communities, the AWF found itself in conflict with the local communities because it sided with representatives of the Wildlife Division who allegedly manipulated and misled Village Councils to push for their acceptance of the WMA [15]. Accordingly, proper identification of actors, their roles and interests from the very beginning of a conservation-through-decentralisation project is quite vital. The reason being that this allows other actors and stakeholders to assess the likely consequences (for them) when the kind of decentralisation they are asked to support entails a transfer of specific bundles of powers from the centre to particular institutions that are more or less downwardly accountable [16].

Powers of key actors in Burunge WMA
From these results, the key actors, that is, Village Councils, the Authorised Association, District and Wildlife Division officials, NGOs and TANAPA, and investors exercised different types of powers ( Table 2). Essentially, there were three types of powers possessed by these actors. The Village Councils had structural power, while investors possessed strategic power. The Authorised Association, District Council, TANAPA, NGOs and the Wildlife Division had both institutional and strategic powers. The structural power relates to cultural or social positions such as being an elder in the society, while strategic power relates to level of education and/or wealth. The institutional power relates to positions in the government, organisations or associations [2][3][4]6].
The communities that collectively make decisions at the Village Assemblies are represented by the Village Councils, that is, the democratically elected Village Governments. Therefore, the structural power possessed by the Village Councils is in line with the findings of Kajembe et al. [5] who assert that structural power is widespread in traditional societies. This is because structural power is associated with social position such as being a household, clan or tribal head. It is further associated with a cultural position such as being a traditional healer in a village [6]. However, in this context, the village's liberty to influence the WMA management decisions was limited. Structural power cannot override strategic or institutional powers unless full discretionary powers are devolved [18]. Therefore, to influence multiple levels of governance, structural power needs the backing and support of institutional or strategic powers. Unfortunately, the exact opposite tends to happen, which increasingly restricts those holding structural powers to realms of cultural and religious rather than socio-economic significance.
With democratic decentralisation, it was expected that institutional power over decision-making in the WMAs would be shifted to the Village Councils whose members are democratically elected and thus downwardly accountable to the local communities. However, contrary to this expectation, the Authorised Association, District Council and Wildlife Division have the institutional power to make most decisions regarding the WMA management at the village level. These findings suggest that the rights of local communities to exercise institutional power and manage the WMA were undermined. Thus, a Village Council's ability to influence WMA management  decisions was limited by the institutional setup of the WMA. This may be associated with the lack of enforcement of rules that are centrally enacted on the ground together with central government interests to re-centralise (while decentralising) WMA management. In support of this argument, Kicheleri et al. [15] found that WMA related meetings were not conducted in the villages to communicate WMA management performance and progress.
The WMA Regulations Section 17 describes the responsibilities of the Village Council in the management of Wildlife Management Areas [13]. Section 17 (g) states that the Village Council shall 'monitor the activities of the Authorised Association and report to the Village Assembly and District Council'. Section 17 (i) states that the Village Council shall 'ensure that the Authorised Association implements sectoral policies while entering into agreements on the management of a Wildlife Management Area'. Contrary to these statements, however, the WMA Regulations' Section 24 excludes Village Councils as members of the District Natural Resources Advisory Board (DNRAB) [13]. We argue that excluding Village Councils from this Board attenuates their ability to influence Authorised Associations' governance of WMAs.
Investors hold strategic power, which gives them access to investment opportunities in the WMA. Mbeyale [4] identifies wealth among the sources of strategic power, whereas Kelly [35] sees the market value of tourism business as high and, therefore, a key area of investment by actors with strategic power. Essentially, tourism investors want to stay in the market; hence, they employ their strategic powers to stay in business and pay taxes to the Wildlife Division rather than to Village Governments as they did before the WMA was established. Hence, tourism investors often find themselves hijacked in a struggle over tourism-based revenues between the central government and Village Councils.
The NGOs, namely the AWF and the World Wide Fund (WWF) for nature both held institutional as well as strategic powers. In the establishment of WMAs, NGOs are claimed to have used their position to manipulate local communities to accept WMAs [15]. Comparable observations are apparent for REDD+ projects in Tanzania where financial resources were used to earn NGOs higher bargaining power [25]. Similarly, instead of facilitating the process of establishing Burunge WMA through meaningful public participation, the AWF alienated local communities by favouring interests of the central government and donors in expanding the area for nature conservation [34].

Struggles over revenues from the WMA investments
In the WMA, benefits stem from non-consumptive utilisation such as tourist lodges, photo- According to one of the Wildlife Division officials, this change was instituted after learning that the Authorised Associations had inadequate capacity to collect the revenues. Allegedly, this deficiency resulted in minimal revenue collection from investors. These observations are in line with those of Ribot et al. [20] on how governments justify re-centralisation control over revenue streams from community-based programmes. Whether the alleged poor capacity of the Authorised Association to collect revenue is true or not, an effective community-based programme must include fiscal devolution, otherwise, the discretionary powers of locally elected leaders become irrelevant to their constituency's livelihoods [16]. The changes in revenue collection also provoked complaints from the Village Councils who claimed that their share of revenue had decreased, which indeed was the case from 2011/12 to 2012/13, c.f. Table 3.
However, the most important messages from and related to Table 3 are that: (i) it was difficult to get access to the data, which is held at the Authorised Association's office, (ii) in most years, the Authorised Association seemed to accumulate substantially larger amounts of cash from the WMA than what any Village Council received and (iii) during Burunge WMA's 12-year history, the Authorised Association appears to have accumulated a surplus of more than TZS 833 million or on average just above TZS 69 million per year, which is more than three times the average annual amount received by the individual villages. Accordingly, the first author attempted to access the Authorised Association's audited accounts to verify its bank balance, but that was promptly stopped. Her subsequent requests to access the books of the Authorised Association were categorically denied.
After 2012/13, when the central government had re-centralised WMA revenue collection, official revenues going to the Village Councils picked up. Although positive in a narrow sense, this fiscal re-centralisation is problematic, c.f. above. Further, our respondents generally agreed that the Authorised Association's capacity to govern the WMA and associated revenue flows needs improvement as does public access to scrutinise the Authorised Association's accounts.
The re-centralisation of WMA revenue governance is evident in the WMA Regulations and associated legislation, which specify how revenues from a WMA must be shared. The Non-Consumptive Wildlife Utilisation Regulations Section 19 (b) stipulates a sharing mechanism of revenues from the non-consumptive operations in WMAs, stating that '5% shall be directed to the District Council, 25% to the Director of Wildlife and 70% to Authorised Associations' [36]. Non-consumptive utilisation includes game viewing and other ecotourism investments in WMAs including tourist lodges and hotels. Likewise, the WMA Regulations Section 48 (8) states that 'the income generated from resident hunting in WMAs shall be shared as follows: (a) Authorised Association 40%; (b) responsible District Council 60%' [13,31]. Moreover, the Twelfth Schedule of the WMA Regulations provides a   Table 3. Therefore, to prevent anger and frustration among villagers, the investor had to renegotiate his position and contribute to village development activities in Minjingu over and above previously agreed annual payments that went to the Authorised Association and subsequently to the central government. This enabled the investor to stay in business but reduced his profit margin.
Benefit-sharing programmes that do not address problems at the household level are likely to become unpopular no matter how much money is injected into them [38].
Unfortunately, costs in the form of wildlife causing crop damages and restrictions on traditional uses of resources within areas set aside for the WMA rather than benefits have materialised at the household level in Burunge WMA [39]. Nevertheless, the WMA Regulations Section 17 (e) specifies that the central government is mandated to determine the share of generated benefits that local communities are to receive from village land investments [13]. This clearly contradicts Section 26 of the Local Government District Authorities Act No. 7 of 1982 that mandates Village Councils to be corporate bodies, which are free to earn benefits from investments on their land and to determine how to use the revenues for their own development [40]. In this context, however, the structural power of the Village Councils cannot override the strategic and institutional powers possessed by the Authorised Association, the District Council and the Wildlife Division. This imbalance is bound to cause conflicts.

Power struggles over the management of Burunge WMA
The general narrative presented to Village Councils in Burunge WMA suggested that the decision to set aside part of their village lands for conservation under a WMA would not affect their control over these lands. However, the strategic powers that the Authorised Association share with the District Council and the Wildlife Division has largely made it the, de facto, governing body of the WMA area. Even though the villagers officially form it, the Authorised Association distanced itself from the communities and acted mostly in the interest of outsiders. The WMA Regulations' Section 17(a) states that village councils shall 'provide land for the designation and establishment of a Wildlife Management Area' [13]. Accordingly, Vilima Vitatu's Village Council committed 65% of its village land to the WMA during the WMA initiation process, that is, 12,830 ha out of 19,800 ha. Simultaneously, the village population increased from 1,323 in 2003 to 3,281 people in 2014. Therefore, the village wanted to reclaim part of its land from the WMA to offer livelihood opportunities to its citizens. That, however, proved to be far from straightforward. The WMA Regulations Section 37(a) states that a WMA will cease to exist when there is 'a change of use in the Village Land Use Plan' [13,26]. However, since WMAs are formed by many villages, one or several of these are not free to independently change the status of the part of their land that falls within the WMA. The WMA Regulations are silent on what should happen in a situation where one or several villages change their land use plan and continues to be party to the WMA, but Kicheleri et al. [15] found that when a WMA is gazetted, the land within the WMA appears to change legal status from village land to reserved land. Thus, if Village Councils like that of Vilima Vitatu want to modify their land-use plan, they wake up to a realisation that they have lost control over, the part that falls within the WMA. Meanwhile, the WMAs generate substantial revenues from different investment ventures. As mentioned, 50% of these revenues are controlled by the Authorised Association, not the Village Councils. Hence, although Village Councils hold de jure institutional power, the WMA institutional arrangement severely limits their control over WMA revenues (strategic power), thereby undermining their relevance to their constituencies among whom many feel under-compensated for losses they have suffered because of Burunge WMA. Adding pain to injury, the Authorised Association leaders are, in comparison to Village Council leaders, paid generous allowances to attend meetings and workshops in different parts of the country which makes their positions more prestigious and financially attractive than those of Village Council leaders.
An interview with the Wildlife Division officials regarding the management of WMAs revealed that the central government's approach was to progressively build the capacity of Authorised Associations to fill their roles. Nevertheless, Village Council members from the study villages claimed that the Authorised Association had thus far failed to communicate its mandate and obligations to local communities. These findings correspond with those of Kicheleri et al. [15] who found noncompliance with WMA Regulations by Burunge Authorised Association, which failed to hold meetings with villagers. Village Councils and the Authorised Association both hold institutional power, but they differ in the sense that Village Councils have structural power over the people, while the Authorised Association has strategic power (money). Hence, including Village Councils in Authorised Association meetings could challenge the latter's strategic power. Formally, nothing prevents Village Councils from seeking such influence, but information gaps and lack of clear communication channels among the WMA actors have thus far prevented this from happening in a structured manner, which is a major cause of frustrations and conflicts [41].
In short, Burunge WMA has, contrary to its official objective, attenuated the democratically elected Village Councils' powers over their village lands as well as their power over revenue Wildlife Management -Failures, Successes and Prospects streams from such lands. Village Councils lost their original authority over land and revenue to a newly created and supposedly democratic institution, the Authorised Association, which has subsequently lost part of its authority over WMA revenues to the Wildlife Division and District Councils. Notably, this followed a 'rule through law' process the dangers and consequences of which Village Councils did not realise before it was too late. Intended or not, this looks like 'divide and rule' tactics by the central government to re-centralise while decentralising wildlife governance. Following Ribot's [16] definition of democratic decentralisation: The transfer of meaningful powers over resources and associated revenue streams from the centre to responsive, representative, and downwardly accountable local institutions, Burunge WMA has done the exact opposite vis á vis Village Councils.

Power struggles over accessing and using resources
The establishment of Burunge WMA has imposed opportunity costs on local communities by prohibiting their access to natural resources such as water, thatching grasses, building poles, dry season grazing areas, firewood and Doum Palm (Hyphaene compressa). However, extraction of some resources such as firewood was allowed on special occasions such as burial ceremonies. Bluwstein et al. [14] reported a similar situation as well as the resulting resistance from local communities within Burunge. In this respect, Kakoi village is a good example. Here, people applauded the idea of establishing the WMA following a verbal promise by the Wildlife Division's representative that the villagers' access to dry season fodders within the WMA area would be maintained. However, unknown to the villagers at that time, plans for tourist hunting in the very same area were also under development. This may explain villagers' subsequent claim that they were hardly involved in the detailed and formalised processes of developing a management plan to legally establish Burunge WMA, c.f. Kicheleri et al. [15]. Village Assemblies and Councils were involved during the sensitisation process but excluded during the preparation of the Management Zone Plan and General Management Plan for the WMA. As mentioned, the official recognition of the WMA area as well as the Authorised Association's authority to manage it is contingent on the Wildlife Director's official endorsement of the Management Zone Plan, which identifies areas for specific uses and set limitations on other forms of use depending on the purpose and sensitivity of the resources. Unsurprisingly, tourist hunting turned out to be incompatible with local resource uses due to risks of accidental injuries or casualties and disturbance of the wildlife targeted for hunting. Accordingly, when the investor complained to the Authorised Association that villagers' collection of fodder and grazing of livestock impacted his tourist hunting business negatively, the Authorised Association decided to criminalise these activities through a revision of the WMA Management Zone Plan. So, on the (verbal) promise of continued access to key resources, the people of Kakoi village were mobilised to support the establishment of Burunge Authorised Association, which subsequently used its legally defined authority to deny people access to these very resources.

Conflict management in Burunge WMA
Power struggles have fuelled numerous conflicts in Burunge WMA. However, our respondents in the four case villages distinguished between 'conflicts', which they identified as situations where conflicts of interests led to open confrontations and 'problems', which they identified as circumstances that had made villagers' lives harder but not led to open confrontations. Rules on rights to land and environmental resource use were identified as the source of conflicts among pastoralists, the investor (Maramboi and Burunge Tented Camps) and the Authorised Association ( Table 6). The conflicts were primarily over access to grazing areas and fodder. Frequently these conflicts were associated with problems of access to (potential) agricultural land, water, building poles, firewood, and thatching grass as grazing lands provide these opportunities and environmental products too.  [42] and Fairhead [43] who assert that land use conflicts are ubiquitous in CBNRM in Africa. However, this was not expected in the WMA context as CBNRM programme principles aim at giving local communities more rights to resources and associated revenues as well as rights and opportunities to manage conflicts, compared to 'fence-and-fine' conservation approaches [44].
This study also uncovered several problems that villagers perceived were directly related to the WMA and the Authorised Association. Villagers claimed that crop raiding had increased since the WMA was established. As described by a man in Kakoi village: '… the elephants come in a herd at night before raiding a pumpkin farm. A few would come close to the farm and walk around. When they realise there is no danger, they will give a trumpeting that will call the rest of the herd into the farm'. Villagers complained about lack of compensation following crop raids. Likewise, fishermen claimed that the Authorised Association had taken over the responsibility of the Beach Management Unit (BMU) [45] denying them their livelihoods, where in reality the BMU should have been run by the communities. They further accused the Authorised Association of assuming control over Burunge Lake.
Villagers also accused investors of fuelling conflicts in Burunge WMA, presumably due to the inherent conflict potential of their conflicting land-use interests. Thus far, none of the Source: household interviews (n = 140). and the Police. Also, some villagers deal with problems simply by extracting products in contravention of the WMA rules, that is, through everyday forms of resistance. This multitude of approaches to deal with conflicts and diverging interests is rather bewildering and contrary to the recommendations of, for example, FAO [46], according to which participatory mechanisms for conflict management are critical for successful natural resources management.

Conclusion and recommendations
In the management of Burunge WMA, actors had different types of power and to some degree diverging interests, which features among the major causes of resource conflicts and power struggle between the key actors: the Village Assemblies, Village Councils, Babati District Council, NGOs, TANAPA and the Wildlife Division [47,48]. Overall, Village Councils, the pre-WMA existing village-level democratic institutions, lost powers over their village lands and associated revenue streams-mainly to the Authorised Association but also to the Wildlife Division. All other actors held and gained, via legislative and management rule revisions (rule through rather than the rule of law), strategic and institutional powers that enabled them to influence WMA management decisions and dominate village citizens' interests. Thus, and quite contrary to its official objectives, Burunge WMA has de-democratised wildlife governance by undermining the importance and relevance of Village Councils to their constituencies. Investors had strategic powers (money) that enabled them to survive amid struggles and conflicts that inevitably involved their business interests. NGOs' strategic and institutional powers enabled them to promote their conservation interests. Specifically, power struggles among the actors were about the generation and distribution of costs and benefits, which resulted from the WMA rules and regulations. Due to power struggles, conflicts emerged over the access to and use of WMA resources. However, conflict management mechanisms pertinent to the WMA were absent.
Consequently, this study recommends that institutional powers must be returned to the Village Councils, whose responsibilities include all aspects of their constituencies' livelihoods. Accordingly, Authorised Associations must be formed by, not in addition to, the group of Village Councils that have set aside parts of their village lands to form a WMA. Based on the fundamental principle that WMAs should always be more profitable to Village Councils than realistic alternative uses of their village lands, such new Associations (authorised by the Village Councils as well as the Wildlife Division) should collect and redistribute WMA revenues. Relevant district-level and state-level actors should also get a share if and only if this does not violate the above principle. Accordingly, legislative amendments must secure individual Village Councils permanent and widely defined rights to withdraw from WMAs.
Also, the group of Village Councils that contribute land to a WMA should hold legal rights to dissolve the WMA. Both instances must result in the reinstatement of Village Councils' pre-WMA rights to their village lands and related resources. Such legal rights of Village Councils to pull out of or dissolve a WMA would fundamentally change the current economic incentive structure. First, if the WMA yields an overall income that surpasses the sum of perceived net benefits from realistic alternatives, then the group of Village Councils would face a strong incentive to negotiate viable short-and long-term compromises for governing the WMA including how to spend and share common funds among them. Second, the central government, conservation NGOs and potential donors would face a compelling incentive to uphold the moral justification for WMAs: that they remain the most attractive land use for rural communities-in practice, not just potentially or in theory.