Expressing agency in antagonistic policy environments

ABSTRACT Studies have shown how centralized institutions in resource management lead to adverse impacts on communities. However, important questions remain about the mechanisms through which people thrive under unfavorable policy environments. This paper examines how locals around the Mount Cameroon National Park (MCNP) carve space for articulating their agency despite unsympathetic management initiatives. Using focus group discussions and thematic analysis of data from 17 villages, the results identified arrangements that do not entirely solicit local consent in natural resource management. In this context, locals express their agency holding onto cultural traditions through religious engagement and the use of economic incentives to enhance livelihoods. They do so with the aid of traditional institutions, awareness of system challenges, acting more wisely, and initiating constructive needs in remote areas.


Introduction
Implementing conservation policies for natural resource management requires a proper inclusion of indigenous/local people who hold local knowledge, i.e. experiences adapted to the local culture and environment, embedded in community practices, institutions, rituals and relationships, that change over time (FAO 2019). Such knowledge is integral to resource use practices, held by societies that have a long history of interaction with their natural surroundings, which is crucial for decision-making (UNESCO 2017). This has been the view shared in global policy frameworks such as the Akwé: Kon Guidelines (Markkula, Turunen, and Kantola 2019) and the UN Sustainable Development Goals (SDG) (Tegegne et al. 2019). Both frameworks advocate for the social inclusion of indigenous/local people in development that might have an environmental, social, and cultural impact on local communities. However, these legalities have often not lived up to expectations when state institutions apply policy initiatives for resource management at national and local levels of society (Hirsenberger et al. 2019;Markkula, Turunen, and Kantola 2019;Owuor, Icely, and Newton 2019;Ramcilovic-Suominen 2019;Sen and Pattanail 2019;Sloan et al. 2019).
It is noted, for example, that in this unfavorable policy context, management institutions request local people to comply with regime practices rather than values and beliefs of the local community (Nadasdy 2003(Nadasdy , 2005. Even so, where an overlap exists between cultural ideologies of locals and political interest of the state, practices of natural resource management end up reflecting the priorities of powerful parties instead of local people (Borrini-Feyerabend et al. 2004). Scholars have equally shown that management institutions that obtain shared power from central governments become more powerful and are inclined to exert control over marginalized people (Spaeder and Harvey 2005). Without a critical study of these antagonistic processes, we risk extending the difficulties communities face in resource management.

Centralized resource management
Several cases on the top-down management of natural resources seem consistent with the views outlined above. For instance, in countries of former Yugoslavia, the lack of open constructive criticism and feedback between conveners of natural resource management projects and local people has led to ineffectiveness in conserving cultural heritage (Hirsenberger et al. 2019). In Laos, ethnic minorities are losing access to fundamental subsistence livelihoods due to the Laos Central government's reinvention of Reduction of Emissions from Deforestation and Forest Degradation (RED+) initiativesto issue funding for forest bureaucracy, large-scale land concessions, and activities of development industries (Ramcilovic-Suominen 2019).
Studies also observed that despite legalities for conserving National Parks in Northern Finland, the Sámi continue to face an increase in land pressures from industrial activities, due to inadequacies of state laws that do not fully protect Sámi lands from encroachments (Markkula, Turunen, and Kantola 2019). Consider the example of the Trans-Papuan Corridor in New Guinea. Here, the Papuan government backed a mega rice project on peatlands, including mining, oil, and gas extraction projects in protected areas. This use of the land has been part of a state-led development agenda to promote resource-extraction (Sloan et al. 2019). In effect, this has purported indigenous Papuans to put forth customary land claims as many of them await formal concessions for their land loss in state-protected forests.
Further, on the example of the El Vizcaino Biosphere Reserve in Mexico, several stakeholders jointly formed Advisory Councils (ACs) since the 1990s, for collective decision-making and implementation of natural resource management, though, steered under centralized institutions. A recent study, however, showed that, despite this move, ACs have little power because they rely on resources from government agencies which are often reluctant to approve decisions that do not reflect the political interest of the state (Brenner 2019). Thus, even with the participation of persons who represent local resource users, ACs fail to integrate the genuine interests of local people.
The posing of these case examples is important for revealing the adverse outcomes of policies for natural resource management in different communities. However, we suggest that, within this setting, it will be useful to explore the mechanisms that enable people to thrive under undesirable resource management. This could help offer ideas for policy improvement.

Choice of study and the niche for collective agency
We find the case of Mount Cameroon National Park (MCNP) and adjacent communities significant because, given that UN protocols promote biodiversity conservation in ways that should properly include local/indigenous people in decision-making, the regime in this case, however, implements natural resource management in ways that seem antagonistic to cultural heritage. There is an urgent need to investigate whether locals adversely affected by state initiatives for natural resource management, find other means to convey their agency.
Agency is not a new concept. In contemporary Sociology, Anthony Giddens' work on agency describes the contingent relations between structure and action, and, its impact on the cultural formation (Karp 1986). Here, the structure comprises of all arrangements that stimulate or limit opportunities and choices in societyand eventually shapes human behavior. In general, agency can be defined as a transformative concept about how a group or individual is capable of acting in the face of challenges, through the lens of various contexts. This might include relations of deriving power (Drydyk 2013;Hanmer and Klugman 2016); assigning responsibilities to bring change (Dowding 2008); habits that develop in response to dynamic environments (Ransan-Cooper (2016); and collective action in instances of community resistance and dialogue (Stammler and Wilson (2006, 22).
Previous scholars in the conservation and resource management literature, have used agency to explore the degree to which locals shape their engagement with different systems, and how this engagement may enable or disable the capacity of local people to act (Chirozva 2015; Cole et al. 2019). Also, how locals cultivate awareness and motivation to participate in conservation programs, such as the World Wildlife Fund (WWF) and Wine Initiative (BWI) in South Africa (Honig et al. 2015). Accordingly, agency is not always overt action or active community engagement to change something, but, includes practices of alternative behavior on the part of local people. In this paper, we explore this explanation based on collective agencya concept where behavior is not only due to discontent against an undesirable management system, but also, how economic benefits and religious tradition justify the behavior of local people. In doing so, we include more passive approaches such as expression of frustration or grievances in our definition of agency by targeting three specific objectives: (1) To identify the complex nature of the MCNP system of natural resource management.
(2) To examine how locals living around MCNP convey their agency beneath this adversarial policy environment.
(3) To determine the enabling mechanisms of agency under which locals carve space for thriving in unfavorable management situations.

Materials and methods
Study site and governance structure MCNP is located in the southwest region of Cameroon, sub-Saharan West Africa. The park is host to Mount Cameroon, which is part of the eight-biodiversity hotspots in the Gulf of Guinea. Its peak is 4,100 meters above sea level and is located at latitude 4°N of the Equator and longitude 9°E of the Meridian. The annual precipitation within MCNP is between 2,000 and 10,000 millimeters, while the temperatures range between 4°C at the top of the mountain and 32°C at the base of the mountain (Charlotte 2014, 17-18). Volcanic soils influence rich biodiversity with some estimated 2,435 plant species recorded in the area. There are 41 villages adjacent to the park with an estimated population of 100,000 inhabitants. The Bakweri, who have long been in the area, constitute less than 30% of this population estimate. Activities of managing MCNP influence the livelihoods of local people, both directly and indirectly. The socio-economic activities of villagers range from hunting, timber exploitation, subsistence farming and keeping cattle, to small trading in agricultural products, herbal plants, and wild fruits.
In terms of the governance structure, the Bakweri inhabited the lower slopes of Mount Cameroon many years before the advent of colonialism in the 19 th century. Around the 1850s, Bakweri patrilineal relatives and families lived in settlements led by lineage elders where they grazed livestock, cultivated crops, and hunted games (Ardener 1970, 140-144). However, the formal context for governing the land came after a series of events, including German colonization in 1884, the loss of Bakweri land to German colonialists following a war in 1891and subsequent land reforms such as the 1974 land tenure law that created state lands while eradicating former claims for Bakweri land ownership (Assembe-Mvondo et al. 2014).
Following Cameroon's signatory to the 1992 UN Convention on Biodiversity, the state adopted a 1994 forestry and wildlife law for the sustainable management of natural resources through the development of protected areas such as national parks. Section 72 of this law prohibits unauthorized entry, hunting of certain animals and the harvesting of various plants perceived as endangered in protected areas. Article 8 of this law, however, grants some usage rights to the locals stating: 'Customary rights are those recognized by local populations to exploit all forest, wildlife and fish products, except for protected species, for their personal use. They may be temporarily or permanently suspended when the need arises for reasons of public interest' (Republic of Cameroon 1994).
The 1994 law (Part 1, section 7) further regulates rights to land ownership stating: 'The state, local Councils, village communities and private individuals may exercise on their forest and aquacultural establishments all the rights that result from ownership subject to restrictions laid down in the regulations governing land tenure and state lands and by this law' (Republic of Cameroon 1994).
In 2009, the state, through a Prime Ministerial Decree (No. 2009/2272) created MCNP. This establishment came with the demarcation of 58,178 hectares of land into four cluster conservation zones including Buea in the southeastern flank of Mount Cameroon, Muyuka in the northeast, Bomboko in the northwest, and West Coast to the west of the mountain (Figure 1). For the state to administer MCNP and its adjacent villages, a Ministerial decision of the Ministry of Forestry and Wildlife (a state institution that oversees the management of protected areas in Cameroon) in August 2014 introduced a plan for natural resource management (Charlotte 2014). In effect, the decision endorsed the task for a unit (MCNP Service) to ensure the preservation of biodiversity on Mount Cameroon through the involvement of local communities.
MCNP Service officials conduct resource management duties along with the participation of stakeholders from Divisional, Sub-divisional, Regional and National levels of the society. For instance, through agreements known as 'memorandums of understanding', the MCNP Service collaborates with local and international institutions (non-governmental and governmental) to meet its management objectives. For instance, the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ), a German institution, has been at the core of implementing the MCNP Service's goals for the sustainable development of natural resources over the last decade. MCNP Service also collaborates with five Sub-divisional Councils (Buea, Limbe II, Mbonge, Muyuka, Idenau) of the Fako Division in the southwest region of Cameroon.

Data collection
We used Focus Group Discussions (FGD) to collect data between August and December 2017. FGD is a form of qualitative research that involves gathering a diverse group of people from a community to discuss their opinions, beliefs, and perceptions about a topic of interest. In recent years, the use of focus groups has been a cornerstone of studies associated with protected areas (Mutanga, Never, and Edson 2017). Understanding that collective agency includes passive behavior among locals that might not entirely be of active community engagement to change something, FGD was important for us to uncover data about attitudes, desires, and reactions not conveyed by other techniques in research. Determining the focus groups entailed a process of obtaining authorization from Cameroon's Ministry of Forestry and Wildlife, during the year 2017. Followed by a review of documents archived at the MCNP Service office in Buea. These documents comprised of Agro Socio-economic Assessment (ASEA) reports for 41 villages around Mount Cameroon.
Our intention for review was to understand how the livelihoods of villages around MCNP might differ, and for us to develop a purposive sample from them. By definition, purposive sampling is a random selection of sampling units from a population with the most relevant information on characteristics of interest to the researcher (Guarte and Barrios 2006). In doing so, we focused on selecting focus groups from villages whose livelihoods connect to MCNP, i.e. they share common practices of timber exploitation, hunting, and farming around the park. These are activities equally regulated by the management regime.
The chosen villages are host to the Bakweri ethnicity with a communal lifestyle, where village groups/  committees and membership, led by Chiefs (village heads), form the traditional basis for enhancing the everyday activities in communities. This communal nature explains our choice for targeting focus groups ( Figure 2). In consultation with Chiefs and park officials to gather participants for FGD, the Chiefs assigned 16 individuals to represent eight groups for every study village. In these villages, residents traditionally honor their Chiefs for leadership skills, knowledge, and their ability to appoint persons they see as competent to share ideas with visitors in support of their village. The participants belong to various community groups that included: distributing seeds, equipment, farm products and sharing funds from the sale of farm produce.
Before meetings, we obtained informed consent directly from participants using letters signed by Chiefs on behalf of the locals. We then briefed participants on the subject matter of natural resource management. The FGD was guided by the following questions: • What opinions do locals have regarding the benefits and worries of the state system of natural resource management? • How do locals move on with their everyday life of attaining cultural needs amid difficulties brought by this system? • What circumstances enable local people to act and decide on ways to satisfy their needs regardless of unfavorable policy arrangements?
The study groups promote a sense of communal support, make decisions for community benefit, and participate in various activities (Table 1). The gender variation attributes to the fact that local women in the villages have a vital role in their homes, where they oversee family issues, while the men perform influential roles within administrative settings away from home and in support of their communities. The lead author moderated the focus group meetings and a research assistant took charge of taking notes during the FGD. Four cluster facilitators (park officials tasked to intercede in support of village inhabitants concerning issues related to MCNP) assisted us with translations during meetings. The languages of discussion were Mokpwe and Pidgin English.

Data analysis
To summarise the bulk of information, we employed the views of Costa et al. (2016) in the thematic analysis of focus group data ( Table 2). The analysis was mainly inductive by questioning multiple opinions from field data, looking at various assumptions, patterns, explanations, and integrating them into the results consistent with the research questions. This approach entailed reading transcripts, field notes, listening to voice recordings from focus group meetings, exploring different responses of participants, and, identifying and defining themes in the data collected. Data responses were the first subject to open coding adapted from Costa et al. (2016, 38) to produce themes and raise questions about various extracts from FGD which we then integrated into the results below.

Results
For us to examine the collective agency of residents in the study area, we need to understand the difficulties posed by the state's resource management procedure.
In this section of the paper, we use two main headings. First, to show the complex nature of the management system, and second, to demonstrate how locals express their agency under this system.

Complexities of the management system
Arrangements for participation Although dissimilar views exist among locals regarding the challenges of participating in state plans for resource management, some observations can be made about the management system. For instance, with the creation of MCNP in 2009, park officials endorsed the creation of Village Forest Management Committees (VFMC)discussion forums through which village inhabitants discuss ideas about the management of natural resources. Through management activities, locals receive various state-led initiatives, including employment in management schemes for harvesting Prunus Africana (a wild medicinal plant on Mount Cameroon that contains anti-cancer properties, exported to other countries) and wild honey, partaking in sensitisation programmes for agricultural development, participating in the organisation of cluster platform meetings, park boundary clearing and monitoring together with park officials.
These arrangements, however, come with difficulties. For instance, despite efforts by park officials to integrate locals into their management agenda, the regime strictly prohibits some of the traditional/subsistence activities such as animal trapping, hunting, and fishing in protected areas of Mount Cameroon, activities which the locals were culturally entitled to for many years prior to the advent of MCNP.
Further, to determine local/indigenous rights to use MCNP, the current management plan uses Conservation Development Agreements (CDA) and stipulations of the 1994 state forestry and wildlife law (Charlotte 2014). Both frameworks do not provide a clear and solicited basis for exerting the customary rights of local/indigenous people to use protected areas. Unfortunately, the terms for locals to exercise their customary rights have not been appropriately recognized by decree. The lack of any procedures to institute customary law in practice leads to disputes among forest stakeholders.
Even so, the CDA warrants locals to obtain state assistance based mostly on how the locals comply with state laws. In other words, the management regime promotes local interests primarily in those activities that serve to achieve visions of the state, and this is why some of the locals perceive the management system as conflicting to their cultural needs. This process of governing the rights of people indicates the complex nature of the resource management system.

Negative impacts on the community
We see the consequence of this complexity in the complaints local people have against interferences on their farmlands. For instance, during group discussions, respondents spoke about clashes between their knowledge of the land and the views of park officials. In this situation, the locals attempted to protect their use of the land against the desires of the management regime. We observed local responses to these negative impacts in three instances: (a) where private farms, which locals own, share boundary with state-protected areas; (b) where animals from the park intrude on farmlands to feed on crops, and (c) where locals express their grievances against unfulfilled promises by the regime.
In the first instance, the locals cultivate crops on farmlands inherited from their ancestors that are in close contact with the MCNP boundary. At times, during periodic alterations of the park boundary, new boundary lines trespass into these farmlands without the knowledge of local people. As a result, the respondents furiously blamed park officials for not consulting with them during changes in park boundaries. We see such a state of frustration in the following words of a farmer in Farmers Group: 'I have always owned and used my farm in the past years. But now, the park people come, they beat us in the farm and say we are farming on park boundary land. During the time we have been collaborating with the park authorities, they have not been working to the advantage of the villagers but rather to their advantage . . . ' (FGD 2017).
On the second point, reports about animals intruding on private farms in Bibunde, Njonje, and Batoke raised the fears of local people. Our focus group meetings showed that crop damage on farmlands close to park boundaries led to physical confrontations between the locals and animals from the park. As a member of the Traditional Council puts it: 'At night, the elephants come down from the mountain and eat our crops and we are left with the very little harvest at the end. I have seen monkeys eating my crops, but I am not able to stop them because the park authorities do not let us hunt them. If we do, they punish us. They should find a solution!' (FGD 2017).
According to this example, the locals whose crops were heavily damaged felt that the regime does very little to address the problem of animal encroachment on their farms but instead imposes heavy sanctions for persons who injure state-protected animals. To quote the view of another member of the Vigilante Group: 'The punishment from park staff for persons caught in attacking or injuring animals which are protected in the park is . . . heavy and unaffordable fines' (FGD 2017).
During fieldwork, we met a hunter at Kuke Kumbo, who spoke of having stumbled upon monkeys destroying crops on his farmland. He explained that it was frustrating to have no reliable means of intervening, due to fear of legal charges by the regime and his inability to pay unaffordable fines for hurting animals.
The third way of responding to negative impacts was the expression of grievances relating to unfulfilled state promises. In Munyenge, a respondent of the Village Development Group who was aware of promised development by the regime claimed that park authorities had procured more from village residents and invested less in the community. Also, in the Bomboko area, following claims about land ownership by the regime, some respondents were unpleased by park restrictions against harvesting forest resources, which they perceived as an important source of livelihood to the local community. Unsatisfied by the activities of the regime, a member of the Farmers Group argued as follows: 'My farm, which was close to the park, has now been claimed as government land and so I have to depend on what the government can give us to survive on' (FGD 2017).

Another local in the Hunters Group added:
'I do not harvest much from my farm anymore because the park people are there. They say my farm is within park boundary land. The park people must consult with our traditional authorities so we can bring this problem to an end' (FGD 2017).
Consequently, the locals address the above grievances by engaging in different ways to safeguard their knowledge of the land. They do so by 'negotiating for compensation' of the land, requesting for development support in exchange for their will to comply with state forestry laws. For example, in Big Koto, Efolofo, and Bonakanda, some respondents demanded park authorities to provide drinkable water to their communities. A local at Bomboko emphasized the fact that the regime ignores the interests of locals during the execution of projects. We observed another example from a farmer at Big Koto, who complained of the low wages paid to him under the Prunus Africana management scheme. These examples reflect the complex nature of the resource management system and how the locals respond.

Expressing collective agency
In this section, the evidence presented about expressing agency reflects in practices of alternative behavior among locals, to hold onto certain livelihoods and cultural traditions, rather than renegotiating the subordinate role they occupy in the resource management regime.

Using economic incentives for livelihood progress
For instance, although some village groups were unfavorably impacted by the system of resource management, the results revealed that some focus groups in other localities had found ways of getting around the system by using economic incentives of the regime to make their livelihoods work better in spite of antagonistic policies on the community. When we questioned focus groups at Muyuka and Buea concerning gaining from their interaction with park officials, they spoke about using the economic incentives of the management system to satisfy local needs as noted in the views of a member in the Village Forest Management Committee: 'When we co-operate with the park authorities, we get to learn other ways of harvesting honey, planting cassava and bush mango which we sell by ourselves to the community and generate income through membership in co-operatives. They also train us on conservation procedures by teaching us how to plant useful trees and create propagators for nursing plants' (FGD 2017).
Doing so, however, requires a degree of compliance with rules of the management regime. For instance, at Bonakanda, Bomana, and Lykoko, respondents indicated that some of them had gained employment by co-operating with the regime in ecotourism activities. In the same way, respondents at Bonakanda were optimistic, insinuating that such collaboration would help them preserve forest resources for future generations.
Based on discussions at Batoke, Bakingili, Lower Boando, Njonje, Sanje, Bomana, Kuke Kumbo, Munyange, and Lykoko, the groups welcomed the supply of social amenities by the state for the wellbeing of village inhabitants. These villages receive machinery from park officials to process farm crops such as cassava and plantain. With the help of conservation bonuses, which are funds from the state, these groups were able to secure the supply of drinkable water, community halls, and to purchase chairs and canopies they use for village meetings. This was the opinion of a local in the Health Committee: 'We now benefit from our village market, which the park authorities built and now it is giving us some hope. We have been able to have water supply in the village, community halls, and canopies that we use for public meetings' (FGD 2017).
In Muyuka and Bomboko, respondents used development incentives as sources for generating income. In these cluster zones of MCNP, the locals convert fragments of the land for cocoa cultivation. Cocoa accounts for about 50% of agricultural income to farmers living in these zones. Further, FGD at Muyuka indicated that people profit more from agriculture in comparison to hunting due to state prohibitions in the park. Similarly, in Bibunde and Batoke, the villagers engage in fishing as an alternative source of food and income. Many farmers in these villages use farm equipment and gain training skills from park officials to produce food crops sold in nearby markets. Accordingly, we see how even where regimes for natural resource management do not entirely meet the expectations of communities, the locals can profit from other sources of income in the system. We see agency in semi-formalized arrangements that help locals decide on attaining other needs using the income they earn from gathering non-timber forest products through a state-backed organization, Mount Cameroon Prunus Management Company. Within this organization, the locals work as harvesters in a state-induced management scheme for Prunus Africana. A member of the Health Committee in Muyuka made the following observation: 'Unions help the villagers to work together and jointly plan finances earned from harvesting Prunus for purchasing healthcare services and to pay for education for their children.' (FGD 2017) Although some participants in Buea and Muyuka were unhappy about this level of collaboration for reasons of low income paid to harvesters, the locals indicated that the scheme enabled them to create Prunus Africana harvesters' unions, i.e. joint economic initiatives that locals use as a means to support livelihoods other than those seen as visions of the state. The fact that a group of people chose not to fight the park system but rather used alternative strategies to circumvent it, is in our opinion also an expression of collective agency.

Maintaining cultural practices through religious expression
The use of state forestry laws to limit locals from entering MCNP changed the forest-based subsistent activities of local people. There were, however, some exceptions to this view. Among the Bakweri, sacred societies and Traditional Councils of elders are crucial institutions for decision-making on issues of preserving cultural heritage, although such power is gradually declining with the introduction of committees. Group discussions showed that dominant sacred societies such as the Maale and the Liengu societies played an active role in villages to meet cultural needs aside from the will of the management regime.
On the example of Maale, this sacred society represents a medium of religious activity for justifying the need to preserve natural resourcesupon worshiping a spiritual being, Efassa moto, whom the Bakweri see as the protector of their land. The locals equally believe this spiritual being has the mystical power to cause natural disasters when park officials fail to support ritual practices in the park. Thus, the loss of biodiversity on MCNP will imply a manifestation of the wrath of Efassa moto for failure by the community to perform certain ritual rites (Monono et al. 2016). Maale has equally been not only a sacred society but also, a symbolic religious tradition of the Bakweri (Ofege 2007) which, nowadays, appears to be a metaphysical space for locals to exert their agency.
In this religious situation, ritual practices take place on the land. Often, knowledge about the sites used for rituals are secretly kept away from the general public except for native people initiated in sacred societies as stated in the words of a local in Quarter Head: 'You must consult with the Chief to know clear spots of cultural importance. There are places only known by native people. Sometimes, to visit these spots, you must also consult the gods of the land. Failure for any local to do so means disrespect to village authority and culture' (FGD 2017).
In this example, the agency develops from people's knowledge of preserving sacred sites. The respondent also reiterates the importance of acknowledging village authority by consulting with Chiefs before visiting sacred sites. In all the study groups, participants specified that they performed rituals through traditional festivals and sacred ceremonies. Again, the Maale ritual is an example performed annually in an appeal for Efassa moto to bring peace and prosperity to the community. A member of the Traditional Council explained: 'During this period, we go into the park and perform our Njoku tradition. You dance spiritually, while in the village, you do it physically, i.e. in public' (FGD 2017).
Information obtained during focus group meetings indicated that rituals help to prevent naturally induced damages on the land. For instance, whenever a volcanic eruption occurs on the mountain, locals perform rituals requesting Efassa moto to safeguard vegetation and farmlands from destruction.
Consider the words of a local in Hunters Group: 'The elders go there to consult the gods when we have problems in the community like the widespread death of people in the villages and the quick dryness of streams and rivers.' (FGD 2017) In the narratives of the Njoku tradition, respondents mentioned totemic practicesthe act of perceiving humans as having mystical relations with a spiritual being through transforming into a plant, an object, or an animal. Here, locals who practice the Njoku tradition go into the forest to ask their ancestors to intervene in times of need. As a member of the Village Development Committee added: 'Our local culture and beliefs move along with a reliance on the natural environment. There are localized traditional manifestations that warrant masquerades to move into the forest. When they do so, they collect barks of trees and herbs that are necessary for performing the traditional rights of appeasing the gods to intervene in times of trouble and need' (FGD 2017).
Along the line of Njoku tradition, a local in the Hunters Group noted: 'We have been told of the stories that people who are active in the sacred societies have certain powers obtained from our ancestors to spiritually transcend at night into the bodies of elephants in the park and influence life events. By doing so, the people can react against intruders on their farmland by using the elephants to chase persons away' (FGD 2017).
In other words, even where resource management by the regime does not entirely address the need for spiritual intervention in complex situations, the locals can flout certain visions of the regime to preserve their heritage. They specifically do so to keep intruders away from their land as stated in the above quotation from a local. Another example of using the religious expression to maintain cultural practices connects to caves and water. In the Bomboko area, there is a cave known as Isuma, located in the peripheral zones of MCNP. According to a hunter in this area, only native people who visit the cave to contact ancestral spirits know the location of Isuma. Similarly, on the West Coast, there is a spot known as the red hill. A member of the Village Development Group explained: 'During the Maale ritual, community members use the caves to carry out animal sacrifices and libations. Only a few individuals know the location of the sacred sites in use' (FGD 2017).
Moreover, in Bakingili and Lower Boando, knowledge of the mamid tradition exists under a Liengu sacred society. According to this example, locals are 'responsible caretakers of the sea'. The respondents talked about worshiping a goddess of the sea, locally referred to as Liengu-la-mwanja. In this instance, the locals maintain ritual practices despite the inadequacy of the resource management system to recognize the customary rights of using the land. The locals see this collective act as crucial for retaining harmony and solidity in their community, such as in situations, following unexpected natural disasters and the widespread death of people.
On the example of Womba (a huge tree that has grown for hundreds of years close to park boundaries), participants stated that because the Womba tree hosts spirits of ancestors, they need to preserve it. Every night before traditional festivals, ritualists (also known in mokpwe as Sango y a bando) sacrifice goats and pour libation beneath the Womba tree. They stated that this was an activity used to plead for the intervention of ancestral spirits to protect people during cultural events. Further, respondents were mindful of sustaining such relationships with spiritual beings at the family level, which they believe influenced the nature of everyday life. These were the words of a member of the Farmers Group concerning family worship: 'In some villages, family members worship a spiritual figure of the family camp called Ikomu Njewoka. This figure represents an object of spiritual value, dug from the ground by the camp leader, who sees and/or comprehends the object as one that protects his family and children. Some families do have another tradition called Lisomelele. This means that every year, the family head pours libation on doorposts, from the right to the left side of the door using his right hand . . . ' (FGD 2017).
Thus, one could say the locals possess their agency through religious engagement, often, in secret, to preserve relevant traditions that help maintain life stability in their communities.

Discussion
The above results identified difficulties of a state-led natural resource management system and examined instances of locals conveying their agency. This is what we describe as expressing a collective agency in antagonistic policy environments. From this study, we can draw three interpretations of the literature on resource management regimes and notions of agency: First, on the complexity of resource management, previous scholars have explored the setbacks of topdown management systems (Borrini-Feyerabend et al. 2004;Nadasdy 2003Nadasdy , 2005Spaeder and Harvey 2005). They note that in such systems, political interests and eagerness for local compliance with visions of governments tend to overshadow the local values and needs of communities. A few cases such as in Laos (Ramcilovic-Suominen 2019) and Papua New Guinea (Sloan et al. 2019) are partly a reflection of the above critiqueswhere both management regimes, advocate for development plans of economic integration and industrial investment at the detriment of local/ indigenous rights in land use.
Consistent with the views outlined above, the example of MCNP and adjacent communities suggest another complex form of this centralized structure. Here, though, the state shares its power to local units towards the management of natural resources. It adopts certain arrangements for local participation issued through initiatives that do not entirely solicit local consent for customary rights. This lack of reasonable consultation with locals partly reflects on occasions of land encroachment and grievances against the regime.
Second, on examining expressions of agency, previous literature explains agency based on how locals determine their engagement with systems, and how such engagement either allows or restricts their capacity to act (Chirozva 2015; Cole et al. 2019). Scholars also relate agency to how locals develop awareness and motivation to participate in conservation programs (Honig et al. 2015). However, our analysis revealed that agency also pertains to processes by which locals attempt to thrive where the tasks of natural resource managers do not seem to work in their favor. In such processes, the agency is not entirely of active community engagement to change something, but also includes passive behaviors where community members reshape economic incentives of the regime towards livelihood improvement, as well as, using religious tradition to justify their cultural needs. This shows the potential for locals to cultivate flexible ways of living on the land amid undesirable situations of resource management.
Third, regarding the enabling mechanisms under which locals carve space for thriving in unfavorable management systems, we found that certain factors/ conditions were essential. Previous studies have identified determinant factors of agency such as: where environmental changes trigger the habits of people (Ransan-Cooper 2016); where people assign responsibilities to act (Dowding 2008); and in practices that stimulate resistance or dialogue among people (Stammler and Wilson (2006), to name a few. While we agree with this scholarship, our study suggests other mechanisms for agency in antagonistic policy environments. For instance, where traditional institutions, i.e. sacred societies and Traditional Councils have an active role in communities. These institutions are often a source of 'local capital' for the agency for maintaining culturally relevant needs, even when regimes do not legalize such institutions in their management agenda.
Another enabling factor for agency is the differences in state monitoring activities among villages across cluster conservation zones of MCNP. In localities within the West Coast and Buea that are closer to urban districts with greater road accessibility, group discussions showed that park officials regularly visit these localities to monitor the extent to which communities comply with state forestry laws, giving little space for locals who intend to maintain their subsistence use of forest land. On the contrary, villages in Bomboko and Muyuka, which were less accessible to park officials due to poor road infrastructure and remoteness, had greater space for locals to maintain certain religious and subsistence practices.
Related to the above factor, scholars have observed 'power' as a mechanism for agency (Drydyk 2013;Hanmer and Klugman 2016). In the resource management literature, power often reflects the ability to exert control or influence management decisions (Borrini-Feyerabend et al. 2004). Our study showed that situations of less state control, i.e. where state forestry laws are less active in a community, locals were able to cope with difficulties by using contradictory management processes to initiate constructive needs. In Bomboko, for instance, a farmer began negotiating compensation for the land he lost to the regime. Another local in the same area spoke about the need to sustain the Njoku tradition for the harmony and solidity of his community. In such instances, the collective agency occurs in people's capability to utilize unfavorable situations of management in advancing issues that are of relevant need to them. Honig et al. (2015) concluded that the agency of locals to participate in conservation programs are defined by either motivation or awareness about the provisions of such programs. By comparison, the MCNP case indicated that awareness enabled respondents to express various complaints against the regime. In this example, locals knew shortcomings in the management system, which incited them to voice their dissatisfaction. For example, in Munyenge, the locals whose farmlands were encroached upon through unannounced changes in park boundaries insisted that park officials consult with them on such activities in the future. In Bomboko, the locals bargained for support of state forestry laws on a condition that the regime enhances the well-being of their community and recognizes locals' rights to use their land.

Conclusion
We set out to gather partial insights on processes of collective agency to address questions that remain about understanding how communities attempt to thrive under unfavorable policy environments of resource management. We revisited previous literature on challenges encountered in processes of top-down resource management (Borrini-Feyerabend et al. 2004;Hirsenberger et al. 2019;Markkula, Turunen, and Kantola 2019;Nadasdy 2003Nadasdy , 2005Owuor, Icely, and Newton 2019;Ramcilovic-Suominen 2019;Sloan et al. 2019;Spaeder and Harvey 2005). Our premise was that understanding that previous studies present examples that are important for revealing the adversarial effects of state-led natural resource management procedures, there is a need to understand how people find other ways to meet their needs even in these undesirable settings of management.
To bridge the above gap, we examined how a natural resource management system for MCNP can be antagonistic for communities adjacent to the park. For instance, one of the difficulties came from the participatory approach of the regime assisting locals mostly when their needs closely match state aspirations of management. In effect, we saw that the locals responded in passive ways such as expressions of frustration and grievances voicing their dissatisfaction against the regime. However, we observed that even where people might be victims of the management system, they carve space for exerting their agency where they devise alternative strategies to attain various needs. Rather than renegotiating their subordinate role in the management regime, some locals get around the system by using economic incentives to improve their livelihoods. Further, they hold onto their traditions through religious beliefs. These, based on the analysis, are the expressions of collective agency in antagonistic policy environments.
These responses, based on the analysis, are expressions of agency in antagonistic policy environments. We reiterate that local people, although subjected to challenges in the institutional management of natural resources, do have a capacity for collective engagement in attaining their needs. This capacity, indeed, lies in locally embedded mechanisms for agency, such as the activeness of traditional institutions in communities, the ability for locals to act more wisely and initiate constructive needs, where management regimes have less monitoring and control on the land, and, where people are much aware of the difficulties posed by resource management systems.