1. Introduction

Private ecotourism enterprises, partnered with communal conservancies, are purported to be central to the success of Namibia's conservancies in achieving biodiversity conservation and poverty alleviation (World Resources Institute et al. 2005; NACSO 2006). Ecotourism, premised on stunning wildlife and scenery, has attracted an international, ‘up market’ clientele to Namibia and generated revenues, employment and additional benefits for participating conservancy communities.

My purpose in this paper is to show the place of ecotourism enterprise partnerships with communal conservancies in community-based conservation, revealing features and implications of power in such partnerships. I argue that while ecotourism enterprise partnerships have generated community benefits from wildlife-based ecotourism, such partnerships have also entrenched power asymmetries between and among partners.

I use the term power to mean the application of action, knowledge and resources to effect decisions, resolve problems and further one's interests (Adger et al. 2006). I examine the Torra Conservancy and the Ehi-rovipuka Conservancy cases in northern Namibia (Figure 1) Torra is lauded as the ‘flagship’ of the conservancy system (UNDP and Equator Initiative 2004; World Resources Institute et al. 2005) premised mainly on a partnership with an international ecotourism enterprise. The Ehi-rovipuka Conservancy, established three years after Torra, has attempted to attract tourism investment partners and replicate the ecotourism partnership model, but has yet to achieve this for reasons that will be discussed.

My work is informed by a growing body of commons scholarship recognizing that community-based conservation alone is not a panacea for community development and biodiversity conservation. Rather, commons institutions for conservation operate in a multilevel world and necessarily link with different scales and levels of organization (Adger et al. 2006; Cash et al. 2006; Berkes 2008). While multiple partnerships with the community level have been characteristic of ‘successful’ community-based conservation cases, including business enterprise partnerships (Berkes and Adhikari 2006; Berkes 2007), I suggest that there can also be more sinister consequences for communities in these arrangements.

Community-based conservation is based on the idea that if conservation and development can be simultaneously achieved, the interests of both are served (Berkes 2004). Community conservation stresses the role of local residents in decision-making for natural resources (Adams and Hulme 2001). Community-based conservation has been practiced in many forms, but in the broadest sense includes conservation by, for, and with the local community. The co-existence of people and nature, as distinct from protectionism and the segregation of people and nature, is its central characteristic (Western and Wright 1994). Recent commons scholarship stresses that local levels of collective action are necessarily linked to higher levels of social and political organization in a globalized world (Cash et al. 2006; Berkes 2007). Indeed, Namibia's conservancies have many and increasing cross-scale and cross-level linkages (Young 2002; Cash et al. 2006), including important linkages with international tourism enterprises.

Centrally and internationally conceived approaches in community-based conservation emerged in the 1980s in Southern Africa to buttress national parks as wildlife reserves, and better conserve wildlife as an economic development alternative to agriculture in semi-arid regions (Adams and Hulme 2001). These have been termed community-based natural resource management or CBNRM (Fabricius et al. 2004). CBNRM has featured the devolution of certain bundles of wildlife use rights to local communities, premised on making wildlife pay, with benefits exceeding the costs of living with wildlife such as crop and property damage. The central theory is that benefits from wildlife for local and indigenous peoples will promote conservation.

The revenue and resource sharing devolution to communities under CBNRM was led by Zimbabwe and Namibia in Southern Africa and was a direct outgrowth of wildlife management on the private land estates in both countries preceding independence (Jones and Murphree 2001). In the 1970s, Zimbabwean legislation was passed that conferred strong proprietor rights over wildlife to private, white landowners. This same type of legislation was passed in Namibia in the 1970s under the South African administration, prior to independence. There was strong political demand after independence in both Zimbabwe and Namibia to transfer the economic success of wildlife management on private lands to communal lands and relieve a racist divide in rights to use wildlife, where indigenous use of wildlife had been criminalized under colonial rule (Gibson and Marks 1995; Gibson 1999).

Design principles for long-enduring common property institutions at local levels have been recognized at varying levels of detail (Ostrom 1990; Agrawal 2002). These are norms and rules determining who is excluded from a particular resource use or area, and how participants deal with subtractability in ways that sustain collective agreement and mutually shared benefits. Namibia's wildlife is a common property resource for which excludability, or the control of access is difficult, given the mobile nature of wildlife, and where collective use involves subtractability, where each user is capable of subtracting from the welfare of others (Feeny et al. 1990; Ostrom 1990).

Under Namibia's CBNRM program, communal area residents form a common property institution called a conservancy and enjoy rights in wildlife and related tourism development devolved under national legislation. Conservancies are approved by and registered with the Ministry of Environment and Tourism. Registration requires a defined conservancy boundary, voluntary registered membership, a representative conservancy management committee, a constitution and a commitment to producing a benefits distribution plan (Long 2004; Jones 2006 interview).

Namibia's CBNRM program drew lessons from regional experience (Jones 2006 interview; Owen-Smith 2006 interview) especially Zimbabwe's Communal Areas Program for Indigenous Resources, CAMPFIRE (Sangarwe 1998; Bond 2001; Murombedzi 2001) and Zambia's Administrative Management Design for Game Management Areas program, ADMADE (Gibson 1999; Marks 1999). There was a deliberate effort to avoid pre-determined administrative boundaries such as CAMPFIRE's use of rural district ward boundaries and ADMADE's use of nationally defined Game Management Area boundaries. Rather, Namibia's conservancies self-organized and negotiated their own boundaries, to help reinforce devolution of wildlife use rights and benefits to community level. Formal registration of conservancy members, and legal gazetting reinforced external recognition (Ostrom 1990), again a significant departure from both CAMPFIRE and ADMADE. As well, the wildlife revenues and other benefits developed under conservancies were intended to accrue solely to the conservancies and were not to be shared with central or regional government, as they were under CAMPFIRE and ADMADE (Marks 1999; Bond 2001; Murombedzi 2001).

2. Methods

I followed a Participatory Rural Appraisal (PRA) approach in this research, including case study and a variety of qualitative field tools (Chambers 1997; Berg 2004). Semi-structured interviews were conducted in Namibia in 2006 and 2007 with key informants involved in CBNRM from NGOs, government ministries, conservancy communities and the national research community (Appendix 1). Interview questions were general, open-ended, and explored the development of CBNRM, institutional linkages and roles. As well, site visits, participant observation, and structured villager interviews were carried out in 2007. A total of 40 structured villager interviews were conducted in Ehi-rovipuka Conservancy. I posed a standard set of questions to each respondent (Appendix 2). I employed a community interpreter and all interviews were in Otjiherero, the language of the resident Herero people. I compared and triangulated the field data acquired with other regional sources. This research was part of a broader doctoral study investigating social-ecological systems linkages between community-based conservation and protected areas in Namibia (Berkes and Folke 1998; Hoole 2008).

3. Results

I first describe the evolution of Namibia's CBNRM program and formation of communal conservancies, situating ecotourism enterprise partnerships in this evolution. The term ‘partnership’ as it applies here means any cooperative relationship between two or more parties involving shared responsibility, authority, and resources: human, financial or material. I focus on the place of partnerships between local conservancies and international tourism enterprises and discuss examples from the Torra Conservancy and Ehi-rovipuka Conservancy. Power relationships between and among conservancies, private enterprise and the state are elucidated and certain conclusions are drawn.

3.1. Institutional linkages and partnerships – an evolution

The roots of Namibia's CBNRM program date back to 1982. The Namibian Wildlife Trust was concerned about severely depleted wildlife in northern Namibia. This conservation NGO deployed a conservationist to collaborate with four local headmen, who shared concern about the dramatic loss of wildlife in the region (Jones 2001; Owen-Smith 2006 interview). The headmen appointed community game guards who were knowledgeable hunters and trackers from local communities. The aim was to stop poaching (Jacobsohn 2006 interview). The game guards monitored wildlife, reporting wildlife incidents and suspicious activities to the headmen, who in turn informed the government wildlife enforcement agency. The community game guards, in contrast to other CBNRM approaches emerging in Southern Africa (Gibson 1999; Marks 2005) were not enforcement personnel employed in salaried positions by the state.

By the late 1980s, wildlife populations had noticeably recovered and the community game guard program was considered a major contributing factor. This coalition of local conservation effort, featuring leadership and collaboration in a basic NGO and community partnership, spurred wider application. Increased demand for the program led to formation of a new Namibian NGO, Integrated Rural Development and Nature Conservation (IRDNC), to facilitate and support the development of CBNRM in the wider Kunene and Caprivi regions. IRDNC remains the leading NGO dedicated to CBNRM in Namibia. This early period of simple, bilateral linkages in CBNRM is depicted in Figure 1 for the 1980s.

The next stage in the evolution of Namibian CBNRM followed independence in 1990. The new black majority government extended rights in wildlife to communal area residents that previously had been granted only to white farmers on private lands by the colonial South African administration. IRDNC's leadership, based upon their knowledge and experience gained in the community game guard program, were engaged by the government to help design and conduct community surveys. This led to policy and legislation for a national CBNRM program under the Nature Conservation Amendment Act, 1996 (Jones 2006 interview; Owen-Smith 2006 interview).

USAID provided donor assistance through the World Wildlife Fund (WWF US). USAID and WWF (US) have remained the main donors to CBNRM, although other international donors have come in. During this same period the Namibian Association of CBNRM Support Organizations (NACSO) was formed as a national umbrella organization for other NGOs supporting institutional development, natural resources management, business enterprises and livelihoods at the local conservancy level (Louis 2006 interview). Thus, as shown in Figure 2 for the 1990s, we see a further evolution of cross-scale linkages and partnership formation that started locally in a remote part of northern Namibia, moved to national level, and in turn attracted and mediated international donor support.

The national conservation NGOs, led by the IRDNC, filled various CBNRM facilitation and support roles, including technical support for annual game counts, wildlife monitoring and reporting, capacity-building in project planning and budgeting, and funding for salaries, vehicles, field offices and equipment at the local conservancy level. In a number of instances, NGOs have served as boundary or bridging organizations (Cash and Moser 2000; Cash et al. 2006; Berkes in press) linking international donor support to the local conservancy level.

The support and facilitation of local conservancies by national NGOs has centred on locally elected conservancy management committees and the community game guards, who remain the central feature of conservancy conservation efforts. Community game guards are now full-time staff monitoring wildlife on behalf of the conservancy management committees, with technical support from NGOs (Stuart-Hill et al. 2005). The accumulated evidence indicates that wildlife conservation has been achieved under CBNRM. Subject to ranges of natural variability, especially due to drought, overall wildlife numbers have recovered and increased since the 1980s (Gibson 2001; Nott et al. 2004; CONINFO Information System 2006).

The national NGOs are headquartered in Windhoek and some have regional field stations or a regional mandate. NGO operations feature professional and technical staff cadres such as biologists, GIS technicians, social scientists and project managers, equipped with all-wheel drive vehicles, modern offices and sophisticated tools, including the latest in computing and remote sensing. Funding for this CBNRM network and establishment comes from the international donor community through multi-lateral and/or bilateral programmes.

Communal conservancy formation has increased exponentially over a relatively short period. There were only four conservancies in the late 1990s and over 50 by 2007 (Weaver 2007 interview). This is depicted in Figure 2 for the period 2000–2010. Figure 2 also illustrates a stronger clustering of linkages within the local level of organization. The first conservancies now regularly participate in regional forums to share experiences and develop planning and management skills with the support of NGOs. Thus, we see an emergence of both multiple cross-scale and cross-level interactions (Cash et al. 2006). Social learning, knowledge sharing and trust have been emergent properties between certain conservancy and NGO partners, reflecting the greater duration of their interactions and experiences working together (Armitage et al. 2009; Berkes in press).

It is noteworthy that the evolution of CBNRM in Namibia has taken almost three decades. A growing network of national and regionally-based NGOs facilitating and supporting CBNRM has helped mediate and channel international donor support and purportedly offset competition among NGOs for donor funds (Louis 2006 interview). This network has also appropriated significant CBNRM funding from international donors to sustain itself. Partnerships between conservancies and ecotourism enterprises have played an increasing role in the last decade and I now focus attention on this particular aspect

3.2. Torra Conservancy and ecotourism enterprise partnership

The Torra Conservancy has received international recognition as a successful case of community-based conservation (UNDP and Equator Initiative 2004; World Resources Institute et al. 2005). Torra is premised on conserving and commercially exploiting the wildlife endemic to the spectacular and remote arid wild lands of the Kunene region. Wildlife move seasonally through the wider region that Torra shares with other conservancies, Skelton Coast Park and Etosha National Park (Figure 1).

The Torra Conservancy was one of the first communal conservancies gazetted in June 1998, with 450 registered members drawn from Damara and Riemvasmaker, Herero and Owambo pastoral villages in the conservancy area (NACSO 2006). It achieved operational self-sufficiency in 2002 following support from international donors and national NGOs. Revenues from wildlife conservation cover the annual operating costs of the conservancy: staff salaries and other annual program operational expenses. The main revenue-generating enterprise is Damaraland Camp, an up-market, exclusive ecotourism resort (daily rates range from US$560–$670/person) owned and operated by Wilderness Safaris, a South African tour company.

Ecotourism is environmentally responsible travel to relatively undisturbed areas in order to enjoy and appreciate natural and cultural features, while promoting conservation, low negative visitor impact, and beneficial socio-economic involvement by local populations (Ceballos-Lascuráin 1996). Damaraland Camp has fulfilled this concept, receiving international awards in recognition. The camp occupies an exclusive, wilderness site with dramatic scenery. It is accessible only by light aircraft or all-wheel drive vehicle. The main lodge and tented accommodations are luxuriously appointed, using natural materials that blend unobtrusively into the setting. Solar power and other ‘green’ practices for wastewater and solid waste management are featured. The camp offers local natural and cultural history tours delivered by local village members hired and trained from the conservancy membership.

Wilderness Safaris pays an annual land rent and monthly bed levies to Torra and employs over 20 conservancy members full-time (Long 2004; Florry 2006 interview). Training and employment for conservancy members has also been achieved in some 40 additional jobs in the wider Wilderness Safari lodge network (Florry 2006 interview; Weaver 2007 interview). Damaraland Camp's annual income contribution to the conservancy has grown steadily from ˜N$50,000 in 1997 to over N$300,000 in 2005 (Wilderness Safaris 2005). The conservancy has earned well in excess of N$2,000,000 from the camp since its inception ($1 US equalled about $7.5 N in this period). Damaraland Camp has been the single largest contributor to conservancy revenues and the single largest employer (Wilderness Safaris 2005). Indeed, up-scale ecotourism lodges generated over N$7.6 million for all of Namibia's conservancies in 2005, amounting to nearly 56% of overall conservancy income (NACSO 2006).

A key feature of the partnership between Wilderness Safaris and Torra Conservancy is the land tenure arrangement for the Damaraland Camp. Wilderness Safaris was first introduced to the community by IRDNC. IRDNC acted as a broker and facilitated ‘role playing’ with a local community committee to prepare them for negotiations with Wilderness Safaris (Owen-Smith 2007 interview). Wilderness Safaris negotiated a lease for the Damaraland Camp (Salole 2003) on communal lands that form part of the Torra Conservancy.

Wilderness Safaris assumed 100% financing and risk for Damaraland Camp under an initial 10 year joint venture agreement with the Torra Conservancy which has been extended. An end goal was 100% ownership of Damaraland Camp by the conservancy. This has not progressed and there is little prospect for this in the foreseeable future. Reportedly, the significant capital re-investments required to maintain an exclusive and remote operation like the Damaraland Camp have exceeded the conservancy's fiscal capacity to attain outright ownership or even a strong equity position in the development (Wilderness Safaris 2005; Van Smeerdijk 2006 interview).

My site visits, field observations and key informant interviews suggest that while the ecotourism joint venture has produced significant training, employment and conservancy revenues, most conservancy households remain highly impoverished, with no apparent benefits from the ecotourism enterprise partnership. I learned about a lack of transparency in conservancy decision-making regarding community investment priorities for tourism revenues. Other allegations were that ‘local elites’ emerged who appropriated jobs, vehicles and other conservancy benefits for themselves. One anonymous informant remarked that many community meetings and household visits were conducted to consult with villagers during the start-up of the conservancy and the enterprise partnership, but many fewer opportunities were now available to learn about how much revenue was in the conservancy bank account and how funds could be allocated for community benefit. Another anonymous informant complained about the slow pace to achieve management training, higher paying jobs in Damaraland Camp and the ongoing deferral in conservancy ownership of the enterprise.

The Torra Conservancy case illustrates a substantive partnership between the local community level and an international tourism enterprise. The conservancy has made an exclusive and attractive site available for the ecotourism development. Local conservancy members have been trained and hired to run the operation, bringing their local knowledge and culture into the ecotourism product and activities. The international partner has invested its capital to develop the property and applied its expertise, global reach and resources to promote and attract international tourist patronage, generating significant local wages and revenues for those privileged with jobs and the conservancy management committee. These features draw on the respective strengths of the partners (Berkes 2007). However, partnership arrangements have also perversely served to limit empowerment and benefits at the local community level. Further discussion follows to elaborate on these observations.

3.3. Ehi-rovipuka Conservancy and community conservation partnerships

Ehi-rovipuka was officially registered as a conservancy in January 2001. Ehi-rovipuka means ‘the place of wildlife’ in Otjiherero. The overall area of the conservancy is 1975 km2 (NACSO 2006) and Herero people make up the conservancy population of ~2500.

Ehi-rovipuka Conservancy has about 700 registered members currently (Ujaha 2007 interview). Membership, similar to other conservancies, is open voluntarily to all adults 18 years old or older who have lived in the conservancy for at least three years (Ehi-rovipuka Conservancy 2000). The Ehi-rovipuka Conservancy possesses a variety and abundance of wildlife. The conservancy shares boundaries with Etosha National Park and several other conservancy and tourism concession areas. Boundaries were negotiated with neighbouring communities, a process that took over three years (Ujaha 2007 interview).

I earlier detailed an evolution of CBNRM in Namibia. A principal interest in my wider doctoral research was to further understand the place of ‘Community’ in this community-based conservation process. I sought to learn more about how villagers actually participated in and benefited from CBNRM and the conservancy. I conducted structure interviews with 40 villagers in Ehi-rovipuka Conservancy (Appendix 2). I wanted to probe understandings and perceptions about how the conservancy actually started up. What level of community self-organization and participation had there been?

The collaboration of the NGO conservationist Garth Owen-Smith with Herero headman Kephas Muzuma was especially noted by villagers, reinforcing the importance of individual leadership, cross-cultural collaboration, shared knowledge, norms and visions for community-based conservation (Stern et al. 2002). Kephas Muzuma was one of the four headmen mentioned earlier who started the community game guard program in the early 1980s.

I inquired about whom from the community was involved in conservancy start-up. Most of the 40 villagers interviewed (85%) noted that a task force of villagers was created by the headman and council, receiving training from IRDNC. The task force took the conservancy idea out to the villages, built understanding and support and helped negotiate the boundaries, a protracted process lasting over three years. I asked about how the boundaries of the conservancy were established. Those that could reply (63%) recognized the process of negotiations with surrounding communities and traditional authorities. A relatively large proportion (43%) did not know how the conservancy boundaries had been formed. Important points of understanding made by some villagers noted that boundaries defined rights of access to wildlife only and the conservancy included those communities who had agreed to share wildlife. Grazing, water rights and other resource access are not subject to the exclusionary role of the conservancy boundaries. The boundaries are well known at the community level; 80% of the villagers interviewed indicated they knew the boundaries, or at least, the villages that made up the conservancy.

Therefore, conservancy start-up featured top-down and bottom-up dimensions (Berkes 2002; Young 2002). The idea originated and was enabled from outside and at higher levels of organization than the local community level. Yet, there was a high degree of self-organization at community level, especially in negotiating and finalizing the boundaries.

I earlier portrayed how dense social networks of NGOs evolved in Namibia to support CBNRM, conservancies and mediate international donor support (Figure 2). I wanted to learn how aware local villagers were about partnerships in community-based conservation. Most (85%) identified IRDNC as the main partner, followed by the Namibia Community-Based Tourism Organization (NACOBTA) mentioned by 43% of respondents, then the Ministry of Environment and Tourism, the ministry's Integrated Community-Based Ecosystem Management Project and the World Wildlife Fund (WWF US), at 33% each. Several other partners or cooperating groups were mentioned once or twice.

IRDNC has been a supporting and ongoing partner from the beginning, raising and channelling donor funding into the conservancy to support management committee operations, community game guard salaries, construction of a conservancy office, purchase of a vehicle and enterprise development, as well as project planning and wildlife management training. NACOBTA has tried to assist the conservancy to find an investor for a joint venture tourist lodge (Katjiuongua 2007 interview), but without success. WWF US has actively supported annual game counts and related data handling and reporting. The Ministry of Environment and Tourism (MET) has played a central role in the original registration and gazetting of the conservancy, as well as approving annual wildlife quotas based on conservancy game counts and recommendations. Overall, there was high community awareness about the involvement of multiple partners and cooperating groups, but specific roles of the various partners were less well known by villagers.

A community-level perspective of vertical and horizontal partnership linkages is derived from the structured villager interview responses and meetings with community key informants (Figure 3). It differs from the earlier picture portrayed in Figure 2. Regional level linkage emerges more clearly, with IRDNC serving as the lead NGO at the regional level, coordinating support to Ehi-rovipuka Conservancy from other national NGOs. Once IRDNC brokered initial relationships with other NGOs, they then formed direct bilateral relationships with the conservancy to provide various technical supports. Relationships to central government are mainly with the Ministry of Environment and Tourism. Relationships with the conservancy member villages, a trophy hunting enterprise and traditional authorities are also shown, illustrating the greater importance attached to cross-level linkages at the local level.

Ehi-rovipuka Conservancy's benefits from CBNRM are much more modest than in the Torra case. The conservancy still depends on donor support channelled through NGOs, especially IRDNC. A trophy hunting enterprise has started up, generating some revenue for the conservancy and wild meat sharing with households. The community has foregone its own harvest of wildlife to support the start-up of the trophy hunting enterprise. Villagers reported that they received only limited amounts of meat from the trophy hunter and some households had received none. I learned that a benefits distribution plan had yet to be finalized for the conservancy and there were complaints from villagers about both the lack of skills and transparency for financial management by the conservancy management committee (Hoole 2008).

3.4. An ecotourism enterprise vision

Ehi-rovipuka Conservancy shares a boundary with Etosha National Park (Figure 1). Etosha is 22,270 km2 in area and is the major tourist destination area in Namibia, attracting at least 156,000 visitors a year (Turpie et al. 2004). Yet, neighbouring communal areas receive few benefits from tourism and the park. Indeed, national parks are recognized as important ecotourism destinations globally and guidelines have been developed for sustainable tourism in national parks (Eagles et al. 2002; Campbell et al. 2008; Eagles 2008). However, in Namibia a parastatal tourism agency and the national parks authority have appropriated most ecotourism development opportunities. In fact, most of the tourism destination facilities and services are concentrated in Namibia's national parks, a common pattern in Southern Africa (Child 2004a).

The Herero of Ehi-rovipuka Conservancy were forcibly relocated from the western part of Etosha in the 1920s by the colonial South African administration. The Herero had traditionally used and occupied the park, including grazing stations, hunting areas and ancestral grave sites. When they tried to return to the park area in the 1970s this effort was spurned by a park conservation agenda and the building of the park fence (Hoole 2008). My structured interviews with villagers revealed that most (88%) believe they receive no benefits from the park to this day. Several villager commentaries are illustrative: “We cannot even bury our dead there any more” and “The colonial system gave a lot of pain. We had hoped at independence that we might get some rights but nothing has come. We are crying from the past until now.” And, “There you come to the wound. People get much pain when they hear of the park.” I have characterized the dislocation of a community such as the Herero from traditional lands in a national park as a de-coupling of indigenous people from their local ecosystems (Hoole 2008).

I inquired further about the benefits villagers would like to receive from Etosha National Park. A joint venture tourism enterprise is highly sought inside the park. The conservancy has made repeated requests to the Ministry of Environment and Tourism for entry passage and rights to develop an enterprise based upon wildlife viewing inside Etosha (Uaroua 2007 interview; Ujaha 2007 interview). This vision derives from the ecotourism model described for the Damaraland Camp partnership with Torra Conservancy, but has not been realized. The conservancy has long sought to attract a private partner for an ecotourism venture on its own conservancy area albeit without success.

4. Discussion

My results demonstrate that Namibia's CBNRM program and conservancies have evolved as a dense network of institutional arrangements. National NGOs serve as boundary and bridging organizations both appropriating and mediating international donor contributions that provide the core funding for CBNRM. Multiple linkages and networks, both horizontal and vertical, engaging local, regional, national and international levels of organization are evident. Partnerships have supported enterprise development in the tourism sector and partnerships have also been featured in capacity-building and institutional strengthening programs for the conservancies.

Ecotourism enterprise partnerships premised on a stunning wildlife spectacle and wild scenery have proven important for community-based conservation in Namibia. The Damaraland Camp, the Torra Conservancy joint venture model, is much sought after by other conservancies like Ehi-rovipuka. However, it is likely that potential investors in Ehi-rovipuka's case are not interested because the best sites for such an enterprise are found inside the Etosha National Park. Market competition from other ecotourism lodges and camps, as well as site exclusivity, are likely reasons for the lack of success in Ehi-rovipuka attracting an ecotourism enterprise investor. So too are issues of scale mismatch inherent in the relatively small sizes of conservancies, contrasting with the regional scale of wildlife distributions, scenic and cultural attractions that ecotourism is premised upon (Cash et al. 2006). An opportunity to develop a joint venture ecotourism enterprise inside Etosha could overcome such constraints and serve to re-couple social-ecological linkages that have been de-coupled by successive park management and tourism development policies. Indigenous groups seek control over traditional lands as an essential element in restoring their societies and indigenous entrepreneurship is often used as a tool towards self-governance (Berkes and Adhikari 2006).

I suggest that there has been asymmetrical power sharing in ecotourism enterprise partnerships (Adger et al. 2006). To illustrate this, it is important to think first about how tourism functions as a social and economic activity. Obviously, a first requirement is the tourists themselves. In the case of ecotourism enterprises in Namibia, these are an affluent clientele of international origin. These tourists must travel long distances to the remote destinations in Namibia. They are enticed to do so by both the wildlife and wilderness spectacle, as well as the assurance of high quality, reliable and secure services. In order for the prospective tourist clientele to become aware of these opportunities and make a choice to come to Namibia, they must first learn about the attractions and services and then be sold on the choice to select this particular tourism product in a highly competitive market place. I apply a classic model of the tourism functional system (Gunn 1979) in order to frame the consideration of power in an ecotourism enterprise partnership such as Damaraland Camp.

Table 1 summarizes the different roles and sources of power among private enterprise, the state and the conservancy in the tourism functional system. The private enterprise partner takes the lead role and effectively controls most components of the system. It is the private enterprise that markets the destination and ecotourism resort internationally through sophisticated lure pieces, the worldwide web and a bookings infrastructure. It is also the private enterprise that manages the transportation linkages to the destination, and of course, it is the private enterprise that capitalizes the resort development itself. The communal conservancy, in assigning land tenure to its private sector partner, holds important leverage in providing the exclusive development site for the enterprise. In recognition, the conservancy has received annual land rent, employment and a share of revenues. However, there is limited to no engagement of the conservancy in the other parts of the tourism functional system. These are effectively controlled by the private enterprise partner, including the ownership of the capital investment. This inherent power imbalance ensures that most of the revenue generated from the partnership, most of the ‘high-end’ management employment and attendant governance is externalized to regional and headquarters operations of the private enterprise in Windhoek and Johannesburg. I suggest that this situation severely limits the benefits to the local communities and their effective role in governance. One key informant indicated that certain private tourism partners with conservancies in the Caprivi region of Namibia seem to have forgotten obligations to increase the management roles and ownership stake of conservancies, once they secured the land tenure for their operations (Owen-Smith 2007 interview).

There is also the power imbalance earlier described between the state and conservancies regarding access to and the share of potential benefits from ecotourism in the national parks. The state, through its Ministry of Environment and Tourism, denies the Herero of Ehi-rovipuka Conservancy rights of access to traditional territory inside Etosha National Park. It is ironic that the same ministry actively promotes CBNRM and the conservancies, while continuing to deny the Herero of Ehi-rovipuka the access and opportunity to develop a joint venture ecotourism enterprise inside the park.

Conclusion

A prevailing and distinguishing characteristic of Namibia's CBNRM program is the facilitation and support of CBNRM by a dense network of national NGOs. Multiple cross-scale and cross-level linkages have developed in a dense CBNRM network and many partnerships have emerged. Critical convergences of persons and events were featured in the early evolution of CBNRM and the rapid scaling up of conservancies from an initial 4 in 1998 to 50 in 2007. My findings show early and significant community conservation effort and self-organization of conservancy formation under CBNRM. Also, certain Namibian NGOs have evolved as bridging organizations at national and regional levels, mediating the contributions of international donors and the legal requirements of central government with local conservancies, and facilitating capacity-building at communal conservancy level.

National NGOs have successfully attracted international donor funds in the name of community-based conservation. Wildlife conservation has been achieved under CBNRM in Namibia since its beginnings in the early 1980s. It is also evident, from the large and diverse establishment of professional and technical cadres, vehicles, offices and equipment in Windhoek, that NGOs have appropriated a good deal of CBNRM donor funding for themselves. This stands in sharp counterpoint to weakly managed and resourced conservancy management committees, weak capacities to manage and influence partners, and persistent, grinding poverty in conservancy villages (Hoole 2008). Community empowerment and development have not been achieved simultaneously with conservation (Berkes 2004).

Ecotourism enterprise partnerships have generated community employment and revenue benefits from wildlife-based ecotourism. Yet, wider community realization of such benefits has been blunted by local elites who have apparently captured most of the local benefits. The prevailing perspective that emerged from interviews with villager and community key informants was that few if any benefits are realized at the village household level. Power asymmetries inherent to ecotourism partnerships between private enterprises and local conservancies have changed very little since the partnerships were first created. In fact, these have become entrenched, with private enterprises receiving most of the benefits, thus limiting community benefits and empowerment. Village households remain highly impoverished and disconnected from the conservation successes achieved, common to other partnerships featured under CBNRM (Long 2004; Central Bureau of Statistics 2005).

Managing power relations and creating capacities to retain the place and voice of ‘Community’ remain big challenges (Lachapelle et al. 2004) in CBNRM. My findings suggest that most of the power that develops under CBNRM is externalized to national and international operations of NGO and private enterprise partners. This has been demonstrated by reference to the tourism functional system. Table 1 illustrates that management of most key components of the tourism system is held by the private tourism enterprise and government. The community role is mainly relegated to providing local labour and land tenure for the tourism enterprise development. As well, my field observations suggest that NGOs and government retain most of the donor funding or revenues generated in the name of CBNRM to support their operations. While conservation objectives and the NRM part of CBNRM may have progressed, the ‘Community’ part has been overtaken and is losing its place (Blackie 2006). More equitable power sharing at community level is needed if CBNRMs theoretical and promised benefits for ‘Community’ are to be realized.

Multiple partnerships in a globalizing world have been shown to be important in successful community-based conservation cases. This has been equally demonstrated for CBNRM in Namibia. However, partnerships may do more harm than good for communities if they are rigid and entrench power imbalances. Further inquiry into dimensions of power such as knowledge sharing, benefit and revenue distribution and how parties in partnerships advance their respective interests are highly topical. My prognosis is that more dynamic, flexible partnerships featuring equitable power sharing with communities are needed to effectively alleviate poverty and sustain community-based conservation.