Dynamics of common pool resource management in the Okavango Delta, Botswana

Abstract

This paper discusses the historical dynamics of common pool resource use and management in the floodplain of the Okavango Delta by comparing Ikoga and Seronga, two multi-ethnic villages located along the Delta panhandle. During colonial and post-colonial times, many local institutions for managing and using common pool resources were dismantled. Despite this trend, open access has not led to overuse of common pool resources. The paper argues that despite the marginality of the area there is relatively little interest in the commercial use of common pool resources since the diamond industry and tourism provide a relatively high income. While Ikoga residents fail to capture gains from tourism, in Seronga some gains do come from community-based natural resource management. However, these gains, or the failure to receive them, can lead to conflicts that take an ethnic shape when local elites benefit differently and inequalities are perpetuated.

Keywords:

• Community-based natural resource management
• common pool resources
• ethnicity
• Botswana

1. INTRODUCTION

There is a broad literature arguing that a tragedy of the commons occurs when common pool resources become de facto open access resources (e.g. Acheson, 1989; Feeney et al., 1990; Becker & Ostrom, 1995; Agrawal, 2002; Haller & Merten, 2008). When local institutions are dismantled and replaced by state institutions, for example, common pool resources can become de facto open access if the state lacks the funds to enforce its own rules, laws and regulations (Haller, 2005). This may be followed by degradation and stress on natural resources due to a lack of effective management. In Botswana, however, such pressures on natural resources do not appear to be as prevalent as this line of argument would suggest, and this paper explores the reasons for this difference. On the one hand, Botswana has developed a relatively strong economic base due to income from diamonds and tourism (Parson, 1984; Valentine, 1993; Mbaiwa, 2003; Poteete, 2007). On the other hand, many local institutions have been dismantled or disempowered in many respects, with the result that there is de facto open access in some sectors (such as fisheries, pasture, veld products) while local access is heavily restricted in others (such as wildlife hunting). Reserving wildlife for tourism purposes can redistribute some income to areas with substantial wildlife populations, and it benefits the tourism businesses, but this is not consistently true in the other cases described in this issue or, necessarily, in other comparative studies (see the African Floodplain Wetlands Project cases in Mali, Cameroon, Tanzania, Zambia and Botswana compared by Haller, 2005).

This paper analyses common pool resource use in Botswana through a close examination of two contrasting villages, Ikoga and Seronga, in the panhandle of the Okavango Delta. It shows that historical pressures on common pool resources in the panhandle area have been reduced in Ikoga because of direct benefits (remittances, government payments and infrastructure) that have made local residents and the immigrants less interested in the commercial use of wildlife, fisheries and tourism. In this case, however, there are no real, sustainable benefits from tourism through community-based natural resource management (CBNRM) because of the difficulty of drafting a local management plan, which requires technical expertise the community does not have. Delays in this process have caused frustration, as it is evident that game viewing, recreational fishing and lodging could provide benefits. For their part, the various ethnic groups blame ethnic differences for the problems with implementing CBNRM. These disagreements about CBNRM, however, have not had a negative impact on common pool resources. Benefits from tourism are more evident in Seronga, although unevenly distributed and often reflecting elite capture of the gains from tourism while many residents face the costs of increasing human—wildlife conflict. The paper, then, explains the constraints and opportunities that emerge from government land planning and from CBNRM schemes in the two cases. Examining CBNRM in this broader frame of common pool resources offers a different picture of the effects of land use policy than many previous discussions of CBNRM, which often focus more exclusively on harnessing income from tourism and participatory approaches to conservation (e.g. Hulme & Murphree, 2001; Swatuk, 2005; Blaikie, 2006; Haller & Galvin, 2008).

To explain these developments, we present a historical description and use an institutional approach by examining changes over time in local and state structures and policies for managing natural resources. The paper shows that various ethnic groups – including indigenous Basarwa who relied on hunting and gathering, along with Bayei fishermen and Hambukushu agro-pastoralists in the Okavango Delta – established local institutions, which included rules for coordinating their use of common pool resources. In particular, arrangements were made between semi-permanent extended kinship groups and installed chiefs of the dominant Tswana ethnic group during the British Protectorate period (1885–1966). Furthermore, administrative colonial and post-colonial policy changes have largely replaced common property regimes with primarily open access tenure on communal (formerly called ‘tribal’) land, including the largest portion of land in Botswana. Under such conditions, exploitative and detrimental land and resource use would often be expected to develop. However, negative effects are rarely observed among communities in remote parts of the Delta, especially in the panhandle floodplain on which this paper focuses. The cases of Ikoga and Seronga illustrate this process through secondary gains from the diamond-based and tourist-based economy that exclude local people from direct resource use but deliver limited financial benefits to some of them. Although this process does not lead to overexploitation of common pool resources, it may perpetuate poverty while increasing conflicts over benefits.

The paper first discusses methods used and the location and characteristics of the study area. This is followed by three sections providing background and context relevant to both case studies: first, the various ethnic groups and the historical shifts in land use and settlement patterns; second, the main livelihood resources; and, third, the political and economic changes that have affected land and resource use over approximately the past 50 years. The case studies of Ikoga and Seronga follow, and the concluding section revisits the paper's central argument that, despite persistent inequalities in the distribution of benefits in geographically and politically marginal communities, changes in access to the commons in the Delta have not put an excessive strain on natural resources.

2. METHODS AND STUDY LOCATION

This study draws on a combination of fieldwork conducted and studies published by the authors as well as secondary literature on livelihoods and resources in the Delta. These include ethnographic fieldwork done by Roland Saum in Ikoga from 2003 to 2004 for the African Floodplain Wetlands Project led by Tobias Haller (Haller, 2005; Saum, 2006). Methods used in Saum's study included participatory observation, interviews with members of all ethnic groups in the village, biographies, oral history, household surveys, archival work and secondary literature research. The Ikoga case contrasts with Seronga, which is on the opposite side of the panhandle and is much less accessible. Data for this study were gathered by Parakh Hoon during primary ethnographic fieldwork in 2000, with additional research by Hoon and Rachel DeMotts in 2007 and 2008. The cases also reflect insights from work conducted at the Harry Oppenheimer Okavango Research Centre (e.g. Mbaiwa, 2003; Kgathi et al., 2004, 2005; Swatuk, 2005; Thakadu, 2005; Magole, 2008).

The two village areas where fieldwork was conducted form the basis for comparison of benefits from common pool resources in the panhandle of the Okavango Delta. The Okavango River brings water into Botswana from the central Angolan highlands, entering northwestern Botswana after cutting through Namibia's Caprivi Strip. The river then follows a narrow rift for about 90 km, forming a panhandle that fans out into wetlands before terminating in the sands of the surrounding Kalahari Desert savanna. The change in altitude is just sufficient to limit the river to fanning out over not more than 15 km. The distance from the apex to the base of this section is about 170 km. Estimates of the size and extent of the Delta vary from 10 000 to 16 000 km² (Hughes & Hughes, 1992). Scudder et al. (1993) note a maximum aerial extent of 20 000 km². The climate of the Okavango River floodplain is semi-arid, with an average rainfall varying from 400 to 600 mm in the southwest. Minimum monthly temperatures around 15–17°C from June to August contrast with 22–24°C between October and January (Ministry of Lands and Housing, 2001).

The Okavango Delta is one of Africa's most significant areas for fish and wildlife (Larson, 1971; Scudder et al., 1993). It is home to a wide variety of species: 164 mammals, 157 reptiles, 540 birds, 38 amphibians, 80 fish, 38 amphibians, and 5000−10 000 insects (South African Savannas Network, 2001). Seasonal movements of fish take place in response to flood changes, as many synchronise their breeding activity in response to water levels and temperatures. The areas on either side of the panhandle are forested with woodlands dominated by acacia (Acacia erioloba) and mopane (Colophosermum mopane). The summer rains in the Angolan highlands, heaviest in January and February and up to three times as high as in the Delta area, generally bring a flood in the winter months (June–September). Agricultural opportunities, however, are constrained due to a shortage of arable land and the variable presence of water (Ministry of Lands and Housing, 2001).

3. HISTORICAL AND ETHNOGRAPHIC BACKGROUND

The Ngamiland region has the broadest linguistic and cultural diversity of any district in Botswana. Signs of human habitation date back as far as 100 000 years to the nomadic San hunter-gatherers, who are now a minority in the settlements (Tlou, 2000). The largest ethnic group in the area are the riverine-oriented matrilineal Bayei, who emigrated after 1750 following the expansion of the Lozi Kingdom in Zambia. They settled on islands and along riverbanks and brought new riverine technologies such as canoes and fishing nets to the area, and have generally maintained good relations with the Basarwa. Slave traders from Angola caused the pastoral Hambukushu to migrate to Ngamiland after the 1850s. Unlike the Bayei, they preferred the drier mainland with its better grazing conditions, which had been allocated to them by the dominant Batawana.

The pastoral Setswana-speaking Batawana were the last migrants to the area and, along with the Herero and Ovambo from what is now Namibia, were responsible for the local introduction of cattle. Arriving at the end of the eighteenth century, they became the politically and economically dominant ethnic group in the region (Tlou, 2000). Much later, in the 1970s, several thousand Hambukushu fled from the colonial and civil wars in Angola and settled around the Etsha villages south of Ikoga. Until recently, both Bayei and Hambukushu lived in spatially separated family units in scattered communities and were very mobile (Larson, 1970, 1971, 1977; Kirkels, 1992; Kuru Development Trust [KDT], 2000; Tlou, 2000; Saum, 2006). The population in the study area also grew as a result of a reduction in this mobility. The overall growth percentages (ranging between 4.3 per cent in 1981 and 3.2 per cent in 2001), however, have not significantly affected population density in the rural parts of Ngamiland District (1.06 persons/km² in 1991), as it has been outweighed by urbanisation, especially in the regional capital of Maun (Ministry of Lands and Housing, 2001).

While the relationships among these different groups are crucial to understanding resource use, it is beyond the scope of this paper to interrogate the deep historical meaning of either the ethnic terms used for different groups of people in Botswana or how these have evolved over the past several hundred years (for such work, see Larson, 1970, 1971, 1977; Kirkels, 1992; Hitchcock & Biesele, 2000; KDT, 2000; Tlou, 2000). Identities are addressed briefly here, rather, to allow for later comment on how membership in ethnic groups is asserted relative to the creation of CBNRM projects in both Ikoga and Seronga. It is, however, crucial to note that ethnic identities in this context cannot be understood separately from historical and current patterns of immigration, land and resource use strategies, and the impacts of shifting state policy. Identities are not fixed, but rather subject to continuous renegotiation and construction. They may be asserted in specific political settings, as we will show, and are not the basis for the knowledge differences and conflicts that we describe in this paper, but rather may be mobilised when interests come into conflict and unfairness is perceived.

4. RESOURCES AND LIVELIHOODS

Arable land here is not widespread and land has mostly been used for grazing. Signs of overgrazing (which vary, but can include trampled or minimal grasses and increasingly large patches of exposed soil) are visible in some areas near the riverbanks and close to permanent settlements, but since cattle lung disease in the 1990s resulted in government killing and compensating for most cattle, pressure on pasture is now minimal. Fish in the Delta are used locally and not for large-scale commercial sale. According to Mosepele (2001),1 there are currently no signs of over-fishing, decreases in fish stocks or changes in occurrence of species. During the Protectorate period, wildlife numbers were decreasing because of commercial hunting, but most species are reported to have recovered quickly or even increased since post-colonial hunting regulations were established and enforced. Most of the species in the Delta and surrounding areas are abundant and it has become one of the most popular regions in Africa for photographic wildlife tourism (Saum, 2006).

As in other remote rural areas in the region, residents of Ikoga and Seronga use mixed-subsistence strategies including farming, fishing, collecting wild plants, and livestock (cattle and goat) herding. Hunting has become limited, since there is no legal local market for bush meat and because the area is zoned for non-consumptive wildlife use (photographic tourism) (Mbaiwa, 2003; Thakadu, 2005). People practise rainfed agriculture (sorghum, millet, and maize) with ploughs and hoes in the drylands, and flood recession and irrigated agriculture (fruits and vegetables) on the river fringes and on islands.

Livestock density in Ngamiland has always been lower than in other parts of Botswana (Kirkels, 1992) and it diminished even more after the eradication of 300 000 cattle in the district in 1996 because they were affected by contagious bovine pleuropneumonia (CBPP). In 2002 the total herd had still failed to reach even one-half of the original size before the eradication (personal communication, Hannalore Bendsen, Harry Oppenheimer Okavango Research Centre, to Saum, March 2002). Over 30 species of fish are caught by local fishermen, including African pike (Hepsetus odoe), bream (Cichlidae family), tilapia (Tilapia ruweti), and tigerfish (Hydrocynus vittatus) (see also Larson, 1971). Traditional fishing methods in arriving or receding floods may include the use of grass mats, traps and baskets, depending on location, water level and species sought. The much more profitable gill net fishery is only practised by those few who own or have access to mekoro (canoes) or boats and nets.

5. POLITICAL AND ECONOMIC SHIFTS SINCE THE 1950S

5.1 Pre-independence local institutions

From 1750 until the beginning of the twentieth century, local institutions regulated access to fishing and hunting grounds in the economies of the Bayei and the Hambukushu in Ngamiland, but according to Tlou (2000) these were largely dismantled during the Protectorate period. The British administration of Bechuanaland declared most of the land to be ‘tribal’ and its administration was vested in the powers of the chiefs, assisted by appointed local Bayei headmen. Access to demarcated fishing grounds required negotiation with resident groups. The Bayei or Hambukushu fished collectively with their own kin and had access to separate pools and channels; these areas were monitored and assessed before local headmen assigned use rights in collective meetings (kgotla) – meaning that, until early independence, fisheries were partly coordinated by the local Bayei headmen.

Early in the nineteenth century, hunting laws came under the jurisdiction of the dominant Batawana chief. This allowed people to hunt small game in the entire tribal area, including the Delta. No reimbursement was requested for small game, and permission to hunt big game was issued through the chief's intermediary headmen. Communities were allowed to devise their own rules and sanctions such as designating hunting grounds. Ritual observance of rules for the Hambukushu and the Bayei was also important. Villagers did, however, ‘give permission to fish or hunt in each other's territory’ (Tlou, 2000:25). Access to resources, then, worked in part on the basis of invitations between local groups and reciprocal access based on the location of particular resources and knowledge about how to use them (Cassidy, 2000; Saum, 2006).

5.2 Post-independence institutional changes

After independence in 1966, laws for allocating residential and arable land were revised by the Tribal Land Act and subsequently fell under the jurisdiction of the Land Boards. According to the Tribal Grazing Land Policy, introduced in 1975, land has been zoned into tribal land for communal use, commercial land for individual use (mostly private large-scale grazing farms) and state land (national parks and reserved areas). Most subsistence economic activities took place on tribal land, which accounted for most of the land in Botswana. Under the Tribal Grazing Land Policy (and subsequent regulations) every citizen has the right to settle on tribal land and use any resources attached to it, which means that grazing and other resources (except wildlife) are open access.

The Wildlife Conservation and National Parks Act of 1992 is the legal basis for reallocating Wildlife Management Areas to community-based organisations for CBNRM. Wildlife Management Areas are demarcated areas that create the opportunity for communities to acquire lease rights to areas rich in wildlife and other common pool resources that can then be developed for tourism. Under CBNRM, the Land Boards, the Department of Wildlife and National Parks, and the Tourism Department can grant exclusive access and management rights to wildlife and other resources in demarcated areas to any village or group of villages that organise under a constitution to form a community trust. Most villages involved in CBNRM focus on tourist activities in the hope of generating income at the local level. While people may fish or collect veld products to satisfy some of their basic needs, hunting−which might be much more lucrative−is largely prohibited or too expensive because the fees are high. The focus on commercial wildlife-based tourism is also furthered by the Tourism Act of 1999 and the Joint Venture Guidelines of 1999 that functionally confine the use of common pool resources inside Wildlife Management Areas to commercial purposes.

5.3 Major economic changes

During the Protectorate period, the population of Botswana was almost entirely dependent on cattle farming, subsistence crop production and wage remittances from migrant workers in South Africa (Colclough & Fallon, 1983). People were forced to move from rural areas because of the introduction of hut taxes, restrictive hunting and pastoral policies, demarcation and alienation of the land, restrictions on selling cattle in South Africa, and restrictions on black Africans in trade and other commercial ventures. Households managed to put together a subsistence income only by combining agriculture with wages and migrant remittances (Hopf, 1991; Saum, 2006).

At the national level, the discovery of diamonds and other minerals, an increase in beef exports and massive foreign investment in mining, in partnership with the state, led to an unexpected boom in Botswana's extractive and export-driven economy. The country's Gross Domestic Product increased almost 2000 per cent between 1965 and 1980; the economy diversified and the increasing variety in markets diminished dependence on South Africa (Parson, 1984; Bock, 1998). Increased national income facilitated investment in public and social infrastructure, resulting in a dramatically enlarged government apparatus (Parson, 1984) while reducing agricultural production to lower levels than anywhere else in Africa.

While the number of cattle grew with better marketing and trade agreements, only one-half of the rural population has directly benefited, and the gains have been disproportionately captured by those who are already more wealthy (Colclough & Fallon, 1983; Parson, 1984; Kirkels, 1992). Today beef is Botswana's second biggest export good, mainly to the South African Customs Union and the European Union (under the Lomé Trade Agreement), where it can be sold above market rates. Since the early 1970s, however, cattle in the study area have not proven to be a secure means of investment, particularly after the severe outbreak of CBPP in 1996 that led to the slaughter of nearly all the cattle in Ngamiland and the subsequent closing of the regional abattoir in Maun. In the short term, the eradication of CBPP increased rural income through compensation payments, but in the long run it reduced agricultural outputs (KDT, 2000; Hoon, 2004). This dynamic was reinforced by public labour programmes, drought relief measures, financial assistance policies and food handouts that increased after the killing of the animals. These measures can also be understood as resulting from the constraints under which Botswana beef is allowed to enter the European market. The veterinary department demarcated a buffer zone (including Ngamiland) for the diseases (CBPP and foot and mouth disease) to improve disease control and meet export standards.

Finally, after mining, wildlife-based tourism is the second largest contributor to the Gross National Product today (around 5 per cent). It has increased steadily over the past few years and is considered to be an engine of growth, generating substantial employment, particularly in Ngamiland (Rozemeijer & Van der Jagt, 2000).

6. IKOGA VILLAGE CASE STUDY: RESOURCE TENSIONS, ETHNICITY AND LOSS OF CBNRM POTENTIAL

This section of the paper stems from original anthropological research conducted by Saum in 2002 and 2004 involving participant observation and interviews. It reveals the conflicts, resource tensions and problems that are expressed on an ethnic level but that are based on the frustration of not being able to profit from CBNRM schemes. Ikoga is situated at the southern end of the panhandle, 5−7 km west of the perennial Thaoge tributary, which may at times flood the village. Although it is remote, the village is accessible by a tarred road. The total population was approximately 800 people in 2002, including 41 per cent Bayei, 26 per cent Hambukushu, and 32 per cent people of Tswana origin (Bakgalagadi, Batawana, or Kalanga; see Saum, 2006). Pasture is no longer of major importance since the eradication of cattle, but the 30 richest households still have herds of up to 50 animals. According to a livestock officer in Gumare, livestock donated after the eradication, however, are not considered well adapted to the floodplain vegetation or resistant to the diseases.

Although access to common pool resources is generally open and no longer based primarily on ethnic affiliation, it lies in the hands of individual Bayei and Hambukushu headmen to recognise each individual's group membership. Headmen are chosen on the basis of descent, although local residents say they can choose who to go to for assistance. However, high mobility and newly established formal settlement schemes make it difficult for residents of Ikoga to claim a particular territory and its resources as their own. Ikoga is in an area gazetted by the state for conservation and protection of wildlife, which means that only limited individual or household use of common pool resources, including arable land, pastures, fish, water, and veld products, is allowed under formal law.

There are no other territorial agreements about grazing access between cattle-owning user groups in the research area. People from Ikoga and the surrounding villages are currently using pastures on dry land as well as on the river fringes and floodplains. While farmers usually keep their cattle in kraals (enclosures) at night, livestock grazes freely during the day and strays into farmed plots in Ikoga and other villages, causing damage and necessitating compensation payments. Fishing has also become open access. Laws and regulations concerning fisheries are very general and weakly enforced, regulating only net size (Mosepele, 2001), and there are no territorial agreements excluding people from neighbouring villages. Residents of Ikoga fish during all seasons on a first come, first served basis, in contrast to the past traditional management of collective fishing between August and December. There is little competition or conflict between local people for these resources, however, as fish and cattle have no market value in the area (Saum, 2006). The only known case in Ikoga where fisheries were said to be in decline was as a result of what were claimed to be inappropriate fishing techniques used by people living upstream, who were held responsible for a local decrease in fish in 2002–2003. Ikoga residents stated that their neighbours in Sepopa (and further upstream) were building barriers and dams that prevented the fish from moving back to the main river to breed. Mosepele (2001) also argues that uncontrolled use of collective fishing methods that were traditionally sanctioned by local authorities might interfere with the natural movements of fish in the system and have a detrimental effect on fish stocks.

At the same time, there are occasional conflicts as a result of competition with tourists and the potential to rent out the area for tourism under CBNRM. Tensions do arise in the gill net fishery between residents and recreational fishermen (tourists). Local fishermen and tourists claim access to the same fishing grounds, and thus may disturb each other. Recreational fishermen (represented by camp owners and managers) accuse the gill net fishermen of overexploiting fish stocks, but this is questionable since these concerns remain unsupported by clear evidence of depletion (Mosepele, 2001) and the Okavango fishery is currently under-utilised. Other difficulties are created because of government protection of wildlife as tourism increases. Populations of mammals, reptiles, birds and fish are reported to be growing (Applied Development Research Consultants, 2001; Mosepele, 2001). This can cause problems for local people, who are endangered by living in proximity to wildlife such as elephants and other animals that become habituated to human settlements. This can make it risky for people to use pastures and fisheries, and they can be fined for defending themselves and their crops (Saum, 2006). The Department of Wildlife and National Parks administers a compensation policy in the panhandle for loss of crops and wildlife, but the policy is unpopular with many residents, partly because it does not offer replacement value for what is lost (DeMotts & Hoon, 2008).

The Department of Wildlife and National Parks and the Trust for Okavango Cultural and Development Initiatives (TOCaDi), a local non-governmental organisation, began introducing CBRNM in the area in 1998. People from Jao, Ikoga and Etsha 1−13 were informed that they would be allocated portions of NG-24, a Wildlife Management Area rich in wildlife and flora, provided that they presented a management plan to detail the proposed use of the area. With the assistance of TOCaDi, the villages registered as Jakotsha Okavango Community Trust (the name ‘Jakotsha’ comes from the names of the constituent villages: Ja from Jao, ko from Ikoga, and tsha from Etsha). In 2000 a land use plan was submitted to the Land Board to confirm the lease of part of NG-24, and different land areas were allocated to the participating villages. With support from TOCaDI, the community trust initially did generate some income from tourism, but this has essentially disappeared. The stipulation for lease was that the community must negotiate with a tourist operator to establish a partnership. In this context, the larger project has stalled for several reasons.

Through the process of implementing CBNRM, the inhabitants of Jao, the village in the middle of the Delta, became aware of the enormous economic potential of their natural environment. Most of the economically valuable natural resources, such as wildlife, are found in the area around Jao. But, according to the guidelines of the management plan for the larger area (KDT, 2000), they cannot profit from large parts of the area without the involvement and agreement of other villages. This became a source of frustration for Jao residents, who perceived that other previously uninterested villages suddenly wanted to control their area once it was clearly shown to have value. When this concern was exacerbated by a lack of clear local understanding that the Board is supposed to use the benefits for all the villages involved, CBNRM inflamed the conflict in the eyes of Jao residents. At the heart of this process is the drafting of a so-called management plan – but in Ikoga, as in most of the involved villages, very few people knew about the concept of the management plan and neither did all Board members (Saum, 2006). This lack of knowledge, however, is the crux of many of the problems of implementing CBNRM in a fair, participatory manner.

In addition to the lack of power to make decisions about using resources, CBNRM as a form of governance has also given rise to disagreements that have an ethnic dimension (Saum, 2006). In ethnically diverse villages like Ikoga, resident ethnic groups have articulated different visions of what was happening in the CBNRM process, which are largely reactive and arise out of frustration at the loss of expected benefits. Repeated accusations and complaints by both Bayei and Hambukushu people against each other have created the impression that this sort of ethnic division is rather new to the area, and that it comes from the frustration about the stagnation of CBNRM. It appeared that people began to accuse others because it was almost impossible for them to discern why the CBNRM process was slow and did not benefit them. For example, board members of the Jakotsha Community Trust accused Bayei residents of Jao and members of Kamanakao, a Bayei cultural association, of trying to sabotage its work. The accused were said to be undermining confidence in the Trust by saying it was Hambukushu-dominated, endangering the good relationships among member communities involved in CBNRM (Botswana Daily News, 13 December 2001, as discussed in Saum, 2006). The board members argued that relationships had worsened, especially after the formation of Kamanakao. Residents of the Bayei-dominated Jao village, on the other hand, complained about poor working relations with the Hambukushu-dominated board.

In this case, lack of training and capacity-building caused misunderstandings and were explained with powerful appeals to ethnic divisions, which previously did not appear to be as salient. Adding to the ethnicisation of disputes was the lack of transparency at the board level. The villagers blamed the members of the board for passing resolutions without first consulting the communities that they ought to be representing (Botswana Daily News, 24 July 2000, as discussed in Saum, 2006). There were also rumours that financial issues were not transparent and that access to information was only available through informal contacts. Board members were reluctant to raise financial issues before their communities because they feared that public discussion of finances could increase complications (TOCaDI employees interviewed by Saum, December 2005). The Board has not developed sufficient legitimacy in the eyes of local residents and does not have a reputation for addressing problems fairly.

Furthermore, there is disagreement between the Trust and other village-level institutions such as the Village Development Committee in Ikoga. The Trust is the key actor in CBNRM and considers itself responsible for decentralised natural resource management. The Village Development Committee, however, is functionally the only recognised entity for village-level development and is the more powerful body because of its relationship with higher ranks of the administration and because it manages food-for-work, destitute rations, and other government support schemes for local communities. Development Committee members understand their authority as primary, saying ‘We govern the village’ and ‘We are responsible for development’ (Village Development Committee members, Ikoga, 2002). Although the two committees work with different central authorities, they are nonetheless expected to collaborate. Trust members complained about a lack of cooperation by the Village Development Committee, and the Committee in turn feared that the Trust might challenge its position in the village as soon as CBNRM generated income and jobs, which could disturb development as intended by the Committee.

In Ikoga, then, it is the failure to implement CBNRM or bring anticipated benefits that has been a source of discontent for local residents. This frustration has manifested itself in ethnic terms, although it is not about ethnicity per se. Rather, ethnicity has become a tool local residents use to help them understand why they have been unable to benefit as promised.

7. SERONGA VILLAGE CASE STUDY: THE PROMISES OF DIFFERENT CBNRM APPROACHES

Seronga is located on the opposite side of the Okavango River from Ikoga. The eastern side of the panhandle is more difficult to reach because the roads are unpaved and there is no bridge; it also has not yet been electrified. Like much of Ngamiland, livelihoods in this area are diversified and locally based on agriculture, livestock, fishing, and veld products (Plantec Africa, 2003). In 2001 Seronga's population numbered 1681, which represents nearly 150 per cent growth over the previous 10 years. There is, however, no land use planning to speak of in Seronga or the surrounding area (Plantec Africa, 2003), and growth is projected to slow substantially in the coming years to a rate as low as 1.4 per cent (Plantec Africa, 2003). The remoteness of the area contributes to the relative lack of government services in the area compared with the opposite side of the river, which means that residents of the Seronga area are less able to depend on this kind of assistance than those in the Ikoga area.

Seronga is the main village in NG-12, a Controlled Hunting Area. Residents of Seronga have been involved with CBNRM programmes through the Okavango Community Trust (OCT) and the Okavango Polers Trust (OPT). The OCT includes five villages (Beetsha, Eretsha, Gudigwa, Gunitsoga and Seronga) and is overseen by a manager and a board of village representatives who are mostly Hambukushu. The OPT, on the other hand, was privately initiated by the local Bayei, who felt excluded from the OCT and had the interest of the private sector in establishing mokoro (dugout canoe) trips in the area. The OPT is not a trust under CBNRM, but rather consists of individual members – the polers themselves – who join, pay a membership fee, and are then employed to take tourists on photographic safaris in the Delta by boat (Sorensen, 2003).

CBNRM here has had varying effects on local views of conservation and resource use (Hoon, 2004). On one hand, as in Ikoga, the increasing presence of elephants, which are responsible for the majority of human−wildlife conflict in the area, is a clear problem (DeMotts & Hoon, 2008). Consultations under the planning process for the Okavango Delta Management Plan showed there was widespread concern about the resulting danger and loss of livelihoods. Compensation for damage is also a sore point, as residents feel it is inadequate and slow in coming (Bendsen, 2005; DeMotts & Hoon, 2008). During public meetings as part of the process of writing the Management Plan, residents suggested creating distant boreholes to draw elephants away; some also thought that elephant populations should be redistributed or culled (Bendsen, 2005).

On the other hand, there are some local benefits from tourism in Seronga that residents of Ikoga do not receive. Both the OPT and the OCT employ individuals, creating at least some jobs in an area where formal employment is limited. This is a significant benefit for those able to access it. The OPT, for example, provides income for its approximately 75 members, which is crucial in addressing basic household needs (Sorenson, 2003). The OPT has been able to reach the economic margins by providing job opportunities for people who in general are less well-off, which for CBNRM schemes is a rare achievement (Sorensen, 2003). At any given time the OCT also employs 50−60 people and also provides sitting allowances to board members for meetings. Some of these positions are considered superfluous and the sitting allowances have risen dramatically in the past few years, but those who manage to hold these seats are certainly benefiting from them, as are their families.

Both projects, however, have suffered from management problems and elite capture. The OCT's agreement with Wilderness Safaris generates 2.5 million pula annually (approximately US$360 000 as of May 2009), but many local residents see no real benefit from this income. While there have been some community projects, such as the establishment of tuck shops, too often these fail. A mortuary opened by the OCT in 2007 to offer a service that government and the private sector would not provide was struggling by mid-2008 and its future was unclear. The project, then, may be criticised by some local residents because elites may benefit without having to share the spoils with the larger community. At the same time, it is not surprising that the leadership of CBNRM trusts is often dominated by local elites. Selection processes that take place in kgotla meetings, for example, are likely to yield the same ‘winners’ who benefit from traditional methods of participation and decision-making normally used in the kgotla (Thakadu, 2005). But both the OCT and the OPT have shown signs of struggling to create new processes that engender space for more voices to be heard and to encourage training (Sorensen, 2003). In recent years, the OCT has seen several waves of increased participation from younger leaders in the community, who challenged the old elite in 2001 when they believed a re-tendering process with the previous tourist operator was unfavourable to the larger community (for further details, see Hoon, 2004). These efforts may reflect a coming shift in authority that will challenge the way benefits are distributed and lead to a more equitable use of profits from CBNRM.

8. CONCLUSIONS

Comparing Ikoga and Seronga shows that the direct use and management of common pool resources may not be as crucial as other resources for the livelihoods of local people, who clearly subsist on a variety of alternatives. In particular, assistance may come from the state directly (Ikoga) or via CBNRM schemes and trusts (Seronga). Although in the Ikoga case CBNRM schemes are meant to be established, local stakeholders remain marginal due to a lack of community empowerment, which results in part from a lack of local technical expertise in establishing a management plan to enable a partnership with a safari operator. This failure has led people to make ethnically based accusations about possible reasons for CBNRM's stagnation. At the same time, even successful tourism can create other tensions; for example, between local and sport fishermen or between local farmers and wildlife. Ikoga is sustained largely by remittances and rental schemes, and benefits from being located on a road network with reasonably good infrastructure. Seronga, on the other hand, has not received the same services and infrastructure development as Ikoga, but is much more involved in tourism. Here as well there is ethnic separation in CBNRM, which in this case may actually help distribute benefits more widely, especially in the case of OPT. And yet in both cases access to information about how communities could benefit from CBNRM rests primarily with local political elites, and actual participation in conservation is extremely limited. While local elites may profit more than most residents, for legitimacy's sake they are also expected to deliver some broader benefits to the community. This expectation and the availability of some funds from CBNRM, for example, may sometimes result in projects and spending that the state did not intend, such as the OCT's efforts in Seronga to address social service needs with CBRNM funds.

Botswana provides interesting counter-examples to what much of the commons literature might lead one to expect. Common pool resources were generally not found to be under stress in the study area, because residents of the panhandle are not entirely dependent on the direct use of the commons. Shifts in local and national institutions over time have changed the way people relate to and use these resources, and factors such as livestock disease, changes in the national economy, and population distribution have also played important roles. Despite efforts to implement CBNRM, however, access to and use of gains from tourism remain unequally distributed. Economic changes nationally through diamond revenue and locally through tourism and institutional land rights lead in both cases to higher standards than in otherwise comparable rural places in floodplain areas (especially Malawi and Zambia; see other articles in this special issue). Overuse of common pool resources is not as contentious at the moment as might otherwise be expected in similar circumstances. At the same time, CBNRM's effects on local livelihoods and common pool resource use in the panhandle of the Okavango Delta vary, illustrating the importance of considering CBNRM in the context of other land use policies.

This special issue was produced with the support of the European Union's Sixth Framework programme through the Cross-Sectoral Commons Governance in Southern Africa Project No. 043982. This work does not reflect the Commission's views and in no way anticipates its future policy in this area.

Notes

¹This is supported by statements from villagers, fishery officers and researchers from the Harry Oppenheimer Okavango Research Centre, as communicated to Saum during fieldwork and to Haller in 2007.