Tourism in the Great Limpopo Transfrontier Park

Abstract

The Great Limpopo Transfrontier Park (GLTP) is a transboundary protected area that straddles the borders of Mozambique, South Africa and Zimbabwe. The park's development was partly motivated by the ecological objective of re-establishing traditional migratory wildlife routes once fences between the three countries are dismantled. Besides biodiversity conservation benefits, the park may also provide a basis to generate revenue for conservation and local economic development through tourism. This paper describes current state and private sector tourism within the GLTP and planning initiatives that may promote responsible tourism, and describes the achievements by community-based tourism enterprises and public–private partnerships in generating economic, social and environmental benefits. The livelihoods of people living in the park are outlined in relation to government policies on land redistribution, resettlement and options for the future, and progress in biodiversity conservation and responsible nature-based tourism development within the GLTP over the past five years is evaluated.

¹Research Fellow, Transboundary Protected Areas Research Initiative, University of the Witwatersrand, South Africa. This research was supported by the BMW Chair for Sustainability, University of the Witwatersrand and the Transboundary Protected Areas Research Initiative (TPARI), an IUCN-SA programme funded through the Centre for the Integrated Study of the Human Dimensions of Global Change, Carnegie Mellon University, by way of a co-operative agreement with the National Science Foundation (SBR-9521914). Any opinions, findings, conclusions or recommendation expressed herein are those of the author and do not necessarily reflect the views of the National Science Foundation or Carnegie Mellon. The author would like to thank the anonymous reviewers who commented on a draft of this paper. This paper is largely drawn from Spenceley (2005); the full paper is available at http://www.anna.spenceley.co.uk/files/Tourism%20InvestmentGLTFCAMar05.pdf

1. INTRODUCTION

Transfrontier Conservation Areas (TFCAs) are described as relatively large areas encompassing one or more protected areas which straddle frontiers between one or more countries (World Bank, 1996). In comparison to national parks, TFCAs have the potential to conserve a greater diversity of species within larger geographical areas and to promote co-operative wildlife management between nations (BSP, 1999). TFCAs may also improve opportunities for tourism, by allowing visitors to disperse over greater areas and obtain better quality experiences (Singh, 1999), and by offering more diverse attractions (van der Linde et al., 2001). The World Wide Fund for Nature (WWF) suggests that TFCAs have the capacity to strategically develop sustainable tourism that may support the costs of conservation management, while also providing employment and entrepreneurial opportunities for poor people in developing countries (BSP, 1999).

The 99 800 km² Gaza–Kruger–Gonarezhou Transfrontier Conservation Area was formally established in November 2000, when an agreement was signed by the governments of Zimbabwe, South Africa and Mozambique. In 2001 a smaller area containing the ‘core protected areas’ of the Kruger National Park in South Africa, Gonarezhou National Park in Zimbabwe, and the Limpopo National Park in Mozambique was recognised as the Great Limpopo Transfrontier Park (GLTP) (Figure 1) (Wolmer, 2003). A formal treaty for this 35 000 km² transboundary area was signed in November 2002 (Huggins et al., 2003) as the first phase towards the wider, and re-branded, Great Limpopo TFCA.

Figure 1: Great Limpopo Transfrontier Park and Transfrontier Conservation Area. Source: GoM, GoSA & GoZ (2002).

The TFCA has generated interest amongst conservationists because of its potential to become one of the largest conservation areas in the world. The area could conserve the widest variety of wildlife on Earth, with areas of great cultural and historical value, making it an area of global conservation significance (Hanks, 1998). However, from the initial planning stages, it was clear that there were disparities between the three countries' policies and law enforcement, access to finance, approaches to natural resource management, threats to biodiversity, relationships with neighbouring communities and levels of tourism infrastructure (Gelderblom et al., 1998; Afrosearch, 1999).

Underlying the TFCA process is the Southern African Development Community (SADC) Wildlife Policy. This promotes the establishment of TFCAs as a means for interstate co-operation in the management and sustainable use of ecosystems that transcend national boundaries. In addition, the SADC Protocol on Wildlife Conservation and Law Enforcement promotes regional co-operation in the development of a common framework for the conservation of natural resources and the enforcement of laws governing these resources and their sustainable use, and it provides the framework for transfrontier development and management (Mtisi & Chaumba, 2001).

The institutional structure of stakeholders working on the GLTP has a hierarchical arrangement that incorporates a Trilateral Ministerial Committee (TMC), a Joint Management Board (JMB), a co-ordinating party and management committees.

The TMC consists of the Ministers designated and mandated by the participating countries and is responsible for overall policy guidance in the development of the transfrontier park. The committee is chaired on a two-year rotational basis and meets at least annually (Codex, 2001–4; Schuerholz, 2003). The JMB includes senior representatives of the competent authorities of each country. The four representatives of each member country are seconded from the public sectors of Planning and Finance, Security/Home Affairs, the provincial/regional authority and the national agency responsible for protected area management (Schuerholz, 2003). The board is responsible for interpreting the political directives of the TMC into a set of operational guidelines and policies. It is also in charge of approving development and management plans, harmonising the expectations of the various parties with respect to the transfrontier park, and monitoring the implementation process of the establishment, development and management of the park. As with the TMC, the JMB is chaired on a rotational basis and meets biannually (Codex, 2001–4). The chair is currently based in Zimbabwe.

The JMB is supported by four management committees that advise and assist in the implementation and day-to-day management of the GLTP. They are composed of representatives appointed by the competent authorities of the participating countries and/or representatives delegated by the relevant ministries (Codex, 2001–4). The management committees address legislation, finance and human resources, conservation, tourism, and safety and security (Schuerholz, 2003). They are responsible for implementing the JMB's action plan and need to ensure full participation by all appropriate stakeholders when preparing policy recommendations and resource management plans (Codex, 2001–4).

The Great Limpopo Transfrontier Park Joint Management Plan (JMP) encourages the park to work closely with the tourism industry within a framework of sustainable tourism. The plan aims for a market-related strategic plan and the harmonisation of tourism-related regulations (e.g. entry fees, speed limits) and border access for certain activities (e.g. guided walking, trails by canoe, horse, camel or elephant, and balloon safaris). The JMP also specifies that activities should be transparently outsourced to the private sector, and also that an element of preference should be given to local entrepreneurs. The plan states that local people should not only be perceived as potential employees, but should also have an equity stake in tourism businesses and have access to business opportunities and the capacity to tender for contracts (JMB, 2002). The plan also recommends that cultural tourism be developed and marketed within local communities (JMB, 2002).

This paper not only describes the status of tourism development in the Great Limpopo TFCA, but questions whether existing and proposed tourism is compatible with sustainable development. The TFCA provides an opportunity to catalyse responsible tourism that contributes both to biodiversity conservation and the alleviation of poverty, and therefore this paper addresses the contribution of public–private partnerships and community-based tourism in achieving this.

2. NATURE-BASED TOURISM AND INVESTMENT

During 2000, nature-based tourism was estimated to generate an aggregate of $US3.6 billion from Africans and non-Africans, and contributed nine per cent of total gross domestic product (GDP) for the SADC region in 1999 (Table 1). The nature-based tourism sector was responsible for an estimated 2.8 million non-African arrivals and 6.1 million African arrivals (Scholes & Biggs, 2004).

Table 1: Annual tourism estimates in the GLTP countries

	South Africa	Zimbabwe	Mozambique
Nature tourism arrivals ('000s)^a	4634.5	1494.4	42.0
Total tourism arrivals ('000s)^b	5872.3	1966.6	400.0
Indicative % nature arrivals ^c	78.9	76.0	10.5
Nature tourism income ($US m)^a	2298.8	143.5	8.4
Tourism income ($US m)^d	4561.7	184.1	—
Indicative % income from nature tourism ^a	50.3	77.9	—
Travel and tourism industry GDP ($US m)^d	3 563.9	158.8	—
^aScholes & Biggs (2004). Data for 2000 for African, non-African and domestic tourists.
^bSouth Africa data: SA Tourism (2004); Zimbabwe data: ZTA (2001). Data for 2000 used. Mozambique data for 2001 drawn from Christie (2004).
^cThe proportion is indicative only, as the figures compared are for different years, and figures were drawn from different techniques: border entry counts, estimates and satellite accounts.
^dWTTC (2003). Data for 2003, but unavailable for Mozambique.

The characteristics of the nature-based tourism market in South Africa, Zimbabwe and Mozambique indicate a relatively strong demand in two of the participating countries, with 78.9 per cent of South African and 76 per cent of Zimbabwean tourism arrivals participating in nature tourism. Income from nature-based tourism across the three countries generates an estimated $US2.45 billion each year. Given this level of income generated from nature-based tourism, the motivation to channel revenue into conservation and poverty alleviation is understandable.

2.1 South Africa

In South Africa, the GLTP consists of the Kruger National Park (KNP). The KNP has well-developed tourism infrastructure, with 25 rest camps of various sizes providing 4056 beds, in addition to 405 caravan and camping sites (UNEP/WCMC, 2004). Although the overall standard of accommodation is not considered high by international standards, accommodation ranges from luxury tented camps to self-catering guesthouses and cottages (KPMG, 2002). The KNP has approximately 1.4 million visitors per year (Stevens, 2005) and generates an annual income of around $US40 million (JMB, 2002).

Until 2000, all tourism infrastructure inside the KNP was developed, owned and operated by South African National Parks (SANParks). At this time SANParks embarked on a commercialisation process that allowed the parastatal to grant concessionaires rights for the use of defined areas of land and infrastructure within National Parks, coupled with the opportunity to build and operate tourism facilities over 20-year concession periods (SANParks, 2001). The aim of the process has been to increase the net revenue that commercial activities contribute to SANPark's core function of nature conservation. The programme led to the transfer of management of tourism operations to commercial operators, who were considered to be more qualified and equipped to manage such facilities than SANParks (van Jaarsveld, 2004). Major objectives included the promotion of economic empowerment of the formerly disadvantaged, the promotion and provision of business opportunities to emerging entrepreneurs (in particular local communities adjacent to national parks) and the application of SANParks' environmental regulations and global parameters to all concessions (SANParks, 2000a).

Seven accommodation concession contracts in the KNP were agreed in December 2000 that guaranteed SANParks a minimum income of R202 million ² over a 20-year period, with all assets reverting to SANParks. Three of the concessionaires were black-controlled consortia, and all of the others had significant percentages of shareholding by historically disadvantaged individuals (HDIs). ³ The average percentage of HDI shareholding within the seven concessionaires (either immediately or contractually bound to be in place within three years) was 53 per cent (SANParks, 2001). The concessions added approximately 250 beds to the KNP's existing accommodation facilities (KMPG, 2002).

However, the concessionaires have not performed as well as they had anticipated over the first three years. In 2004 SANParks undertook an investigation of the concessionaires' financial progress and found that concessionaires had effectively bid too much for their sites. For the first three years occupancies were 22 per cent lower than the bid forecast, and the fixed concession fees amounted to R29 million despite concessionaires incurring losses of R34.9 million. The total actual capital investment for the KNP concessions was R218.8 million; 65 per cent more than the anticipated R132.3 million. The companies had invested between 22 and 118 per cent more than they had predicted in their bids (van Jaarsveld, 2004). The upmarket game lodge industry in South Africa is highly volatile and risky, characterised by sensitivity to market fluctuations, a highly competitive market, high capital requirements, high fixed costs and a requirement for predictable and high quality game viewing. Apparently the lack of information about the market for concessions inside national parks at the time of the tender process, coupled with concessionaires bidding in a competitive environment, led them to overestimate their potential performance. In addition, van Jaarsveld (2004) reported that tourism demand in South Africa has become increasingly price driven, and growth is being experienced in the three-star market owing to the high end four- and five-star markets being less affordable. SANParks is currently undergoing a consultation process with concessionaires to address and ameliorate some of their constraints. The problems faced by the concessionaires in the KNP need to be understood so that future concession programmes in Mozambique and Zimbabwe can build on the South African experience.

With regard to socio-economic impacts, no detailed audits of the accommodation enterprises have yet been successfully completed, although a pilot study is currently being undertaken by Empowerdex (A van Jaarsveld, Business Development, South African National Parks, personal communication, 2004). However, it has been reported that commercialisation of the restaurants and retail outlets will lead to a guaranteed annual spend of R6.9 million in adjacent HDI communities. Within the restaurants there is 40 per cent ownership by black individuals, ten per cent ownership by a community trust and ten per cent equity among employees (Varghese, 2005).

Other tourism investment in the KNP to support the TFCA has included money from the national Department of Environmental Affairs and Tourism (DEAT). The DEAT's Poverty Relief Fund funded two major projects within the KNP between 2002 and 2004: upgrading staff accommodation (R10 million) and partly upgrading the western boundary's foot-and-mouth fencing (R2.5 million). Poverty Relief Fund money has certain socio-economic conditions attached to it, such as the following (DEAT, undated):

• Between two and four per cent of the project budget must be dedicated to training and capacity-building, and all training must be accredited to enable the trainee to either manage the funded project or seek employment elsewhere with the skills or knowledge gained.

• Projects must promote the creation of small, medium and micro enterprises (SMMEs) and favour previously disadvantaged individuals (PDIs).

• Ninety per cent of temporary jobs created must go to local people, and 60 per cent of the temporary and permanent jobs should be reserved for women, 20 per cent for youth and two per cent for disabled people.

The DEAT also used R40 million granted by the National Treasury to invest in tourism infrastructure within the GLTP at the end of 2002. This investment was used to improve the quality of roads in the KNP, construct a new border post, develop research infrastructure, interpretation facilities and picnic sites, build bridges, and to remove a section of fence between the KNP and Limpopo National Park (LNP) (Swanepoel, 2004; Theron, 2004).

2.2 Zimbabwe

In Zimbabwe the GLTP consists of Gonarezhou National Park (GNP), and is proposed to include the Sengwe communal area. Gonarezhou has one rest camp, Mwenezi, which has rondavels with 16 beds. The transport infrastructure, although available in and around GNP, is in a state of disrepair owing to poor funding and maintenance. Roads are untarred in the park, poorly maintained and mainly accessible by 4 × 4 (JMB, 2002; KPMG, 2002). Between 1996 and 1998 there were approximately 6000 visitors to Gonarezhou annually, of whom about 20 per cent were foreigners. By 2004 this number had declined steeply (along with other areas in the country) to just over 2600 visitors per annum, of whom 7.7 per cent were international (Stevens, 2005) and more than half were day visitors (JMB, 2002).

The 1998–2004 GNP Management Plan proposed the creation of 13 ‘undeveloped’ campsites, 15 developed sites and 20 day-visitor picnic sites, in addition to new hides and viewing points (KPMG, 2002) which have gone out to tender (F Mutepfa, Coordinator, Great Limpopo Transfrontier Park, personal communication, 2005). Reports from the ZTA indicate that during 2004 the Ministry of Finance allocated $Z17.9 billion (R26 million) to upgrading international airports, including refurbishing the Buffalo Range Airport that is used to reach Gonarezhou. The state also allocated $Z2.2 billion (R3.2 million) to upgrade existing tourism facilities in GNP. It was indicated that they had begun with three main areas – road development, electrification and communications – to provide for accessibility for interested investors to develop campsites. However, the ZTA indicates that a lack of funds has put de-mining programmes along the borders on hold (Chikanga, 2004). There has been no evidence of tourism investment in Gonarezhou since 2000.

2.3 Mozambique

The GLTP in Mozambique is proposed to include the LNP (JMB, 2002), which was a hunting zone called Coutada 16 until November 2001 when it became a national park (Huggins et al., 2003). Apart from tented camps constructed by Gaza Safaris – the previous hunting concessionaire – there is no tourism infrastructure inside the LNP (G Vincente, Director, Limpopo National Park, personal communication, 2004) and the park is not open to tourists (Stevens, 2005).

Tourism in the LNP is predominantly at the planning stage. A tourism master plan, zoning plans and an infrastructure development plan have been drawn up. The tourism framework identifies five forms of tourism that can be accommodated in the LNP: recreational, adventure, consumptive, cultural-historical and ecotourism (Limpopo Tourism Consortium, 2004). KfW has helped to finance the construction of a network of safe roads, basic infrastructure for the establishment of private tourist camps and lodges, fences to protect the villages from roaming wildlife, and the construction of health stations and village schools in the fringe areas of the park (KfW, 2004).

The Peace Parks Foundation (PPF), an NGO working on the TFCA, summarises progress in Mozambique as follows (Codex dds, undated).

• An anti-poaching drive has been launched, leading to the confiscation of 40 firearms. The anti-poaching unit in the LNP works closely with its counterpart in the KNP.

• A de-mining contractor has completed 70 per cent of the de-mining.

• A total of 1987 animals have been translocated into a wildlife sanctuary as a first phase of restocking the LNP.

• Seventy-three field rangers have been trained and deployed in the park, and workshops between communities and field rangers have been held to enhance collaboration and co-operation.

In addition to facilities for administration, game rangers and environmental education, the LNP Tourism Development Plan proposes four entrance gates: Massingir, Giriyondo Border Post, Mapai and Pafuri Border Post. A road from Massingir to Giriyondo is proposed to establish access from the KNP into the LNP, in addition to a series of loop roads in the moderate-use zones of the LNP for tourists. Tracks for 4 × 4s are proposed for safaris and park management around Limpopo, Shingwedzi, Makandezulo, Mapai and the Sandveld area. View points are proposed along these routes at suitable locations. This management infrastructure will be developed in phases over a period of time, depending on the availability of funding. It is predicted that in addition to the 2184 potential overnight visitors, 160 day-visitors will also be able to enter the LNP. At capacity, and with a two-night average stay, it is anticipated that the park may have a capacity of 486 180 visitors per annum (Limpopo Tourism Consortium, 2004).

3. LOCAL COMMUNITY INVOLVEMENT IN THE TFCA DEVELOPMENT

The involvement of local communities in the planning and development of tourism in the TFCA has been sporadic. So far, tangible examples of community-based tourism enterprises and joint ventures with the private sector are few and far between, but there are some notable developments that are described below. Issues relating to the participation of inhabitants of the TFCA are also reviewed.

3.1 South Africa

In 1969 the Makuleke people were forcibly removed by the state from a 24 000 ha area that they inhabited in the north of the KNP. They were compensated for their relocation in 1998 with the restitution of their land and the creation of a contractual park (Elliffe, 1999) in the Pafuri Triangle (Figure 2). A 25-year agreement was forged between the Makuleke and SANParks to return the ownership and title of the land to the people, although the title specifies that the land may be used only for wildlife conservation (Steenkamp, 1998; Steenkamp & Grossman, 2001). The contract that governs the incorporation of the Makuleke land in the KNP enables them to make sustainable use of specified natural resources (E Koch, Director, Mafisa, personal communication, 2002). They have the option to construct six small camps with a cumulative occupancy of 224 beds (Grossman & van Riet, 1999).

Figure 2: Makuleke's Pafuri Triangle and the Sengwe Corridor. Source gl_makuleke_al_1q. Not to scale.

Trophy hunting was able to deliver quick revenues to local people during the period when tourist lodges and camps were being built (THETA, undated). Between 2000 and 2003 the Makuleke used trophy hunting to generate both income and venison for households. Four hunts earned the Makuleke the equivalent of roughly $US230,000 during this period (Ford Foundation, 2002; Collins, 2003; THETA, undated). The Communal Property Association (CPA) wants to phase out hunting as the lodges and camps begin to bring tourists and revenue to their area (THETA, undated).

The Makulekes' CPA has a contract with a private sector partner, Matswani Safaris, who developed a luxury 24-bed lodge called The Outpost at the confluence of the Luvhuvhu and Mutale Rivers (THETA, undated). The lodge development began in 2002 and cost between R10 and 15 million (c. $US1 million to $US1.5 million) to construct (Ford Foundation, 2002). The lodge is an upmarket establishment, designed to have low environmental impacts. An initial fee was paid to the Makulekes on signing of a contract, and Outpost Lodge has provided a bond to guarantee its performance for the duration of the agreement (THETA, undated). Rather than playing a fixed rental fee, The Outpost pays the CPA eight per cent of the gross earnings generated by the business, a traversing fee for game drives, and two per cent of gross earnings are placed in a social development fund earmarked for educating the Makuleke youth (Ford Foundation, 2002). They also pay a monthly traversing fee for every vehicle based in the Makuleke region (THETA, undated). Projections have indicated that when running at 60 per cent occupancy the lodge will pay a total annual concession fee of $US75 000 to the Makuleke community, and around $US150 000 to the 30 local employees in wages (Koch, 2001; THETA, undated). This would generate an estimated $US400 per family through the initiative, which according to Koch (2001) is significant in relation to the average annual wage of around $US750.

In July 2003, the CPA announced a R45 million concessionary agreement with Wilderness Safaris to build three luxury low-impact tented lodges (Turner, 2004). Wilderness Safaris invested R20–30 million in the camp (C Bell, Chief Operating Officer, Wilderness Safaris, personal communication, 2004) and have established 20 tented rooms with en-suite bathroom facilities and views over the Luvuvhu River. For the course of the 45-year lease, Wilderness Safaris state that, as in the case of The Outpost, the Makuleke will receive a percentage of turnover (rather than a flat rental fee), training, employment (up to 200 jobs) and some scholarships will be provided (Turner, 2004). Under this agreement the Makuleke will no longer be allowed to hunt in their contractual park (L Maluleke, South African National Parks and community leader, personal communication, 2005).

Both tourism partnerships have similar variable concession fee payment systems and are build–operate–transfer arrangements. This means that their partners will build and operate for a specific number of years and then they will transfer the ownership of the lodge to the CPA. The Makuleke can then decide whether they want to run it themselves or invite the partners to operate on different terms (as they will then own the facilities) (Maluleke, 2003; Collins, 2003).

3.2 Zimbabwe

In the run-up to the 2000 general election, various coalitions of actors, gathered under the banner of ‘war veterans’, stepped up a previously low-level campaign of occupying commercial farms and some state-owned land. ZANU (PF), the ruling party, fought the election under the slogan ‘Land is the economy and the economy is land’ (Wolmer, 2003). In February 2000, a national referendum was held over a draft new Constitution for Zimbabwe and this was followed in June by a nationwide parliamentary election. These two processes dramatically changed the country's political environment. The changes were traumatic and reported as such, resulting in a poor international image for Zimbabwe, which had adverse implications for the whole economy, most particularly the tourism industry (de la Harpe, 2001). Farm occupations and a ‘fast-track’ land reform process picked up momentum after the election, underpinned by a policy emphasis on the importance of small-scale peasant agriculture at the expense of white-dominated commercial agriculture in general, and the wildlife industry in particular (Wolmer, 2003).

3.2.1 Resettlement in Gonarezhou

When Gonarezhou was declared a National Park back in 1975, the inhabitants of the park were forced to resettle outside the park's boundaries (Ferreira, 2004). In 2000, reoccupations began as the Chitsa people used the invasion of commercial farms in Zimbabwe to pursue their historical claim for the land, driven by the need for access to grazing and hunting (Chaumba et al., 2003). By November 2000 areas of the park were being occupied, cleared and burned by residents of neighbouring communities (Sharman, 2001). By May 2001 Agritex, Agricultural and Rural Extension Department, had planned ten villages in Gonarezhou and had also allocated separate arable plots and a communal grazing area. Provisions were made for 750 settlers on 520 plots covering 11 000 ha. Although there was an immediate increase in settlers and bush clearing, a combination of drought and elephant crop raiding led to most settlers returning to communal areas by June 2002 (Chaumba et al., 2003).

Since November 2002 the Governor of Masvingo Province has reportedly encouraged families of those previously evicted, and opportunistic ‘war veterans’, to occupy land within the park located to the north and south of the Runde River. Cattle are grazed in the park and cattle fencing has been pulled down. Although the incursions have only taken place in some parts of Gonarezhou, poaching and decline in tourist numbers have affected the GLTP (Ferreira, 2004). However, Wolmer et al. (2003) indicated that it is not only cattle and agriculture that are forming the mainstay of enterprise development in Gonarezhou. There appears to be a unique case of local entrepreneurs allocating themselves land in a former veterinary corridor, ⁴ with the objective of developing commercial wildlife tourism. Apparently 50-hectare self-contained plots have been allocated to 56 people, who are members of a relatively wealthy and politically well-connected elite, including councillors, war veterans' leaders, army personnel, policemen and National Parks staff. Although none of the people have physically moved back into the veterinary corridor, they propose to operate it as a mini-conservancy, where revenues from a safari concession would be disbursed to the 56 landowners (Wolmer et al., 2003). More recently, in January 2004, the government of Zimbabwe began to relocate illegal settlers out of Gonarezhou and its buffer zones, because of the proposed TFCA (Staff Reporter, 2004).

3.2.2 The Sengwe Corridor

Sengwe Communal Land (SCL) is in the southern part of Chiredzi District and located on the southern international border of Zimbabwe. Owing to its strategic location, any linkage between the KNP and Gonarezhou must lie within the SCL, and the ‘Sengwe Corridor’ could form this direct link (CESVI, 2002) (Figure 2).

A proposal has been put forward to demarcate a wildlife corridor between the KNP and Gonarezhou. The area does not currently contain any settlements or cultivation and therefore would not require any relocation of households. Through a process of consultation, the communities, the TFCA committee and the Rural District Councils (RDCs) agreed to set aside a strip of land parallel to the Limpopo River for wildlife, which would not be settled, cultivated or used for livestock. The corridor varies between 0.6 and 2.3 km wide, and has an area of approximately 20 km² (note that the boundaries of this area are not clearly indicated in Figure 2 as the area is still under discussion). If a border crossing is installed between the northern end of the KNP and the Limpopo strip, this would allow tourists from the KNP to access the Sengwe area on their way to Gonarezhou (CESVI, 2002). Economic activities proposed for the Sengwe corridor include (CESVI, 2002):

• Safari hunting.

• Wildlife and landscape-based tourism, i.e. photographic tourism, with small high quality wilderness-style lodges (e.g. 24 beds) with joint venture agreements between the private sector and communities. Three potential lodge sites have been identified by communities.

• Cultural tourism, based on the traditional lifestyles of the Hlengwe communities in Sengwe and Chipise.

• Craft and veld products. Sengwe area already has a craft association facilitated by the NGO SAFIRE. Products may include Marula and Zyziphus jams, and cosmetics made from Kigelia (Sausage Tree) pods.

• Agricultural produce. Farm produce, including vegetables, fruit and poultry, would be sold to the lodges. This would require small-scale irrigation projects and rehabilitation of transport infrastructure.

It is anticipated that the TFCA will bring into the corridor additional wildlife resources that can be exploited by people for revenue through trophy hunting, and that additional tourists attracted to the area by the wildlife will spend money in the local area (e.g. on craft) (Dzingirai, 2004). However, people in Sengwe have indicated that their experience of CAMPFIRE is that the state and rural district councils (such as Chirezi RDC) are not interested in sharing benefits, but rather aim to retain as much revenue as possible for their bureaucratic processes (Murombedzi, 1992). There are few institutional structures in the area that can claim revenue, or can engage with the state (Dzingirai, 2004). SAFIRE is promoting the development of meaningful businesses by local people, including ownership of tourism enterprises (G Kundhlande, Director, SAFIRE, personal communication, 2005). People are also concerned that their existing ability to cross the border to South Africa will be restricted by the need for passports and visas. Since Zimbabwean travel documents are difficult to obtain, other than through corrupt means, locals see this inevitable requirement as one intended to inhibit their mobility. They are also concerned about potential fees for importing goods and the likelihood of goods being impounded by officials (Dzingirai, 2004). Importing goods has been a strategy used by the poor to survive in a marginal environment, and they thus feel threatened by the TFCA.

3.3 Mozambique

3.3.1 Covane Community Lodge

A community-based tourism enterprise near the Massingir Dam has been facilitated by the Swiss NGO Helvetas. Covane Community Lodge is located about 7 km from Massingir Township, and was opened in May 2004. The lodge has two five-bed chalets, three three-bed tents (19 beds) and additional space for people to bring their own tents. Besides accommodation, the lodge offers traditional dances, traditional food, hiking trails, village visits and viewing by boat, and the opportunity to purchase local crafts (Helvetas Moçambique, undated). There are self-catering facilities, a restaurant, ablutions and a seating area overlooking the Massingir Dam. Between July and September 2004 the lodge had 95 domestic and foreign visitors (G Palalane, Programme Officer, Helvetas, personal communication, 2004).

The lodge was financed jointly by USAID ($US50 000) and Helvetas ($US20 000) in 2002 but is owned by the Canhane community. After presenting the concept of a community lodge to the Provincial Ministries of Tourism and Agriculture, the district-level administration and the Departments of Agriculture and Tourism, the concept was raised with the Canhane and Kubo communities. There was a low level of awareness of tourism at the time, but Helvetas talked to the communities about CBT (community-based tourism), potential locations for a lodge, the land law and land delimitation. ⁵ The community identified a suitable location and, at the suggestion of Helvetas, organised a steering committee of ten local volunteers. Community members chose members of the steering committee during a community meeting (G Palalane, personal communication, 2004). Helvetas facilitated the delimitation of an area of 7024 hectares so that the community ‘owned’ the land the lodge was to be built on. The community chose 20 local people to work on the construction between February 2003 and November 2003, without assistance from the private sector. Helvetas organised a constitution for the association (G Palalane, personal communication, 2004), which addresses the financial responsibilities of the Canhane community and Helvetas (Helvetas Moçambique, 2004).

Nine members of the community are now employed at Covane Lodge (G Palalane, personal communication, 2004) with indirect employment of a further 40 people (e.g. primary school students produce and sell handcrafts for their own benefit) (Palalane, 2005). Helvetas suggested the type of qualifications that employees would need, and the community searched for these qualifications within the community. Helvetas sent these people on tourism and hospitality training courses, and the manager is about to attend a hotel management course. The NGO also organised visits by community members to community-based tourism enterprises in Swaziland and at Xai-Xai (G Palalane, personal communication, 2004).

The lodge accrued eight million Metacais (c. R260,000) from the accommodation and services sold to tourists between May and October 2004, and plans to hold a community meeting to decide how to spend the money within the 3033-strong population of Canhane. The agreement currently states that 50 per cent of the money should be spent on community infrastructure and 50 per cent on investment for the camp (G Palalane, personal communication, 2004). To date, two wells have been rehabilitated and a classroom has been constructed (Palalane, 2005). The camp investment proportion of revenue will eventually be spent on salaries when Helvetas completes its involvement (G Palalane, personal communication, 2004).

3.3.2 Limpopo National Park

When the Government of Mozambique (GoM) proclaimed the Coutada 16 as Limpopo National Park in November 2001, it did so despite the fact that people were living within it. The World Bank reported that

… despite their efforts to resolve the fate of the communities living in Coutada 16, and the assurances given by GoM that the issue would be adequately resolved prior to any action, in 2001, the LNP was gazetted as a national park and some animals were allowed to enter the park area. This created a lot of media attention and controversy around the project.(World Bank, 2004: 11).

In all, it is estimated that there are 6500 people living along the Shingwedzi River and close to Massingir Dam, with an additional 20 000 living along the Limpopo River (Huggins et al., 2003). The people living in the Shingwedzi River basin inhabit an area of approximately 3700 km². They are clustered in nine villages, the biggest amounting to about 2000 inhabitants, and the smallest to less than 150. Each homestead cultivates an average of about three hectares of land, and there are more than 5000 head of cattle (GFA Terra Systems, 2002). Communities living along the Limpopo River and along the Olifants River in the south-east of the park number approximately 20 000 people, settled in the support zone throughout about 40 villages (GFA Terra Systems, 2002).

A resettlement process is underway in the LNP to determine whether the inhabitants will remain there, or whether they will be relocated to areas outside the park and compensated. The LNP has aligned its resettlement policy with the World Bank's safeguard policy on involuntary resettlement (Operational Policy 4.12). The first principle of this policy is that ‘Resettlement must be avoided or minimised’ (Huggins et al., 2003). A Resettlement Working Group (RWG) has been appointed which is composed of district administrators, provincial government officials, local leaders and community representatives. The RWG will draw up the Resettlement Action Plans (Limpopo Tourism Consortium, 2004). The resettlement policy considers the ‘no resettlement’ options with regard to creating fenced enclaves within the park, or remaining unfenced. Huggins et al. (2003) suggested that in the enclave scenario land available for farming and grazing would be limited and villagers would not be able to leave the enclave on foot, but would be dependent on arranged transport. If homesteads remained unfenced, there will be competition between people, livestock and wild animals over resources. Living with wildlife would also increase the danger of spreading diseases between domestic and wild animals. Although homesteads would be free to move, they would live in a more dangerous environment. However, the LNP tourism development plan also states that tourism may have positive impacts on local communities through job creation and other opportunities (Limpopo Tourism Consortium, 2004).

4. CONCLUSIONS

Despite being formally agreed on in 2001, the GLTP is still in the early stages of development. South Africa has a well-developed tourism industry and infrastructure within the park; Mozambique has minimal infrastructure and is at the planning stages for tourism; and Zimbabwe's investment has been heavily constrained by a lack of funding ensuing from political unrest and land instability. Although this study has revealed the level of planned, developing and existing tourism in the participating countries, the implications for sustainable tourism development cannot yet be evaluated. Simply put, there has been too little tangible tourism development in either Zimbabwe or Mozambique since the TFCA was agreed on to be able to formulate solid conclusions. As the TFCA develops, as tourists return to Zimbabwe, as tender processes for concessionaires evolve in Mozambique, as socio-economic problems are addressed in addition to conservation management, and as linkages with other destinations in regional circuits are realised, the implications of the area for sustainable development will become clearer.

The early stages of the TFCA have indicated that the South African state, in particular, is promoting responsible tourism development. The SANParks concessioning process has been designed to promote biodiversity conservation and the economic empowerment of HDIs, within the parameters of commercial viability. Investments from the government's Poverty Relief Fund have also promoted the creation of small businesses, local employment, gender equality and training for HDIs. Aside from guidance within the Joint Management Plan, there is currently no tangible evidence of similar investment conditions being facilitated in either the Zimbabwean or Mozambican parts of the Great Limpopo TFCA.

Land tenure clearly has direct implications for the abilities of all stakeholders (including communities, the private sector, NGOs and the state) to develop commercially viable tourism. Where there are restrictive rules and bureaucracy, it is difficult for communities to take advantage of tourism opportunities, but there are examples of proactive partnerships with NGOs and the private sector that may facilitate this (e.g. Helvetas facilitated the delimitation of land for the Canhane community prior to the development of Covane Community Lodge; and Wilderness Safaris and The Outpost have engaged with the Makuleke to develop tourism in the KNP). However, land tenure insecurity directly affects people's livelihoods and the achievement of conservation objectives, as is particularly illustrated by the situation of people living in Gonarezhou in Zimbabwe and the LNP in Mozambique. There is uncertainty for these people as to whether they will realistically be able to remain within the park or be relocated to other areas, which has implications for agricultural practices and tourism alike. Protected areas are motivated to establish positive relationships with neighbouring communities, and tourism can be used as a mechanism for local people to derive direct benefits from conservation areas. Building positive economic relationships can act as a driver to reduce poaching pressure.

Public–private partnerships have high transaction costs, and power relations between various partners influences their benefits and responsibilities associated with tourism development. NGO and private sector partners working with members of rural communities need to facilitate tourism development processes so that ventures will succeed in the short and long term, and which include exit strategies for leaving the venture when sufficient capacity has been built in the community.

What is obvious is that through the process of planning and development stakeholders are learning more about the complex issues involved in catalysing a sustainable nature-based tourism industry. They are learning more about the market demand for nature-based tourism (e.g. SANParks concessioning process); more about how to address local community issues – especially with regard to people living within the TFCA (e.g. the Sengwe Communal Area, the LNP); more about developing viable community-based tourism ventures (e.g. Makuleke, Covane Community Lodge); more about the environmental impacts of nature-based tourism (e.g. SANParks concessioning process); and more about the fragility of the industry in the face of political instability and uncertainty of land tenure (e.g. Zimbabwe). An adaptive approach to the TFCA development is compatible with sustainable tourism development, and Hunter (1997) argued that it must address widely divergent situations, different goals and different mechanisms of utilisation. The degree to which tourism in the GLT will develop in a responsible manner remains to be seen. Ideally it will not degrade the environment to an extent which prevents the successful development of other activities, and also will promote poverty alleviation in the area.

All three countries are facing similar demands, some of which conflict with the development of the Great Limpopo TFCA. These include the demand from communities for productive agricultural land; the demand from conservation organisations wishing to expand the land under conservation; and the requirement for greater socio-economic returns from land, particularly in relation to investment, infrastructure and employment. In the future, transactions between the TFCA, the private sector and rural communities will be more conducive to biodiversity conservation and poverty alleviation if they promote a responsible approach to investment.

Notes

¹Research Fellow, Transboundary Protected Areas Research Initiative, University of the Witwatersrand, South Africa. This research was supported by the BMW Chair for Sustainability, University of the Witwatersrand and the Transboundary Protected Areas Research Initiative (TPARI), an IUCN-SA programme funded through the Centre for the Integrated Study of the Human Dimensions of Global Change, Carnegie Mellon University, by way of a co-operative agreement with the National Science Foundation (SBR-9521914). Any opinions, findings, conclusions or recommendation expressed herein are those of the author and do not necessarily reflect the views of the National Science Foundation or Carnegie Mellon. The author would like to thank the anonymous reviewers who commented on a draft of this paper. This paper is largely drawn from Spenceley (2005); the full paper is available at http://www.anna.spenceley.co.uk/files/Tourism%20InvestmentGLTFCAMar05.pdf

²In real Net Present Value terms.

³HDIs are defined by SANParks as individuals who are citizens of the Republic of South Africa who, according to racial classification, did not have the right to vote or had restricted voting rights immediately prior to the 1994 elections (SANParks, 2000b).

⁴A strip of land left between wildlife and livestock to avoid disease transmission.

⁵Delimitation is a Mozambique legal process by which people are legally allocated the right to use land. This is not the same as a title deed, because rights to all land in Mozambique are retained by the State.