Place, belonging and population displacement: new ecological reserves in Mozambique and South Africa

Abstract

This paper investigates the effect the proclamation of conserved areas in southern Africa may have on the ability of inhabitants in these areas to retain control of ancestral territory and to access long-standing livelihood options in the future. In particular, it examines how two national parks (in South Africa and Mozambique) effect a change in ownership and land use, and the resulting impact of such a change on local socio-cultural patterns of identity, or ‘place’. This is achieved by examining the dilemma confronting social and ecological scientists in planning these parks – where interventionist policies often deny customary tenure of land and thus prevent a thorough understanding of any historical claims to land prior to actual proclamation of the area. Accordingly, this paper argues for greater cooperation between social and ecological researchers in order to prevent the politicisation of national reserves and the intense opposition that has accompanied most instances of population displacement in the subcontinent.

1. Introduction

The planned Greater Addo Elephant National Park (GAENP) in the Eastern Cape, South Africa, and the Transfrontier Conservation Area (TFCA) in Mozambique (adjoining the Kruger National Park), like many other parks in Africa, function within two contextual frameworks. The first is a macro-political one, where the creation of ecological reserves in southern Africa can be associated with a history of direct state intervention in the lives of rural inhabitants. In South Africa, the creation of illegitimate homeland states and the forced resettlement of people in post-colonial states such as Mozambique (Munslow, 1984), has resulted in the forced separation of kin over national borders and the removal of the ancestral lands of numerous ethnic and linguistic groups.

Current parks in southern Africa function within a slightly different milieu. Both the TFCA and GAENP have been partly funded by global corporations, and have been marketed to a largely foreign clientele. The impact of globalisation upon local access to land has been described by Appadurai (1996: 39) as the ‘deterritorialised world’ – where, ironically, the lack of space required for globalised transactions has resulted in a marked loss of territory for many ethnic and linguistic groups. Many claimants of lost land have often tried to reclaim their heritage, despite the fact that these claims are not always authentic, and may be based more on indeterminate feelings of belonging than on a factual historical occupation of land. In spite of this, the increased number of such instances indicate that those who have lost access to a home-based territory in the past thrive on the need for some form of memory of such a homeland, and will try to recapture such a loss, often through violent means (Sharp, 1996; Comaroff, 1998).

The impact of deterritorialisation thus holds important consequences for newly created ecological reserves, as these sites may well become localities of competition and conflict, resulting in the increased marginality of local inhabitants in global tourist ventures. This paper holds that the social and cultural consequences of land loss in new reserves in southern Africa have not been fully gauged by ecological managers and that the complex historical meanings associated with local land use and occupation must be understood by both social and ecological scientists before these reserves are formally proclaimed. The paper draws extensively upon the information provided by people in villages in a border zone of Mozambique – Pafuri (or Crooks Corner), as well as by farmers and farm workers around the planned GAENP in the Eastern Cape, South Africa. The name ‘Crooks Corner’ originated between 1900 and 1920, where groups of professional elephant hunters and bandits from the former Portuguese frontier and slave traders in search of native children used the dense bush around the confluence of the Limpopo River as a hideout (Stevenson-Hamilton, 1937: 96; Murray, 1995). As an introduction, the implementation of ecological theory is first discussed in order to acquaint the reader with the southern African context.

2. At an impasse? social and ecological science

The relationship between the natural environment and cultural change has long been the subject of social enquiry, particularly in the anthropological use of theories of cultural ecology and cultural materialism. These ‘old ecological’ theories, as Kottak (1999: 23–4) has termed them, were criticised for their widespread concern with equilibrium. Systems were analysed according to the response and adaptation of human beings to an external natural environment, a space that was somehow removed from the rigours of everyday life – rather than an embedded interaction with a specific (eco)system.

At present, the study of the relationship between the natural environment and the cultural representation and human use of such a space is notably different, as it is ecology has entered a far more power-laden context. Now, the ‘environment’ is one of the most contested areas of debate (Milton, 1993) and conserved areas of land, in particular, are subject to high levels of disagreement regarding the amount of human control allowed in natural systems (Croll & Parkin, 1992). The public nature of these debates and the reliance on global media, tourism and international financial corporations (such as the World Bank) as a means of support emphasise the controversy surrounding conservation: who, and under what circumstances, may regulate the control of natural resources? Social and ecological scientists have different opinions about the impact of ecological reserves.

Environmental anthropologists and sociologists focus on giving voice to the indigenous inhabitants of landscapes and often provide detailed examinations of local mechanisms of landscape control and natural resource utilisation (Kottak, 1999: 25). More specifically, it is the use of the theory of post-materialism which holds that conserved landscapes are only luxuries which are available to the few – those who have the highest degree of control over their environments (Guha & Martinez-Alier, 1997: xv). This agrees with the work of Grove (1995), as well as Carruthers (1995: 100) in the Kruger National Park (KNP), where thousands of residents were forcibly removed. The use of these theories has thus compared the creation of many new conservancies and the corresponding search for ‘wild’ areas of the globe to many ill-fated development projects which have attempted to increase northern consumption of goods produced in poorer countries of the south (Hobsbawm & Ranger, 1983). To anthropologists, the challenge is not only to devise a way in which indigenous models of knowledge can be incorporated into mainstream ecological science, but also to put forward a more socially sensitive means of conservation than is currently in practice.

Ecological debates in reserves in southern Africa have made extensive use of two ideas – ‘deep ecology’, and ‘wilderness’ theory (Guha & Martinez-Alier, 1997: 82–93). According to post-materialism, deep ecology is premised on a biocentric perspective, which assumes that an anthropocentric use of nature is essentially exploitative. In this view, the effects of industrialisation or western methods of farming are rejected in favour of a restoration of unspoilt wilderness, as untrammelled by the effects of human influence as possible. In a similar vein, the wilderness movement draws upon mainly American ideas of exploration into the frontier zones during the 18th century. Rob Marshall, the founder of the wilderness society in the USA, rejected organised western methods of agriculture and culture, and favoured an image of the archetypal aboriginal wanderer, living in perfect harmony with their surroundings. In this view, the role of the human race in ecosystems is that of stewardship – controlling the amount, type and quality of human influence that may be allowed into reserves. These basic principles have become the primary means of landscape management in many reserves in post-colonial Africa.

In response to these ecological ideas, anthropologists and other social scientists have traced the use of conservation as a science to the expansion of imperial power in Africa during the first half of the 20th century. Studies such as those in Guinea (Fairhead & Leach, 1996), by Carruthers (1994, 1995) and Beinart (1984, 1996) in South Africa, and by Moore & Vaughan in Zambia (1994) have explored the impact of conservation theories in Africa, theories which targeted African farmers and pastoralists as agents of desertification, overgrazing and soil erosion. These cases have pointed out that these theories of land protection have either led to the expulsion of residents from newly declared conservation areas, or placed a severe limitation on the use of natural resources by local inhabitants.

The debate between social and ecological scientists has continued to cause a great deal of controversy in conservation, and has resulted in a number of heated exchanges. Given their tendency to support local action, anthropologists have been accused of using Marxist and neo-populist dogma to create a misrepresentation of ecology and ecological theory (Spinage, 1998: 274). Ecologists have not excluded social action from reserves on a mere whim, Spinage writes, but because human influences such as pollution and poaching are clearly created by people within or directly adjoining a protected area – and ecology's main purpose is clearly to prevent or manage such anthropocentric influences.

What is true, however, as anthropologists have pointed out, is that the ecological management of reserves does not emanate from a value-neutral context. Many existing reserves are deeply embedded in political and social contexts of dispossession and poverty – the very existence of these impoverished communities creating the problems of pollution and poaching that ecologists are trying so hard to prevent. Accordingly, with the pressures of the deterritorialised world, and the many land claims that accompany the creation of reserves, some ecological scientists have sought to revise previously held theories of landscape management. Theories of disequilibrium, in particular, doubt the assumption that natural systems are to be managed as ‘pristine’ relics of the past (Fairhead & Leach, 1996: 119). Forman (1995) views elements of human landscape patterning, such as fire, grazing, agriculture and human settlement, as playing a dynamic role in promoting system diversity and flux. Practically, Fourie (1994) and Loader (1994) have commented that these theoretical changes have been echoed in the very manner in which conservationists have responded to the social challenges of their neighbouring communities. Besides the Communal Areas Management Programme for Indigenous Resources (CAMPFIRE) initiative in Zimbabwe, some ex-residents of reserves in South Africa have entered into profitable agreements with conservation authorities, especially in the northern KNP and the Northern Cape. CAMPFIRE allows local communities to share the profits of tourism and hunting in natural reserves, and provides selective access to natural resources within parks.

However, it still appears that the planning of new reserves in southern Africa has been characterised by the marginalisation of social research. As Cernea & Schmidt-Soltau (2003: 4) mention, intensive and in-depth social commentary is often sidelined in favour of token recommendations and consultancy work, while biological concerns have gained significant policy and economic backing. Thus, expert social commentary is not integrated into planning and management procedures from the very start and this has aggravated problems such as displacement and poverty among those affected by new reserves. In addition, the creation of larger national parks such as the KNP, TFCA and the GAENP, is often touted as a nation-building enterprise, so that political wrangling and negotiation often replace genuine consultation and participation by local land users. As the two case studies below illustrate, inhabitants of new protected areas have genuine claims to these landscapes as a valuable resource in times of need, and have articulated their claims through historical ideas of belonging and home-making. The proclamation of protected areas may effect a fundamental change or limitation in land use and ownership and thus will severely restrict the local residents' ability to effectively manage and construct their homes and obtain their livelihoods in the future.

3. Mozambique and the transfrontier conservation area

Despite a decision to leave human settlements inside the reserve intact and to provide residents with the choice of leaving (with compensation) or remaining, media coverage of the TFCA reported some inconsistencies in the treatment of local residents in Mozambique. According to a survey conducted by the University of the Witwatersrand, (Mail & Guardian, 2002), only 40 per cent of the 84 household heads surveyed in the upper Limpopo valley had ever heard about the Transfrontier initiative. Another 83 per cent of people in the same sample indicated that they had never been consulted about the possible consequences that the Park could hold for their homes and families.

The TFCA in Mozambique has erected a 52-km fence on the southern side of the Limpopo River, around what is termed the Limpopo Reserve. This fence, which encloses an area of 35 000 hectares, serves to limit the movement of large mammals over the Limpopo river and confine them to the ambit of the Transfrontier Zone. However, the fence has also excluded residents from an important source of water and arable land near the river, which also contains valuable natural resources, such as ilala palm-wine plantations and thatching reed. This fence will make it particularly difficult for families to continue agricultural activities and harvest natural resources, both of which are essential for the survival of these communities along the Limpopo river, especially in times of drought.

Despite the conspicuous nature of these issues, the situation in Mozambique is far more complex than is suggested by an explanation of the TFCA as an exclusionary and hegemonic symbol of global power (in the guise of conservation). In Mozambique, people do not respond congruously to an ‘outside’ influence (i.e. that of conservation) but, as the history of local settlement demonstrates, have developed imaginative strategies for dealing with instability, war and natural disaster. As this section explains, both the people living within the present TFCA and the conservationists who are involved in the project have developed very similar ideas of Mozambique as a ‘ancestral home’ and a ‘wilderness’ site, which have proved essential in influencing the way both parties have perceived their respective historical identities.

The ecological version of life in Pafuri may be illustrated by a short history of the Kruger National Park (KNP) in South Africa. The KNP is the oldest and the largest national park in Africa – it was established in 1926 and covers a total of 1 948 528 hectares. As Carruthers (1995) illustrates, the region of the 19th century Eastern Transvaal has often assumed the image of an uncivilised wildlife paradise in the minds of white South Africans. Historically, the creation of the KNP, like national parks in the United States (such as Yellowstone), provided a mechanism for promoting national feeling and unifying South Africans against the ‘evil’ influences of wildlife decimation and land degradation – imperfections which were placed squarely on the shoulders of African occupiers of land. Many white visitors to the KNP, such as myself, have subsequently enjoyed the absence of human interference in the park. This is a romantic notion that is easily exaggerated through a contrast with the overcrowded settlements in the areas surrounding the park, through which one must inevitably travel before reaching one's destination.

During the 1990s the KNP experienced a sharp increase in the number of elephants, which resulted in a decline in the quality and quantity of forage available to these large herbivores. Subject to great public debate and concern, the KNP chose to cull a number of them on a yearly basis. In 2000, however, this practice was stopped, owing to a proposed amalgamation of the KNP with other reserves in Zimbabwe and Mozambique, essentially creating a huge park that would span the borders of all three countries. After extensive political lobbying in all three countries, this Transfrontier Park, which had been in the pipeline for at least a decade prior to 2000, was eventually formally proclaimed in 2001 – the occasion being marked by the release of a number of elephants from the KNP into Mozambique. (These elephants subsequently escaped back into the Kruger National Park.)

For conservationists, the suitability of Mozambique for a TFCA has been strengthened by the fact that the residential populace in this country are fairly thinly spread over a somewhat dry savanna, and mainly active in the riparian areas of major watercourses. Furthermore, the absence of large mammals such as elephant, rhino and buffalo (which have been decimated by poaching and war) in this area has motivated the extension of the KNP into Mozambique – the creation of which presumably would recreate a pre-colonial abundance of wildlife. The Mozambican side of the TFCA includes a core region that is similar in size to the KNP in South Africa. The main conservation area is the hunting reserve of Coutada 16, excluding the Limpopo river to the north-east, the Massingir dam to the south and Gonarezhou reserve in Zimbabwe. A second area, which will be managed for conservation use as a private concession zone, includes Gonarezhou (which will possibly be linked to the main TFCA with a corridor), as well as Bahine, Zinhave national parks and the Massingir area in Mozambique.

Similarly, like white conservationists and middle class tourists in South Africa, people resident in the previous Bantustan state of Gazankulu (which was designated as a semi-independent ‘homeland’ for Tsonga speakers) have also harboured romantic notions about Mozambique. ‘Bantustan state’ refers to the apartheid vision that native reserves should be transformed into national and ethnically separate states – or ‘proto-nations’ – and that these states could be led towards sovereign independence. They are also known as homeland states (Sharp 1988: 85).

Oral and written historical sources (Harries, 1987) reveal that Gaza province in Mozambique (which is mainly composed of Tsonga speakers) has long functioned as an image of home to many displaced communities in surrounding countries. The border region of Pafuri in particular (Connor, 2003) has attracted a particularly high number of returnees (displaced by the Mozambican civil war) from South Africa and Zimbabwe, more so than other parts of the province. The reason why Pafuri has attracted more returnees can mainly be ascribed to an extremely porous border region, where the high number of commuters and clan members in Mozambique, South Africa and Zimbabwe has resulted in a marked relaxation of visa requirements for ‘crossing over’. For many commuters who have been removed from their country of origin, the Maluleke in particular, visiting Mozambique is a reminder that Pafuri is an ‘unspoilt’ area of ancestry. These romantic ideas are reinforced by the prominence of cattle and subsistence farming in Mozambique, in contrast to the many overburdened settlements in South Africa and Zimbabwe. Similarly, the abundant tracts of riparian forest hold special significance for healers (or n'angas) in South Africa.

One of these shared clan groupings, the Maluleke, stems from a common geographical area spanning the borders of Mozambique, South Africa and Zimbabwe, which was inhabited before and during the upheavals associated with the mfeqane (and particular that of Soshangane) in the late 19th century. However, owing to a dispute over succession between two brothers, this clan then split in two – one brother occupying the eastern side of the present state border in Mozambique, the other residing on the western side of the border in the present northern KNP area (around Punda Maria).

Soshangane (or Manucusse) was a Ndwandwe (far northern Zulu clan) warlord who is credited with establishing the Gaza empire in southern Mozambique. During the wars associated with the mfeqane – a Nguni word referring to the conflict created by the expansion of the Zulu empire (Sotho: dfeqane) – he created a system of loose sovereignty over tribute-paying subjects such as the Maluleke and devastated the south of Mozambique through raids and warfare (Newitt, 1995: 295)

Although resident outside the proclaimed Mozambican State in 1897, the Maluleke in South Africa remained part of the social and cultural life of their dispersed clan. These Maluleke were eventually removed from their land in the KNP in 1976, and were resettled in a village (by the name of Makuleke) outside the park, 200 km south-west of their original home. Despite these divisions, however, the Maluleke in South Africa and Zimbabwe still foster an image of the upper Limpopo river as an ancestral home of origin, and as Harries (1987: 111) writes, ‘their golden age, which is always contrasted with the bleak conditions of their present existence, provides the community with a sustaining and guiding myth’.

This myth is cemented by border commuting, which assumes many forms – women from Zimbabwe regularly harvest reeds and make palm wine from ilala plantations in Mozambique, while Mozambicans crossing over to South Africa have vested employment opportunities in South Africa. A few men in Pafuri even indicated that they had wives on both sides of the border, a practice that provided them with a safe haven during the Mozambican civil war. Unlike South Africans, Mozambican residents said they did not suffer the disasters associated with agricultural ‘betterment’ practices, which involved stock culling and forced resettlement into the previous homeland states of Venda and Gazankulu in neighbouring South Africa.

In addition, the remoteness of Pafuri as a frontier region has also served to create very different features of political power in Mozambique, compared to South Africa. Residents in Mozambique are subject to an authoritarian chefe do posto (regional administrator), whose administrative area is geographically isolated from the district capital in Xai Xai. Notably, the history of Mozambique cements this image: the Portuguese did not exert as powerful a political force as did the British, and later the apartheid government in South Africa, but still managed to influence native opinion by undermining the authority of the chiefs. This was not done as directly as was the case in South Africa, however, where land was purchased for the creation of labour reserves and customary practices used as justification for ethnic and racial separation. Instead, colonial force was exerted indirectly through a replacement of chiefs with Portuguese settlers or sympathetic indígenas. Village headmen were often used as agents for Portuguese rule, which created a distinct class of government-supported régulos (who later came to be classed as district governors). While Portuguese colonies were still considered part of the mother country, the inhabitants of Mozambique were classed as ‘uncivilised’ indigena and não-indigena. (Newitt, 1995: 387.)

However, it is the extent to which labour migration has featured in the history of most individual working lives in Mozambique that has cemented the image of Mozambique as a ‘home’ to many Tsonga speakers who came to reside in South Africa (Harries, 1994) and has played a prominent role in shaping the use of land in Gaza province. The Witwatersrand Labour Association (Wenela), which was created in 1901, was granted concessionary rights to use the region south of the Save River (notably the same region occupied by the current TFCA) as a labour reserve for South African mines. Gaza province became the primary source for over 65 per cent of all mineworkers in South Africa, and continued to dominate the lives of residents from 1901 until its closure in 1975 (Covane, 1996).

Labour migration in Mozambique has followed very different patterns from those of South Africa. Mozambican workers could not return to their homes regularly, and were subject to a restrictive labour agreement in a foreign country, which tied them down for at least three years. This meant that they could not send money home as often as their South African compatriots could, and increased the reliance of Mozambican families on agriculture, rather than cash, for food. Consequently, this has strengthened the role of women as rural labourers and providers. In many cases, long-term labour contracts popularised polygyny, which provided a larger and more reliable female labour force in the absence of a male provider (Covane, 1996: 146). The reliance on female labour and subsistence agriculture can still be seen today – women are responsible for hoeing, weeding and food production, and the largest tracts of farming land are occupied by those men with the most wives and daughters.

Collectively, the main features of life in Pafuri – where residents are dependent on dryland farming and natural foods for their survival, as well as representations of the ‘good life’ in Mozambique in the minds of many South African visitors – has reinforced the idea of Pafuri as an untouched and wild area. However, upon deconstruction of this myth, it appears that the images of conservationists and local communities in South Africa, Zimbabwe and Mozambique may be based more on memories of the ‘good life’ rather than on hard evidence. Residents in Mozambique have certainly not been exempt from social disruption, and have been badly affected by war, dislocation and state resettlement. Border villages, in Pafuri in particular, have been subject to communal farm ventures by the Frelimo government since independence in 1976. Munslow (1984: 211) refers to the creation of these villages as a form of ‘crisis management’ – a response to a variety of factors such as the civil war (where Frelimo would group villagers together for purposes of defence), natural disaster and flooding (in 1977, 1984 and 2001), and the self-motivated physical movement of residents to the fertile upper Limpopo valley. Such overwhelming evidence of movement to and from geographical locations has created settlement and residence patterns that are transitory and heterogeneous.

Even when fieldwork was done during 1999 and 2000, it was well known to outsiders that the residents of one particular settlement near the national border of Zimbabwe and South Africa preferred to move their houses to different sites, depending on their proximity to fertile patches of land and pools of water, near the river. Although the Limpopo regularly floods its banks, the deluge in 2000/2001 presented a particular challenge to residents of this village – not only had their fields disappeared under the floodwaters, but their huts and possessions as well. After this calamitous event, residents once more had to shift their residential and agricultural sites to more suitable locations.

The innate mobility of ‘settlement’ in Mozambique does not invalidate the myth of origin as perceived by displaced communities. On the contrary, memories of the ‘good life’ in Mozambique are bolstered by the continual movement of visitors to and from Pafuri. It is not because of their fixed positions in their home countries that people in South Africa have harboured ideas of Mozambique as a home away from home, but because the permeability of border zones, as well as shared kinship linkages, encourage mobility. In Pafuri itself, it is also clear that physical movement of people in Mozambique, partly because of the instability of the climate and partly because of numerous instances of forced removal, is far safer than a state of immobility. Covane (2001: 55) mentions that many people in the Limpopo valley have developed survival strategies in order to cope with climatic instability, such as marriages between families from different ecological zones. Furthermore, commuting between countries is intrinsically beneficial because goods are cheaper and more accessible in South Africa than in Pafuri.

More importantly, Pafuri offers diverse opportunities for many impoverished communities to retain control of their lives, and make choices that stand apart from the controlling influence of politicians and state officials. Residents in Pafuri are very much aware of the limitations that political processes, such as the Transfrontier Park, may impose upon their lives. However, this is vindicated by the number of economic possibilities created through free movement across borders and a continued investment in dryland agriculture as the primary means of subsistence. These opportunities present residents with an alternative: that they are agents who do have a certain amount of economic and political choice in their lives and are not irrevocably caught up in issues over which they have no control.

The creation of the Transfrontier Park, however, has put the control of land and the right of ownership beyond the reach of most ordinary citizens of Mozambique. Formal ownership of land is limited to the modern Mozambican state, and the land can be leased out to potential owners and contractors (O'Laughlin, 1995). In this respect, Myers (1994: 607) and Myers & West (1996: 29) note that land concessions in Mozambique have been rather haphazardly distributed in the last five years, which has created a new category of landless and marginalised individuals. In these circumstances, where land is treated as state property and the customary use of land denied, it appears that residents in Pafuri will once again have to negotiate the physical and imaginary boundaries that have been created around them.

4. The eastern cape and the greater addo elephant national park

As with the TFCA, the creation of the GAENP has evolved from a similar concern with the rapidly shrinking habitat of the Addo elephants, who are the remnants of far larger herds of animals that once inhabited the bushy thickets of the Eastern Cape during the 19th century. Moreover, the process of land acquisition for the GAENP, like the TFCA in Mozambique, has been a long-term project of conservation since the early 1990s and both projects have been heralded as essential for nation building in South Africa. Similarly, Addo has also experienced a rapid increase in the numbers of elephants, and owing to the small size of the reserve, South African National Parks (SANParks) have chosen to expand conserved land into neighbouring farming areas.

SANParks have established a formal initiative to create the GAENP within the next five years, involving a number of social, economic and ecological experts, as well as the World Bank. Kerley & Boshoff (1999: 46) describe the proposed park as being one of the most diverse in the world – including six of the seven biomes that occur in South Africa – and as having the potential for increasing socio-economic development in the economically depressed Eastern Cape. The proposed GAENP will be roughly 398 000 hectares in size, consisting of a 341 000 hectare terrestrial zone and a 57 000 hectare marine reserve – forming the third largest park in South Africa (Kerley & Boshoff, 1997).

Unlike the TFCA, however, the process of land acquisition in the GAENP has been much more complicated, largely because land ownership (as well as the right to occupy land) in South Africa has been subject to an exhaustive land reform process since 1994. In terms of the Restitution of Land Rights Act of 1994, claimants may register a land claim on a property that they were dispossessed of after 1913. The Labour Tenants Act of 1996 holds that a person who was a labour tenant (not a labourer) has the right to occupy that part of the property on which he was residing. In terms of this act, labourers are paid employees, and are therefore not legally entitled to land restitution or reoccupation of ancestral territories. The Extension of Security of Tenure Act 62 of 1997 was created to protect the rights of farm labourers with regard to eviction and working conditions, but does not directly comment on issues related to the rights of labourers as occupants or claimants. In the Sundays River Valley there may be an overlap between the application of these two acts, as some labourers can be defined as being tenants in the past, before their transition to paid employment.

In the GAENP, these land reform measures created the expectation (with the academic consultants involved in the consultancy process) that SANParks, as an organ of the state, would adhere to policies of land reclamation. However, this appears not to be the case. The GAENP had been conceived by conservationists in South Africa before laws of land reclamation were instituted after 1994 and has subsequently been purchasing properties on a ‘willing seller/willing buyer’ basis, which requires that a farm be cleared of all previous inhabitants upon transfer to a new owner.

A number of social and economic consultants have expressed their concern about the impact of these evictions on local farm workers, who have been left without employment, homes, stock and a variety of privileges accumulated while in the employ of a farmer (Connor, 2001; Huggins et al., 2003). Unlike people in Mozambique, farm workers in the Eastern Cape do not have access to cultivated land and the economic privileges associated with border movement. On the contrary, they have been subject to a vested relationship of domination by white landowners since the mid-18th century, a system which has been historically reproduced through the forced necessity for employment (on the part of workers) and the need for cheap and landless labour (on the part of farmers). As the history of the Eastern Cape illustrates, the lives of farmers and workers have been irrevocably tied together as the users and occupiers of land (Manona, 1988).

For farmers, the use of local labour and the expansion of successful small stock ventures in the GAENP area have been closely associated with the rise of Afrikaner nationalist identity in South Africa. During the 18th century, after the imposition of Dutch rule in the Cape, large numbers of trekboers (literally ‘mobile farmers’) found new independence in the frontier regions of the Eastern Cape. Among the local inhabitants, the Khoi Khoi were represented by various groups of mobile small stock herders in the area, none with a strong enough ruler to unify them, or to present any great threat to the security of the trekboers. The relative weakness of political rule among the Khoi Khoi, and a lack of stable territory, meant that these mobile herders were the first to be dispossessed of land. Frontiersmen used varied ways to force the Khoi into service – such as by apprenticing children or seizing captives (Newton-King & Malherbe, 1981: 7) – but the most common means of securing labour was through simple agreement between Khoi and farmers. In this manner, the Khoi could retain their stock, stay with their families on particular farms, and foster payments in kind (such as labour indenture).

Groups of Xhosa-speakers who were resident in the Eastern Cape frontier during the mid-18th century were in a much more powerful position than their Khoi neighbours. Despite the fact that most Xhosa communities had fled earlier wars of succession in the east, small nodes of relatively stable communities were created – based largely on a system of barter with farmers, which enabled many Xhosa to enter into sharecropping arrangements with trekboers (Giliomee, 1989). Owing to their stability and freedom of movement, residence in these settlements (especially in the upper Sundays River Valley around the present town of Kirkwood) became attractive for both Khoi and Xhosa.

However, it is the history of the 19th century that impinges more directly upon the present lives of farmers and workers in the GAENP area. During the early 1800s, farmers came under increased pressure from an influx of European settlers, and from a series of laws imposed by the British, which restricted the availability of land. With the expansion of larger farms, smaller farmers started sinking into debt, while others lost their land and reverted to labour tenancy on more economically secure territories. During this period, the activities of what may euphemistically be termed ‘frontiersmen’ became entrenched in the oral history of the upper Sundays River Valley area.

Tales such as that of the notorious Hans Weyer (all individual person names referred to in this document are pseudonyms), a German immigrant who accumulated property behind what many refer to as ‘Weyer's berg’ (mountain), can be used to illustrate the conflicting positions of many farmers during the late 19th century. Left alone with his mother and siblings after their father abandoned them in Port Elizabeth, young Hans heard of some money to be made through trading in the Sundays River Valley area. He then met a Jewish trader, who taught the young man all he needed to know about trading and gave him his ox wagon. Hans then started a trading post in Darlington (which was unfortunately flooded through the creation of Lake Mentz in 1929), from where he gradually started accumulating properties. Local farmers, however, still harbour resentment against the family of Weyer, whose ownership of property was established through a process of extracting debt from local farmers. The ‘character’ of Hans, according to some farmers, has been carried through to the present generations of Weyers in the area, who are said to be just as cunning (and wealthy) as their ancestor.

The many versions of this particular narrative, and many others like it, illustrate the close relationship that many farmers have with their land, their right to occupation (according to them) having been established through generations of struggle and economic poverty. The value that farmers place upon their property also explains why many feel embittered by the sale of property to the future GAENP. In this sense, the relationships many farmers in the GAENP have with their property are similar to those that many residents in Mozambique have with theirs. Albeit in different ways, both parties have occupied land for generations, and regard their occupation of land as their ancestral right. Similarly, many myths of occupation, such as that of Hans Weyer, and that of the Maluleke, have been embellished and expanded in order to justify their varied forms of ownership.

However, for farm labourers in the future GAENP area, the situation is very different, since the forebears of current farm ‘workers’ have been dispossessed of land at numerous stages during their past. For those who can recall a childhood on a site that had been occupied for at least three generations (for more than 100 years), the event of complete displacement from these sites is more recent. For others who have been involved in intermittent farm labour since the Christian mission stations of Bethelsdorp and Enon were established in 1812, the area of Lake Mentz (the dam which consumed Weyer's trading store) has become particularly significant owing to the development of long-term employment prospects in the region.

These missions provided safe havens for groups of people who were dispossessed of land at an early stage of colonial occupation, and for those who became disillusioned with working conditions on farms (Newton-King & Malherbe, 1981: 25). In addition to missions, people also came to reside in many settlements around towns such as those between Uitenhage and Somerset East, which were home to a number of Xhosa, Khoi and a few white bywoners (poor tenants) who had found employment in the area. However, the increased numbers of people who came to settle in mission stations, and the existence of a few tracts of private and municipal land occupied by subsistence Xhosa landholders, clashed with the needs of the newly established Sundays River Company. This company passed a resolution in 1906 which stated that ‘no native, coloured person, coolie or chinaman [sic] should be allowed to hold land as owner, tenant, or sub-tenant with out the permission of the Company’ (Meiring, 1959: 45). The building of Lake Mentz and the Sundays River Irrigation scheme in 1929 provided life to the district – not only were farmers assured of the provision of water and irrigation for the burgeoning citrus industry, but people were assured of farm work for the next few decades. Ironically, however, the growth of white farms meant that those individuals who had occupied freehold land around the Sundays River had to be dispossessed to make way for the expansion of irrigated land.

Despite these measures, some land users after 1929 held onto their land by entering into relationships with surrounding farmers as labour tenants or sharecroppers (or a mixture of the two). However, government records of the 1950s to 1980s reflect the rising uncertainty of the Native Commissioner and his employees concerning the situation of five particular ‘black spots’ around the town of Kirkwood. According to the 1945 Native Consolidation Act, people residing in these dispersed settlements had no status or right to reside in the Kirkwood area at all, since most were only itinerant workers, not permanent employees on farms.

In 1957, 3,500 people were recorded as residing in five unwanted ‘black spots’ around Kirkwood, along the Sundays River. Two main settlements were subject to particular attention from the Native Commissioner in Cape Town, who did not want the ‘black’ residents of Beersheba interfering with the lives of ‘coloured’ communities in Enon (a German mission station). The co-existence of people who found themselves in conditions of economic deprivation created a slow dissipation of ethnic and linguistic boundaries. Although the boundaries of race and ethnicity were mostly indeterminate, people such as these were subsequently classified by the apartheid state as ‘coloured’.

The remaining four ‘black spots’ around Kirkwood presented a similar problem to local government structures, as natives were occupying land without the approval of the state. Without the overt influence of the state and the control of the local municipality, people in these settlements were subject to a greater degree of autonomy. However, this also meant that they were largely unwanted residents in the area, classified as ‘squatters’ and living on land that could be used for agricultural purposes. The state thus chose to destroy these settlements.

Among the narratives collected thus far from workers in the Lake Mentz/Kirkwood area, some individuals can recall their co-residence on properties outside Kirkwood (some of which have been purchased by SANParks), from which they were displaced. In late 2002, one particular farm worker, Alec Draghoender, related to me that:

every family had their own little piece of land – we did not live on top of one another. Then a jeep came and said that we should get rid of our animals, so the cattle and sheep were sold from five cents to about two Rands per animal. We took a few years to leave, because we didn't know where to go, because all our stock had been sold and our lands taken over by a white farmer.

Grant (1993: 32) says that by April 1960 it was estimated that over 2,000 people had been moved from these sites. (Most population displacements in South Africa occurred as a result of the Natives Land Act of 1913 and the Group Areas Act of 1950, 1957 and 1965, which resulted in the forced removal of many urban residents either to locations on the outskirts of cities, or to ethnically suitable Bantustan states in South Africa.) Residents were not given the choice of leaving or staying, but were systematically forced out of the area over a period of three years, some leaving once or twice, and then returning to build shacks on their old housing sites. Some pursued farm employment elsewhere, but most were simply dumped in the locations surrounding Kirkwood. The following excerpt from a letter addressed to the Secretary for Native Affairs in Pretoria, from the Secretary of Miskraal Location in 1958, reflects the conditions under which people were forced to move:

Although he say that we [should] go to Bonterug [an informal settlement in Kirkwood] he divided us others to the farms – he want only old mothers to the location. We have no wagons to load our articles … that place have no water and firewood and the building woods, there is no road even the doctors hardly go there, that place it has no veld for our goats and cows, it four times removing us from this place and yet its our expense to build these houses and we pay for the water …

However, for those who did not move to Kirkwood, but to farms near Lake Mentz, these working (i.e. farm) sites have been put on a par with their previous occupation of ancestral land. After forced removal, permanent employment on farms, though scarce, offered a home similar to those occupied on past sharecropping and labour tenancy sites – with a house, a menial salary and rations. Despite suffering from a distinct lack of free movement from the confines of a controlling farmer, as well as being subject to forced labour indenture, an elderly farm worker (interviewed near Lake Mentz) still considered the farm where he had worked for more than 40 years as his ‘true home’. Here, he was allowed to keep cattle and gained valuable farming expertise from the landowner. In his words: ‘Ek is geleer – ek is gaar – daardie slimgeid hulle het my gevoer’ (‘I have been taught – I am cooked – they (the farmers) fed me on their cleverness’).

Although many people did find a second home on farms near Lake Mentz, it would nevertheless not be appropriate to view the homes of people who are currently working and living in this area as stable and permanent sites of habitation. Farmworkers have multiple ‘homes’ and are negotiators and participants in many different worlds: some are displaced regularly from farm to farm, and do not settle in one farm long enough for their children to receive a stable education, or to accumulate possessions or stock. Many of these unwanted labourers are simply dropped off on the nearest main road by employers, from where they can search for their next contract. Others who have access to stable employment often have homes in town (built for their retirement) as well as on a farm. In addition, most workers can recall their childhood and removal from past tenancy sites, which were occupied by traditional ‘red’ Xhosa around the town of Kirkwood, or in mixed settlements further north towards Graaff-Reinet. Lastly, there are those individuals – both farmers and workers – who have been removed from properties purchased by SANParks during the past six years.

However, as Hastrup and Olwig (1997: 33) mention, despite the fact that the lives of farm workers are diverse and multi-sited, it would be erroneous to view their varied ‘homes’ as being characterised by complete disjunction. By focusing on the relationships that people have with others – on nearby farms or in towns – and with other places (such as past sites of removal) it is possible to detect places of more permanence. Here people have had more opportunities for economic independence and less interference from an employer – whether on a farm, in a nearby town, or on an ancestral site. All of the different categories of ‘home’ and place among people in the Sundays River Valley thus create an impression of an intricate network of overlapping sites, connected by travel, memories and social relationships.

Colson (1999: 4) has argued that the phenomenon of rootlessness is extraordinarily diverse. In the GAENP, there are a great variety of people ‘who find themselves in qualitatively different situations and predicaments’. In the GAENP these predicaments arise not only through disruption from one site to another, but from many geographical sites and throughout different eras of the past. The connection of workers to various sites in and around the future GAENP area thus does not only represent a connection to a single ‘place’, or property, but a connection to a district, with a number of intersecting relationships with people in a variety of peri-urban and rural sites. Unfortunately, however, the continued presence of workers in the district of Lake Mentz, in the upper Sundays River area, is under threat and declining rapidly owing to the loss of small stock farms to conservation-related enterprises – of which the GAENP is the major player.

Owing to the economic and social decline of this area of the future GAENP, the displacement of farmers and farm workers from their properties in the future GAENP area represents something different than the more regular moves from farm to farm, and from towns to farms. Displaced workers are not only left in economic decline, but stripped of their homes, ancestral gravesites and familial assets (such as stock) on farms occupied for decades. The creation of the GAENP can thus be likened to those instances of removal perpetrated by previous reserves such as the KNP. Like many occupiers of communal land in the 1950s, current farm workers displaced by the GAENP often do not know where to go once a property has been transferred, usually because their stock has to be sold, and their social connections with the district are terminated once they leave.

With the start of a process of public participation in 2000, SANParks have recently embarked on a process of work provision in the GAENP area in 2002 and, subsequent to this, have also created a national framework policy document for conservation-related resettlement. Locally, a number of ex-farmworkers have been absorbed into a few jobs in the planned conservation area, and a few work teams (on six-monthly contracts) have been organised to break down homesteads and fences on those farms that have been purchased. However, although short-term employment offers a measure of relief, the loss of land, income and security suffered will definitely have to be compensated for by a more long-term solution to replace the advantages that workers have lost.

5. Place, belonging and the consequences of displacement

This paper has raised a number of issues. The first of these is the central importance that land use and the historical occupation of landscapes have for notions of belonging and identity. The varied geographical sites occupied by people in the GAENP, as well as the places of home and belonging in Mozambique, are intrinsically valuable to people as sites of homecoming and empowerment. In addition, as much of the work dealing with refugees and displaced people indicates, the very experience of helplessness in the event of displacement creates the motivation for people to somehow recapture the ideal of such a home.

In this respect, Lovell (1998: xii) uses the idea of ‘belonging’ as an idiom to refer to people who ‘care more about sentiments than origins’ which is common to many people who have historicised experiences of dislocation, displacement and spatial movement. Like farm workers in the GAENP, this idea of belonging is much more complex than the identity connected to an ethnic group – in that even through complete and ongoing displacement from a territory or home a meaning can still be derived from the event of removal itself. Similarly, Malkki (1995) applies the idea of belonging to refugees in Tanzania, who, like the farm workers in the proposed GAENP area, have been stripped of the specifics of culture, place and identity. These observations hold important consequences for new reserves such as the TFCA and GAENP, in that the memories of the ‘good life’ in these landscapes will not be diminished by removal from these areas, but will in fact increase resistance to the creation of a biological reserve.

Like those individuals affected by wars and political displacements (such as those created through apartheid), the residents affected by new ecological reserves are highly likely to associate conservation with disempowerment and exclusion unless they are given a more prominent role in the planning and management of protected areas. As Cernea (2000), de Wet (1995, 2002) and Scudder (1993) point out, those affected by displacement face a range of obstacles. Displacement from established homes and neighbourhoods often disrupts social relationships, which means further isolation and economic impoverishment in a new environment. In cases where people are not displaced by a project, as is the case in the Transfrontier initiative, people still experience a distinct lack of autonomy and control over their daily lives, political decisions being out of their control. Often these people are already marginalised by illiteracy and land hunger, as the residents in the GAENP are, which further reduces their chances of being recipients of the benefits attached to a Park.

However, incorporating social issues in new reserves involves far more than the creation of economic opportunities and the implementation of projects for ‘development’. New reserves will also have to acquire a fine understanding of the ways in which indigenous land users and occupiers have constructed their particular landscapes, both historically and in the present day. As a basic premise, environmentalists should consider that a landscape is not only influenced by complex issues of a ecological nature, but that land has a fundamental social and cultural value, entrenched by historical claims of occupation and a contemporary need for identity and belonging. These ideas of belonging have meant that Pafuri, in particular, has one of the highest rates of return for those displaced by war and conflict, simply because the area offers more freedom of movement and agricultural possibilities than others. Similarly, the displacement of people in the GAENP, both as labour tenants on sharecropping (or tenancy) land and as farm labourers has not diminished the people's attachment to the district. On the contrary, loss of land has intensified the desire to recapture lost territory, accumulate stock, seek employment and retain kinship ties in the district.

In the ‘deterritorialised’ world described by Appadurai (1996), the removal of familiar markers of belonging, through the combined effect of a new park and historical experiences of displacement, may have far-reaching ramifications for the future of these reserves. As was the case with the Maluleke in the Limpopo Province of South Africa, who were removed from the KNP in the 1970s and then successfully claimed back a portion of the Park in 1999, intense opposition to conservancies, and the ensuing land claims, may be the future results of exclusion from ecological decision-making. In many other instances involving people who are not as empowered as the Maluleke to voice their opposition (such as the subsistence cultivators on state land in Mozambique and farm workers in the GAENP), new parks may well serve to estrange the very people they seek to accommodate. It is therefore a matter of urgency that the social, economic and cultural rights of land occupiers in large wildlife areas be understood and addressed before these reserves are actually proclaimed.

Archival sources

Eastern Cape Archives. Port Elizabeth Depot. Kirkwood files (unmarked, 1948–85): Letters addressed to and from the Secretary of Plural Development, Minister of Bantu Administration, Secretary of Coloured Affairs and the Native Commissioner.

Additional information

Notes on contributors

Teresa Connor

The author wishes to thank Michael Whisson and Janet Cherry for commentary on a previous draft, as well as the editors of DSA. The National Research Foundation in South Africa, as well as Rhodes University, partly funded this project in 2000, 2003 and 2004. All opinions expressed in this paper are the author's own.