Conservation and Society

: 2015  |  Volume : 13  |  Issue : 2  |  Page : 129--140

Producing Gorongosa: Space and the Environmental Politics of Degradation in Mozambique

Michael Madison Walker 
 Current affiliation: Department of Anthropology, California State University, Sacramento, CA; Research undertaken at: Department of Anthropology, Dowling College, Oakdale, NY, USA

Correspondence Address:
Michael Madison Walker
Current affiliation: Department of Anthropology, California State University, Sacramento, CA; Research undertaken at: Department of Anthropology, Dowling College, Oakdale, NY


This article examines the spatial production of the greater Gorongosa ecosystem, linking the production of space with scientific discourses on environmental degradation. Ecological research conducted in Gorongosa National Park (GNP) in the 1960s established the spatial contours and produced the greater Gorongosa ecosystem that is continually under threat from Mozambican cultivators. This discursive production and its material effects obscure a long history of human occupancy and transformation of the landscape that is now categorised as a national park. The use of aerial surveys and satellite imagery by conservationists to chart biophysical changes in the landscape is central to the spatial production of the greater Gorongosa ecosystem. The knowledge produced through these ways of seeing the landscape is used to justify various socio-technical and legal interventions to protect the environment. Through analysing the discourse on human-induced environmental degradation in GNP between 2005 and 2010, I suggest that when nature and space are taken as self evident by conservation practitioners, there is a danger of reproducing narratives of environmental degradation that simplify historically dynamic interactions between people, institutions, and their biophysical surroundings, and serve as further justification for intervening in the lives and livelihoods of adjacent residents.

How to cite this article:
Walker MM. Producing Gorongosa: Space and the Environmental Politics of Degradation in Mozambique.Conservat Soc 2015;13:129-140

How to cite this URL:
Walker MM. Producing Gorongosa: Space and the Environmental Politics of Degradation in Mozambique. Conservat Soc [serial online] 2015 [cited 2020 Jun 1 ];13:129-140
Available from:

Full Text


In July 2010, the Government of Mozambique signed a new decree expanding the borders of Gorongosa National Park (GNP or 'the park') in central Mozambique. The new boundaries increased the overall size of the park by roughly 10% to 4,067 sq. km and established a buffer zone of 3,300 sq. km that limits human activity near the park. These new delimitations include all land parcels above 700 m on Mount Gorongosa, located 20 km west of the park, incorporating the mountain as a satellite protected area of the park. These changes in the park's boundaries coincide with the 50 th anniversary of Gorongosa's recognition as Mozambique's first national park. The combination of GNP and Mount Gorongosa, as protected areas, completes a long-held vision of officially recognising the greater Gorongosa ecosystem.

The Gorongosa Restoration Project (GRP), started in 2008, has ignited debates and revealed competing perspectives on African land use, biodiversity conservation, human settlement, and rural development in Mozambique. The impetus for expanding the protective auspices of the park derives from concerns over human livelihood activities on the mountain, which are believed to have deleterious effects on the mountain's hydrology that supplies water to GNP. Conservation scientists and volunteers with the Carr Foundation ('the Foundation' hereafter), an American philanthropic organisation working in partnership with the Mozambican government to rehabilitate and co-manage GNP, contend that smallholder agricultural practices on the upper elevations of the mountain are endangering the mountain's ecosystem through deforestation and soil erosion; and these ecological changes will have negative consequences for the park's flora and fauna and efforts to restore the park following Mozambique's 16 year war. 1 In post-war Gorongosa, many mountain residents experience the park restoration project as an attempt to usurp control over land and natural resources, to undermine their autonomy over where people live and cultivate, and how they organise their lives. There are roughly 2,000 people on the mountain's upper elevations, and its protected status now makes human land use and occupancy above 700 metres illegal. At the moment it is unclear to what extent the criminalisation of mountain land use will be enforced, or if and how people will be relocated from the upper elevations.

Protected areas not only establish boundaries and attempt to regulate social relations between people and nature, 2 but also work in the material and discursive production of space (West et al. 2006). Conservation, through protected areas, not only preserves landscapes, but also actively remakes them (Brockington 2009: 13), implying the salience and mutual constitution of spatial relations, material practices, and environmental discourses. In this article I examine the links between the production of space, protected areas, and discourses on environmental degradation in a national park in central Mozambique. More precisely, I explore the use of aerial photographs and satellite imagery as evidence used to mobilise legal and socio-technical interventions in the name of protecting the environment. I analyse important ecological research conducted in the park during the 1960s and the 1970s at a time when scientific management reached its zenith, and I explore aerial representations of and discourses about deforestation disseminated between 2005 and 2010 to justify action against rural cultivators living on the slopes of the mountain. In doing so, I want to consider the relationship between knowledge production about the environment and the production of space. Smith (1990) notes that nature and space have largely been taken as self evident. I problematise nature and space in Gorongosa to understand the micro-practices through which space is being remade in Mozambique, and the role of conservation through the protected area model in this process. When nature and space are taken for granted, the creation of protected areas in Mozambique often reproduces narratives of environmental degradation that simplify historically dynamic interactions between people, institutions, and their biophysical surroundings, and serve as further justification for intervening in the lives and livelihoods of people marginalised from decision-making over land use, conservation, and development.

 Protected Spaces

Since the mid 1980s, the number and the size of protected areas have expanded to encompass nearly 11% of the earth's land cover (West et al. 2006) with the largest increase in protected areas occurring between 1985 and 1995 (Brockington and Igoe 2006). This growth in the number and size of protected areas as a means of preserving biodiversity is linked to emerging partnerships between conservation and capitalism whereby new types of value are created for further expansion and accumulation (Brockington et al. 2008; Brockington and Duffy 2010; Igoe et al. 2010). These partnerships have resulted in various working alliances between conservation, environmentalism, government departments, capitalism, and philanthropy. The growing involvement of the private sector in conservation activities coincides with a return to the more exclusionary practices of fortress conservation (Milgroom and Spierenburg 2008: 436) premised on keeping people who are dependent on land and other natural resources out of protected areas. Conservation NGOs are at the forefront of this 'conservationist mode of production', deepening the relationships between biodiversity conservation and capitalist accumulation (Brockington and Scholfield 2010). The symbolic and physical separation of people from nature, on which many national park models are based, obscures the role people have played in shaping the natural contours of landscapes. Historically embedded engagement with the environment, mediated through human labour, social relations, and direct experience, is replaced by landscapes of consumption premised on observational detachment and picturesque views that bear no traces of human use or occupancy (Neumann 2002). Many of the aforementioned scholars have noted the highly uneven spatial distribution of protected areas with central and South America, East Asia, and eastern and southern Africa having the highest proportion of land dedicated to protection (West et al. 2006: 254). As protected spaces, these areas not only regulate human occupancy, livelihood activities, and access to resources, but also constitute ways of "seeing and governing the world" (West et al. 2006: 255). The partitioning of space becomes essential in demarcating protected areas and buffer zones, and in the governing practices instituted to regulate human-environment interactions within these landscapes. Spatial production becomes critical in the creation of these new types of value, thus spatialised struggles lie at the heart of debates over protected areas, biodiversity conservation, ecotourism, and development.

Conservation distributes fortunes and misfortunes (Igoe and Brockington 2007), and in the context of protected areas, what is categorised as 'protected,' who gets access to it, and under what conditions, is one means through which these fortunes and misfortunes are distributed. Duffy (2010) highlights numerous problems with conservation within the protected area model: 1) restriction of land and resource use by adjacent communities; 2) human displacement through forced removal or 'voluntary' resettlement; 3) the erasure of local histories and understandings of the surroundings; 4) privileged and often commodified use of the landscape and resources by tourists and scientists; and 5) the propensity to devalue nature outside of protected zones. While all of these issues present challenges for the various actors engaged in conservation activities, displacement is one of the most controversial aspects of protected areas (West et al. 2006). Displacement is often premised on an understanding of nature as separate from humans, and thus in order to protect nature, human activity must be curtailed, strictly controlled, or prohibited. "However, too often people can be moved from places without clear evidence that their actions and livelihoods are the cause of the problem. They are simply moved because people in parks are a category error" (Brockington 2009: 133). Protected areas, as people-less spaces, are inextricably linked to Euro-American conceptualisations of wilderness as uninhabited space (Cronon 1995) that serve as the model and basis of many national parks around the world, and contemporary conservation efforts reflect the values and assumptions embedded in western conceptualisations of nature. While the fate of mountain residents as well as people living near the newly enlarged buffer zone is uncertain in Gorongosa, conservation through the re-creation of people-less space is unfolding in other communities adjacent to national parks in Mozambique. Resettlement initiatives, through restrictions on livelihood activities and 'voluntary' relocation, are being implemented in both the Limpopo National Park and Banhine National Park in southern Mozambique. In these contexts, resettlement efforts are born not only from conservation policy and practice, but also emerge from government efforts to decentralise political authority and reorganise rural settlement patterns in the name of national development (Lunstrum 2008; Milgroom and Spierenburg 2008; Dear and McCool 2010).

Protected areas are constituted through ideological, discursive, and material processes. West (2006) argues that spatial production is essential to the creation and reproduction of national parks, and she chronicles the spatial production of the Crater Mountain Wildlife Management Area (CMWMA) in Papua New Guinea, emphasising that space is produced by the "social and material relations between peoples" (West 2006: 28). In this formulation, ideas, location, and social practices bring notions of space and spatial relations into existence. Once space is created through these social and material relations it produces "place, people, society, environment and so on; it comes to take part in the process of production" (West 2006: 28). West's work on the social, material, and discursive relations that produce and reproduce CMWMA highlights how the production of space and nature are intertwined within a particular conservation and development project. In this way, space is not synonymous with emptiness, but rather constituted through social relations, ideational expression, and material qualities of biophysical processes and human acts of production and reproduction. As Smith (1990: 95) argues it is only "in the advanced capitalist world today we all conceive of space as emptiness, as a universal receptacle in which objects exist and events occur, as a frame of reference, a coordinate system (along with time) within which all reality exists." National parks, such as GNP, do not exist because they are spatial containers filled with important flora and fauna worth protecting, but rather are produced through social practices that render certain ecological relations intelligible that are supported by a diverse network of scientists, conservation organizations, funding sources, media coverage, government officials, park managers, volunteers, and tourists.

Temporality is inextricably linked to the spatial production of protected areas, as the justification for protection is often premised on predictions of impending degradation or extinction or ecological changes that have already occurred. In Gorongosa, the often repeated time frame, before the mountain's hydrological network is irrevocably damaged, is three to five years. This time frame was first articulated in 2005 and has structured conservation activities aimed at reducing and mitigating the impact of human land use on the mountain. Consequently, conservation through national parks, game reserves, and protected areas not only produces space, but also 'makes time' through what anthropologist David Hughes (2005: 157) terms 'third nature'. In contrast to first and second nature, Hughes argues that third nature is a speculative and conditional representation laden with "assumptions or predictions regarding human action or inaction" (Hughes 2005: 158). In the debates over environmental degradation on Mount Gorongosa, crisis narratives articulate what 'could' or 'should' happen based on a particular view of people's impact on the biophysical world, reflecting Hughes' position that third nature "is not past, present, or future, but conditional; not a tense but a mood of speech" (Hughes 2005: 158). Whereas in the case Hughes describes the third nature articulated by conservationists represents a desired future, but in Gorongosa, future projections are undesirable and can be avoided by changing land use practices and human-environment relations. Thus, temporal speculation embodied in time-sensitive discourses of environmental degradation help to justify various socio-technical and legal interventions on the mountain and serve as an effective fundraising tool directed at environmentally conscious 'Westerners' who want to help protect and restore an important site of biodiversity.

'Environmental orthodoxies' and the production of space

Gorongosa's natural boundaries have been produced through multiple social practices at various scales. During the 1960s and early 1970s ecological research conducted in the park (discussed below) helped establish the contours of a greater Gorongosa ecosystem worthy of protection. Some of this work incorporated the use of light aircraft to conduct an aerial survey of the park and surrounding areas. Since 2005 the Foundation has used aerial surveys and transects conducted by helicopter and remote satellite imagery to build a case for the incorporation of Mount Gorongosa as a protected area. Seeing and interpreting the landscape from above, whether through one's own eyes or through the use of digital imagery, has played an important role in producing knowledge about socio-ecological relations in central Mozambique. Before I turn to the ecological research in the 1960s and the more recent discourses on the 'crisis' on Mount Gorongosa, I want to consider some links between the construction of knowledge about the environment and the production of space.

Euro-Americans have a long record of intervention in African environments that were often justified in the name of protecting nature from 'backward' and 'primitive' livelihood practices believed to cause irreparable ecological harm. Claims that African agricultural and livelihood practices degrade the environment have been used to justify various forms of sociotechnical intervention in colonial and postcolonial contexts (Beinart 1984). Contemporary thinking about African environments "has its roots in the European colonial period and is dominated by global narratives of degradation" (Bassett and Crummey 2003: 24). During the colonial period the environment emerged as a domain, through which colonial states and white settler populations attempted to govern African rural populations by restructuring livelihood practices in attempts to make peasant agriculture more 'rational' and 'efficient.' Colonial environmental discourses raised alarm over soil erosion, deforestation, siltation, and desertification and provided the pretext for state intervention. Many colonial land use regulations derived from perceptions and unsubstantiated evidence of environmental degradation attributed to African production and colonial administrators, agronomists, and settlers attempted to address these changes through the imposition of contour ridges, bans on wetland cultivation, destocking measures, or the resettlement of rural populations into linear grids organised around particular forms of land use.

Leach and Mearns (1996) refer to the institutionalized, 'common sense' understandings of environmental degradation as 'environmental orthodoxies.' These understandings are powerful discourses because they often maintain their persuasion in the face of counter evidence. In turn, these explanations are used to justify land use regulations and environmental policies that penalise smallholders or other socio-economically marginalised resource users. In an age where conservation and development efforts are articulated and implemented by transnational networks of African states, donors, international financial institutions, and private organisations, 'environmental orthodoxies' remain highly influential explanations of ecological change for their ability to simplify complex historical, social, economic, political, and ecological dynamics, and often serve to depoliticise land use, conservation, and development policy by framing the solutions to these problems through socio-technical interventions.

Orthodox explanations of environmental degradation are premised on an unchanging and static representation of the environment that ignores or minimises anthropogenic factors in shaping and maintaining landscapes. From this perspective, human alteration of the environment upsets its equilibrium and threatens the stability of the ecosystem. However, landscapes are not objective descriptions of a biophysical reality, but rather synthesise material changes in the environment and imaginative understandings of these ecological processes (Beinart and McGregor 2003). Thus landscapes and their interpretations are imbued with power relations and contested knowledge over the causes and directions of ecological change. Fairhead and Leach (1996) note that in colonial and postcolonial Guinea, discourses about deforestation and savannisation served to justify the usurpation of local control over natural resources. But as Fairhead and Leach demonstrate, forest cover has increased, not decreased, through peasant settlement and livelihood practices. In numerous contexts across the continent, African landscapes have been 'misread' by colonial and postcolonial governments, aid organisations, and development professionals looking to implement their definitions of sustainability and development.

Interventions into African land management often use spatial frameworks that inscribe new types of social-ecological relationships. Over the last decade Mozambican donors, NGOs, and government officials have embarked on a decentralised rural development model focused on making the countryside more legible (cf. Scott 1998). Numerous districts have been defined as national development priorities. These areas are then further divided into 'development zones,' placing emphasis on improving infrastructure and basic services. In order to intervene and promote desirable social and economic outcomes, Mozambique is broken down and conceptualised in terms such as development zones, agroecological regions, national parks, game reserves, buffer zones, and rural communities. Smith's (1990: 96) observation that "Consciousness of space is a direct efflux of practical activity" implies that one's notion of space is internalised through particular ways of knowing, and knowledge produced about Gorongosa and its surroundings has often come from looking down.

Between 2005 and 2010, park authorities relied heavily on aerial views of the mountain and the park to build a case for incorporating the mountain as a protected area. Aerial surveys and remotely sensed satellite imagery are indicative of what Haraway (1991) terms the 'god trick,' a disembodied view from everywhere and nowhere simultaneously that produces 'truth' by purporting to offer an objective look or a point of view beyond a position of representation. As Adey (2010: 86) explains, "The aerial view is in fact positioned, often taken to be from above looking down". He continues that "from above, the eye is neither clouded with subjectivity nor distracted by feelings; it is calculated, it is truthful." In the context of politicised environments, such as Gorongosa, maps, aerial photographs, and satellite imagery are powerful representations because of their ability to circulate while remaining unchanged, allowing multiple audiences to view the 'devastation,' thus stabilising the facts of the narrative of deforestation, soil erosion, or environmental degradation (cf. Braun 2002). Aerial perspectives are not only portable, but because they are seemingly objective, they can erase other ways of knowing born out of direct experience. One of, but not the only, ways knowledge has been produced and disseminated about Gorongosa has been through the use of pictures and satellite imagery of the mountain.

 Producing A Greater Gorongosa Ecosystem

Protected areas and their boundaries are as much social, political, economic, and legal as they are ecological, and as suggested earlier, reflective of particular ways of 'seeing and governing the world'. Article 109 of the Mozambican Constitution and Article 3 of the Land Law (Lei de Terra 19/97) define all land as state property. National parks fall under Article 7 of the land law and are defined as total protection zones. According to the law, total protection zones are areas intended for nature conservation and preservation activities where land use rights and benefits are prohibited. In addition to land, forest and wildlife resources are also state property (Lei de Florestas e Fauna Bravia 10/99). The Land Law confers numerous rights of use and occupancy to Mozambican smallholders and stipulates that in order for outside interests to use, occupy, or develop community lands, a public consultation process must be held and the community must consent to authorising the use of its lands. However, in the case of lands designated for protection, no consultation is required. The state can exercise its sovereignty by claiming the lands for national interest. Once this happens, national parks are accorded the highest protective status of all land use categories in the country. Lunstrum (2008) describes these protected areas as 'neoliberal state space' whereby the state regulates resource use and access, in effect, producing a private space, where prior residents no longer have legally recognised claims to their land or legal means for addressing the appropriation of their land. As Lunstrum (2008: 350) notes, the forest and wildlife law offers some room for negotiating land and resource use concessions in these areas, but these concessions amount to a 'state-granted privilege', not a legal right.

Mozambique has dedicated nearly 40,000 sq. km to national parks, 44,000 sq. km to national reserves, 42,000 sq. km to hunting blocks or coutadas, and 125,342 sq. km to game farms. At the turn of the twenty first century, national parks and reserves constituted about 11% of the country (Anesty 2001). Protected areas have expanded since then, including the creation of three transfrontier conservation areas since 1999, bringing approximately 16% of the national territory under some form of protection. While the overall percentage of land set aside for conservation related initiatives is less than other East African countries, Mozambique's increase in the area of land dedicated to conservation over the last 10 years has important implications for the spatial politics of land use and rural development, and reopens longstanding questions about where rural Mozambicans should live and cultivate (cf. Hughes 2006: 168-185).

GNP's boundaries are not self-evident. As described above, they have been produced through a set of social, legal, and economic relations. Perceptions of environmental degradation and human influence shape them as much as biophysical processes. The park's boundaries have been enlarged and reduced since Gorongosa reserve was demarcated in the 1920s, and the making and remaking of these boundaries often had little to do with the science of ecology. This would change in the 1960s when a new era of scientific management became the guiding framework for administering the park. The revised boundaries of 2010, in part, owe their dimensions to the vision of a South African ecologist who, for his doctoral thesis, traced out the interrelations between Gorongosa's hydrology, soils, geology, flora, fauna, and topography (Tinley 1977). The result was a comprehensive and holistic view of these various biophysical properties and processes as constituting a unified ecosystem linking space and scale. The ecological research conducted in the late 1960s and early 1970s continues to influence perceptions of the Gorongosa environment as a unified and stable system, and the latest revisions to the park's boundaries come closer to categorising the amount of land as protected that was envisioned for the viability of the Gorongosa ecosystem. However, despite the saliency of conservation and ecological sciences to the expansion of GNP, the park's origins are embedded in another logic.

Mozambique has experienced limited state intervention on behalf of wildlife (Anesty and De Sousa 2001). Portuguese settlers introduced the first interventions that targeted the relationship between people and wild species. GNP's origins are intertwined with the period of company rule, which combined colonial extractive activities organised around forced labour and administrative governance. In 1921 the Companhia de Mocambique, a Portuguese charter company with administrative rights over central Mozambique from 1891 to 1942, established a game reserve of 1,000 sq. km to protect wildlife for hunting by notable European visitors and control the supply of ivory and bush meat. The park's origins derive, not from concerns for the total preservation of wildlife, but rather as a space reserved for the aesthetic and leisure activities of particular racial and class interests of colonial society, and to control the trade in meat and ivory. In 1935, the company increased the reserve's size to 3,200 sq. km, designating three times the amount of land for European interests as the original demarcation. The colonial government awarded Gorongosa Reserve national park status in 1960, making it the country's flagship national park 15 years before Mozambique's independence. At the time of recognition as the country's first national park, the boundaries were increased to 5,300 sq. km. National park status ushered in a new era of 'scientific management' premised on the idea of wilderness as space devoid of people (French 2009: 212). In keeping with the notion that national parks are premised on the separation of people from nature, in 1966, the boundaries were reduced to 3,770 sq. km to ensure that neighboring people did not reside within the park (Tinley 1977: 6). During the colonial period the park's boundaries shifted, in response to European hunting interests, concerns over African land use and cultivation, and finally as Gorongosa garnered national and international recognition as a protected area. However, each of these demarcations failed to adequately enclose enough land and natural resources for what would become known as the greater Gorongosa ecosystem.

In 1968, a South African ecologist, Kenneth Tinley began four years of ecological research in Gorongosa for his doctoral thesis at the University of Pretoria, South Africa. As Tinley notes in the opening of his thesis, several previous ecological researchers had worked in the area prior to him, including an ornithologist, a pedologist, and several Portuguese botanists, but these scientists often operated within a single disciplinary framework. Tinley's research and thesis 'Framework on the Gorongosa Ecosystem' is the first to synthesise multiple fields within the natural sciences to construct a comprehensive analysis of the dynamics and what he termed 'reciprocal relations' between flora, fauna, moisture, climate, geology, soils, and topography. Tinley used a 'holistic ecological approach' that examined "the salient reciprocal relations of and succession of important biotic communities and their components with landscape processes" (Tinley 1977: 7). Tinley immersed himself in the ecology of Gorongosa, writing that, "Ecosystems study requires the worker's complete immersion and empathy with the subject, to 'feel' in the Taoist philosophic sense by becoming the ecosystem oneself-I am the inselberg, the plains and the totality of the elements of life at play on them" (Tinley 1977: 8). His immersion yielded a rich understanding of these biophysical dynamics and implanted the idea of a greater Gorongosa ecosystem into the collective consciousness of present and future conservationists.

Tinley was very much attuned to spatial relations and sought to delineate the interrelationships between the various environs that he believed comprised the Gorongosa ecosystem. He outlined four interconnected topographical environments: Gorongosa Mountain, a 1,863 m inselberg rising from the plains; the Barue or Midland Plateau stretching from the mountain's eastern slopes; the southern trough of the Rift Valley that encapsulates Lake Urema; and the Cheringoma Plateau that forms the Rift Valley's eastern escarpment. Together these physiographic elements made up the greater Gorongosa ecosystem, worked in conjunction to shape precipitation patterns, regulate moisture essential to the life, and viability of the area's flora and fauna. He also situated these topographical zones in relation to all of central Mozambique between the Zimbabwean border and the Indian Ocean, and the Zambezi River to the north and the Pungue River to the south, in his efforts to understand the totality of the interrelations.

Tinley argued that the mountain was the catalyst for producing the area's moisture. The mountain's presence in the midst of the surrounding plains influenced the southeasterly winds causing them to drop precipitation on the mountain's upper elevations. These heavy orthographic rains supplied the mountain with more than 2,000 mm of annual rainfall, supporting rainforests and montane grasslands that act as a "sponge capturing and releasing water in a constant radial pattern of flow"(Tinley 1977: 5). The moisture captured on the mountain's upper elevations fed a perennial river system responsible for directing water through the midlands and channelling it to Lake Urema at the southern tip of the Rift Valley and within the confines of the newly demarcated park.

Tinley's interest in the scalar and spatial relations of reciprocal ecological processes also resulted in the first aerial survey of the park. From the vantage point of a light aircraft, he mapped the contours of the Urema watershed, outlined rainfall zones, counted animals including Waterbuck, Zebra, Wildebeest, Hippopotami, Elephants, and Buffalo as well as calculated their total biomass. He represented his findings on maps where he sketched the Vunduzi and Mucosa Rivers in thick dark lines descending from Gorongosa Mountain draining into the Urema basin and filled in the Nhanduhe, Muanza, Nhandinde Rivers, marked with dotted lines. The water channelled by these perennial and seasonal rivers was essential to the water levels of Lake Urema, a permanent freshwater lake at the southern tip of the park from where a diversity of wild species stopped to drink. From the air it became apparent to Tinley that because the rivers, vital to the park's water supply, originated outside of the park's boundaries, the boundaries must be expanded to protect the river systems from human influences that could alter the balance of the park's waters essential to the wildlife. The 'aesthetic sensibility' produced from the air magnified some aspects of the landscape while it minimised others (cf. Hughes 2010: 84). From a mobile, omniscient perspective the mountain and its network of rivers could be understood and valued in new ways. Thus by stepping back or moving up, a reciprocal and self-regulating environment began to take form. In a 35 page report entitled 'Gorongosa National Park: An Ecological Sketch,' Ken and Lynne Tinley propose the ecological boundaries of GNP as much larger than any of the previous demarcations, including the expansion of the park's boundaries in 1960. They summarise the ecological limits and the centrality of Gorongosa Mountain to the park:

In 1968 and 1969 a special study to determine the ecological boundary limits of the park revealed that the entire Gorongosa Park Ecosystem and population of over 20,000 African cultivators west of the Park are all dependent solely on the perennial water resource born on the isolated Gorongosa Mountain. The ecological limits proposed in the above study will enlarge Gorongosa National Park to 8,700 sq. km, and include the greater part of Gorongosa Mountain which will be protected as a water catchment area quite apart from its tremendous scientific and recreational value (Tinley and Tinley Undated: 3).

Through Tinley's work, GNP was no longer understood as an island surrounded by other landscapes, but intimately woven into much larger ecological processes contingent on the distribution of water from the mountain. The ecological limits of Gorongosa became synonymous with the Urema drainage basin and thus necessitated the same type of protection extended to the park. The understanding of the greater Gorongosa ecosystem as a stable and self-regulating landscape reflects the equilibrium model of ecosystems dominant within ecology of the 1960s. From this view, people were seen as 'unnatural' and disruptive to an ecosystem's natural functions. Tinley's vision of the park's ecological limits as extending beyond the historical and contemporary boundaries of the park continues to shape conservation policy and practice in Gorongosa today. In particular, efforts to 'restore' GNP now rest on initiatives to save Gorongosa Mountain through restricting human use and occupancy above 700 metres. Ecological science and location have melded to produce place-a place in need of restoration.

The 'crisis' on Mount Gorongosa

Mozambique's 16 year war destroyed Gorongosa's infrastructure and killed off the majority of the park's wildlife. When the Frelimo state and Renamo (the South African backed rebel movement) signed a peace accord in 1992, GNP was no longer the tourist destination that it was in the 1960s and early 1970s. During the war, Renamo's main base was located in Gorongosa District, and the area experienced extensive fighting. The district's population today is 117,129 (INE 2009) with over half of the population living in or around Gorongosa town. Today, Renamo (the largest opposition party) still receives considerable support in Gorongosa, further politicising questions around land use, state resource allocation, and national development. It is in this postwar environment that the Mozambican government and donors have attempted to restore and revitalise the country's flagship national park.

The park's most comprehensive restoration efforts have come under the guidance and financial assistance of the Foundation in cooperation with the Mozambican government. In 2008, the Foundation and the government signed a 20 year agreement to comanage the park. The park's management and restoration is premised on combining biodiversity conservation, for-profit tourism, and local socioeconomic development. Tourism is particularly important to this model, as it will fund the long term restoration of the park and serve as the engine for local economic development. The park's restoration efforts are also supported by United States Agency for International Development (USAID), the Government of Portugal, the Mozambican Museum of Natural History, Kruger National Park in South Africa, numerous travel and tour operators, and several Mozambican universities.

While enlarging the park and buffer zone boundaries are essential to the vision of restoration, other conservation activities have included the reintroduction of wild species killed off during the war (many of them airlifted to GNP from other countries in the region) and a reforestation programme on the slopes of Mount Gorongosa. The Foundation has also contributed to local educational and health infrastructure. Individuals, regardless of place, can also participate in these efforts through making donations on the park's website that help pay for the reforestation programme or animal relocation costs. The Foundation has also proposed irrigated agricultural settlements to entice people from the lands on Mount Gorongosa. In the near future students will be able to learn about the earth's rich biological patrimony through a digital textbook called Life on Earth, segments of which were filmed alongside evolutionary biologist, E.O. Wilson's visit to Gorongosa in July 2011. All of these initiatives draw on and reproduce the idea and materiality of Gorongosa as an important site of biodiversity and an endangered place.

Although Gorongosa's restoration involves multiple conservation and development activities conducted by local and translocal actors, in the remainder of the paper, I want to focus on the 'crisis narrative' regarding Mount Gorongosa that was first publicised in 2005 when the Foundation became involved with the park. This narrative is instructive for thinking about how space is produced in the context of a conservation and development project that, at least initially, took nature as self-evident. The global circulation of this narrative and related images told the story of an impending ecological crisis created by rural cultivators that reproduced long-standing depictions of African farmers as threats to nature. The continued production of a greater Gorongosa ecosystem became intertwined with environmental crisis discourses. In the context of Gorongosa, it is not just capital that remakes space in its own image, but narratives of environmental degradation also reproduce space and serve as the rationale for reordering socio-ecological relations. In the last section of the paper I will discuss the more recent use of satellite imagery and aerial surveys as important technologies in the production of this crisis narrative and in the production of space, but first I turn to what became the orthodox explanation for a series of environmental issues on the mountain.

I first became aware of the crisis on Mount Gorongosa in the dry season of 2006. I was living in Manica Province in central Mozambique, several hours from GNP. In late September and early October, following a meeting between the Foundation, and various communities and government officials in Gorongosa, the story of the mountain's degradation circulated through the donor and expatriate community of central Mozambique. The story I heard at the time was that agriculturalists living on the mountain were clearing land through 'slash and burn' agriculture, causing deforestation and erosion that threatened the park located to the east. What at first seemed to be a discussion contained within the social world of aid and development organizations working in the country was soon circulated on a global scale. Articles appearing in The New Yorker and Outside Magazine and a National Geographic documentary carried all or part of this story of environmental degradation caused by mountain residents. The argument presented in foundation-sponsored reports, updates on the park's website, in documentary films, and print media claimed that residents living on the upper elevations of the mountain were responsible, through their land use practices, for endangering the mountain's ecology and hydrological network that fed the park below.

Consequently, saving the mountain became an imperative for restoring the park. However, at the time this narrative of degradation was supported by little ecological evidence documenting the effects of human activity on the mountain and subsequent relation to the park. Instead, the narrative worked to depict a crisis and the explanation for it. The refrain publicised by the Foundation, "To save the park, we have to save the mountain", reflects a shift within biodiversity conservation from a concern with individual species to preserving entire ecosystems (Takacs 1996), and underlies the restoration project's vision and efforts to incorporate the mountain within the park's boundaries.

Material disseminated by the Foundation argues that agricultural and livelihood practices, glossed as 'slash and burn', are responsible for deforestation, soil erosion, and siltation that threaten the mountain's network of rivers and streams and the transfer of water from the mountain to the national park below. The Foundation contends, "It will take no more than three to five years before the ecosystem is degraded to a point from which it is unlikely to recover," and identifies the primary threats to the mountain as "deforestation resulting in erosion, the loss of old growth forests and the biodiversity (flora and fauna) they contain, the degradation of the water supply due to sedimentation, debris, and riparian agriculture, which is believed to cause erosion, for people downstream and GNP" (Carr Foundation 2006). The document continues, "Slash-and-burn encroachment and uncontrolled fire escaping from hunting or clearing are the two main destructive forces behind the loss of the forest ecosystem on Gorongosa Mountain." These 'destructive forces' are believed to accelerate land clearing. They estimate the total forest loss between 1992 and 2000 at around 2.3%, but state that "recent clearing and burning will dramatically revise that figure upwards." The time frame of three to five years before irrevocable ecological damage, along with the causes, have been well publicised on the park's website as well as within the various print and video coverage the park and foundation have received over the last three years. 3

In other materials posted on the park's website, explicit links are made between water and the vitality of the park. For example, "The survival of Gorongosa NP depends on the sustainable management of water resources throughout the entire catchment of the park" (Beilfuss 2006: 9). Water management includes the rivers above 700 metres as well as a 10 kilometer catchment zone from the park's boundaries. The project draws on Integrated Water Resource Management (IWRM), a water management model premised on stakeholder participation, safeguarding water for its ecological importance, and implementing demand-driven regulations for the use and allocation of water and the construction and maintenance of infrastructure that has gained prominence across southern Africa. "As most of the water resources of Gorongosa National Park originate outside the park's boundaries, water resources management must be integrated with regional planning and development efforts, especially the Regional Water Authority of Central Mozambique and various stakeholders of the Pungwe [sic] Basin." (Carr Foundation 2006: 8).

Despite the assertions of an environmental crisis, proponents of the mountain restoration project acknowledge the lack of a research and monitoring network to assess the biophysical changes in the landscape (Beilfuss et al. 2007). In addition to the lack of empirical evidence, the project is largely silent on the government's proposals to build several dams within the Pungue basin that are likely to have an effect on the park's water resources. Instead, mountain residents are presented to wider publics as the primary threat to the mountain's forests and rivers and the biodiversity of the park. Depicting farmers as threats to the mountain is contingent on representing African landscapes as static, stable, and separate from the convergence of anthropogenic and biophysical processes that have shaped them for millennia. In the next section I explore the use of aerial perspectives to document and understand landscape change and how disembodied views from above reinforce the 'common sense' explanations of ecological change.

A 'black box' in the sky

Degradation is not a neutral term; it is contextual and embodies a particular point of view (Bassett and Crummey 2003: 18). The way in which environmental degradation is often deployed by various conservation organisations constitutes what Bruno Latour (1987) terms a 'black box.' A 'black box' signifies the foreclosure of interactive processes and replacing them with a concern for understanding only inputs and outcomes. As Latour (1987: 3) suggests "no matter how controversial their history, how complex their inner workings, how large the commercial and academic networks that hold them in place, only their input and output count." In the context of GNP and debates over anthropogenic impacts on the mountain, concepts such as deforestation and soil erosion conceal more than they reveal. They serve as foregone conclusions and justifications for reconfiguring people's relationships with the environment, and they reinforce an a priori reasoning that substitutes for rigorous empirical analysis. Furthermore, by treating these concepts as self-contained with predictable outcomes, conservationists obscure many of the uncertainties that accompany their projections of ecological change. I use the following quotes concerning agricultural impacts on the mountain's hydrology from the Carr Foundation (2006) to exemplify:

The 'likely' effect on the hydrology of land clearing and deforestation will be to reduce the total evapotranspiration and to increase the direct runoff while decreasing the baseflow. The expected hydrological effects can be summarized as increased volume of surface runoff, increased peak flows in surface runoff, decreased dry season flows, and decreased groundwater recharge. As a result, some perennial streams may become seasonal (11: author emphasis).

'If' surface flows are responsible for maintaining Lake Urema and wetland areas during the dry season, then the changes to the flow regime identified above could result in partial drying out of wetlands and the lake in the late dry season. 'If' the lake and wetlands are mainly fed by groundwater flows, then the effects of land clearing on Gorongosa Mountain may not be as severe if some of the additional surface runoff is captured as groundwater recharge at the rift margins. Reduced groundwater infiltration on the upper reaches of the mountain 'may' affect more distant groundwater discharge points (11: author emphasis).

In either case, the Rift Valley system acts to extend the duration of flows from the lower Urema, and 'may' mitigate the effects of increased peak run-off from the Gorongosa Mountain (11: author emphasis).

Words such as 'if,' 'could,' 'may,' and 'expected' are conjectural, and depend on a range of factors before becoming true, what Hughes (2005) refers to a 'third nature'. Ironically, these words implicitly acknowledge contingency and uncertainty and perhaps better characterise the dynamism within ecologies and between human-environment relations. However, in spite of the recognition within the foundation's own reports of the ambiguity between mountain land use and the mountain's hydrology and the implications for GNP, mountain residents were demonised within much of the conservationist discourse surrounding GNP from 2005 to 2010, and their land use and occupancy rights criminalised under the new government decree and expansion of the park's protective boundaries. This is not to suggest that natural or social scientists engaged in environmental research should wait until negative biophysical outcomes are apparent before policies are devised and actions implemented, but it does warrant careful consideration of particular historical, political, and social values concerning people's relationship with the environment that informs what constitutes 'negative' outcomes and asks the question, 'negative' for whom? The lack of baseline empirical ecological data on soil formation and loss, tree and vegetation change, and socio-economic data on people's livelihoods and agricultural practices, that goes beyond misnomers such as 'slash and burn,' raises epistemological questions concerning the basis on which the decision to annex the mountain is premised. As Stott and Sullivan (2000) remind us, national and international environmental policies often acquire legitimacy by framing environmental threats in the language, not the actual practice, of science.

The concepts such as deforestation and soil erosion are premised on an understanding of nature as inherently stable, orderly, and unchanging and that natural environments function to maintain their equilibrium in the face of change. The deforestation narrative assumes tropical forests are pristine and fragile while the concept of soil erosion intertwines multiple biophysical processes. Forsyth (2004: 32) questions the accuracy of the word 'erosion' partly because of its linkages with crisis and ponders how well it indicates "the causes of soil degradation, or the most fitting policies to address it." In the Foundation's claims that soil erosion threatens the mountain and its waterways, there is no distinction made between the removal of soil by wind or by water or data indicating the rate of soil formation. Furthermore, are there acceptable levels of soil loss that allow the ecosystem to continue its current 'functions' while altering its 'structure'? To what extent do intense storms, common in central Mozambique, shape soil loss? By fixating on anthropogenic changes to the environment and framing them as 'soil erosion,' project scientists collapse the multiple biophysical processes producing soil formation and loss with human alteration of the environment.

The evidence of deforestation is equally problematic. To borrow again from Forsyth (2004: 34) "Asserting 'deforestation reduces biodiversity' therefore depends in part upon the particular definitions of deforestation and biodiversity." In Gorongosa, mountain deforestation has come to signify changes in tree cover associated with the felling of trees to clear land for agriculture. Throughout central Mozambique field clearance is done by hand, making it labour intensive with smallholders rarely clearing more land than they can cultivate. Depicting smallholder agriculture as 'slash and burn' ignores the ways in which farmers actually prepare fields, plant and harvest crops, and let unused fields lie fallow as well as how farmers' livelihoods practices are embedded in broader political and economic structures. This is not to suggest that every transformation of the environment made by mountain residents is without ecological consequences, but rather the discourses of soil erosion and deforestation that circulated from 2005 to 2010 seemed to be based on long-standing representations of rural Africans and less on empirical evidence.

Much of the evidence used to make a case for deforestation on the ground comes from the air. Satellite imagery and helicopter flyovers have been used to document changes in forest and land cover and to support the claim that mountain forests are in decline. Remote sensing, satellite imagery, and geographic information systems (GIS) can be important tools for recording landscape change. However, these technologies, and the maps and images they produce, privilege certain ways of seeing and knowing the environment over others and consequently must be triangulated with embodied forms of knowledge to provide more complete accounts of what unfolds on the ground. In addition, aerial views are still subject to competing interpretations and how the categories of interpretation are constructed is important for how one sees the landscape from above (Robbins 2003). Aerial views of the park, the mountain, and surrounding areas were selected early on because conservationists did not think it was important to talk with people on the ground. The crisis rhetoric also made it appear there was no time for a more in depth study and consultation. People who have lived on the mountain were not seen as potential sources of knowledge. 4 Aerial views are powerful. Conservationists first used light aircraft in Africa to count animals, particularly the wildebeest migration in the Serengeti (Adams and McShane 1996). Their omniscient quality, while an illusion, supplies the audience and onlookers with a literal point of view. Transects and satellite imagery often carry the veneer of scientific authority and the use of these tools and the studies that emerge, are powerful representations. Aerial transect maps and satellite images of the mountain were often positioned alongside calls for donations on the Gorongosa park website operated by the Foundation.

Time and scale are important variables in deciphering changes in forest cover. It is important to distinguish between long-term gradual change and short-term, punctuated changes. Because changes can unfold at multiple scales and within varying temporal frames, it is difficult "to infer a causal relationship with observed changes in forest cover" (Walker and Peters 2007: 76). Assessing changes in forest cover also depends on the temporal reference points. Temporality is particularly important in contexts where seasonality shapes and reshapes the appearance of the landscape. Throughout central Mozambique, it is common for the landscape to become dry and brown from August to November only to become verdant once again after the onset of the first significant rains. Thus, measurements of forest and vegetation changes calculated through satellite images, remote sensing, aerial photography, and helicopter fly-overs must be attentive to these seasonal dimensions of landscape change. Since 2006 the Foundation has asserted that only three to five years remain before irrevocable change is done to the mountains ecology and hydrology and thus the time for action is now. French (2009: 75) notes that conservationists today use Gorongosa in the 1970s, when conservation efforts were most successful and tourism development the highest, as the reference point for 'restoring the park.' Contemporary observations of the park's landscape are compared with a time period when environmental planning in colonial Mozambique was at its highest and when ecological scientists began to articulate the narrative of 'people as a threat' to GNP.

Time and scale are also vital considerations when interpreting satellite or remotely sensed images. There are several problems associated with using satellite images to document ecological changes on the mountain. There is not one satellite that photographs the entire mountain, thus images are spliced together to produce a representation of the mountain. Cloud cover also limits what satellite images can reveal during the rainy season. The mountain is most visible by satellite when the landscape is the most barren. It is difficult to distinguish between primary and secondary forests and other micro-ecological changes within forested landscapes from satellite photographs. Nor do satellite images reveal the causality of suspected changes. People must interpret them, thus they are subject to multiple and competing interpretations.

The Foundation conducted six preliminary aerial transects by helicopter that also helped shape the crisis narrative. Helicopter transects are also problematic because they often cover narrow tracts of land, and then observers make generalisations about the landscape from these narrow strips. Both techniques for representing changes in the landscape emphasise a detached and omnipotent point of view that may contradict, or fail to capture, ecogenic and anthropogenic changes in forests, fields, soil composition, and river systems. Relying on these technologies also fails to address how people on the ground understand and experience changes in the landscape. The use of satellite images and other aerial depictions of landscape change in the hands of bureaucratic and regulatory agencies often penalises or marginalises local resource users who are defined as the cause of these changes. It appears this has happened in Gorongosa. Conservationists' selective reading of the park's history and landscape are informed by tacit assumptions concerning the pristine and fragile nature of the Gorongosa's environment. Contemporary ecological changes are compared against an imaginary past environment. This 'misreading' of Gorongosa's landscape has engendered conflict between park authorities, conservationists, and mountain residents who now feel their livelihoods, land, and sovereignty are under threat.

Finally, what impact the change in the forest cover and soil composition will have on the hydrology of the mountain and park is equally uncertain. There was not a system to measure evaporation or evapotranspiration or much information on groundwater recharge and the seasonal variance in water flow and extraction. The Carr Foundation acknowledges these uncertainties:

Although the present extraction of water from the catchment is insignificant (<1%) and likely to remain minor relative to the total annual inflow volume to the Urema floodplain for some time, the impact of water extraction may be significant at certain specific locations, especially at the end of the dry season or during prolonged drought periods. Little is known about groundwater recharge rates (Carr Foundation 2006: 15).

In addition to the lack of data on hydrological processes, the potential for large-scale dams along the Pungue River is not condemned with the same vigour as smallholder land use practices. One of the few places where the Carr Foundation mentions the dams:

Any large dams or water extraction projects that alter natural runoff patterns in the Pungwe River pose a significant threat to Gorongosa NP. Of particular concern are the proposed Bue Maria and Pavua Dams, located immediately upstream of the park, which would reduce or eliminate the influx of floodwaters from the Pungwe River, lowering water tables and reducing the carrying capacity for wildlife in the southern part of the Park where tourism and wildlife are concentrated (Carr Foundation 2006: 13).

The discourse of irrevocable environmental degradation on the mountain became prominent in 2005 when the Foundation became involved in rehabilitating the park. With little empirical evidence to support such a claims conservationists drew on and reproduced long-standing stereotypes and images of rural Africans as harmful to nature and justified Western, scientific intervention. Aerial views of the area's landscape were used to support these claims and deployed strategically to garner international support and attention to the restoration project.

Gorongosa has long been a politicised environment, and its current environmental politics are linked to ways of knowing and valuing the landscape that produce competing claims to land. Since the mountain's incorporation as part of the park in 2010, land and forest clearance above 700 metres has accelerated. 5 By contesting residents' long-standing land use rights and removing their legal right to land, the park restoration project has helped create the type of 'crisis' that was to be avoided. Aerial surveys and methods of viewing the ground from above evolved alongside disciplinary regimes of governance (Adey 2010), and the use of aerial techniques in biodiversity conservation are likely to increase. 6 Thus, a critical examination of the production of space, nature, and knowledge in biodiversity conservation, particularly regarding the protected area model, is necessary for understanding how these technologies are used to produce an 'objective' depiction of ecological relations to further specific conservation goals and values, especially when they conflict with residents' land use practices.


In this paper I have tried to suggest that although Gorongosa National Park's borders existed in 1960, and as a hunting reserve since 1921, comprehensive ecological research conducted in the area in the late 1960s and 1970s brought to life a greater Gorongosa ecosystem that in order to remain sustainable needs protection from human influences. Thus it was through the spatial production of Gorongosa that reciprocal ecological relations emerged and the park's ecological limits became real to actors and institutions tasked with managing the park, and more recently, restoring the park after Mozambique's protracted civil war. Central to the spatial production of the Gorongosa ecosystem is the use of aerial views, whether through maps, satellite imagery, or helicopter and aircraft vantage points, as representations of a unified and symbiotic landscape. In addition to the visual representations from above, long-standing narratives of environmental degradation associated with African land use continue to inform perceptions of anthropogenic changes in environment with little ecological evidence. Scientific discourses situated within an equilibrium framework for understanding the environment have played a prominent role in the perpetuation of the view that rural Mozambicans are a threat to nature and biodiversity conservation. These assumptions derive from constructions of a pristine environment as stable and orderly and fail to recognise the role human settlement and labour, through livelihood activities, has played in shaping and maintaining the landscapes of central Mozambique. The narratives of deforestation and soil erosion function as 'environmental orthodoxies' in explaining and placing blame for ecological change, and attempts to 'restore the park' are premised on a 'third nature' or speculative and imaginary future contingent upon human (conservationist) intervention that is likely to have negative consequences for MaGorongosi, that are defined as threats to the mountain and park ecosystem.

Conservation through the protected area model in Mozambique raises multiple ecological, social, and political questions and highlights the inequality and power relations between multi-million dollar organisations, ecologists, hydrologists, international tourists, Mozambican officials, and rural residents who depend on land and water for their livelihoods. Furthermore, the emphasis on environmental degradation by the project's advocates fails to understand the historical, political, and economic factors shaping mountain settlement and cultivation and is likely to further alienate and marginalise mountain residents who are suspicious of and do not trust the government or those associated with the park.


I presented some of the arguments in this paper on the panel, 'Protecting an African Eden?: Conservationists, Communities, and Collaborations in Mozambique,' at the 2010 African Studies Association Annual Meeting in San Francisco, CA. I would like to thank the panelists and audience for their comments and questions. Comments by Christy Schuetze, Todd French, and Rozenn Diallo strengthened this article at an early stage. Christy Schuetze deserves special thanks for sharing her in-depth knowledge of Gorongosa with me. Three anonymous reviewers and editorial staff for this journal provided valuable suggestions and guidance for which I am grateful.


Rhodesia and apartheid South Africa played significant roles in Mozambique's armed conflict between Frelimo and Renamo from 1977 to 1992. Over the course of the conflict, the fighting also drew on and produced local tensions far removed from the racial and Cold War politics of southern Africa.I have resisted placing nature or greater Gorongosa ecosystem in quotation marks, as it will become evident these are not natural categories, but rather social constructions. The Gorongosa website no longer includes a time frame predicting when the Gorongosa ecosystem is likely to be damaged beyond repair. See Schuetze (Forthcoming) for an examination of mountain residents' voices and competing narratives around Gorongosa.See Schuetze (Forthcoming).For example, Google recently donated US 5 million to the World Wildlife Fund to buy drone aircraft to fight poaching in Africa. See:; Accessed on March 6, 2014.[45]


1Adams, J.S. and T.O. McShane. 1996. The myth of wild Africa: conservation without illusion. Berkeley, CA: University of California Press.
2Adey, Peter. 2010. Aerial life: spaces, mobilities, affects. West Sussex: Wiley-Blackwell.
3Anesty, S. 2001. Necessarily vague: the political economy of community conservation in Mozambique. In: African wildlife and livelihoods: the promise and performance of community conservation (eds. Hulme, D. and M. Murphree). Pp. 74-87. Oxford: James Currey.
4Anesty, S. and C. de Sousa. 2001. Old ways and new challenges: traditional resource management systems in the Chimanimani Mountains, Mozambique. In: African wildlife and livelihoods: the promise and performance of community conservation (eds. Hulme, D. and M. Murphree). Pp. 195-207. Oxford: James Currey.
5Bassett, T. and D. Crummey. 2003. Contested images, contested realities: environment and society in African savannas. In: African savannas: global narratives and local knowledge of environmental change (eds. Bassett, T. and D. Crummey). Pp. 1-30. London: James Currey.
6Beinart, W. 1984. Soil erosion, conservationism, and ideas about development: a southern African exploration, 1900-1960. Journal of Southern African Studies 11(1): 52-83.
7Beinart, W. and J. McGregor. 2003. Introduction. In Social history and African environments (eds. Beinart, W and J. McGregor). Pp. 1-24. Oxford: James Currey.
8Beilfuss, R. 2006. Proposal for Gorongosa National Park buffer zone delimitation and management of the greater Gorongosa ecosystem. Presented at the Fourth Gorongosa National Park Stakeholders Meeting. Carr Foundation.
9Beilfuss, R., F. Steinbruch, and R. Owen. 2007. Long-term plan for hydrological research: adaptive management of water resources at Gorongosa National Park. Carr Foundation.
10Braun, B. 2002. The intemperate rainforest: nature, culture, and power on Canada′s west coast. Minneapolis, MN: University of Minnesota Press.
11Brockington, D. 2009. Celebrity and the environment: fame, wealth and power in conservation. London: Zed Books.
12Brockington, D. and J. Igoe. 2006. Eviction for conservation: a global overview. Conservation & Society 4(3): 424-470.
13Brockington, D., R. Duffy, and J. Igoe. 2008. Nature unbound: conservation, capitalism and the future of protected areas. London: Earthscan.
14Brockington, D. and R. Duffy 2010. Capitalism and conservation: the production and reproduction of biodiversity conservation. Antipode 42(3): 469-484.
15Brockington, D. and K. Scholfield. 2010. The conservationist mode of production and conservation NGOs in sub-Saharan Africa. Antipode 42(3): 551-575.
16Carr Foundation. 2006. Gorongosa mountain: status and recommendations for conservation and sustainable development.
17Cronon, W. 1995. In search of nature. In Uncommon ground: rethinking the human place in nature (ed. Cronon, W.). Pp. 23-66. New York, NY: Norton.
18Dear, C. and S. McCool. 2010. Causes and consequences of displacement decision-making in Banhine National Park, Mozambique. Conservation & Society 8(2): 103-111.
19Duffy, R. 2010. Nature crime: how we′re getting conservation wrong. New Haven, CT: Yale University Press.
20Fairhead, J. and M. Leach. 1996. Misreading the African landscape: society and ecology in a forest-savanna mosaic. Cambridge: Cambridge University Press.
21Forsyth, T. 2003. Critical political ecology: the politics of environmental science. New York, NY: Routledge.
22French, T. 2009. Like leaves fallen by the wind: resilience, remembrance, and the the restoration of landscapes in central Mozambique. PhD thesis. Boston University, Boston, USA.
23Haraway, D. 1991. Simians, cyborgs, and women: the reinvention of nature. New York, NY: Routledge.
24Hughes, D.M. 2005. Third nature: making space and time in the Great Limpopo Conservation Area. Cultural Anthropology 20(2): 157-184.
25Hughes, D.M. 2006. From enslavement to environmentalism: Politics on a southern African frontier. Seattle, WA: University of Washington Press.
26Hughes, D.M. 2010. Whiteness in Zimbabwe: Race, landscape, and the problem of belonging. New York, NY: Palgrave Macmillan.
27Igoe, J. and D. Brockington. 2007. Neoliberal conservation: a brief introduction. Conservation & Society 5(4): 432-449.
28Igoe, J., K. Neves, and D. Brockington. 2010. A spectacular eco-tour around the historic bloc: theorising the convergence of biodiversity conservation and capitalist expansion. Antipode 42(3): 486-512.
29INE (Instituto Nacional de Estatistica). 2009. Mozambique census of 2007.
30Latour, B. 1987. Science in action: how to follow scientists and engineers through society. Cambridge, MA: Harvard University Press.
31Leach, M. and R. Mearns. 1996. The lie of the land: challenging received wisdom on the African environment. Portsmouth, NJ: Heinemann.
32Lunstrum, Elizabeth. 2008. Mozambique, neoliberal land reform, and the Limpopo. National Park. The Geographical Review 98(3):339-355.
33Milgroom, J. and M. Spierenburg. 2008. Induced volition: resettlement from the Limpopo National Park, Mozambique. Journal of Contemporary African Studies 26(4): 435-448.
34Neumann, R. 2002. Imposing wilderness: struggles over livelihoods and nature conservation in Africa. Berkeley, CA: University of California Press.
35Robbins, P. 2003. Fixed categories in a portable landscape: the causes and consequences of land cover categorization. In Political ecology: an integrative approach to geography and environment-development studies (eds. Zimmerer, K. and T. Bassett). Pp. 181-200. New York, NY: Guilford Press.
36Schuetze, C. Forthcoming. Narrative fortresses: crisis narratives in the conservation of Mount Gorongosa, Mozambique.
37Scott, J. 1998. Seeing like a state: how certain schemes to improve the human condition have failed. New Haven, CT: Yale University Press.
38Smith, N. 1990. Uneven development: nature, capital, and the production of space. Athens, GA: University of Georgia Press.
39Stott, P. and S. Sullivan. 2000. Political ecology: science, myth and power. London: Arnold Publishers.
40Takacs, D. 1996. The idea of biodiversity: philosophies of paradise. Baltimore, MD: The Johns Hopkins University Press.
41Tinley, K. 1977. Framework on the Gorongosa Ecosystem. DSC Wildlife Management. University of Pretoria, South Africa.
42Tinley, K. and L. Tinley. Undated. Gorongosa National Park, Mozambique: An Ecological Sketch. Unpublished paper.
43Walker, P. and P. Peters. 2007. Making sense in time: remote sensing and the challenges of temporal heterogeneity in social analysis of environmental change-cases from Malawi. Human Ecology 35: 69-80.
44West, P. 2006. Conservation is our government now. the politics of ecology in Papua New Guinea. Durham, NC: Duke University Press.
45West, P., J. Igoe, and D. Brockington. 2006. Parks and peoples: the social impact of protected areas. Annual Review of Anthropology 35: 251-277.