Community-Based Forest Management in the Lucite (Rusitu) Valley : People and Policies of a Proposed Mozambique-Zimbabwe Transfrontier Conservation Area

The World Bank Maputo, Mozambique Community-Based Forest Management in the Lucite (Rusitu) Valley: People and Policies of a Proposed Mozambique-Zimbabwe Transfrontier Conservation Area FILE Wo,!t,1! n ~,r ,-,LC- C^rl,f .L . by David M. Hughes Department of Anthropology University of California Berkeley, California 94720 USA tel: +1 510-642-3391 fax: +1 510-643-8557 September 1995 Copyright by author  TABLE OF CONTENTS Acknowledqements i Abbreviations i Summary of Recommended Activities ii Maps iv Part I Introduction and Purpose of the Studv 1 Part II Policy Aspects of Transfrontier Conservation in Mozambique 2 The land question in Mozambique 2 Community-managed conservation 4 Border-related issues 6 Development in Manica and Chimanimani 7 Part III Communities and Land in the Lucite Valley 8 Research Methodology 9 Forests and other natural resources 10 Fields and other commercial interests 13 Migration and Demography 15 Political dynamics on the Zimbabwe side 15 Political dynamics on the Mozambique side 20 Part IV Management of Lucite's Protected Areas 23 Nyakwawa / Rusitu Botanical Reserve 24 Chizire / Haroni Botanical Reserve 27 Chimanimani National Park 29 Part V Recommended Activities 31 Notes 37 Bibliographv 42 Officials Consulted 47 Note: This report differs from the May 1995 version only in the correction of a number of minor typographical and grammatical errors and in the use of a different font, altering pagination. The content of the two versions is identical.  ACKNOWLEDGEMENTS Numerous individuals and organisations rendered substantial assistance in the course of my field and archival research. First and foremost, residents of the Lucite Valley gave unreservedly of their time, of their knowledge and of their sadza. The Vhimba Area Development Committee and Zvidzai Chidhakwa deserve especial thanks for providing a roof over my head and friends in the field. My assistant, Elias Muhanyi, carried out a strange job with competence and humour, although he bears no responsibility for the conclusions I draw from our wanderings. SAFIRE made my stay in Lucite possible by introducing me to the community and transporting me there. In Kopa, the Chigogo family and staff of the Rusitu Valley Smallholder Tea Project gave me further logistical support. The Centre for Applied Social Sciences at the University of Zimbabwe allowed me to present some of findings as a seminar and provided invaluable comments thereon. Finally, I am grateful to James Bannerman for access to his large collection of documents on Manica Province, Mozambique and to the staffs of the Arquivo Hist6rico de Mogambique (Maputo), the Arquivo do Patrim6nio Cultural (Maputo) and of the National Archives of Zimbabwe (Harare). ABBREVIATIONS AHM Arquivo Hist6rico de Mogambique (Historical Archives of Mozambique) ARPAC Arquivo do Patrim6nio Cultural (Cultural Patrimony Archives) CAMPFIRE Communal Areas Management Programme for Indigenous Resources CEF Centro de Experimentacao Florestal (Centre for Forestry Research) DNFFB Direcgao National de Florestas e Fauna Bravia (National Directorate of Forestry and Wildlife) DPA Direccao Provincial de Agricultura (Provincial Directorate of Agriculture) Frelimo Frente de Libertagao National de Mogambique (Front for the National Liberation of Mozambique) GTZ Gesellschaft fur Technische Zusammenarbeit (German Technical Cooperation) IUCN International Union for the Conservation of Nature and Natural Resources NAZ National Archives of Zimbabwe NGO non-governmental organisation Renamo Resistencia National Mocambicana (Mozambican National Resistence SAFIRE Southern Alliance for Indigenous Resources SPFFB Servigo Provincial de Florestas e Fauna Bravia Provincial Services of Forestry and Wild- life) Vidco village development committee WWF Worldwide Fund for Nature i  SUMMARY OF RECOMMENDED ACTIVITIES Stage 1 - Immediate Activities 1. Form a Chimanimani collaborative group. Relevant agencies and a contingent of Mozambican organisations selected by SPFFB Chimoio should hold a workshop in Chimoio to decide on their roles and responsibilities in the Chimanimani Transfrontier Conservation Area. If the workshop cannot be held entirely in Portuguese, then simultaneous translation will be required. 2. Visit Regulo Mafussi: The collaborative group should send a two-person team to conduct a one-week visit Regulo Mafussi's area, in nothern Mossurize District. The visit would have three objectives: first, explain the project to Mafussi and seek his permission to proceed; second, determine the boundaries of Mafussi's land; and, third, map the sacred forests and other cultural sites in that area. 3. Propose a concessions ban: Based on the information gathered above, DNFFB can write a justification for protecting important cultural and natural resources by prohibiting timber and other concessions within the boundaries of Chief Mafussi's area. Stage 2 - Contingent upon Funding for Implementation of the Transfrontier Project 4. Obtain necessary staff: Based on decisions taken in the workshop above, a Mozambican agency - probably SPFFB Chimoio - will hire two fieldworkers for the Chimanimani region. These individuals will make extended visits on a regular basis to rural communities and establish a lasting relationship between them and agencies implementing the Project. The fieldworker responsible for Lucite will be based in Rotanda or Mavita. 5. Encourage sustainable fuelwood harvesting: The Lucite fieldworker will meet with Mozambican women who sell fuelwood in Zimbabwe. He or she will encourage these individuals to exploit the increasing Zimbabwean demand by establishing a selling point, a marketing cooperative or a savings club. The collaborative group might send a CEF employee to help these women design a "management plan" for the sustainable harvesting of their secondary forests-cum-"extractive reserves." Stage 3 - Contingent upon an Exemption Allowing Cross-Border Tourism 6. Help form a local management committee: The fieldworker should stimulate Mozambican Lucite residents to form a committee ii capable of making decisions and handling finances in regard to cross-border tourism. The committee could contract with the CAMPFIRE-related local committee on the Zimbabwe side of the Lucite or collect monies directly from tourists crossing the river. Stage 4 - Contingent upon an Exemption Allowing Cross-Border Commerce 7. Assist agricultural marketing cooperatives: Once allowed to sell fruit and other products in Zimbabwe, Mozambican residents could join the Rusitu Valley Fruit Growers' Association or, with the help of the fieldworker, establish their own structures for selling in bulk across the border (the only market accessible given existing infrastructure in northern Mossurize).  A ,A -4.tb W \ t \t Lx ~ ~ (A ji;t,~~~ ,~ * 4 g 7- 0 *~ ~ town. t, -r I 2~~~~~~~~~~ t°^ tiivof ' ~~~~~~~~~~~~~~~~~A t- a C, lf^ n _L '4- N) st'-\c&-4 ) R. .tw>\ 4. + - ^A .. LkM P\* A As e 9 !±K L ' % ivsK _ _  A tw;29Kfoks s+) - 1'al~ ~ ~ ~ -~i b~~ il- -js^ - ' - - / 9 Bss 16v9 0 } a) \A/ I~~~~~/ / *1'J5 / 8 I. INTRODUCTION, SUMMARY AND PURPOSE OF THE STUDY This study explores some the ways in which smallholder farmers in the Lucite River Valley (called the Rusitu River in Zimbabwe) manage forests and agricultural land. Previous, mostly ecological studies of this area have focussed on deforestation and on the degradation of protected areas. This study uses anthropological methods to investigate cultural aspects of resource management, and its conclusions differ markedly from those of some of its predecessors. The Ndau-speaking people who live in this area (Ndau is a dialect of Shona.) have been managing its natural resources for centuries. To a large extent - an astonishing extent given the pressures of colonial land alienation and of recent refugee flows - the people of Chiefs Ngorima and Mafussi have conserved the forests they consider sacred. Part III explains in detail the system of land use and political leadership among Lucite Valley residents. Such local practices and institutions, as opposed to externally-imposed restrictions, can form an appropriate basis for future forest conservation in this area. This finding will assist the World Bank in promoting the Mozambique Transfrontier Conservation Areas Project (hereafter "Transfrontier Project"). This anthropological study is one of a number of social science research projects intended to orient the implementation of the Transfrontier Project in three areas of Mozambique. Those areas are southern Maputo Province (including the Maputo Elephant Reserve), northern Gaza Province (including Zinave Park, Banhine Park and Coutada 16 Reserve in Mozambique as well as Kruger National Park in South Africa and Gonarezhou National Park in Zimbabwe) and, in Manica Province, the Chimanimani Mountain region. This report suggests means of carrying forward a strategy of community-managed conservation in the Lucite Valley of the Chimanimani Transfrontier Conservation Area. Yet, first the following, second section of the report situates this strategy in the contexts of land reform, community- based conservation and border-related policies in Mozambique. That section also discusses development problems and prospects for Manica and Chimanimani, and, finally, evaluates plans for the Transfrontier Project in light of the approach herein recommended. The report's intended audience is, however, by no means limited to the World Bank. A major goal, perhaps the major goal, of the Transfrontier Project is to the strengthen the Mozambican National Directorate of Forestry and Wildlife (DNFFB) and its Provincial Services for Forestry and Wildlife (SPFFB). Since the World Bank is a funding and not an implementing agency, other organisations must eventually own and manage the Transfrontier Project if it is to succeed. For reasons that will be explained later, the DNFFB and especially the Provincial Services are the most logical agencies to implement the Transfrontier Project. 1  The author has, therefore, tailored the recommendations in Part V to the needs and capacities of this arm of the Mozambican government. As a result, some suggestions in the report - such as, the reduced size of the target area and the gradualist approach to tourism development - may not conform to the original spirit of the Transfrontier Project. Overall, the author presents a modest conservation plan, one that outside agencies can assist effectively and residents of the Lucite Valley can control. That plan reflects close consultation with residents of the Lucite Valley. The author met repeatedly with the Vhimba Area Development Committee and once with a group of Mozambican sellers of fuelwood. II. POLICY ASPECTS OF TRANSFRONTIER CONSERVATION AREAS IN MOZAMBIQUE The future progress of the Transfrontier Project depends on reform in a number of areas of policy in Mozambique. Government and NGOs are currently debating the two most important of these areas, land tenure and community-based approaches to conservation. Previous World Bank reports, especially the "Preparation Studies" (Environment and Development Group 1994), address these topics. This study supplements and, in the case of conservation, somewhat redirects their observations. This section also examines border policies, a hithertofore neglected concern of the Transfrontier Project. Finally, the author summarizes development problems and institutions particular to Manica Province and to the Chimanimani Transfrontier Conservation Area. The land question in Mozambique Currently in most parts of Mozambique, individual titles provide the only secure access to land. Since the procedure for acquiring a title is complicated and expensive, the vast majority of farmers in the family sector have not and will not be able to obtain individual titles. Under these conditions, physical presence on the land is the best means of defending rights, and many rural people assert ownership primarily through occupation. Since the end of the war, however, Mozambique's rural population has been extremely mobile. Displaced people and refugees are returning to their homes or finding new, more attractive homes. These migrations are causing ripple effects and reverse flows as migrants search for areas of unoccupied, fertile land close to infrastructure. Many conflicts arise between smallholders.' Conflict over land is also heightened by the continuation of resource-grabbing in post-war Mozambique. During the war, responsible government authorities lost control of large parts of rural Mozambique. In the absence of effective regulation, 2 Mozambican and foreign entrepreneurs exported minerals, timber and wildlife products, including ivory. In some areas of the country, there was a violent free-for-all in which smallholders lost land and resources. Unfortunately, some characteristics of this period have outlasted the war. The state now issues concessions to timber, hunting and other enterprises. Again, in some areas of the country, different levels of government - national, provincial and district - have granted concessions on easy terms with little oversight. It is the responsibility of district and provincial agricultural officers (the DPA) to vet and reject those concessions which would displace family sector farmers. Since such land is not titled or centrally registered, staff must visit the area in order to make such a determination. The DPA has neither sufficient staff or vehicles nor does Mozambique have sufficient roads to make comprehensive inspections possible. Therefore, there is reason to believe that concessions have been granted for land already occupied and that smallholders have been or will be evicted. This kind of second, post-colonial alienation of land threatens to dispossess and underdevelop Mozambican smallholders. If unchecked, concessions may create a class of landless, destitute people, on the brink of famine and/or civil unrest. Timber, hunting and mining concessions may also deplete Mozambique's natural resources. It appears that many concession- holders are still influenced by a war-era get-rich-quick mentality. Sensing that peace and/or the legality of their concessions is uncertain, many timber or hunting companies may choose to extract resources fast rather than to use them sustainably over the longer-term. This kind of business is not only socially and ecologically destructive but also it discourages investment in the long-term development of Mozambique. The Transfrontier Conservation Areas Project may be able to forestall such effects in its proposed areas. The project calls for a moratorium on all concessions in the three areas, and there are currently moratoria in the Futi Corridor in southern Maputo Province and in the entire proposed Chimanimani conservation area. These provisions will enable government and NGOs to devise long-term solutions to the problem of land insecurity. In rural Maputo Province, the Swiss NGO Helvetas has assisted in the formation of peasant associations, corporate entities that may more easily obtain land titles (see Pereira and dos Santos Alves 1994). Perhaps, in some areas of the country, the existence of sacred forests can substantiate residents' claims to long-term occupation and consequent rights. Yet, this approach may not be feasible or desirable in parts of Mozambique where smallholders wish to preserve their ability to migrate and move livestock, both crucial strategies for coping with drought and war. Additionally, smallholders in different areas and those in the same area will likely disagree on the appropriate role of 3 "traditional" leaders in land tenure and land management. In any case, the moratoria on concessions are the beginning of a process of working with communities to ensure their rights with the instruments and institutions they desire. Community-managed conservation This kind of process - one through which communities gain control over forces affecting their lives - should characterise the entire approach of the Transfrontier Conservation Project. To a large extent, it already does. The World Bank and DNFFB have embraced principles of community-based conservation. CAMPFIRE (Communal Areas Management Programme for Indigenous Resources) in Zimbabwe provides a model for this effort, and many in Mozambique are attempting consciously to adapt CAMPFIRE to the conditions of that country. Initiated in the mid-1980s, CAMPFIRE intends to make the residents of wildlife-rich areas partners in the conservation and sustainable use of wildlife. These "producer communities" should, first, earn income from wildlife and other resources and, second, participate in management decisions. Where these conditions obtain, and the producer community is small enough for individual households to earn substantial income, people are likely to maintain the integrity of their local environment (Murphree 1991). They should opt for the most profitable, sustainable use of resources - an option that usually excludes commercial poaching (but might include illegal local hunting for meat and crop protection). CAMPFIRE has, therefore, made communities "stewards" of natural resources and actually decentralised the management of some resources, chiefly wildlife. The Department of National Parks and Wildlife Management has granted "appropriate authority" to many district councils, allowing them to sell rights to use a quota of wildlife. Most districts have thereupon contracted with sport- hunting firms. Sport-hunting concessions bring tremendous revenue for the development of CAMPFIRE areas. Communities have responded most positively where they have received tangible and immediate benefits from the programme. Where the district council has devolved the control of revenues to the ward or village level, as in Beitbridge District (Child and Peterson 1991), residents have obtained cash payments and development projects from wildlife- related funds. In the case of Chikwarakwara village in Beitbridge, smallholders of the producer community met and actually voted on the allocation of their wildlife income. In order to increase that income, these and other smallholders have undertaken to conserve and restore the wildlife habitat. In sum, CAMPFIRE can promote conservation in a fashion that gives communities greater power. 4  Mozambican agencies are wise to try to adapt this aspect of CAMPFIRE. Yet, they should also note some of the areas in which they can improve upon CAMPFIRE. While CAMPFIRE is undoubtedly community-based, it is often not community-managed or community- controlled. In many areas outside Beitbridge, district councils often manage natural resources on behalf of communities. In so doing, they may redistribute revenues from the wildlife-rich producer communities to the general population of the district (Murombedzi 1991). These actions reduce communities' incentive not to poach or hunt for crop protection or meat. More fundamentally, in few CAMPFIRE area does the producer community actually manage wildlife as opposed to revenues alone. Officially, CAMPFIRE has restored to communities the right to hunt for subsistence or crop protection, both management decisions removed from them by early Rhodesian legislation. De facto, however, many communities have not taken advantage of these rights, and some district councils have discouraged them from doing so. In such cases, communities cannot be said to control CAMPFIRE projects. In the worst cases, concession-holders exert power over communities. If local people's herding, fishing or fuelwood collecting diminish the hunting habitat or the "wild Africa" aesthetic, many sport-hunters attempt to restrict those activities. These individuals are armed, and sometimes they are supported by government anti-poaching units; when conflicts 2 ensue, local people feel intimidated and victimised. Zimbabwean communities in this predicament cannot be expected to conserve the very wildlife for which they suffer. Mozambican authorities have the opportunity to avert such situations by undertaking community-managed conservation. Indeed, in Mozambique, the rather more attenuated structure of local government may necessitate greater control on the part of the producer communities themselves. Many rural areas of Mozambique lack entities equivalent to Zimbabwe's district councils, that is, government institutions representing the inhabitants of clearly defined geographical areas. Proponents of community-based conservation are, therefore, almost forced to work with grassroots leaders and associations. The "Mozambique Resource Management Initiative," implemented by DNFFB and funded by the Ford Foundation, has begun this process in the Bawa area of Tete Province. The Transfrontier Project may soon do so in Chimanimani, Gaza and southern Maputo Province. First, it must overcome the ambivalence expressed in statements advocating, to paraphrase, the "participation" of communities as "stakeholders" in the project. This approach treats residents as just one of a number of interested parties. Rather than assuming a false equality between residents and outsiders, the Project should prioritise the needs and plans of rural committees. The Project can then begin to work with the longstanding and emergent grassroots institutions governing resource use in the three 5 transfrontier areas. This report proposes a means to initiate such collaboration with traditional leaders and local committees in the Lucite Valley. Border-Related Issues While planning documents of the Transfrontier Project have considered conservation policy and, to a lesser extent, the land question, they have devoted much less attention to issues pertaining to the frontiers themselves. In the Gaza Transfrontier Area, the Project does recognize the tourist potential of a bi-national or tri-national park. Yet, it is no accident that Kruger, Gonarezhou, Chimanimani and other regions of high biodiversity in Southern Africa lie along its international frontiers. Historically, British and Portuguese settlement expanded to natural barriers of mountains (such as, Chimanimani) or tsetse-fly belts (such, as the Zambezi and Limpopo Valleys). Such areas already contained rough, scenic terrain or large populations of wild mammals. Further aspects of settlement and colonisation concentrated wildlife in border regions. As they alienated the rich land of the central plateau, Rhodesian settler farmers systematically hunted and otherwise eradicated its big game. A similar result was achieved in white agricultural areas of South Africa. Conservation projects in Southern Africa are, thus, likely to touch or cross borders, and community-based conservation must confront the particular social and economic problems (as well as opportunities) posed by borders. Residents of frontier communities often live, grow crops, seek wage labour, or collect wild resources in two countries. Many of these practices are illegal under the laws of one or both countries. To the extent that border residents evade arrest, they may enjoy economic advantages not available to those living in core areas of their countries. Even so, these kinds of "gray market" activities rarely lead to long-term growth and development. Because these ventures are semi-legal and hence insecure, people seldom commit substantial investment to them. Many Mozambicans in the Lucite Valley, for example, refrain from planting fruit tree orchards although there is an unsaturated market for fruit on the Zimbabwean side. Indiscriminate rules against "smuggling," therefore, can hinder even laudable forms of agricultural development in border communities. The Transfrontier Project should aim to expand the range of development options legally available to resident communities. At the very least, it should avoid narrowing those options. There is a danger that this project will invite unsympathetic government scrutiny of cross-border trade and travel in the three Transfrontier Areas. The Project should, therefore, raise these 6 issues as matters of development planning before they become cases for law enforcement. Development in Manica and the Chimanimani Area of Mozambique The Lucite Valley forms only a small, rather peripheral part of the proposed Chimanimani Transfrontier Area. In fact, the precise area of this study is just outside the current boundaries of the proposed area. Previous studies related to the Transfrontier Project have considered much larger sections of Chimanimani, although exclusively within Mozambique. The "Preparation Studies" (Environment and Development Group 1994) establish the boundaries of the Chimanimani Conservation Area as well as those of a core protected zone above the 1000 meter contour line (see Map #1). A joint Mozambican-Zimbabwean research team recently carried out participatory rural appraisals in the localities of Mavita and Dombe (Makuku, et al,1994). That study identifies communities in the Chimanimani Area and describes their use of resources. Both studies point out that the Chimanimani region contains faunal species and mountain scenery that could hold touristic value. These studies of Chimanimani have served to gather information rather than to establish relationships with rural communities. Yet this kind of link and a permanent rural presence is precisely what Mozambican agencies need most. The need is especially urgent in Manica Province and in the Chimanimani Mountain region. Renamo occupied the bulk of the province from Dombe southwards, including the lower one-third of the proposed Transfrontier Conservation Area. Government services and personnel, as well as those of NGOs, did not reach people in those areas from the mid-1980s to 1992 or even until this year. Much of the rest of province was insecure, contested territory for much of the war. There has thus been a widespread rupture of the links between rural communities and external sources of development assistance. The rupture is also political: people fear that government will not deliver a fair share of benefits to Manica Province. Manica residents, like those in the other central provinces, registered this disaffection in the 1994 elections by voting overwhelmingly in favour of Renamo. For the continuation of peace and security, as well as for Manica's development, foreign and local efforts must enable the civil service to function throughout the province. By supporting the DNFFB, the Transfrontier Project has the potential to contribute towards this goal. DNFFB is the agency best suited to oversee the sustainable use of resources in the Transfrontier Conservation Areas. As a government department, DNFFB can grant or influence other branches of government to grant management rights to communities. It also has the authority - that NGOs lack - to enforce contracts between 7  communities and sport-hunting or tourism firms. DNFFB's ten provincial offices are, of course, in an even better position to work in areas distant from Maputo. DNFFB is likely to decentralise and pass greater responsibility to the Provincial Services for Forestry and Wildlife (SPFFB). Provided these offices obtain funds for fieldworkers - a recommendation of this report for Manica - they will be authorised and equipped to implement community-based conservation. When it does start to work in the Chimanimani area, the SPFFB may collaborate with a number of institutions and groups of individuals. The Mozambique Agricultural Rural Reconstruction Programme, funded by GTZ, recently completed an "Integrated Rural Development Plan for Manica Province." When implemented, this plan will enhance the capacity of government services tremendously. If insufficiently assisted, however, civil servants may find GTZ's ambitious goals unattainable and eventually alienating. The Ford Foundation's work in Manica Province has combined research and action rather more effectively. Ford is sponsoring staff of the Centro de Experimentagao Florestal (CEF) to receive academic training and conduct applied research. One or more of these individuals could greatly assist an SPFFB fieldworker in the Chimanimani area. SPFFB could similarly collaborate with the new Chimoio office of ARPAC, the Arquivo do Patrim6nio Cultural (Cultural Patrimony Archives). As Part V suggests, an ARPAC staff member could undertake a culturally-sensitive visit to Mossurize District. Finally, because of its proximity to the border, Chimanimani should contain many trained repatriated refugees. During the war, the Zimbabwean NGO Christian Care offered courses in the refugee camps in agriculture and other income generating skills. Many of the these people have returned to Manica Province, and many are eager turn their skills and general experience with development towards the betterment of their areas (see Christian Care 1993-1995). In Chimanimani, an SPFFB fieldworker might seek these people out and work with them in supporting communities' plans for resource use. Finally, SPFFB itself should receive funds for staff and/or vehicles as part of a project sponsored by the African Development Bank. III. COMMUNITIES AND LAND IN THE LUCITE VALLEY The Lucite Valley and its inhabitants are the main subject of this study. This section illustrates the ways in which these people manage the valley's natural resources. Detailed information of this kind can support the case for community- managed conservation proposed above. Parts IV and V of this report will draw on the material now presented to advocate that outside agencies support institutions and practices of forest use in the valley. Like previous ecological works on Lucite, this 8 cultural analysis devotes great attention to forests. Yet it draws a distinction - one supplied by local residents - between mature and immature woodland. People manage these woodlands differently. The ways in which they tend to extract certain resources from certain forests provide a basis for sound conservation. Such practices of sustainable use comprise only one aspect of a community's interaction with its environment. The leadership and institutions of a given community also play a vital role in determining who consumes what resources. Indeed, studies that focus on resource use alone overlook crucial political issues of access to and control over land. The bulk of this section describes the function of various local leaders in allocating agricultural land: chiefs (r&qulos in Portuguese terminology), or madzimambo (singular, mambo); headmen, or masadunhu (singular, sadunhu); and village heads (chefes de grupo in Portuguese), or masabhuku (singular, sabhuku). Since land clearing and forest preservation are flip sides of the same coin, these individuals effectively manage Lucite's woodland. Many writers and policy-makers in Southern African conservation are now suggesting that governments devolve greater power over land to traditional leaders.3 While the author is sympathetic to these findings, this analysis does not accept on faith either the effectiveness or the immutability of traditional leadership in Lucite. As will be demonstrated in an extended historical discussion, traditional leaders and their allocation of land in Lucite arise from a complex dynamic of state administration, land alienation and migration. Nonetheless, these land allocators have overcome tremendous pressures successfully to conserve substantial amounts of primary forest. Part IV presents further evidence for this conclusion. Research Methodology In March and April 1995, I conducted a total of 1 1/2 months of anthropological fieldwork on both sides of the Lucite. Living on the Zimbabwe side of the Lucite, I was the official guest of the Vhimba Area Development Committee and a participant observer in its and other community functions. More formal fieldwork involved interviews in Shona without translation. In Zimbabwe, a local assistant and I conducted these interviews mostly at people's homesteads. Because of unforeseen difficulties in crossing the border4 , I interviewed most Mozambicans during their frequent visits to Zimbabwe. Many of these interviews, especially those with masabhuku and masadunhu, were repeated as I gained people's trust and confidence. In the course of these activities, my assistants and I traversed and mapped much the forested and farmed areas of the Lucite landscape. I also attended meetings of the Vhimba Area Development Committee and convened a meeting of Mozambican women. These meetings allowed 9 me to consult Lucite residents on their plans for forest management and to base this report's recommendations thereon. In this connection, I also met with officials of governmental and non-governmental organisations, a list of which is included at the end of this report. Finally, I reviewed a small number of relevant documents at the National Archives of Zimbabwe (Harare), the Arquivo Hist6rico de Mogambique (Maputo) and at the Arquivo do Patrim6nio Cultural (Maputo). The bibliography provides a description of these materials. Forests and other natural resources The Lucite landscape is formed by the Lucite River and the slopes bounding it on both sides. In the area studied, the river flows from west to east at an elevation of 300-350 meters. On the nothern or Zimbabwean bank - an area known locally as "Vhimba" - the slope rises steeply, at a gradient of as much as 40% (SAFIRE 1995: 7), forming a virtual escarpment at 1100 meters elevation. The southern, Mozambican bank contains more lowland, broken by uneven hills. Nowhere do these hills form a ridge or plateau, as they do in Zimbabwe, except further downstream where the Lucite turns southward into Mozambique. In that area, which is, in any case, outside the immediate study site, the left or eastern bank abuts the Sitatonga ridge. To the north of the study area lies the Chimanimani Mountain Range with peaks of up to 2436 meters (Mt. Binga). The Chisengu, Haroni and Mukurupini rivers all rise from these mountains and meet the Lucite at the edge of the study area (see Map #2). This varied, well-watered terrain supports diverse floral and faunal communities. Lowlands along the Lucite, Haroni and Mukurupini Rivers contain moist forest. In strips and larger patches, Newtonia buchanii (mufomoti) and Khava nyasica (muwawa) trees provide a closed, high canopy. Further up the slope, forest cover is lower and more open, dominated by Uapaca kirkiana (muzhanie) and Parinari curatellifolia (muchaka). According to Zimbabwe's Department of National Parks and Wildlife Management, these two forest types combined hold up to 800 floral species (Timberlake 1993: 1). Additionally, tremendous bird, amphibian and mammal life is present within the forest and in the ecotonal edge zones between forests and fields. One study found 80 bird species in the lowland forest, and they are known to contain species found nowhere else in Zimbabwe (although perhaps elsewhere in Mozambique). Finally, small numbers of various buck and monkey species, including the rare Grant's night ape inhabit the valley (Robertson 1986: 104). Lucite residents use the resources of the forests for a wide range of purposes. A study conducted by Steven Mavi lists 111 plant species collected for food, medicine, fuelwood, poles and other building materials and for the carving of tools and 10 household furniture (Mavi n.d.: 5-10). I also found that nearly all the most common trees served one or more practical ends. In particular, my assistants cited as sources of fuelwood and/or of poles nearly all species encountered during one walk in the Haroni Valley. In the bulk of cases, residents collect dead and dry wood for fuel while they must cut poles live. As per the relatively strict gender division of labour in Lucite, women collect fuelwood, and men cut poles (Hobane 1993: 14). With the notable exception of Uapaca kirkiana and Parinari curatellifolia, my assistants and I found relatively few wild species yielding edible fruits.5 Mavi's and the bulk of Zimbabwe-based botanical research has concentrated on Lucite's primary forest. Secondary forest, which is much more represented on the Mozambican side of the Lucite, has been almost totally ignored. Residents themselves distinguish between two types of forest: sango (plural: masanqo) or in Ndau dialect, gwasha (plural: makwasha) is the general term for all forest with large trees. Forests in the process of regrowth after having been cut and burnt for a field are termed makura (probably from the verb -kura: to grow). Eventually, makura become mature as makwasha. Thus the indigenous categories differentiate between maturing and fully mature forest, as opposed to primary and secondary. This distinction may be more useful given that regeneration can render a cut parcel indistinguishable from surrounding uncut woodland within ten years. While regenerating, woodland may actually yield more of some resources than mature forest. Makura contain a high proportion of saplings, which are ideal for poles. In Lucite, residents identified makura as by far the better source of poles. Many, perhaps those with a more restrictive definition of the forest type, did not consider makwasha to contain even a worthwhile number of poles. As less productive, climax forest, makwasha should contain mostly tall, mature trees and leafy undergrowth. When these tall trees die and dry out they become a convenient source of fuelwood. According to many women, makwasha produce more fuelwood than makura, but makura still provide some dry wood. In other words, makura are an inferior, but still viable, source of fuelwood, while makwasha may not furnish even measurable quantities of poles (fieldnotes 15/4/95). Further ecological research could confirm or modify these generalisations. If makura are the better source of resources overall, it is not surprising that they should fall under a strict tenure system. On the Zimbabwean side of the Lucite, in Vhimba, where there is comparatively little forest, tree tenure is quite strong. All makura are considered to be long-fallowed fields and are owned by the person who last cut the forest. Presumably the owner (muridzi) will burn that makura in a number of years and 11 plant it again. Therefore, anyone wishing to harvest poles or fuelwood from that makura must obtain permission, usually from the male head of the household. Most people in this position will grant open-ended permission, although some, who apparently lack ample makura, will only allow harvesting on a one-time basis. In all cases, access to the forest requires neither a payment nor any sort of gift (fieldnotes 14/4/95, 15/4/95). The makwasha in Zimbabwe are all sacred forests owned by masabhuku and masadunhu responsible for propitiating the spirits in them. Part IV will elaborate upon these beliefs and upon the management of makwasha. The Mozambican bank of the Lucite presents quite a different mix of woodland. There is, on this side, considerably more woodland than in Vhimba where the population is more dense and people appear to farm more extensively. Yet, of the Mozambican woodland, a smaller proportion is makwasha than is the case in Zimbabwe. In absolute terms, Vhimba may have more acres of makwasha than a comparable area south of the Lucite. Mozambicans may have cultivated their makura because they appear to consider much less of it to be sacred. (I could only confirm the existence of two sacred groves in Mozambique of comparable size to the smallest of the sacred forests to be discussed in Part IV). Additionally, depopulation due to the recent war is likely to account for both the sizeable makura and the low population density in Mozambique. Aerial photographs show that, before the war in Mozambique, that area was much more thickly settled and less wooded.6 In the course of the war, many Mozambicans fled to Zimbabwe or deeper into Mozambique. Those who remained behind protected themselves from Renamo by hiding in the bush during the nights. These movements, including the strategy of part-time residence in the forest, greatly reduced the amount of land cultivated on the Mozambican side of the Lucite. Forest crept back over abandoned fields. Only with the security of peace, are returning residents and those who did not leave reopening and expanding fields. The makura on the Mozambique side are, therefore, likely to shrink, although they will undoubtedly continue to bulk larger than all forest in the Zimbabwe side of the study area. This discrepancy puts Zimbabwean residents at a certain disadvantage. Not only is much of their forest at the less usable climax stage, but also much of that climax forest is considered sacred. Many if not most residents refuse to collect resources from sacred forests. Part IV of this report will consider these forests and the cultural mechanisms that conserve them. From the point of view of resource use, the existence of these "protected areas" in Zimbabwe compels many residents to seek fuel and poles in Mozambique. Many women cross the Lucite to gather wood, and many men do so on a less frequent basis to cut poles. As in Zimbabwe, such people seek the permission of families living adjacent to makura. Even though these makura are 12 relatively abundant, Zimbabweans - who are considered, in some sense, as outsiders - must obtain permission each time they wish to collect. There is another, more widespread means by which Zimbabweans obtain Mozambican fuelwood. Many Mozambican women collect fuelwood for sale. Especially in the cold, dry season (chando) (perhaps because people use wood both for cooking and heating) women deliver wood directly to Zimbabwean households. They charge Z$2 or Z$3 for bundles of indeterminate size and weight (fieldnotes 14/3/95, 12/4/95). (Bundles which I purchased for Z$3 were said to be sufficient for three days of use by the average family.) This fuelwood trade underpins the strategy for development and sustainable use of the Mozambican makura suggested in Part V of the report. Fields and other commercial interests The fact that Mozambicans sell fuelwood to Zimbabweans indicates the economic disparity between the two sides of the Lucite. Residents on both banks agree that Zimbabwe is much richer. A road passes through the northern side of the study area. That side also has shops, a school, the Vhimba Primary School, and a grinding mill recently donated by SAFIRE. The nearest clinic, secondary school and bus stop are, however, at least 17 km distant. The Mozambican bank offers none of these amenities, although Mozambican children do attend the Vhimba school, and women grind maize at the grinding mill. Mozambicans' only store and grinding mill were destroyed during the war, and are not soon likely to be rebuilt. Zimbabweans take advantage of their more proximate - and legal - access to markets. Nearly all farmers sell some of their crops to buyers outside the valley. Lucite receives an average of between 1400 and 1600 mm of annual rainfall (among the highest of any agricultural area in Zimbabwe), rain falls in every season. Thus, smallholders are able to grow without irrigation bananas, pineapples, avocado-pears, oranges, cassava and sweet potatoes, not to mention maize, finger millet, and sorghum. Trucks from Bulawayo and other large towns ply the road on a daily basis, paying cash for fruit, especially for bananas. Since these operators pay very little, many smallholders have joined the Rusitu Valley Fruit Growers Association. The Association how has three groups in the study area, comprising 160 families.7 Members own this marketing cooperative (although most with whom I spoke did not fully understand this fact), while staff of the Agricultural and Rural Development Authority manage it, and the European Union has helped finance it. Most of the commercially successful farmers in Vhimba are members of the Association, although everyone sells some produce to private operators for immediate cash. The richest farmer in Vhimba has 13 solved the problem of late payments from the Association by using his own pick-up to market his crops personally. Zimbabweans consider this individual prosperous because they measure wealth in terms of income. According to virtually all the Zimbabweans I interviewed, rich people are those who can afford consumer goods and are able to pay school fees for all their children. By this standard, Zimbabweans in Lucite are well-off even compared to those in less remote, better serviced areas. The young and old women and men interviewed all agreed that Vhimba's fertility made them privileged. These economistic understandings of wealth may also be partly explained by the absence of cattle in Lucite. Historically, this section of the Lucite Valley was infested with tsetse fly, although the 8 Chimanimani upland was tsetse-free. Even if tsetse fly is now eradicated or eradicable, the valley's slopes are too steep for cattle grazing. People do keep goats, but livestock is not a store and symbol of capital as it can be in other parts of Southern Africa. (Grooms in Lucite pay lobola, or bridewealth, in cash, rather than in cattle.) Furthermore, people may understand wealth in commercial terms because of their experiences in migrant labour. Many men from both sides of the Lucite have worked for forestry plantations, notably Border Timbers and for nearby tea and coffee estates in Zimbabwe. Perhaps more so than other forms migrant farm labour, employment in this kind of high- value commodity production stimulates smallholders (while providing them with start-up capital) to grow their own cash crops. Mozambicans' comparative lack of commercial success cannot, therefore, derive from any shortage of commercial inclination. Nor do rainfall or soil conditions differ on opposite banks of the Lucite. Mozambicans' relative poverty is due to political factors: instability and economic uncertainty caused, first, by the recent war and, second, by the obstacle of the international border. During the insecurity of the war, Mozambicans were in no position to invest in tree crops or in long-maturing banana plants. Since flight frequently separated them from their fields, Mozambicans mostly planted small plots of annuals, chiefly maize and millet. At the conclusion of the war, some - many fewer than on the Zimbabwe side - did plant bananas, and these plants are now productive. Mozambican banana growers carry their fruit across the Lucite and sell, in a buyers' market, to the passing trucks. These farmers are not permitted to join the Rusitu Valley Fruit Growers Association. Already, they are engaged in low-level smuggling, and, unless border conventions change, authorities will not allow them to develop trade with Zimbabwe. The commercial potential of markets within Mozambique is virtually nil: the only goods people head-load between Lucite and the road are lightweight, expensive products, such as capulanas (headscarves) from Beira - certainly not bananas. The 14 border, thus, virtually seals off half of the Lucite Valley from its only source of income and commercial potential. Migration and demographv If one can speak of "Zimbabweans" and "Mozambicans" as different groups in terms of their economic and commercial linkages, in other respects, especially that of kinship, these communities are more closely wedded. Almost all families have some relatives, often immediate kin, across the border. Mozambicans who fled to Zimbabwe during the war married Zimbabweans.9 Some men have wives living on both sides of the Lucite. These individuals rotate between two or more homesteads and plant fields in both countries. Cross-border marriages occur frequently among members of the area's two independent Christian churches. The Jekenaishun and Zion Apostolic churches draw believers from both sides of the border and hold well-attended ceremonies in both countries. Finally, the border is "soft" because some Mozambicans were born in clinics in Zimbabwe and, hence, hold Zimbabwean birth certificates. Others obtained Zimbabwean working papers during employment on timber or tea plantations. For all intents and purposes, these people are dual nationals, and traditional leaders affirm their rights to immigrate to Zimbabwe. More recent migrants, however, are distinguishable from the more longstanding Zimbabwean population. Many Mozambicans, from both nearby and from deep within Mozambique, crossed the Lucite in the mid-to-late 1980s as the war intensified in Mossurize and Sussendenga Districts. Most of these families returned in the early 1990s and would have crossed paths with a second wave of migrants fleeing drought. These latter refugees, many of whom came to Zimbabwe in order to become eligible for relief supplies, have or will probably soon return.'1 Given the mobility of Lucite residents, it is extremely difficult to calculate either the size or the density of population. The two masabhuku and one sadunhu on the Zimbabwe side, that is, in Vhimba, claim to rule a total of 183 households (misha).11 My own exceedingly rough calculations based on the boundaries of territories of one sabhuku and one sadunhu indicate a household density of 12.8 and a population density of 128 per square km, quite a high figure. 12 As for the Mozambican side of the Lucite, in the absence of reliable information from either traditional leaders or other researchers, one can only say that the population is decidedly less dense than that of Vhimba. Political dynamics on the Zimbabwe side 15  The migrations and demographic changes discussed above are a consequence of and, at the same time, an influence upon local political conventions. Local authorities, mostly masabhuku, make decisions regarding the residence of Mozambicans in Zimbabwe and vice-versa. They adjudicate people's access to agricultural land and their rights to open fields in previously forested areas. Yet, over the course of this century, land allocators have carried out this responsibility in different fashions and with different practical consequences for the distribution of forests and fields. The role of the local political leaders in Vhimba has been shaped by two specific demographic movements and by state administration in general. The first involuntary migration took place as Rhodesian settlers grabbed large tracts of land in the 1890s (Palmer 1977: 262-63). This dispossession and the later establishment of a national park determined the extent of land now available to land allocators in Vhimba. The second series of migrations in the 1980s and 1990s brought to Vhimba people from Mozambique who needed fields - providing, in other words, occasions for political leaders to allocate land. Yet refugees would not have so affected local political relations had not British colonial governance politicised land allocation in certain ways. Vhimba's political hierarchy immediately before its incorporation into the British Empire probably comprised two levels: the chief or mambo (plural, madzimambo) and headmen, or masadunhu (singular, sadunhu). Both of these offices are hereditary, although in some cases succession is collateral (brother to brother), and in others is through primogeniture (to the first-born son). The present chief of the Zimbabwe side of the Lucite is a member of the Ngorima lineage. Formerly known as Sahodi, Ngorima is one a number of dynasties with the mwoyo (heart) totem (mutupo) that migrated from the Rozvi Empire in the late 17th century. They subjugated the existing dziva (pool) polity.1 3 As a result, chieftaincies and established the Sanga most of the chiefs in the Chipinge-Chimanimani area are mwoyo while the leading dziva chief, Musikavanhu, holds the ritual office of rainmaker. This pattern has been quite common in Bantu conquests: rulers of subjugated polities, who have a longstanding tie to the land and its spirits, are incorporated as ritual specialists in the new polity.' 4 (Part IV discusses an instance of this phenomenon still evident in the Lucite Valley.) By contrast, the Nguni conquest and occupation of the Lucite in the second half of the 19th century appears to have been much more superficial. Its most noticeable trace is the use of Nguni names for the preexisting totems - sithole or nkomo for mwoyo and mlambo or muyambo for dziva. These conquests and changes in chiefly politics appear to have preserved the most important aspects of the second political 16 tier. Masadunhu are considered to be owners and administrators of the land. One historian refers to makes a distinction between two kinds of masadunhu, vazvare or those related to the chief's clan vabatiwenyika, or those of the ruling clan of a previously conquered polity.'5 If this distinction still carries weight today, it affects the emphasis accorded different leadership roles and not the roles themselves. All the second-tier leaders in Vhimba are and have been responsible for land allocation, for propitiating ancestral spirits (-dira midzimu), and for collecting offering (makoto)for Chief Ngorima and for the rainmaker, Musikavanhu. Individuals may, of course, devote greater attention to one of these functions. As we shall see in Part IV, they have often used their power in the sphere or spheres that give them strategic advantages in the complex land politics of the Lucite Valley. British indirect rule transformed the roles of chiefs and masadunhu. Chiefs became government employees and were tasked with mobilising their people in support of state policies. At times, for example during a "traditionalist" phase of Rhodesian overrule in the 1960s and 1970s, the state tried to use chiefs as popular, supposedly anti-nationalist administrators. At other times, as in the more technical 1940s and 1950s, agricultural extensionists and conservationists sought to bypass traditional leaders and remake African farming practices. At present, the position of chiefs is as enigmatic as ever. As the lowest level of Zimbabwe's Ministry of Local Government, Rural and Urban development, chiefs receive monthly allowances and meet regularly with the District Administrator. Yet they play not play a substantial role in policy-making, especially not as regards the sensitive issues of land distribution and conservation.'6 Like chiefs, second-tier leaders have also partially entered the state bureaucracy. Early in the century, Rhodesian administrators instituted a hut tax and made some traditional leaders responsible for its collection. Colonial administrators of Vhimba gave record books to masadunhu, and they were renamed masabhuku (singular, sabhuku), literally "owners of the book." In many parts of Zimbabwe, masabhuku served under masadunhu, who continued to work closely with the chief. In Vhimba, however, masabhuku and masadunhu appear to occupy the same level of a two- tier political hierarchy. According to older residents, two of Vhimba's four masadunhu received books and became masabhuku. Few residents articulated a clear distinction between the two offices, except to say that the sabhuku is the one with the book. People further disagreed on the hierarchy of these offices. Some asserted that the sadunhu rules the sabhuku and his territory; others argued the exact reverse; and the masadunhu and masabhuku themselves invariably insisted on their own supremacy. From my observation, these leaders all allocate land, propitiate spirits and collect offerings for Musikavanhu. The collection of taxes, which responsibility distinguishes masabhku, is largely formal. 17 Taxes were suspended between 1972 and 1994, and many people have still not paid the new "development levy." Most residents seem to regard taxation as a symbolic - if powerfully symbolic - connection between the masabhuku and the state. In other spheres, by contrast, Vhimba's second-tier leaders have largely struggle against state policy to retain pre-colonial duties. The Native Land Husbandry Act of 1951, which marked the height of the technical development phase mentioned above, outlawed traditional land allocation. The Act charged government's, presumably more conservation-minded, agronomists with demarcating arable and grazing lands. Demarcation barely got off the ground in Vhimba. The area has no large livestock, and, more importantly, much of its cultivated land is on slopes steeper than 15%, the maximum gradient considered feasible for farming. Also, the density of the area's population and the presence of permanent tree crops left very little room for reshuffling fields (Alexander 1993: 53-54). Older residents did mention maraini, or the "lines" in which demarcation placed homesteads (fieldnotes 15/4/95). Yet, it appears that resettlement was minimal and/or ephemeral. Since Independence, the government of Zimbabwe has tried again to implement formal land use planning. In Vhimba, these efforts have met with similar resistance and achieved even less success. Chiefs, masabhuku and masadunhu have opposed land use planning, first, because its threatens their social positions, but also because it undermines their claims to alienated lands. The density of population in Vhimba and the (perceived) problems of conservation are a direct result of the seizure of the bulk of Chief Ngorima's land by Rhodesian settlers. Beginning in 1893, land alienation caused the first of the demographic shifts mentioned above: the removal and confinement of most of the Ngorima population to the narrow, crowded Rusitu Valley. Prior to British rule, Chief Ngorima's land had extended from the Lucite northwards almost to the present-day Chimanimani Village. Indeed, the Chimanimani Plateau contains many sacred sites, the most important of which are at a place called "Jiho" on the present Tilbury estate. 17 At the turn of the century, Rhodesian farmers occupied the plateau, substituting for it, the "tribal trust land" in the Rusitu and Haroni Valley's (see Map #2). Later amendments in 1925 converted the Haroni and lower Lucite Valleys to "native purchase area" for Africans under freehold 8 tenure." Faced with the Native Land Husbandry Act of 1951, Chief Ngorima and other local leaders demanded land redistribution, or the return of alienated land, prior to any consideration of land use planning. Neither objective was achieved, squatting increased, and a faction of the Ngorima family actively supported ZANU guerrillas during the Liberation war (Alexander 1993: 144-48). In the meantime, Chimanimani National Park was extended southwards in the late 1960s to cover the entire Haroni Valley. Chief Ngorima's people lost the narrow 18 arable strips of the Haroni and Chisengu Valleys. As explained in Part IV, the relevant sabhuku still claims these areas and is resisting the eviction of his people therefrom. These politics of land distribution and land use planning shaped the outcome of the second major demographic movement affecting Vhimba: the arrival of waves of Mozambican migrants and refugees in the 1980s and 1990s. Government's procedures for land use planning included a process of allocating land to migrants. Masadunhu and masabhuku sit on Village Development Committees (Vidcos) along with the (local government) district councillor and various civil servants. The "traditional leaders" serve to guard sacred areas from demarcation as arable land while, in theory, government agricultural extension agents assume the remaining responsibility for allocating land where technically suitable. Zimbabweans and Mozambicans with Zimbabwean identity documents were eligible for land in Vhimba under this procedure. Mozambicans without Zimbabwe papers were to be reported for transfer the Tongogara refugee camp. In the event, masabhuku and masadunhu followed none of these procedures. Even neglecting to consult Chief Ngorima, they allocated land where and to whom they wished. In the process of allocating land - and often deliberately - masabhuku and masadunhu have undermined both land use planning and, to a lesser extent, the distribution of land in Vhimba. The Vidcos never succeeded in regulating land use in Vhimba. I attended one meeting of the Mutoradiya Vidco whose main purpose was to discuss needs for services and infrastructure, a common residual role of Vidcos elsewhere in Zimbabwe (fieldnotes 18/4/95). Almost all informants accepted that masabhuku and masadunhu were legitimate land allocators, and few wished to see the Vidco assume any influence in this matter. By the same token, I found almost no evidence of self-allocation of land. Only two people opened or extended fields without authorization, and these people were fined and forced to abandon the field by the given sabhuku or sadunhu (fieldnotes 5/3/95, 14/3/95). If masabhuku and masadunhu have used land allocation to defeat land use planning, they have also used it to contest the delimitation of their "tribal trust land." One sabhuku, in particular, placed Mozambican families, as well as migrants from within Zimbabwe, just inside the official boundary of Chimanimani National Park. Further strategies of land allocation have challenged the territorial boundaries of neighboring leaders or asserted power by withholding sacred areas from allocation. 19 These multifaceted struggles have, to a large extent, determined the disposition of field and forest in Vhimba. It is likely that masabhuku and masadunhu will continue to control the use of land in Vhimba, notwithstanding the state's efforts. Part IV analyzes the implications of these land politics for the conservation of sacred forests and officially protected areas. 19 Political dynamics on the Mozambican side Politics on the southern bank of the Lucite have not centered on land issues to nearly the same extent as they have across the border, in Zimbabwe. Historically, Portuguese colonial administration did not attempt to transform either the use or the ownership of land in this area. Frelimo's programme of villagisation was short-lived in the extreme, and the civil war soon shifted the terms of rural politics away from land and towards more basic issues of security and survival. Also during the war, as in most of this century, demographic factors mitigated against conflict over land. People fled from Lucite, or northern Mossurize District, into Zimbabwe or towards government-controlled areas of Mozambique. Hence, Regulo Mafussi and other local leaders have not been preoccupied with land allocation and related boundary disputes, as have their Zimbabwean counterparts. With the conclusion of the war, communication between the Mozambican state and Lucite farmers has apparently not reawakened land debates. For this reason, the Mozambican side of the Lucite presents, in some ways, more favourable conditions for community-based forestry that does Vhimba, where such a programme is already planned. The pre-colonial history of northern Mossurize is essentially identical to that of Vhimba. The Sanga polity of dziva and mwoyo dynasties encompassed both areas. Mafussi, neighboring Gogoi, Musikavanhu (now in Zimbabwe) and other dziva chieftaincies had split from the Rozvi state in the first wave of 17th-century out-migration. As discussed above, Ngorima and the mwoyo chiefs followed close on their heels. For most of the 19th century, the mfecane, or migration from Zululand, brought the Sanga polity under Nguni control. The Gaza Nguni state exacted tribute from its subject peoples. Lucite paid its share while the state's capital was less than 30km to the south, in the Mossurize River Valley. When King Gungunyana moved south to Bilene in 1884, however, he retracted most of his tribute- collecting noblemen. In the long run, therefore, Nguni rule did not substantially alter the pre-existing political structure of the Lucite Valley.2 0 With the defeat of Gungunyana in 1895, Portugal gained sovereignty over Mossurize. Portugal and Britain had already demarcated their border in 1893. Lisbon promptly granted a concession for all of Manica and Sofala to the Mozambique Company. Unlike the British South Africa Company in Rhodesia, the Mozambique Company did not encourage European settlement and agriculture. In southern Manica, the Company established plantations only in the vicinity of present-day Sussundenga town. (Government authorities later created a colonato, or settler 20  colony, centered on Sussendenga town, in the early 1960s.) (Alexander 1994, Bannerman 1993) The Company also required smallholders around Dombe to grow cotton for sale. From the remaining territory, the Company extracted forced labour at the rate of four or six months per year for the cotton plantations of -Sussundenga or places further afield. Reaulos were responsible for delivering and annual quota of male workers. They found this job increasingly difficult, as Mozambicans evaded compulsory labour by leaving Mozambique. Even when a minimum wage law was passed in 1927, Lucite residents continued to avoid the requisite service. Instead, they sought higher-paid and voluntary employment outside the Colony of Mozambique, notably in the nearby Rhodesian timber and tea estates and most frequently in the Transvaal mines. 21 Out-migration was one of the factors that limited the politicisation of land-use in northern Mossurize. Whereas in Zimbabwe, evictions forced smallholders into increasingly crowded reserves, Portuguese interventions did not result in elevated population densities. In the absence of land alienation, pre- colonial settlement patterns largely persisted. Also, because emigration blunted the effect of natural population increase, Mossurize farmers were not forced to intensify cultivation, through terracing, for example, as were farmers in Ngorima Tribal Trust Land. Finally, Portuguese colonial policy did not address conservation problems, real or imagined. The Mozambique Company and the Colony's much attenuated civil service were simply much less interested than were the British in agricultural practices or rural development (Vail 1976) (perhaps explained in part by the fact that the Portuguese did not have to alleviate the effects of land alienation or advance the interests of a large settler farmer class). Thus, throughout present-day Manica Province, smallholders continue to cultivate streambanks and wetlands and to practice other farming methods long ago outlawed and suppressed in Zimbabwe (Bannerman n.d.). Given these differing demographic and political conditions, the role of local leaders on the southern bank of the Lucite has diverged noticeably from that of masabhuku and masadunhu on the Zimbabwe side. The Mozambique Company did give the existing second-tier leaders - probably then called masadunhu - books for tax collection. As in Zimbabwe, these leaders are now called masabhuku, although further from the border they go by the Portuguese label chefe de crupo. (I heard no mention of living masadunhu on the southern side of the Lucite.) Mafussi's masabhuku, like those of Ngorima, administer individual pieces of his lands. Yet, the territory of the sabhuku immediately along the southern bank of the Lucite is as large as the three territories of Vhimba combined. This sabhuku is able to administer his land by rotating residence between the homesteads of three wives (fieldnotes 14/3/95). Even so, he does not exercise nearly the same degree of authority over land as do his 21 Zimbabwean counterparts. In northern Mossurize, masabhuku's power of land allocation to newcomers is largely hypothetical. Former refugees and displaced people have mostly returned to their former field sites. Since no one has occupied those plots in the interimZ 2 , their arrival has not increased the brief local leaders either as land allocators or as dispute resolvers. As opposed to these roles, Mozambican residents of Lucite give much greater emphasis to function of masabhuku in propitiating spirits. Masabhuku give offerings in a number of sacred groves and collect offerings for the rain-making chiefs Musikavanhu in Zimbabwe and Dima in Mozambique.2 3 In sum, whereas British indirect rule alternately expanded and restricted the secular powers of local leaders, Portuguese rule preserved a more even balance between secular and religious responsibilities - if not a greater emphasis on the religious. In northern Mossurize, Frelimo attempted the first (and perhaps only) transformation of existing local leadership. Following Independence - the war not having affected Mossurize directly - Frelimo embarked on a programme of rural villagisation. Nucleated settlements, it was felt, would enable a more efficient provision of services and infrastructure. Likewise collective farming of a communal plot would exploit agricultural economies of scale and generate surplus production. On a more political level, villagisation aimed to weaken the powers of regulos and chefes. The leaders, whom Frelimo often accused of colonial collaboration, had no place in a system of formalised settlement and state-controlled land allocation. In Regulo Mafussi's area, a Frelimo party secretary persuaded people to open fields in one or more communal villages. Informants viewed collective agriculture as a form of theft perpetrated by the state. Through the aldeais, in one man's metaphor, Samora Machel was saying, "Mukadzi wenyu ndewangu" - "Your wife is mine" (fieldnotes 15/3/95). In any case, by the late 1980s, aldeias communais across southern Manica Province had become mere garrisons, whose principle purpose was to protect their inhabitants (Dirk Effler, pers. comm.). In fact, Renamo deliberately targeted the aldeias and any other government edifices. In the mid-to-late 1980s, Renamo gained effective control over Mossurize District as far south as Espungabera and Sussundenga District as far north as Sussundenga town. In Regulo Mafussi's area, Renamo military officers established a system of indirect rule and tribute collection relying upon regulos and masabhuku. Masabhuku instructed their people to contribute food and clothing to Renamo, and Renamo's vaiibha - essentially locally-recruited adjuncts - delivered these supplies to bases in the Sitatonga Mountains. 24 This administration offered a modicum of security. Nonetheless, many, if not most, residents still slept every night in the bush. These people seem to have been as afraid of Renamo attack as were of incursion by the Zimbabwe 22  National Army, which was known to resent residents as Renamo sympathizers. To the extent that residents did or did not sympathize with Renamo, this sentiment appears to have derived, first, from the minimal stability provided by Renamo rule and, second, from resentment of Frelimo's aldeias.25 Since the 1992 peace accords, Lucite residents have not fully recovered from the isolation and fear created by the war. The valley has had little or no recent contact with either Renamo or Frelimo. Residents could recall no recent visit by an administrator, government or NGO representative (fieldnotes 19/4/95). There is an authority at the Dacata post whom people believe to have contact with Regulo Mafussi. Other outsiders are suspected of more malicious intentions. In February 1995, a group of whites crossed over from Zimbabwe with unknown intentions. Many Lucite residents believe that these visitors were armed and intended to buy land and evict the inhabitants. Even if these rumours are wildly exaggerated, Mafussi's people clearly desire protection from the granting of concessions on their lands. Part V discusses a means to allay their fears by establishing a moratium on concessions in this area. IV. MANAGEMENT OF LUCITE'S PROTECTED AREAS Most outside agencies have been interested in the system of land allocation discussed above only as it affects certain protected forests. They have often identified agricultural land allocation and land allocators as a threat because masabhuku and masadunhu licence people to clear and plant wooded areas. Clearly, these leaders do accommodate in-migration and a growing local population by converting forests to fields. How could they do otherwise? Yet, they also act positively to preserve certain woodlands. Indeed, the masabhuku and masadunhu have an extremely strong record of conservation: they have withheld from cultivation sacred forests now additionally protected by Zimbabwe's Department of National Parks and Wildlife Management. In so doing, Lucite residents have contributed more to forest conservation than have the bulk of the other land managers in the Chimanimani area of Zimbabwe. After taking the plateau from Chief Ngorima's people, most Rhodesian settler farmers removed the existing trees. Tilbury, Welgelegan, Dunstan and others planted highly profitable timber forests of pine and gum. The parastatal Forestry Commission manages its Chisengu and Tarka Forest Lands in the same commercial fashion. Lucite residents, then, have protected some forests in the narrow belt of land that remained to them after colonial land allocation; governmental and private authorities, on the other hand, have almost entirely destroyed indigenous forests on the lands they alienated. 23  Lucite residents have conserved a number of woodlands they consider sacred. Nyakwawa and Chizire, in Zimbabwe, and Gorokota (fieldnotes 9/3/95) and probably other groves, in Mozambique are sacred forests (makwasha anoera) and are inhabited by ancestral spirits (midzimu or vadzimu, singular: mudzimu). The sabhuku or sadunhu, in whose territory a given forest lies, provides offerings to the resident spirits. In an annual or biannual ceremony, the sabhuku or sadunhu deposits, at a specific site in the forest (often by a pool), home-brewed beer and chicken or goats. The subjects of that sabhuku or sadunhu make a contribution (makoto or dundu) to the offering, most often by brewing beer. In order not to offend, the spirits residents also refrain from defecating in the forest (-ita tsvina). Many residents also believe that spirits forbid the collection of forest resources. To varying degrees, people avoid cutting poles or collecting fuelwood from core areas or from the whole extent of the makwasha. These prohibitions appear to offer some room for manoeuvre: they depend largely on the level of agreement over the boundary of the sacred part of the forest. Furthermore, it is not clear that the given sabhuku or sadunhu dictates explicitly what people may or may not collect from which parts of his forest. Yet, at a more fundamental level, his allocation or non-allocation of fields determines the extent of forest preserved. Nyakwawa and Chizire are officially considered Botanical Reserves. In neither case, do the sacred forests extend to the wider margins of areas administered by the Department of National Parks and Wildlife Management. In fact, the Lucite Valley contains or abuts four protected areas: the Rusitu Botanical Reserve, encompassing Nyakwawa; Haroni Botanical Reserve, encompassing Chizire; Chimanimani National Park; and, the only such area on the Mozambican side, the Maronga Forest Reserve. In all of these cases, the boundaries of protected areas are either in dispute or utterly in doubt. Residents had no knowledge of the Maronga Forest reserve. For years, the war prevented SPFFB Chimoio from conducting any activities at all in this area. As regards the remaining three protected areas, conservation politics in Lucite largely express debates over the location of boundaries. This section reviews those conflicts and assess their impact on plans for community-based natural resources management in Vhimba. Nyakwawa / Rusitu Botanical Reserve Nyakwawa is particularly sacred because of its place in the social history of Lucite. The ancestors which inhabit this forest belong, not to Chief Ngorima's people, but to an earlier group of occupants. According to one historian, Ngorima arrived in the early 17th century (Young 1970: 50). At that time, people called matsonnu or masabunga lived in the valley and on the 24 upland north of it.26 Their ancestral spirits resided in Nyakwawa, and they already treated it as a sacred forest. Upon conquering the area, Ngorima drove out the ruler of masabunga. Shortly thereafter, however, Ngorima recalled him from exile so that the customary rituals in Nyakwawa could continue. Ngorima made this person one of his masadunhu, governing a much smaller territory than he had before the conquest (fieldnotes 13/3/95). The present sadunhu for Nyakwawa and its environs descends from this vassal. This sadunhu considers as crucial leadership duties the preparation of offerings to the spirits of Nyakwawa and the general stewardship of this sacred forest. Many times he insisted forcefully, "Nyakwawa ngerangu," "Nyakwawa is mine!" He is the living person who knows best what the spirits do or do not permit. When pushed, he believes that the spirits forbid only defecation, hunting and the clearing of fields in the forests.27 In practice, many people play it safe: they do not cut poles or collect fuelwood within the boundaries of the sacred forest.28 (As explained already, the primary forest does not produce as many poles as does immature secondary forest.) The sadunhu does not allocate land within these boundaries, and with the exception of only one case in memory, people have not self-settled in Nyakwawa. This singular individual, a migrant from Mozambique, was fined by the sadunhu and forced to abandon his field in favour of land outside Nyakwawa (fieldnotes 14/3/95). Nyakwawa's conservation, therefore, depends on the delimitation of the sacred forest. People with whom I spoke agreed unanimously that a certain small stream marks the eastern border, and another dry streambed marks the western border. The sacred forest extend northward to the headwaters of these streams. To the south, the Lucite River bounds Nyakwawa, although mature forest does extend slightly into Mozambique. (Informants differed as to whether this Mozambican extension is sacred (fieldnotes 18/4/95, 19/4/95).) On its eastern and western edges, the forest abuts cultivated fields. In one or two places, fires from fields being cleared have escaped and burned forest undergrowth and even mature trees. The sadunhu, who allocated those fields, is displeased by this infringement, the result, he says, of unexpected whirlwinds (fieldnotes, 22/3/95). Government authorities consider people to have encroached on protected areas since long ago. In 1974, the Department of National Parks and Wildlife Management declared Nyakwawa and a much larger area of primary forest surrounding it to be the Rusitu Botanical Reserve. The Department felt that it was essential to protect the only woodland in Zimbabwe containing moist lowland and middle altitude forest as well as a transition zone. In the event, it gazetted 150 ha., a much larger area than that of the sacred forest. So wide were the boundaries of the reserve that they encircled fields already under cultivation in 25 1974 (turning those farmers overnight into "squatters"). Had people wished to observe these boundaries, they would have found compliance quite challenging. With the exception of the Lucite River, all the boundaries of the Rusitu Botanical Reserve are artificial surveyor's lines. In 1974, National Parks staff did mark those boundaries with cuts in trees. Yet, rapid regeneration quickly obliterated any trace of them until National Parks installed beacons in 1994. In the meantime, the sadunhu continued to allocate land, and people "encroached." This "squatting" demonstrates dramatically that local residents actually control the forest. The sadunhu himself exercises full authority over the extent to which Lucite smallholders clear the reserve, or any woodland in his territory, for fields. National Parks staff now realise that they cannot enforce their own restrictions against an unwilling community. They have abandoned the strongarm tactics of arrest and, on some occasions, of physical abuse that characterised their previous mode of engagement with Vhimba residents. A number of National Parks officials in Harare are seeking actively to cooperate with local leaders under the umbrella of CAMPFIRE. In anticipation of this change in attitude, the Southern Alliance for Indigenous Resources (SAFIRE) began operating in Vhimba in 1994. A participatory rural appraisal revealed that Vhimba's residents wished to obtain a grinding mill (SAFIRE 1994: 7). SAFIRE donated a grinding mill and stationed a field officer in Vhimba. With the encouragement of that field officer, residents formed the Vhimba Area Development Committee. The members of this committee meet regularly with each other and with the general community. The Committee is now considering a draft constitution and is debating whether to officially become a CAMPFIRE-related subcommittee of the Rural District Council. The Committee is eager to participate in the CAMPFIRE programme. SAFIRE took many of its members to visit a CAMPFIRE project in Mashonaland East Province, a community-run lodge. Vhimba residents who previously might have perceived tourists as a nuisance, now realise their potential contribution to Vhimba's development. The sadunhu of Nyakwawa, who is an ex-officio member of the Committee and himself visited Mutoko, supports tourism provided guests do not excrete in the sacred forest. Because the Rusitu and Haroni Botanical Reserves contain rare bird species, bird-watchers come in groups from as far away as South Africa. Under CAMPFIRE, the Vhimba Area Development Committee could collect gate receipts, hire out guides and guards and even run a tourist lodge or camp.29 Concessions sold by the Committee to bird safari operators in Chipinge or Johannesburg could bring in substantially more revenue. Although, the amounts could not nearly equal those earned in sport-hunting areas, Vhimba is a small producer community; so each household could receive noticeable income. This incentive should encourage 26 people to refrain, even more than they already do, from collecting resources in the primary forest. More importantly, the Committee can work with the sadunhu to enforce limits on forest cultivation. The Department of National Parks is, thus, ready to take a giant step forward in its relationship with the Vhimba community. According to staff of its Harare office, National Parks would like to contract officially with the Vhimba community - most likely with the Vhimba Area Development Committee - to perform essential functions for which it lacks capacity. Under his scenario, local residents would provide all services to tourists and enforce restrictions on resource extraction from the botanical reserves. The Chimanimani Rural District Council must, of course, approve entrepreneurial activities in communal lands, but they are likely to do so once they understand the intention of CAMPFIRE. The only unresolved issue outstanding is the old subject of the boundaries of the forest to be protected. National Parks will probably try to broker an arrangement in which the Vhimba Area Development Committee officially leases fields within the botanical reserves to the farmers cultivating them for a nominal Z$1 per year. National Parks is clearly most concerned that further fields not be allocated, and the community, especially the sadunhu, is in a position to enforce such a management regime. Ultimately, these proposals constitute not only a major development improvement for the Vhimba community; they also represent a major advance in CAMPFIRE. Within the boundaries eventually negotiated, the Rusitu and Haroni Botanical Reserves will be the first community-managed parks in Zimbabwe. Chizire / Haroni Botanical Reserve The CAMPFIRE package makes Haroni a junior partner to the Rusitu Botanical Reserve. It is striking that both National Parks and the Vhimba Area Development Committee consider the two forests subject to the same agreement. The Haroni Reserve has come under much greater pressures from both local and external land users. The politics of this forest are more complex and involve more parties. Outside agencies would be wise to examine the particular conditions of this woodland and the sacrifices that may be required to secure local cooperation in its management. The Haroni Botanical Reserve lies at the confluence of the Haroni and Mukurupini Rivers. Like the Rusitu Reserve, Haroni is bounded on two sides by artificial lines, but is much smaller than Rusitu, only 20 ha. National Parks now estimates that only 4 ha. remain. Yet, this fragment lies immediately across the Haroni River from the Mukurupini Forest in Chimanimani National Park. The Mukurupini Forest is huge in relation to Haroni - 170 27 ha. Both forests contain comparable moist lowland vegetation. (They both lack the middle-altitude forest and mid-slope transition zone particular to the Rusitu Botanical Reserve.) It is not clear why the Department of National Parks has devoted such attention to the Haroni relict when adjacent woodland abounds. Bird-watchers and other tourists prefer the Haroni to the Mukurupini Forest, but for reasons unrelated to conservation: there is no convenient ford to cross into Mukurupini, and many believe that the Rhodesian government minded this forest during the Zimbabwean war (a rumour disconfirmed by local residents as well as staff of Chimanimani National Park). Like Rusitu, the Haroni Reserve contains a sacred area, called Chizire. To date, this core has been conserved. Yet, the sabhuku responsible for Chizire does not appear to consider its care to be his top priority. As a member of the Jekenaishun independent Christian church, he does not make offerings or otherwise worship ancestral spirits. His younger brother, still a traditional believer, acts in the sabhuku's stead. Whereas the sadunhu for Nyakwawa asserts his power, to a large extent, by presiding over and conserving Nyakwawa - in other words, by not allocating land - the sabhuku for Chizire asserts his power precisely by allocating land. This is not to say that Chizire's sabhuku willfully allows its destruction. On the contrary, he recently fined a neighbor of Chizire who had expanded his fields into the sacred forest. That farmer was forced to abandon the unauthorised field (fieldnotes, 5/3/95). Nonetheless, this sabhuku devotes much attention and political capital to land issues and disputes not directly related to Chizire. At present, the most visceral controversy involves a banana plantation recently established by the Rural District Council. In 1992 Council allocated to itself 60 ha. in the Masonga neighborhood immediately adjacent to the Haroni Botanical Reserve. As the trustee of all communal land in Chimanimani District, the Rural District Council has to legal right to alienate land for businesses, development projects, infrastructure or other purposes. In this case, Council even obtained (if under false premises) the permission of the sabhuku discussed above. The plantation displaced the fields of seven families - for which they may have been compensated (fieldnotes 3/3/95). (Although still unclear, it appears that residents may have been promised or given a small amount for the loss of tree crops.) The second phase of the banana project, for which Council seeking funding from the European Union, would displace many more families. With regard to both phases, Council has not communicated effectively to outside agencies the fact that the plantation is a Council-owned project, generated funds for the Council's use in providing services throughout Chimanimani District.3 0 It is not a cooperative or community-owned project and provides no immediate benefit to Vhimba residents apart from the employment of 30 wage labourers (at Z$80 per month) 28 (fieldnotes 2/3/95). In fact, local people resent the Council for having taken land for what they perceive to be selfish purposes. In the words of aggrieved residents, "Vakaisa purazi mukati meReserve yedu" - "They put a commercial farm in the middle of our Communal Land;" and, more generally, "Ivo veCouncil vanotitsika" - "Those in the Council, they step on us" (fieldnotes, 5/3/95). Resentment boiled over in March 1995 as unknown suspects destroyed 200-250 banana plants. It is not clear how these contentious politics affect the Haroni Botanical Reserve. Since the Reserve's boundary is an artificial line, no one can be sure whether or how deeply the plantation has encroached. National Parks staff believe that the encroachment is only a matter of a few square meters (of already encroached and cultivated land). Local people and at least some Council personnel assume that the overlap in much greater. Indeed, one councillor argued that Council had done National Parks a service by interposing its plantation as a barrier to further encroachment by smallholders (fieldnotes 31/3/95)! This claim aside, the plantation has already complicated efforts to conserve the Haroni Botanical Reserve. Residents are questioning why they should observe boundaries that local government does not respect. In practical terms, the plantation reduced by 60 ha. the amount of land available to neighboring farmers. Expansion of the plantation will further tighten the squeeze on land in Vhimba. As the allocator of agricultural fields, the sabhuku is caught in the middle of these three-way debates between the community, the Council and National Parks. It is not clear how he will act or what powers he may have in this dispute. What is certain is that the establishment of the plantation has not made him more inclined to cooperate with the Department of National Parks in managing the Haroni Reserve. Chimanimani National Park The Haroni banana controversy is only the latest incident in an ongoing series of battles between this sabhuku and the Department of National Parks. Those disagreements date back to the expansion of the Chimanimani National Park in the late 1960s. At the time of its creation in the early 1950s, the Park boundary lay roughly 20 km. north of Vhimba.3 ' Its southern addition, however, cut into the sabhuku's territory in two areas: the Mukurupini Forest and the southern bank of the Chisengu River, known as the Mukwiratunhu neighborhood (see Map #2). The sabhuku considers this alienation of land illegitimate and duplicitous. Shortly after moving the Park boundary, Park staff evicted smallholders from Mukurupini. It appears that the sabhuku did not allocate land there again until the time of the Zimbabwean liberation war. Perhaps the weakening of government control or people's own desire to hide themselves in the bush prompted 29  resettlement in the Mukurupini Forest. In any case, the Department of National Parks of independent Zimbabwe evicted them again in roughly 1983. At present, the sabhuku has surrendered Mukurupini: he will allocate no more land in that part of Chimanimani National Park (fieldnotes, 22/3/95). As regards the Chisengu Valley, however, the sabhuku is sticking to his guns. This steep slope, known as the Mukwiratunhu neighborhood (from -kwira: to climb; and dunhu: territory), lies north of the artificial east-west line marking the border of the park. Even so, until the 1992-1993 drought in Mozambique, the sabhuku and National Parks appear to have observed a truce. Only two or three families lived in Mukwiratunhu, including a largely powerless and often unacknowledged sadunhu. Famine brought a number of Mozambican families across the Lucite to the sabhuku's territory (where store-bought and/or relief food was available). As refugees, they needed land, and, for a number of reasons, Mukwiratunhu was an ideal site for their resettlement. The area was much less densely populated than other parts of Vhimba; many Mozambicans wanted to live in areas like Mukwiratunhu that were relatively distant from the continuing insecurity of the border; finally, by allocating land in Mukwiratunhu, the sabhuku could demonstrate his authority over Mukwiratunhu's resident sadunhu. Aside from these selfish motives, the sabhuku and his subjects feel that Mozambicans fleeing drought and those who fled the war in the 1980s must be accommodated on humanitarian grounds. Vhimba's generosity has not only supported otherwise destitute people but it has saved the government of Zimbabwe and international agencies considerable expense. Moreover, refugees are clearly better off as self-sufficient farmers than as recipients of relief in camps. In the event, the sabhuku did give land to five families in Mukwiratunhu. In 1994, National Parks evicted some of these families, but six households, whose claims were apparently supported by then-Acting Chief Sanyero Ngorima, remain in Mukwiratunhu (fieldnotes, 19/3/95, 20/3/95). These kinds of turf battles will likely persist unless National Parks reaches a settlement with the sabhuku, the sadunhu and their people. The clearest and surest way to improve relations with these residents would be to degazette the southern bank of the Chisengu. Not only would such a step demonstrate National Parks's goodwill and sense of fairness; it would also ease the land pressure on the Haroni Botanical Reserve. If National Parks would miss the Chisengu strip of (relatively abundant) middle-altitude forest, it might consider gazetting adjoining indigenous forest on the Hayfield B estate. Such a trade - Mukwiratunhu for part of Hayfield B - would render practical as well as poetic justice. A land swap would shift some of the burden of conservation onto those large landowners able bear it with the least amount of hardship. The exchange would also return land originally taken by force (although not by 30  the present owners) to the community of its original proprietors. This approach to conservation, of course, lies entirely outside the powers or mandate of the Department of National Parks alone. Perhaps higher authorities will recognise that land redistribution can be a fair and feasible solution to some conservation problems. V. RECOMMENDED ACTIVITIES These recommendations concern the Mozambican side of the Lucite and relevant agencies active in Mozambique. The preceding section has suggested some slight modifications in the policy of the Zimbabwean Department of National Parks towards the Zimbabwe side of the Lucite. Overall, this report supports current moves leading towards community-managed botanical reserves in Vhimba. The recommendations below are intended to enable Mozambican residents, assisted by Mozambican organisations, to undertake some income-earning and management roles related to the Zimbabwean reserves. The first stage of activities - equivalent to "Phase 3" of the preparation cycle of the Transfrontier Project - will initiate this process. The second stage of activities can begin when the Transfrontier Project enters its implementation cycle. Further development in Lucite, in a third and fourth stage, will depend upon policy changes detailed below. In this connection, DNFFB and the World Bank could negotiate with the governments of Mozambique and Zimbabwe to obtain legal exemptions to certain border controls. Stage 1: Immediate Activities Activity #1: Form a Chimanimani collaborative group All the agencies involved in community-based natural resource management in the Chimanimani region on both sides of the border should hold a workshop or similar meeting to decide on priorities and on the ultimate composition of a bi-national collaborative group of organisations. From the Zimbabwe side, this group would include SAFIRE, Zimtrust, IUCN, WWF and possibly Christian Care. Among Mozambican organisations, SPFFB de Manica would serve as the core agency, assisted by DNFFB staff from Maputo. According to its needs, SPFFB could invite, to the workshop, other agencies working in Manica Province, such as, the Centro de Experimentacao Florestal (already linked to SPFFB through DNFFB), the Arquivo do Patrim6nio Cultural and/or GTZ. (Additional individuals whom I would suggest are indicated in the "officials Consulted" section.) This workshop should determine the long-term and short-term needs of SPFFB and other agencies in 31 terms of staff and training necessary to implement the Transfrontier Project. Ideally, this and all future bi-national meetings would take place in Portuguese. (All relevant documents should appear in Portuguese or in both Portuguese and English - starting with the translation of this report.) Alternatively, the collaborative group should engage the services of a professional simultaneous translator. Otherwise, meetings held primarily in English will prevent the full participation of SPFFB in the collaborative group and, hence, make implementation of the Chimanimani project impossible. Activity #2: Visit Requlo Mafussi Research related to this report determined that a large proportion - and probably virtually all - residents of the Mozambican side of the Lucite acknowledge Regulo Mafussi and his masabhuku as the leading authorities in issues of land and forest use. It is necessary, therefore, for the representatives of the Transfrontier Project to meet with Regulo Mafussi, explain the aims of the Project to him, and seek his permission to undertake further activities in this area. To this end, a two-person team could make a one-week visit to his home and surrounding communities. In addition to the aims just mentioned, the visit would obtain information necessary to justify the imposition of a concessions moratorium in Mafussi's area. The team would determine, first, the location of the boundaries of Regulo Mafussi's area32 and, second, the resources therein - particularly sacred forests - that warrant protection. Given this agenda, the team should contain one member familiar with the role of r6gulos in Mozambique, perhaps a current employee of the Chimoio office of the Arquivo do Patrim6nio Cultural. The second team member should be capable of conducting resource mapping exercises (a method of participatory rural appraisal). For this job, I would suggest a field officer of the Zimbabwean agency SAFIRE, perhaps the field officer presently working just across the Lucite, in Vhimba. The common language of this bi- national team would have to be Shona/Ndau, although the report, included in the next activity, would be written in Portuguese. The team would report to the Manica SPFFB. Activity #3: Propose a concessions ban Staff in the Manica office of SPFFB could spend the week following the visit to Regulo Mafussi converting the team's findings into a proposal for a ban on concessions in Mafussi's area. The proposal should argue convincingly that the natural and cultural 32 and, hence, make implementation of the Chimanimani project impossible. Activity #2: Visit Requlo Mafussi Research related to this report determined that a large proportion - and probably virtually all - residents of the Mozambican side of the Lucite acknowledge Regulo Mafussi and his masabhuku as the leading authorities in issues of land and forest use. It is necessary, therefore, for the representatives of the Transfrontier Project to meet with Regulo Mafussi, explain the aims of the Project to him, and seek his permission to undertake further activities in this area. To this end, a two-person team could make a one-week visit to his home and surrounding communities. In addition to the aims just mentioned, the visit would obtain information necessary to justify the imposition of a concessions moratorium in Mafussi's area. The team would determine, first, the location of the boundaries of Regulo Mafussi's area32 and, second, the resources therein - particularly sacred forests - that warrant protection. Given this agenda, the team should contain one member familiar with the role of r6gulos in Mozambique, perhaps a current employee of the Chimoio office of the Arquivo do Patrim6nio Cultural. The second team member should be capable of conducting resource mapping exercises (a method of participatory rural appraisal). For this job, I would suggest a field officer of the Zimbabwean agency SAFIRE, perhaps the field officer presently working just across the Lucite, in Vhimba. The common language of this bi- national team would have to be Shona/Ndau, although the report, included in the next activity, would be written in Portuguese. The team would report to the Manica SPFFB. Activity #3: Propose a concessions ban Staff in the Manica office of SPFFB could spend the week following the visit to Regulo Mafussi converting the team's findings into a proposal for a ban on concessions in Mafussi's area. The proposal should argue convincingly that the natural and cultural endowments of this area merit protection from timber and other enterprises. The floral and faunal species found on the Zimbabwe side (and documented by references in the Bibliography and in SAFIRE's possession) are doubtless also present across the Lucite, although forests differ in structure. SPFFB's proposal can indicate the rare birds in Lucite as well as the existence of sacred forests and groves, sites that are both natural and cultural resources. 33 Stage 2 - Contingent upon Funding for Implementation of the Transfrontier Proiect Activity #4: Obtain necessary staff The workshop (Activity #1) will have determined the hiring needs of SPFFB and other agencies in the Chimanimani collaborative group. I would suggest that the organisations involved locate a total of two fieldworkers for the Mozambican side of the Chimanimani region. The workshop will have determined to which agency will supervise these fieldworkers. I would suggest that they report directly to SPFFB Chimoio, although they might have been seconded by other agencies. These individuals must speak Ndau/Shona and Portuguese. They should be capable of facilitating community meetings and motivating people towards development and conservation activities. The fieldworkers need not possess an academic or high- level professional training. It is much more important that they exhibit the social skills and personal dynamism necessary to win smallholders' trust and encourage their ideas. Based in Dombe and Rotanda/Mavita, the two fieldworkers could reach communities by motorbike. The Lucite Valley will actually be more accessible to the northern fieldworker, in Rotanda or Mavita. Crossing the border at Cashel, this person can reach Vhimba by road and ford the Lucite on foot (provided appropriate permissions are obtained for informal border-jumping). Alternatively, he or she could travel to Regulo Mafussi's home by driving through the border at Tamandayi, in which case most vehicles first obtain an immigration stamp at the official Espungabera border post. Time permitting, this individual will spend 1-2 weeks per month living in Lucite (he or she may wish to build a house or hut) and will, therefore, establish a lasting relationship between Lucite and SPFFB. Activity #5: Encourage sustainable fuelwood harvesting The fieldworker responsible for Lucite could support residents in any number of development and conservation activities they might wish to undertake. Mozambican women might opt to formalise their present sale of fuelwood by establishing a market stand in Zimbabwe. With income thus received, they might wish to form a savings club or some kind of income-generation project. If and when, Rusitu and Haroni Botanical Reserves in Zimbabwe become community-managed parks, Zimbabweans will probably wish to protect their woodland more than they already do. The demand for Mozambican poles and fuelwood - and the money Zimbabweans have to purchase 34  them - is likely to increase. If they choose to, Mozambican women could reap indirect benefits from CAMPFIRE by controlling access to their secondary forests and then harvesting them commercially. In this connection, one or more researchers might provide shorter-term services when needed. A social forester from CEF, for example, could investigate the productivity of immature secondary forest. He or she might determine the extent to which Mozambican residents harvest poles and fuel from this woodland. He or she might then train local women and men to establish levels of sustainable use (a "management plan") and to recognise signs of over-harvesting. Stage 3 - Contingent upon an Exemption Allowinq Cross- Border Tourism Activity #6: Help form a local management committee The Mozambican side of the Lucite will receive further income once the governments of Mozambique and Zimbabwe permit cross-border tourism. In this event, the fieldworker might be able to stimulate the formation of a "development committee" equivalent to the Vhimba Area Development Committee. The fieldworker might even organise a training visit to Regulo Mafussi's area by the Vhimba Area Development Committee (paying fees to these "grassroots consultants"). Through such a counterpart committee, Mozambican residents could cooperate formally with CAMPFIRE activities in Zimbabwe. Bird-watchers would visit the small section of Mozambican primary forest alongside the Rusitu Botanical Reserve. Birders with whom I spoke were eager to explore this forest patch and others on the Mozambican side of the Lucite (fieldnotes 6/3/95). The Mozambican development committee could, therefore negotiate its own concessions and gate fees (on terms similar to those explained in the "Nyakwawa" section of Part IV), either directly with tourists or with the Zimbabwean committee. Stage 4: Contingent upOn an Exemption Allowinq CroBerder Commerce Activity #7: Assist agricultural marketing cooperatives Perhaps the success of this minimal cross-border commerce would persuade the governments of Zimbabwe and Mozambique to allow trade of locally-produced goods across the Lucite. Specifically, the Mozambican bank would develop substantially - indeed this may be the only way for it to do so - if its residents were 35 permitted to join the Rusitu Valley Fruit Growers Association or to form their own agricultural marketing cooperative. The Lucite fieldworker could encourage farmers to join and sell their produce to a marketing coop. To this end, DNFFB and the World Bank could encourage government authorities to create formal procedures, even involving customs duties, for Lucite and other border communities to export their products. 36  NOTES 1. The Land Tenure Centre of the University of Wisconsin, USA, in cooperation with the Ministry of Agriculture, leads the field in research on this topic. See, for example, Myers (1994); Myers, Eliseu and Nhachungue (1993); and West and Myers (1994). 2. For an analysis of conflicts between local residents and a sport-hunting firm in the Coutada 16 reserve of Gaza Province, Mozambique, see Hughes and Barbossa (1995). 3. For example, Matowanyika (1991). 4. I fell victim to the frontier-related restrictions discussed above. There is no formal procedure in either Zimbabwe or Mozambique to grant either citizens or non- citizens official permission to cross borders where there are no border checkpoints. Nonetheless, the Zimbabwean police in Chimanimani granted me permission to cross, and the Provincial Department of Agriculture for Manica provided me with a "Guia de Marcha" authorizing my presence in the Lucite area of Mozambique. These documents, accompanied, of course, by my passport and Mozambican multiple entry visa, satisfied all parties except the village head on the Mozambique side of the Lucite. This person insisted on referring the matter to Chief Mafusi and to the nearest administrative post, Dacata. Much later, I learned from a messenger of the chief that the matter was still unsettled. Dacata and the nearest road in Mozambique are at least a full day's walk from the Lucite. Unless a track motorbike can cross this terrain (worth a try), fieldworkers and researchers who wish to transport equipment and supplies to this area will have to do so through Zimbabwe, as do local residents. I would urge that DNFFB and the World Bank seek special exemptions for any staff that may work in Lucite. 5. Mavi (n.d.: 5-10), on the other hand, found a number of species with edible leaves. Yet, his report does not clarify whether leaves might be medicinals or undesirable, slightly toxic "famine foods." 6. I consulted the 1:25,000 scale aerial photographs of the 1963, 1975, 1981 and 1987 Chipinge series from the Surveyor General, Harare. 7. Fieldnotes 12/4/95. Interestingly, the boundaries of these groups were established along those of the territories of masabhuku and masadunhu. They, therefore, cut across the boundaries of the Village Development Committees established in the early 1980s (see the section on "Political dynamics on the Zimbabwe side"). 37 8. Letters from Division of Entomology, Salisbury to Secretary, Department of Agriculture, 12/11/1931 and 14/7/1943 (In NAZ public record S 3024/9/1 Veterinary Department file on trypanosomiasis: Melsetter and Chipinga, 1923-1950); and Relatorio da Inspeccao Ordinaria a Circunscricao de Mossurize, 1943-44, p. 3 (in AHM collection of Inspeccao dos Servicos Administrativos e dos Negocios Indigenas Caixa 17-39). Swynnerton (1921) provides information on tsetse-prevention practices used by the Gaza Nguni during their occupation of the nearby Mossurize Valley from 1865 to 1895. 9. According to residents, Zimbabwean men more often married Mozambican women than vice-versa. Some people explained this discrepancy by the fact that Zimbabwean men, who mostly have attended school, make more attractive husbands than the largely uneducated Mozambican men (fieldnotes 4/3/95). If the war did bring Mozambican wives to Zimbabwean men, it may partly account for the high incidence of male polygamy (polygyny) reported by Hobane (1993: 7). 10. Hobane (1993: 4) found that, of a sample of 28 households established between 1980 and 1993, 8 came from Mozambique, 9 from elsewhere in Zimbabwe, and only 11 actually originated from within the Zimbabwean side of the Lucite. 11. One sabhuku oversees 60 households (a figure further substantiated by a map of the same area showing 47 households, drawn by an informant with less complete information that the sabhuku) while the single sadunhu oversees 68. The remaining sabhuku gave me a figure of 164 resident "parents" (vabereki) (fieldnotes 21/3/95, 22/3/95, 24/3/95). If one takes Hobane's (1993: 7) calculation of roughly 2 wives for every husband, 164 parents should be equivalent to approximately 55 households. The total number of households, therefore, reaches 183, admittedly very much of a "ballpark figure." 12. On the 1: 50,000-scale Zimbabwe Surveyor General's map, I was able to locate the boundaries of the aggregate territory ruled by one sabhuku (with 68 households) and of the single sadunhu (with 60 households). As the territory covers (very approximately) 10 square km. on the map, there should be 12.8 households per square km. 13. Rennie (1973, 1977) and Bannerman (1981) provide the most comprehensive accounts of this history. Supporting evidence may be found in Young (1970), Meredith (1976), and, in the National Archives of Zimbabwe: transcript of interview with Mr. Herbert Dzvairo, 30 January 1979 (oral history file AOH/47); and delineation reports for Melsetter Distriction (public records file S 2929/1/6). 38 14. See, for example, Monica Wilson (1979). Rennie (1977) offers hypotheses as regards shifts in the secular and religious valences of the Musikavanhu chieftaincy. 15. Rennie (1973: 75). In Lucite, I found only one person who mentioned vazvare (fieldnotes 22/3/95). No one spoke of vabatiwenyika. 16. Alexander (1993) provides novel insights on the ambiguous, shifting roles of traditional leaders from 1940 to 1990. She focusses partly on Chimanimani District. 17. See Young (1970). At the installation ceremony of the present Ngorima chief, his uncle made a speech in which he asked for the return of the most important sacred sites of the Ngorima lands (written text of the speech by Phineas Philip Ngorima, 31 March 1995). I quote at length: Tinokumbirawo Cde. Minister, kubudikidza nge- vanokumirirai kumatunhu ano NZVIMBO dzedu dzinokosha (Tribal Monuments) dzakadai nge 1. PaRupango - pana makuwa mashanu emadzibaba edu. 2. Makwasha ari muJiho - ane makuwa manomwe madzi- baba edu nevakunda vedu. 3. Ruvhayo 4. Chingwekwe 5. Gwasha reMarirwe ne Hode 6. Pabuwe rehwiro muNdakopa 7. Gwasha rekwa Nyakwaa kwa Vimba ... nedzimwe dzakadaro. Tingachengete idzi nzvimbo, nyika yedu inozogara yakanaka nekuramba yakapuwa donhodzo ndiMusiki. We ask you, Comrade Minister [Minister of Lands Kumbirai Kangai, the guest of honour], through those who represent you in these lands, for our important places (Tribal Monuments), such as: 1. PaRupango [on the Tilbury estate] - there are graves of our ancestors. 2. Groves in the Jiho [on Tilbury] - there are seven graves of our ancestors and of our chief's rela- tives. 3. Ruvhayo [grassland east of the Rusitu Mission]. 4. Chingwekwe [a hill close to the Rusitu Mission]. 5. Forest of Marirwe and Hode [probably at the con- fluence of the Nyamarirwe and Nyahode Rivers]. 6. The settlement by the graves of Ndakopa [between the Hlabiso and Kopa Business Centres]. 7. Nykwaa Forest in Vimba [discussed at length in Part IV] ... and similar important places. 39  We can care for these places so that our land can stay good and be given blessings from the Creator.(emphasis in original; for locations, fieldnotes 12/4/95, 13/4/95) The Jiho graves are the only one of these sites officially recognised by the state, under the 1936 Monuments and Relics Act (fieldnotes 11/3/95). The owner of the Tilbury estate permitted chief Ngorima and his immediate family to live close to Jiho from 1907 to 1956 (Young 1970: 58-59). Rituals are still permitted and regularly observed at the Jiho. 18. See NAZ, Chief Native Commissioner File N 3/24 on "Native Reserves"; Government of Rhodesia, Report of the Widlife Commission, 1969, p. 50; Palmer (1977: 262-63); Goodier and Phipps (1962); and various Land Apportionment maps of the Rhodesia Surveyor General 1909-1970 (available in the "La Rhodesia" map series at the National Archives of Zimbabwe). 19. See Wilson (1986) for a description of such processes involving Angolan migrants in western Zambia. I did not find evidence of the kind of exploitation of refugees by more longstanding residents noted in McGregor (1995). 20. Liesegang (1967) provides an excellent description of the society and politics of the Gaza Nguni state while Omer- Cooper (1966) offers a more general overview of the mfecane. 21. Gogoyo station annual reports for the years 1921, 1927 and 1928 (NAZ United Church Board for World Ministries, Rhodesia Mission files UN 3/20/1/16/5, UN 3/20/1/22/7, UN 3/20/1/23/13). Portuguese reports drew a clear connection between forced labour and the "exode" of "emigraca'o clandestina" ("Relat orio da Inspec9ao Ordinaria a CircunscriVao de Mossurize" for 1943-1944 and 1946 - AHM Inspeccao dos Servicos Administrativos e dos Negocios Indigenas Caixa 17-39). 22. As opposed to circumstances in some other parts of post- war Mozambique (see West and Myers 1994). 23. Fieldnotes 15/3/95, 13/4/95; Distrito de Mossurize (n.d.: 12) 24. Fieldnotes 15/3/95, 23/3/95; Mozambique Agricultural Rural Reconstruction Programme (1993: 33-34). 25. I did not encounter the kind of visceral anti-government feelings Geffray (1990) documents. In part, Geffray's informants may have been incensed by the death of one of their reaulos at the hands of Frelimo soldiers. Residents with whom I spoke reported no such event and frequently associated 40 Renamo with the horror of "kutema," or cutting (with a machete). 26. Matsonnu could be a derogatory name. Masabunga literally means "owners of Bunga," that is, Mount Bunga, now in the nothern corner of Tarka Forest Land. According to Young (1970), these people, while relatively short in stature, were not San bushmen. Like those who conquered them, masabunga were agriculturalists. Nvakwawa is actually the Sena word for chief, but this derivation does not mean that masabunpa were Sena. The Sena could have influenced them to adopt the term. 27. According to residents, a bird-watcher who fell from a tree or ledge and died recently in Nyakwawa, had either defecated or attempted to capture a bird or both (fieldnotes 4/3/95). 28. This is a rather one-dimensional approach to Shona religion. Space does not permit, for religion, the more holistic treatment here attempted in the realm of politics. The same caution applies to both subjects: social scientists concerned with natural resources management risk distorting the full meaning of these aspects of culture when they describe them as bodies of rules and regulations. 29. As a sacred-forest-cum-park, the site could be marketed as a cultural, as well as a natural, attraction. For interested tourists, and for a fee, an already-active local dance troupe could perform the mutsvongoyo, a striking Ndau dance. 30. Council is forced to raise funds through projects in the communal lands due to the unusual demographic structure of Chimanimani District, Zimbabwe. Chimanimani is the only district without a sizeable town. Officially a "village," rather than a "town," the settlement of Chimanimani contains very few residents who pay rates for electricity and water services. Ironically, the rural district council has very few other means of raising revenue from a patently rich area; it receives very little, if anything, from the timber businesses in the district. 31. See Surveyor General's land apportionment maps, 1909- 1970, in the NAZ "La. Rhodesia" map series. 32. On the question of Mafusi's boundaries, I received ambiguous information: his territory is bounded on the north and east by the Lucite, on the west by the Zimbabwean border, and on the south by either the Puizice or the Chicambue River or by some line connecting the two (fieldnotes 20/3/95, 15/4/95, 18/4/95). 41  BIBLIOGRAPHY A. Non-Archival sources Alexander, Jocelyn. (1993) "The state, agrarian policy and rural politics in Zimbabwe: case studies of Insiza and Chimanimani Districts, 1940-1990." D. Phil. Dissertation, Oxford University. -- (1994)"Terra e autoridade politica no p6s-guerra em Mogambique: o caso da Provincia de Manica.n Arguivo 16: 5- 94. Arquivo do Patrim6nio Cultural. (1993) "A organizarao politica Ndau." Recortes 5:25-30. Bannerman, J.H. (1981) "Helgweni: the history of the Hlengwe of the lower Save and Lundi Rivers, from the late eighteenth to the mid-twentieth century." Zimbabwean History 12: 1-45. -- (1993) "Land tenure in central Mozambique: past and present." Chimoio: Mozambique Agricultural Rural Reconstruction Programme, mimeo. -- (n.d.) "Farming systems in relation to conservation: Manica Province." Chimoio: Mozambique Agricultural Rural Reconstruction Programme, mimeo. Broadley, Donald G. (1974) wThe Lusitu Forest: a new protected area." Wild Rhodesia 3: 17. Child, B. and J.H. Peterson. (1991) "CAMPFIRE in rural development: the Beitbridge experience." Harare: Centre for Applied Social Sciences, University of Zimbabwe, Occasional Paper 1/91. Chipinge Wildlife Society of Zimbabwe. (1992) "Report on the Haroni Rusitu Botanic Reserves 91-92." Christian Care. (1993-1995) "Mozambican Repatriation Reports." Mutare: Christian Care. A series of status reports by Shirley DeWolf, sometimes with Timothy Kondo. Crook, A.O. (1956) "A preliminary vegetation map of the Melsetter Intensive Conservation Area, Southern Rhodesia." The Rhodesia Agricultural Journal 53(1): 1-25. de Vletter, Rod. (1995) "The GEF tranfrontier conservation areas and institution strengthening project: review to date." draft. 42 -- (1995) "Integrated rural development strategy plan for Manica Province - agriculture." Chimoio: Mozambique Agricultural Rural Reconstruction Programme, mimeo. Murombedzi, James. (1991) "Decentralising common property resources management: a case study of the Nyaminyami District Council of Zimbabwe's wildlife managment programme." London: International Institute for Environment and Development, Issues Paper No. 30. Murphree, Marshall. (1991) "Communities as institutions for resource management." Occasional Paper of the Centre for Applied Social Sciences, University of Zimbabwe, Harare. Myers, Gregory W. (1994) "Competitive rights, competitive claims: land access in post-war Mozambique." Paper published by the Land Tenure Center, University of Wisconsin-Madison, USA. Myers, Gregory, Julieta Eliseu and Erasmo Nhachungue. (1993) "Security and conflict in Mozambique: case studies of land access in the post-war period." Paper published by the Ministry of Agriculture, Maputo and the Land Tenure Center, University of Wisconsin-Madison, USA. Omer-Cooper, J.D. (1966) The Zulu Aftermath: A Nineteenth Century Revolution in Bantu Africa. London: Longman. Palmer, Robin. (1977) Land and Racial Domination in Rhodesia. London: Heinemann. Pereira, Carlos Raposo and Rui Baltazar dos Santos Alves. (1994) Reflexoes sobre o Regime Juridico da Terra. Maputo: Helvetas. Phipps, J.B. and R. Goodier. (1962) "A preliminary account of the plant ecology of the Chimanimani Mountains." The Journal of Ecology 50(2): 291-319. Rennie, John K. (1973) "Christianity, colonialism and the origins of nationalism among the Ndau of Southern Rhodesia." Ph.D. dissertation, Northwestern University. -- (1977) "From Zimbabwe do a colonial chieftaincy: four transformations of the Musikavanhu territorial cult in Rhodesia." In J.M. Schoffeleers, ed. Guardians of the Land: Essays on Central African Territorial Cults. Gweru: Mambo. Pp. 257-85. Rhodesian Schools Exploration Society. (1966) "Chironi expedition (Chimanimani area), 1965/1966." mimeo. 44 Distrito de Mossurize (n.d. [early 1980s]) "Povoacao guerreira: monographia do Distrito de Mossurize." unpublished mimeo. Dutton, T.P. and E.A.R. Dutton. (1973) "Reconhecimento preliminar das Montanhas de Chimanimani e zonas adjacentes com vista a criacbo duma &rea de conservagao." Lourengo Marques: Direccao dos Servigos de Veterinaria, mimeo. Environment and Development Group. (1994) "GEF transfrontier conservation areas and institution strengthening project." Maputo: DNFFB, mimeo. Fuelwood Crisis Consortium. (1993) "A needs assessment study of the communities adjacent to the Haroni and Rusitu Botanical Reserves, Chimanimani District, Zimbabwe." unpublished report. Geffray, Christian. (1990) La cause des armes au Mozambique: Anthropolocie d'une querre civile. Paris: Karthala. Goodier, R. and J.B. Phipps. (1962) "A vegetation map of the Chimanimani National Park." Kirkia 3: 2-7, map. Hobane, P.A. (1993) "A socio-economic report on Haroni and Rusitu Reserves." unpublished paper. Hughes, David M. and Filomena Barbossa. (1995) "The socio- economic impact of the Coutada 16 hunting concession on resident communities." Maputo: the World Bank, unpublished report. Liesegang, Gerhard Julius. (1967) "Beitr'age zu'r Geschichte des Reiches der Gaza Nguni im siudliche MoTambique." Ph.D. dissertation, Cologne University. Mavi, Steven. (n.d.) "Report on the use of plants in the Haroni/Lusitu Valley." unpublished draft. McGregor, JoAnn. (forthcoming) "People without fathers: Mozambicans in Swaziland." Journal of Southern African Studies. Meredith, L.C. (1976) "Melsetter District - history of native tribes and chiefs." Native Affairs Department Annual 16: 338-44. Mozambique Agricultural Rural Reconstruction Programme. (1993) "Mossurize District re-integration, resettlement and reconstruction project." Chimoio: Mozambique Agricultural Rural Reconstruction Programme, mimeo. 43  Robertson, F. (1986) "A study of the conservation status of botanical reserves in Zimbabwe." The Zimbabwe Science News 20(7/8): 102-6. SAFIRE. (1994) "A particpatory rural appraisal exercise in Mutemanerangu Village, Vhimba Ward, Chimanimani District, Zimbabwe." -- (1995) "A Project proposal for the development of water resources in Vhimba communal lands, Chimanimani District, Zimbabwe." Swynnerton, C.F.M. (1921) "An examination of the tsetse problem in North Mossurise, Portuguese East Africa." Bulletin of Entomological Research 11: 315-85. Timberlake, Jonathan. (1992) "The status and conservation of the Haroni, Rusitu and Mukurupini forests." unpublished draft. Vail, Leroy. (1976) "Mozambique's chartered companies: the rule of the feeble." Journal of African History 17(3): 389-416. West, Harry and Gregory Myers. (1994) "A piece of land or a land of peace?: land access and tenure security for smallholders in post-settlment Mozambique." Journal of Southern African Studies (forthcoming). Wilson, K.B. (1986) "The integration of Angolan refugees in western and north western Zambia." unpublished draft. Wilson, Monica. (1979) "Strangers in Africa: reflections on Nykyusa, Nguni and Sotho evidence." In William A. Shack and Elliot P. Skinner, eds. Strangers in African Societies. Berkeley: University of California Press. Young, John. (1970) "The legendary history of the Hodi and Ngorima chiefs." Native Affairs Department Annual 12(2): 49- 60. B. Archival Sources 1. Arguivo Hist6rico de Mocambique (AHM), Maputo Branquino, A. Melo. (1967) "Prospeccao das forcas tradicionais - Manica e Sofala." Lourengo Marques: Servigos de Centralizagao de Coordenacao de Informagoes. Dirrecgao Provincial dos Servigos de Planeamento e Integracao Econ6mica da Provincia de Mogambique. (1970) "Divisao administrativa de Mogambique por regedorias." Lourenco Marques: government printers. 45  OFFICIALS CONSULTED Agritex, Chimanimani, Zimbabwe Mr. Ushe Agritex, Kopa Clever Mundeta, Extension Worker ARPAC, Maputo Renato Matusse, Director Chimanimani Tourist Association, Chimanimani, Zimbabwe Jane High, Chair *Christian Care, Mutare Shirley DeWolf, National Refugee Programme Coordinator Department of National Parks and Wildlife Management, Harare George Pangeti, Deputy Director Cathy Rogers, CAMPFIRE unit Department of National Parks and Wildlife Management, Chimanimani, Zimbabwe Gilbert Moyo, Sargeant-Major District Administrator's Office, Chimanimani, Zimbabwe Mr. Dhlawayo, Assistant District Administrator *DNFFB, Maputo Augusto Urbano Correia Judy Oglethorpe, Wildlife Planning Advisor Alfonso Madope Michael Murphree Bartolomeu Soto, Director of Wildlife Department Ford Foundation, Harare Ken Wilson, Programme Officer Forestry Commission Harare Saiti Makuku, Research Division Frank Matose, Research Division IUCN, Harare Simon Anstey Joseph Z.Z. Matowanyika Land Tenure Center, Maputo and Madison, Wisconsin, USA Gregory Myers * Suggested participant in the first workshop of the Chimanimani collaborative group (see recommended activity #1 above.) 47  Governo Geral, Processos, Distrito de Manice e Sofala. Cards of "regedores" from 1964. Pasta 900. Inspeccao dos Servicos Administrativos e dos Neg6cios Indigenas. Various reports of inspections in Manica and Sofala, 1943- 1944, 1946 and 1954-57, Caixas 17°-39 and 12Q-41. 2. Arquivo do Patrimonio Cultural (ARPAC), Maputo Manica files of interview and oral history transcripts collected in 1982 as part of the Campanha de Preservacao e Valorizacao Cultural of the Servigo Nacional de Museus e Antiguidades. 3. National Archives of Zimbabwe (NAZ), Harare Delineation Reports, Tribal Trust Lands in the Melsetter District, 1965. Public records file S 2929/1/6 Government of Rhodesia. (1969) "Report of the Wildlife Commission." Salisbury: Government Printer. Interview with Mr. Herbert Dzvairo (born, 1905) on 30 January 1979. Oral history file AOH/47. Land apportionment maps, intermittently 1909-1970. Map file "La. Rhodesia." Native Commissioner, Melsetter. Outgoing correspondence, 1895- 1897. Public Records files NUE 2/1/1 and NUE 2/1/2. United Church Board for World Ministries, Rhodesia Mission. Gogoyo Station annual reports, 1919, 1921, 1927. Files UN 3/20/1/14/5, UN 3/20/1/16/5, UN 3/20/1/22/7. Veterinary Department, file on trypanosomiasis, Melsetter and Chipinga, 1923-1950. Public records file S 3024/9/1. 46  Manica Agricultural Rural Reconstruction Programme / GTZ, Chimoio *James Bannerman, consultant Dirk Effler, consultant Michael Froude, Agricultural Project Manager Member of Parliament, Chimanimani, Zimbabwe Michael Mataure Rural District Council, Chimanimani, Zimbabwe Mr. Harawa, Chief Executive Officer Mr. Maringe, Assistant Planning Officer Mr. Mukwayo, Administration Officer Mr. Mundoringisa, Council Chair Rusitu Valley Fruit Growers' Association / Agricultural and Rural Development Authority, Kopa Elias Nyamunda, Manager for Vhimba area William Mushayi, Field Supervisor, Technical Department Rusitu Valley Smallholder Tea Project, Kopa Nison Chigogo, Director *SAFIRE, Harare Gladman Kundhlande, Assistant Director Gus LeBreton, Director SAFIRE, Vhimba Zvidzai Chidhakwa, Field Officer *SPFFB and CEF, Chimoio Ana Paula Reis, Provincial Director Universidade Eduardo Mondlane, Maputo *John Hatton, Department of Biological Sciences Vhimba Area Development Committee, Vhimba all committee members and officers World Bank, Maputo Rod de Vletter, Operations Officer for Environment *Zimtrust, Harare Champion Chinhoyi 48