Navigation – Plan du site

Community forest management on the agricultural frontier : charcoal makers, immigrant associations and land claims in Ankarafantsika, North-West Madagascar

Frank Muttenzer
p. 249-272


La gestion communautaire d’une frontière agro-forestière : charbonniers, associations d’immigrants et demandes de terre à Ankarafantsika (nordouest de Madagascar)
En raison des opportunités de travail agricole, des immigrants ruraux se sont installés dans le nord-ouest de Madagascar depuis le XIXe siècle. Ce processus migratoire a pris de l’essor dans les années 1930. Il a débouché dans les années 1980 et 1990 sur l’occupation permanente des terres anciennement forestières aux abords et à l’intérieur d’une aire protégée. Plus récemment, un projet régional de développement rural a autorisé des associations paysannes à produire du charbon de bois dans la zone périphérique du parc national Ankarafantsika. Ces contrats de gestion communautaire des forêts sont proposés comme un dernier recours pour préserver les forêts protégées, procurer des revenus alternatifs aux occupants illégaux déplacés et freiner l’afflux de nouveaux cultivateurs pionniers. Mais les immigrants eux-mêmes ne semblent pas être concernés uniquement par la gestion durable de la filière bois énergie. Ils considèrent la gestion communautaire des forêts aussi comme un moyen pour légaliser les prétentions coutumières fondées sur une première occupation.

Haut de page

Texte intégral

  • 1 Many associations were established prior to the externally funded negotiation of community manageme (...)

1Livelihoods of rural immigrants in the lower Betsiboka region, North‑West Madagascar, consist of subsistence cultivation of dry crops on new or ancient forest burns and flooded rice in the lower lying areas, combined with charcoal making for nearby urban markets. Charcoal is produced in a framework of loosely connected village associations. Such associations were first set up by local people themselves consisting of first or second-generation immigrants from elsewhere in Madagascar, to regulate issues of common interest such as charcoal burning and securing cultivation rights on formerly forested lands1. Local communities are often multi-ethnic yet each immigrant group maintains its own specific migration patterns and ideals. The Betsirebaka for instance, a local term denoting different peoples from the South-East (such as Antaimoro, Antaifasy, Antanosy), describe themselves as strangers “who search for a livelihood” but who want to “return to the ancestral village” if only to be buried there. Many inhabitants are in fact locally-born descendants of immigrants and consider themselves as having full property of their agricultural lands.

2Besides the large scale migration from South-East to West that has been going on for several generations, there are other forms of mobility within the host region, which follow typical paths of social ascension, or regular seasonal shifts in land-use patterns. Some lands cannot be inhabited during the rainy season while other cannot be cultivated during the dry season. In some cases diversification of family labour is such that certain individuals are part of the territorial group (and the village association) only during some months in order to work in charcoal before leaving to their fields which are situated elsewhere in the region. In other words, charcoal producer associations may fulfil, alternatively or at the same time, several social functions: they provide the administrative framework for economic activity based on state‑owned resources; they informally distribute individual parcels of land to each member of the association once the forest has been cleared; and they facilitate the integration of new immigrants into local society.

3In what follows, I shall be concerned with “mobile” as much as “local” communities when describing the complex and multiform relations between village associations and customary territorial groups. Community-based resource management initiatives carried out in the buffer zone by both the Park administration and a regional energy wood management project, entails a repositioning of local actors’ strategies through participation in forest management associations. Yet these interpretations of local custom are not random because the process of selecting and combining legal rules of different kinds is informed by relatively stable social representations of labour, ancestral origins and inter-ethnic relations.

The study region: geographical and historical context

  • 2 Before 2002 when the National Park was established, 2 150 inhabitants used to live on lands inside (...)

4Immigrants from other parts of Madagascar have been settling in the lower Betsiboka since the 19th century because of labour opportunities in agriculture (Deschamps, 1959). In the 1980s and 1990s immigration has led into the permanent illegal occupation of formerly forested lands inside a protected area and its immediate surroundings. The Ankarafantsika National Park is the largest remaining dry forest in the region and covers a total surface of 120 000 ha. Besides recreation for tourists, it provides invaluable ecosystem services to a floodplain of national importance situated some way downstream where irrigated rice is cultivated. According to the National Park administration, 27 300 persons lived in the buffer zones of the park and were distributed over 108 villages and hamlets when the new park was established by merging to previously protected forest reserves2. These people are largely immigrants belonging to different ethnic groups from the Southern, South-Eastern and Northern part of the island.

Settlement history

  • 3 Assisted by the British, Radama I pursued a policy of expansion of the Merina kingdom to strategica (...)

5A problem of integration might arise from the fact that these immigrants claim to be customary owners of land that fall within the ancestral domain of the Sakalava ethnic group (Jacquier-Dubourdieu, 2002, p. 289). However the possible links between contemporary land-tenure and territorial claims based on Sakalava origins should not be overstressed. Contemporary Sakalava ethnicity refers to the pre-colonial kingdoms in Western Madagascar established in the 17th and 18th centuries through conquest by an in-coming dynasty. To this day, to be Sakalava means to be a subject of former Sakalava rulers and is predicated on people’s ritual work, including possession by royal ancestors, and multi‑form ideological expressions of former political allegiances. Following this definition, immigrants from other parts of Madagascar would become Sakalava by performing Sakalava ritual work. Moreover, in the Marovoay floodplain close to which our two study sites of Manaribe and Marolambo are located, the traditional Sakalava economy based on cattle has been in competition for more than two centuries with a system of permanent agriculture imposed by outside forces. The first migrations followed the pathways of King Radama’s military expedition in 18243. Merina colonisation of the fertile lands of the Betsiboka floodplain was pursued right until the end of the 19th century. The French colonial administration also took interest in the Marovoay plain and converted it into one of Madagascar’s rice granaries. During the 1920s, the land improvement schemes set up by the French attracted huge numbers of migrants from the Centre, the South-East and the South of the island.

6A specialisation in economic activities then took effect among the immigrants. The Merina and Betsileo were encouraged by the French administration and settlers to take as sharecroppers on the land developed for wet rice cultivation. Migrants from the South and South-East, collectively referred to as “Betsirebaka” or “Korao”, were employed in the industrial plantation zones deserted by the Sakalava who refused salaried labour. The first wave of migration in the 1930s was followed in the 1950s and 1960s by that of the Tsimihety arriving from the North mainly for demographic reasons. Unlike the migrants from the South-East, who usually intend to return to their lands of origin, the Tsimihety are known to pursue a model of territorial expansion. But we shall see that those patterns only reflect general tendencies that do not create stable separations along ethnic lines because they vary greatly according to local conditions. Sakalava identity in the region today is reproduced quite independently of land tenure relations. Given the long history of migration from the South to the North-West, customary forms of land tenure are no longer thought of in terms of an indigenous mode of production in competition with that of the immigrants. In spite of ethnically specific mobility patterns, the descendants of immigrants constituted at the end of the colonial period around one half of the regional population. This has meant that since 1960, the new arrivals have had to adapt themselves to a society already transformed by more than a century of Malagasy nation‑building.

Contemporary immigration and environmental issues

7At present, the villages near Ankarafantsika are the main purveyors of charcoal for the city of Mahajanga, as well as the smaller towns of the lower Betsiboka (Duhem et al., 1999). The importance of the region for charcoal has increased due to exhaustion of wood resources in the rural communes closer to the provincial capital. In the surroundings of the National Park, charcoal is produced in two rural districts, Ambato-Boeni and Marovoay. Our case material refers to fieldwork with charcoal producers’ associations from two villages of Marovoay district on the Northern side of the Park. These villages sell their charcoal either to Marovoay, a secondary town of 30 000 inhabitants or to Ankazomborona, a small town of less than 10 000 inhabitants located on the national highway connecting Mahajanga to Antananarivo. We chose to study these two associations in detail after having done a survey on most associations around the Park, especially the ones set up by the energy wood management project. Usually these associations have little or no influence on the price levels at which charcoal is sold because of the monopoly position of buyers who transport charcoal to Mahajanga. In these conditions, the production chain is controlled by these intermediaries and local associations have little leeway to re-organise the production chain. In the cases we studied, charcoal is sold independently local and regional markets. Prices are to some extent negotiated on a case by case basis when burners own cart and oxen to bring charcoal from village to town. Those without means of transportation have to sell their charcoal to others at a lower price in the village itself.

8Producers usually give two reasons for adhering to village charcoal associations. On the one hand, it facilitates the recognition by government authorities, namely the forest service and the commune rurale, of an activity that is essential for local livelihoods. On the other hand, the administration does not have to deal with each producer individually to collect taxes, which makes the relation between villagers and the authorities more comfortable. The issue of access to land is usually not mentioned spontaneously, although it appears to be of central importance in the working of those associations. The arguments related to the livelihood complement in the form of monetary income and to administrative recognition both point to the need to regularise the insecure situation of local communities consisting mainly of migrants. Most of the inhabitants have not lived in the region more than ten or fifteen years and they came to find land for cultivation in and around a forest that has recently been transformed into a National Park.

New aspects of conservation policy in Madagascar

9Since the late 1980s Madagascar has received substantial amounts foreign aid to create an environmental program. Prior to establishing a National Environmental Action Plan (NEAP) based on a World Bank model followed elsewhere in Africa, Madagascar only had a forest service but there was neither a Ministry of Environment nor specialised agencies for the implementation of environmental policies. During the first years of the NEAP (1990-1996), foreign aid was directed mainly towards a network of protected areas consisting of 50 national parks and natural reserves, about half of which had existed since colonial times while the other half was to be set up from scratch. In the view of international donors, protected areas were to be taken out of the hands of the understaffed and corrupt forest service and administrated by a private organization less influenced by government interests. As long as projects were confined to protected areas, ideas on new public management and public-private partnerships did not entail changes in land tenure policy and natural resource governance. The objective was to strengthen sector-based management of public land by central government or by donor controlled agencies, rather than to decentralize power over land and resources by devolving it to local government.

10In the second phase of the NEAP (1997-2002) substantial efforts went into contractual management of state forests by user associations at the village level. Community forestry in Madagascar is a case of aid project coordination through transnational policy discourses, a process involving international donors, the government, NGOs and village associations. Much expert knowledge in the field of community forest management clusters around the idea of integrated landscape conservation, which was experimented first in the 1990s in the context of buffer zone management and has gained momentum ever since. As in other aid-dependent states, the objective of conservation policies in Madagascar is to involve local communities in nature protection while at the same time take into consideration local livelihood needs. Natural scientists consider community forest management as a way to go beyond the fortress conservation approach by extending protection to forests outside protected areas (Nicoll, 2003). Extension appears necessary because the protected areas of the colonial period were established in view of protecting certain spectacular landscape features and of strictly separating human activity from the domain of ‘nature’, meaning that the designated surfaces of existing reserves and parks are too small to allow for effective biodiversity conservation (Kremen et al., 1999). For environmental economists, more equitable benefit sharing will alleviate rural poverty and thus enable the potential trade-offs between productive uses and environmental services of forests to actually take place. They look at community forest management as a means to allocate resources more efficiently. For other social scientists, community forest management is not confined to benefit sharing but entails power sharing between the administration and local communities (Wily, 1999). Local forest user associations are seen as a first step to ‘decolonise’ tenure relations and to sort out conflicting land claims, both of which enhance overall tenure security and act as incentives to integrate at the landscape level sustainable resource extraction with environmental conservation.

Participatory approaches

  • 4 To preserve the remaining forest and to stem the influx of more and more migrant cultivators from t (...)

11Despite the political consensus on the importance of community‑based management, current approaches and pilot projects seem to overlook that the objectives pursued by peasant associations are usually linked to securing rights on forest land that is being cleared for cultivation. A major problem faced by ANGAP has been how to curb human pressures without destroying the livelihoods of the 27 000 villagers that inhabit the immediate surroundings of the Park. To implement national conservation policy goals, ANGAP has in the main relied on a narrative of benefit sharing4. Approaches were designed to gain local cooperation with the existing state-controlled management regime, the focus being set on providing alternative sources to forest income, employment opportunities, improved legal access to certain resources and shares from revenue earned from the forest. These approaches were implemented both in buffer zones and in so-called zones of controlled occupation inside the park itself.

12Participation, conservation and livelihoods have taken yet another meaning when a development project for the management of energy wood became active in the park’s buffer zone. The project was based on a narrative of power sharing and was designed to devolve forest resource control to community level. It assisted the community, in bringing the charcoal product chain, under its own control and even land tenure, on the grounds that only this level of empowerment will enable the communities to conserve the forest for lasting livelihood and local environmental benefit. The energy wood project interventions in some villages located in the buffer zone had unexpected consequences. While the project encountered important resistance from the field-level park management agents, it was extremely popular with recent immigrants in the buffer zone. This was not only due to the improvements it was to bring about in the local charcoal producing chain but also because the community forestry associations were seen by villagers as a form of recognition of prior occupation of the land.

Environmentally sophisticated land reform

13The participation of rural communities in managing ‘integrated forest landscapes’ is difficult to justify while it is acknowledged at the same time that deforestation is a way of securing traditional claims to land (Muttenzer, 2006a). The ‘human occupation of protected areas’ first emerged as a problematic issue in the environmental policy debates of the early 1990s. Challenging the conventional neo-Malthusian explanations, social scientists pointed to open access as the major cause of human occupation in protected areas, and to the lack of administrative recognition of customary property rules and practices in adjacent zones (Weber, 1995). This led international donors and the government to opt for a policy of community-based management of resources located in buffer zones, as well as other “forests outside protected areas” (Nicoll, 2003). The issue of customary land tenure has once again come to the fore during the third phase of the NEAP (2003-2008), following former President Ravalomanana’s 2003 landmark announcement to triple the surface of protected areas from 2 to 6 million hectares. The new kinds of protected areas to be created under this policy are eligible, at least in part, for management through forms of community conservation currently still under discussion, but which would ideally be based on pre-existing tenure arrangements.

  • 5 Despite the revisionist tendency in the political ecology literature, it appears that Madagascar’s (...)

14A century after the introduction of colonial land law in Madagascar, the majority of state-owned land, and even part of privately owned land, continues for all practical purposes to be governed by customary tenure relations. There also is evidence that rural populations take advantage of this legal pluralism by securing new land for cultivation to cope with soil degradation and social and economic inequality5. Although they constitute the predominant form of law in rural Madagascar, customary rules and arrangements have enjoyed very limited statutory recognition, at least until recent land legislation created two mechanisms to recognise aspects of customary tenure on public lands. The first such mechanism is the contractual management of forests outside protected areas, which was designed for sustainable management of village commons by community associations. The second mechanism is registration by local government of customary ownership rights in cultivated land, which was designed to privatise joint lineage or family property.

15Although the purpose of an “environmentally sophisticated land reform” (Geisler and de Sousa, 2000) is not to replace one solution with another, but rather to enlarge the range of available options to secure equitable access and ensure resources are used sustainably, it appears that the new land legislation in Madagascar is set to re-enact the spatial separation of agricultural and forest domains (Muttenzer, 2006b; 2010b). The effectiveness for environmental conservation of community forest management is as uncertain as that of the earlier state-centred forest policies, particularly in places where poverty reduction entails cultivating land that is being acquired by “first occupants” through clearing a piece of forest, a claim not recognised by community management contracts. By contrast, local registration of customary tenure is expected to encourage agricultural intensification by recognising labour efforts invested in the land. But the registration of customary private property applies only to permanently cultivated lands, such as irrigated and flooded rice-fields, and excludes future inheritance claims by joint family and/or lineage members. On formerly forested lands, both individual and joint property claims, are altogether excluded from registration by local authorities. To date, the existing legal options for recognising customary tenure are too limited in scope to bring about or even initiate a significant transformation of prevailing land use management patterns.

16The following analysis of livelihoods and institutional change shows that clearing forest is a more effective way for families to deal with pressing problems of rural poverty, generating revenues by selling charcoal to securing first occupancy rights of migrants, and integrating later arrivals into the existing social fabric. The fact that immigrants’ associations deal with such diverse issues as regulating livelihoods and integrating foreigners within host communities is no proof that community forest management works to achieve its stated goals. What there is on people’s minds may be more accurately described as the strategic use of a community forestry discourse to defend a “folk conceptualisation of property” (Muttenzer, 2006a).

Community forest management on the agricultural frontier

17The immigrants voluntarily adopt development notions related to associations, rather than avoiding or resisting participation in environmental policies. However, given that the goals pursued by projects contradict the goals of settler communities, the favourable attitude displayed by villagers should not be misinterpreted as genuine environmental concern. Whereas development projects try to fit migrant communities within an externally conceived spatial grid to reduce human impacts in protected areas, immigrants ask for the recognition by government authorities of a temporary state of affairs in an ongoing process of conversion of land use from forest to agriculture. The comparison of land and resource use patterns indicates that there in this respect no difference between an association who received project assistance and one who did not.

Marolambo: an immigrant association with project support

18In the first case studied, individual members of charcoal burner association continue to clear forests for cultivation, and incidentally burn charcoal, on parcels other than those designated by the management plan. The papers required to transport charcoal from the village to the town are issued locally even if the legal origin of produce is in doubt. Spatial zoning based on ecological criteria as imagined by the project for the forests where energy‑wood is produced, and by ANGAP for the buffer zone and larger surroundings of the National Park, is not effectively implemented by the village association. The membership of later immigrants is said to be superficial because there are conflicts with earlier members, who were already living there before the contracts were elaborated by the project and signed by the forest service. The interviews with field-level agents of ANGAP reveal a distinctly negative perception of charcoal burners in the buffer zone. In their accounts, the members of the association are accused of not living up to their promises with respect to protecting forest in the buffer zone, of hiding their true intentions, and of taking advantage of the physical presence and moral support of the charcoal management project.

19These problems were further aggravated by the fact that since 2003, the World Bank funded project, which temporarily associated the Forest Service, the Ministry of Energy and a French agricultural development agency, has disappeared from the local scene. As a consequence, the associations received neither financial nor technical support after the forest management contracts had been signed in 2002. The forest service is competent to follow through contracts and evaluate outcomes but its role is rather unobtrusive in comparison with ANGAP. This is obvious in the discourses of villagers who project their traumatisms on this new authority perceived as all-powerful and even willing to put villagers’ lives in danger. Independent of aid-supported community forest management contracts, the forest service issues administrative authorisations to burn charcoal on state lands other than protected areas, both to immigrant associations and individual families. The amount of taxes paid by an ordinary charcoal burner’s association was between 1 and 2 million FMG (Franc malgache) per year at the time of our enquiry. Tax revenue generated through the project scheme was expected to reach the double or triple if taxes are paid regularly, although the cheaper option of ordinary authorisations remained available to producers.

Mangatelo: an immigrant association without project assistance

20This association had initially received the same attention by project staff as the previous. But later on, although villagers had applied for a community forestry contract, their case was dropped without further explanation, possibly due to a latent conflict between the charcoal project and park managers over project activities. The concerned piece of land was adjacent to the protected area and remaining resources there were scarce in the eyes of the villagers. Given that the project failed to support their request and that in no instance the forest service has devolved management without external project funding, this charcoal burners’ association was not to be recognised under the new community forestry policy. The association had been set up as early as 1996 in response to initiatives by Conservation International, an international NGO that was managing the Ankarafantsika forest reserves prior to establishment of the new National Park. They were interested in identifying alternative income possibilities and more generally in talking to people living in Ankarafantsika, rather than setting apart forests for community-based charcoal burning. A second difference with respect to the recent settlers discussed earlier is that a local community had been in place for much longer including several generations of immigrants.

21The area supposed to be managed by the association is rather large (about 2 000 ha) and dwellings are dispersed in hamlets and small villages inhabited by one or several extended families. There is no zoning plan defining different land uses nor is there a specific forest set apart for charcoal production. Plots with trees that can be used for charcoal are found in several locations not too far away from habitations and agricultural fields of a given hamlet. The situation is different from the previous one where a primary forest is cut down to make way for agricultural land. One could describe it as a mainly agricultural system with including a significant charcoal component from secondary forest growth.

22An observation made frequently by villagers from Mangatelo is that forest resources were not sufficient to allow charcoal production both by permanent residents and occasional burners arriving in large numbers from other villages of the region. In spite of resource scarcity, one does not find the polarization and conflicts observed in Marolambo. The charcoal burner association is only one among many elements that structures the relations between families and is hardly decisive in creating orderly relations at the level of the local community. The contrast between the two associations therefore cannot be explained only with reference to the organisation of rural charcoal markets. It is dependent on the more general social role played by the village associations. In Mangatelo, this role is limited to charcoal making and the association deals only accessorily with integrating new arrivals, whereas in the case of Marolambo, the very identity of the pioneer community in competition with other such communities is being negotiated via the charcoal association. The criteria of membership in Mangatelo are therefore less rigid than those in Marolambo. The difference is nicely illustrated by the contrasting notions of affiliation to an association and belonging to a community. In field interview we usually asked villagers whether there were particular rules governing participation, whether charcoal burning was reserved for certain categories of people or whether it was an activity open to all.

23The responses we obtained show that the distinction between affiliation and belonging was not clear-cut, given that the qualification of members of the association varies according to their individual objectives. Some of them both plant rice, corn and cassava and burn charcoal, others only plant. There are people who plant elsewhere but visit the place regularly to burn charcoal. And there are the landless who do not plant anywhere but who come to burn charcoal for a limited period before leaving to other destinations. Yet in Mangatelo each of those individuals may in some respect be considered part of the local community as soon as he informs the president of this presence and pays a minor fee. Whether the said individual has the intention to look for a more permanent status or to leave after having worked for some time is his personal decision and of no concern to other people. Thus while membership in the charcoal association is a modern legal construct, its underlying purpose conforms to longstanding rules of hospitality and of creating a customary community by attracting immigrants.

24This observation confirms a conclusion drawn on many other occasions in rural Madagascar. The attachment to certain traditions, although they may be transformed in the process, does not prevent peasants and local officials from imitating modern legal categories to give legitimacy to their practices. On the contrary, charcoal burners make use of community associations to display respect and conformity towards state authorities, even without the presence of development projects that encourage them to do so. In other words, there is an imitation and addition of new forms rather than a substitution or replacement of indigenous law by transplanted law (Chiba, 1987). This seems to be a reaction to the ambiguity and insecurity that may result from the weakening of descent based authorities, be it because of rural mobility or because of the influence of bureaucratic governance. It translates a superficial modernization of law where new legal concepts, such as community forest management, are adopted in the spirit of existing categories, such as prior occupation of the land.


Forest user associations and livelihood strategies

25Charcoal burners’ associations have existed in the region before the energy wood management interventions of the project. The development project only gave official status to the village grouping that existed already before. At the time of fieldwork, only the first association had a contract for energy wood based on a simplified management plan, the members of the second association burned charcoal based on annual authorisation by the forest service in Marovoay. The participatory approach pursued by development projects, here defined in terms of community based forest management, coincided with an existing tradition of associations grouping immigrants of common ethnic origin in search of lands and livelihoods. Behind the screen of development and integrated conservation discourses, the charcoal burners’ associations fulfil a whole range of functions related to the transformation of subsistence economies, such as securing a regular complementary monetary income essential for landless immigrants, minimising conflicts with local government authorities through conforming to the law and allowing for the collection of rather symbolic taxes, or to regulate land claims between immigrants who settle on previously unoccupied land.

26According to the official view of integrated conservation, the purpose of community forestry associations is to contribute to ecosystem conservation through sustainable resource use. In ordinary practice, there are significant differences between conservation projects and types of participation that allow for a more productive engagement with the environment, including commercial uses. The dissimilarities in approach, which seem to reflect a division of labour between conservationist and pro-poor aid agendas, are more pronounced in the present case of resource management in the immediate surroundings of a National Park. But there is potential for conflict where aid projects are set up without any reference to spatial planning and management by local government, that is potentially anywhere in rural Madagascar. As already mentioned, the association of Marolambo was set up (or at least adapted to a new purpose) in the framework of a region-wide effort to control charcoal burning. Following the legal procedures, villagers addressed a request for a community forestry contract to the district forest official, after an information campaign by the project.

27In this particular case, the villagers’ request lead indeed to the elaboration of management contract with help from the project. A forest plot was delimited for harvesting according to a simplified management plan authorising a sustainable yearly quota of charcoal and requiring yearly rotation. The contract and management plan also mention agricultural land and areas for grazing. These provisions to some extent reflect existing relations among members, but they have no further bearing on third parties. The provision concerning rotational harvesting echoes the local perception according to which one harvests the trees where they stand and goes elsewhere once all trees are gone. But villagers do not consider that wood for charcoal burning is anywhere close to exhaustion.

28In our interviews, we raised the issue of occasional charcoal burners who are not members of the association. The answers suggest that the difference in status and rights between members and non members is well understood, which is exceptional in regional comparison. The reason is that the association is also taken by villagers as a tool of social control way beyond charcoal burning, because membership points to the discourse of justification of land rights by first occupants. Community-based user associations are a symbol of modernity, and confer in the eye of the field-level park officials a degree of respectability even to poor, landless immigrants. In the view of the pioneers settling at Ankarafantsika, forming an association is the first step towards the recognition of human occupation of the area by local government authorities. Unless immigrant communities are able to acquire regular administrative status (after reaching a certain population threshold), or to register the occupied land under collective title (following long and complex land titling procedures), forest user associations are the only means to give the customary territorial groups some form of administrative existence. The search for administrative recognition also explains why members of the association display a lot of goodwill to cooperate with park managers in matters such as controlling the movement of persons in the buffer zone or preventing forest fires. In some cases, the collaboration with authorities to “preserve the forest patrimony of the nation”, as was often repeated by informants, may be directed against the residually pastoralist land uses of the ‘indigenous’ Sakalava. In other cases, similar arguments are used against other, more recent immigrants pursuing resource appropriation strategies that are perceived as aggressive by the earlier settlers.

29One of the associations illustrates the social project of a pioneer community. This is the usual case of the Tandroy from the South of Madagascar, whose strategies of occupation of land are straightforward with little regard to the pre-existing natural and human environment. Even before the legal recognition of charcoal associations, an immigrant association in Marolambo had obtained clearing permits for a surface ten times larger than the current charcoal production forests. Once the forest was cleared, parcels of 2 ha were attributed individually to each family head member of the association. In situations like this, charcoal producer associations are at the same time “immigrant associations” (Rajaonarison, 2002). The descendants of earlier inhabitants, whether indigenous to the region or not, usually refuse to adhere. The pioneer mentality that is particular to the Tandroy although not exclusively contrasts with the transmigration model of the so-called Betsirebaka from South-Eastern Madagascar. In the case of the second association, there is no polarisation between the indigenous and the migrants and everyone will join as a member whether he is a foreigner or not. There are indeed no strictly speaking indigenous claims to land because local communities consist of several immigrant groups with complementary histories.

30To understand this kind of social structure it is necessary to recall the colonial economic history of the region. In order to attract and keep their salaried migrant labour, the colonial concession owners let them cultivate on unexploited lands of the concessions, or beyond. While working on the concession, migrants at the same time tried to establish themselves as small peasant producers, settling on government lands and on indigenous reserves attributed to the local populations (Jacquier-Dubourdieu, 2002, p. 295). The present strategies of immigrants follow a similar pattern. Many charcoal producers we spoke to said they first came to work as day labour in the large rice fields of the Marovoay plain. While working there, they prospect the close-by hills for land of lesser quality that is not yet occupied. After working for a year or two as day labour, they would decide to settle more permanently in the region cultivating maize, manioc, beans and dry rice, on the newly established plots considered as their personal customary property. Small agriculture of the sort is mainly for self-consumption. Yet at the same time, migrant cultivators would regularly produce some charcoal to generate minimal but more or less stable money income. In some other areas, where there is sufficient water to cultivate tomatoes and other vegetables sold in Marovoay, people would produce proportionately less charcoal.

31In the places where most of the lands are already occupied, migrant would invest their effort in charcoal, exclusively or combined with sharecropping. The settling on the slightly elevated plateau between the Marovoay plain and the limit of the Park is the second stage of a trajectory of social ascension from to landless migrant to small peasant. Based on interviews with about one hundred individuals, we infer that this model of immigration applies to at least one half of local inhabitants, and to most of the charcoal burners. The similarity with migrations during the colonial period is not coincidental. Many of our informants, especially the Betsirebaka who are the majority in the second association we studied, say they do exactly as their parents did. They arrive “in search of a livelihood” and it is their custom to “return to the village” once they found what they were looking for, only to come back to search again. The same pattern of mobility is found among other ethnic populations than the Betsirebaka. But one could argue that mobility is an actor ideology rather than an actual way of organising migration. Whatever individuals may express concerning their customs, intentions, hopes and ideals, many of our informants are in fact locally born children of immigrants. While they continue to be attached to kin in the ancestral villages far away, the population transfer from the South-East to the North-West of Madagascar is substantial over time and hardly reversible.

32As an ideology, mobility helps to define the community structures in the newly settled territories, more so perhaps than the actual migrations between the place of departure and the place of arrival. Besides that, there are also migrations within the study region that contribute to the shape of local communities. These are due to climate hazards, work opportunities and displacement of settlers by the Park authorities. The contrasted ways in which ethnicity is played out by pioneer communities and mobile communities suggest that one point all types of charcoal producer associations have in common is to provide a safety net. They are part of a risk minimising strategy which consists in selling charcoal to live on when subsistence cultivation of maize and cassava is insufficient to survive.

The role of forest user associations in land claims

33Ideas of prior occupation of the land differ substantially with respect to the period of arrival of immigrants, ethnic representations of mobility, as well as pre-existing social structure in the host territories. Discourse analysis suggests that there are at least two ideal types of prior occupation: original acquisition and derived acquisition land rights (Muttenzer, 2010a; Rarijaona, 1967). As a consequence of those diverging narratives, the local use made of associations and participatory mechanisms put in place by the Park management agency and by the energy wood management project are locally specific. While in the case of pioneer communities, the objective of user associations is to secure land rights, in the case of trans-migrant communities it is to secure alternative livelihoods to new arrivals during the time they establish and improve relations with earlier occupants.

34Pioneers explain, and justify, their land rights with reference to a material act of appropriation followed by cultivation, rather than with reference to a negotiation with earlier occupants. This does not mean that pioneers have no need at all to secure land rights by appealing to a third party, but simply that the objective of installing the group on the territory and the appropriation of family fields are pursued directly through the immigrants or charcoal burners associations, rather than through contractual relations with Sakalava tompontany (masters of the land) who may claim customary rights, especially for pasture, over the pieces of land colonised by immigrants. We asked the members of the association of Marolambo why they had chosen to settle on the previously forested plateau of Belavenona rather than elsewhere. They answered that the choice is due to the fertility of the land that « promised to be a way to avoid famine and suffering of families »

35In the case of immigrant association of Marolambo, later converted into a charcoal burners’ association, the occupation of large pieces of forest land by several pioneer groups from the South took place between 1990 and 1995. The process entailed, or indeed consisted of, the appropriation of family properties because personal lots were distributed to individual family heads as members of the association. As there were several groups of pioneers, as well as earlier immigrants pushed back towards the river plain, a competition between several community-based associations was the logical consequence of the conquest of new lands. On top of that, numerous families already settled in the lower-lying areas have seen some of their plots destroyed by inundations and changes in the river-bed of the Betsiboka. Usually these families consist of earlier immigrants, who do not appreciate the late-comers taking most fertile lands in the higher areas. These families thus compete with the « foreigners » for the Belavenona forest, while continuing at the same time to cultivate fields further down. This competition may explain the seasonal movements mentioned earlier between Belavenona and the Betsiboka plain.

36The community forest management initiative, which legalised already existing immigrant associations as charcoal burners associations, though it was not the cause of this competition for land, contributed to the conflict because villagers understood that the energy-wood management contract confers a title that can be invoked against other pioneers, or park officials who try to restrict productive uses of adjacent forest. All kinds of actors we interviewed confirm that after the intervention of the project, huge local enthusiasm for charcoal burners associations ensued. The energy-wood project is popular in the view of villagers not so much because it legalizes the commodity chain linking them to the growing charcoal market of Marovoay than because of the administrative recognition it confers on the otherwise illegal settlement of state lands.

  • 6 The bureaucratization of village and inter-village social relations by forest user associations has (...)

37The main purpose of the community forest association is to legitimise the conquest of open land by a pioneer community as well as individual appropriation of plots by its constituent families. While in the case of Mangatelo, community forestry’s only stake is charcoal burning, in Marolambo, it is the economic viability of their families that is at stake because, in this community, membership in the village association is considered a condition to become an individual customary land holder. Somewhat surprisingly association members reproduce the ‘totalitarian philosophy’ of colonial forest law. Ordinary members say they need an authorisation by the president to cut a tree even when that tree is located on the individual lots attributed to each family, outside the plot for which decision-making power has been devolved to the association according to the management plan6.

38In spite of difficult relations with park managers whom they perceive to be “responsible for the forests of all Madagascar”, villagers speak of their forest user associations in terms of localized branches of the park service, although they are officially about charcoal. The internalisation of the postcolonial order may be for them a means of social control on the frontier, a way to unite disparate settler groups by giving them a common local identity. It may also be a means to demonstrate to the outside world their being in conformity with the goals of a larger community, personified by the forest service, the park officials and the charcoal project. The role of the charcoal burner association is not limited to regularise relation with the forest service, but also to substitute and prepare for the future administrative recognition of a new territorial group. But the competition between diverging uses of space for production and conservation is likely to remain.

39In contrast, the origin of inhabitants of Mangatelo is mainly from the South-East (Betsirebaka) and the North of Madagascar (Tsimihety). We have seen earlier that the Betsirebaka are characterised by pattern of immigration where the identification as “foreigner” is artificially prolonged far beyond the time objectively required to realise their integration into the local community. Although the latter is a relatively recent construct resulting from successive migrations since the late colonial period, it is perceived by new arrivals as a pre-existing social unit with which they have to come to mutually agreed terms. Contrary to pioneer communities who use the category of association to establish themselves autonomously in the host territory, the local integration of ‘trans-migrants’ does not primarily rely on charcoal burners associations but on agrarian contracts relating senior and junior migrants. The overall goal of immigrants is the same everywhere, to seek access to land for cultivation, but the trans-migrants’ way of going about it is quite specific.

40New arrivals are expected to become the clients of earlier ones who act as tutors. It usually takes about five years to move from client (junior immigrant) to tutor (senior immigrant) status. During that time, access to personal rice fields is restricted but they may work on the fields made by others. As a consequence, they depend during that time more than others from collecting forest products and burning charcoal to satisfy family consumption. Families which have not yet gained access to a plot for cultivation, or only through relations with other families, will concentrate their labour force on exploiting the forest. Charcoal is a major source of income while waiting to become a full member of the local community and acquiring customary ownership of the fields needed for subsistence. Charcoal burner associations contribute to the process of acquiring land for cultivation and permanent settlement, but only indirectly. Rules of access to agricultural plots in Mangatelo are in continuity with the personal histories of trans-migrants. Before establishing themselves on that piece of land, most have already worked for some time as day labourers or sharecroppers in the nearby Marovoay plain. After some time, they try to get their own property and establish themselves indefinitely in the region. This kind of biography is frequent and the corresponding tenure arrangements have several implications for the role played by forest user associations.

41The first, and most obvious, objective pursued is procurement of monetary income. According to local officials of ANGAP, 95 % of cash circulating in villages stems from the sale of charcoal. As already mentioned, the participation in associations is a way of generating income especially for those who have only derived access to agricultural land. But it is only accessory to agricultural colonisation, and people continue to burn charcoal even after they have their own plots because they will always need cash “to pay for other work”. Access to land, that is unequal relations between senior and junior immigrants, is the less obvious variable which accounts for the amount of labour a family will invest in exploiting forest resources. If cultivation is possible, it will be less, if it is not possible or conditions are discouraging, it will be more. Charcoal is complementary to the progressive conquest of territory by several generations of trans-migrants. But in Betsirebaka terms, irreversible migration from the South-East to the North-west is thought to never have started because each generation “only does what their parents did”, and it will never come to an end because they will always “return to their village” even though they have been living in the host region for several generations and have acquired permanent rights to the land. Rather than indefinitely reproducing second-class citizens excluded from access to land, the concept of derived rights is used to justify not only use-rights indeed the full right of customary ownership usually reserved for tompontany (masters of the land).

42At the same time, there are original claims to certain types of plots. The second objective of charcoal burner associations is to authorise clearing of new lands. This is not in contradiction with the derived rights conceptualisation of appropriation just presented, because sharecropping arrangements only deal with rice fields, while the rules of access to less fertile lands are more permissive. Before delimiting a plot and “tidying”, i.e. clearing it before planting, villagers must ask the charcoal burners’ president for an authorisation, to make sure that the plot is not yet occupied by somebody else. However, land distribution is much less explicit here than in the case of pioneer associations where delimitation of individual plots was decided by an assembly and a map was drawn up.

43Whatever the differences in detail, the sociology of rural charcoal markets in Madagascar confirms the observation made already some years ago that there has been a revival of associations in rural Africa since the 1980s (Olivier de Sardan, 1994). Although immigrants’ associations may not necessarily lead to efficient forest management, they contrast sharply with the passive resistance and avoidance by local communities of relations with the state and other external actors. On the contrary, these communities seek legal recognition to engage with local government representatives, NGOs and international aid projects, who take their attitude as a proof for their organisational capacity and self-promotion. International donors joined the NGOs and grass-roots developers in trying to establish an open dialogue with the civil society, a term often used to refer to associations, and to make corrupt and inefficient public administrations accountable. Actors external to the rural society think that peasant associations may encourage production, manage product chains, spread knowledge, and participate in public policy more efficiently than state bureaucracies. The energy wood management project, for instance, was based on the certainty that forest resources will be exploited anyway, and that instead of prohibition, the administration had better control it in view of sustainable management. But this objective is shared neither by pioneer and trans-migrant settlers, who form associations because it serves their colonisation project, nor by park officials working in the buffer zone, who intend to mobilise associations for conservation.


44Work in the political ecology and environmental anthropology of Madagascar suggests that the received wisdoms of conservation and development practitioners may be influenced by a world view which overrates the opposition between modern and traditional forms of political-legal control (Corson, 2011; Kull, 2000; McConnell, 2002). In reality local legal practice may be much more hybrid than the corresponding legal discourses, as the bureaucrat’s model can be internalized by the villagers and orient their day-to-day conduct in the face of local officials (Blanc-Pamard and Rakoto Ramiarantsoa, 2007; Goedefroit, 2006). The assumption then is that hybridized law, because of its apparent flexibility, is also more amenable to negotiated policy solutions balancing local livelihood interests with global conservation interests. The customary law I have described in this article is certainly most often hybrid, even in situations where local populations may insist on its purity and forest officials on its backwardness. But does it follow therefore that hybrid law is more negotiable than its purely traditional predecessors or more likely to avoid stalemates between Park managers and buffer zone populations?

45Community forest management contracts amount to a de facto recognition of prior occupation by state authorities. By the same token, the state recognises existing orders of precedence between first occupants and later migrants. However the state, and indeed the forest user associations themselves, do not recognise exclusive community rights, because forest guards continue to issue authorisations to people who are not members of the community associations. Rewards for personal labour investment are granted irrespectively of time of arrival, and in exchange generate a rent of non enforcement of forest laws. In contrast to this, the park managers recognise neither the contractual agreements nor the authorisations of the forest service but threaten to impose penalties on charcoal burning, although it is an essential livelihood component of people having been relocated from the park, as well as more recent immigrants.

46The purpose of these legal procedures is to reframe power relations in order to protect nature, or to control land and resources in new ways. The necessary condition of such an approach is that all stakeholders commit to following the procedure and to accepting its outcomes. The condition is not met when definitions of nature and rights over land and resources are interpreted by participants as a means to maintain power relations, or to control people in new ways. In a world where relations between humans change only as a consequence of constructions of nature, the scientific justifications of environmental policy are meaningless or considered irrelevant. In such a world, the purpose of negotiating resource access is to make real the idea of a ‘naturally given’ order of precedence among first occupants, later settlers and political authority. Rather than trapped in an unchanging tradition, people are committed to a different kind of procedure, which contains both an explicit social contract and a set of tacit conventions (Lewis, 2002) by which the contract is implemented.

Haut de page


Association nationale pour la gestion des aires protégées (ANGAP), Antananarivo, 2000 - Plan d’aménagement et de gestion du Parc national d’Ankarafantsika 2001-2005. Antananarivo: ANGAP, novembre, 126 p.

Blanc-Pamard Ch. and Rakoto Ramiarantsoa H., 2007 - Normes environnementales, transferts de gestion et recompositions territoriales en pays Betsileo (Madagascar). Natures Sciences Sociétés, n° 15, p. 253-268.

Chiba M., 1987 - Three dichotomies of law: an analytical scheme of legal culture. Tokai Law Review, n° 1, p. 279-290.

Corson C., 2011 - Territorialization, enclosure, and neo-liberalism: non-state influence in struggles over Madagascar’s forests. Journal of Peasant Studies, vol. 38, n° 4, p. 703-726.

Deschamps H., 1959 - Les migrations intérieures passées et présentes à Madagascar. Paris: Berger-Levrault.

Duhem C., Razafindraibe R. et Fauvet N., 1999 - Le schéma directeur d’approvisionnement en bois énergie des villes de Mahajanga, Marovoay et Ambato-Boeni. Antananarivo: Ministère de l’Energie et des Mines, Groupement CIRAD Forêt-FOFIFA, 116 p.

Geisler Ch. and de Sousa R., 2000 - From Refuge to Refugee: the African case. Madison: LTC Working Paper, n° 38, University of Wisconsin, 24 p.

Goedefroit S., 2006 - La restitution du droit à la parole. Études rurales, Paris, n° 178, p. 39-64.

Jacquier-Dubourdieu L., 2002 - Les modes d’intégration de l’étranger dans les sociétés sakalava du Boina: impact des migrations de travail. In : Raison‑Jourde F. et Randrianja S., dir. - La nation malgache au défi de l’ethnicité. Paris: Karthala, p. 289-303.

Kremen C., Razafimahatratra V., Guillery R.P., Rakotomalala J., Weiss A. and Ratisompatrarivo J.-S., 1999 - Designing the Masoala National Park in Madagascar based on biological and socioeconomic data. Conservation Biology, vol. 13, n° 5, p. 1055-1068.

Kull Ch., 2000 - Deforestation, erosion and fire: degradation myths in the environmental history of Madagascar. Environment and History, n° 6, p. 423-450.

Laney R., 2002 - Disaggregating induced intensification for land-change analysis: a case study from Madagascar. Annals of the Association of American Geographers, vol. 92, n° 4, p. 702-726.

Lewis D., 2002 [1969] - Convention. Oxford: Blackwell.

McConnell W., 2002 - Misconstrued land use in Vohibazaha: participatory planning in the periphery of Madagascar’s Mantadia National Park. Land Use Policy, vol. 19, p. 217-230.

Muttenzer F., 2006a - The Folk Conceptualization of Property and Forest-related Going Concerns in Madagascar. In : Benda-Beckmann K. von, Benda-Beckmann F. von and Wiber M., eds. - Changing Properties of Property. New York: Berghahn Publishers, p. 269-292.

——, 2006b. Fiscalité, corruption et culture de l’État forestier à Madagascar. In : Froger G., dir. - La mondialisation contre le développement durable? Brussels, New York, Oxford: PIE Peter Lang, Coll. Ecopolis, p. 235-256.

——, 2010a. Déforestation et droit coutumier à Madagascar. Les perceptions des acteurs de la gestion communautaire des forêts. Paris: Karthala, 350 p.

——, 2010b. Politiques foncières et pluralisme juridique à Madagascar. Taloha, n° 19, p. 29-54 : Online version accessed 11 january 2012 :

Nicoll M., 2003 - Forests outside protected areas. In : Goodman S. M. and Benstead J.P., eds. - The Natural History of Madagascar. Chicago: University of Chicago Press, p. 1432-1437.

Olivier de Sardan J.-P., 1994 - Préface. In : Jacob J.-P. et Lavigne Delville Ph., dir. - Les associations paysannes en Afrique. Paris: Karthala, p. 5-6

Pollini J., 2007 - Slash-and-Burn Cultivation in the Malagasy Rainforests: Representations and Realities. Ithaca NY: PhD dissertation, Cornell University.

Rajaonarison A., 2002 - Les associations d’originaires et leur impact, entre solidarité familiale et influence politique. In : Raison-Jourde F. et Randrianja Solofo, dir. - La nation malgache au défi de l’ethnicité. Paris: Karthala, p. 257-276.

Randrianandianina B.N., Andriamahaly L.R., Harisoa F.M. and Nicoll M.E., 2003 - The Role of the Protected Areas in the Management of the Island’s Biodiversity. In : Goodman S.M. and Benstead J.P., eds. - The Natural History of Madagascar. Chicago: University of Chicago Press, p. 1423-1432.

Rarijaona R., 1967 - Le concept de propriété en droit foncier de Madagascar. Paris: Cujas, Coll. Études malgaches, 306 p.

Weber J., 1995 - L’occupation humaine des aires protégées à Madagascar: diagnostic et éléments pour une gestion viable. Natures Sciences Sociétés, vol. 3, n° 2, p. 157-164.

Wily L., 1999 - Moving Forward in African Community Forestry: Trading Power, Not Use Rights. Society & Natural Resources, vol. 12, p. 49-61.

Haut de page


1 Many associations were established prior to the externally funded negotiation of community management contracts to control charcoal production in and beyond the Park’s buffer zone.

2 Before 2002 when the National Park was established, 2 150 inhabitants used to live on lands inside the protected area and had therefore to be re-grouped in 12 controlled occupation zones (ANGAP, 2000).

3 Assisted by the British, Radama I pursued a policy of expansion of the Merina kingdom to strategically important coastal regions including the Sakalava kingdoms, in the period between 1810 and 1828, and was the first Merina ruler to be recognised internationally as “King of Madagascar”.

4 To preserve the remaining forest and to stem the influx of more and more migrant cultivators from the South and South-East of Madagascar, Conservation International (CI) designed a management plan in 1996. In 2005 the Reserve was transformed into a National Park and responsibilities transferred from CI to the National Association for Management of Protected Areas (ANGAP).

5 Despite the revisionist tendency in the political ecology literature, it appears that Madagascar’s last remaining natural forests are cleared at the expense of both its exceptional biodiversity and the long term sustainability of rural economies, although it cannot be assumed that the two problems are identical. Peasant agriculture in humanly transformed landscapes may be ecologically sustainable yet incompatible with maintaining high rates of species endemism (cf. Laney, 2002; Muttenzer, 2010a; Pollini, 2007).

6 The bureaucratization of village and inter-village social relations by forest user associations has been noted by several scholars having studied the impacts of global environmental norms on customary orderings of territory and landscape (Blanc-Pamard and Rakoto, 2007; Corson, 2011; Goedefroit 2006; Pollini, 2007). The debate is whether the observed bureaucratization phenomenon is a valid indication of increased state control over local resource access and property relations or whether it simply points to the fact that potentially adverse effects of increased state control are being sidestepped by local people.

Haut de page

Pour citer cet article

Référence papier

Frank Muttenzer, « Community forest management on the agricultural frontier : charcoal makers, immigrant associations and land claims in Ankarafantsika, North-West Madagascar », Les Cahiers d’Outre-Mer, 258 | 2012, 249-272.

Référence électronique

Frank Muttenzer, « Community forest management on the agricultural frontier : charcoal makers, immigrant associations and land claims in Ankarafantsika, North-West Madagascar », Les Cahiers d’Outre-Mer [En ligne], 258 | Avril-Juin 2012, mis en ligne le 01 avril 2015, consulté le 29 mars 2017. URL : ; DOI : 10.4000/com.6604

Haut de page


Frank Muttenzer

PhD, Department of ethnology, University of Luzern, Frohburgstrasse 3, Postfach 4466, 6002 Luzern ; mél :

Haut de page

Droits d’auteur

© Tous droits réservés

Haut de page
  • Logo Presses Universitaires de Bordeaux