Skip to content
Environmental programmes with refugees: abandon the blanket approach?

Approaches to environmental problems in refugee emergencies have traditionally focused on two main areas: promotion of tree planting and dissemination of fuel-efficient stoves. This is done with the intention of increasing wood supply and simultaneously reducing the level of demand. Such activities are relatively easy for non-specialists to implement. They also produce two visible (and hence quantifiable) assets in a short space of time: trees and stoves.

A case study from western Tanzania illustrates the diminished utility of such approaches where natural resources are locally abundant. Tree planting is of questionable ecological value under such circumstances and stoves have had little effect on reducing wood demand. Cost-effectiveness of both activities is low. Therefore a blanket promotion of similar activities is of uncertain value.

An alternative approach is proposed. Environmental strategies can be based on a fuller consideration of the actual local resources on a camp-by-camp basis. Where refugees are in resource-rich areas, then the environmental imperative is for policies that protect, enforce and regulate. Where refugees are located in resource-deficient areas, programmes can concentrate on support, education, assistance and environmental awareness-raising with the refugee community.

There are 235,000 refugees from Burundi and the Democratic Republic of Congo housed in eight camps across Kigoma Region in western Tanzania. The majority having arrived from war-torn countries in late 1996, these people now make up 20 per cent of the region's total population. The camps are located in a narrow strip of land between the Burundi border to the west and a block of protected reserves to the east.

Two of these camps, Mtabila and Moyovosi, lie in areas already degraded over a long period by the local population, while the remaining six are in sites further east that were relatively undisturbed prior to the refugee influx. Nduta and Mtendeli camps are two such camps. Being located in dense miombo woodland adjacent to the Buyungu Forest Reserve and Moyowosi Game Reserve, they provide a useful environmental contrast with the more degraded sites.

A series of environmental data collection exercises have been commissioned by UNHCR, CARE and the Royal Dutch Embassy and these have made it possible to compare the resource rich Nduta/Mtendeli camps with the resource poor Mtabila/Moyovosi camps. This paper is based on the findings of this research, and particularly upon natural resource surveys conducted by CARE between September 1997 and March 1998.

The paper addresses issues related to the technical approach followed in the refugee environment programme in Kigoma and is not intended to compare the merits of different implementation systems, institutional structures or participatory methodologies. Nevertheless, the importance of identifying and working with local and community-based institutions to promote a situation of efficient, sustainable environmental management should not be overlooked.

Nduta and Mtendeli

Nduta and Mtendeli camps have a combined refugee population of approximately 60,000 Burundians. They are located in uninhabited miombo woodland, sporadically cleared by Tanzanian farmers practicing shifting cultivation prior to the villagisation period of the 1970s, but undisturbed since then, allowing for over 25 years of regeneration. The average weight of usable wood in these forests is an exceptional 120 tonnes/hectare.

Mtendeli camp is located on the boundary of the Buyungu Forest Reserve, which acts as a buffer for the Moyowosi Game Reserve 3 km further east. Nduta lies 4 km from the Moyowosi reserve.

The reserves have both ecological and economic significance. The 20,000 sq km Moyowosi ecosystem is believed to contain 20 per cent of the world’s shoebill stork population (listed as “suspected to be threatened” by the International Union for the Conservation of Nature???). Additionally, it is one of the few protected habitats of the sitatunga antelope and is a major transit corridor for large mammals such as elephant. It generates $170,000 in annual hunting revenues for the government of Tanzania. The Buyungu Forest Reserve is an important protective buffer for this Game Reserve, limiting eastwards expansion of cultivation by local people and illegal poaching and timber extraction. It also generates nearly $20,000 annually in government royalties for the collection of wood, honey and beeswax.

The large refugee populations have brought a variety of threats to these natural resources. Cultivation is expanding rapidly around the camps, thinning the forest cover and converting the sensitive areas adjacent to the reserve to agricultural land. Cultivation along watercourses disturbs downstream supplies and increases the risk of soil erosion. Cheap refugee labour is employed in unlicensed and illegal charcoal and timber operations, both of which result in further loss of ecologically and economically valuable trees. Poaching of game meat to supply refugee markets poses a potential risk to wildlife populations up to 30 km inside the game reserve.

Mtabila and Moyovosi

Mtabila and Moyovosi are adjacent camps and contain 84,000 Burundian refugees. They are situated some 12 km west of the nearest forest reserve and over 40 km from the Moyowosi Game Reserve. The environmental situation here is very different.

There are three Tanzanian communities within 5 km of the camps with a combined population of 22,000. For several decades these communities have exploited the area for grazing, farming and extraction of wood products. The area is characterised by large open spaces cleared of trees, remnant forest patches and plots of land cultivated on a rotational basis. While some of the valley bottom land is relatively fertile the area is, on the whole, characterised by natural resources of low economic and ecological value. The forests have historically been depleted of the largest and most valued species, while the wildlife has long since moved eastward to more remote and undisturbed habitat.

Environmental programmes

Environmental activities in each camp in Kigoma are implemented by the refugee community services and camp management agencies, under UNHCR coordination and funding and with technical support from CARE. All agencies are expected to work in collaboration with the Tanzanian government’s Natural Resource Officers and their staff at district level. Few management responsibilities have been delegated to local community institutions, in large part due to their non-existence or lack of capacity.

The environmental approach of agencies and local government has been uniform across the four camps described. The principal elements are: promotion of household tree planting, dissemination of improved cooking systems and protection of standing trees.

Tree Planting: The tree planting component is centred on agency-run tree nurseries in each camp which employ refugees and local people to produce seedlings. The tree species raised reflect a mixture of agroforestry and woodlot varieties, with a prevalence of Eucalyptus, Casuarina, Pinus, Leucaena, Grevillea, Black Wattle, Papaya and Passion Fruit. Seedlings of all types are distributed freely to refugee households for planting on garden plots and cultivated land around the camps.

Stoves: Improved stoves constructed of mud, ash and straw are promoted in the camps through a programme of training and awareness-raising. The programme includes the introduction of improved cooking techniques to save energy, such as the use of lids on pots, pre-soaking of hard beans and pounding of maize grain prior to cooking.

Tree protection: A network of approximately 30 forest guards in each camp guides refugees to designated areas for the collection of firewood and building poles, while attempting to protect standing trees and regenerating stumps within the immediate vicinity of the camps.

Impacts of environment programmes

The effectiveness of the environment programmes has varied across the camps. There has been notably more success in achieving environmental goals in Mtabila and Moyovosi (in degraded areas) than in Nduta and Mtendeli. This may come as a surprise. It might be expected that in a damaged area environmental management programmes might be less successful in controlling further degradation than in a relatively undisturbed area under the influence of new human pressure, whereas the converse seems to have been the case.

Tree planting

The tree planting programme has been relatively easy to implement, monitor and account for in all four camps. Material assets are produced in the nursery and distributed. Every aspect can be quantified, which may be desirable in a relief-oriented programme with short-term planning cycles. True environmental benefits, however, have been more difficult to identify.

In the degraded camps (Mtabila and Moyovosi) refugees have generally been willing to plant and nurture seedlings that they have received. Survival rates and community participation are higher. Trees are being mixed with horticultural crops on garden plots. People have engaged in direct sowing of their own Sesbania sesban seeds. There is a perceived need to plant trees to grow products for domestic use and for sale – such as poles, firewood and fruit.

In the forested camps (Nduta and Mtendeli) there has been a lower seedling survival rate and less active participation in seedling management. Important questions have been raised over tree ownership, protection responsibilities and harvesting rights. Additionally there are serious ecological concerns over the practice of planting economically preferred exotic species in an indigenous woodland and its impact on future succession and species mix. Fire is a major threat to these exotic species in the event of the refugees’ repatriation or re-settlement.

Applying a standardised forestry approach across camps in markedly different ecosystems has brought significantly different results.

Stove promotion

The energy efficiency campaign has resulted in high adoption rates of improved stoves in all camps. Seventy-one per cent of households have stoves in Nduta and Mtendeli, and 75 per cent in Mtabila and Moyovosi. However, in common with the tree planting programme, such statistics make for simple monitoring at only a superficial numerical level. Simply counting the number of stoves fails to reflect the different energy efficiencies actually achieved at household level and the environmental benefits accrued.

An important observation here is that a stove will not be used efficiently if the conditions under which it is used are not conducive to fuel efficiency. Thus in Mtabila and Moyovosi, where firewood is in relatively short supply, the refugees have achieved a considerable reduction in fuelwood consumption to an average per capita weight of 1.8 kg per day. In the forested Nduta and Mtendeli camps, the corresponding figure is a drastically higher 4.3 kg per person per day.

The straightforward promotion of fuel-efficient stoves seems to have had much less influence on energy consumption than has the sheer availability of firewood. Even though over 70 per cent of the households in all camps have built stoves, those in the resource-rich areas consume an average of 140 per cent more firewood. The siting of the camps has pre-determined the likely energy consumption, and stove construction campaigns have had little influence on this pattern of energy use, other than enabling refugees to save energy if they so wish.

Tree protection

Protection activities, though significant, have been relatively low profile in comparison with the tree planting and stove programmes. One reason for this is that protection activities do not produce new assets. They retard the process of forest clearance and as such merely ensure a 'least negative’ level of impact. The activities are less visible and much harder to quantify and monitor than stoves and tree seedlings. In Nduta and Mtendeli, the presence of forest guards has been crucial in the protection of trees both within and outside the camp areas. In Mtabila and Moyovosi they have served more as facilitators of refugee access to distant fuelwood and building pole resources on a rotational basis. The number of forest guards is approximately the same in each camp and has not depended upon the value of assets to be protected.


Environment programmes in the Kigoma camps have been remarkably similar from one location to another. A blueprint has been followed that focuses mainly on tree planting, stove promotion and tree protection. This blueprint has paid little attention to local ecological differences that might dictate other priorities. The result is inefficient use of funds on activities such as tree planting within forested areas, stove promotion in situations of abundant firewood where wood users lack incentives to conserve, and a standardised resource protection approach across camps with natural assets of significantly differing values.

An alternative to trees and stoves?

The Kigoma case study suggests that a different outlook to environmental programmes might be worthwhile. A camp-by-camp reconsideration of supply, demand and protection of natural resources should be made to reach environmental goals. It could save on inefficient expenditure and better respond to environmental threats.

Broadly speaking, a division can be made between camps with abundant natural resources and those with limited natural assets. The project approach in each should be philosophically and operationally distinct.

Camps with abundant natural resources

Where natural resources are abundant, the promotion of tree planting and fuel-efficient stoves runs into serious constraints. Refugees show little interest in either conserving or replacing assets that are freely available in large quantities. Education and training programmes have failed to confront successfully the physical reality of abundant resources available at no (financial) cost to the refugees.

Under such circumstances it is argued that the principal option available is to boost protection, enforcement and regulation of access to resources. Meaningful deterrents must be put in place to create a situation of restricted access to resources that are physically abundant. This may involve systems of forest guards, alongside concerted support to government and local institutions to better enforce regulations on wood cutting, charcoal making and poaching.

The emphasis is on control and containment of refugee access to natural resources, a situation to which they will inevitably respond with greater efficiency in care and use.

Camps with scarce natural resources

Where natural resources around a camp are already degraded the focus of environmental programmes should be radically different. Refugees should be helped to gain access to limited forest products on a sustainable basis. Guided cutting in carefully identified source areas can help meet domestic demand in an environmentally sensitive manner. Environmental education can supplement existing knowledge on energy-saving techniques. The emphasis is on the replenishment and efficient use of depleted resources by the refugees.


Environment programmes could benefit from leaving behind the traditional focus on tree planting and stoves. Though simple and easily monitored from the point of view of hardware dissemination, such programmes are unlikely to bring real benefits where resources around a settlement are abundant. Environment programmes should be adaptable, locally specific and based on a long-term vision for each refugee settlement site. Blanket regional policies are not helpful in this regard.


Greg Grimsich is former Project Manager with CARE International’s Kigoma Environmental Management Project, implementing a range of environmental activities in and around refugee camps in western Tanzania (email

Matthew Owen is a freelance environment consultant based in Nairobi, specialising in the mitigation of the impact of refugees on natural resources (email

This site is registered on as a development site. Switch to a production site key to remove this banner.