A Comparative Review of Agroforestry Practices in Two Forest-adjacent  Parishes on Mount Elgon, Uganda

Learning from the failings of previous technological and core-based approaches to conservation and rural poverty, many reversals have taken place during the last decade in rural development agendas. The approach which has emerged has attempted to move away from previous seasonal, gender, accessibility, wealth and power biases, in which the most disadvantaged people were often overlooked. It has been recognised that if long term conservation and development are to succeed, local people must be involved directly and actively at all levels of the research, management and decision making process:

The Mount Elgon Conservation and Development Project (MECDP) was set up in 1987, and its current aim is to "promote community development and conserve Mount Elgon’s ecosystem for present and future use" using a "community based resource management approach" involving the participation and empowerment of local communities in the development process (MECDP, 1995). Working in conjunction with Mount Elgon National Park (MENP), park regulations have been formulated with reference to the needs of local people and their resource use levels, and enforced in conjunction with a comprehensive education programme (family planning and agricultural extension programmes have been established in the majority of forest-adjacent parishes). Collaborative management has been piloted in two parishes, with the aim of extending it to all forest-adjacent parishes before the project ends in 2000. Uganda National Farmers Association (UNFA) and UWA-Face (Forests Absorbing Carbon dioxide Emission Foundation) are also active in the area.

Population density is high in the majority of parishes adjacent to the forests of MENP, and land shortage and fragmentation are an increasing problem as the population grows (Scott, 1994). Agricultural encroachment was widespread until authorities began to move people out of the forest during the 1980s and 1990s. Plans are underway to move the small number of remaining forest dwellers from Benet parish. However, people from forest-adjacent parishes are continuing to graze cattle and collect live wood from the forest. Current grazing by cattle in Benet and grazing which took place in the past has been shown to alter the composition of plant and animal communities considerably (Cameron, 1997; McPharland & Opus, 1998; Reed and Clokie, sumitted). In particular, grazing in the forest has been found to suppress tree regeneration. Tree growth continues to be suppressed in areas which have not been grazed intensively for fifteen years, due to dense colonisation by the woody herb Mimulopsis alpina in these areas. This represents a serious threat to the continuation of forest in grazed areas.

Agroforestry has the potential to provide a sustainable supply of tree products which were formerly harvested from the forest, as well as improving the sustainability and productivity of local agriculture. In conjuction with family planning and the spread of other modern production techniques, agroforestry can play a significant role in reducing encroachment and is a key component of the MECDP extension programme. The aim of this study was to analyse the contribution of trees to the production strategies of local farming households in forest-adjacent parishes, in the context of local agricultural systems, constraints to production and the work of locally active development organisations. This was done using elements of rural and participatory rural appraisal methodologies to gather information in the pilot parish, Tegres (where the project had worked for four years (1992-1996)) and in the neighbouring parish, Kwoti (in which the project had been working since the end of 1997 (three months before the study commenced)). The contribution of development agencies to changes in production strategies was evaluated, constraints to change were determined and the attitudes of farmers to the development process was assessed. It is hoped that this information will enable those working in the area to gain a fresh insight into the potential use of agroforestry in the conservation and development Mount Elgon, as well as providing an insight into the obstacles which may prevent this potential being realised.
 

The following objectives were formulated for the appraisal:

  to learn about the production strategies of local farming households and how they are constrained;

  to describe and analyse the contribution of trees to these strategies;

  to evaluate the role of locally active development organisations in changing these strategies;

  to determine constraints to change;

  to assess farmer perceptions of MECDP
 

In order to develop successful and adoptable agroforestry interventions as part of a rural development strategy, it is essential to base research on the knowledge and attitudes of local farmers. Rapid Rural Appraisal (RRA) (Khon Kaen University, 1987), and more recently Participatory Rural Appraisal (PRA) (Mascarenhas et al., 1991; Macarenhas, 1992; Chambers, 1992), have been developed to optimise the input of local people to the research and development process. The aim of this research was to provide information which could facilitate decision-makers and local people to make appropriate decisions within the current development process. Information was extracted using a range of RRA techniques, whilst applying the behavioural lessons learnt from PRA.

The parishes of Kwoti and Upper Tegres were comparable in their physical and human environments. However, MECDP had been active for different amounts of time in each parish (the project had only worked in Kwoti for three months in comparison to the four years of work in Upper Tegres). Upper Tegres was selected for study in preference to Lower Tegres due to its greater similarity to Kwoti in terms of area, altitude, climate, crops grown, population size and density, soil fertility and distance from the National Park. The slope and aspect of cultivated land in each parish was similar (approximately 4 degrees and predominantly north to north-west facing in both parishes). The same agricultural extension officer (AEO) and MECDP project co-ordinators carried out the work in each parish. Although Upper Tegres was more similar to Kwoti than Lower Tegres, there were differences between the two parishes, which are discussed in Section 3.4.

Data was collected over six weeks from 5 January to 13 February 1998, after a two week acclimatisation period in which the research team learnt about aspects of the local culture and language, and began to build relationships with the local community. Twenty farming households were sampled at random from each parish. This represented 7% of households from each parish. Semi-structured interviews (using a checklist) were conducted with the assistance of an interpreter (fluent in English, Kup-Sabine and Khi-Swahili). The interview checklist was approved and augmented by the MECDP Kapchorwa District Co-ordinator, and progressively modified throughout the interviewing process. Interviews were conducted with the main decision maker of the household, who tended to be male. Partners and others involved in the decision making process (e.g. the mature children of single parents) were interviewed whenever possible. Where the presence of these people was not possible throughout the interview, they were asked questions relating to problems and perceptions (see checklist (section F)). Each interview lasted between one and two hours, and was conducted in accordance with behavioural principles developed in the literature (e.g. expressing humility through body language, listening skills and the desire to learn from farmers) (Chambers, 1983). During the interviews, farmers guided the research team around the property. Where this was not possible, diagrams were drawn by the farmer. This served to bring to attention aspects of the production system which the owner may have otherwise overlooked or perceived to be unimportant. Interviewees were asked to rank their problems, and their perceptions of the development organisations.

Plot gradients were estimated on a visual scale (from 0 (flat) to 6 (steeply sloping)). Estimates of farm area were given by interviewees and corroborated by an estimate from the interpreter. Soil fertility was estimated using a variety of techniques: by calculating maize productivity in bags per acre, by verbal reports from farmers and by a visual assessments of soil colour, compaction and organic matter content. Soil erosion was estimated on a visual scale (from 0 (none) to 6 (severe)) and from verbal reports from farmers. Wealth was measured by calculating household incomes, and corroborated by visual indicators (presence/absence of an iron roof, condition of health and clothing). Visual indicators were scored on a scale from 2 (richest) to -2 (poorest).

Species identification was carried out in the field by the research team using the key in Katende et al. (1995). Vouchers were taken for any species which could not be identified in the field. Vouchers are currently being identified at Makerere University Herbarium where a number of specimens have been lodged permanently.

Five days of extended interviews with key informants in Kapchorwa and Mbale followed the work with farmers. The Training Officer from MECDP headquarters in Mbale who is responsible for much of the day to day running of the organisation was interviewed as well as the AEO for Kwoti and Tegres, MECDP project co-ordinators for animal and crop husbandry, soil conservation and agroforestry were interviewed, and representatives from UNFA and UWA-Face.

Location, climate and geology are described.

There are two districts in Uganda which are adjacent to the National Park. In Mbale District there are 5 counties, 15 sub-counties and 35 parishes bordering the forest, and in Kapchorwa District there are 3 counties, 8 sub-counties and 23 parishes bordering the forest.

Two main tribal groups live on the northern slopes of Mount Elgon. The Sabei, of Nilo-Cushtic origin, were originally pastoralists living in the forest between about 2500 and 3000m where they grazed their cattle, sheep and goats on pastures within the forest and on the high moor lands. The Bagisu, of Bantu origin, were originally agriculturists, and have gradually moved further up the lower slopes of Mount Elgon, having moved into the area in the sixteenth century (Were & Wilson, 1970). Bagisu are a small minority in the two study parishes. The Sabei are the only recorded traditional residents of the area gazetted as Mount Elgon National Park.

There has been a long history of forest utilisation by the local communities of Mount Elgon. The most commonly collected forest products are firewood, ropes, polewood, vegetables (including mushrooms), bamboo shoots, bamboo stems, crop stakes and grazing (Scott, 1994). The religious and cultural significance of the forest is still stressed by the local communities who use the forest for circumcision ceremonies and medicinal plants on a regular basis (Ransom, 1998). A group known as the Kony have lived in the forest until recent relocations moved the majority of these people out of the forest. They were a pastoralist community, grazing cattle, collecting honey and bamboo (for basketry) and hunting. Relocated Kony have had to learn agricultural skills. The majority of these people are content living outside the forest due to the improved educational and health facilities available (Ransom, 1998).

High immigration rates and encroachment proceeded unchecked throughout the years of civil unrest (17% of the reserve had been deforested by 1990 according to Howard (1991)) until the resettlement programmes took place in 1983 and 1990. A large area of Kwoti (6000 ha) was gazetted from the forest for the purpose of resettling forest dwellers in 1983. However, in 1990 the gazetted area was found to have been measured inaccurately, and the area was regazetted later that year. This left many people who had been allocated ground in 1983 without land, and these people have continued to farm the area illegally since this time, petitioning the government to reinstate the 1983 park boundary. Considerable confusion was caused over this matter by a government minister who visited the area in 1997. He promised people in Kwoti that the 1983 boundary would be restored, and permitted them to graze cattle in the forest until stability had returned to the areas from which many had fled from cattle rustling. He had no authority to make these statements, and has complicated the work of development organisations in the area.
 

The conservation of MENP is important for a number of reasons. The combined area of the Ugandan and Kenyan parks are sufficiently large to maintain viable populations of many of the larger and rarer species of large mammals which are vulnerable to extinction in smaller National Parks. Bird surveys conducted on the mountain (Van Someren, 1922, 1932; Britton, 1980; Pearson and Turner, 1986; Howard, 1991; Kings, 1997) show that avifauna of Mount Elgon is diverse and includes a number of rare and threatened bird species which are restricted to Mount Elgon and a few other East African mountains. In addition to this a number of studies have found a higher proportion of forest generalist species than specialist species, which can be attributed to the effects of human disturbance on forest bird communities (Katende et al., 1989; Mathews, 1996; McPharland & Opus, 1998). Overall, IUCN have listed 37 faunal species in the area as "globally threatened" (22 mammal, 2 insect and 13 bird species, of which nine species are endemic), making the area a priority for species conservation (IUCN, 1995a,b).

The majority of the plant species in the forest zone above 2000m have been shown to be endemic to the Afromontane Region (White, 1983), and a number of species in this zone are endemic to Mount Elgon (Van Heist, 1993). Mount Elgon is thought to have been a radiating point for a number of elements of the afro-montane flora (Coe, 1969).

The most recent and the most extensive survey of Mount Elgon’s biota (including vegetation, birds, butterflies, moths, and small mammals) was conducted as part of the Mount Elgon Biodiversity Survey (Davenport et al. 1996) between 1991 and 1995 by the Forestry Department as part of the National Forest Biodiversity Inventory Programme. As a result of this programme, Mount Elgon was provisionally ranked amongst the top ten most species rich forests and was identified as a priority for the conservation of Uganda’s small mammals (Davenport et al. 1996).
 

The conservation of Mount Elgon is vital to the economic functioning of an extensive surrounding area. The mountain plays a crucial role as a water catchment, supplying approximately one million people to the north and west with fresh water (Howard, 1991).

Educational facilities in the area are limited, as are job opportunities locally and externally. This leaves many young people with little choice but to depend on a small and usually inadequate portion of their parent’s land (Scott, 1994).

There is however a high potential for the development of tourism in the area. Since National Park status was given to the area, MENP and the Uganda Wildlife Authority (UWA) (which was formed in 1996 from the amalgamation of Uganda National Parks and the Uganda Game Department) have moved increasingly towards the use of tourism as a tool for development. The government targeted ecotourism in its Integrated Tourism Master Plan of 1993 because revenues from ecotourism are high in relation to the number of tourists involved. Amooti (1996) describes ecotourism as ‘purposeful travel that creates understanding of culture and natural history, while safeguarding the integrity of the ecosystem’. On the other hand, the fact that ecotourism reach places which were formerly untouched by outsiders increases the potential impact of tourism. In addition to this, the tourist trade is highly dependant on foreign markets (Cochrane, 1993). Nevertheless, MENP management view tourism as a potential source of funding which could sustain its conservation programme after donor funding has ceased Ransom, 1998).

Legislation was put in place in order that revenues from ecotourism would be used in rural development. Introduced in 1996, the Uganda Wildlife Statute states that,

 No payments have yet been made however, and few people know about the initiative. In MENP, those who knew about the new funding believed the figure was only twelve percent of park entry fees (Ransom, 1998). Revenues from tourism on Mount Elgon have been growing at an increasing rate according to information collected by MENP about visitors to the park between 1994 and 1996 and a personal communication about numbers in 1993 from Wagstaff in Scott (1994). The number of people visiting the park remained similar between 1993 and 1994 (approximately 300 visitors per year) but it increased by 60% between 1994 and 1995 and had increased by a further 65% by 1996 (850 tourists visited the park in that year) (Ransom, 1998).
 

Mount Elgon has been controlled by the Forest Department since 1929. It became Mount Elgon Crown Forest in 1940, and became a Forest Reserve in 1951. The main objective of the working plan for 1968-1978 (Synott, 1968) was to protect the forest, with the extraction of timber as a secondary objective. Since the restoration of civil stability in Uganda, the government has become increasingly concerned with conservation issues, and in 1988, a forest rehabilitation project was initiated on Mount Elgon. In 1992, the area was handed over to Uganda National Parks and renamed Mount Elgon National Park. Since this time, the MECDP have been assisting management authorities with in-forest and community issues, and more recently, UWA-Face and UNFA have been involved in the development process. In the last eight years, management policies have shifted from the protection and extraction of forest resources, to an increasing involvement of local communities in management decisions, and an emphasis on sustainable utilisation of resources. This culminated in a new management plan for the park in 1995. MECDP have commissioned a number of resource inventories and assessments to facilitate management planning. Katende et al. (1989) carried out a biodiversity inventory for woody perennials and birds and a Land Mapping and Biodiversity Survey of Mount Elgon National Park was carried out in 1993 to assist the development of a long term management plan (van Heist, 1994). The survey described numerous aspects of the mountain’s ecology with an emphasis on plant biodiversity. A "resource use assessment" was commissioned for the same purpose detailing resource use by people groups across the mountain through a series of semi-structured interviews and group discussions (Scott, 1994).
 

The Mount Elgon Conservation and Development Project was initiated in 1987 with three main long term objectives:

to ensure the conservation of the biological diversity and ecological processes within the natural forests
   of Mount Elgon, Kibale and Semliki;

to meet human needs on the land surrounding the forest on a sustainable basis;

to safeguard the quality and rate of flow of water to all areas affected by the forest.

The main work of the project initially concentrated on assisting the Government of Uganda to regain control over the protected area, and to protect and demark the boundary line which was established in 1963. As global attitudes to conservation began to shift, the emphasis of the project work began to shift, and in 1991 work with the local communities began to form a major component of the project. At end of Phase II (1993/94), the future direction of the project was assessed. As a result of this, sustainable forest resource use and the role of the community in forest management became key components of the project. The main aim of phase III (the current phase) was to "promote community development and conserve Mount Elgon’s ecosystem for present and future use". This aim has been broken down into five main objectives:

improved management and use of Mount Elgon’s ecosystem with the effective development of all
    interested groups;

community based resource management approach based on participation and empowerment of
    communities for development;

management and protection of MENP and its financial and economic viability to grow;

community undertaking of conservation and sustainable development activities outside the park with
    reduced resource use pressure on the park;

strengthening of the planning and management capacity of district and local councillors;

mechanisms put in place to monitor and evaluate the impact of critical activities which develop. 

In 1995, Uganda’s National Park Authority gave permission for collaborative forest management to be piloted in two forest-adjacent parishes. Both parishes had been involved in the resource use assessment of 1994. The roles and responsibilities of local people and National Park staff were negotiated, and a committee was elected to represent different categories of resource users. Collaboration between the committee and National Park staff over five months produced a draft agreement which was signed by both parties. By 1998, collaborative management was deemed to have succeeded in one parish and failed in the other (Lwanga, I. pers. comm.). The parish in which collaborative management was deemed to have failed had a long history of conflict with the Forest Department and National Park Authority, arising from persistent encroachments in the Reserve and later in the National Park (Onyango, 1996).

The work is co-ordinated from headquarters in Mbale, and is operational at a district and parish level. At the district level, there are co-ordinating officers for crop and animal husbandry, agroforestry and soil conservation. These officers train and co-ordinate a team of agricultural extension officers (AEOs). Both co-ordinating officers and AEOs divide their working hours equally between the Department of Agriculture and MECDP. Co-ordinating officers and AEOs spend 10 and 12 days each month in the field respectively, with the remainder of their time taken up predominantly with administration work. AEOs are selected from the local communities in which work will be done because they can more easily relate to the target group as equals, they will defend the interests of the community, and the community will feel more confident to criticise the project (MECDP Internal Communication, 1994).
 

The Role of UWA-Face and UNFA
 

The following text and photographs are taken from http://www.facefoundation.nl/:

"The project organisation UWA-Face was set up by the Uganda Wildlife Authority to implement reforestation projects. The projects provide work for over 1000 people during the planting seasons. Work started on building two central nurseries for the two separate project areas, Ngasire and Buwabwala, in August 1994. Each of these nurseries has a capacity of 300 000 trees. The supply of plant material from these project nurseries is gradually being taken over by local nurseries established as private enterprises by workers from the project nurseries.

The border between the agricultural area and the forest reserve is at an altitude of roughly 1900 metres, but had become indistinct because of the activities of the neighbouring people. It was therefore resurveyed over a 120-kilometre stretch during the 1990-1994 period and defined by planting a 10-metre-wide strip of recognisable fast-growing eucalyptus trees. It is now clear for everyone where the National Park begins."

The Uganda National Farmers Association (UNFA) is a governmental organisation under the Department of Agriculture, with additional assistance in development issues from Investment Development Export Agriculture (IDEA). UNFA started work in Kapchorwa District in 1993, and now has four hundred members throughout the district (including parishes which are not adjacent to the park). The main aim of the organisation is "to improve productivity and meet farmer’s needs". Their approach is "demand driven", requiring farmers to approach them for assistance and pay for work done. Only after paying the UNFA membership fee can farmers make use of the association’s services. Workshops take place in each parish (82 per year throughout Kapchorwa District at planting time, at the first and second weeding times, and at harvesting time) teaching agricultural and agronomic skills and recruiting members. UNFA teach a wider range of subjects, particularly in the field of farm management, and tend to cover subjects in more detail than MECDP. They do however reach fewer farmers than MECDP. Although membership figures were not available at the time of the study, but district management stated that membership was "low" in Tegres and "very low" in Kwoti.
 

Kwoti and Tegres are situated on the northern slopes of Mount Elgon adjacent to the National Park at the altitudes of 1900-2000 m and 1700-1900 m respectively. Tegres is a much larger parish than Kwoti, covering approximately 5 km² in comparison to 2 km² for Kwoti. Tegres is commonly divided into Upper and Lower Tegres, due to the natural differences between these areas. A higher proportion of land in Upper Tegres has recently been brought into cultivation from the forest, so soils tend to be more fertile than in Lower Tegres, a lower proportion of people in the area grow matoke (Musa sapientum), and people are more reliant on the forest for tree products and grazing. Although Upper Tegres is slightly smaller than Lower Tegres, the population size is approximately equal in each part.

Soil erosion has become a problem in both parishes due to the high population and the gradient of the land, compounded by poor soil conservation by most farmers. Levels of soil erosion were similar in Kwoti and Upper Tegres. Soil erosion was positively correlated to gradient in 60% of sample farms in each parish. The average land-holding size (including rented lands) was 2.0 acres in Kwoti (range: 0.25-4 acres) and 4.7 acres in Upper Tegres (range: 0.25-35.00 acres). Although this would suggest that land-holdings are significantly larger in Upper Tegres than Kwoti, the figure for Upper Tegres included one farm of 35 acres which was unusually large. If this farm is excluded from the data, the average land-holding in Upper Tegres was 2.7 acres, which is only slightly larger than the figure for Kwoti. Whilst there was a clear positive relationship between land-holding size and income in Upper Tegres, no such relationship existed in Kwoti. This may be accounted for by the small sample size of sample households earning over one million Uganda Shillings (10% of households). The gradient of farmland in Kwoti and Upper Tegres was also very similar.

There were two distinct soil types in Upper Tegres which were most easily defined by colour. Soils in farms nearer the forest were light brown in colour and known to be fertile. Soils further from the forest were orange-brown in colour and known to be much less fertile. This difference can probably be accounted for by the different amounts of time that these areas have been used for agriculture. From the community’s history of gradual encroachment into the forest, it can be postulated that areas closer to the forest will have come into agricultural use most recently. In comparison, Kwoti was gazetted from the forest in 1983 and the soils in all farms have been cultivated for the same amount of time. Apart from the degree of compaction there were few visual differences between the soils of farms in Kwoti, between farms in Upper Tegres. Soils were generally more fertile in Kwoti than in and there was less difference in the maize yields of farms in Kwoti than there was Upper Tegres. Soil fertility in Kwoti was lowest in farms that experienced the most soil erosion. The relationship between soil fertility and erosion was not clear in Tegres.

The majority of farmers in both parishes acknowledged a decline in the yields of their main crops since they had started cultivating the area (44% and 55% had observed declining yields in Kwoti and Upper Tegres respectively) (Figure 1). In Upper Tegres the most probable explanation for this was the extended time over which the majority of farms in the parish had been cultivated. The main reason given to explain declining yields by farmers in Upper Tegres was over-cultivation of maize and the absence of rotation with other crops. In Kwoti, the main reason was soil erosion. Data collected for this study showed that farms with planted contour bunds have lower levels of soil erosion. As there are very few farms with planted contour bunds in Kwoti, the lack of erosion control measures in the parish is probably a significant factor contributing to the decreasing productivity of the area.

Although a paved road is in the process of being built from Mbale to Kapchorwa, progress has been slow, and the road is not expected to be complete for many years. The lack of proper roads makes markets for maize (Zea mays) and other produce inaccessible for much of the year, and has prevented large-scale marketing of fruit and vegetables. Access to the inputs such as seeds, fertilisers and other materials necessary to practice intensive agriculture is impeded by road conditions. The majority of farmers purchase hybrid maize seeds which are imported from the Kenyan Seed Company in Kitale. Delays in the arrival of seed stocks occur commonly, and detrimentally affect maize yields.

The main differences between the human environments of each parish were that the population density was greater in Upper Tegres (329 people per km²) than Kwoti (134.8 people per km²), and the community in Kwoti was composed mainly of immigrants (80%) whereas immigrants were in a minority in Upper Tegres (20%). Immigrants had stayed in Kwoti for an average of 13 years. The majority of these people had moved to the area due to cattle rustling activities at their previous homes, although some were attracted by the fertile soils and forest grazing. They had most commonly come from Sipi, Bukwa, Kaproron and Kapteret. Immigrants in Upper Tegres came from a wide range of places, and most of them had moved in preference to their previous residence. Immigrants had stayed in Upper Tegres for an average of 8 years. 1806 people currently live in Kwoti and 1876 in Upper Tegres (based on figures from the 1991 Population Census compounded at an increase rate of 2.5% (national average at time of census) to give an estimated figure for 1998). The growth rate is similar to national rates. The average size of household was 6.3 in Kwoti and 6.2 in Upper Tegres (there are 287 households in Kwoti and 303 in Upper Tegres). The landless population is difficult to estimate, but is thought to be increasing in the area.

The tribal composition of the two parishes was very similar, with Bagisu being a small minority in both parishes. The majority of people in each parish belonged to the Sabei tribe. There was very little difference in the demographic composition of the sample households in either parish. Figures for each parish were therefore combined to show the age and gender structure of the communities studied. The age structure, dominated by people under the age of fifteen, is typical of most developing countries (Erlich and Erlich, 1990). Family planning education was conducted by governmental organisation Reproductive Educative and Community Health (REACH) in Upper Tegres from 1992, and started in Kwoti at the end of 1997. The majority of couples in Kwoti were either intending to use family planning in the future (35%) or not practising family planning with no intention to start in the future (35%) (Figure 2). The majority of couples in Upper Tegres were intending to use family planning in the future (40%). 25% of couples were practising family planning and 15% were not practising, and had no intention of starting in the future. Fertility has traditionally been valued highly, and there is little indication that this has changed significantly. Polygamy is valued and is a goal for many men, but monogamy is the most common marriage pattern.

People in Upper Tegres were better educated than people in Kwoti. Males and people under the age of forty were better educated in both parishes. There was no correlation between the level of education reached by decision makers in households and the annual income generated.
 

Marketing plays a crucial role in land use and agricultural development. As agricultural production systems develop and specialisation increases, wider markets are sought, and marketing systems become increasingly important (Kenya Web, 1998). National and local market prices are highly variable seasonally and annually. Coffee is bought by a local co-operative society and sunflower seeds are purchased by local industry. A lack of local market outlets for maize and the poor quality of roads force many farmers to rely on middlemen to market their produce. The effects of this on the profitability of maize growing is compounded by the unpredictability of production inputs (seeds from Kenya), outputs and prices. There is no external marketing system for milk, although it is marketed locally. The economy of Kwoti is more highly dependant on the forest than Tegres, but the economy of Upper Tegres is more dependant on the forest than that of Lower Tegres. Other than providing a cheap source of firewood, ropes, polewood, vegetables (including mushrooms), bamboo shoots, bamboo stems, crop stakes and grazing, some of these resources are marketed. Products which are low in availability or involve high risk to obtain are most marketable. Building poles for example, are difficult to obtain as deadwood, and risky to obtain as live wood due to National Park restrictions, and are commonly sold for relatively high prices.

The economy of both parishes is lead by agriculture, making incomes in the study area highly variable both annually and seasonally. The average income in sample households of Kwoti at the time of the study was USh 314,447 (£170) (range: USh 50,000 (£27.03) - USh 1,091,500 (£590.00)) and USh 1,176,973 (£636) in Upper Tegres (range: USh 25,000 (£13.51) - USh 7,540,000 (£4,075.68)). Considering the broad range of incomes in Upper Tegres, the figure for average income should be treated with caution. Households were considerably less wealthy than households in Upper Tegres however. This is supported by data collected from visual indicators, as households in Kwoti had an average score of 0.1 and households in Upper Tegres had an average score of 0.8 (for indicators, see Section 2.2).

Labour demand is highest at weeding and harvesting times. Labour is readily available at these times of the year. However, at financially secure times of the year (e.g. after harvest) there are very few people in the labour market. The unavailability of labour after harvest time has limited the cultivation of sunflowers in both parishes.

Households in Kwoti and Upper Tegres hire on average 7 and 9 labourers per year respectively. Labourers are most commonly hired to weed the maize crop. Weeding takes place twice between April and July each year. The use of labour in the production system is limited by income. Therefore, poorer households can afford to employ less labour than wealthy households (Figure 3). In Upper Tegres, larger farms use more labour than smaller farms, but there is no clear relationship between land size and labour use in the farms of Kwoti.
 

The scale of agricultural production in the area is small. With the exception of one farm with thirty acres of land in Upper Tegres, the average land-holding size is under three acres in each parish. There were few internal linkages between components of the production systems studied, although cattle manure was commonly applied to cultivated soils. In particular there was little interaction between trees and animal or crop systems (the exception to this was the traditional use of shade trees in banana plantations). External linkages existed with common property resources. The forest, common land and fallow fields after harvest were commonly utilised for grazing, and represent an important source of nutrients in the production system. Climatic variations occur seasonally and annually and cause production output to vary unpredictably. The production systems in the area are susceptible to a range of risks from factors such as climate, prices, social and physical disasters and sickness. Many farmers had incomes from other sources to subsidise financially difficult times of year. These enterprises consisted of shops, sale of building poles from the forest, labouring, sale of herbal medicines and various part-time professions such as birth attendance and circumcision.
 

Maize was the main cash crop, followed by coffee and matoke in Upper Tegres, and potatoes and wheat in Kwoti. It is usually not possible to grow maize profitably unless over ten acres is cultivated (Cheptagei, W. & Lwanga, I. pers. comm.). The majority of farmers are therefore working at a loss, although few them acknowledge this. The preference for maize may be due to dietary preference or ease of cultivation (farmers are highly familiar with the cultivation of this crop). Late arrival of seeds, poor marketing and transportation systems and the unpredictability of climate and prices can further reduce the profitability of maize.

Maize was the most significant source of income in both parishes. The most significant difference in household income sources between the parishes was the importance of coffee and matoke in Upper Tegres. The importance of matoke in Upper Tegres may be due to the inclusion of matoke cultivation as a education topic in the first year of extension work by MECDP. Income generating projects such as this were dropped from the project in 1993 in favour of concentrating on topics which were more directly linked to forest conservation. Matoke is a subsistence crop in Kwoti. The three most important income sources in Kwoti were maize, potatoes and wheat, and in Upper Tegres were Maize, coffee and matoke. People in Kwoti relied more on maize than other sources for income than people in Upper Tegres. In addition to this, people in Kwoti were dependant on fewer crops for subsistence and income than people in Upper Tegres. In Kwoti, the average farm had five crop species, whereas in Upper Tegres the average farm cultivated seven species. Table 1 shows that households in Upper Tegres also grew a wider range of minor crops. Not suprisingly, maize is planted over the largest area, but the area of coffee in Upper Tegres was small in relation to its economic importance (Figure 4).

There was little mechanisation in any of the systems studied, and there are no reports of mechanisation throughout either parish. The only mechanisation seen was spray pumps for fertilisers and pesticides.

Pesticides are used by a slightly higher proportion of people in Upper Tegres (45% in Kwoti and 53% in Upper Tegres). In Kwoti they were predominantly applied to potatoes, but in Upper Tegres, pesticides were applied to a range of crops. In Upper Tegres, pesticides were most commonly applied to potatoes and green vegetables, and less commonly applied to beans, coffee, maize, tomatoes and wheat. There was no correlation between income and pesticide use in either parish. In Kwoti, farmers that had been educated to secondary level were twice as likely to use pesticides as those who had no formal education, and farmers who had been educated to primary level were four times as likely to use pesticides as those who had no formal education (Figure 5).

Farmers in Upper Tegres who had reached secondary level were more likely to use pesticides than farmers that had reached primary level (55% of primary educated and 78% of secondary educated were using pesticides). In Upper Tegres, the majority of pesticide users had spoken to an AEO, whilst in Kwoti, pesticide users that had spoken to an AEO were in a minority (Figure 6). This is probably a reflection of the lower number of people that had spoken to the AEO in Kwoti. This might therefore suggest that pesticide use has spread to Kwoti from users in Upper Tegres or other parishes where usage is high. Pesticide use declined with age in Kwoti, whereas pesticide use was lower in the youngest and oldest age groups, and highest amongst people in their forties in Upper Tegres (Figure 7). When compared with national statistics for the use of pesticides and fertilisers, usage in Kapchorwa District is low. A recent FAO report (FAO, 1997) stated that Kapchorwa District had one of the lowest rates of fertiliser and pesticide use in Uganda’s eastern region, and blamed this for the low agricultural output of the district in relation to other districts of Uganda.

Fertilisers were not used by any of the farmers sampled in Kwoti. 10% of farmers in Upper Tegres were applying fertilisers to their maize crop and obtaining yields which were between 30% and 65% higher than average yields. There was no correlation between income and fertiliser use amongst these farmers, however the number of farmers using fertilisers in Upper Tegres was low, and a larger sample might show a correlation. Farmers using fertilisers had reached secondary level education. 74% and 88% of farmers in Kwoti and Upper Tegres respectively were applying organic fertiliser (animal manure) to their land. Manure application was generally restricted to cattle owners, and as 70% and 80% of people in Kwoti and Upper Tegres respectively owned cattle, this could explain why more people in Upper Tegres were applying manure.

In Upper Tegres, half of the sample farms situated on sloping land had contour bunds planted with Napier Grass or S. sesbans(and in one case the shrub "Agamiandet" (Plectranthus sp.)) whereas in Kwoti, only 18% of farms situated on sloping land had situated contour bunds and planted them with Napier Grass. In Upper Tegres, maize fields with planted contour bunds were almost twice as productive as maize fields without contour bunds (15.6 bags per acre as opposed to 8.6 bags per acre (average maize productivity in Upper Tegres was 12.1 bags per acre)). Maize fields with contour bunds in Kwoti were also more productive, although to a lesser extent (16.7 bags per acre as opposed to 12.8 bags per acre (average maize productivity in Kwoti was 14.2 bags per acre)). Planted contour bunds were found to decrease soil erosion. A third of farms in Kwoti with contour bunds (planted with Napier Grass and/or S. sesbans) were experiencing reduced soil erosion as a result, and half the farms in Upper Tegres with contour bunds planted with Napier Grass were experiencing reduced soil erosion.

Contour bunds situated in Upper Tegres may be more effective in soil conservation than those situated in Kwoti for two reasons. Firstly, some of the farms in Upper Tegres had planted the nitrogen fixing tree S. sesbans which could explain the increased productivity of maize fields with planted contour bunds in Upper Tegres. Secondly, local government in Upper Tegres have passed and effectively enforced laws to restrict free-range grazing by cattle. As a result of un-checked free-range grazing by cattle in Kwoti, Napier grass is constantly grazed and was generally less well established than Napier grass in Upper Tegres. Free-range grazing also acts as a disincentive to plant Napier grass. Alternatively, Napier grass in Upper Tegres may be better established due to its greater age. This cannot be confirmed as the age of Napier grass stocks was not established in any of the farms studied. A third of the farms that claimed to have contour bunds actually had "acres". These are low ridges of earth formed to demark each acre of land owned (for tax purposes). They are usually colonised by a range of local grasses, of which Kikuyu Grass (Pennesetum clandestinum) is usually the most prolific.

The majority of farmers in Kwoti were practising rotation (70%). None of the farmers that were not practising rotation had plans to start rotating crops in the future. The most common rotations was maize with potatoes or wheat. A similar proportion of farmers were rotating their crops in Upper Tegres (75%), but a significant proportion of those who were not rotating crops had plans to do so in the future (40%). There were many different rotations used, involving most crops. Beans and sunflowers were commonly used in rotation, due to the ability of these species to improve soil fertility.

A quarter of farmers in Kwoti were mulching crops, and in Upper Tegres almost half of the farmers were mulching (45%). In Kwoti maize was the most common species to be mulched, usually with maize stems, leaves and cobs. Maize was commonly mulched in Upper Tegres also, but usually with the leaves of nearby trees (most commonly Ekebergia ruppeliana and Markhammia lutea (which are relatively abundant in the parish) and Ficus natalensis (in another East African study, crop yields were found to be higher near F. capensis (Balderrama et al., 1988).). Tomatoes were commonly mulched with maize stems and matoke leaves, and matoke was commonly mulched with matoke leaves, wheat straw and the leaves of various nearby trees.

Few farms in Kwoti or Upper Tegres used irrigation (15% and 20% respectively). The only irrigation system in use was the manual removal of water from rivers to sprinkle on tomatoes, green vegetables or nurseries, and was restricted to farms that had streams running through them.

20% and 40% of farms in Kwoti and Upper Tegres respectively had fallow land. One farm in Upper Tegres had enriched the fallow by planting S. sesbans. Otherwise, fallow land was grazed and dominated by grasses. Kikuyu grass (Pennisetum clandestinum) was most abundant with Andropogan amethystinus, Carex sp., Digitaria scalarum, Eragrostis ternuafolia, Juncas sp. and Poa annua occurring in varying lower abundances. Depending on the length of the fallow, there were occasional herb, shrub and pioneer tree species in addition to grass.
 

The most commonly kept domestic animals in Kwoti in descending order of abundance were goats, sheep, pigs and donkeys, whilst in Upper Tegres these were goats, pigs and sheep. Donkeys are primarily used for transport, but the other animals were mostly kept for sale. Money from the sale of livestock was rarely invested in agricultural improvements. Proceeds were most commonly used to pay school or medical expenses. Most farms kept chickens, but rarely for commercial purposes. Most farms also kept dogs for guarding, and on occasions for illegal game hunting in the National Park. MECDP has been promoting the use of exotic and cross bred cattle in Upper Tegres, and is beginning to start the promotion in Kwoti. Farmers are encouraged to sell their indigenous cattle and buy exotic cows or subsidised exotic bulls (project pays half the price). Alternatively, MECDP inexpensively hire exotic bulls to villages for mating.

Only one of the farmers sampled in Kwoti owned an exotic cow, although two farmers were planning to buy one in the future. However, in Upper Tegres half of the farmers sampled owned exotic or cross-breed cattle. Households with higher incomes were more likely to own exotic or cross-bred cattle (Figure 8). Exotic cattle owners had five animals on average, while cross-breed owners had three animals on average. People in Upper Tegres that only owned indigenous breeds of cattle owned two animals on average, whilst people that owned exotic or cross-breed cattle also on average owned five (range: 1-27) indigenous breed cattle. Indigenous cattle owners in Kwoti kept three animals on average. One of the primary objectives of introducing exotic breeds was to reduce the number of indigenous cattle in forest-adjacent parishes, whilst maintaining milk yields to the local population, with the ultimate aim of reducing grazing pressure within the National Park. It is not possible to comment on the success of this objective without cattle ownership data from previous years.

Zero grazing systems were being used on few of the sample farms in Kwoti (10%). Zero grazing was practised more commonly in Upper Tegres (40% practising zero grazing). A quarter of the farms in Upper Tegres that were practising zero grazing used barbed wire in the system. The remainder employed people to watch the cattle. Napier grass was the main food supplement for cattle in zero grazing systems, although matoke leaves and stems were frequently used, as well as the leaves of N. congesta and S. sesbans. The leaves of C. edulis, A. tortilis and Ricinus communis were used as fodder for goats in both parishes. The supply of Napier grass was low in Kwoti, but adequate for the demand from zero grazing systems (15% of farms had Napier Grass planted on contours and further 10% had Napier Grass planted elsewhere on the farm). In Upper Tegres, half of the sample farms had Napier Grass planted on contours and further 10% had Napier grass planted elsewhere on the farm.

The majority of cattle owners in Kwoti grazed their cattle primarily in the forest of the National Park (62%). People who grazed their cattle in the forest took them to graze there on average three times a week. The next most common grazing strategy in Kwoti was to supplement the use of fallow land with a diet of Napier grass, and free-range grazing on common land and fallow fields between the harvesting and planting seasons. The majority of cattle owners in Upper Tegres grazed their cattle in fallow land, and some farmers supplemented this with Napier grass or wild cut grass and the type of free-range grazing described above. A minority of cattle owners in Upper Tegres (19%) grazed cattle in the forest, taking them to graze there five times a week on average.

Cattle owners that took their animals to graze in the forest were asked what their response would be if forest security were tightened to prevent all forest grazing. In both parishes farmers said that they would sell their indigenous cattle. The majority of farmers were aware of zero grazing techniques and would buy exotic cattle to replace them. Most of these farmers also said they would grow feeding supplements or leave land fallow. A number of farmers would not have bought exotic cattle with the proceeds from the sale of their indigenous cattle. One farmer in Kwoti was typical of this group when he answered "We would die; we depend on cattle". When questioned further, his only response was to sell the cattle. This suggests that a significant proportion of cattle owners in each parish were aware of exotic cattle breeds and zero grazing techniques. It is therefore clear that for a significant proportion of the population, knowledge alone will not motivate them to alter the way in which they manage their livestock. It seems that if forest grazing is still easily accessible people will have little motivation to invest their resources in setting up a zero grazing system.
 

Adoption of agroforestry practices has the potential to make significant contributions to the conservation of MENP as well as to the welfare of the adjacent rural communities. Apart from reducing the demand for forest tree products by replacing forest resources with equivalent on-farm resources, agroforestry can sustain and improve crop yields as well as providing a range of additional resources and services. By utilising trees for animal fodder and green manure, the integration of trees into crop and animal systems can introduce substantial nutritional inputs. Trees also have the potential to maintain and improve structural and nutritional properties of soils through nitrogen fixation, soil aeration, contributions to the organic matter content and structural stability of the soil (Young, 1989). In addition to this agroforestry can contribute towards the satisfaction of subsistence needs (e.g. food, fuelwood, timber), supplement incomes (sale of raw or processed tree products), replace farm inputs which would otherwise be purchased (e.g. green manure, live fencing) and provide social benefits such as boundary demarcation, shade and privacy (Gregersen et al., 1989; Falconer, 1990; Arnold, 1997).

Tree planting was taking place in every farm sampled in Upper Tegres, and 60% of farmers in Kwoti were growing their own trees. Farmers closest to the forest in Kwoti were less likely to plant trees. This is because the ownership of the area is still under dispute. Until land tenure is settled in this area, many farmers will be reluctant to plant trees. Nevertheless a significant proportion of people in the area were planting trees. Farmers in Kwoti had on average 11 planted trees per acre, and in Upper Tegres farmers there were on average 16 planted trees per acre. Including wild trees (but not including distant woodlots), farms in Kwoti on average contained 41 trees (one in ten of these were wild Acaciasp. under one year old) and farms in Upper Tegres contained on average 59 trees. In Kwoti, a significant proportion of on-farm trees were under one year old in both parishes (38%), whereas in Upper Tegres much fewer trees were under one year old (19%). The majority of on-farm trees in Upper Tegres had been planted (80%), whereas 46% of the trees growing in Kwoti had been planted.

A variety of species were being grown in each parish, of which Eucalyptus grandis, Sesbania sesbans, R. communis and Persea americana were most common in Kwoti. In addition to these species, Cypress and Markhammia lutea were also being grown commonly in Upper Tegres. The most common wild trees in the Kwoti in decreasing order of abundance were Acacia tortilis, A. lahai, Vangueria apiculata, Erythrina abyssinica, Dombeya goetzenii, Nuxia congesta, Bersama abyssinica ssp. abyssinica and Syzygium cordatum. The most common wild trees in Upper Tegres were M. lutea, R. communis, A. tortilis, Ekebergia ruppeliana, Senna didymobotyra, V. apiculata, Croton macrostachys and Vernonia auriculifera (which can be a shrub or tree).

Despite the number of people planting trees and the number of trees growing on each farm, the proportion of people relying on the forest for as their main source of firewood and construction materials is still high. 95% and 85% of households sampled in Kwoti and Upper Tegres respectively rely on the forest for their main source of firewood, and 80% and 65% of households in Kwoti and Upper Tegres respectively rely on the forest for their main source of construction materials. The collection of wood for construction is a sensitive issue, as this is usually collected live, so the proportion of households actually relying on the forest for their main source of construction materials is probably higher. As such, on-farm trees have tended to augment rather than substitute tree products from the forest. This may represent a strategy in which farmers are saving on-farm tree resources for sale or times of forest resource scarcity.

Although the benefits of trees to animals and crops were not good incentives for farmers to plant trees (many AEOs omit to educate farmers about these effects for this reason), awareness about the role of trees in soil conservation was high. The majority of decision makers sampled in Kwoti and Upper Tegres (75% and 67% respectively) believed that certain tree species could improve the fertility of the soil or increase crop yields (these farmers acknowledged that other species could adversely affect crops by shading them or attracting insect pests). 20% and 19% in Kwoti and Upper Tegres respectively believed trees to have no effect on either soils or crops, and 5% and 14% in the two parishes believed all kinds of trees were harmful to crops if grown nearby.

Despite this high level of awareness about the potential of trees to improve soil conservation and crop productivity and the high proportion people growing trees in each parish, integration of trees with animal or crop components of the production system was limited in both parishes. The main exception to this was the traditional use of shade trees in banana and coffee plantations. Due to the higher incidence of banana and coffee plantations in Upper Tegres, interactions between trees and crops were more common in Upper Tegres than in Kwoti. The use of M. lutea, F. natalensis, Ricinus Communis and other species in banana plantations is one of the few traditional agroforestry systems documented in Uganda (UNT and ICRAF (1988:31). Trees have been historically used for shade in banana plantations in the study area, and were used for this purpose before MECDP started education programmes in the area. The popularity of banana cultivation in Upper Tegres is partly due to the inclusion of banana cultivation in early MECDP education programmes aimed at increasing the incomes of local farmers. Although few farmers were growing trees in coffee plantations, many of these plantations had been established under mature wild trees. The most common tree to be grown in coffee plantations was Cordia africana. This species is commonly used to provide shade in coffee plantations throughout Uganda (Katende et al., 1995), and use of this system probably predates MECDP involvement in the area.

With the exception of shade trees in banana and coffee plantations in Upper Tegres, trees were most commonly situated on internal or external boundaries or near the homestead (Table 2). These trees were often being used to demark boundaries or provide a wind break. E. grandis was commonly grown in woodlots in Upper Tegres. 35% of farms had E. grandis woodlots. Woodlots containing up to 50 trees were situated on the main farm, and larger woodlots (50-500 trees) were usually situated on land further away from the main farm. A significant proportion of the trees planted in Upper Tegres were situated in woodlots (35%). E. grandis woodlots on the main farm wereusually situated along streams and in marshy, steep or rocky areas where annual crops could not be grown. Woodlots were much less common in Kwoti. Although only one farmer had planted a woodlot in Kwoti the trees in it accounted for 12% of the trees sampled in the parish. It was uncommon for trees to be grown in cultivated fields. The main exception to this was one farmer in Upper Tegres who was a traditional healer. Due to small field size and the higher value of medicines relative to small scale maize production, this farmer had planted many trees amongst his maize crop. He experienced low maize yields, but blamed low soil fertility for the yields.

There are a number of factors contributing to the absence of trees from crop systems. Most importantly, the ploughing system employed by farmers in both parishes involves the use of bulls to pull the ploughs. This method is considered to be incompatible with intercropping with intercropping. In Kwoti particularly, grazing of livestock in fallow fields after harvesting is a disincentive to intercropping. In addition to this, the shade of many trees species would be counter-productive in farms with small fields. Light shade trees however can be compatible with small field size. S. sesbans has been intercropped successfully in small fields in the Kakamega District of Kenya (Bradley, 1990) (the ploughing method was not described).

The position of trees was ranked in descending order of frequency (Table 2). Wild trees were most commonly found in maize fields, near homesteads, in fallow land or on boundaries. Given a random distribution of wild trees, it is most likely that the majority will be situated in the most common land use. Maize cultivation takes up the largest area in each parish, and hence the majority of wild trees are situated in maize fields. People have chosen to situate homesteads near mature wild trees to provide shade, and have used large wild trees to mark boundary lines, which accounts for the high incidence of wild trees in these positions. Trees are the climax species of natural successions which take place in fallow land.

Indigenous knowledge of trees and their uses was examined by comparing the number of tree species present on different farms and the number of uses known for the trees in those farms. People in Upper Tegres were more likely to have more species present on their farms and know more uses for them than people in Kwoti (farms in Kwoti had seven species present on average and farms in Upper Tegres had nine species present on average. People in Kwoti knew eight uses for the trees on their farm, and people in Upper Tegres on average knew eleven uses for the trees on their farm). In addition to this farmers in Upper Tegres were more likely to weed, prune and irrigate their trees than people in Kwoti (Figure 9). Knowledge was concentrated in the older generations (Figure 10) and was greater in people that had reached secondary level education than in those reaching primary level or with no formal education. People in Upper Tegres that had been in contact with the AEO had more tree species present on their farms and knew more uses for them than those that had not spoken to the AEO (Figure 11). In Kwoti, there was no correlation between the number of species present or the number of uses known and contact with the AEO. People that had spoken to the AEO had also planted more trees than people that had not spoken to him (Figure 12).

Table 3 describes the main uses of on-farm trees. Planted and wild trees in both parishes were most commonly used to provide firewood and construction materials. Although most trees are used for firewood, this is not an incentive for the majority of decision makers (who are male) to grow trees. Female decision makers commonly grow trees for firewood because the collection of this resource is their responsibility. Male decision makers are more likely to grow trees for resources that they are responsible for providing, for example income and construction materials. However, the provision of construction materials was rarely the primary incentive to plant trees. Many farmers that had planted trees for other purposes were planning to use them for construction when they became too old to provide the former resource. As such, despite their promenace in Table 3, both firewood and construction materials may not be important incentives for people to plant trees. The growth of trees for sale was more important in Upper Tegres than Table 3 would suggest, as the majority of trees grown for sale were situated in woodlots, the size and importance of which was difficult to represent in the ranking system used. People commonly grow trees for shade in both parishes, however in Upper Tegres these trees are usually grown to provide shade for banana plantations, and in Kwoti these trees are usually grown to provide shade around the homestead. Trees from Upper Tegres are more likely to be sold than trees from Kwoti. This may be due to the tighter restrictions on forest access in Upper Tegres, or because there are more commercially valuable species grown in Upper Tegres. Trees were more commonly used for stakes in Upper Tegres due to the higher proportion of farms with banana plantations in which stakes are used. Trees may have been more commonly used for fodder in Upper Tegres due to the forest grazing restrictions imposed by UWZ-Face adjacent to the parish. The lower incomes of most people in Kwoti may have given them a greater incentive to use trees for medicinal purposes than people in Upper Tegres, as the costs of conventional medicine can be high.

Knowledge of medicinal tree species was concentrated in middle aged people (Figure 13). A quarter of the species sampled had known medicinal uses. In Kwoti, medicinal knowledge was concentrated in the immigrant community. In considering the significance of this, it should be noted that although 89% of the medicinal species were named by immigrants, the community in Kwoti is composed of 80% immigrants. It should also be noted however that the majority of immigrants that knew of medicinal species had come from nearby parishes where it is likely that they were familiar with the same tree species as they were in Kwoti. In Upper Tegres, the balance of knowledge between immigrants and locals corresponded to the size of these groups within the community. This list does not include the many medicinal species used which grow in the forest, on common ground or in the plains.

The main tree species promoted by IUCN were Calliandracalothyrus, Cordia africana, Ekebergia ruppeliana, Eucalyptus grandis, Grevillea robusta, Leuceana leucocephala, Sesbania bispinosa and S. sesbans. The proportion of farmers adopting these species in Kwoti and Upper Tegres can be see in Table 4. With the exception of S. sesbans in Kwoti, where the majority of farmers that were growing the tree had not spoken to an AEO, there was no correlation between growing promoted species and having had contact with the AEO. This is probably because MECDP have promoted these species through a network of community nurseries. AEOs train interested farmers in the collection of tree seeds. They pay the farmers for the seeds they collect and sell them on to the nurseries. Nursery proprietors are given free fertilisers and pesticides, gerry cans and watering cans. Community nurseries give farmers a financial incentive to promote the benefits of trees to their community. As members of the community, these people are more socially acceptable and respected than outsiders, and providing demand for trees continues, these nurseries should be sustainable beyond the end of the project work of MECDP. Whilst AEOs can give advice to farmers, they cannot give seedlings away freely as this would undermine the work of community nurseries, although there are a number of unconfirmed reports of this happening in both parishes (the second most common source of seedlings in Upper Tegres and the third most common source in Kwoti was from the AEO (Table 5) and the distribution of free seedlings was ranked by farmers in Kwoti as one of the most important benefits of the project (Table 8). The majority of farmers in Upper Tegres now obtain their seedlings from nurseries. Community nurseries have not been so successful in Kwoti. Demand for trees is lower in Kwoti, and people that are planting trees are mainly planting indigenous species. The majority of people therefore obtain their seedlings from the forest, and this has made nursery establishment difficult in the parish. The only working nursery had under a hundred seedlings when we visited, although plans were being made to expand the enterprise. People in this parish are under the impression that the AEO is planning to create a nursery in the forest, but this was not confirmed by the AEO himself.

The poorest people in Kwoti were growing more Grevillea robusta, E. grandis and S. sesbans than more wealthy people in the parish. There was no correlation between other promoted tree species and income in Kwoti, and there was no correlation between income and the growing of promoted tree species in Upper Tegres.

The average age of planted trees in Kwoti and Upper Tegres was 3.9 and 3.0 years respectively. A fifth of farmers in Upper Tegres had been planting trees before the project started work there in 1992. Farmers that had been planting trees before the project began had used five different species, two of which were exotic (M. lutea, Cypress sp., E. grandis, C. africana and P. americana). All these were being used for firewood and construction, and in addition to this, two species were used for shade, one species for sale, and one species for fencing. It is not clear whether the uses for these trees have changed during the time since they were planted.

In comparison, 75% of farmers in Kwoti had been planting trees before the project began there in 1997. As Kwoti and Upper Tegres border each other, it is possible that the practice of tree planting had spread to Kwoti before the project began working there. However, as the AEO was a resident of Kwoti throughout the work in Upper Tegres, it is not possible to differentiate between spread to Kwoti from farmers in Upper Tegres and spread through existing relationships and casual interactions with the AEO in Kwoti. Figure 14 shows that there were few farmers in Kwoti planting E. grandis before the project started working in the parish, but a significant proportion were planting the species in Kwoti before the project started work there. It should be noted that this figure is based on information about the age of trees on farms at the time of the study and does not take into account the fact that older trees may have been harvested prior to the study (although few of the farmers sampled were selling trees).

Figure 14 does however show that farmers in Upper Tegres have been planting less trees since the AEO left the parish. Trees planted in 1997-1998 would not have been mature enough to harvest prior to the study, so this figure should be more reliable than figures for previous years. If figures for previous years were depressed due to harvesting, the drop in tree planting after 1996 is more significant than Figure 14 shows. This is corroborated by records of tree age which showed that only 13% of the planted trees on farms in Upper Tegres had been planted in the last year.

Few of the trees planted in either parish were able to fix Nitrogen (on average, the proportion of planted trees which were nitrogen fixing was 7.5% in Kwoti and 11.0% in Upper Tegres). Few farms sampled in either parish had planted a significant proportion of Nitrogen fixing trees (Nitrogen fixing trees accounted for more than a quarter of the total trees grown in 10% and 25% of farms in Kwoti and Upper Tegres respectively).

In order to examine the possibility of a relationship between the growing of Nitrogen fixing trees and soil fertility, it was necessary to use the more ambiguous soil fertility data collected by verbal farmer accounts and visual descriptions. Maize productivity data is of little relevance in relation to the effect of Nitrogen fixing trees on the soil, as no farmers in the sample had planted them in maize fields. The only recorded instance of intercropping with a Nitrogen fixing tree was S. sesbans with bananas in two farms in Upper Tegres. According to this data, farms in Kwoti with Nitrogen fixing trees were more likely to have fertile soil than farms with Nitrogen fixing trees in Upper Tegres (83% and 20% of farms with planted nitrogen fixing trees in Kwoti and Upper Tegres respectively had "fertile" or "moderately-fertile" soils). One reason why farms with nitrogen fixing trees in Upper Tegres were less likely to have fertile soil could be that farmers with the naturally less fertile orange-brown soils were more likely to grow nitrogen fixing trees in an attempt to increase the fertility of their soil. Whilst the trees may have increased soil fertility in these areas, it is still unlikely to match that of naturally more fertile soils of Kwoti. On the other hand, the majority of people in Kwoti had "fertile" of "moderately fertile" soil, and the presence of Nitrogen fixing trees in these farms may not be significant. There was much more natural variation in the fertility of soils in Upper Tegres.

Three quarters of tree species identified in the two parishes were indigenous species, and a quarter were exotic. Of the ten most abundant species, two were exotic (Eucalyptus grandis and Persea americana). 90% of all the trees that harboured insect infestations were indigenous. It is known that the more "apparent" a species is, the more insect herbivory it will attract. Apparancy can be increased by increased tree species density, visibility and length of stay (for example, a small tree species which has recently grown in the middle of large forest with no other trees of the same species nearby has a low apparancy). The insect community on any tree species takes many years to accumulate (this is especially true for climax species which tend to invest more resources in chemical and physical defences), so exotic species are less apparent than indigenous species. Indigenous species which are found commonly in the forest are particularly apparent due to the high regional density of the species.

People spend less time collecting firewood in Kwoti than in Upper Tegres (on average 3.1 hours per week in Kwoti compared to 4.3 hours per week in Upper Tegres). There could be a number of reasons for this. A significant portion of the farmland in Kwoti was reclaimed from the forest in 1992, and in this area there is a substantial supply of firewood from old forest trees still standing in the fields. In addition to this, the Z-shaped nature of the forest line in Kwoti gives the parish a higher proportion of forest edge to the area of the parish than Upper Tegres. The restrictions on firewood collection imposed by FACE in Upper Tegres may also increase the time necessary to collect firewood.

The use of domestic trees for firewood was linked to the number of trees present on each farm. Firewood was more likely to be a major use of trees on farms with many trees. Farms with fewer trees were more likely to use them primarily for other purposes (Figure 15).

More people had fuel saving stoves (FSS) in Upper Tegres than people in Kwoti (10% of farms in Upper Tegres, and none in Kwoti). This could be due to the greater time required to collect firewood in Upper Tegres, or due to the fact that more people had been in contact with the AEO in Upper Tegres. Income did not seem to be a barrier to obtaining a FSS (owners had very different incomes). Indeed, the main reasons given for not owning a FSS were that there were none available for sale (or that it was not known where a FSS could be purchased) or it was not known how to use one. A small number of farmers said they were too expensive or had not heard of them. Half the people that did not own FSSs in Upper Tegres were planning to obtain one in the future.
 

Farmers from each parish were asked to rank their problems and needs (Table 6). A number of needs were shared by communities in both parishes. Soil erosion, land shortage and lack of education were consistently stated as major problems by males in both parishes. Soil erosion, low yields and labour shortage for weeding were consistently ranked highly by females in both parishes, and only females ranked the difficulty of firewood collection as a problem. With the exception of soil erosion, which was an important problem for both males and females, gender differences in problems tended to relate to the different responsibilities ascribed to males and females. For example, it is a female responsibility to feed the family, weed the crops (although weeding is also done by children and external labour) and collect firewood, and men are the decision makers (and would like education to help them make appropriate decisions) and land is passed through the male line (and fathers fear they will not have enough land to distribute amongst their sons). Females typically stated less needs than males.

Soil erosion was ranked as the second most important problem by males in both parishes, and ranked as first and fourth most important problem by females in Upper Tegres and Kwoti respectively. Indeed, 95% of farms sampled were experiencing soil erosion. From the context of other problems stated by respondents, it can be suggested that males and females ranked this problem highly for different reasons. For males, who are responsible for income generation, soil erosion decreases profit, but for females, who are responsible for feeding the family, soil erosion decreases crop yields (males ranked lack of capital above than low yields in both parishes, and females did not rank lack of capital).

AEOs are perceived as having a key role in the education of farmers. The majority of farmers in both parishes felt that their education was inadequate and attributed this to the shortcomings of the AEO.

Lack of capital was a typically male problem (income generation is a male responsibility). The inaccessibility of loans was believed by many to exacerbate the problem and widen the gap between rich and poor. There was widespread ignorance about the loan system used by banks, non-banking financial organisations and development organisations. Most people believed that wealthy people had gained access to credit through dishonest means. In addition to this, the unpredictability of the agricultural sector and the high interest rates charged by some institutions has made many farmers afraid of repayment difficulties and they are wary of loans. Whilst lack of capital to invest in the production system was ranked as a problem by farmers in each of the parishes studied, when capital did exist, it was rarely invested directly in the agricultural system. Many farmers in Upper Tegres had formed marketing co-operatives in which the majority of the profits were given as a bonus to one farmer each year. Farmers involved in co-operatives most commonly spent their bonus on vehicles.

The problem of crop pests was ranked higher by farmers in Kwoti than farmers in Upper Tegres. Although more people in Upper Tegres were using pesticides, it is not certain whether such a small difference in usage would adequately account for the different priority of the problem in each parish. It is possible that farmers in Kwoti are less well trained in pesticide application than farmers in Upper Tegres. Stray animals were more of a problem in Kwoti than in Upper Tegres. This is probably due to the fact that local government in Upper Tegres passed legislation, which has been effectively enforced, to prevent cattle roaming.

The importance of illness (which prevented work) in Kwoti may be a function of the poverty of the parish. Poor families can less easily afford medical expenses and are more likely to leave illnesses to become serious before seeking medical help. Illnesses are most prevalent in the wet season, when labour is highly valuable, and may represent a significant constraint to production in the parish, although it was not possible to quantify this. The late arrival of maize seed from Kenyan suppliers was also a problem which was restricted to Kwoti. This is probably due to the fact that Upper Tegres is far more accessible than Kwoti by vehicle.

The importance of fertiliser shortage in Upper Tegres may be due to the low soil fertility of many farms in the parish. Low soil fertility was a problem restricted to farmers in Upper Tegres.

There were a variety of possible solutions to the problem stated by the farmers. There was a potential for agroforestry to contribute to the solution of problems put forward by 40% of male and 50% of female interviewees in Kwoti. When these farmers were asked how they were planning to solve their problems, 63% of males and 50% of females were planning to use trees as part of the solution to their problem (a quarter of the males (and none of the females) only planned to use trees to generate income). When those who were not planning to use trees were asked if growing more trees could help solve any of their problems, a further 25% of males and 50% of females specified problems which they could use trees to solve, although the majority of these people would only use the trees to generate income with which to solve the problems. There were possible agroforestry solutions to the problems put forward by 60% of the male and 71% of the female interviewees in Kwoti. When these farmers were asked how they were planning to solve their problems, 58% of males and 20% of females were planning to use trees as part of the solution to their problem. When those who were not planning to use trees were asked if growing more trees could help solve any of their problems, a further 41% of males and 61% of females farmers specified problems which they could use trees to solve (again, the majority of these people would only use the trees for financial purposes).
 

People were asked to rank the main benefits and problems of work of MECDP (Table 7). From the answers collected, it is possible to build up a picture of how people perceive the project. The project was perceived primarily in terms of the direct benefits it could bring to individual farm units in Upper Tegres. In Kwoti there was a higher appreciation for more indirect, less personal benefits, such as the maintenance of the hydrological cycle and the conservation of the forest in order to maintain supplies of firewood and construction materials. Therefore, although people in Kwoti had less detailed knowledge about the project (more people held inaccurate beliefs about details of the project in Kwoti), they appeared to have a wider base of knowledge about the aims of the project. This may be due to the stage Kwoti has reached in the extension program, as the wider aims of the project are explained at an early stage in order enable people to recognise their need to change their land-use practices. During later stages of the programme, these aims may have become obscured by aims of personal gain.

Despite having a wider knowledge base, people in Kwoti had more misconceptions about the project than people in Upper Tegres. The inaccuracy of perceptions in Kwoti was identified by analysing the stated benefits and problems of the project for inaccurate statements. 22% of statements made about the project by people in Kwoti were false (for example some people stated that MECDP had banned firewood collection, when in fact only live wood collection has been banned), whereas only 2% of these statements were false in Upper Tegres. Much of this difference can be explained by the fact that less people in Kwoti had spoken to the AEO or attended a seminar, coupled with the fact that those who had attended seminars had attended half the number of seminars attended by people in Upper Tegres. In addition to this, policies and techniques may not have been explained in detail during the early stages of the extension programme. However, this does not account for the magnitude of difference between the accuracy of the perceptions of people in the two parishes. The difference in accuracy of perception might further be explained by the fact that the data for AEO contact does not take into account the number of times a person has spoken to the AEO, the length of time spent or number of topics covered in the course of the meeting. The majority of farmers in Kwoti (63%) that an inaccurate perception of MECDP had spoken to the AEO previously. The only farmer that had an inaccurate perception of MECDP in Upper Tegres had not spoken to the AEO.

Positive perceptions differed in Kwoti and Upper Tegres according to people’s experience of the project. Farmers in Upper Tegres had received more information about zero grazing, exotic cattle and beeping than people in Kwoti, and therefore were more likely to include these subjects in their list of benefits.

Negative perceptions of the project in Kwoti centred around lost privileges such as the right to collect live wood, graze cattle or cultivate land in the forest. Farmers in both parishes complained that restrictions had been imposed before it was possible for planted trees to have reached maturity. However, considering E. grandis can be used for construction and firewood at one year old, this argument is unfounded. Whilst some farmers had been given free seedlings against MECDP policy, those who had not been given them freely were as a result dissatisfied with MECDP policy on this matter.

In Upper Tegres however, negative perceptions were more complex and tended to be based on greater contact with the AEO. The AEO was perceived negatively by almost half of the farmers sampled. The last stages of the project work before the extension team stopped working in Upper Tegres appear to have produced some of the negative perceptions. Few farmers were aware that the extension team had moved on to other parishes and some had been expecting for a farm visit for years. Many people were not satisfied with the amount of education they had received from the AEO. Many important problems had been left unsolved and productivity had not increased as much as they had expected. The lack of follow-up by AEOs was criticised by a number of people. Advice and training was given, but farmer’s progress was not checked. As a result, failures to implement advice were not picked up, for example some farmers had planted trees which had died, and did not know how to prevent this happening in the future. A number of people blamed these problems on the insufficient knowledge and expertise of the AEO. In addition to these criticisms, a small minority of farmers in Kwoti complained that the AEO was discriminating between immigrants and indigenous people, and that he was not setting an appropriate example of tree growing in his own farm.

Unrealistic expectations of the project in Upper Tegres may have produced disappointment amongst farmers who were expecting more help than they received. The poor perception of the AEO may be due to people apportioning blame for their disappointment. With the exception of a poor follow-up programme and the number of trees planted on the AEO’s farm, it is not possible to comment on whether other allegations about poor AEO practices are true. The negative perception of the AEO in Kwoti may represent farmer’s frustration that they cannot get the information they need. At this stage in the project work, the AEO clearly has not had time to meet all the educational demands of the people in Kwoti.

It should be noted that the salary of AEOs is low (lower than primary school teachers) and only just enough to live on (Cheptagei, w. pers. comm.). AEOs must have other income to be able to afford to send their children to school. The only other income for the AEO in Kwoti and Upper Tegres was from agriculture, which takes up considerable time. The AEO stated that more days in the field would be possible if he was better paid. As a result, MECDP does not expect AEOs to be able to carry out all the tasks they are responsible for.

Overall, perceptions about MECDP were generally positive, but more so in Kwoti than Upper Tegres. Greater contact with the project in Upper Tegres appeared to create more negative perceptions, which along with the higher incidence of misconceptions about the project in Kwoti, suggests that perceptions in Kwoti were less informed and may change with greater project contact.

We would like to thank all the staff from the Mount Elgon Conservation and Development Project who made this research possible: Cherop Sandei, Cheptagei William, Nyangas Simon Vincent and Kakwaka Bosco in Kapchorwa, and Imelda Lwanga in Mbale. Thank you for all the time and invaluable help you gave us. We greatly appreciate the help given to us by staff from UNFA and FACE.Thank you Olivia Maganyi (Makerere University Herbarium) for identifying the plant specimens which reached you in a shameful conditionconsidering the number of times you explained how to collect them properly last year! Finally, we want to specially thank Toskin Michael for the kindness and patience he showed us, for his impressive interpretation skills and the invaluable advice he shared with us throughout the research.
 

Amooti, N. (1996). Eco-Tourism, Uganda: Tourism Today, August 1996.
Arnold, J.E.M. (1997). Framing the Issues, In Farms, Trees and Farmers: Responses to Agricultural Intensification, Arnold, J.E.M. and Devees, P.A. (eds.), Earthscan.
Barlett, P.F. (ed.) (1980). Agricultural Decision Making: anthropological contributions to rural development. Academic
        Press, New York.
Barron, R., Logan, H., Lilly, J., Inman, N. and Brennan, M. (1994). Negative Emotion and Message Processing, Journal of
        Experimental Social Psychology 30, 191-201.
Baum, K.H. and Schertz L.P. (1983). Modelling farm decisions for policy analysis. Westerview Press, Boulder.
Binswanger, H. and Ruttan, V. (1978). Induced Innovation: technology, institutions and development. John Hopkins
        University Press, Baltimore.
Boserup, E. (1965). The Conditions of Agricultural Growth. Aldine Publishing, Chicago.
Bradley, P. (1990). If I had known the discussion was to be about Kuni I wouldn’t have wasted my time coming. In
        Afgroforestry for Sustainable Production: Economic Implications, R.T. Prinsley (ed.) Commonwealth Science
        Council.
Britton, P. L. (Ed.) (1980) Birds of East Africa. East African Natural History Society, Nairobi.
Cameron, A. (1997).Small Mammal Community Composition on Mt. Elgon, Uganda, with Reference to Habitat
        Modification by Humans, Aberdeen University BSc Hons Dissertation, unpublished.
Chambers, R. (1983). Rural Development: Putting the Last First, Longman, Harlow.
Chambers, R. (1992). Rural Appraisal: Rapid, Relaxed and Participatory, IDS Disscussion Paper No. 331, Institute of
        Develpoment Studies, Brighton, Sussex.
Cochrane, J. (1994). Ecotourism Surveys, Expedition Advisory Centre, Royal Geographical Society, London.
Coe, M.J. (1969). Microclimate and animal life in the equetorial mountains. Zoologica Africana
        4: 101-128
Davenport, T. Howard, P. and Dickenson, C. (1996) Mount Elgon Biodiversity Report. Republic of Uganda Forest
        Department, P.O. Box 1752, Kampala, Uganda.
Davies, K. A. (1952). The building of Mount Elgon. Geological Survey of Uganda, Memoir VII Geological Survey
        Department, Entebbe.
Erlich, P. and Erlich, A. (1990). The Population Explosion, Simon and Schuster, Sydney.
Falconer, J. (1990). Agroforestry and Household Food Security. In Agroforestry for Sustainable Production: Economic
        Implications, R.T. Prinsley (ed.), Commonwealth Science Council, London.
Gregersen, H., Draper, S. and Elz, D. (eds.) People and Trees: The Role of Social Forestry in Rural Development,
        Economic Development Institute of the World Bank, Washington D.C.
Van Heist, M. (1994). Land Unit Map of Mount Elgon National Park. IUCN Technical Report, unpubl.
Holden, D., Hazell, P. and Pritchard, A. (eds.) (1991). Risk in Agriculture, Proceedings of the Tenth Agriculture Sector
        Symposium. World Bank, Washinton D.C.
Howard, P. C. (1991). Nature Conservation in Uganda’s Tropical Reserves. Forest
        Department/Ministry of Environment Protection, Uganda.
Hull, C.L. (1943). Principles of Behaviour, Appleton-Century-Crofts, New York.
Hull, C.L. (1951). Essentials of Behaviour, Yale University Press, New Haven.
Katende, A., Birnie, A. and Tengnas, B. (1995). Useful Trees and Shrubs for Uganda: Identification, Propogation and
        Management for Agricultural and Pastoral Communities, Regional Soil Conservation Unit.
Katende, A., T., Ipulet, P., Rodriques, R. and Dranzoa, C. (1989). Birds and Woody Perennials Inventory: Mount Elgon
        Forest Reserve, Sustainable Development and Forest Conservation in Uganda, IUCN Technical Report No. 1.
Kenya Web (1998). Bungoma District: Constraints on Development,
        http://www.kenyaweb.com/economy/bungoma/bu_const.html
Kerkhof, P. (1990). Agroforestry in Africa: a survey of project experience, Panos Institute, London.
Khon Kaen University (1987). Proceedings of the 1985 International Conference on Rapid Rural Appraisal, Rural
        Systems Research and Farming Systems Research Projects, Khon Kaen, Thailand.
Kings, S. (1997). A comparative study of the forest avifauna of Mount Elgon National Park,
        Project Elgon ‘96 Final Report, unpublished.
Mascarenhas, J. (1992). Participatory Rural Appraisal and Participatory Learning Methods: recent experiences from MYRADA and South India’s forests, Forests, Trees and People Newsletter, No. 15/16.
Mascarenhas, J. et al. (1991). Participatory Rural Appraisal: Proceedings of the February 1991 Bangalore PRA Trainers
        Workshop, RRA Notes, No. 13, IIED and MYRADA, Bangalore.
Matthews, R. (1996). Chapter 3: Birds. In: Dickinson, C., Davenport, T., and Howard, P. (eds.) (1996). Mount Elgon
        Biodiversity Report (No. 1), Uganda Forest Department, Kampala.
McPharland, C. and Opus, C.J.C. (1998). Study of Species Richness and Diversity of the Avifauna of the Piswa Area,
        Project Elgon ‘97 Final Report, unpublished.
MECDP (1994). Advatages of employing members of the local community as Agricultural Extension Officers, internal
        communication, unpublished.
MECDP (1995). Phase III Objectives, IUCN unpublished.
National Population Census (1991). Government Press, Entebbe.
Onyango, G. (1996). Local Participation for the Conservation and Management of Natural Forests: Uganda,
        http://www/worldbank.org/html/extdr/offrep/afr/afr_for/interim/sugand-2.htm
Pearson, D. J. and Turner, D. A. (1986) The less common Palaearctic migrant birds of Uganda. Scopus 10: 61-82
Ransom, R.W. (1998). Local Community Perceptions of the Socio-economic and Cultural Consequences of
        Conservation and Tourism in the Mt. Elgon National Park, Uganda. Aberdeen University BSc Hons Dissertation,
        unpublished.
Reed, M.S. and Clokie (submitted). Effects of Grazing and Cultivation on Afromontane Plant Communities on
        Mount Elgon, Uganda, submitted to Journal of Appied Ecology, April 1999.
Rocheleau, D.E. (1987). The Suer Perspective and the Agroforestry Research and Action Agenda. In Agroforestry:
        realities, possibilities and potentials, H. Gholz (ed.), Martinus Nijhoff, Dordrecht.
Ruthenberg, H. (1980). Farming Systems in the Tropics. Clarendon Press, Oxford.
Scott, P. (1994). Assessment of Natural Resource Use by Communities from Mount Elgon National Park, Technical
        Report No. 15. IUCN unpublished.
van Someren, V.G.L. (1922) Notes on the birds of East Africa. Novitates Zoologicae. 29: 1-246.
van Someran, V.G.L. (1932) Birds of Kenya and Uganda, being addenda and corrigenda to my previous paper in Novitates
        Zoologicae 29, 1922. Ibidem37:252-380.
Scherr, S. J. (1997). Meeting Household Needs: Farmer Tree-growing Strategies in Western Kenya, In Farms, Trees and
        Farmers: Responses to Agricultural Intensification, Arnold, J.E.M. and Devees, P.A. (eds.), Earthscan.
Singh, I., Squire, L. and Strauss, J. (eds.) (1986). Agricultural Household Models: extensions, applications and policy.
        John Hopkins University Press, Baltimore.
Synnott, T. J. (1968). Working Plan for Mount Elgon Central Forest Reserve. 1st Revision. Period 1968 to1978.Uganda
        Forest Department, Entebbe.
Uganda National Taskforce (UNT) and ICRAF (1988). Agroforestry Potentials for the Land-use Systems in the Bimodal
        Highlands of East Africa: Uganda. AFRENA Paper No. 4
Uganda Wildlife Statute (1996). Statute No. 14.
UNDP/WTO (1993). Integrated Tourism Masterplan, Republic of Uganda, FinalReport, UGA 91/010, Madrid.
Were, G. S. and Wilson D. A. (1970). East Africa through a Thousand Years, Evans Brothers Limited.
Witte, K. (1992). Putting the Fear Back into Fear Appeals, Communication Monographs 59, 329-349
White, F. (1983). The vegetation of Africa. UNESCO, Paris.
 

Project Co-ordinators: Mark Reed and Joyce Reed
Field Work: Mark Reed, Joyce Reed, and Toskin Michael
Advisor:Dr. H. Parry
Report Writer: Mark Reed