People, Park and Plant

PEOPLE AND PLANTS WORKING4 PAPER - DECEMBER 1996 This series of working papers

is intended to provide information and

to generate fruitful discussion People, park and plant use on key issues Recommendations for multiple-use zones and develop- in the sustainable ment alternatives around Bwindi Impenetrable National and equitable use Park, Uganda of plant resources.

Please A.B. Cunningham

send comments

on this paper

and suggestions

for future

issues

People and Plants Initiative,

Division of Ecological Sciences,

UNESCO, 7 Place de Fontenoy, The designations employed and the presentation of material throughout this publication do not imply the expression of any opinion whatsoever on the part of UNESCO concerning the legal status of any country, territory, city, or area of its authorities, or concerning the delimitation of its frontiers or boundaries. The opinions expressed in this paper are entirely those of the authors and do not commit any Organization.

This publication is based on a report prepared in 1992 for CARE-DTC. CARE's Development Through Conservation (DTC) project, initiated in 1988 under an agreement between USAID, WWF and CARE, has supported environmental conservation in south-west Uganda with particular focus on the two national parks, Bwindi Impenetrable National Park and Mgahinga Gorilla National Park and the adjacent communities. DTC has supported Uganda Wildlife Authority in setting up the innovative multiple-use progamme in which communities harvest non-timber forest products from the two parks. The People and Plants Initiative of WWF, UNESCO, and the Royal Botanic Gardens, Kew, has supported CARE- DTC's work through sponsorship, research and information gathering, ethnobotany training, and publication of the project's experience in implementing multiple use.

Author’s address:

A.B. Cunningham P.O. Box 42 Betty’s Bay 7141 SOUTH AFRICA

Photos: All photos by A.B. Cunningham, except for cover photos 2 and 3 and photos 4 and 5 by R. Höft

Cover illustration: Contours of Bwindi Impenetrable National Park overlaid on (1) Granary constructed from Loeseneriella apocynoides (omujega) near Bwindi; (2) Women carrying Cyperus latifolius (ekigaga) bundles used for weaving mats; (3) Smilax anceps (enshuli) used for granaries and stretchers.

Published in 1996 by the United Nations Educational, Scientific and Cultural Organization 7, place de Fontenoy, 75352 Paris Cedex 07 SP, FRANCE Printed in France by EGOPRIM on chlorine-free recycled paper

Edited by Martin Walters and Robert Höft Design: Ivette Fabbri Layout: Martina Höft

SC-97/WS/48

Recommended citation: Cunningham, A.B. 1996. People, park and plant use. Recommendations for multiple-use zones and development alternatives around Bwindi Impenetrable National Park, Uganda. People and Plants working paper 4. UNESCO, Paris.

This publication is also available in French. People, park and plant use RECOMMENDATIONS FOR MULTIPLE-USE ZONES AND DEVELOPMENT ALTERNATIVES AROUND BWINDI IMPENETRABLE NATIONAL PARK, UGANDA Abstract In the large forested areas of the Zaïre and wood or other wooden products), and minor Amazon basins human densities are low, and dis- forest products (all non-wooden products). The turbance by "forest peoples" creates rather than results and recommendations of this report are reduces diversity, forming a mosaic of vegetation presented first for the latter category, involving types at different stages of recovery after distur- mainly specialist users of non-wood products, bance. including wild plant resources, honey, basketry Afromontane forests are at the opposite and bamboo use. The various uses of wood, the extreme. Situated in one of the most densely pop- major forest products, (e.g. blacksmiths, carved ulated areas of Uganda, the remaining forests wooden handcrafts, beer boats, building poles, formerly occupied by the Batwa have become the bean stakes) are then considered. focus for harvesting of plant resources by the These recommendations need to be seen as farmers who cleared them. They have also part of an ongoing process of interaction between become the subject of national and international the rural community surrounding Bwindi conservation efforts. Impenetrable National Park and the park man- Afromontane forests in western Uganda, and agement, with DTC (Development Through Bwindi Impenetrable Forest in particular, are Conservation) project staff at the interface now fragmented islands, surrounded by rural between the two groups. farmlands. Under these circumstances, sustain- Specialist user groups within rural communi- able forest management differs greatly from the ties surrounding Bwindi Impenetrable National use of low species diversity, highly productive Park, having a good knowledge of plant reedbeds or thatch-grassland, where harvesting is resources, can form an important interface seasonal, obvious and easy to manage. Recovery between the National Park or DTC staff and the from harvesting in productive annual systems is rural community in general. They also represent also short, due to annual production of above- groups of resource users with a common interest ground biomass. Instead of the short rotation in beekeeping, traditional medicines, basketry or applied in reed cutting, sustainable harvesting of other uses. All of these are recognized for their forest for timber is usually aimed at rotation skills within communities and by the Resistance times of 50-200 years. Council (RC) system. Many are already mem- In Bwindi Forest, the effects of pitsawing and bers of organizations established either on com- agricultural clearing in the past are superimposed munity initiative or through the combined inter- on differences in vegetation caused by topography ests of the community and Ugandan government and soil type. This results in patchy distribution departments. both of species and size class categories of the trees Key species to specialist users are Faurea or shrubs used for bellows, building poles, beer saligna (omulengere) and Sericostachys scan- boats or bean stakes. It also influences the avail- dens (omuna) for hives and honey; Rytigynia ability of these resources, either increasing the kigeziensis (nyakibazi) to treat internal parasites; number of young saplings (bean stakes, building Loeseneriella apocynoides (omujega) and Smilax poles) or colonizing species (e.g. Polyscias fulva anceps (enshuli) for basketry, bamboo for build- (omungo)) in disturbed sites. It has also ing and granaries and Rapanea melanophloeos decreased stocks of large hardwoods, due to (omukone) for carved sticks. The use of edible overexploitation and competing uses for timber plants is generally limited to famine periods, (e.g. Prunus africana (omumba) and Newtonia with Myrianthus holstii (omufe) fruits and buchananii (omutoyo)), despite the "reserved" Dioscorea (ebikwa) tubers most favoured. status of hardwoods. Three size classes of woody plants are This report focuses on resource use and man- favoured for beer boats (>50 cm dbh), building agement issues relating to wild plants and multi- material (5-15 cm dbh) or bean poles (1.5-5 cm ple-use zoning in Bwindi Impenetrable National dbh). These size classes represent successive Park. Foresters usually group products into two stages of trees forming the forest: canopy, sub- categories for forest management purposes: canopy and understorey. Straight hardwoods are major forest products (such as timber, fuel- selected for their durability for building (e.g.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 1 People, park and plant use A.B. CUNNINGHAM Drypetes (omushabarara) species) and beer land-use conflict and putting off the real solu- boats (e.g. Newtonia buchananii (omutoyo)), tions to the problem. At present, this is con- although less durable Ficus species (ekyitoma) sidered to be the case for some wood uses (beer are also used as they are easier to carve. boats, bean stakes and building poles) due to the Bean stakes are selected on the basis of size combination of past impacts, high demand and and abundance rather than species. Alchornea limited staff for complex management. hirtella (ekizogwa) is favoured, however, as it Cultivation of trees is widely practised in the readily produces a high density of thin stems. If DTC area already. In a recent survey conducted saplings of canopy species fit into this size class in this area, for example, Eucalyptus (88% of 120 then they are also cut. These species are not only a respondents) and Acacia mearnsii (49%) were useful resource to local people, they also represent the species most preferred for building and had the forest canopy of the next century. respectively been planted by 77% and 36% of Recommendations for forest product use by respondents. From field observation, it is clear specialist groups within multiple-use areas fall that many homes in the DTC area are built from into four categories: these cultivated tree species (particularly (1) open access to specialist users (e.g. bee- Eucalyptus), with the use of exotic species keepers, non-commercial harvesting of increasing with distance away from the forest. It medicinal plants); is recommended that self-sufficiency in these (2) seasonal access to popular plant resources categories of wood use is facilitated through with limited distribution by harvesters development of nurseries and by the supply of elected within user groups (e.g. Marantaceae seedlings to interested growers. used for basketry); Cultivation efforts could also involve produc- (3) seasonal and rotational management by spe- tion of favoured and effective medicinal plants cialist harvesters (e.g. bamboo; and craftwork species, through a collaborative Loeseneriella apocynoides (omujega) and effort between DTC, IFCP/ITFC (Impenetrable possibly Rapanea melanophloeos Forest Conservation Project/Institute for Tropical (omukone)), with potential users involved in Forest Conservation), primary health care initia- resource assessments and setting of quotas; tives and local ICRAF (International Centre for (4) continued closed access to resources where Research in Agroforesty) groups. sustained use is not possible, due either to Tree planting activities need to be focused complexity, high demand or slow growth particularly on steep sites, high human popula- rates (beer boats, building poles, bean poles tion density parishes, and adjacent to critical and possibly fuelwood) and where the parts of Bwindi Impenetrable National Park (e.g. emphasis needs to be placed on providing the Kitahurira corridor and Ngoto Swamp areas). alternatives outside the National Park. The current problems faced by generally landless Additional recommendations are made for Batwa people and their need for land have been future research and monitoring, including the highlighted in previous reports. Involvement in involvement of resource users and traditional forest conservation, beekeeping, research and experts as research partners. Special mention is eco-tourism activity as specialist guides and con- made of the valuable role that Batwa people can servation staff can provide an alternative means play in research on forest ecology and in inven- of income from a remnant of the forest their fore- tory work as "parataxonomists". fathers formerly occupied. At least for some, If resource harvesting is not sustainable, then income through these activities could enable land it is a false solution, providing brief respite from aquisition and a continued link with the forest.

2 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Contents 1 Abstract 3 Contents 4 Introduction

6 Bwindi Impenetrable Forest: conservation importance and vegetation change 6 Conservation importance 8 Climate change 8 People and vegetation change

11 Creating guidelines for multiple-use zones 11 Which are priority species ? 13 Local knowledge of resource users 13 Life form, plant parts used and the effects of harvesting 14 Sustainable in practice?

15 Results and recommendations 16 Minor forest products and specialist user groups Edible wild plant resources Beekeeping and honey collection Medicinal plants Basketry Bamboo 31 Major forest products: wood Blacksmiths and bellows Canoes Wood carving - household items Building poles Bean stakes Fuelwood 38 Sustainable use of wood products

47 New natural products with commercial potential 48 The future

49 References 51 Personal communications 52 Acknowledgements 53 Acronyms 54 Appendix I

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 3 People, park and plant use A.B. CUNNINGHAM Introduction

Proclaimed as a Forest Reserve in 1932 recognizes land-use zones for different purposes, (Howard, 1991) and managed by the Forest including fully protected areas for conservation Department until recently proclaimed as a of biodiversity. Addressing the land-use conflict National Park (1991), Bwindi Impenetrable between National Park and people, this approach Forest is internationally recognized as site of out- stems from a major concern expressed by sur- standing importance for the conservation of bio- rounding people that the 1991 change of status of diversity (Butynski, 1984; Hamilton, 1981; Bwindi Impenetrable Forest from a Forest Kingdon, 1990). Bwindi Forest is the only site in Reserve to a National Park would result in their East Africa, and one of the few forests in Africa losing access to forest resources. The objective as a whole, with continuous forest cover over an of this study was to provide detailed information altitudinal range from 1190-2607 m (Howard, on wild plant use and resource management 1991). Considered to be a Pleistocene refugium, issues relating to Bwindi Impenetrable National Bwindi Impenetrable National Park (BINP) is Park and the DTC area. The terms of reference the highest biodiversity site in East Africa for were to: birds, plants, butterflies and primates, and con- (1) review existing surveys and information on tains half of the world's endangered mountain the ethnobotany of Bwindi Impenetrable gorilla (Gorilla gorilla beringei) population Forest, identifying gaps in the information (Butynski, 1984; Howard, 1991). previously collected, and train Ugandan staff Farmers surveyed around Bwindi Impene- to collect the data needed; trable Forest recognized the importance of the (2) conduct ethnobotanical surveys (including forest for forest products, "bringing rain" and interviews with local traditional medical catchment protection (Forbes, 1991; Scott, practitioners); 1992). Despite these perceptions, Bwindi (3) focus on two or three critical species identi- Impenetrable Forest (321 km²) is all that remains fied by prior studies as being of high priority of forest that formerly covered most of Kigezi for including in a multiple-use programme; (Figure 1). With a long boundary (114 km) sur- (4) train Ugandan staff in ethnobotanical tech- rounded by nearly 100,000 people, the long-term niques and management during the course of future of this forest depends on a strong partner- fieldwork; ship between conservation and development. The (5) investigate ethnobotanical research priorities Development Through Conservation (DTC) pro- for Ugandan staff and/or graduate students, ject (CARE International) based in Ikumba and and provide an ethnobotanical research plan; working since 1988 through Conservation (6) visit the project area in and around Bwindi Extension Agents in parishes around Bwindi Impenetrable National Park and conduct a Impenetrable National Park (Figure 2) is an two-month field training workshop for example of such a partnership. Government of Uganda staff and post-gradu- Aimed at a grass-roots approach to rural ate students, including the management of development and conservation, the DTC project multiple-use areas.

4 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Figure 1. Locality of Bwindi Impenetrable National Park, Uganda in relation to other forests in Uganda, showing the former extent of forest cover (from Howard, 1991).

Figure 2. The DTC area surrounding Bwindi Impenetrable National Park, showing location of 1 dot = 1 person per km2 parishes and population densities determined from an unpublished 1991 census data and number of households (data from 1992). Parishes with the five highest population densities are immediately adjacent to the National Park (Nyamabale: 298 people per km2; Mpungu: 279 people per km2; Remero: 247 people per km2; Kashasha: 230 people per km2; and Rutugunda: 217 people per km2). No data are available for Kifunjo and Nyarurambi.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 5 People, park and plant use A.B. CUNNINGHAM Bwindi Impenetrable Forest: conservation importance and vegetation change

Conservation importance Uganda, or represented in Uganda only in Kabale-Rukungiri. (Although Allanblackia kim- The importance of conserving Bwindi and other biliensis (Photo 2), Brazzeia longipedicellata, forests in western Uganda has been explained by Grewia mildbraedii, Strombosiopsis tetrandra, Butynski (1984) and Struhsaker (1987). Detailed Maesobotrya floribunda (plus Chrysophyllum comment here is limited to aspects relating to pruniforme, which Howard (1991) has since forest plants. recorded from Budongo and Itwara forests) were Although tree species diversity of Bwindi thought to be confined to Ishasha Gorge Impenetrable Forest is low compared with high (Hamilton, 1991), this is probably an artefact of diversity rain forest, it is important not only as a previous plant collecting surveys. Allanblackia representative of the Afromontane centre of kimbiliensis, for example, was recorded during endemism for plants (Photo 1), but also for ani- this survey, not only in Ishasha Gorge, but also as mals restricted to this habitat (Butynski, 1984; a relatively common tree along the Ihihizo and Howard, 1991) (Tables 1 and 2). A 1ha plot sur- Ivi River valleys). veyed for trees >10 cm dbh (diameter at breast Bwindi Forest is a Pleistocene refugium height) in Amazonian rain forest in Peru, for containing not only plants typical of Afromontane example, contained 275 species, representing 50 forest but also representatives of the families (Peters et al., 1989), compared to only Guineo-Congolian flora, such as the secondary 45-50 tree species >10 cm dbh in 1 ha of Bwindi forest tree Musanga leo-errerae (Cecropiaceaea), Impenetrable Forest at 2000-2200 m asl, and the shrub Agelaea pentagyna (Connaraceae), only 20 tree species per ha in forest at 2400 m asl herbs such as Ataenidia and Marantochloa (Howard, 1991). (Marantaceae) and parasitic plants such as Bwindi Impenetrable Forest contains tree Thonningia sanguinea (Balanophoraceae). genera endemic to Afromontane forest, and Bwindi Forest is a major catchment area and many tree species that typify Afromontane rain a source of water to surrounding rural communi- forest are also represented. Although Lovoa ties, and through Lakes Edward and Mutanda via swynnertonii (Meliaceae) is the only tree species the Nile to the Mediterranean. It can also provide listed as endangered, Bwindi Forest contains a economic benefit from non-consumptive uses of number of tree species not found elsewhere in the forest (e.g. eco-tourism) and consumptive

Table 1. The seven centres of endemism in Africa, with numbers of seed plants, mammals (ungulates and diurnal primates) and passerine bird species in each, and the percentage of these endemic to each unit (in Huntley, 1988).

Biogeographic Area Plants Mammals Birds Unit (1000 km2) No. of species % endemic No. of species % endemic No. of species % endemic Guineo-Congolian 2815 8000 80 58 45 655 36 Zambesian 3939 8500 54 55 4 650 15 Sudanian 3565 2750 33 46 2 319 8 Somali-Masai 1990 2500 50 59 14 345 32 Cape 90 8500 80 14 0 187 4 Karoo-Namib 692 3500 50 13 0 112 9 Afromontane 647 3000 75 50 4 220 6

6 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM uses. Consumptive uses may be of resources meeting basic needs of the surrounding community (e.g. plant resources) or on a wider scale (e.g. genetic values of wild relatives of crop and forage plants, and chemical structures for new pharmaceuticals).

Photo 1. An unidentified Memecylon Photo 2. Fruit of Allanblackia kimbiliensis (Clusiaceae). (Melastomataceae) species.

Table 2. Tree species in Bwindi forest with particular conservation importance.

Tree genera in Tree species in Bwindi Tree species in Bwindi not Tree species found elsewhere in Bwindi endemic to that typify Afromontane found elsewhere in Africa but restricted in Uganda to Afromontane forest forest Uganda (Butynski, 1984; the south-west (Butynski, 1984; Howard 1991) Howard 1991) Afrocrania Entandrophragma excelsum (Cornaceae); (Meliaceae); Myrianthus Allanblackia kimbiliensis Cassipourea congoensis Hagenia (Rosaceae); holstii (Cecropiaceae); (Clusiaceae); Brazzeia (Rhizophoraceae); Chrysophyllum Ficalhoa, Balthasaria Podocarpus latifolius longipedicellata pruniforme (Sapotaceae); Drypetes (Theaceae) (Podocarpaceae); Ocotea (Scytopetalaceae); Grewia bipindensis and Sapium leonardii-crispi and Xymalos usambarensis (Lauraceae); mildbraedii (Tiliaceae); (Euphorbiaceae); Oncoba routledgei (Monimiaceae). Agauria salicifolia Strombosiopsis tetrandra and Dasylepis racemosa (Ericaeae); Aningeria adolfi- (Olacaceae); Maesobotrya (Flacourticeae); Tabernaemontana friedericii, Chrysophyllum floribunda (Euphorbiaceae); odoratissima (Apocynaceae); Cola gorungosanum Xylopia staudtii bracteata (Sterculiaceae); Pauridiantha (Sapotaceae); Hallea (Annonaceae), Balthasiaria callicarpoides (Rubiaceae); (=Mitragyna) rubrostipulata (=Melchiora) schliebenii Pittosporum spathicalyx (Rubiaceae); Parinari excel- (Theaceae), Guarea (Pittosporaceae); Millettia psilopetala sa (Chrysobalanaceae); (=Leplaea) mayombensis (Fabaceae); Dichaetanthera corym- Prunus africana (Meliaceae) and an unidenti- bosa (Melastomataceae); Musanga (Rosaceae); Syzygium fied Memecylon species leo-errerae and Myrianthus holstii guineense (Myrtaceae) and (Melastomaceae) which (Cecropiaceae); Ocotea usambarensis Strombosia scheffleri occurs on alluvial terraces in (Lauraceae); Ficalhoa laurifolia (Olacaceae). the Nteko and Buhoma (Theaceae). areas.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 7 People, park and plant use A.B. CUNNINGHAM To facilitate informed decision-making, plant content. From field observation, the density and use and forest conservation policy have to be seen species abundance of edible wild plants (the prin- against the background influences of climate and cipal species being Myrianthus holstii (fruits) human disturbance of forest ecosystems. Both and Dioscorea spp. (ebikwa) tubers) appear to be have had a major influence on African vegetation even lower than in the Zaïre lowland forest stud- in the past and will continue to do so in the future, ied by Hart and Hart (1986). None of the Guineo- perhaps even more so with the effects of global Congolian zone edible fruit trees (e.g. Irvingia, warming and human population increase. Ricinodendron heudelotii) that are major food sources to Mbuti people occur in Bwindi Forest. Climate change It appears likely therefore that Batwa subsistence would have been dependent on plant and animal Massive oscillations in the Pleistocene climate, resources of savanna and wetlands, in addition to caused by expansions and shrinking of the polar those of forest. ice-caps, resulted in long, cool, dry periods alter- However, Batwa hunter-gatherers may have nating with shorter, warmer, moist periods. manipulated forest and savanna vegetation. Equatorial forests, as indicators of world climatic Although there is no direct evidence from the conditions, are believed to have expanded out- Rukiga highlands on this, it may be that Batwa wards from, or shrunk into, Pleistocene refugia. hunter-gatherers achieved this through the use of Detailed pollen analysis from cores taken in the fire. Fire would have been used seasonally in for- Rukiga highlands near to Bwindi Forest has pro- est during honey hunting and possibly in savan- vided evidence of vegetation dynamics and cli- na to attract game. mate change over the past 40,000-50,000 years, Fire could also have been used as a tool in including forest expansion around 10,600 BP forest during dry periods, to create disturbance into the Ahakagyezi catchment along the Ishasha and stimulate production of Dioscorea tubers. river south-east of Bwindi Forest (Taylor, 1990). Dioscorea climbers are most commonly found in During the most recent glacial phase secondary forest or forest margins (Hart and (pre-12,000 BP), forests were restricted to a few Hart, 1986; this study). Hunter-gatherers in refugia, later expanding outwards with moister, southern Africa, for example, use fire as a tool to warmer conditions (Hamilton, 1981). Hamilton increase below-ground production of edible (1981) has stressed the importance of conserving Iridaceae corms (Deacon, 1983). "Fire-stick forests which retained forest cover during the farming" is also thought to have been used by earlier arid phase. Bwindi Impenetrable Forest is hunter-gatherers in forests in New Guinea for considered to be one of these refugia in Uganda. edible resources, including yams (Dioscorea The decline of forest tree species-richness across spp.) (Groube, 1989). Dioscorea tubers are Uganda from west to east is considered to be thought to have been a major food resource of indicative of this, and forests in western Uganda, Mbuti Pygmy peoples in the past (Tanno, 1981). particularly those of Bwindi-Kayonza and In the Ituri Forest, Mbuti Pygmy population den- Bwamba are thus considered to have the highest sity was approximately 1 person per km². The national conservation priority (Hamilton, 1981; hunter-gatherer population density in the Rukiga Howard, 1991). highlands in the past is unknown, but was probably no higher than this. It would be expected People and vegetation change therefore that with low human densities, their impact on vegetation would have been localized Dating of archaeological remains from Matupi and subtle compared to the clearing of forest by Cave in the eastern Ituri Forest, Zaïre, indicates agriculturists. Butynski (1984) estimated that the human occupation 32,000-40,700 years ago (Van Batwa Pygmy people accounted for less than Noten, 1977). Similar data for the Rukiga high- 0.5% of the total population. This would be con- lands are not available, but it is likely that, like sistent with a population density of (less than) 1 the Mbuti hunter-gatherers in the Ituri region, person per km² today. Batwa Pygmy people originally occupied the Pollen analysis has not only provided evi- forests and savanna of south-west Uganda and dence of shifting forest cover in response to cli- northern Rwanda. mate change over the past 40,000-50,000 years, From their study of Mbuti hunter-gatherer but also of the clearing of forests in the Rukiga subsistence in the Ituri Forest, Hart and Hart highlands. Although previously thought to have (1986) suggest that it is unlikely that hunter-gath- started before about 4800 years ago (Hamilton et erers would have lived independently in the for- al., 1989), a reassessment of core material sug- est interior, as for five months of the year virtu- gests that clearing took place after about 2200 ally no nutritionally important wild edible plants years ago (Taylor, 1990), coinciding with the are available, honey is not abundant and, influx of Bantu-speaking agriculturists with although game meat is available, it has a low fat iron-smelting technology (Van Noten, 1979)

8 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM (Photo 3). Farming was probably based on finger millet (Eleusine coracana), sorghum (Sorghum sp.) and possibly cow-peas (Vigna sp.) and pigeon peas (Cajanus cajan) which originate from the Horn of Africa. Agriculture was established in Rwanda (and probably the Rukiga highlands) by about 2000 years ago, resulting in more permanent settlements and concentrating the effects of human occupation on the surrounding vegetation due to burning, clearing and cutting of fuel (for iron-smelting (Photo 4) and household use and other purposes). This would also have stimulated trade in forest products (e.g. bush meat) for cultivated starches between Batwa and Bakiga. Since then, a wider range of crops has gradually been introduced from Central and South America (sweet potatoes, tomatoes, cassava, pineapples, chilli peppers, groundnuts, potatoes and tobacco), south-east Asia (bananas, sugarcane), south-central China and northern India (tea), the near East (peas), Central Asia (carrots) and the Mediterranean (cabbages). Agricultural production occupies 83% of the population of Uganda, accounts for nearly all export earnings, and contributes 60% to GDP (World Bank, 1986). In 1921, the Ugandan population was 3 million people (Howard, 1991), and by the year 2000 it is projected to be 23.8 million (Bulatao et al., 1990). At the same time, the area of forest that formerly would have been used for harvesting of plant resources has rapidly decreased, due to agricultural clearing and burning. Harvesting intensity therefore concentrates on the remaining vegetation, ultimately focusing on species within core conservation areas. Photo 3. Iron smelting technology introduced into the Rukiga high- Extensive transformation of the Rukiga highlands c. 2000 yr ago remains essentially unchanged today by black- lands landscape has occurred since the 1900s, due smiths (omuhesi) in this area. Wooden bellows (omuzuba), made from to natural population increase and migration from Polyscias fulva (omungo) wood and the clay tuyère (encheru). Rwanda, where population density is 480 people per km² of arable land (Balasubramanian and Egli, 1986). Between 1948 and 1980, the population of the Kabale and Rukungiri districts increased by 90%, from 396,000 to 752,000 (Butynski, 1984). Today, the Rukiga highlands are one of the most densely populated areas of Uganda, with population densities in the DTC area surrounding the forest ranging from 102-320 persons per km² (Figure 2, page 5) (data from an unpublished DTC 1991 census). Intensive agriculture by a high density of rural farmers has resulted in removal of indigenous woody plants, shorter and shorter fallow periods and a reduction in species diversity. Situated in one of the most densely populated areas of Uganda, with a 115 km long boundary surrounded by almost 100,000 people (Anon, 1992) Bwindi Impenetrable National Park became an island in a sea of rural farmers, gold-miners Photo 4. Polyscias fulva (omungo) tree (Araliaceae) favoured for wood- and pit-sawyers (Photos 5-8). en bellows (omuzuba) used in traditional iron-smelting technology.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 9 People, park and plant use A.B. CUNNINGHAM Photo 5. Bwindi Forest, intact apart from disturbance Photo 6. The impact of farming on forest: fields in what due to tree falls (canopy gaps) and fire. was forest in 1950.

Photo 7. Species-selective over-exploitation and gap Photo 8. Disturbance to river valley forest due to illegal formation: pitsawing. panning of alluvial gold.

10 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Creating guidelines for multiple-use zones

Multiple-use zoning, with the assistance of the between people and the National Park. Prior to Impenetrable Forest Conservation Project proclamation as a National Park, Butynski (IFCP), has formed part of the DTC project, with (1984) estimated that 10-20 people enter the for- preliminary zoning developed by Butynski and est each day for beekeeping purposes or to hunt later by Scott (1992) (see Wild and Mutebi 1996, for wild beehives, whilst 25-50 people a day col- Working paper 5). lect fuelwood, bamboo and building materials. These zoning plans were developed in the Wild plant resources from the forest provide context of previous regulations on resource har- rural people with a wide range of basic needs: vesting. Under the 1964 Ugandan Forest Act building materials, fuel, binding material, house- relating to Kigezi, regulations were aimed at the hold utensils, medicines, food supplements and a controlled harvesting of forest plant resources, source of income from the sale of baskets, honey, and were based on a permit system, intended to carvings or from practising as a herbalist or mid- generate government revenue and facilitate mon- wife. itoring (Butynski, 1984). Mention might be made At present, beekeeping takes place in part of of the following regulations. the park, but other uses previously permitted (1) Any produce, except private or reserved under Forestry Department regulations are not trees, may be taken from Kigezi forest land catered for under current National Parks legisla- without license or fee by any African in rea- tion. Managed use of resources within national sonable quantities for his own personal parks or in buffer zones around them has, how- domestic use. Such produce includes timber, ever, become a widespread strategy as a means poles, bamboo and fuelwood. No produce, of defusing land-use conflicts (McNeely, 1988). however, is to be cut or removed without a Concern over loss of access to wild forest plant permit, and shall not be removed from the resources is an important local issue, and the place where it was cut or taken until checked need to take these concerns and local needs into and marked by a forest officer or forest guard. account has been identified (DTC, 1991). Both (2) A licensee can only fell trees or collect forest National Parks and CARE-International have produce if the trees or produce have been policies aimed at sustainable resource use. The marked or otherwise designated for felling or question is, once resources have been identified, collection by a forest ranger or forest guard. how does one decide whether uses are sustain- (3) Within seven days of felling trees or collect- able or not ? ing the produce, the licensee must notify a forest ranger or forest. Which are priority species? Additional regulations related specifically to timber production, pitsawing and cattle grazing. As outlined above, this study aims to produce Control of harvesting of "minor forest products", guidelines for multiple-use zones around Bwindi as well as timber and goldmining was inadequate Impenetrable National Park, and to identify gaps (Butynski, 1984; Hamilton, 1984; Struhsaker, in previous work. It must be recognized at the 1987) for economic and political reasons outset that data on abundance, productivity and described by Howard (1991). In 1983, an esti- population biology of even the African equatori- mated 140-280 people were involved in pitsaw- al forest trees with major economic importance ing and carrying timber, with an estimated are limited. There are even fewer data on the 100-200 people involved in panning gold from hundreds of species providing "minor forest river-valleys (Butynski, 1984). As a result, an products". Nevertheless, it is possible to identify estimated 10% of the forest reserve remained plant species vulnerable to over-exploitation and intact, 61% had been heavily exploited by pit- to identify use categories (e.g. hardwood timbers sawers and 29% "creamed" of the best hard- used for beer boats) where there is a narrow mar- woods by selective pitsawing (Howard, 1991). gin between sustainable use and over-exploita- Since then, following recommendations made by tion. Identification is based on: Butynski (1984), these major threats to the forest (1) indicators of demand, from field data outside have been stopped. the forest (e.g. density of bean stakes per ha, With agricultural clearing, pitsawing and number of building poles per home, basket gold mining stopped, the major threats to BINP sales at markets), data on favoured "indica- now are increased fire damage through runaway tor" species inside the forest (e.g. bark fires or arson, and in the longer term, possible removal from Rytigynia kigeziensis (nyak- deproclamation in the face of increasing conflict ibazi) for medicine, proportion of Alchornea

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 11 People, park and plant use A.B. CUNNINGHAM hirtella (ekizogwa) bean stakes) and pub- by traditional birth attendants or for tradition- lished information from long term studies al veterinary medicines and little information (e.g. fuelwood consumption per capita per was available on edible wild plants (including year, construction materials); fungi) or on plants important to bees or bee- (2) information from observation at local mar- keepers. kets and local resource users, or from devel- (3) No quantitative data were recorded on key opment of commercial trade elsewhere in resources, whose use has important implica- Africa, indicating whether species are, or will tions for maintaining forest structure and sus- become subject to a greater intensity and fre- tainable resource use policy. There was also a quency of use for commercial sale; need to go beyond visual assessments of (3) life form categories, as indicators of popula- resource abundance. tion biology (Cunningham, 1991); For these reasons, the present survey took a (4) part used (roots, bark, leaves, whole plant, complementary approach to that used by Scott wood, etc.) as an indicator of impact on indi- (1992). Instead of dealing with people encoun- vidual plants; tered at random, information was collected using (5) information from published records, local a more "targeted" approach through: resource users and field observation on scarce (1) work with specialist user groups (e.g. black- key resources; smiths, traditional birth attendants (TBAs), (6) experience of resource management prob- beekeepers, herbalists, cattle owners); lems that have arisen elsewhere that could be (2) selection of particularly knowledgeable local avoided in this case. Batwa and Bakiga people in each area, the This approach is a useful tool in a situation most knowledgeable of whom (J. Bandusya) where management guidelines are required, and worked as part of the survey team during the there is neither the time nor the financial or entire 2 month period of fieldwork; human resources available to undertake studies (3) doing as much quantitative work as time per- of the population biology and biomass produc- mitted. Good voucher specimens were col- tion of the species involved. It must, however, be lected whenever possible. regarded as a "first approximation", which can be Identification of specimens was done at the developed on a finer scale through subsequent Royal Botanic Gardens, Kew and Makerere longer-term studies. University herbarium. Additional specimens During a four-month study, Scott (1992) were sent to the Royal Botanic Gardens, interviewed respondents encountered randomly Edinburgh and the East African Herbarium, during fieldwork outside Bwindi Impenetrable Nairobi. Forest and conducted group interviews. Fieldwork with women specialist groups Information from these interviews was combined (basket makers, TBAs) was done by women with visual assessments by local people of the forming part of the survey team at different times abundance of key species, which were assigned a (M. Mehanda, M. Cunningham, R. Badaza and J. rating on a six-point scale. Scott (1992) recog- Tumusime). Information was collected from nized that further study relating to the "ecologi- TBAs and traditional medical practitioners cal sensitivity" of key plant resources would be (TMPs) on plants with symbolic ("magical") and necessary. To build on this useful background physiological actions. work several gaps in Scott's study were identi- Although a distinction is made in this survey fied. The most important of these were as between plant species with symbolic (or psycho- follows. somatic) uses and those with active ingredients, (1) Only a limited number of plants were identi- both are important in health care as: fied by botanical name, and there were no * species that have a purely symbolic value are herbarium voucher specimens for many important ingredients of traditional medicines plants. These are important in setting conser- for their psychosomatic value and are at least vation or resource use policies for plant as effective as placebos are in urban industrial species and use of local folk taxonomy society; (Rukiga names in this case) is not sufficient. * the majority of traditional medicines have not This is because some local names are applied been adequately screened for active ingredi- at a "generic" level. For example, the single ents, and a number of species (e.g. Rapanea local name "bitindi" corresponds to two melanophloeos (Myrsinaceae) in southern Memecylon species, M. jasminoides and an Africa) which are primarily used for symbolic undescribed species only found in Bwindi purposes also have active ingredients. Forest; "omushabarara" is applied to at least Conservation efforts must therefore be three Drypetes species, including the rare D. directed at all species vulnerable to bipindensis; and "omurara" covers at least over-exploitation, and the main priority for this four Macaranga species. study was identifying whether the species are (2) No information was recorded on plants used vulnerable to over-exploitation or not.

12 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM As part of the research training exercise, Despite such pitfalls, local knowledge repre- quantitative assessments were made of three key sents a practical and cost effective method for plant resources (hardwood building poles, bean identifying possible key species. In some cases, stakes and beer boats) and for bamboo, for which such as for small herbs or plants occurring in lim- there is scope for experimental rotational man- ited areas (e.g. Marantochloa leucantha, agement in selected sites. Quantitative assess- omwiru), it provides the main evidence of comments in 20 m x 20 m plots were made with mercial trade, and can direct specialist monitor- resource users, assessing density and suitability ing and conservation programmes (Cunningham, (on a five-point scale) of trees for building poles 1990, 1991). (all trees >5 cm dbh) and bean stakes (>1.5 cm diameter). Level of cutting was also assessed. Life form, plant parts used A similar approach was taken in the bamboo zone (10 m x 10 m plots). At Nteko, in a single and the effects of harvesting site where no pitsawing or harvesting had taken It is generally accepted that a relationship exists place, suitability for beer boats was assessed in a between resource stock or population size, and 100 m x 100 m plot of all trees >30 cm dbh. Plot the sustainable rate of harvest. Low stocks are sizes conform to those used by Muir (1991) and likely to produce only small sustainable yields, recommended by Alder and Synnott (1992). In particularly if the resource is a slow growing order to put bark use by herbalists into perspec- plant that takes a long time to reach reproductive tive, bark damage assessments were carried out maturity. Species with large populations, high by A. Tsekeli, B. Otim, R. Baragira and J. biomass production and a short time to reproduc- Bandusya, comparing this to elephant damage. A tive maturity would similarly be expected to pro- seven-point scale of bark damage was used duce high sustainable yields, particularly if com- (Cunningham, 1990) in two 100 m x 100 m plots petition was reduced by ‘thinning’. There is also of all trees >10 cm dbh in the Bamboo zone and a clear relationship between the part of a plant at Mubwindi swamp. being harvested and the impact on the plant. The response of plants to exploitation and the impli- Local knowledge cations of declining productivity under a high of resource users frequency or intensity of exploitation is critical to policy development. The knowledge and perceptions of resource Life form categories are useful for establish- users, such as traditional healers, craftworkers ing resource management principles in conjunc- and commercial medicinal plant harvesters, pro- tion with other factors such as demand and part vides valuable insights into the scarcity of useful of the plant being used. These categories help plant species and the development of conserva- bridge the gaps in knowledge about plant demog- tion and resource management proposals. This is raphy, enabling a first approximation of cate- particularly useful where we are dealing with gories of vulnerability to commercial exploita- hundreds of species, as in Bwindi Impenetrable tion. Life form categories represent a natural National Park, or trade in traditional medicines sequence from large trees to annual forbs and (Cunningham, 1990). grasses (Rutherford and Westfall, 1986), in other Such knowledge is particularly useful as it words from K-selected to r-selected species. has been gathered over many years of harvesting, Large trees are often the most vulnerable as buying and selling these plants. With increasing people choose them for their thick bark (e.g. for scarcity, a commercial trade develops, such as the national or international trade in medicinal the sale of bean stakes or building poles. If plant material) from large, old plants which have scarcity increases, the distance covered and time a long period to reproductive maturity, a low used to collect the scarce resources also increase, ratio of production to biomass and specialized with corresponding increases in price. This has habitat requirements. occurred in many parts of Africa with fuelwood Destructive harvesting affects habitat struc- (Leach and Mearns, 1989) and in southern Africa ture and the three ‘vital attributes’ essential for with rising prices for certain medicinal plants, replacement of plant species (Noble and Slatyer, many of them from Afromontane or coastal for- 1980), namely: est (Cunningham, 1990). (1) the means of dispersal or persistence at the Validity of local knowledge can be tested site before and after disturbance; against data in herbaria and in the literature on (2) the ability of the species to establish and the geographical distribution, rarity and extent of grow to maturity in a developing community; exploited species, so that local traders' perception (3) the time taken to reach critical life stages. of scarcity that may be an artefact of limited geo- Degree of disturbance to the species popula- graphical distribution can be distinguished from tion and vulnerability to over-exploitation scarcity due to over-exploitation (Cunningham, depend on demand, supply, part used and growth 1990). form. Coppicing ability and the vulnerability of

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 13 People, park and plant use A.B. CUNNINGHAM trees to bark removal are important attributes tion of annual stems which are resilient to har- which vary with the physiology of different vesting for hut-building purposes (Cunningham, species. Although able to withstand fire due to 1985; Shackleton, 1990). Managed harvesting of their thick bark, Faurea saligna (omulengere, reeds and thatch-grass is also facilitated by the Proteaceae), and Podocarpus latifolius (omufu, fact that cutting takes place in late autumn and Podocarpaceae) are at one extreme, being species winter when disturbance to nesting birds is min- sensitive to bark removal and susceptible to fun- imized. The same applies, in African savanna, to gal infection and borer attack. Prunus africana common, fast-growing medicinal plants or (omumba, Rosaceae) and Ficus natalensis (ekyit- encroaching species such as Acacia karroo, oma, Moraceae) are species at the other extreme, Acacia nilotica, Dichrostachys cinerea where bark regrowth occurs after the lower trunk (Fabaceae) and Euclea divinorum (Ebenaceae) has been severely de-barked. But even resilience which can often be removed as an aid to meeting after ring-barking does not enable trees to sur- management objectives for the savanna parks. vive when demand exceeds supply. Commercial Forests are considerably more complex sys- medicinal plant gatherers continue to debark tems than reedbeds, with multiple uses. Muir favoured trees when bark is only partially (1991) working with local wood-cutters in regrown because of the its scarcity and commer- Afromontane forest in southern Africa, has cial value, finally debarking large roots and demonstrated that cultivating alternative sources killing the trees. of building material outside indigenous forest In extreme cases (not yet recorded in western can be over ten times cheaper than the cost of an Uganda), genetic erosion may occur in the long intensive monitoring programme for sustainable term, as the plants cannot cross-pollinate because use of that resource. In most cases, conservation bark stripping for medicines continually keeps the bodies in developing countries do not have the population in a vegetative phase (e.g. Warburgia financial or human resources to carry out such ugandensis) or due to the felling of all but the programmes. Thus in cases where demand is smallest or remotest individuals (e.g. of Dalbergia high, and resources are both slow growing and melanoxylon for craftwork). In both of these exam- popular, "mining" rather than "managing" ples, this over-exploitation has occurred through resources occurs, and the narrow border between much of the range of these species, from South sustainable use and over-exploitation is crossed. Africa to Tanzania, and this emphasizes the impor- The higher the number of harvesters and tance of the minimum viable populations (MVP) uses of a species and the scarcer it is, the greater debate in conservation biology. the chance that resource managers and local people will get embroiled in complex juggling of Sustainable in practice? uses and demands in an attempt at a compromise that could end up satisfying nobody. One of the primary objectives of national parks Examples of increased demand leading to and reserves is the maintenance of habitat and over-exploitation include the felling of Mauritia species diversity. Park managers attempting to flexuosa and Jessenia bataua palms for their defuse land-use conflicts cannot allow fruits in the Peruvian Amazon (Vasquez and over-exploitation of natural resources within Gentry, 1989; Peters, 1990), the killing of those parks without compromising this primary favoured medicinal plants and dye resources by objective. It is therefore essential to establish ring-barking or uprooting in Africa for the local guidelines for the harvesting of wild plant or international trade (Cunningham, 1987, 1990), resources in national parks. the over-exploitation of Aquilaria crassna Since the publication of the World (Thymeleaceae) for export to Hong Kong as Conservation Strategy (IUCN, 1980) the term incense (Payapyipapong et al., 1988), the use of "conservation" has become almost interchange- Parkia roxburghii trees for jamu medicinal able with "sustainable use". This approach is fine preparations in Indonesia (Rifai and Kartwinata, for species or vegetation types with a high bio- 1991) and the depletion of copal and rattan mass production, low species diversity and sources in Palawan, Philippines (Conelly, 1985). resilience to harvesting, or where human popula- Forests are distinguished by high species tion densities are low. Demand for fast-growing diversity, limited distribution in eastern Africa, species with a wide distribution, high population low biomass production and multiple uses (med- density and high reproductive rate is easily met. icines from leaves, roots, bark and fruits, tradi- This is possible in the gathering of leaves or tional dyes from bark and roots, poles and laths fruits for medicinal purposes and as dietary sup- for hut building and the gathering of edible plements, thatch, weaving materials and reeds. fruits). Almost no published data are available on Examples of easily managed vegetation types are root or bark production of African woody plants Phragmites australis wetlands and Cymbopogon (Rutherford, 1978), making it virtually impossi- validus stands, which have a wide distribution, ble to set sustainable limits. Costly research in low species diversity and high biomass produc- obtaining such data would have little practical

14 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM value anyway due to the limited human resources to place major emphasis on developing alterna- available for monitoring and management. The tives to consumptive use of forest resources. At intensive management of forests practised in the the same time, an "Environmental Protection southern Cape for valuable hardwood (mainly Society" has credit co-operatives, education, Ocotea bullata (Lauraceae), Seydack et al., health and business components to channel ben- 1982) is unlikely to occur within national parks efits to village members. These include income or forest reserves due to financial constraints and to villagers from non-consumptive uses the number of species involved. (eco-tourism), loan schemes, income generation This is one of the main reasons why the and primary health care benefits, and has been approach taken in tropical forest conservation highly successful (Payapyipapong et al., 1988). sites like Khao Yai National Park, Thailand, has It has also avoided all the complexities of multi- been to stop all harvesting of forest products and ple-use forest management.

Results and recommendations

"...where funds are short, it is often tempting to impose a programme by official edict. Without necessary consultation and negotiation with the people, failure is almost certain to result." (Martin, 1986; Zimbabwe CAMPFIRE Programme).

Forest products are usually grouped by foresters Ugandan government departments (e.g. tra- into two categories for forest management pur- ditional medical practitioners (TMPs) and poses: major forest products (timber, fuelwood traditional birth attendants (TBAs)), or or other wooden products) and minor forest prod- through interaction between the community, ucts (all non-wooden products such as grass, National Parks and DTC staff (e.g. bee- resins, fruits, etc.) (Osmaston, 1968). keepers associations). In addition, the rural Results and recommendations of this study population around Bwindi Impenetrable are presented first for "minor forest products" Forest is large (c. 100,000 people) and (which despite the term, have major value to growing, whilst the National Park has rela- local people), secondly for bamboo, and finally tively few staff. Dealing with resource users for wood use (blacksmiths, carved wooden hand- by specialist user group enables interaction icrafts, beer boats, building poles, bean stakes). and more focussed discussion with smaller These are part of an ongoing process of interac- interest groups within parishes. tion between the rural community surrounding (2) Although a non-wood "minor forest product", Bwindi Impenetrable National Park and the park bamboo use is presented separately because it management, with DTC project staff at the inter- is so important to many people, rather than to face between the two groups (see Wild and a few specialist users. It is found within a Mutebi 1996, Working paper 5). localized area of Bwindi Impenetrable The reasons for presenting the results in this National Park, has high growth rates com- way can be justified as follows. pared to most tree species, and because of the (1) Specialist user groups within rural communi- low species diversity of woody vegetation ties surrounding the Bwindi Impenetrable within bamboo forest. All these are factors National Park can form an important inter- which influence management recommenda- face between the National Park or DTC staff tions. and the rural community in general. They (3) Although results are presented separately for have a good knowledge of plant resources. each category of wood use (e.g. beer boats, They also represent groups of resource users bean stakes, etc.), recommendations are made with a common interest in beekeeping, tradi- for all categories of wood together. This takes tional medicines, basketry or other uses, all of into account competing uses for the same tree whom are recognized for their skills within species, and the cumulative effects of differ- communities and by the village Resistance ent categories of wood or tree use on forest. Council system. Many are already members Due to the slow growth rates and long gener- of organizations established on community ation times of hardwood trees, and the devas- initiative (e.g. the Bakiga stretcher-bearer tating effects of pitsawing on Bwindi Forest society, ekyibinachengozi), or through the in the past, timber harvesting is included in combined interests of the community and this section.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 15 People, park and plant use A.B. CUNNINGHAM Minor forest products markets. With the exception of Dioscorea tubers, gathering is seasonal and practised only by the and specialist user groups poorest sector of the community, except perhaps during famine periods. Efugwe (Urticaceae), a EDIBLE WILD PLANT RESOURCES common understorey nettle in the bamboo zone Edible wild plants characterize disturbed for- is gathered from the forest, but edible leaves are est sites, and represent a resource seasonally used usually gathered from "weeds" (e.g. Amaranthus as dietary supplement or as a subsistence source spp., dodo) in fields. Physalis and Solanum of income from the sale of fruits such as nigrum are also found in fallow fields and dis- Myrianthus holstii (omufe) or edible fungi. Few turbed areas outside the forest. Use of species are used, and the impact of collection is Myrianthus (omufe) fruits encourages conserva- low. It is recommended that collection be tion of the female tree of this dioecious species allowed in multiple-use areas. when forest is cleared for agriculture, and The most important wild plant foods to peo- Myrianthus trees are a significant woody compo- ple are Myrianthus holstii (omufe) (Photo 9), two nent of agricultural fields, until they are pollard- Dioscorea species (ebikwa and ebihama) with ed and finally chopped out for fuelwood. tubers that can be eaten without detoxification, Removal of Dioscorea tubers usually kills and four types of edible fungi (Lentinus prolifer the individual plants collected. Although the (ebishanja) (Photo 10) and three species identi- tubers are long-lived and slow-growing, removal fied by local name only (ensabili, obushokore of tubers was only observed in two sites on the and a mushroom, obutusi) (Table 3). forest margin during this survey (three tubers in Lentinus prolifer fungi are more common Buhoma, two in Ishasha) and is considered only where forest has been cleared for agriculture, and to have localized impact. This is because: grow mainly on dead Polyscias fulva (omungo) (1) harvesting is selective, with rejection of small wood. Together with ensabili and obutusi fungi, or very large (and therefore woody and they are usually found and gathered when weed- unpalatable) tubers; ing sorghum and millet fields. Lentinus prolifer (2) gathering is low-intensity (restricted to is a particularly popular and tasty food source, famine periods or by a small sector of the "as good as meat". population not self-sufficient in cultivated These wild foods are useful supplements foods, particularly the Batwa); during famine periods, and a source of subsis- (3) there is regeneration from wind-dispersed tence income from the (seasonal) sale of Dioscorea seed blown in from adjacent zones Myrianthus holstii fruits and edible fungi in local or from smaller plants.

Table 3. Edible wild plant resources recorded in and around Bwindi Impenetrable National Park.

Family Plant species Rukiga Name Life Form Part used Amaranthaceae Amaranthus sp. dodo annual herb leaves Cecropiaceae Myrianthus holstii omufe tree fruit Dioscoreaceae Dioscorea sp. 1 ebikwa climber tuber Dioscoreaceae Dioscorea sp. 2 ebihama climber tuber Dioscoreaceae Dioscorea bulbifera climber tuber Lentinaceae Lentinus prolifer ebishanja fungus whole fungus Moraceae Ficus sur ekyitoma tree fruit Rosaceae Rubus sp. emerembwe scandent climber fruit Solanaceae Solanum nigrum entakara herb fruit, leaves Solanaceae Physalis peruviana entutu herb fruit Urticaceae Laportea sp.? efugwe herb leaves Zingiberaceae Aframomum sp. omatahe geophyte fruit indet. ensabili fungus whole fungus indet. obutusi fungus whole fungus indet. obushokore fungus whole fungus

16 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Photo 9. Edible Myrianthus holstii (omufe) fruit. Photo 10. Lentinus prolifer (ebishanja) edible fungus.

Box 1. Recommendations for edible wild plant use

∗ Enable continued gathering of edible wild plant resources by local people within multiple-use zones around Bwindi Impenetrable National Park. ∗ As part of the primary health care (PHC) strategy, encourage the continued conservation of Myrianthus holstii (omufe) trees in cultivated lands as a source of vitamin-C rich edible fruits, since vitamin C is a nutrient that can be deficient in a starchy staple diet.

BEEKEEPING AND HONEY Impenetrable Forest area. African honey bees COLLECTION (Apis mellifera adansonii, enjokyi) are kept in beehives, with two varieties recognized, the Beekeeping is a form of land-use which is aggressive brown ekitaka and the darker enyum- dependent on and can complement forest conser- bu. These bees provide the main source of honey. vation. There are two conservation concerns Honey-hunters (particularly Batwa) get about beekeeping, however. First, the possibility honey from wild hives of these bees and from a of fires associated with honey harvesting, and number of ground- or tree-nesting stingless bees second, the felling of Faurea saligna trees for (Trigonidae) known as ebihura. Six types of wooden hives. Trigonid bees are recognized: obugashu, obwiza, The possibility of runaway fires can be obuganza and obuhumbamga, all of which nest addressed through the formation of beekeeping in hollow tree stems or branches, obugazale bees societies, so that rather than being a source of which nest underground as well as in trees, and forest fires, beekeeping societies will bring obwahashi which only nest underground. strong social pressure to bear against arson or the Beekeeping is also an important seasonal careless use of fires. activity for Bakiga agriculturists, who use tradi- Hive construction materials and designs are tional beehives constructed either of timber or being used in the DTC area that provide an alter- bamboo, or woven from forest climbers (Table 4, native to Faurea wooden hives. There is an Photo 11, page 18). Top-bar hives, common in opportunity here for constructive intervention Kenya and Tanzania, are not used. Honey is usu- and local income generation through the forma- ally taken from the hives twice a year. tion of beekeeping societies, improved hive con- Batwa people in particular, as well as bee- struction and marketing of honey. keepers, have a rich knowledge of bees, plants Honey-hunting and beekeeping are impor- favoured by bees and hills that provide the best tant seasonal activities in the Bwindi sites for placing of productive hives. Forest

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 17 People, park and plant use A.B. CUNNINGHAM Table 4. Plant materials used for construction of wooden and woven beehives.

Family Plant species Rukiga name Life form Use Alangiaceae Alangium chinense omukofe tree wooden hives Fabaceae Albizia gummifera omushebeya tree wooden hives Poaceae Arundinaria alpina omugano bamboo woven hives Myrtaceae Eucalyptus spp. * ? tree wooden hives Proteaceae Faurea saligna omulengere tree wooden hives, popular Theaceae Ficalhoa laurifolia omuvumaga tree wooden hives Celastraceae Loeseneriella apocynoides omujega climber woven hives Araliaceae Polyscias fulva omungo tree wooden hives Musaceae Musa sp. * enjagata banana popular fibre for rain cover Celastraceae Salacia sp. bwara climber woven hives Tiliaceae Triumfetta macrophylla omunaba shrub twine for binding Note: exotic species are marked (*)

plants are an important source of nectar and that honey from some Trigonid bees also has this pollen to bees, and also provide resin for effect. Trigonid beehive construction. Forest tree Much of this knowledge is of use not only to species susceptible to heart-rot are also important beekeeping, but also to forest ecology in general. sites for wild bee nests. Some 21 species are Trigonid stingless bees, for example, are valu- recorded as favoured by bees as a source of able specialist pollinators of certain forest tree nectar or as nesting sites or by Trigonid bees as a species, yet most knowledge about this important source of resin for nest construction. aspect of forest tree biology has not been docu- Beekeepers and honey-hunters also know mented by formally trained scientists, but is which plants have toxic pollen, producing honey unrecorded knowledge held by Batwa causing diarrhoea, such as pollen from the giant honey-hunters. lobelia, Lobelia gibberoa (entomvu) and the for- An observation of many beekeepers in the est climber Urera hypselodendron (omushe), and northern sector of Bwindi Impenetrable Forest also worth following up is that there is an association between the establishment of tea plantations and a marked drop in honey production. This is attributed by some beekeepers to insecti- cides used to spray the tea bushes and needs further investigation. Areas of forest with high densities of prolif- ically flowering trees (e.g. Faurea saligna), shrubs (e.g. Brillantaisia, Mimulopsis in valleys) and climbers (e.g. Sericostachys scandens in disturbed montane forest, and an unidentified species, orumaga, Malphigiaceae in lowland forest) are well known to beekeepers, and hills near to these areas are favoured for placing of hives. There are four other reasons why forests are favoured places to site hives in comparison to putting hives in fields or near to homesteads: (1) hives are away from the smoke that is often present in fields due to people burning cleared vegetation; (2) the hives are protected from wind in forest, but not in fields. This facilitates movement of bees to and from the hives; (3) hives can be hidden in the forest away from the general public and are therefore less susceptible to theft or spells placed by jealous people on the hives of a man with many high- yielding hives; Photo 11. Woven hive made from forest climbers (4) keeping hives away from the homestead pre- and banana fibre, Ishasha area. vents family members being stung.

18 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Butynski (1984) encountered hives in 4% tor of Bwindi Impenetrable National Park. Large (five 1 km² blocks) of Bwindi Forest. Most of trees (>30 cm dbh) are felled for hive construc- these were placed, as they are today, in the tion, and in the Rihija area, the majority of hives bamboo zone, above 2300 m. (93.5%, 106) measured by Mwesigye (1991) Concern was expressed over the following were made from Faurea saligna, and most of the impacts: remainder (4.5% (5)) from Polyscias fulva wood. (1) it was thought that the use of fire when gath- It would be ironic if increasing numbers of bee- ering honey may have been responsible for keepers felled the very trees (Faurea saligna) most of the fires in Bwindi Forest; which provide one of the major nectar sources in (2) on three occasions, Butynski (1984) found the forest. large trees felled to obtain honey from wild Although wooden hives from Faurea can hives; reportedly last for over 20 years, hives poorly (3) large trees were felled for construction of protected from rain are said to last 5 - 8 years; hives and 3 - 5 smaller trees felled when mak- those from Alangium chinense (omukofe) 6 years ing a clearing around each hive. when covered, 4 years uncovered; Polyscias These are valid concerns, which need to be fulva, 5 years when covered; Albizia gummifera, addressed. Fire is now the major threat to con- 4 years when covered, 2 years when exposed to servation of Bwindi Impenetrable National Park. rain. Woven hives, if well constructed, can last Forest fires during the very dry period November 5 - 6 years, and if not protected from rain, only 1991 - March 1992 were not caused by beekeep- 1 - 2 years. ers, but were due to runaway fires from outside Although no data are available on growth the forest, or from deliberate arson. Fire is cer- rates of Faurea saligna, it is likely that the tainly a threat to beehives kept within the forest, resource replaces itself faster than the rate of and in at least one case (Wild pers. comm., deterioration of hives, whilst this is not the case 1992), beekeepers tried to prevent one of these for hives woven from bamboo, papyrus or some runaway fires. climbers. From discussions with beekeepers at a meet- During this survey, woven hives were ing at Ruhija (April 1992), it would appear that recorded in the Ishasha (forest climbers) and beekeepers are genuinely prepared to prevent Katojo bamboo (Arundinaria alpina) areas. runaway fires and as a group would be an impor- Only two trees were recorded felled to get to tant influence against arson or careless use of wild hives: a Carapa grandiflora (omuruguya) fire. of 70 cm dbh felled for Trigonid bee (ebihura) Mwesigye (1991) recorded 63 beekeepers honey, and a Markhamia lutea (omulembwe) of with a total of 469 hives around the southern sec- 48 cm dbh.

Box 2. Recommendations for beekeeping and honey collection

∗ The existing initiative of DTC, IFCP and Bwindi Impenetrable National Park staff to stimulate the formation of local beekeeping societies should be encouraged. This includes the printing of identity cards suggested by beekeepers, and should be on a parish basis. If a beekeeper wants to keep hives in a multiple-use zone in an adjacent parish, he should belong to that beekeeping society as well. Identity cards should not be sold and should not be transferable. ∗ DTC/CARE-International can facilitate development of new markets (including export markets) for forest honey, as is being done in the Oku forest region, Cameroon. Reviving the marketing of beeswax, which existed in the DTC area in the 1960s, should also be considered. ∗ Clearing a 4-5 m diameter area around hives should be permitted. This is important to beekeepers to clear a flight-path for the bees, and it also reduces the risk of runaway fires. Large trees (>15 cm dbh) should not be felled during this clearing. ∗ Trees should not be felled to rob wild hives. Within multiple-use zones, wild hives can be climbed to obtain wild honey. ∗ The DTC project should consider giving special attention to involvement of the Batwa in beekeeping activities. They are extremely knowledgeable about bees. Few Batwa own land, and this would give successful beekeepers an opportunity to earn money from the sale of honey from the forest. ∗ DTC should consider reviving the beekeeping programme that formerly operated in the DTC project area. This should include training courses on improved hive construction such as the "top-bar" hive, which, as far as is known, is only being made by one person in the DTC area at present. ∗ Beekeepers should take responsibility for preventing fires and protecting their hives against chimpanzees and other wild animals.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 19 People, park and plant use A.B. CUNNINGHAM MEDICINAL PLANTS doctor: total population ratio of 1:20,000 cited Agricultural clearing, whilst increasing the by the Ministry of Health (Kakuru, 1990). amount of weedy species used medicinally, has In the Kilungu district, Kenya, for example, been the major cause of depletion of medicinal rural populations of TMPs averaged 1:224 plants from forest. Commercial trade in herbal (herbalists 1:665; traditional birth attendants medicines is not well developed, and no sign of 1:1640 and diviners also 1:665), while in urban large-scale bark or root harvesting was observed Mathare, the overall ratio was 1:883 (Good, in Bwindi Impenetrable National Park. 1987). Economic deterioration in the past, and a Collection of herbal medicines is highly 3.4% per annum population growth rate, increase species-specific, with over 100 plant species the difficulty of providing adequate health ser- involved, primarily for local use. Leaf material vices. Although clinics are found in most sub- forms the major component of plant parts col- counties, hospitals in the Kabale-Rukungiri dis- lected for medicines used to treat people or live- tricts, located at Kabale, Kisoro, Kisizi and stock. "Chewing sticks" used for dental care, are Kambuga are difficult for many people to reach. mainly from stems of plants occurring in dis- Many of the traditional medical practitioners turbed areas outside the forest. Most species col- and traditional birth attendants are registered lected are fast-growing herbs gathered from out- members of TMP or TBA associations, the TBAs side the forest. The impact of medicinal plant using cards associated with UNICEF. Traditional harvesting is low, even for favoured species such birth attendants (TBAs) (or traditional midwives) as Rytigynia kigeziensis (nyakibazi) (Photo 12). play an important role in assisting home births in East Africa. In Tanzania, for example, 75 -80% of births were assisted by TBAs (Anderson and Staugard, 1986). Similarly, in the Machakos area, Kenya, only 26% of pregnant women went to the hospital for delivery (Voorhoeve et al., 1982). With few exceptions in the DTC area, (the use of naturalized Physalis peruviana, Ricinus communis, or cultivated Pennisetum purpureum, Eucalyptus spp., etc.) traditional medicinal plants are all gathered from the wild. A good example is the use of Rytigynia kigeziensis (Rubiaceae) bark to treat intestinal parasites ("worms"). Parasite loads are extremely high in the Bwindi/DTC area. In an examina- tion of stool samples from 35 people, 89% (31) were infested with roundworm (Ascaris), and 34% (11) with whipworm (Trichuris) nema- todes, in addition to ten other types of intestinal parasite (Ashford et al., 1990). Many local people encountered during this survey expressed the view that if they did not have continued access to Rytigynia kigeziensis (nyakibazi) bark, "they Photo 12. Rytigynia kigeziensis (nyakibazi), an effective worm remedy, would die". Sustainable use of medicinal plants rated as the most important medicinal plant to people in the DTC area. is therefore essential, and the focus here remains on key medicinal plants or on those that are com- It is recommended that traditional medical mercially traded. practitioners (TMPs) and traditional birth attendants (TBAs) be allowed to collect medicinal Sustainability of medicinal plant use plant material from multiple-use zones and that Outside the national park, Afromontane for- medicinal plant gardens are promoted by both est which was a source of traditional medicines DTC, ICRAF and local clinics through the in the past has declined drastically through clear- Ministry of Health. ance for agriculture. Exclusion from Bwindi Traditional medical practitioners (TMPs) Impenetrable National Park thus affects TMPs and traditional birth attendants (TBAs) play an who previously gathered medicines there. TBAs, important role in rural communities throughout who gather most of their medicinal plants in dis- Africa. They also have a useful role to play in turbed sites outside of the forest, are less affect- primary health care (PHC) initiatives. ed by restrictions. In addition, supplies of herbal No data are available on the ratio of TMPs or medicines to TMPs are affected by competing TBAs to total population in Uganda. There is no uses such as timber logging (e.g. for doubt, however, that TMPs and TBAs are more Entandrophragma excelsum, omuyovi). There is numerous than medical doctors, with a medical therefore growing demand for fewer resources.

20 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Although a large number of plant species are [22% (34 species)]; to assist women either prior used by TMPs, impact on medicinal plant to, or shortly after childbirth (preventing prema- resources is reduced by three factors: ture labour, assisting labour, removal of placen- (1) the low ratio of TMP and TBA to total popu- ta, treating swollen breasts or improving lacta- lation, and small quantities of medicinal tion) [18% (29 species] and for treatment of plants gathered for use by TMPs and TBAs; internal parasites ("deworming") [7% (11 (2) the low level of urbanization in Uganda (6% species)]. of total population), compared to many other By comparison, in this preliminary analysis, African countries, and the correspondingly leaf material (37% (17 of 46 species) was also low level of commercial trade in traditional the largest plant part category used by TMPs, medicines. In Owino market, Kampala, for but came from a wider range of life forms example, 10 - 12 men sold traditional medi- (trees, shrubs and climbers, in addition to herba- cines in small quantity. By contrast, tradi- ceous plants). The three major categories of tional medicines are sold in bulk in large use were to treat stomach aches [15% (7 species)] cities such as Abidjan, Côte d’Ivoire (107 and internal parasites ("worms") [13% bulk sellers) and Durban, South Africa (392 (6 species)] and for use as antidotes to poisons bulk sellers, over 100 retail shops), generat- (often those administered by jealous rivals) [11% ing a long-distance, species-specific trade (5 species)]. which threatens popular medicines within From field observation, rural areas (Cunningham, 1990). It is worth it appears that the impact of noting, however, that a small-scale trade in medicinal plant gathering is traditional medicines has developed, with low. Only two large speci- bark, roots and other parts of favoured tradi- mens of over 70 Rytigynia tional medicines ordered from people living kigeziensis trees seen near to Bwindi Forest (Table 5, Photo 13). In showed signs of bark 1992, a single herbalist was buying herbal removal, despite the popu- material; larity of this species as a key (3) the fact that (in this study at least), leaf mater- medicinal plant. Both of ial was the most common part of the plant used these were large plants adja- by TBAs (Photo 14, page 22) and cattle own- cent to paths. Debarking of ers. Although data are incomplete, a prelimi- Hallea rubrostipulata was nary analysis of life forms and parts used by observed outside, but not TBAs suggests that of the plant parts used, at inside the national park. By least 62% (96 of 154 species) is leaf material, comparison, debarking by mostly from herbs growing outside the forest. elephant of trees such as The three most important categories of uses Macaranga kilimandschari- Photo 13. The start of commer- of medicinal plants by TBAs were for symbolic ca (omurara), Prunus cial trade in the DTC area: bark or magical purposes (as protective charms africana (omumba) and from Entandrophragma excel- against bad omens, to ensure safe journeys or Rytigynia kigeziensis (nyak- sum (omuyovi) sold by a herbal- assist in love affairs and court cases) ibazi) is much greater. ist at the main village centre.

Table 5. Commercially collected traditional medicines for sale at Batogota.

Family Plant Species Rukiga name Life Form Part Used Araliaceae Polyscias fulva omungo tree bark Clusiaceae Symphonia globulifera omusisi tree bark (**) Euphorbiaceae Croton macrostachyus omurangara tree bark (**) Lauraceae Ocotea usambarensis omwiha tree bark (**) Meliaceae Entandrophragma excelsum omuyovi tree bark (**) Myricaceae Myrica salicifolia omujeje tree bark (**) Myrsinaceae Maesa lanceolata omuhanga shrub roots ? (**) Phytolacaceae Phytolacca dodecandra omuhoko climber leaves Rubiaceae Hallea rubrostipulata omuziku/ngomera tree bark (**)

Note: (**) denotes species collected in maize bag (50 kg size) lots. Other material sold was from the following species, identified by local name only: omurama (**), kashosho, omuhe, omurahusyo, omukoko, kaboha and omuhaka. Information gathered by R. Badaza and J. Tumusiime (16 April 1992).

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 21 People, park and plant use A.B. CUNNINGHAM Seven species used for dental care were identified in discussions at Buhoma. All were obtained outside the forest, from disturbed sites near to villages and homes; these were stems of omuhuke (Lantana trifolium), omusinga, omukyindezi, omuchundura, omusambya (Dodonaea viscosa) and the leaves of two species (ekarwe and omufumbwa).

Traditional veterinary medicines Although pastoralism is not the major form of land-use in the DTC area, cattle still have important economic and social significance in this area. Veterinary medicines and formally trained veterinary doctors are scarce and expensive, and wild plants are commonly used to treat a wide variety of livestock diseases in cattle and goats. Knowledge of traditional veterinary medicines is more widespread than knowledge of medicinal plants used to treat people. Over 20 different types of ailments affecting cattle are recognized. Leaf material forms the bulk of plant parts used (64%, 16 of 25 species, %%) of this small sample. Many of the plant species occur outside the forest. There is no commercial trade in plants for veterinary medicines, impact is considered low and the current level of use sustainable. More detailed work is required on traditional veterinary medicines before firm conclusions can be drawn on major ailments treated in livestock. This preliminary survey showed that the most Photo 14. Herbalist Mr. M. Mafurira with leaf material which commonly used remedies were for diarrhoeal dis- forms a major part of herbal medicines. ease (khitwa), swellings on the legs, ears or groin of livestock (ekibagarila), mastitis and swellings of the udder and teats of cows. Dental care: chewing sticks Dentists are scarce in many parts of Africa, Traditional medicines and hunting dogs particularly in rural areas. The ratio of dentist to Hunting to provide meat and reduce num- total population in Ghana for example, was bers of crop raiding animals such as bush-pig has 1:150,000 (compared to 1:3000 in Great Britain) been an important recreational activity in the (Adu-Tutu et al., 1979). DTC area in the past. In the past, dogs were used Although diet plays a major role in incidence to run animals down, or to drive them into nets of dental caries, dental hygiene is also important. (Butynski, 1984). Dogs are still used for hunting While toothpaste and toothbrushes are widely used in small patches of scrub and forest outside by the sector of the population with a high level of Bwindi Forest, with one hunt seen during this formal education, toothpaste is expensive and survey. sometimes not available in rural areas. Traditional medicines, with symbolic and In the DTC area, in common with many possibly also physiological value, are widely rural areas in Africa, traditional toothbrushes or used in Africa to improve the aggressiveness, "chewing sticks" from thin branches or roots of sense of smell and hunting ability of dogs. In the local plants are still in common use. DTC area, the use of five plant species was Continued access to popular and effective recorded: the bark of Tabernaemontana odor- sources of chewing sticks, many of which have atissima (kinyamagozi) and Schefflera barteri anti-bacterial properties, is important. Unlike the (omwamira), the fruit of Coccinia mildbraedii situation in West Africa, where a thriving trade and another Coccinia species (omutanga), and in chewing sticks has developed, use here is for the leaves of Thalictrum rhynchocarpum (omwi- local purposes only and the impact on vegetation tango). Small quantities of these plants are used, is negligible. and impact is considered to be negligible.

22 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Box 3. Recommendations for medicinal plant use

The aims should be the improvement of self-sufficiency of TMPs, the reduction of harvesting demand on wild populations of vulnerable traditional medicinal plant species in core conservation areas, and the reduction of future harvesting pressure on wild populations. This may be brought about through: ∗ mass production of rooted cuttings and seedlings of key species for home gardens; ∗ dissemination of information on appropriate cultivation methods for medicinal plants in demand locally (e.g. Rytigynia kigeziensis - nyakibazi, Piper spp. - rokokota) or which are known to be effective herbal medicines and are used regionally but not locally (e.g. Warburgia ugandensis (= W. salutaris)), which occurs in Kibale but not in Bwindi forest); ∗ planting of popular medicinal plants (particularly trees and shrubs) as an important part of the reafforestation of slopes where soil loss is high and sustainable agriculture is not possible (buffer zone management around core conservation areas); ∗ supporting traditional healer societies or associations by national parks management. Very little goes unnoticed in communally owned areas, and if problems arise regarding depletion of valued local resources, TMP associations or community leaders can play an important role in resource controls; ∗ formation of rural TMP associations (where these do not exist already), possibly through local health services with support of the Ministry of Health/UNICEF as a focal point for establishment of medicinal plant cultivation with assistance of DTC and possibly ICRAF. ∗ investigation of the practicality of using facilities of government or commercial companies (e.g. the Uganda Tea Co-operative) to produce scarce and popular medicinal plant species from cuttings, using commercially available rooting media as an appropriate means of boosting initial stocks for distribution at cost to herbalists and interested farmers.

Through the Medical Faculty at Mbarara University of Science and Technology, investigate: ∗ the efficacy of popular herbal remedies, e.g. the effects of different concentrations of Rytigynia kigeziensis bark decoctions on intestinal parasites; ∗ toxic species that are known to cause problems locally through over-dosage, e.g. due to liver and kidney toxins; ∗ the desirability of producing medicinal plants of known toxicity through cloning as a means of standardizing dosage and producing a quality end product. The approach taken to cloning Urginea maritima (Liliaceae) by Gentry et al. (1987) is appropriate here. The DTC project provides an opportunity for CARE-International, with its expressed interest in primary health care (PHC) adapted to local technologies and responding to local customs (Anon, 1991), to play a greater role in the PHC and medicinal plants issue (see Wondergem et al., 1989; Desawadi, 1991). A similar approach is taken on a smaller scale by Brother A. Wassawa at Kyotera, who produces herbal recipes for local "first aid kits". Monitoring is needed of: ∗ the development of commercial trade. Permanent plots may need to be set up in a few selected sites to monitor the status of commercially traded "indicator" species, e.g. Hallea rubrostipulata populations; ∗ the success of cultivation, possibly through a register of growers and of species under cultivation.

BASKETRY Loeseneriella apocynoides (omujega). The sec- Basketry combines traditional skills and ond of these species is of greatest concern from local materials to produce a range of woven arti- a sustainable use viewpoint. Clearing of forest cles for storage and processing of agricultural for agriculture, followed by recent development crops or home use. Plant materials used vary of the tea industry led to increased use of from fast-growing, productive wetland species Loeseneriella apocynoides for tea-baskets in (e.g. Cyperus papyrus, efundjo) to scarce, slow addition to other competing uses (for granaries, growing climbers that are found at low density in stretchers and as a general purpose binding mate- forest (Loeseneriella apocynoides, omujega). rial), resulting in over-exploitation. Many species favoured for basketry are Almost without exception, households in the common in disturbed sites and over-exploitation DTC area use baskets for harvesting, drying, win- is unlikely. Specific recommendations are made nowing, grinding and storing agricultural produce. regarding two climbers that are popular for cer- Basketry techniques and plant materials are also tain woven baskets: Smilax anceps (enshuli) and used to weave granaries, fish-traps and stretchers.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 23 People, park and plant use A.B. CUNNINGHAM Although alternatives (e.g. plastic bags) are occasionally used, indigenous plants are the major source of basketry fibre (Table 6). Despite their widespread use, skilled basket-makers are relatively few in number, particularly for specialized baskets made by men, such as winnowing trays and stretchers. The most widely used baskets are the flat, circular baskets which are placed adjacent to the grinding stone to collect ground flour (orugali), the deep, bowl-shaped millet basket (echibo), a larger and shallower bowl-shaped basket for grain (entemere), a larger basket for carrying headloads of crops (etchitukuru) and the winnowing basket (entara). The first three basket types are mainly made by women, using a coil-foundation technique with grass (Eleusine indica, enchenzi) and papyrus (Cyperus papyrus) for the coils, and Plantago palmata flower-stalks for the dark design. Women also make mats (omucheche) for sit- ting or sleeping on, as well as for drying millet (etchigaru). These are made from Cyperus latifolius leaves, bound with twined Triumfetta (omunaba) bark, sometimes with dark banana fibre woven in for a decorative pattern. The etchitukuru basket is woven using a chequer- weave by both men and women, using either Smilax anceps (enshuli) or bamboo (Arundinaria alpina), and is the type of basket most common- Photo 15. Carrying baskets (etchitukuru) for sale at ly sold at markets (Photo 15). Batogota market.

Table 6. Main plant materials used for basketry, woven stretchers and granaries in the DTC area.

Family Plant species Rukiga name Life form Part used Use Agavacae Dracaena laxissima enchence climber stem stretchers Araceae Raphia farinifera ekihunje palm leaf basketry Araceae Phoenix reclinata enkindu palm leaf-stem basketry, stretchers Celastraceae Loeseneriella apocynoides omujega climber stem basketry, stretchers, granaries Celastraceae Hippocratea odongensis oruyangaro climber stem granaries Celastraceae Salacia elegans orudyangara climber stem granaries Cyperaceae Cyperus latifolius ekigaga sedge leaf basketry Cyperaceae Cyperus papyrus efundjo sedge leaf cuticle,culm basketry granaries Marantaceae Ataenidia conferta ebitatara geophyte leaf basketry Marantaceae Marantochloa leucantha omwiru geophyte leaf basketry Malphigiaceae Flabelleria paniculata ? climber stem granaries Plantaginaceae Plantago palmata embatambata herb flower stalk basketry Poaceae Arundinaria alpina omugano bamboo stem basketry, stretchers, granaries Poaceae Eleusine indica enchenzi grass leaf basketry Poaceae Pennisetum purpureum ? grass stem granaries Poaceae Setaria plicatilis ekikoka grass leaf basketry Smilacaceae Smilax anceps enshuli climber stem basketry, stretchers, granaries Tiliaceae Grewia sp. omutahendeka scand. shrub stem basketry, stretchers, granaries Urticaceae Urera hypselodenderon omushe climber stem granaries Note: Species used for supporting poles and thatched roofing for granaries are not included.

24 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM The entara winnowing basket is made mainly by men, using a chequer-weave, often from Smilax anceps (enshuli) weft and Grewia sp. (omutahendeka) warp. A basket that is probably made more commonly now than in the past is the "tea-basket" (orutete), which is preferentially made from Loeseneriella apocynoides (omujega), but as this climber has become scarce, it is also being made from leaf-stems of Phoenix reclinata (enchindo). Tea-baskets are also made mainly by men. Thick (2-4 cm) climbers are used, and split into thinner strips. All basket fibres can be stored in a dry place for later use, Photo 16. Woven stretcher, with weft of Smilax kraussiana (enshuli) and are soaked before weaving. and warp (length wise) of Phoenix reclinata (enchindu), supported by bamboo poles (the exotic Bambusa sp). Ishasha area, May 1992. Stretchers (engozi) Woven stretchers (engozi) (Photo 16) play a made: a smaller, tightly woven type for beans, and very important role in rural communities in the a larger granary for grain (finger millet, sorghum) Rukiga highlands for the transport of sick (or and root crops (sweet potatoes, potatoes). dead) people. With the exception of Phoenix Materials used, and quality of construction reclinata (enchindu) leaf-stems they are entire- vary considerably, however (see Table 6). These ly constructed from forest plants (see Table 6) include crop surplus such as sorghum stalks, culti- and are a mainstay of stretcher-bearer societies vated plants like Pennisetum purpureum (elephant (ekyibinachengozi (ekyibina = society; engozi = grass) or forest plants like bamboo and various stretcher)) that are something like a local med- climbers (Photo 17). ical aid association. These associations are well organized, with a chairman and secretary. Monthly financial contributions are made by men (c. 200 shillings per month) and women (c. 50 shillings per month) to cover the cost of food for journeys and for buying new stretchers, which last 2-4 years, depending on the materials used, the toughest material being Loeseneriella apocynoides (omujega). There is usually one association per "cell" and usually 7-8 per parish. Stretchers are made mainly by men, who sometimes also weave winnowing baskets, and there are very few of these specialists within the DTC area. There are none in the Ruhija area, for example, and society members have to get stretchers from the Rubanda area. Similarly, the single stretcher weaver in the Rushaga area reportedly supplied stretchers to stretcher-bearer societies in the Rubuguli, Remero and Kaara parishes.

Granaries Post-harvest loss is a worldwide problem, with losses of cereal crops of between 10-20% in developing countries (FAO, 1981). From discussions with local farmers, the DTC area is no exception, whether due to insects, fungi or less commonly, baboons breaking into a poorly made granary to eat finger-millet. Resolving this problem complements attempts to increase food production and improve food security being undertaken by CARE-International. Granaries are the major means of storing crops in the DTC area and are found at every homestead. Photo 17. Granary in the DTC area constructed from the forest Men construct most granaries, with two basic types climber Loeseneriella apocynoides (omujega), thatched with Imperata cylindrica (Bujengwe parish, adjacent to Kitahurira).

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 25 People, park and plant use A.B. CUNNINGHAM Sustainability of basketry resources root-stock, combined with dispersal by birds into For resource management purposes, basketry forest margins and disturbed sites is thought to materials could be divided into five categories: avoid depletion of plants within forest patches. (1) species common in disturbed sites such as Regeneration time for Smilax was considered to roadsides and old fields: Eleusine indica be 6-12 months by resource users. Bamboo use (enchenzi), Plantago palmata (embatambata); is discussed below (page 27). Use of Urera (2) locally common wetland species: Cyperus hypselodendron and Flabellaria is limited to papyrus (efundjo), C. latifolius (ekigaga); granaries, and impact is considered negligible. (3) bamboo, climbers, and a scandent shrub of Category 4: The cuticle of leaf-stems (both disturbed scrub and forest, with Smilax Marantaceae) or of young leaves (Raphia) is anceps (enshuli) and the scandent shrub used. This limits harvesting impact on the plants, Grewia sp. (omutahendeka) most popular, although they are only locally common in moist followed by the high altitude bamboo, valleys. If commercial harvesting is developed to Arundinaria alpina, and then to a lesser satisfy large-scale commercial production, then a extent Urera hypselodendron (omushe) and management problem could arise due to high Flabellaria paniculata, which are used for intensity and frequency of defoliation, and possi- "bush rope" or for granaries; bly also to trampling or uprooting of these (4) members of the Marantaceae restricted to plants. This has been recorded elsewhere for moist valleys and gullies in lower altitude usually resilient mat-rush (Juncus) and palm (1500-1750 m) sites in the forest - (Hyphaene) species in southern Africa Marantochloa leucantha (omwiru) and (Cunningham and Taylor, 1983; Cunningham Ataenidia conferta (ebitatara); or the palm and Milton, 1987). Baskets from Raphia cuticle Raphia farinifera (Arecaceae) which is also are only made on a limited scale, and impact is restricted to moist valleys in the Nteko and considered to be negligible. Buhoma areas at 1500 m; Category 5: The impact of cutting Salacia (5) climbers in the Celastraceae found in older sp. (bwara), Hippocratea odongensis (oruyan- secondary forest and mature forest. The garo) and Salacia elegans (orudyangara) for most widely used is Loeseneriella apo- granaries is unknown, but from discussions with cynoides (omujega) and Salacia sp. local resource users is judged to be far less than (bwara), with others being Hippocratea that on Loeseneriella apocynoides (omujega), odongensis (oruyangaro) and Salacia ele- which is considered to be the slowest growing of gans (orudyangara). these forest climbers. It is also in greatest These categories represent increasingly scarce demand for a wide variety of uses, from a gener- species with more and more specific habitat al purpose "bush rope" of exceptional strength, requirements. to use for tea-baskets, stretchers and granaries. Categories 1 and 2 are common and wide- According to resource users, it takes 10-20 years spread, and their use would be sustainable. for Loeseneriella apocynoides to reach a high Basketmakers select patches with a high density quality useable diameter (3-4 cm), although of Eleusine indica or shady sites where Plantago there are no data on climber growth rates to con- plants have longer flower-stalks more useful for firm this. From field observation, however, it is coil-built basketry, or cut small quantities of clear that L. apocynoides occurs at a low density Cyperus latifolius leaves or young culms of C. in lower altitude (1500-1750 m) forest (e.g. papyrus. Buhoma area), with single specimens seen in the Category 3 is limited to forest or forest mar- Ihihizo and Ishasha valleys. Plants appear to gins, with the climber species common in sprout after cutting of the large stems, and pre- canopy gaps. Of these four species, Smilax is in sumably this removal of large stems represents the greatest demand, despite the difficulty of loss of mature stems producing flowers and working with this tough material. Although seeds. widespread in disturbed sites in East, Central and Loeseneriella has come under markedly southern Africa, it is considered scarce by bas- increased harvesting pressure with the develop- ketmakers in the DTC area due to the intensive ment of the tea industry around Bwindi land-use around the forest, with fallow periods Impenetrable Forest, as it is the most favoured too short to enable its dispersal and growth. species for making tea-baskets. According to Exceptions to this is the Nteko area, where pop- data from the Uganda Tea Growers Corporation ulation densities are lower, and forest patches (UTGC) factory at Batogota, there are 1597 still occurring outside Bwindi Impenetrable out-growers in the DTC area. Assuming that National Park. Harvesting is probably higher there are three pickers per out-grower, each with along paths or roads where disturbed sites are a basket made from L. apocynoides (weighing c. more accessible. Thicker stems are used, and 500 g and said to last four years), this would rep- although local over-exploitation takes place, resent the use of 2.4 tons of this climber every recruitment with regeneration from the four years (or approximately 600 kg per year).

26 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM With the low density of this species in the forest, Bamboo is a widely used forest product of presumably indicating old canopy gaps, tea bas- great importance to rural communities in many ket making has had a major impact on this tropical forest areas, particularly in Asia, but also climber, in addition to its use for stretchers and in East and Central Africa. Arundinaria alpina granaries. A stretcher maker in the Rushaga area, thickets are found in Afromontane forest in East for example, reported having to travel an esti- Africa from 2400-3000 m, occurring to 3200 m mated 5 km into the forest (almost to Mubwindi on Mount Kenya and as low as 1630 m in the swamp), spending two days to collect enough Uluguru mountains (White, 1983). material to make a stretcher. Scott (1992) also In Uganda, bamboo is cut in the Rwenzori, encountered many resource users, including Mount Elgon, Mgahinga, Echuya and Bwindi tea-pickers, who were having increasing difficul- Impenetrable forests (Howard, 1991). It is also ty in obtaining this species. cultivated on a small scale in the DTC area. Bamboo thicket occurs in a limited 0.4 km² high BAMBOO altitude area in the south-east of Bwindi Bamboo (Arundinaria alpina), restricted in Impenetrable National Park (Butynski, 1984). the wild to the high altitude south-east of Bwindi Forty-one percent (48) of respondents surveyed Impenetrable National Park, is a very important in the DTC area used bamboo, probably from resource for home and granary construction in cultivated sources (Arundinaria alpina and the the Katojo, Mushanje and Nyamabale parishes. exotic bamboo Bambusa sp.) (Table 7) as well as Young bamboo culms also provide a valuable from the wild (Kanongo, 1990). material for basketry, for commercial sale and From field observation in the parishes adja- home use. Bamboo for this purpose is obtained cent to the south-eastern corner of Bwindi from Echuya Forest Reserve, and probably also Impenetrable Forest it is clear that home con- from Bwindi. In contrast to Mount Elgon, bam- struction (cross-pieces) is a major use of bamboo shoots are not eaten by people in this area. boo, followed by use for granaries and baskets Although theoretically not permitted when (Table 7). this study was carried out, harvesting of bamboo Bamboo was one of the most important still took place within Bwindi Impenetrable "minor forest products" sold by the Forest National Park. Quantitative resource surveys Department, with almost 500,000 bamboos sold indicate a low level of cutting in the past. Work annually from former Central and Local Forest with resource users also showed that a high pro- Reserves between 1961 and 1962 and between portion of bamboo culms (stems) were unaccept- 1963 and 1964 (1961-62: 515,000 bamboos; able for building purposes due to the high inci- 1962-63: 450,000 bamboos; 1963-64: 459,882 dence of borer attack (moth larvae). Although it bamboos) (Forest Department, 1964). is a key resource for people and wildlife, few Although not mentioned in the former data are available for Uganda on the biology and forestry working plan for Bwindi Forest (Leggatt biomass production of A. alpina, and research and Osmaston, 1961), plans were drawn up for work is recommended. It is also suggested that regulated cutting in Mgahinga Forest, where an an adaptive management approach is taken to average of 77,400 bamboos were cut annually bamboo harvesting in multiple-use zones cover- from 1955 until 1966-67 from within four ing the bamboo thicket. This could be undertak- coupes, one harvested per year (Kingston, 1967). en seasonally by licensed bamboo harvesters The previous extent of use in Bwindi from the Katojo, Mushanje and possibly Impenetrable Forest is unknown, although a sur- Nyamabale parishes, and permitted on a trial vey by Kanongo (1990) indicates that out of 54 basis. Development of edible bamboo shoot har- respondents, 92% (50) obtained bamboo from vesting for local or external markets is not rec- Bwindi Forest, with or without licensing ommended. (Table 7).

Table 7. Extent of use, users and source of bamboo in the DTC area (data from Kanongo, 1990).

No. users (n = 116) Use (n = 52) Bamboo source (n = 54) Yes 48 (41%) Other (fences, firewood and From forest (freely taken): 26 (48%) No 68 (59%) home construction): 40 (77%) From forest (bought Granaries: 7 (13%) through Forest Department): 24 (44%) Baskets: 5 (10%) Grown at home: 4 ( 7%)

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 27 People, park and plant use A.B. CUNNINGHAM Box 4. Recommendations for basketry species

∗ Quality of granary construction varies considerably, from excellently made granaries enabling effective protection of crops to flimsy granaries that do not. Skilled granary weavers are well known within each community, and it is suggested that these skilled local people are involved in teaching improved granary design to DTC farmer groups, for example by using Pennisetum purpureum (elephant-grass). ∗ Research into major causes of stored crop losses and appropriate solutions to this problem needs to be undertaken, possibly by Mbarara University of Science and Technology (MUST). ∗ Form basket-making societies in each parish and meet with the existing stretcher-bearer societies to discuss resource use issues covered here. Licenses enabling members to collect forest plants from multiple-use areas would be on a similar for- mat to those issued to beekeepers. ∗ CARE-DTC could facilitate the commercial marketing of finely made baskets (e.g. finger-millet baskets) to improve local income and keep traditional skills alive, either through export or sale to tourists at camps being established with the development of the Tourism plan for Bwindi Impenetrable Forest National Park. The quality of these baskets can compete with basketry worldwide. Materials for commercially produced baskets are discussed below. ∗ Unrestricted use of Eleusine indica (enchenzi), Plantago palmata (embatambata), Cyperus papyrus (efundjo) and C. latifolius is suggested. These are the most commonly used species apart from bamboo, and are the materials which could be used for finely made, export-quality basketry. ∗ Unrestricted local use of Raphia farinifera leaves for basketry. ∗ It is recommended that Smilax anceps (enshuli), Marantochloa leucantha (omwiru) and Ataenidia conferta (ebitatara) are harvested seasonally by two "specialist harvesters" selected by the parish society. For Smilax, this would relate to most parishes, but for Marantochloa and Ataenidia, this would only apply in lower-lying parishes (e.g. Mukono, Karangara and Rubimbwa). An open season would be during the peak basket-making time of the year (at a less busy point in the agricultural cycle, probably May-August). ∗ Discussions should be held involving management staff from Bwindi Impenetrable National Park, Uganda Tea Growers Corporation (UTGC) and DTC regarding the management of Loeseneriella apocynoides (omujega). This is a problem of common concern to all organizations, as availability of this species affects tea-picking, income to local out-growers and forest. At present, UTGC may be unaware of the over-exploitation of this forest climber. They also appear mistakenly to regard this tea-basket weaving material as a "free good", whereas UTGC supply yellow "picking jackets" and gum-boots to their out-growers. L. apocynoides (omujega) is a relatively scarce species which is important for many other purposes in the surrounding community, particularly for stretchers (engozi), and not only to tea-growers. Sustained management of this species is needed, and this should involve the tea-growers and UTGC. For example, the UTGC may have nursery facilities and staff for tea plantations who could also assist in experimental cultivation of L. apocynoides (omujega), and probably more effectively, collection of seed and cultivation of Phoenix reclinata (enchindu, wild date palm), which is already used as a substitute for L. apocynoides in weaving tea-baskets in the Ishasha area. The palm leaf-stems can be cut without damage to the plant. Phoenix reclinata is faster growing, occurs along alluvial plains of the Ishasha river and is a common palm in swamp forest or on termite mounds in seasonally flooded grassland in Uganda. The UTGC staff may also be able to assist with cultivation of other species (see below). ∗ Subject to more detailed field research, a closed period of at least four years should be considered for Loeseneriella apocynoides (omujega), after which the stretcher-bearer society (ekyibinachengozi) groups should have precedence in controlled harvesting of this species. There are very few stretcher-makers in the DTC area, and they collect their L. apocynoides (omujega) in forest as close to their homes as possible. It is unlikely that they would be prepared to travel long distances in difficult terrain to fit in with a rotational system, yet a single multiple-use area for a parish is unlikely to have enough material for sustained use of this species, and rotational harvesting is a useful alternative to either over-exploitation or a complete ban on use of this species. It is suggested that the following approach be implemented on a trial basis: (i) through the stretcher bearer societies, determine the number and distribution of stretcher makers; (ii) after discussion about this problem of common concern, involve the stretcher-bearer society members from the 18 parishes immediately adjacent to the forest in collection of material on a rotational (20-year rotation) basis by selected society members, for supply to the society's stretcher-makers; (iii) through these L. apocynoides gatherers, and if possible, fixed forest plots inside and outside multiple-use areas, monitor the status of this species. ∗ DTC, through community extension agents (CEAs) and nurseries (e.g. Buhoma and Kitahurira nursery), should assess the potential for cultivation of Marantochloa leucantha (omwiru) and Ataenidia conferta (ebitatara). Some farmers are already cultivating Cyperus latifolius in the Ishasha area (May 1992).

28 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Management of bamboo thickets for sustain- * a percentage of culms are crooked, broken able use is relatively simple compared with or too young. Afromontane forest, as bamboo thickets have a (3) Young shoots, which are produced annually lower diversity and complexity in terms of during the rainy season (Were, 1988), break age/size classes and uses. Bamboo is an impor- off easily, and would be affected by harvest- tant resource which is fast-growing and relative- ing activity. Young shoots are also eaten by ly resilient to harvesting, with new culms pro- animals (mainly primates), with an average duced from underground rhizomes. Current lev- of 15.5% of shoots eaten out of four sample els of harvesting are relatively low, even in plots. favoured sites (Figure 3, page 30). (4) Cutting intensities reportedly affect regrowth It is suggested that harvesting be considered rates. With clear-cutting, it took 8-9 years to within multiple-use zones, with potential har- obtain full-sized culms; if 10% of old culms vesters and DTC/national parks staff involved in were left standing and evenly distributed, resource surveys and setting quotas as an "action full-sized new culms could be produced after research" project. The following factors need to 7-8 years, and if 50% of culms are left stand- be taken into account, however, in considering ing, the recovery period may be reduced to this proposal. 3-4 years (Wimbush, 1945). Although these (1) Harvesting within the bamboo thicket in plots were placed near to the road, the level of Bwindi Impenetrable Forest is not uniformly harvesting was low. This could be attributed spread, but is largely restricted to three parts to "closure of the national park to bamboo of the bamboo zone (Mpuro, Omushenje and cutting", but a similarly low level of bamboo Kasule). This is a result of the difficult cutting was also observed along the main access from surrounding parishes, the prob- path to the densely populated parish adjacent lems of transporting long (often >5 m) bun- to the bamboo zone. dles of cut bamboo through forest in steep (5) The previous system of selling licenses as a terrain, and possibly also due to fear of ele- means of controlling and monitoring the har- phants. Setting of quotas and a rotational vesting of forest produce (including bamboo) management system should therefore not be failed for the reasons given by Howard based on the entire area of bamboo, or even (1991), where declining purchasing power of that within multiple-use zones, but on a far forestry salaries led to lack of commitment of smaller area with a consequently smaller staff to their work and unofficial sale of bam- carrying capacity. boo or cutting licenses to earn supplementary (2) Although above-ground biomass of income. It would be important to avoid this Arundinaria alpina is high (100 tons per ha problem in the future if sustained harvesting (Wimbush, 1945) and growth rates are fast of bamboo is to be implemented in Bwindi compared to forest trees, with culms reaching Impenetrable National Park. In Mghahinga full height in 2-4 months with stems senesc- forest, it was noted that although bamboo har- ing after 7-14 years (Were, 1988), this study vesting was prohibited within the SNR (strict indicates that a much lower proportion of nature reserve) section of Mgahinga, there stems (or biomass) are suitable for harvesting were extensive signs of illegal bamboo cut- for the following reasons, lowering the carry- ting, and recommended that harvesting pres- ing capacity of bamboo stands: sure be shifted to more abundant and better * although mature culms dominate bamboo quality stands of bamboo in Echuya Forest thicket, the majority of these would be Reserve. The extent of managed use of bam- rejected by harvesters due to a high inci- boo through licensed harvesting in Echuya dence of moth larval borer attack Forest is uncertain, however (K. Sucker, pers. (Figure 3); comm., 1992).

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 29 People, park and plant use A.B. CUNNINGHAM 100 100 90 90 80 80 70 70 60 60 50 50 40 40 Number of stems Number of stems 30 30 20 20 10 10 0 0 Shoots Full Height Dead Cut Shoots Full Height Dead Cut Bamboo category Bamboo category

Undamaged Damaged Eaten Good for building Borer Too small Crooked Dead or cut Other

100 100 90 90 80 80 70 70 60 60 50 50 40 40 Number of stems 30 Number of stems 30 20 20 10 10 0 0 Shoots Full Height Dead Cut Shoots Full Height Dead Cut Bamboo category Bamboo category

Figure 3. Data from four plots in the bamboo zone, assessing shoot density and damage, and the value of bamboo stems within 10 x 10 m plots for building purposes. Note the high proportion of stems that are unsuitable for building purposes (due to borer damage, crooked shape, small size or other factors), the proportion of young shoots eaten by primates and the low level of bamboo cutting.

Box 5. Recommendations for bamboo

∗ Detailed mapping of the bamboo thicket in Bwindi Impenetrable Forest is needed. ∗ Sustainable harvesting should be carried out on a trial basis, with resource users involved in the decision-making process regarding resource assessments and management. Harvesting blocks/coupes need to be established in the Mpuro, Mushenje and Kasule areas, and resource assessments carried out with resource users (bamboo cutters) selected through the RCs of the Katojo, Mushanje and possibly Nyamabale parishes. Quotas should be set on the basis of resource assessments. ∗ Two cutting seasons should be considered (possibly July-August and January-March). ∗ A limited number of specialist harvesters should be licensed to harvest for people in the parish who require bamboo. Licenses should be similar to those issued to beekeepers, and should not be sold or be transferable. Separate licenses need to be issued to basketmakers. ∗ Cultivation of bamboo should be an important component of the DTC agro-forestry programme. ∗ Research is needed on biomass production and effects of harvesting on Arundinaria alpina. Most recommendations and information on these issues (Kingston, 1967; Kigomo, 1988; this study) are based on the results of a short-term study published 50 years ago (Wimbush, 1945). ∗ Additional research on the population biology and gap dynamics of Arundinaria alpina is important for resource use and maintenance of this vegetation type as management objectives for Bwindi Impenetrable National Park. White (1983), for example, suggests that trees scattered in bamboo stands become established in the 30-40 year intervals of bamboo flowering and die-off. According to Glover and Trump (1970), Arundinaria stands are fire-induced in what was formerly Juniperus forest on the Mau Range, Kenya. Whether this is the case in Bwindi Impenetrable Forest or not, or whether establishment of bamboo within forest, or trees within bamboo is due to other factors (e.g. elephant and "canopy gaps" formed in bamboo due to the combined effects of wind and borer attack weakening mature culms, both of which have been observed in this study) is uncertain, and needs to be investigated.

30 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Major forest products: wood Due to their soft wood however, both Polyscias fulva and Musanga leo-errerae are BLACKSMITHS AND BELLOWS avoided for other uses (building, timber, beer boats). Felling of these trees for this purpose Blacksmiths are a small but important group would be on a limited scale and restricted to sec- of specialists who play a valuable role in the ondary forest, and it is recommended that utliza- farming community of the DTC area, producing tion in multiple-use areas be permitted. agricultural implements and tools (photo 3, page Both exotic and indigenous tree species were 9). Smelting of haematite is no longer carried recorded as used by blacksmiths for charcoal. out, but scrap metal is re-worked into tools and Black wattle (Acacia mearnsii) is favoured for hardware (e.g. hammers, locks, dog bells). this purpose, a situation unchanged from that of The secondary forest tree species Polyscias the late 1960s (White, 1969). Favoured indige- fulva (omungo) (photo 4, page 9) is the major nous species are Syzygium guineense (omugote) source of wood for construction of bellows, in low-altitude sites, and Agauria salicifolia whilst exotic species such as black wattle (etchigura). Parinari excelsa (omushamba) and (Acacia mearnsii) have been favoured for char- Sapium ellipticum (omushasha) are also used. coal use since at least the 1960s. It is recommended that blacksmiths are CANOES allowed to continue to harvest Polyscias trees for Dug-out canoes are made and used at only construction of bellows. The role that black- one locality in the DTC area (Lake Bunyonyi, smiths can play in rural development and possi- bordering on Nyarurambi parish). All of these bly sensitively planned specialist tourism needs canoes are carved from cultivated trees, primari- to be recognized. If not, this traditional skill and ly Eucalyptus (82%), and none is cut in Bwindi technology will disappear. Impenetrable Forest. Canoe construction is not Wood is used by blacksmiths (omuhesi) for therefore considered applicable to forest multi- two main purposes. First, in construction of bel- ple-use areas in this study. lows (omuzuba), where large trees with low density ("soft") wood are selected, and second, for WOOD CARVING - HOUSEHOLD ITEMS charcoal, where high density woods are preferred. A range of wooden items is found in most The use of bellows for iron-working repre- households in the DTC area and these are impor- sents an historical link with the technology intro- tant in food processing (stamping mortars and duced to this region some 2000 years ago. Both pestles for grain and groundnuts), collection the technology and the traditional knowledge this (milk pails) and consumption (spoons, beer represents is disappearing, however, due to com- mugs) (Photo 18, page 32). petition from industrially produced goods. In Tree use for these utensils is often selective, 1968, only four blacksmiths interviewed in with hardwoods required for mortars, while soft- Kigezi by White (1969) claimed to smelt er species are acceptable for beer mugs and milk iron-ore. Although favoured sites for collection pails. Hardwoods are also important for walking of haematite and smelting technology are still sticks, while Rapanea melanophloeos (omukone) known, iron-smelting no longer takes place. is used for carved walking sticks for commercial Instead, scrap metal from old cars or from broken sale (Table 8, page 33). agricultural tools is reworked. Although commercial scale harvesting con- Blacksmith numbers are considered to have centrating on a single species (e.g. Rapanea declined, and most blacksmiths are older men melanophloeos) may result in localized over 50 years. White (1969), for example, over-exploitation if not regulated, impact of use records 23 blacksmiths working in the Kitumba for carving is small compared to uses such as cut- area in Kigezi. From enquiries made in this sur- ting for bean stakes or beer boats. vey, it would appear that at most only 2-4 blacksmiths work in each parish, and in some parishes Beer boats (obwato) there are none. Beer boats, the carved wooden troughs used Most blacksmiths have one set of bellows; for brewing banana beer (tonto) are a very impor- most (n = 9) of these are made of Polyscias fulva tant item to banana farmers adjacent to the north- (omungo) wood, and a single bellows was made ern sector of Bwindi Impenetrable National Park. from Musanga leo-errerae (omutunda). Both are As an alternative to selling the bananas, they pro- favoured for their soft wood, which can be hol- vide a means of processing and "adding value" to lowed out to form the drums and pipes of the bel- the crop of embiri, kisubi, musa or endizi banana lows. Despite the low density of Polyscias wood, varieties used to make banana beer, which is then bellows last 20-30 years, which probably transported to village markets. exceeds the time that it would take P. fulva to The irony is that in the process of clearing reach a suitable size for bellows construction land for agriculture, including land for bananas, (40-50 cm dbh). most of the large trees (>50 cm dbh) suitable for

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 31 People, park and plant use A.B. CUNNINGHAM making beer-boats have been destroyed. In addition, stocks of some of the most favoured tree species used for beer boats (e.g. Prunus africana and Newtonia buchananii) have already been over-exploited by pitsawyers. Although not per- ceived as in critically short supply at present, it can be expected that in the forseeable future, Bwindi Forest will be seen as the major remaining source of beer boats. Beer boat production should not be considered for multiple-use areas at present. Emphasis needs to be placed on encouraging more extensive planting of Markhamia, Ficus and Erythrina trees as alternative sources of beer boats, and investigating other alternatives used in tonto producing areas, for example in Ankole District and near to Kampala, where deforestation has already taken place, but tonto is still produced in vast quantities. These alternatives include banana juice extraction in troughs lined with cowhide (J. Baranga, pers. comm., 1992) or cement. Although bananas are grown throughout the DTC area, the most important parishes for banana production, and tonto making in particular, are the Nteko, Mukono, Kanungu and Karangara areas (Photo 19). With the exception Photo 18. Carving spoons from Markhamia lutea of poor farmers with very little land who are (omusavu) wood. unable to produce a banana surplus for marketing tonto, all farmers cultivating bananas have at least one beer boat. At Nteko, in the sample of 35 farmers owning beer boats, 63% (22) owned a single beer boat, 34% (12) owned two beer boats each, and the remaining farmer owned three beer boats. At Nteko, a farmer with a single beer boat brewed twice a month, with 175 litres of tonto produced each time. This represented an income of 20,000 shillings per beer boat (or 40,000 shillings per month, with tonto sold for 2500 shillings per 20 litre jerry-can), a very important aspect of economic production in the DTC area. Indigenous trees were used for all the beer boats measured (n = 79) in Nteko and the Ngoto area. Hardwood trees (Newtonia, Prunus) are favoured for their durability, and Ficus species, with less dense or durable wood, because of their size. Newtonia buchananii and Ficus sur were the most commonly used for beer boats in the Nteko area (Table 9), whilst unidentified Ficus species (probably F. ovata and possibly F. sur) and Prunus africana were most commonly used in the Ngoto area. This results in selective, localized removal of large Ficus trees that are a "keystone" species, being a major food source for frugivorous birds and primates. Eucalyptus trees were considered to be unsuitable because they cracked too easily, Photo 19. Beer boats provide an important although they are used for making canoes at Lake means of processing certain banana varieties in Bunyonyi. Beer boats are constructed and then order to add value and reduce the weight of carried by a group of men to the banana planta- banana product to be transported. This is done tion. Most, 92% (42), beer boats at Nteko, and by trampling the bananas to remove the juice. 61% (17) at Ngoto were less than 9 years old, and

32 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Table 8. Plant species recorded used for wood carving in the DTC area.

Family Plant species Rukiga name Life form Use Alangiaceae Alangium chinense omukofe tree pestles Apocynaceae Pleiocarpa pycnantha omutoma shrub pipe-stems Bignoniaceae Markhamia lutea (**) omusavu tree spoons, mortars Euphorbiaceae Drypetes gerrardii omushabarara tree pestles, sticks Euphorbiaceae Drypetes bipindensis omushabarara tree spear handles, sticks Euphorbiaceae Sapium ellipticum omushasha tree pestles Fabaceae Millettia dura omutate tree hoe handles Flacourtiaceae Rawsonia spinidens omusalya tree combs, walking sticks Moraceae Ficus asperifolia omushomora shrub sandpaper Moraceae Ficus exasperata omushomora shrub sandpaper Myrsinaceae Rapanea melanophloeos omukone tree carved sticks Rosaceae Prunus africana omumba tree mortars Rubiaceae Rothmannia longiflora oruchiraje shrub spear handles Rubiaceae Aidia micrantha orube tree spear handles Note: Species cultivated are marked (**).

few beer boats in either area lasted more than 12 BUILDING POLES years. Durability depends on care of beer boats, Building poles are required throughout the which last longer if stacked off the damp ground DTC area. Selection of poles is based on a need on poles or stones, and kept under a shelter built for straight and preferably durable trees or tree in the banana field. Many beer boats, however, ferns of a suitable diameter (5-15 cm dbh, Table are left on the ground, and merely covered with 10, page 34). banana leaves in between brewing times. Taking Favoured indigenous forest species are all beer boats measured into account (n = 79), Drypetes spp. (omushabarara), particularly 72% (57) were greater than 40 cm diameter Drypetes ugandensis and D. gerrardii, (Figure 4). From these data, and discussions with Tabernaemontana sp. (kinyamate), Harungana local banana farmers, it was considered that madagascariensis (omunyananga) and the these were made from trees >50 cm dbh. tree-fern Cyathea manniana (omungunza) for

Table 9. Number of beer boats made from tree species in the Nteko and Ngoto areas.

Species Beer boats Nteko (n=46) Ngoto (n=32) Newtonia buchananii 23 Ficus sur (= F. capensis) (omulehe) 8 Markhamia lutea (omusavu) 6 1 Ficus spp. (ekyitoma)414 Prunus africana (omumba)2 7 Ocotea usambarensis (omwiha) 3 Entandrophragma excelsum (omyovi) 1 Macaranga monandra? (ekifurafura) 1 Albizia gummifera (omushebeya) 1 ekiko 1 Sapium ellipticum 3 Erythrina abyssinica 1 Allophyllus sp. 1 ekywezu 1 indet. 1

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 33 People, park and plant use A.B. CUNNINGHAM 40

30 Nteko

25 Ngoto

15 Number of stems 10

0 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 Beer boat diameter [cm]

Figure 4. Diameter size class distribution of beer boats measured on small banana farms in the Nteko and Ngoto swamp areas of the DTC project.

durable support poles and roofing, with to the comparatively low density of suitable Arundinaria alpina (omuganu) favoured for poles. In a recent survey conducted in the DTC cross-pieces. area with 120 respondents, Eucalyptus (88%, Compared to cutting of poles from woodlots 106) and Acacia mearnsii (49%, 59) were the (Eucalyptus, Acacia mearnsii or Sesbania), har- species most preferred for building and had vesting from forest is more labour-intensive due respectively been planted by 77% (92) and 36%

Table 10. Plant species whose stems are used for building poles in the DTC area.

Family Plant species Rukiga name Life form Apocynaceae Tabernaemontana holstii kinyamagozi tree (SF) Apocynaceae Tabernaemontana odoratissima kinyamagozi tree (SF) Apocynaceae Tabernaemontana sp. kinyamate tree (SF) Clusiaceae Harungana madagascariensis omunyananga tree (SF) Cyatheaceae Cyathea manniana omungunza tree fern Euphorbiaceae Bridelia micrantha omujimbu tree (SF, CP) Euphorbiaceae Croton megalocarpus omuvune tree (SF) Euphorbiaceae Drypetes bipindensis omushabarara tree Euphorbiaceae Drypetes gerrardii omushabarara tree Euphorbiaceae Drypetes ugandensis omushabarara tree Euphorbiaceae Macaranga kilimandscharica omurara tree Euphorbiaceae Sapium ellipticum omushasha tree (CP) Fabaceae Acacia mearnsii (O **) obulikoti tree Fabaceae Baphiopsis parviflora omunyashandu tree Fabaceae Newtonia buchananii omutoyo tree (CP) Lauraceae Ocotea usambarensis omwiha tree (C, CP) Melastomataceae Dichaetanthera corymbosa ekinishwe tree Myrsinaceae Maesa lanceolata omuhanga shrub Myrtaceae Eucalyptus spp. (O **) uketusi tree (C) Olacaceae Strombosia scheffleri omuhika tree (CP) Rubiaceae Galiniera saxifraga omulanyoni shrub Sapotaceae Chrysophyllum gorungosanum omushoyo tree (CP) Ulmaceae Trema orientalis omubengabakwe tree indet. omukarati tree indet. omuzo tree Note: cultivated species (**), those mainly occurring outside of the forest (O) and those which coppice readily (C). Canopy tree species are marked (CP) and secondary forest species (SF).

34 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM (43) of respondents (Kanongo, 1990). From field (1991) for the Bwamba region, western Uganda observation, it is clear that many homes in the (0.27 m³ wood per household per year, or 0.038 DTC area are built from these cultivated tree m³ per person per year). Although low compared species (particularly Eucalyptus), with the use of to other parts of Africa (e.g. 1.5 m³ per person exotic species increasing with distance away per year in Owambo, Namibia; Erkkila and from the forest. It is recommended that self-suf- Siiskonen, 1992), population densities and num- ficiency in building materials is facilitated ber of households, and consequently the demand through development of nurseries and supply of for building poles, adjacent to Bwindi Forest are seedlings to interested growers. Opening of mul- very high (Figure 2, page 5). tiple-use areas to allow harvesting of building Occurrence of good quality building poles is poles is not recommended. patchy, depending on forest structure and species Housing is a basic need, and in the DTC area composition. Where poles occur, they are often cultivated and wild plant resources are an impor- at a low density (an average density of 207 very tant source of most low-cost housing material. good poles per ha, or a total of 525 useable poles Although corrugated iron is favoured and com- per ha (n = 7 plots) (Figure 5). This is very low monly used, it is difficult to get since the closure compared to a stand density of 1363 stems per ha of the Rwanda-Uganda border, and it is also even in 12 year old (thinned) commercial stands expensive. Many homes are therefore thatched of Acacia mearnsii with a 14.4 cm mean dbh with banana fibre, or if adjacent to wetlands, with (Schönau, 1970) and a far higher density in Cyperus latifolius sedge. local Eucalyptus plots. Harvesting of thatching materials from with- These factors, coupled to the steep terrain, in Bwindi Impenetrable Forest is negligible, and make cutting of building poles from the forest a the two main categories of building material rel- very time consuming activity, and it is not sur- evant to multiple-use areas are building poles and prising that, apart from farmers living close to bamboo. As bamboo is a multiple-use material the forest, most people either cultivate or buy occurring in a limited area of Bwindi building poles of Eucalyptus, Markhamia or Impenetrable National Park, it is discussed sepa- Acacia mearnsii. Cultivation of building poles is rately (see page 27). the major reason for exotic tree planting in the Wood consumption for building purposes is DTC area (see Table 10) and in Bwamba probably similar to that estimated by Howard (Howard, 1991).

60 50 40 Very good 30 Accepted 20 Not accepted

Number of stems 10 0 Plot 1 Plot 2 Plot 3 Plot 4 Ishasha

Figure 5. 60 Acceptability of trees 50 in seven 20 x 20m Very good plots in Bwindi Forest 40 Accepted (Ishasha, n = 4 plots; 30 Not accepted Ngoto, n = 3 plots) 20

showing proportion Number of stems of trees with stems 10 rated suitable for 0 building purposes. Plot 1 Plot 2 Plot 3 Ngoto

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 35 People, park and plant use A.B. CUNNINGHAM BEAN STAKES in Afromontane forest (Figure 6), it was nowhere Although over 20 varieties of bean are rec- near the density of bean stakes in fields and this ognized in the DTC area, these are represented survey showed that a high proportion (35-58%) by two main growth forms: climbing beans and of bean stakes had already been cut. Although bush beans. Beans are one of the important sta- bean stakes must all have been cut from forest or ple foods in this region, and they are cultivated in forest remnants in the past, many farmers now all parishes. Major sites for cultivation of climb- plant Eucalyptus or Pennisetum purpureum (ele- ing beans are the Rubuguli, Rushaga, Nteko and phant grass) as a source of stakes. Nyamabale areas. Climbing beans are more pro- Demand for bean stakes is expected to ductive, easier to pick than bush beans and increase, and this cannot sustainably be met from reportedly softer and easier to cook, and the DTC supplies available within multiple-use areas. It is project is promoting the production of climbing recommended therefore, that in addition to pro- beans for these good reasons. moting the cultivation of climbing beans, the Bean stakes are essential for climbing bean DTC project should encourage existing initia- production, and cutting of bean stakes is an tives taken by farmers to cultivate trees and ele- important seasonal agricultural activity in May phant grass for bean stakes, and facilitate culti- and June. With bean stake density of c. 50,000 vation of other tree species (e.g. Sesbania ses- bean stakes per ha (Photo 20), and lasting only ban) for this purpose. 2-3 seasons, it is clear that huge quantities of Cutting of bean stakes is a labour-intensive bean stakes are needed every year. activity. Use of plants for bean stakes, although The understorey shrub Alchornea hirtella favouring certain species such as Alchornea hirtel- (ekizogwa) is one of the most favoured sources la (ekizogwa), is based more on selection for sites of bean stakes, re-growing readily and occurring with high densities of thin (1.5-4 cm diameter), in high density patches which are focal points for straight stems in order to maximize stakes cut per bean stake harvesting. Despite the high density unit time, than on species-specific selection. of this and other species suitable for bean stakes While a wide range of species (and life forms) is used (Table 11, page 38), favoured sites have a high density of potential bean stakes: either disturbed sites (e.g. scrub dominated by Acanthus arboreus (amatojo) or secondary forest with understorey, dominated by Alchornea hirtella (ekizogwa), where "bean stake density" has been further increased by coppicing, or Brillantaisia stands in river valleys. Similarly, although crop surplus such as cassava (Manihot utilissima) stems are used, cultivated stands of Pennisetum purpureum or Eucalyptus are favoured, as both give almost consistently straight stands of bean stakes of a suitable diameter, and harvesting is quicker and often closer than indigenous forest.

FUELWOOD Although the highest consumption of wood in the DTC area is for fuelwood (an estimated 140,000 m³ per year), one DTC survey (Kanongo, 1990) showed that most of this is from cultivated trees and only a small proportion (9 = 7.5%) from the forest. This is supported by field observation and data from Kanongo (1990) that only dry wood is used, as well as the low prices paid for fuelwood in the DTC area. In common with most rural areas in Africa, fuelwood provides the major source of household energy for cooking and heating in the DTC area (Table 12, page 38). Wood is also used for distilling waragi, and for baking bricks and clay pots. Although certain indigenous species are favoured (Table 13, page 38), crop surplus and Photo 20. Average bean stake density in fields is cultivated trees (e.g. black wattle Acacia mearn- 50,000 stakes/ha. sii, 73% and Eucalyptus, 58%) are the major

36 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM 200 Other 180 Plot 1: 296 trees (n=553 stems) Symphonia 160 Rytigynia 140 Ocotea 120 100 Psychotria 80 Alchornea

Number of stems 60 40 20 0 0-1 1-2 2-3 3-4 4-5 5-10 10-15 15-20 >20 Diameter at breast height [cm]

200 Other 180 Plot 2: 244 trees (n=271 stems) 160 Xymalos 140 Rytigynia 120 Ocotea 100 80 Psychotria 60

Number of stems Alchornea 40 20 0 0-1 1-2 2-3 3-4 4-5 5-10 10-15 15-20 >20 Diameter at breast height [cm]

200 Other 180 Plot 3: 197 trees (n=292 stems) 160 Bridelia 140 Alangium 120 100 ekigaragara 80 Brillantaisia 60 Number of stems 40 20 0 0-1 1-2 2-3 3-4 4-5 5-10 10-15 15-20 >20 Diameter at breast height [cm]

100 Other Plot 4: 145 trees (n=189 stems) Peddiea 80 Xymalos 60 Allophylus

40 Macaranga

Number of stems Psychotria 20

0 0-1 1-2 2-3 3-4 4-5 5-10 10-15 15-20 >20 Diameter at breast height [cm]

Figure 6. Data from four forest plots in the Rushaga area to assess density of stems used for bean stakes, showing the patchy distribution of this resource reflecting variation due to differences in topography, species composition and forest disturbance, from 553 stems in Alchornea (ekizogwa) dominated understorey (Plot 1) to 189 stems in secondary forest with an understorey dominated by Psychotria (Plot 4).

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 37 People, park and plant use A.B. CUNNINGHAM Table 11. Plant species used for bean stakes in the DTC area.

Family Plant species Rukiga name Life form Acanthaceae Acanthus arboreus (O) amatojo shrub Acanthaceae Brillantaisia sp. echunga shrub Asteraceae Vernonia sp. ekiheriheri shrub Euphorbiaceae Alchornea hirtella ekizogwa shrub (C) Euphorbiaceae Sapium ellipticum omushasha tree Euphorbiaceae Manihot utilissima (O **) [cassava] shrub Fabaceae Tephrosia vogelii (O) omukurukuru shrub Fabaceae Acacia mearnsii (O **) obulikoti tree Lauraceae Ocotea usambarensis omwiha tree (C) Myrtaceae Eucalyptus spp. (O **) uketusi tree (C) Poaceae Arundinaria alpina omugano bamboo Poaceae Pennisetum purpureum (O **) grass Rubiaceae Galiniera saxifraga omulanyoni shrub Rubiaceae Oxyanthus subpunctatus ? shrub Rubiaceae Psychotria schweinfurthii omutegashali shrub Rubiaceae Rytigynia kigeziensis nyakibazi shrub (C) Note: cultivated species (**), those mainly occurring outside of the forest (O) and those which coppice readily (C).

Table 12. Uses and attitudes to fuelwood use for cooking in the DTC area (data from Kanongo, 1990).

Main energy source Most suitable wood Wood source Why scarcity? (cooking) (cooking) Dry wood: 120 (100%) Black wattle: 87 (72.5%) From own land: 102 (85%) Little tree-planting: 87 (72.5%) Crop residue: 12 (10%) Eucalyptus: 69 (57.5%) Woodlots: 30 (25%) Over-population: 48 (40%) Charcoal: 15 (12.5%) Others (mainly Forest: 9 (7.5%) Restricted from forest: 21 (17.5%) Kerosine: 9 (7.5%) indigenous): 21 (17.5%) Bought (market): 5 (4.2%) Climate change: 14 (11.7%) Cupressus: 7 (5.8%) Charcoal used: 5 (4.2%) Other (land shortage Other sources: 3 (2.5%) destruction of trees): 9 (7.5%)

Table 13. Plant species favoured for fuelwood, fire making and chacoal in the DTC area.

Family Plant species Rukiga name Life form Use Ericacaeae Agauria salicifolia ekyigura tree charcoal Euphorbiaceae Bridelia micrantha omujimbu tree fuel Euphorbiaceae Sapium ellipticum omushasha tree fuel Fabaceae Acacia mearnsii (**) obulikoti tree fuel Fabaceae Albizia gummifera omushebeya tree fuel Fabaceae Newtonia buchananii omutoyo tree fuel Fabaceae Millettia dura omutate tree fuel Lauraceae Ocotea usambarensis omwiha tree fuel Myrsinaceae Maesa lanceolata omuhanga shrub fuel Myrtaceae Eucalyptus spp. (**) uketusi tree fuel Myrtaceae Syzygium guineense omungote tree fuel Proteaceae Faurea saligna omulengere tree charcoal Rosaceae Hagenia abyssinica omujesi tree charcoal Rubiaceae Galiniera saxifraga omulanyoni shrub fuel Tiliaceae Glyphaea brevis omusingati tree fire tinder Ulmaceae Trema orientalis omubengabakwe tree fuel Note: Cultivated exotic species are marked (**)

38 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM sources of fuelwood. According to one DTC sur- and species composition, resulting from canopy vey only 7.5% of 120 respondents obtained fuel- gap formation (Howard, 1991). In addition to dif- wood from indigenous forest. ferences caused by topography and soil type, log- The impact of dead-wood collection is low ging results in patchy distribution of both species compared to cutting of livewood for fuel, bean and size class categories of trees and shrubs used stakes or building materials. National Parks man- for bellows, building poles, beer boats or bean agement should consider three options in con- stakes. It also influences the availability of these nection with fuelwood: resources, either increasing the number of young (1) People living around the forest be allowed to saplings (bean stakes, building poles) or coloniz- collect dead wood, including dead trees, pos- ing species in disturbed sites (e.g. Polyscias sibly on a twice weekly basis. fulva, Maesopsis eminii), or decreasing stocks of (2) Collection of fallen dead wood be permitted, large hardwoods due to over-exploitation and but not the felling of dead trees, which pro- competing uses for timber. vide important nest sites for barbets and horn- bills, and feeding sites for woodpeckers. Competing uses (3) Attention be focussed on providing alterna- This situation is complicated further through tive sources of fuelwood outside the forest, competing uses for young trees or saplings of the recognizing that dead-wood use from multi- same species, particularly as Bwindi ple-use zones can only meet a fraction of Impenetrable Forest is the most heavily pitsawn of local needs. all principal forests in Uganda (Figure 7, page 40). To foresters, whose objective is hardwood Sustainable use of wood timber production, saplings of "reserved species" products (Table 14, page 41) represent regenerating timber trees. To people from local rural communi- Although the division in forestry terminology ties they also represent an important source of between "minor" and "major" forest products beer boats (>50 cm dbh), building material (5-15 reflects the bias of foresters managing forests for cm dbh) or bean poles (1.5-5 cm dbh), with high timber, the slow growth rates of forest hard- density wood favoured due to greater resistance woods still makes "major forest products" a use- to borer attack or fungal infection. ful category, particularly where the same species For this reason, despite their "reserved sta- is used for different purposes in different age tus", hardwood trees such as Newtonia classes. buchananii (omutoyo) and Prunus africana As discussed above (page 13), forest use dif- (omumba) are, and probably always have been, fers greatly from the use of low species diversity, favoured for beer boats, and Ocotea usambaren- highly productive vegetation types such as sis (omwiha) for building poles. Phragmites wetlands or Cymbopogon Cutting of bean stakes is selective for size thatch-grass, where harvesting is seasonal and rather than species, but if saplings of canopy easy to manage, with obvious clear cutting of species fit into this category, then they are cut suitable stands. (e.g. cutting of Strombosia scheffleri (omuhika) Bwindi Forest is at the opposite end of the and Ocotea usambarensis (omwiha) in secondary scale. Instead of the 1-year rotation applied in forest during this survey). Not only are these reed cutting (Cunningham, 1985), sustainable species a useful resource to local people, they harvesting of forest for timber aims at a rotation also represent the future forest canopy of the next of 100 years (Leggatt and Osmaston, 1961). Tree century. growth rates are slow, so unlike reeds which Trees with low density wood or small shrub- regrow from an underground rhizome within a by species are not generally subject to such com- year, the time between final harvesting and peting uses, however. Examples are: replacement of mature trees is seldom less than (1) Low density timber species used for beer 50 years, and often as much as 200 years (for oak boats but not for timber or building poles trees in Europe for example). (such as large Ficus species such as Ficus sur On the basis of growth measurements from (omulehe) and F. ovata (ekyitoma)); Afromontane tree species in southern Africa, (2) Secondary forest colonizers Polyscias fulva Stapleton (1955) considered that under natural (omungo) and occasionally Musanga leo- conditions, growth time to marketable maturity errerae (omutunda) used for blacksmiths' bel- for timber production from Podocarpus latifolius lows but not for other uses due to the soft wood. was 230 years, Ocotea bullata 220 years and (3) Bean poles cut from understorey shrubs such Olea laurifolia, 200 years. as Psychotria schweinfurthii (omutegashali), Logging, particularly when mechanized, but not for timber or beer boats, and rarely for affects mature forest, changing forest structure building poles.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 39 People, park and plant use A.B. CUNNINGHAM Rwenzori Mount Elgon 0% 235 km2 2 515 km 0% 0% Kibale 17% 432 km2

34% 0%

16% Mabira Semliki 2 66% 306 km 0% 67% 0% 2 100% 209 km 8% 0% 3% 27% 27% 31%

23% 9%

0% 12%

0% 30% 42%

Sango Bay 22% 21% 151 km2 6% 0% 4% 54%

7% 100% Budongo 53% 0% 2 12% 424 km 19%

10%

6% 19% 2% 4% 0% All principal forests 0% 2 71% 3842 km 69% Bugoma Itwara 0% 7% Bwindi Impenetrable 0% 8% 2 1% 300 km 2 2 68 km 12% 321 km 9% 0% 0% 10% 0% 0%

Kasyoha-Kitomi 328 km2 29%

81% 82% 61% Kalinzu-Maramagambo 553 km2

7% Mechanically harvested pre-1950 Severely encroached (> 30% canopy cover removed) 9% Mechanically harvested post-1950 Lightly encroached (5-30% canopy10% cover removed) 12% 2% Intensively pitsawn (> 20 % of trees exceeding 50cm dbh exploited) Essentially undisturbed 6% Selectively54% pitsawn (5-20% of trees exceeding 50cm dbh exploited)

Figure 7. The status of Uganda's principal forests, showing the degree of timber removal by various means, with Bwindi Impenetrable Forest most heavily affected by pit sawing (from Howard, 1991).

40 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Table 14. Competing use of forest tree and shrub species in successive age/size classes for pitsawn timber (>50 cm dbh), beer boats (>50 cm dbh), building poles (5 - 15 cm dbh) and bean stakes (1.5 - 5 cm dbh) in Bwindi forest.

Plant species Rukiga name Life form Pitsawn Beer boats Building Bean poles stakes Alangium chinense omukofe tree *** * * Albizia gummifera omushebeya tree (SF) * * Alchornea hirtella ekizogwa shrub * *** Arundinaria alpina omugano bamboo * *** Baphiopsis parviflora omunyashandu tree *** * Beilschmiedia ugandensis omuchoyo tree * Bridelia micrantha omujimbu tree (SF) *** * Carapa grandiflora omuruguya tree * * Chrysophyllum gorungosanum omushoyo tree (CP) *** * * Croton megalocarpus omuvune tree (SF) * * Cyathea manniana omungunza tree fern *** Dichaetanthera corymbosa ekinishwe tree * Drypetes gerrardii omushabarara tree *** * Drypetes ugandensis omushabarara tree *** * Entandrophragma excelsum omuyovi tree (CP) *** * * Faurea saligna omulengere tree (SF) *** * Ficalhoa laurifolia omuvumaga tree *** *** Ficus sur omulehe tree *** Ficus spp. (F. ovata, etc.) ekyitoma tree *** Galiniera saxifraga omulanyoni shrub * *** Harungana madagascariensis omunyananga tree (SF) *** * Macaranga kilimandscharica omurara tree * * Maesa lanceolata omuhanga shrub *** * Maesopsis eminii omuguruka tree (SF) *** *** Markhamia lutea omusavu tree (SF) * *** * Newtonia buchananii omutoyo tree (CP) *** *** * Ocotea usambarensis omwiha tree (CP) *** *** * Oxyanthus subpunctatus ? shrub *** Parinari excelsa omushamba tree *** Podocarpus latifolius omufu tree *** * *** Prunus africana omumba tree (CP) *** *** *** Psychotria schweinfurthii omutegashali shrub *** Sapium ellipticum omushasha tree (CP) * *** * * Strombosia scheffleri omuhika tree (CP) *** *** * Symphonia globulifera omusisi tree (CP) *** * * Syzygium guineense omugote tree *** Tabernaemontana holstii kinyamagozi tree (SF) *** Tabernaemontana sp. kinyamate tree (SF) *** Zanthoxylum gilletii omulemankobe tree (CP) *** * indet. omukarati tree *** Note: Canopy tree species are marked (CP) and secondary forest species (SF); *** high preference, ** acceptable, * used occasionally.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 41 People, park and plant use A.B. CUNNINGHAM Wood requirements and supplies Despite this, only 3 (2.6%) of the 114 trees In terms of supply and demand for fuelwood >30 cm dbh (or 4.6% of the trees >50 cm dbh) and building poles, the situation in the DTC area were suitable for beer boats. With the exception is similar to that described by Howard (1991) for of moist sites with a high density of Ficus trees, the Bwamba and Bajonjo counties in western pitsawn sites would be expected to have an even Uganda, north-east of Bwindi Impenetrable lower density of potential beer boats. Growth National Park. Population densities are high, and rates of most tree species favoured for beer boats land-holdings in intensively cultivated landscape are unknown for this area. It is likely, however, are similar - 0.2 ha per person in the DTC area that Ficus sur and Ficus ovata would be expect- (Kanongo, 1991); 0.26 and 0.19 ha per person ed to reach the minimum tree diameter (50 cm respectively in Bwamba and Bakonjo counties dbh) suitable for beer boats in 20-30 yr, and (Howard, 1991). Prunus africana or Newtonia buchananii in In Bwamba, Howard (1991) calculated that 40-50 yr, both of which greatly exceed the aver- the 121,600 people (17,000 households) would age life-span of most beer boats (9 yr). require about 151,000 m³ of fuelwood and With improved road networks and urbaniza- 4600m³ building poles every year (on the basis tion, marketing of banana beer, and therefore of a fuelwood consumption rate of 1.24 m³ per demand for beer boats can be expected to increase. person per yr and a building pole requirement of Building poles come from trees intermedi- 0.27 m³ per household per yr or 0.038 m³ per per- ate in size between those used for bean stakes son per yr). The DTC area is occupied by a sim- and beer boats. Resource assessments evaluating ilar number of people (over 99,000 people, trees within 20 x 20 m plots for suitability for 19,000 households). building purposes on the basis of durability, This situation has been worsened by the diameter and straightness, showed that there rapid clearing and burning of indigenous for- were fewer building poles per ha (Figure 8) than est for agriculture outside Bwindi resource users anticipated from visual assess- Impenetrable National Park. In 1954, approxi- ments made before quantitative work was done. mately 120 km² of forest remained in the DTC Visual assessments of resource availability made area within a 15 km radius outside the Bwindi with local people reported in Scott (1992) there- Forest boundary. By 1972 this had been fore need to be considered with caution. reduced to 42 km² of forest, and by 1983, less In the seven plots surveyed, only 20.8% than 20 km² remained (Butynski, 1984). Now, (8.3) poles per plot were classed as very good for apart from forest patches in less densely popu- building poles and 52.7% (21) poles per plot lated parishes such as Nteko, virtually nothing were accepted (see Figure 5, page 35). Although remains. additional plots are required, in the absence of Removal of forest outside of Bwindi other data this would indicate an average density Impenetrable National Park has reduced supplies of 207 very good poles per ha, or a total of 525 not only of fuelwood and building poles, as useable poles per ha. Howard (1991) points out, but also of natural Building pole cutting was not widespread, forest as a source of bean stakes, which although and despite the big demand for poles, high inten- relatively small in diameter are needed in vast sity harvesting was limited to a few patches in quantity, and beer boats, which although secondary forest. This is attributed to the wide- required in smaller numbers are large trees, some spread cultivation of Eucalyptus and black wat- of which are species already over-exploited by tle, and the less labour-intensive harvesting from pitsawyers (Newtonia, Prunus) or which are these cultivated tree species. "keystone species" favoured by frugivorous Counts in 20 x 20m plots in fields of climb- birds and primates (Ficus, Prunus). Through this ing beans in the Rubuguli and Nteko areas dur- process, the national park has become a focal ing this survey showed that there were point of harvesting pressure. 48,000-52,000 bean stakes per ha of climbing beans. By comparison, two forest plots in Wood resources availability Alchornea hirtella (ekizogwa) dominated forest For beer boats large trees are used (usually understorey favoured for bean stakes (and the >50 cm dbh), and most beer boats are greater highest density of plots sampled) (Figure 9, page than 40 cm in diameter (see Figure 4). Density of 44), contained 479-630 per 20 x 20m plot, or large trees suitable for beer boats is low in the approximately 12,000-16,000 bean stakes per ha, multiple-use area, judging from field observa- less than half as many as are required for climb- tion, and a 1 ha plot carried out in an ing bean cultivation. Allanblackia - Syzygium guineense dominated Bean stakes last 2-3 seasons, and beans are a forest site selected as representative of the forest. major crop in the Rubuguli, Nteko, Rushaga, This had a high number of beer boats per ha, due Nteko and Nyamabale areas. Total demand for to the flat terrain and absence of any pitsawing bean stakes in the DTC must represent millions activity there in the past. of saplings per year.

42 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM 40

35 Forest plot Ishasha, building poles

30 Plot 4 Plot 3 25 Plot 2 20 Plot 1

0 4- 6 6- 8 8- 10 10- 12 12- 14 14- 16 16- 18 18- 20 20- 22 22- 24 24- 26 26- 28 28- 30 >30 Diameter at breast height [cm]

35 Forest plot Ngoto, building poles

30 Plot 3

25 Plot 2

20 Plot 1

Number of stems15 Number of stems

0 4- 6 6- 8 8- 10 10- 12 12- 14 14- 16 16- 18 18- 20 20- 22 22- 24 24- 26 26- 28 28- 30 >30 Diameter at breast height [cm]

Figure 8. Combined number of tree stems in 2 cm size-class intervals for four and three 20 x 20 m plots (combined total of 0.28 ha) in secondary forest in the Ishasha gorge and Ngoto swamp areas respectively, showing the number of stems in the size-class range preferred for building purposes (5-15 cm dbh).

Two additional points are significant here. are inadequate to meet either the existing or the First, distribution of Alchornea hirtella dominated future demand (Table 15, page 45). The need for stands is patchy, and density of bean poles in the building poles was also the main reason for tree surrounding areas is significantly lower (3000-6000 planting in Bwamba, but it was considered to bean stakes per ha). Second, the Alchornea hirtella provide only 328 m³ of the total annual demand patches are already heavily utilized, with 58% (429) of the 151,000 m³ of fuelwood, 4600 m³ of build- of stems cut in the plot with the highest density of ing poles and annual increase in demand for tim- bean stakes, and 35% (186) of stems cut in the adja- ber of 5280 m³. It is likely that a similar situation cent, lower density plot . exists in the DTC project area. Elephant grass (Pennisetum purpureum) and Cultivation of wood resources trees (particularly Eucalyptus) are also planted Shortages of fuelwood, building poles and for bean stakes, while Ficus cuttings are planted bean stakes are being experienced in the DTC for beer boats (recorded in the Ngoto area) and area, and a shortage of large trees for beer boats canoes (Lake Bunyonyi). In some areas, even can be expected in the future. Reasons for wood hardwood timber trees have been planted, with scarcity, and solutions to the problem are recog- Entandrophragma (omuyovi) reaching a dbh of nized by local people. 90 cm within 40 years (Photos 21 and 22, page Trees (particularly Eucalyptus) are planted 45). Such local initiatives need to be recognized in the DTC area, mainly for building poles, but and encouraged.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 43 People, park and plant use A.B. CUNNINGHAM 600

Cut 500 Plot 1: 296 trees (n=724 stems) Suitable 400 Not suitable 300

200 Number of stems

100

0 Alchornea Psychotria Ocotea Rytigynia Symphonia Other Species

600 Cut Plot 2: 244 trees (n=419 stems) 500 Suitable Not suitable 400

300

Number of stems 200

100

0 Alchornea Psychotria Ocotea Rytigynia Xylmalos Other Species

200

180 Plot 3: 197 trees (n=292 stems) 160 Suitable

140 Not suitable 120

100

80 Number of stems 60

0 Brillantaisia ekigaragara Alangium Macaranga Bridelia Other Species

200 180 Cut 160 Suitable Plot 4: 145 trees (n=197 stems) 140 Not suitable 120

100

60 Number of stems 40

0 Allophylus Psychotria Peddiea Macaranga Xylmalos Other Species

Figure 9. Data from four 20 x 20m plots in secondary forest in the Rushaga area, showing selectivity and high proportion of stems cut for bean stakes in Alchornea hirtella (ekizogwa) dominated understorey (Plots 1 and 2), and low degree of harvesting in less favoured sites (Plots 3 and 4).

44 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Table 15. Attitudes and approaches to tree planting in the DTC area (from data in Kanongo, 1990).

Reasons for Species Source of Species preferred Planting site planting planted seedlings for building (n = 115) (n = 120) (n = 120) (n = 120) (n = 120)

Building (73) Eucalyptus (92) Own seedlings (89) Eucalyptus (106) Uncultivated land (98) Fuelwood (28) Acacia mearnsii (46) Forest dept. (33) Acacia mearnsii (59) House compound (49) Sale (5) Cupressus (43) Community nursery (13) Cupressus (14) Boundaries (47) Other (9) Markhamia (5) Other (8) Markhamia (10) Among crops (34) Sesbania (5) Other (9) In pasture (29) Other (2) Fallow lands (19) Other (roadsides) (4)

Photo 21. Mr. K. Byarugaba, second generation Entandro- Photo 22. Entandrophragma excelsum (omuyovi) planted in phragma (omuyovi) planter with one of six 1 year old trees at 1950 at the same homestead by the late father of Mr. K. his homestead in the Ngoto area . Byarugaba.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 45 People, park and plant use A.B. CUNNINGHAM Box 6. Recommendations for wood use

∗ Tree planting outside forest reserves should be undertaken as an urgent priority, as recommended by Butynski (1984), Hamilton (1984), Struhsaker (1987) and Howard (1991). The existing single line of Cupressus that marks the boundary of Bwindi Impenetrable National Park needs to be paralleled by planting a strip of fast-growing exotic or indigenous trees useful both for building purposes and fuelwood. ∗ Initiatives already taken by local farmers in tree cultivation need to be supported through greater supply of seedlings and the establishment of nurseries. DTC staff are already involved with an agroforestry programme. In addition to work underway, critical areas with high population densities, little woody cover and steep slopes need to be identified and become a priority, as short-term rotation crop production on steep slopes is unlikely to be sustainable due to high soil losses. ∗ Cultivation of bamboo and elephant grass (Pennisetum purpureum) should be encouraged as a soil conservation measure on bunds and in water-courses, as well as to provide building material and bean stakes. ∗ Subject to further investigation, wood carvers in parishes within the DTC area could be registered and involved in a rotational management system for carving of household utensils (e.g. Rapanea melanophloeos for carved walking sticks). ∗ Felling of Polyscias fulva trees for making blacksmiths' bellows should be permitted within multiple-use zones. ∗ CARE/DTC-Uganda also need to promote the cultivation of trees suitable for grinding mortars and carving (e.g. Markhamia lutea, Rapanea melanophloeos), and investigate the viability of introducing appropriate technology mills for millet and groundnuts as an alternative to hardwood mortars. ∗ No felling of trees for beer boats, building poles or bean stakes should be allowed in multiple-use areas. ∗ Attention should be focused on providing alternative sources of fuelwood outside the forest, recognizing that the use of dead-wood from multiple-use zones can only meet a fraction of local needs, and that staff capacity for multiple-use management is limited. ∗ Involve community leaders from Resistance Council (RC)-1 level (village level) upwards in tree planting, inducing people to plant a target number of trees per year. ∗ Brick-makers, potters and waragi makers should be encouraged to plant a greater number of trees to balance the higher fuel consumption rates of these activities. ∗ Forest destroyed by arson should be closed to any utilization for fuelwood or building poles. Both fallen and standing trees play an important role in preventing soil loss on steep slopes and also in trap- ping seeds, as well as providing perches for birds dispersing seed into disturbed sites. The use of wood for fuel could also provide an incentive to burn the forest more regularly if wood shortages increase, instead of planting trees as an alternative supply. ∗ Recommendations for additions of land adjacent to the Bwindi Impenetrable National Park should be followed through as soon as possible, even if zoned as part of multiple-use areas. Critical sites are the Kitahurira corridor, which needs to be widened through becoming the focus of tree planting activity and Ngoto Swamp, where a strip of land at least 50 m wide around the swamp needs to be negotiated for tree planting. All uses of plants from the Cyperus papyrus (efundjo) swamp should be allowed to continue. Ficus trees should be planted around the swamp from cuttings as a source of beer boats in the future. ∗ Although exotic tree species are commonly planted and are extremely useful, some indigenous species also have potential for fulfilling local needs, and are worth considering. Maesopsis eminii, Harungana madagascariensis, Maesa lanceolata, Dodonaea viscosa, Trema orientalis, Millettia dura and M. lutea for example, all grow well in disturbed sites, have many uses (such as building) and are suited to local conditions. The exceptional coppicing ability of Alchornea hirtella makes it a good candidate for managed coppice rotations in private woodlots.

46 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM New natural products with commercial potential

The potential of tropical forests as a source of returns than from the two species mentioned potential new medicines, other natural products above. or wild relatives of crop plants is a common rea- Bwindi Forest contains representatives of son given for conservation. Bwindi Impenetrable many plant families, and includes some regional- forest is no exception, and a least one commer- ly endemic species of interest as potential cial company is interested in the forest as a sources of new drugs. For example, the families source of new antibiotics, and there is no reason Rubiaceae and Apocynaceae, both contain why this should not proceed, provided certain species which are rich in alkaloids. conditions are met. Edible fungi, for example Lentinus prolifer Unfortunately, although forests are seen as (Photo 9, page17), may also have potential for a source of new natural products with commer- cultivation, perhaps grown on crop surplus as cial value, little thought is given as to how practised in Asia (FAO, 1983b). equitable partnerships could be developed, so A Pleistocene refugium with high biological that some of the profits arising will return to and topographic diversity, Bwindi Impenetrable the region of origin of that product - whether National Park is a rich source of soil micro- this be an oil, resin, an organic chemical struc- organisms such as Actinomycetes, which are a ture used as a basis for a new drug, or genetic potential source of new antibiotics. material for developing a disease-resistant crop Western Uganda, including Bwindi variety. Impenetrable Forest, is rich in plants with horti- Consideration also needs to be given to the cultural potential as well, including regional impact that extractive harvesting of forest prod- endemics in the genus Impatiens. ucts may have. Both issues need to be taken Several wild relatives of crop plants, for seriously by both the development and conser- example the cowpea relative Vigna luteolus, vation partnership surrounding Bwindi many members of the Cucurbitaceae such as Impenetrable National Park. Coccinia mildbraedii, and wild coffee (Coffea) Allanblackia kimbiliensis seeds are a may be useful, in addition to plants in which source of a fat which may have potential for there is international interest for their value in use in cosmetics and cosmetic soaps. In the breeding programmes for forage plants (e.g. Usambara mountains, Tanzania, the seeds of Trifolium and Aeschynomene), particularly as the closely related A. stuhlmannii, which con- this is a high-altitude refugium. tain 51% edible fat, are harvested and sold by Farmers in the DTC area cultivate, and have local people to GAPEX (General Agricultural a rich knowledge of crops, either developed in Products Export Company) for the extraction Africa (e.g. finger millet, Eleusine coracana of "a firm, white and somewhat brittle fat" and Sorghum), or introduced from elsewhere (FAO, 1983a). such as bananas and sweet potatoes. Carapa grandiflora (omuruguya) seeds, A probable reason for the greater variety of which are used on a small scale by local people land-races grown here compared to many other for extraction of an oil used for cosmetic purpos- areas is due to the great altitudinal range of farm- es as a "Vaseline" substitute, may also have ers’ fields. Local farmers have a greater knowl- potential in the cosmetics industry. edge of these local land-races than most formal- Myrianthus holstii (omufe) (Photo 8, page 17) ly trained plant breeders, recognizing at least 20 may have value for genetic enhancement as a new bean varieties, 16 banana varieties, 9 sweet pota- crop plant, just as the closely related M. arboreus to varieties and a number of groundnut, finger has been proposed for planting and fruit production millet and cassava land-races. Steps need to be (FAO, 1983a). Although M. holstii fruits are sold taken in the DTC area to avoid the genetic ero- commercially on a small scale in local markets, sion of these land-races that may take place if development of this plant as a new fruit crop is a they are replaced through the introduction of new long-term project with less immediate potential high-yielding varieties.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 47 People, park and plant use A.B. CUNNINGHAM The future sufficiently, then selective controlled removal of large trees, either by pitsawyers or for beer boats, In the foreseeable future, demand for forest could again be considered as a means of creating resources is going to increase. Apart from meet- canopy gaps, disturbance and diversity. This may ing future wood requirements, a major tree plant- not take place for 40-50 years, but needs to be ing effort is required even to make up the exist- borne in mind. ing backlog in wood requirements. Howard If urbanization patterns in Uganda follow (1991), Struhsaker (1987) and others have all those experienced elsewhere in Africa, then it is emphasized the need for family planning if any also likely that the commercial trade in traditional conservation programmes are to succeed in the medicines will develop. This needs to be moni- long-term. tored, so that pro-active management through pro- If forest protection enables recovery from vision of cultivated supplies of favoured, slow- the timber over-exploitation of the past, then growing species can be implemented. Elephant management recommendations will have to be numbers and forest destruction by elephant need reviewed. Forests are dynamic systems, and in to be monitored and, if necessary, steps taken to the future, if recovery of mature forest occurs avoid this threat to the forest habitat.

Box 7. Recommendations for future research and monitoring

The following suggestions are in addition to those already made in the text: ∗ An updated vegetation map is needed for Bwindi Forest, based on the relatively recent (1990) aerial photographs, and identifying high diversity sites. Attention needs to be directed towards the Ivi and Ihihizo River valleys, which have been sparsely collected in previous surveys compared to the Ishasha Gorge, yet on the basis of this short study, are considered to be equally important sites, with a high diversity of plant species. ∗ Ecological work is required on canopy gap densities in Bwindi Forest, based on terrain evaluation as well as on vegetation type/age, since canopy gap densities differ with slope. ∗ Work is needed on growth rates, biomass production and mapping of bamboo. ∗ Research should be focussed on the biology of Parinari excelsa var. holstii and Newtonia buchananii, which, apart from being useful plant species harvested for timber, are key species for supporting a high diversity of epiphytes (Orchidaceae, Cactaceae (Rhipsalis baccifera), and various pteridophytes). Why is there such poor recruitment of Parinari excelsa, and what implications does this have for epiphyte diversity? ∗ Assessment of elephant damage to forest is needed, as a result of their restricted range, rather than wider seasonal movements out of the forest. What role do the elephants play in creating and maintaining canopy gaps? What is the carrying capacity of the forest for elephant, in terms of a balance between disturbance, diversity and mature forest? ∗ Ecological work is needed on the biology of climbers and their biomass production and spacing in relation to canopy gap dynamics. This work would have wide applicability, as climbers are important both to people and to primates (e.g. Urera hypselodendron as a food item for gorillas). ∗ Disturbed sites of known age, such as pitsawing sites and old mining camps, need to be relocated and studied in terms of species composition and size classes (growth rates) of trees. This provides useful information for forest ecology and reforestation ("restoration ecology"), as well as for maintaining a certain level of disturbance to create diversity and habitat for canopy gap and secondary forest specialists (climbers, and trees such as Maesopsis eminii). ∗ Encourage partnership between Mbarara University researchers, traditional healers and Ugandan natural products chemists such as Mr A. B. Kakooko (Kampala), to evaluate herbal remedies, whether for human use or for livestock, and micro-organisms that may have value as sources of new antibiotics. ∗ Additional work is required that involves resource users and traditional experts in resource assessments and monitoring (see Wild and Mutebi, 1996, Working paper 5). ∗ Local experts, particularly the Batwa, need to be more widely involved as research partners. They have a tremendous store of knowledge that can add greatly to an understanding of forest functioning and ecology, whether on specialist pollinators such as Trigonid bees, canopy gap formation and regeneration, seed dispersal. Employment of local people as "parataxonomists", after short training courses, is also desirable.

48 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM References

Adu-Tutu, M.; Y. Afful, K. Asante-Appiah, D. Cunningham, A. B. 1991. Development of a conserva- Lieberman, J. B. Hall and M. Elvin-Lweis. 1979. tion policy on commercially exploited medicinal Chewing stick usage in southern Ghana. plants: a case study from southern Africa. In: Economic Botany 33: 320-328. Conservation of Medicinal Plants (eds: O. Alder, D. and Synnott, T. J. 1992. Permanent sample Akerele, V. Heywood and H. Synge), pp. plot techniques for mixed tropical forest. Oxford 337-358. Cambridge University Press, Forestry Institute, Oxford. Cambridge. Cunningham, A. B. 1993. African medicinal plants: Anderson, S. and F. Staugard. 1986. Traditional mid- setting priorities at the interface between conser- wives: traditional medicine in Botswana. vation and primary health care. People and Plants Ipelegeng publishers, Gaborone. Working paper 1. Paris. UNESCO. Anon. 1991. CARE: An overview. Annual meeting, Cunningham, A. B. and S. J. Milton. 1987. Effects of September 11, 1991. CARE-International, New basket weaving industry on Mokola palm and dye York. plants in north-western Botswana. Economic Anon. 1992. Population figures for DTC project. Botany 41, 386-402. Unpublished report, DTC project, Ikumba, Cunningham, A. B. and Mbenkum, F. T. 1993. Uganda. Sustainability of harvesting Prunus africana bark Ashford, R. W., G. D. F. Reid and T. M. Butynski. in Cameroon: A medicinal plant in international 1990. The intestinal faunas of man and mountain trade. People and Plants Working paper 2. Paris. gorillas in a shared habitat. Annals of Tropical UNESCO. Medicine and Parasitology 84: 337-340. Deacon, H. J. 1983. The peopling of the fynbos. In: Balasubramanian, V. and A. Egli. 1986. The role of Fynbos paleoecology: a preliminary synthesis. agroforestry in the farming systems in Rwanda (eds: H.J. Deacon, Q.B. Hendey and J.J.N. with special reference to the Lamprechts), pp. 183-204. South African Bugesera-Gisaka-Migongo (BGM) region. National Programmes Report 75, Pretoria, South Agroforestry Systems 4: 271-289. Africa. Bulatao, R. A., Bos, E., Stephens, P. W. and Vu, M.T. Desawadi, P. 1991. The conservation of medicinal 1990. World population projections 1989-1990 plants used in Primary Health Care in Thailand. edition. Short- and long-term estimates. Johns In: Conservation of Medicinal Plants (eds: O. Hopkins University Press, London. Akerele, V. Heywood and H. Synge), pp. 253-258. Cambridge University Press, Butynski, T. M. 1984. Ecological survey of the Cambridge. Impenetrable (Bwindi) forest, Uganda, and rec- Erkkila, A. and H. Siiskonen. 1992. Forestry in ommendations for its conservation and manage- Namibia 1850-1990. Silva Carelica 20. ment. Unpublished report to the Ugandan University of Joensuu. Government. FAO. 1981. Food loss prevention in perishable crops. CARE-International. 1992. Scope of work: ethnob- FAO Agricultural Services Bulletin 43, FAO, otany and sustainable use of forest plant Rome. resources. CARE-International, Kampala, Uganda. FAO. 1983a. Food and fruit-bearing forest species. 1. Examples from eastern Africa. FAO Foretsry Conelly, W. T. 1985. Copal and rattan collecting in the Paper 44/1. FAO, Rome. Philippines. Economic Botany 39: 39-46. FAO. 1983b. Growing mushrooms : Oyster Cunningham, A. B. 1985. The resource value of Mushroom, Jew's Ear Mushroom, Straw indigenous plants to rural people in a low agricul- Mushroom. FAO Regional Office for Asia and tural potential area. Ph.D thesis, University of the Pacific, Bangkok. Cape Town. Forbes, L. P. F. 1991. Cambridge Uganda Forest Cunningham, A. B. 1987. Commercial craftwork: bal- Island expedition. Cambridge University. ancing out human needs and resources. South Forest Department. 1964. Annual report of the Forest African Journal of Botany 53: 259-266. Department, 1963/64, Uganda Government. Cunningham, A. B. 1990. People and medicines: the Government Printer, Entebbe. exploitation and conservation of traditional Zulu Gentry, H. S., A. J. Verbiscar and T. F. Branigan. medicinal plants. Mitteilungen aus dem Institut 1987. Red Squill (Urginea maritima, Liliaceae). fur Allgemeine Botanik, Hamburg. 23b, 979-990. Economic Botany 41 (2): 267-282.

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 49 People, park and plant use A.B. CUNNINGHAM Glover, P. E. and E. C. Trump. 1970. An ecological Martin, R. B. 1986. Communal areas management survey of the Narok district of Kenya Masailand. programme for indigenous resources (CAMP- Pt 2. Vegetation. Nairobi, Kenya National Parks. FIRE). CAMPFIRE Working Document No. 157 pp. (cited by White, 1983). 1/86, Department of National Parks and Wildlife Good, C. M. 1987. Ethnomedical systems in Africa: Management, Zimbabwe. patterns of traditional medicine in rural and urban McNeely, J.A. 1988. Economics and biological diver- Kenya. The Guildford Press, Kenya. sity: developing and using economic incentives to Groube, L. 1989. The taming of the rain forests: a conserve biological resources. IUCN, Gland. model for Late Pleistocene forest exploitation in Muir, D. P. 1991. Indigenous forest utilization in New Guinea. pp. 292-304 in: D. R. Harris and G. KwaZulu: a case study of the Hlatikulu forest, C. Hillman (eds) Foraging and farming: the evo- Maputaland. MSc thesis, University of Natal, lution of plant exploitation. Unwin Hyman, Pietermaritzburg. London. Mwesigye, V. M. 1991. Beekeeping in the Hamilton, A. C. 1981. The quaternary history of Impenetrable forest. Unpublished field-training African forests: its relevance to conservation. report, Institute for Tropical Forest Conservation, African Journal of Ecology 19: 1-6. Ruhija. Hamilton, A. C. 1991. A field guide to Ugandan forest Noble, I. R. and Slatyer, R. O. 1980. The use of vital trees. Makerere University, Kampala. attributes to predict successional changes in plant Hamilton, A. C., Taylor, D. and J. C. Vogel. 1986. communities subject to recurrent disturbances. Early forest clearance and environmental degra- Vegetatio 43, 5-21. dation in south-west Uganda. Nature 320: Osmaston, F. C. 1968. The management of forests. 164-167. George Allen & Unwin, Ltd. London. Hart, T. B. and Hart, J. A. 1986. The ecological basis Payapyipapong, P., Traitongyoo, T. and Dobias, R. J. of hunter-gatherer subsistence in African rain- 1988. Integrating park conservation and rural forests: the Mbuti of eastern Zaïre. Human development in Thailand. Unpublished paper pre- Ecology 14 (1): 29-53. sented at the IUCN meeting, Costa Rica. Howard, P. C. 1991. Nature Conservation in Uganda's Peters, C. M. 1990. Plenty of fruit but no free lunch. Tropical Forest Reserves. IUCN, Gland, Garden 14(6): 8-13 (published by the New York Switzerland. Botanical Garden). Huntley, B. J. 1988. Conserving and monitoring biot- Rifai, M. A. and Kartawinata, K. 1991. Germplasm ic diversity: some African examples. pp. 249-260. and the Conservation of Indonesian medicinal in: E. O. Wilson (ed) Biodiversity. National plants. pp. 281-292, in: Heywood, V., Synge, H. Academy Press, Washington. & Akerele, O. (eds). Conservation of medicinal IUCN. 1991. Biodiversity in sub-Saharan Africa and plants. Cambridge University Press. its islands. Occasional Papers of the IUCN Rutherford, M. C. 1978. Primary production ecology in Species Survival Commission, no.6. IUCN, southern Africa. In: Biogeography and Ecology of Switzerland. southern Africa. (ed: M. J. A. Werger), Kakuru, W. 1990. The potential of the Impenetrable pp. 621-659. W. Junk, The Hague. (Bwindi) forest for medicinal plants. BSc Rutherford, M. C. and Westfall, R. H. 1986. Biomes of (Forestry) project, Makerere University. southern Africa: an objective categorization. Kanongo, M. 1990. Draft report of baseline data, Memoirs of the Botanical Survey of South Africa, report for Development through Conservation 54. Botanical Research Institute, South Africa. project. Phase 1. Dec. 1989-Feb. 1990. Schönau, A. P. G. 1970. Timber/bark weight ratios for Development through Conservation project, black wattle. Wattle Research Institute Report Ikumba, Uganda. 1969-1970. October 1970, WRI, University of Kigomo, B. N. 1988. Bamboo resource in the East Natal. African region. pp. 22-28, in: Bamboos Current Scott, P. 1992. Fringe benefits: National Park bound- Research, Proceedings of the International aries as areas for local community utlization: the Bamboo Workshop, Nov. 14-18, 1988. case of the Impenetrable (Bwindi) forest. Kingdon, J. 1990. Island Africa: the evolution of Unpublished M.Sc. thesis. Africa's rare animals and plants. Collins, London. Seydack, A. H. W., Grewar, S. G. and van Dijk, D. Kingston, B. 1967. Working plan for Mgahinga 1982. Proposals for a management system for Central Forest Reserve, Kigezi district, Uganda. indigenous forests of the southern Cape and For the period 1967-1977. Uganda Forest Tsitsikamma. Department of Environment Department. Affairs, Southern Cape Region. Leggatt, G. J. and H. A. Osmaston. 1961. Working Shackleton, S. E. 1990. Socio-economic importance of plan for the Impenetrable Central Forest Reserve, Cymbopogon validus in Mkambati Game Kigezi district, western province, Uganda. Period Reserve, Transkei. South African Journal of 1 July 1961-June 1971. Botany 56: 675-682. Malpas, R. C. and M. M. Infield. unpublished. The Stapleton, C. C. 1955. The cultivation of indigenous Mgahinga forest and gorilla game reserves, trees and shrubs. Journal of the South African Uganda. WWF/IUCN. Unpublished report to the Forestry Association 26: 11-17. Ministries of Tourism and Wildlife, and Struhsaker, T. T. 1987. Forestry issues and conservation Agriculture anf Forestry, Uganda. in Uganda. Biological Conservation 39: 209-234.

50 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Takeda, J. 1990. The dietary repertory of the Ngandu Personal Communications: people of the tropical rainforest: an ecological and Baranga, J. 1992. Institute for Tropical Forest anthropological study of the subsistence activities Conservation, Bwindi Impenetrable National and food procurement technology of a Park. slash-and-burn agriculturalist in the Zaïre basin. African Study Monographs, Suppl. 11: 1-75. Sucker, K. J. 1992. Mgahinga Gorilla National Park Project, PO Box 723, Kabale, Uganda. Tanno, T. 1981. Plant utilization of the Mbuti pygmies - with special reference to their material culture Wild, R. 1992. Development through Conservation and use of wild vegetable foods. African Study project, CARE- Uganda, PO Box 7280, Kampala, Monographs 1: 1-53. Uganda. Taylor, D. M. 1990. Late quaternary pollen records from two Ugandan mires: evidence for environ- meltal change in the Rukiga highlands of south-west Uganda. Paleogeography, Paleoclimatology, Paleoecology 80: 283-300. Taylor, R. H. and Cunningham, A. B. 1983. The conservation of wetlands. Journal of the Limnolo- gical Society of Southern Africa 9: 141-145. Van Noten, F. 1977. Excavation at Matupi cave. Antiquity 51: 35-40. Van Noten, F. 1979. The early Iron Age in the Interlacustrine Region: the diffusion of iron technology. Azania 14: 61-80. Vasquez, R. and Gentry, A. H. 1989. Use and misuse of forest harvested fruits in the Iquitos area. Conservation Biology 3, 350-361. Voorhoeve, A. M. et al. 1982. Machakos project studies: agents affecting health of mother and child in a rural area of Kenya -- antenatal and delivery care. Tropical and Geographical Medicine 34: 91-101 (cited in Anderson and Staugard, 1986). Were, J. M. 1988. Arundinaria alpina in Kenya. Bamboos Current Research, Proceedings of the International Bamboo Workshop, Nov. 14-18, 1988. White, F. 1983. The vegetation of Africa. UNESCO, Paris. White, R. G. 1969. Blacksmiths of Kigezi. Uganda Journal 33: 65-73. Wild, R. and Mutebi, J. 1996. Conservation through Community use of plant resources. Establishing collaborative management at Bwindi Impenetrabel and Mgahinga Gorilla National Parks, Uganda. People and Plants working paper 5. UNESCO. Paris. Wimbush, S. H. 1945. The African alpine bamboo. Empire Forestry Journal 24: 33-39 (cited by Kingston, 1967). Wondergem, P., K. A. Senah and E. K. Glover. 1989. Herbal drugs in primary health care. Ghana: an assessment of the relevance of herbal drugs in PHD and some suggestions for strengthening PHD. Royal Tropical Institute, Amsterdam. World Bank (1986). Joint UNDP/World Bank Energy Sector Managment Assistance Programme. Activity Completion Report No. 053/86: Fuelwood/Forestry Project feasibility report. Washington, DC (cited by Howard, 1991).

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 51 People, park and plant use A.B. CUNNINGHAM Acknowledgements

This report would not have been possible without the invitation from Dr. Jan Kalina to visit Uganda, the expertise of local plant users and the support of CARE-International. CARE are also thanked for permission to publish the report in this Working paper series. Companionship and assistance in the field from Ben Otim, Alfred Tsekeli, Erastus Mehanda, Aurelia Mehanda, Robert Baragira, Rob Wild, Jacob Bandusya, Rose Badaza, Jovita Tumusime and Michelle Cunningham is greatly appreciated. This survey represents the combination of formal botanical approaches with traditional knowledge and local expertise. In particular, the exceptional ability, acute field observation and humour of Jacob Bandusya made a great contri- bution. I must also thank Kazoka Ruwajiri, James Tumutegyereize, Bernado Ruwenzije, Runago Zimbehere, Benon Twine, Monday Mafurira and other local experts in this regard. I also benefitted from detailed discussions with specialist users, particularly the fledgling Nyamabale Beekeepers Association and beekeepers in the Ruhija area, and craftworkers in the Ngoto area. Rob Wild, Lorna Slade and Dr. Tom Butynski are thanked for their hospitality and for the use of the facilities at Ruhija and Ikumba. Dr. Jonanthan Baranga and Joseph Serugo, John Miskell and Cindy Carlson are also thanked for their assistance whilst I was in Uganda, and Dr. Bernard Verdcourt, Dr. Diane Bridson, and Dr. Brian Schrire are thanked for help with identifying plant specimens at Kew, and Dr. David Pegler for identifying the Lentinus specimen. Tony Katende, Elizabeth Ogwal and Dr. Z. R. Bukenya kindly assisted with identification of specimens at Makerere University herbarium. Dr. Alan Hamilton is thanked for his encouragement, and for copies of several references, including his very useful book on forest trees.

52 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Acronyms

BINP Bwindi Impenetrable National Park CAMPFIRE Communal Areas Management Programme for Indigenous Resources CARE International relief and development agency that helps people in developing nations through humanitarian programs CEA Community extension agent dbh Diameter at breast height DTC Development Through Conservation Project FAO Food and Agriculture Organization GAPEX General Agricultural Products Export Company GDP Gross Domestic Product ICRAF International Centre for Research in Agroforesty IFCP Impenetrable Forest Conservation Project ITFC Institute of Tropical Forest Conservation IUCN The World Conservation Union MUST Mbarara University of Science and Technology MVP Minimum Viable Population PHC Primary health care RC I-V Resistance Council I to V SNR Strict Nature Reserve TBA Traditional Birth Attendent TMP Traditional Medicinal Practitioner UNEP United Nations Environment Programme UNESCO United Nations Educational, Scientific and Cultural Organization UNICEF United Nations Children’s Fund UNP Uganda National Parks UTGC Uganda Tea Growers Corporation WHO World Health Organization WWF World Wide Fund for Nature

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 53 People, park and plant use A.B. CUNNINGHAM Appendix 1

Preliminary list of plants recorded during this survey, noting uses, local names and voucher numbers (AC = A. Cunningham, RB = R. Badaza, AT = A. Tsekeli). A. Tsekeli's collection numbers follows those used by A. Cunningham. Although incomplete at this stage, with additional botanical names expected from herbaria, this Appendix is provided as a background to the report, which concentrated mainly on key species. It is expected that subsequent ethnobotanical work will be done in western Uganda to develop this list further and validate/correct Rukiga names. At present therefore, this Appendix should be treated as a preliminary draft. A = Veterinary medicines, B = Building, BB = Beer boats, Be = Bee nectar and nesting sites, BH = Bee hives, Bl = Blacksmiths, Br=Brooms, BS = Bean stakes, Bw = Bows and arrows, C = Charcoal, Cm = Combs, Co = Covers for food, pots, etc., Cu = Cups, DB = Straw for drinking beer, DC = Dental Care, E = Edible, Fa = Fats, Fe = Fencing, FW = Firewood, G = Granaries, H = Herbalist, Gu = Gums, HD = Hunting dog medicines, HH = Hoe handles, Mo = Mortars, MW = Mid wives, P = Pipes, Pe = Pestles, Po = poison for arrows, R = Rakes, SH = Spear Handles, T = Twine and rope, Tr = to trap moles, Ty = Trays, W = weaving, WS = Walking sticks.

BOTANICAL NAME FAMILY RUKIGA NAME VOUCHER NO. USE Abrus canescens Fabaceae ekyanyamashozi 3223 AC W Acacia mearnsii Fabaceae bulikoti 84 RB BS B T C FW Acalypha sp. Euphorbiaceae enzibirabusha 31 RB MW Acanthus arboreus Acanthaceae amatajo 3149, 4028 AC BS Adenia sp. Passifloraceae echururu 4136 AT H Aeschynomene sp. Fabaceae 4016 AC BS FW Aframomum sp. Zingiberaceae amatehe 3000, 3008 AC E Agauria salicifolia Ericaceae ekigura/ekigwa 3084, 3226 AC MW C Agelaea pentagyna Connaraceae 3035 AC Aidia micrantha Rubiaceae orube 3134 AC SH P Alangium chinense Alangiaceae omukofe BH Be Pe B BS Albizia gummifera Fabaceae omushebeya 4 RB MW B FW BH BS Alchornea hirtella Euphorbiaceae ekizogwa 3109, 3032, 4030 AC BS B Allanblackia kimbiliensis Clusiaceae omutaka/omugus 3098, 3178 AC SH B Amaranthus sp. Amaranthaceae dodo 108 RB E Anchomanes difformis Araceae Anthocleista zambesiaca Loganiaceae omuyamgabe 3052 AC H Arundinaria alpina Poaceae omuganu 3258 AC B W BH BS Asparagus sp. Liliaceae olugawampinga 127 RB MW Ataenidia conferta Marantaceae ekitatara 3200 AC W Baphiopsis parviflora Fabaceae omunyashandu B BS Basella alba Basellaceae enderema 5, 70 RB MW E A Begonia sp. Begoniaceae omwamira 3025 AC H Beilschmiedia ugandensis Lauraceae omuchoyo 3024 AC B FW Be Bersama abyssinica Melianthaceae omukaka, nyakibazi 3143 AC MW H Bidens pilosa Asteraceae enybarashana 4087 AC, 81, 24 RB MW Biophytum abyssinicum Oxalidaceae irango 4066 AC MW Bridelia micrantha Euphorbiaceae omujimbu 3050 AC, 18, 75 RB MW H B C BS Brillantaisia owariensis Acanthaceae omuliwchenje 3027 AC, 22 RB Be Cannabis sativa Cannabaceae enzayi 4126 AT A Carapa grandiflora Meliaceae omuruguya E B Be Fa BS Carduus kikuyorum Asteraceae ekyigyembagyem 4089 AC, 123 RB MW Celtis durandii Ulmaceae omunuka 3111 AC B Chassalia cristata Rubiaceae 3155 AC Chrysophyllum gorungosanum Sapotaceae omushayu 3090, 4023 AC B BS Cissus sp. Vitaceae ibombo 4137 AT H Clerodendrum buchholzii Verbenaceae ekigugunya 3144 AC Clerodendrum schweinfurthii Verbenaceae 3042a AC Clutia abyssinica Euphorbiaceae omubalama 48 RB MW Coccinia barteri Cucurbitaceae ekikunjabutima 4138 AT H Coccinia mildbraedii Cucurbitaceae omwobeire 139 RB HD Commelina capitata Commelinaceae enteija 4125 AC, 3125 AT H Commelina diffusa Commelinaceae enteija 4059 AC, 27 RB MW Conyza sp. Asteraceae eshwiiga 87 RB MW E Conyza sumatrensis Asteraceae ekyizimyamurilo 4061 AC MW Crassocephalum crepidioides Asteraceae ekyizimyamurilo 4038 AC H Crotalaria sp. Fabaceae 167 RB Croton macrostachyus Euphorbiaceae omurangara 7 RB MW Croton megalocarpus Euphorbiaceae omuvune 4012 AC B BS Cucumella sp. Cucurbitaceae akabindizi 4015, 4075 AC MW H Cupressus lusitanica Cupressaceae Be Cuscuta sp. Convolvulaceae olubulameizi 27 RB MW Cyathea manniana Cyatheaceae ekigunju B Cyperus latifolius Cyperaceae ekigaga 104 RB W Cyperus papyrus Cyperaceae efundjo W G Cyperus sp. Cyperaceae entatala 117 RB W

54 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM BOTANICAL NAME FAMILY RUKIGA NAME VOUCHER NO. USE Cyphostemma sp. Vitaceae ekibombwe 3105 AC H Dalbergia lactea Fabaceae 3116a AC Desmodium repandum Fabaceae endebura 3268, 4076 AC MW H Dichaetanthera corymbosa Melastomataceae omunyinju 3038 AC B Dioscorea bulbifera Dioscoreaceae Dioscorea sp. Dioscoreaceae ebikwa 3030 AC E Dioscorea sp. Dioscoreaceae ekihama 3247 AC E Dissotis senegambiensis Melastomataceae omwonyongwente 4040, 4047 AC MW Dodonaea viscosa Sapindaceae omushambya A DC Dombeya kilimandscharica Sterculiaceae omukole 32 RB MW Dombeya torrida Sterculiaceae omukole 3241 AC MW Dracaena laxissima Agavaceae enchence W Dracaena sp. Agavaceae omugurura 3262 AC Fe Drypetes bipindensis Euphorbiaceae omushabarara 3180 AC, 59 RB SH WS B Drypetes gerrardii Euphorbiaceae omushabarara Be Pe BS Drypetes ugandensis Euphorbiaceae omushabarara 3057 AC B Be BS Eleusine indica Poaceae enchenzi 3009, 3202 AC W Entandrophragma excelsum Meliaceae omuyovu H B BB Eriosema montanum Fabaceae omucwafuka 4074 AC MW HD Eucalyptus sp. Myrtaceae uketusi B BS Euphorbia candelebrum Euphorbiaceae entakala Fe Euphorbia schimperiana Euphorbiaceae kamaramahano 128 RB MW Faurea saligna Proteaceae omulegyere BH Be B Ficalhoa laurifolia Theaceae omuvumaga 44 RB B BH BS Ficus asperifolia Moraceae omusomora 3051 AC SP Ficus exasperata Moraceae omusomora 4118 AT SP Ficus sp. Moraceae ekyitoma 3048 AC FW A Ficus spp. (F. ovata) Moraceae ekyitoma E BB Ficus sur Moraceae omulehe E BB Flabellaria paniculata MaIpighiaceae 4021 AC T Galiniera saxifraga Rubiaceae omulyanyonyi 4127 AT, 69 RB BS B FW Geranium arabicum Geraniaceae akanziranzira 92 RB MW Gloriosa sp. Liliaceae 3158 AC H Glyphaea brevis Tiliaceae omusingati 3029 AC T G FW Gouania longispicata Rhamnaceae omufurura 4072 AC MW Grewia sp. Tiliaceae omutahendeka 4094 AC G W Hagenia abyssinica Rosaceae omujesi 3263 AC FW Hallea rubrostipulata Rubiaceae engomera Be Hallea rubrostipulata Rubiaceae omuzibaziba 4109 AT A Hallea rubrostipulata Rubiaceae omuziku H Be Harungana madagascariensis Clusiaceae omunyananga 3091 AC B BS Harungana madagascariensis Clusiaceae omwongolero B Be Helichrysum foetidum Asteraceae enkyeza 90 RB MW Helichrysum foetidum Asteraceae okanyunya 4041 AT MW Helichrysum sp. Asteraceae akatoma 45 RB MW Hibiscus fuscus Malvaceae omusinga 3239, 4069 AC, 131 RB MW G W BH DC Hibiscus sp. Malvaceae omuchibikanumi 3275 AC H T Hippocratea odongensis Celastraceae oruyangaro 3184 AC T Impatiens stuhlmannii Balsaminaceae omulembe 4062 AC MW Indigofera cf. arrecta Fabaceae omushoroza 100 RB Tr Br Indigofera sp. Fabaceae omunyzaba-shum 3227 AC MW Jasminum sp. Oleaceae akababalira H Kalanchoe sp. Crassulaceae enchenanchene 3266 AC, 15 RB MW H Kalanchoe sp. Crassulaceae enjugoto 63 RB MW Keetia molundensis Rubiaceae 3145 AC Lantana triphylla Verbenaceae omuhukye 4090 AC MW Laportea sp. Urticaceae efugwe 3272 AC E Lentinus prolifer Fungi: Lentinaceae ebishanja E Leonotis neptifolia Lamiaceae ekicumucumu 4041 AC MW Leucas deflexa Lamiaceae akanyamafundo 4082 AC, 94 RB MW Linociera iohnsonii Oleaceae omuteze 3129 AC S Lobelia gibberoa Lobeliaceae ontomvu 4113 AT A Loesenerilla apocynoides Celastraceae omujega 3095 AC W G BH Lysimachia ruhmeriana Primulaceae omwisamura 4048, 4070 AC MW Macaranga kilimandscharica Euphorbiaceae omurara 3036, 3260 AC B Be BS Macaranga monandra Euphorbiaceae 2656 AC Macaranga schweinfurthii Euphorbiaceae omukole MW Maesa lanceolata Myrsinaceae omuhanga 4073 AC, 2, 10 RB MW H B FW BS Maesopsis eminii Rhamnaceae omuguruka 3040 AC H Be B Manihot utilissima Euphorbiaceae BS Marantochloa leucantha Marantaceae omwiru 3154, 3201 AC, 4117 AT W A Marattia fraxinea Pter: Marattiaceae ekitumbagire 3018 AC, 13 RB MW H Markhamia lutea Bignoniaceae omusavu Mo B Maytenus acuminata Celastraceae omulembwe 4035 AC H WS A Memecylon sp. Melastomataceae

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 55 People, park and plant use A.B. CUNNINGHAM A = Veterinary medicines, B = Building, BB = Beer boats, Be = Bee nectar and nesting sites, BH = Bee hives, Bl = Blacksmiths, Br = Brooms, BS = Bean stakes, Bw = Bows and arrows, C = Charcoal, Cm = Combs, Co = Covers for food, pots, etc., Cu = Cups, DB = Straw for drinking beer, DC = Dental Care, E = Edible, Fa = Fats, Fe = Fencing, FW = Firewood, G = Granaries, H = Herbalist, Gu = Gums, HD = Hunting dog medicines, HH = Hoe handles, Mo = Mortars, MW = Mid wives, P = Pipes, Pe = Pestles, Po = poison for arrows, R = Rakes, SH = Spear Handles, T = Twine and rope, Tr = to trap moles, Ty = Trays, W = weaving, WS = Walking sticks.

BOTANICAL NAME FAMILY RUKIGA NAME VOUCHER NO. USE Millettia dura Fabaceae omutate 20 RB MW HH B FW Millettia psilopetala Fabaceae 3164 AC Mimulopsis solmsii Acantheceae ekiwisi 4020 AC Be Monanthotaxis sp. Annonaceae omutaru G Monopsis stellarioides Lobeliaceae akararambwe 4055 AC MW Musa sp. Musaceae enjagata BH Musanga Ieo-errerae Cecropiaceae omutunda 3041b, 3013 AC H FW Myrianthus holstii Cecropiaceae omufa/omufe 30, 36 RB MW Myrica salicifolia Myricaceae omujeje 3087 AC, 78 RB MW H Neoboutonia sp. Euphorbiaceae omwaya 25 RB Co Newtonia buchananii Fabaceae omutoyo/omukun 3046 AC B FW BB Ochna sp. Ochnaceae omnzigani 3082 AC H Ocotea usambarensis Lauraceae omwiha 3081 AC H B FW BB Oxalis corniculata Oxalidaceae obunya 4058 AC MW Oxyanthus speciosus Rubiaceae oruchiraje 3157 AC SH HH Oxyanthus subpunctatus Rubiaceae 4037 AC BS Parinari excelsa ssp. holstii Chrysobalanaceae omushamba 3015 AC B C Pavetta abyssinica Rubiaceae omunagatunguru Peddiea fischeri Thymelaeaceae omuzinya 3255 AC, 4124 AT H T Pegularia sp. Asclepiadaceae omulandagasi 79 RB T W Pellaea doniana Pter: Adiantaceae akanyasiru 7 RB MW Pellaea viridis Pter: Adiantaceae orushwiga 3229 AC, 66 RB MW Pennisetum purpureum Poaceae ekyibingo 3134 AT G BS Pentas longifolia Rubiaceae ishangala/esingala 4077 AC, 137 RB MW Peponium vogelii Cucurbitaceae omugoshora 132 RB MW Pergularia extensa Asclepiadaceae ekyikurakura 4129 AT MW Phoenix reclinata Arecaceae ekyindo W Phyllanthus fischeri Euphorbiaceae omulisafumbiri 4067 AC MW Phyllanthus sp. Euphorbiaceae isheru 3224 AC MW Phyllanthus sp. Euphorbiaceae omuturika 3228 AC MW Physalis peruviana Solanaceae entutu 89 RB MW E Phytolacca dodecandra Phytolaccaceae omuhoko 4108 AT, 19, 28 RB MW A Piper (guineense?) Piperaceae rukokota 3122, 3142 AC, 4116 AT H B Piper capense Piperaceae orubogote 16, 65 RB BS BH Pittosporum spathicalyx Pittosporaceae omushekyera BS BH Plantago palmata Plantaginaceae embatabata 35 RB MW G Plectranthus (albus?) Lamiaceae ekyiagaga 4107 AC A Pleiocarpa pycnantha Apocynaceae omutoma 3132, 3159, 3242 AC P Pleopeltis macrocarpa Polypodiaceae orushanjamizi 3100 AC T Podocarpus latifolius Podocarpaceae omufu B Pollia condensata Commelinaceae omabahashiha Co Bs Polygala ruwenzoriensis Polygalaceae egongwe 3141 AC, 61 RB WS HH Polygala sp. Polygalaceae omuseresere 3259 AC, 46 RB Bw Polyscias fulva Araliaceae omungo H BH Premna sp. Verbenaceae omuguna 4011 AC Db Prunus africana Rosaceae omumba 64 RB MW HH Be BB Mo Psychotria mahonii Rubiaceae omukali 67 RB BS WS Psychotria schweinfurthii Rubiaceae omutegashali 4034, 4096 AC BS Pycnostachys elliotii Lamiaceae ekyisindokwa 3222, 4086 AC MW Pycnostachys goetzenii Lamiaceae ekyisindokwa 4019 AC H Pycreus sp. Cyperaceae ekubo 119 RB MW Ranunculus multifidus Ranunculaceae itengye 41 RB MW Raphia farinifera Arecaceae ekihungye 3123 AC W Rawsonia spinidens Flacourtiaceae omusadya 3128 AC, 4122 AT Be Cm Rhipsalis baccifera Cactaceae engurukira 3101 AC, 34 RB MW H W Rinorea ferruginea Violaceae omunyashandu 3096 AC R B Rothmannia longiflora Rubiaceae oruchiraje 3131 AC SH Rubia cordifolia Rubiaceae okaramba 3234 AC MW Rubus sp. Rubiaceae emerembwe E Rubus steudneri Rosaceae omucereli 4053 AC MW Rumex bequaertii Polygonaceae omuku 4085 AC MW Rumex usambarensis Polygonaceae omufumbwa 4130 AT, 12 RB MW DC E Be Rutidea orientalis Rubiaceae 154 RB Rytigynia kigeziensis Rubiaceae nyakibazi 3113 AC, 43, 106 RB MW H Sabicea sp. Rubiaceae endarayiguru 3086 AC H Salacia elegans Celastraceae oruyangara 4095 AC G Salacia sp. Celastraceae bwara 53 RB G BH

56 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM BOTANICAL NAME FAMILY RUKIGA NAME VOUCHER NO. USE Sapium ellipticum Euphorbiaceae omushasha 3105 AC, 68 RB BS B FW Pe Schefflera (barteri?) Araliaceae omuchulaaraga 3041a AC H Schefflera barteri Araliaceae omwamira 3116 AC, 105 RB HD B BE Schefflera sp. Araliaceae omuhamika 4041 AC H Senecio sp. Asteraceae ekizimyamuriro 3230 AC, 140 RB MW Sericostachys scandens Amaranthaceae omuna 4063 AC, 19 RB MW Be Sesbania sesban Fabaceae entahutara H Setaria plicatilis Poaceae ekikoka 141 RB W Sida sp. Malvaceae omuchundezi 3120 AC Gu Smilax anceps Smilacaceae enshuli 54 RB W Solanum Solanaceae ekyihiara 3240a AC MW Solanum (aculeastrum?) Solanaceae omutugunda 4054 AC, 4106 AT MW A Solanum nigrum Solanaceae entakara 103 RB E Solanum sp. Solanaceae enturameshu H Spermacoce princeae Rubiaceae enyabata 83 RB MW Strombosia scheffleri Olacaceae omuhika 3039 AC, 52 RB B Be Symphonia globulifera Clusiaceae omusisi H Be B BS Synadeniurn sp. Apocynaceae omukone 4135 AT, 144 RB BS WS A Syzygium guineense Myrtaceae omugote B C Be Tabernaemontana odaratissima Apocynaceae kinyamangozi B Tabernaemontana pachysiphon Apocynaceae kinyamangozi 3007 AC, 4029 AT MW HD B Tabernaemontana sp. Apocynaceae kinyamate B Tagetes minuta Asteraceae lunuka 4079 AC, 88 RB MW Teclea nobilis Rutaceae omuzo 60 RB WS B Tephrosia vogelii Fabaceae omubarama 159 RB Tephrosia vogelii Fabaceae omukurukuru 3256 AC, 4131 AT BS A Thalictrum rhynchocarpum Ranunculaceae omwintago 3124 AC HD Thunbergia mildbraediana Acanthaceae ekyikurakura 4024 AC MW Trema orientalis Ulmaceae omubengabakwe 3108 AC B FW Trichilia rubescens Meliaceae omununka 3178a AC Bw Trilepisium madagascariense Moraceae omukumbwe 3031, 3153 AC E B Be Triumfetta longicornuta Tiliaceae oruhigura 4068 AC MW Triumfetta macrophylla Tiliaceae omunaba MW T A Urera hypselodendron Urticaceae omushe 3028, 4091 AC MW T BH Vernonia congolensis Asteraceae omwilima 3011, 4084 AC, 29 RB MW H Vernonia iodocalyx Asteraceae omutahendeka 72 RB MW Vernonia kirungae Asteraceae ekigaragara 4111 AT, 142 RB MW A Vernonia sp. Asteraceae echiheriheri 3149 AC, 4028 AT BS Vernonia sp. Asteraceae omukurenju 4088 AC MW Xymalos monospora Trimeniaceae omuhotora 74 RB BS HH B Zanthoxylum gilletii Rutaceae omulemankobe H B Zanthoxylum leprieurii Rutaceae omuchanga/omucharandi 3047 AC H (Bulbostylus?) oburahuka 3240 AC MW (Clematis?) oruzibira 3236 AC MW (Combretum?) enkazi 3233 AC MW (Crotalaria?) eyituza 3238 AC MW (Cyathula?) ekishokonkoro 120 RB MW (Friesadelsia?) omutaro 3099, 3156 AC W (Kosteletzkya grantii?) omuzigonumi 4081 AC MW (Pseudarthria?) omukongoroni 3231 AC MW (Rawsonia?) omushadya 3128 AC B Mo (Ritchiea albersii?) omuheru 50, 62 RB Be Cu (Serkasia?) omunyivi 105 RB B Be gen. indet. Asclepiadaceae omukumbo 3031 AC E gen. indet. Asclepiadaceae omunuka 3175 AC Bw gen. indet. Asteraceae eirarira 3232 AC MW gen. indet. Asteraceae ekyifuramende 4087AC MW gen. indet. Asteraceae ekyoganyaja 4046 AC MW gen. indet. Asteraceae ekyoganyaja 3235 AC MW gen. indet. Asteraceae enyongyera 114 RB MW gen. indet. Asteraceae esemwe 4083 AC MW gen. indet. Asteraceae esununu 4027 AC, 91 RB MW H gen. indet. Asteraceae okatoma 4039 AC MW gen. indet. Asteraceae omushura 3237 AC MW gen. indet. Brassicaceae? mwetango H gen. indet. Lamiaceae ekisindokera 33 RB MW gen. indet. Lamiaceae ekymwa 115 RB MW gen. indet. Lamiaceae omuku 118 RB MW gen. indet. Malpighiaceae orumaga 3102 AC T Be gen. indet. Malvaceae emikungyere 122 RB T gen. indet. Malvaceae olukoma 116 RB A gen. indet. Meliaceae gen. indet. Rhamnaceae akanyarwakasaka 47 RB MW gen. indet. Rubiaceae ekinyamazi 3136 AC H gen. indet. Rubiaceae ohubango 57 RB SH

PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 57 People, park and plant use A.B. CUNNINGHAM A = Veterinary medicines, B = Building, BB = Beer boats, Be = Bee nectar and nesting sites, BH = Bee hives, Bl = Blacksmiths, Br = Brooms, BS = Bean stakes, Bw = Bows and arrows, C = Charcoal, Cm = Combs, Co = Covers for food, pots, etc., Cu = Cups, DB = Straw for drinking beer, DC = Dental Care, E = Edible, Fa = Fats, Fe = Fencing, FW = Firewood, G = Granaries, H = Herbalist, Gu = Gums, HD = Hunting dog medicines, HH = Hoe handles, Mo = Mortars, MW = Mid wives, P = Pipes, Pe = Pestles, Po = poison for arrows, R = Rakes, SH = Spear Handles, T = Twine and rope, Tr = to trap moles, Ty = Trays, W = weaving, WS = Walking sticks.

BOTANICAL NAME FAMILY RUKIGA NAME VOUCHER NO. USE indet. Fungi ebishanja E indet. Fungi ensabili E indet. Fungi obutusi E indet. akahengyeri 4047 AC MW indet. akanyamatta 9 RB MW indet. akanyaruguma 4152 AC MW indet. akasanjamaizi MW indet. akatampihi 4050 AC MW indet. amazirani 95 RB MW indet. bitindi 3137 AC B indet. chumaya 1 RB MW indet. ebifunjo G indet. ebikanga W indet. ebizogwa G indet. echitambampazi MW indet. ecinza 4056 AC MW indet. ekarwe DC indet. ekibingo MW BS indet. ekichumachumu 28 RB MW indet. ekichuruchubi 138 RB MW indet. ekigazura Be indet. ekigorogoro 133 RB MW indet. ekikondogoro 58 RB MW P indet. ekimanki Po indet. ekimara 11 RB MW indet. ekitondore 10 RB MW E indet. ekituruguma 101 RB MW E indet. ekyizimyamuliro MW indet. embungu 55 RB T W G indet. emisese G indet. emitembe 126 RB E indet. enfunjo G indet. engongwe Be indet. enshekashekye 136 RB Co indet. enshirabwiko 13 RB MW indet. entagara 4110 AT A indet. entahutara 4051 AC MW indet. enyamunuka 25 RB MW indet. enyogwa 3248 AC G indet. enzogwa 3248 AC G indet. etangala Fe indet. etaritari G indet. ihoza 130 RB MW indet. ikibonobono A indet. ishesha 20 RB MW indet. kaijajuba 8 RB MW indet. kashundwe 3221 AC MW indet. katamba 8 RB MW indet. katampihi 26 RB MW indet. kitaluwia H indet. kyemerwa 21 RB MW indet. matunda E indet. molobungu 3006 AC G indet. mwatambale Bl indet. obukako G indet. obukaraati 99 RB E indet. omwenyi 4140 AT A

58 PEOPLE AND PLANTS WORKING PAPER 4, DECEMBER 1996 People, park and plant use A.B. CUNNINGHAM Already published in this series:

1. Cunningham, A. B. 1993. African medicinal plants: Setting priorities at the interface between conservation and primary healthcare.

2. Cunningham, A. B. and Mbenkum, F.T. 1993. Sustainability of harvesting Prunus africana bark in Cameroon: A medicinal plant in international trade.

3. Aumeeruddy, Y. 1994. Local representations and management of agroforests on the periphery of Kerinci Seblat National Park, Sumatra, Indonesia. (This publication is also available in French.) The People and Plants Initiative was started in July 1992 by WWF, UNESCO and the Royal Botanic Gardens, Kew to promote the sustainable and equitable use of plant resources through providing support to ethnobotanists from developing countries.

The initiative stems from the recognition that people in rural communities often have detailed and profound knowledge of the properties and ecology of locally occurring plants, and rely on them for many of their foods, medicines, fuel, building materials and other products. However, much of this knowledge is being lost with the transformation of local ecosystems and local cultures. Over- Contact addresses: harvesting of non cultivated plants is increasingly common, caused by loss of habitat, increase in local WWF International use and the growing demands of trade. Long-term Plant Conservation Officer conservation of plant resources and the knowledge Panda House, Weyside Park associated with them is needed for the benefit of the Godalming, Surrey GU7 1XR local people and for their potential use to local UNITED KINGDOM communities in other places. Fax: 44 1483 426409

The diversity of traditional plant-resource management practices runs through a spectrum from “cultivation” through to gathering “wild” plants, all of which are included in the People and Plants Division of Ecological Sciences approach. Man and the Biosphere Programme UNESCO, 7 Place de Fontenoy Ethnobotanists can work together with local people to 75352 Paris Cedex 07 SP study and record the uses of plant resources, identify FRANCE cases of over-harvesting of non-cultivated plants, find Fax: 33 1 45685804 sustainable harvesting methods and investigate alternatives such as cultivation.

The People and Plants initiative is building support for ethnobotanists from developing countries who The Director work with local people on issues related to the Royal Botanic Gardens, Kew conservation of both plant resources and traditional Richmond ecological knowledge. Key participants organize Surrey TW9 3AB participatory workshops, undertake discussion and UNITED KINGDOM advisory visits to field projects and provide literature Fax: 44 181 3325278 on ethnobotany, traditional ecological knowledge and sustainable plant resource use. It is hoped that a network of ethnobotanists working on these issues in different countries and regions can be developed to exchange information, share experience and collaborate on field projects.