Pores;~;ology Management Forest Ecology and Management 77 (1995) 107-117 Ecology of a miombo fruit tree: Uapaca kirkiana ( Euphorbiaceae) Mzoma R. Ngulube a,*, John B. Hall b, J.A. Maghembe ’ a Forestry Research Institute of Malawi, P.O. Box 270, Zomba, Malawi b School of Agricultural and Forest Sciences, University College of North Wales, Bangor LL 57 ZUW, UK ’ International Centre for Research in Agroforestry &XAFI, SADC/ICRAFAgroforestry Project, P.O. Box 134, Zomba, Malawi Accepted 20 April 1995 Abstract Recordsin herbariaandpublishedinformationon the distributionof Uupucu kirkiuna Muell. Arg. were assembled and summarizedas a distributionmap. Soil and vegetationmapsof Africa, meteorological data and informationon ecological accountswereusedin the interpretationof the map.Uupucu kirkiunu is reportedfrom mostcountrieswithin the Zambezian centre of endemismand adjacenttransitionalphytochoria.The specieshasbeen recordedgrowing in Angola, Burundi, Malawi, Mozambique,Tanzania,Zaire, ZambiaandZimbabwe. Within the naturalrange,occurrenceis relatedto a unimodalrainfall regimewith an annualrangeof 500-1400 mm occurringover a 4-5 monthperiodfollowedby a longdry season lasting5-7 months.The meanday timetemperature range is 18-29°C in the hot seasonand 12-24°C in winter. Typically Uupacu kirkiuna growsin well-drainedescarpments at altitudesof 500-2000 m, with infertile sandor gravelly soilsof acidic reaction.As a miombospecies,it occursin mixed communities of Bruchystegiu-Julbernurdiu woodlandasdominantor co-dominantspecies,often gregarious,formingdense groves,commonlyreferredto asUupucuwoodlands.Throughoutthe range,listingsof associated woody speciesfrequently include Albiziu, Anisophylleu, Bruchystegiu, Burkeu, Isoberliniu, Julbernurdiu, Monotes, Purinuri, Proteu, Pericopis, Pterocurpus, Ochnu andother Uupacuspp. Attention is drawn to various relationshipswith mycorrhizaeand animals(including man) and their significancein conservationand management options.Definitive studiesaimedat generatingrelevant data to aid husbandryaction are outlined.Suggestions for positivemanagement prescriptionsanddomestication initiativesare made. Keywords: Uapaca kirkiana; Ecology; Distribution; Conservation; Management; Mycorrhizae 1. Introduction The genus Uupacu occurs in tropical Africa (including Madagascar), mostly in the miombo region and in the closed lowland forests. The most widespread and best known is Uapaca kirkiana Muell. Arg., a typical miombo woodland species * Corresponding author. (Fig. 1). Uupaca kirkiana is an important multipurpose tree especially valued for its edible fruits (Hans et al., 1978; Storm, 1979; Palgrave, 1981; Drummond, 1981; Fox and Young, 1982; Pullinger and Kitchin, 1982; Food and Agriculture Organization (FAO), 1983, 1990; Campbell, 1987; Mwamba, 1989a; Seyani, 1991; Peters et al., 1992; Lovett, 1993). In addition, the wood is durable and useful for general purposes(Goldsmith and Carter, 1981>, while its roots and bark have values medicinally and 0378-1127/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0378-1127(95)03572-9 108 M.R. Ngulube et ul. /Forest Ecology for dye production (Fanshawe, 1968b; Storm, 1979; Palgrave, 1981). The fruits of U. kirkiana have high nutritional values and play an important dietary role (Carr, 1957; Sufi and Kaputo, 1977; Malaisse and Parent, 1985; Saka et al., 1992; Saka and Msonthi, 1994). Indeed, in recognition of its importance as a resource of great potential, it was declared a protected food tree in Barotseland (Western Province, Zambia) as early as the 1950s (Palgrave and Palgrave, 1957). The fruits are collected from the wild and sold at the roadside and in town centres and city markets. They are a significant source of income in rural areas. In recent years, the development of cottage wine industries in Malawi and Zambia has opened an attractive additional market for the fruit. Large quantities of fruits are now required to sustain the winery industries as well as to meet growing demand for whole fruits in urban and city markets. Although U. kirkiana continues to play a vital role where natural stands still remain, sustainability is seriously threatened by the high deforestation rates currently affecting the miombo woodlands of southern Africa (Hyde and Seve, 1993). Despite its potential for development as a food and commercial crop, the tree has received little scientific attention. However, while references in the literature are sparse, Fig. 1. Uupaca kirkiunn and Management 77 (1995) 107-117 scattered and mostly descriptive, ethnobotanical and socioeconomic studies (Maghembe and Seyani, 1991; Kwesiga and Chisumpa. 1992; Grundy et al., 1993) have revealed that local farmers have great interest in the domestication of U. kirkiana as a multipurpose tree to supply growing local markets and to cater for immediate household needs. Data are needed to guide and support the development and management of U. kirkiuna as a resource in southern Africa This paper therefore provides background on the ecology of U. kirkiana, unifying information assembled from published sources, floras, check lists and ecological and vegetation accounts. A distribution map has been prepared using herbarium data and literature, and has been interpreted in terms of emi.ronmental factors, principally climate on a rangewide basis. 2. Range Uapaca kirkiana occurs naturally south of the equator in Angola, Zaire, Burundi, Tanzania, Malawi, Mozambique, Zambia and Zimbabwe (Fig. 2). From its northern limit in Geita district-at Uzinza (02”3l’S, 32”43’E; Burtt Davy 6515, K) in Tanzania -the range of U. kirkiana extends southwards to trees in a natural stand at Malawi College of Forestry, Dedza, Malawi. M.R. Ngulube et al./Forest Ecology Zimbabwe in Lomagudi district-at Mangula (20”58’S, 31”56’E; Jacobsen 1195, K)-as the most southern limit. The most easterly occurrences are in Mpanda district, Tanzania-at Kurasimba (06”5O’S, 39”17’E; Semsei and Herring 44, FHO). The disjunctive Angolan population in Benguela Highlands-at Cutato (09”44’S, 14”20’E; Gossweiler 3801, BM)-marks the most westerly longitudinal limit of the range. and Management 77 (199.5) 107-117 109 10" 10" 0" 3. Relations with environmental 0" factors 3.1. Elevation Occurrence is generally between 500 and 2000 m elevation. No records refer to occurrences below 500 m or above 2500 m. The highest elevation recorded 10" 10" 20" 20" 30" 30" 20" 40" t 17 i 7 30" 40" 10" 0" 10" 20" 30" 40" 50" I Fig. 3. Distribution (dots) of Uapaca kirk&a in relation to rainfall (400, 800 and 1400 mm isohyets shown). Areas at low elevation subject to frost are demarcated by broken lines. 130" 20" 10" 0" 10" 20" 30" I . 20" 10" 0" 10" 20" 30" 40" 50" I Fig. 2. Distribution (dots) of Uapaca kirkiana in relation to the main phytochoria of White (1983): 1, Guineo-Congolian centre; 2, Zambezian centre; 3, Sudanian centre; 4, Somalia-Masai centre; 5, Cape centre; 6, Karoo-Namib centre; 7, Mediterranean centre; 8, Afromontane/afroalpine centre; 9, Guinea-Congolia/Zambezia transition; 10, Guinea-Congolia/Sudania transition; 11, Lake Victoria mosaic; 12, Zanzibar-Inhambane mosaic; 13, KalahariHighveld transition; 14, Tongaland-Pondoland mosaic; 15, Sahel transition; 16, Sahara transition; 17, Mediterranean-Sahara transition; 18, East Malagasy centre; 19, West Malagasy centre. appears to be 2400 m (Benguela Highlands, Angola; Welwitsch 455, BM; Mbeya Highlands, Tanzania; Procter 3108, K, Mbala, Zambia; Burtt Davy 6310 and Richards 5077, K). Populations in MalawiLitende and Nkhota-kota escarpments (550 m) along Lake Malawi (Shorter, 1989)-appear to be the lowest occurrences. 3.2. Climate The whole range of U. kirkiana broadly experiences a 5-7 month main dry season, dry months being those with rainfall < 50 mm. Most reports of occurrence relate to areas enjoying a mean annual rainfall between 500 mm and 1400 mm (Fig. 3). Mean annual temperatures range from 18°C to 29°C with monthly means of daily temperature maximum consistently below 29°C and corresponding minimum above 10°C. Absolute temperature maximum often exceeds 30°C and minimum around 0°C with frequent frosts in the southern areas of the range (Fig. 3). Leaf damage of U. kirkiana through frost scorching has been reported in Malawi (Anonymous, 1955) and Zimbabwe (Anonymous, 1951). 110 M.R. Ngulube et al. /Forest Ecology and Management 3.3. Soils neous rock types, mainly schist, gneiss, granite, granodiorite, magnetite and quart&e (Phipps and Goodier, 1962; Watson, 1964; Lawton, 1978; Mwamba, 1983; FAO, 1983; McGregor, 1994). The soils are Uapaca kirkiana grows on ferruginous or ferralitic soils derived from different metamorphic and igTable 1 List of frquently reported associates of U. kirkiana Species Country A% * Albizia antunesiana Harms Albizia uersicolor Oliv. Anisophyllea pomifera Engl. & Brehm Annona senegalensis Pers * Brachystegia boehmii Aub. Brachystegia bussei Harms * Brachystegia floribunda Benth. Brachystegia glaberrima Re. Fr. * Brachystegia longifolia Bentn Brachystegia microphyila Harms Brachystegia manga De Wild. * Brachystegia spiciformis Benth. Brachystegia taxifolia Harms Brachystegia utilis Bum-Davy & Hutch. * Brachystegia wangermeeana De Wild Bridelia micrantha (Hochst) Bail]. * Burkea africana Hook Catunaregam bispinosa Keay Combretum molle R. Br. ex G. Don Cussonia arborea Seem. Dalbergia nitidula Welw ex Bak. Diplorhynchus condylocarpon (Muell. Arg.) Pith. Diospyros kirkii Hiern Faurea saligna Harv. Faurea speciosa Welw. Flaucortia indica Merr. Isoberlinia angolensis (Welw. ex Bak.) Hoyle & Brenan Julbernandia globifrora (Benth.) Troupin * Julbernandia paniculata (Benth.) Troupin * Lannea discolor (Sand.) Engl. * Monotes africanus A. DC. Ochna schweinfurhiana F. Hoffm. * Parinari curatellifolia Benth. Pericopsis angolensis (Bak.) Van Meeuwen * Protea petiolaris (Engl. ex Hiern.) Bak. Pseudolachnostylis maprouneifoha Pax * Pterocarpus angolensis DC. Strychnos spinosa Lam. Syzygium guineense (Willd) DC. * Uapaca nitida Muell. Arg. * Uapaca robynsii De Wild. Uapaca sansibarica Pax Vangueria infausta Burch. Vitex doniana Sweet a Countries: Ang, Angola; 77 (19951107-117 Bur, Burundi, Mlw, Malawi; a Bur MlW * Moz Ti? * * * * * * * * * ‘i * Zai * * * I * * * * * d * * * * * * * * * Zim * * * Zam * * * * * * * * d * * * * * * * Moz, x * * * * A. * * * * * * 1 * * * * * * * * * * * * * * * * * * * * * * 1: * * * * * * * * * * $ * * * * * * * * * * * * * * * * * 1 * * * * * * c * * * * * * * * Tz, Tanzania; * * * Mozambique; * Zai, Zaire; * * * * * * * * Zam, Zambia; * * Z&n, Zimbabwe. M.R. Ngulube et al. /Forest Ecology generally sandy or gravelly with good drainage. The absence of U. kirkiana on clay dominated or hydromorphic soils with poor drainage (Burn-Davy and Hoyle, 1936; Phipps and Goodier, 1962; Savory, 1963; Fanshawe, 1968a,b; Lawton, 1978; Hans, 1981; Mwamba, 1983) signifies the importance of soil texture in its occurrence. Where U. kirkiana occurs, the soils are generally characterized by low cation exchange capacity, low organic matter content and macro nutrients: nitrogen, phosphorus and potassium (Phipps and Goodier, 1962; Watson, 1964; Mwamba, 1983). Soil reaction is acidic with pH between 4 and 6, but 5-5.5 being the most ideal (Mwamba, 1983). 3.4. Site Uapaca kirkiana favours the freely draining middle slopes of escarpments. A soil profile of shallow gravelly soil overlying weathered rock is typical in these situations. In tree cover, U. kirkiana dominates over Brachystegia spp. (Watson, 1964). On sites with gravelly soils overlain by sandy loam surface material, U. kirkiana is largely replaced by Julbernardia globiflora (Benth.) Troupin and Burkea africana Hook or Parinari curatellifolia Benth. if seasonally high water table conditions exist. Where hydromorphic soils with high organic matter content overlie gravelly soils and seasonal flooding conditions prevail, U. kirkiana is totally absent (Pielou, 1952; Watson, 1964). If, however, the general topography is gentle, clear toposequence relationships inventory of U. kirkiana and related species are not evident (Lawton, 1978). 4. Uapaca kirk&a as a vegetation component 4.1. Chorology and vegetation formations Uapaca kirkiana is a species typical of the Zambezian Regional Centre of endemism and adjacent transitional centres (White, 1983). It is widespread and abundant in mixed communities of Brachystegia-Julbernardia woodland vegetation either as a dominant or co-dominant species, and is usually gregarious, forming dense groves. Where rainfall is high (> 1200 mm), U. kirkiana forms pure woodland communities with either closed or open and Management 77 (1995) 107-117 111 canopies, approaching semi-deciduous forest (Rattray, 1961; Shorter, 1989). In such communities, the ground flora is usually sparse or absent on account of the shade cast by the large crowded leaves of U. kirkiana (Pardy, 1951; Rattray, 1961; Astle, 1968). In natural populations, U. kirkiana is easily distinguished from other associated Uapaca spp. on account of its distinctive broad and leathery leaves and rounded crown. 4.2. Associated species Literature comments suggest associations of U. kirkiana with Albizia, Anisophyllea, Brachystegia, Burkea, Isoberlinia, Julbernardia, Monotes, Parinari, Protea, Pericopis, Pterocarpus, Ochna and other Uapaca spp. throughout the natural range. Records of association at the species level are, however, scanty (Table 1). Most of the species recorded as close associates occur in more than three countries within the natural range. 4.3. Prominence and population levels Few estimates of relative abundance of U. kirkiana in miombo woodlands have been traced. A few available reports (Lawton, 1978; Chidumayo, 1987) record percentage with reference to stems, height, diameter or basal area. Nevertheless, they provide indications of the prominence of U. kirkiana and its contribution to the vegetation community within its natural range. About 54-74% U. kirkiana representation (all sizes) has been recorded in natural stands (Lawton, 1978). For trees of > 10 m height or > 60 cm g.b.h., Chidumayo (1987) recorded 10.3% and 27.4% respectively as U. kirkiana. Where pure stands exist, > 90% representation is not an overestimate, although many sources indicate a non-balanced size class distribution with more individuals within the lower diameter class (Chidumayo, 1987). Following chitemene cultivation (trees chopped/pollarded and the branches burnt over the intended site of cultivation), enumeration of vegetation in a 6-year-old fallow revealed that 42% of the basal area was contributed by U. kirkiana (Stromgaard, 1985). In a mature natural stand, U. kirkiana accounted for 4.8% of the basal area (Hogberg and Piearce, 1986). 112 5. Uapaea kinkianr M.R. Ngulube et al. /Forest Ecology in ecosystem interactions 5.1. Influence on soil environment No information on the effects of U. kirkiana directly or through litter fall and decomposition on the soil environment has been seen. Nevertheless, foliar analysis of U. kirkiana (Ernst, 1975) revealed higher concentrations of copper than associated miombo species and the closely related Uapaca robynsii is an indicator of copper-rich soils (Shewry et al., 1979). An association of U. kirkiana with ectomycorrhizae has been noted in natural stands in Tanzania and Zambia (Hogberg, 1982; Hogberg and Piearce, 1986). Ectomycorrhizae are conspicuous on roots at 5-10 cm depth (Hogberg and Piearce, 1986) and endo-ectomycorrhizal associations also occur in U. kirkiana (Hogberg, 1982) as recorded in Uapaca guineensis (Thoen and Ba, 1989). For U. guineensis, 17 ectomycorrhizal fungi, including Amanita spp., Cantharellus congolensis Beeli, Lactarius gymnocarpus Hein. and Russula annulata Hein., have been recorded (Thoen and Ba, 1989). Amanita, Cantharellus, Lactarius and Russula constitute the major commonest genera of fungi which typically form ectomycorrhizae within the miombo woodlands (Hogberg and Piearce, 1986). Within the miombo woodlands, Amanita zambiana Pegler and Piearce, Cantharellus cibarius Fr., C. congolensis Beeli, Cantharellus densifolius Hein., Cantharellus longisporus Hein., Lactarius gymnocarpus Hein. and Russula delica Fr. are commonly associated with Uapaca-Brachystegia woodlands in general (Willianson, 1975; Pegler, 1977; Pegler and Piearce, 1980; Morris, 1987). Specific association of Cantharellus spp. and Lactarius gymnocarpus with U. kirkiana has been recorded in Malawi and the local names of these-‘Kamsuku’, ‘Nakasuku’, ‘Ngundasuku’ (Morris, 1987)-all bear testimony to the association since the tree is also locally known as ‘Msuku’. Amanita rubescens (Pers. Fr.1 SF. Gray, a putative ectomycorrhiza of U. guineensis (Thoen and Ba, 19891, has also been confirmed as a symbiont of U. kirkiana in Zambia (Mwamba et al., 19921. 5.2. Relations with the natural fauna Reported relationships with animal species in natural vegetation are limited. Uapaca kirkiana serves and Management 77 (1995) 107-117 as host for phytophagous and other insects reportedly collected off the tree (Lee, 1971). In addition, association with vertebrates arises from the dispersal role as source of food and other passive roles. CercopZastes uapacae Hall, Ledapis spp., Microsyugrus rosue Bry. and Euphoria spp. may cause lo-20% foliar damage in U. kirkiana (Parker, 1978). Curpohilus fumatus Boh. and Deudorix spp. attack mature fruits to feed on the pulp, followed by Drosophila enanasse Doleschall and Ceratitis cosyrae Walker which rapidly degrade the ripe fruits (Parker, 19781. Whether pollinators are specific is not reported: floral and pollen characters conform with entomophilous syndrome (Hans and Mwamba, 19821, but no reports on the pollination process have been traced. There are reports of association with. the natural vertebrate fauna in terms of both browsing and dispersal. Elephants (Loxodonta africana Blumenbach) have been noted as browsers, consuming shoots, leaves, bark and fruits in the Kasungu National Park (Jachman, 19891. Eland (Taurotragus oryx Pallas) and zebra (Equus burchelli Gray) have also been noted as browsers (Shorter, 19891. In addition to elephants, fruits are also eaten by baboons (Papio cynocephallus L.), blue monkeys (Cercopithecus mitis Wolff), velvet monkeys (Cercopithecus aethiops Linn.), thick-tailed galago (Galago crassicaudatus E. Geoffroy), lesser galago (Galago senegalensis E. Geoffroy) bush pigs (Potamochoerus porcus E., warthogs (Z’hacochoerus aethiopicus Pallas) and squirrels (Sciurus spp.). Of these, perhaps the most effective dispersers are the baboons and monkeys who only suck the fruit pulp, discarding the seeds. Elephants, bush pigs and warthogs, which ingest fruit, may evacuate these at considerable distance, but how this affects subsequent germination is not known. Passive roles arise from concealment, protective shelter or microclimate generated by the dense crowns of the trees individually or collectively within local populations. In Malawi and Zimbabwe, U. kirkiana trees serve as host to a hemipterous bug (Encosternum delegoruri Spin.) during winter (Makuku, 1993). These bugs benefit from the protective shelter and the morning dew from the trees. Nesting by several species of birds has been identified in natural vegetation where V. kirkiana is prominent. Pseudo-galls on U. kirkiana leaves, M.R. Ngulube et al. / Forest Ecology harbouring unidentified insects, have been noted. Larvae and pupae of unidentified lepidopterous species have also been noted in fruits and seeds of U. kirkiana. 5.3. Casual relations with man Bush fires from a variety of human activities are a feature of the miombo vegetation in which U. kirkiana is prominent. Uapaca kirkiana itself is moderately fire-resistant (Trapnell, 1959; Lawton, 1978; Kikula, 1986) and does not suffer from early dry season burning (Trapnell, 1959). Burns late in the dry season, however, affect U. kirkiana adversely through damage to coppice shoots or young seedlings and crowns of mature individuals (Trapnell, 1959; Lawton, 1978), particularly if repeated. The fleshy fruits contribute a substantial amount of animal feed, albeit for a short period (Walker, 1980). The flush of U. kirkiana leaves appearing at the end of the dry season is utilized by cattle as fodder in the absence of more palatable alternatives such as Albizia, Combretum, Brachystegia and Parinari (Rees, 1974; Lawton, 1980; Walker, 1980). With the growing utilization of miombo rangelands by cattle, U. kirkiana populations have been adversely affected by increased fire frequency. 5.4. Uapaca kirkiana and vegetation succession There has been discussion on fire and succession of the miombo woodland vegetation of southern Africa (Trapnell, 1959; Lawton, 1964; 1978; Kikula, 1986). The importance of fire-resistant species in the recovery of the miombo vegetation was noted in early studies involving controlled burning in Zambia (Trapnell, 1959). In these studies, U. kirkiana was one of the species found occupying an intermediate succes.sional stage between fire-tolerant woodland and fire-sensitive dry evergreen forest. Following exclusion of fire from regularly burnt sites, resilient species, such as Combretum molle, Pterocarpus angolensis, Diplorhynchus condylocarpon and Syzygium guineense, which are able to form a canopy, suppress the herb layer, and subsequently reduce the fire intensity, will establish. This enables species of Uapaca to establish themselves, creating more and Management 77 (1995) 107-117 113 favourable conditions for establishment of Brachystegia and Julbernardia spp. which require some protection to develop beyond the sapling stage. Uapaca spp. are eventually shaded out by the Brachystegia and Julbernardia spp. Stromgaard (1986), however, argues that, although most of the fire-resistant species are recorded in the early stages of succession, they do not persist long enough to make way for the miombo pyroclimax of Uapaca, Brachystegia and Julbernardia spp. In other instances, Stromgaard (1985) suggests a diminution of Uapaca, Brachystegia, Julbernardia and other associated species through the succession cycle. In these early discussions, however, the importance of mycorrhizae in vegetation succession of the miombo (Hogberg and Piearce, 1986) is not featured. 6. Discussion The importance of U. kirkiana arises from its role as a source of fruit towards the end of the dry season and during most of the rainy season when the food status is generally poor. The fruits have particular dietary value in time of famine. For Malawi and Zambia, the core of the distribution range, cottage wine industries have been established based on fruit from this species. Uapaca kirkiana is a tree of well-drained, infertile acidic soils at 500-2000 m altitude subject to a dry season lasting 5-7 months and unimodal annual rainfall of 500-1400 mm. Temperatures generally from 18-29°C are appropriate. Uapaca kirkiana apparently depends on symbiosis with ectomycorrhizal fungi. The species is adversely affected by bush fires late in the dry season. Despite the abundance of U. kirkiana in the miombo region, accelerating deforestation (Hyde and Seve, 1993) and the rarity of domesticated stands underline the need for active conservation. Around the many expanding centres of population within the natural range, extensive replacement of forest by arable and tree crops has been responsible for substantial reductions in Uapaca populations. Clarification of the ecological and geographical aspects of intra-specific variation should enable matching provenances and sites in field evaluation. 114 M.R. Ngulube et al. /Forest Ecology Appropriate morphology, growth and production will determine resource value at any locality, and provenance trials offer the best prospect for distinguishing inherited characteristics from those markedly affected by the environment. The mapping exercise has indicated a number of sources for genetic material likely to be of interest in conservation terms. An integral approach with detailed definitive studies generating data on the phenology, reproductive biology and breeding systems has been initiated to support planned provenance studies. The disjunctive Angolan and Zairean populations (Fig. 21, once considered as a separate species, U. benguelensis, are of obvious interest. Mwamba (1989b) and Hans (1980, 1981) draw attention to the variability of the fruits of U. kirkiana and implications for selection of improved accessions. Variation throughout the range emerges clearly from flora and entries of herbaria specimens from different parts of the range: height and diameter attained at maturity, leaf and fruit sizes. The frequent mention of the species in the literature promoting multipurpose trees, especially indigenous species bearing edible fruits (Fanshawe, 1972; Maghembe and Seyani, 1991; Maghembe et al., 1994; Grundy et al., 19931, anticipates formal management of U. kirkiana. Practical conservation on a local scale can be undertaken through management actions devised for sustained local resource use both at the sites where there are existing natural populations and in areas with no populations where scope for planting in association with prevailing land use systems exist. Effort made where natural populations exist must recognize the risk that the currently acceptable conservation status of U. kirkiana could quickly change if human and animal pressure on the miombo woodlands were to increase sharply. The woodlands can generally be viewed as a continuous phase of openaccess land: a common property resource within which there are isolated protected pockets under identifiable authority (forest and game reserves, national parks and nature sanctuaries, ranches and other estates). Only in such protected areas is early implementation of any innovative exploratory management policy feasible; a view widely held in southern Africa (Piearce and Gumbo, 1993). Nevertheless, socioeconomic studies within the range are increasingly revealing cultural mechanisms with strong pos- and Managemenf 77 (19951107-117 itive resource conservation impact (Makuku, 1993: Coote et al., 1993; Grundy et al., 1993). Progress in successful communication of management ideas through extension activities must take into account indigenous initiatives in the target areas. The eventual objective should be to extend practices of demonstrable value developed within the protected areas to the miombo woodlands in general, enhancing indigenous measures. Where a management authority is identifiable, the need is initially to ascertain the age structure of the U. kirk&w populations and monitor these over time. The immediate conservation objective should be to maintain or promote ideal stocking figures for optimum breeding and production of natural populations within the distribution range. Being dioecious, information on population sex ratios and the spatial distribution of the male and female mature individuals in a population is critical in planning and implementing active conservation programmes. Sites where populations of U. kirkiana existed should be rehabilitated through encouragement of natural regeneration. Initial strict fire control will promote regeneration as long as there are relic living root stocks or nearby seed sources. Failing this, enrichment planting can be undertaken. Extended periods of rigorous protection would be unnecessary if any burning was early. Total exclusion of fire is impracticable, but management should be able to maintain natural conditions and minimise fuel accumulation and fire risk through early controlled burn-ing. Enrichment planting is an option available where there is doubt as to the sufficiency of seed sources or relic individuals capable of responding positively to any fire control. Transplanting of nursery-raised seedlings of U. kirkiana is preferable to direct seeding which results in poor establishment success (Mwamba et al., 1992). Vegetative propagation is so far still undeveloped for U. kirkiana. Ex situ action should be concentrated within the range of Ii’. kirkiana, but in areas where the natural vegetation cover has been replaced by land-use options. In farms where relics of U. kirkiana populations remain, scope for incorporating the species in existing agroforestry systems exist. Of particular interest are spontaneous plantings or tending of seedlings from natural regeneration within the prevailing land-use M.R. Ngulube et al./Forest Ecology systems (Kwesiga and Chisumpa, 1992; Grundy et al., 1993; McGregor, 1994). Conscious selection of provenances displaying characteristics particularly appropriate for the role sought from the tree would be advantageous. Unlike most forest trees in breeding programmes, U. kirkiana is dioecious and this will have implications for breeding strategies. If U. kirkiana is bred for its fruit-it will be grown in a garden environment or any other agroforestry arrangement-the spatial arrangement of the male and female trees will be crucial. Hybridization may improve the attractiveness of U. kirkiuna as a tree crop. Breeding-improved varieties will however, depend upon a well-organized programme of provenance surveys and trials, as well as sufficient data on the breeding systems of the tree. Screening of range-wide provenances should assist with identification of suitable sources of seed or propagation and clonal material. A general interest in growth rate variations can be predicted, but of particular importance is an assessment linking fecundity and fruit yield with growth rate and/or tree morphology . Development of appropriate silvicultural procedures for domestication of U. kirkiunu has, however, received limited attention throughout the entire range. Formal silvicultural experiments investigating various aspects of tree development and subsequent effects upon yield will therefore be required to generate extension packages to guide successful conservation. Vegetative propagation procedures should receive special attention as this is likely to form the basis for future domestication of plants with high yields and may allow effective control of sex ratio within stands. The accumulation of the basic phenological data, upon which vegetative and other propagation procedures will depend, is of paramount importance. Investigations into mycorrhizal associations with U. kirkiunu and implications on conservation potential should also be addressed at the earliest possible stage. Acknowledgements This work was undertaken as part of the Support to FRIM Project funded by the Overseas Develop- and Management 77 (1995) 107-117 115 ment Administration of the United Kingdom of Great Britain. The support is gratefully acknowledged by the authors. The SADC (Southern Africa Development Community) Regional Tree Seed Centre Network Project funded by the Canadian International Development Agency (CIDA) provided partial support to visit some of the herbaria in southern Africa. We greatly appreciate this timely support. Assistance and advice from staff of the various herbaria visited in Britain and southern Africa, without which this work would have not been possible, is highly appreciated. We thank all those who contributed, in whatever way, to the success of this work. References Anonymous, 1951. Southern Rhodesia: Report of the Chief Conservator of Forests for the Year 1951. Salisbury, Rhodesia. Anonymous, 1955. Nyasaland: Annual Report of the Forestry Department for the Year Ending 31 December 1955. Govemment Printer, Zomba, 44 pp. Astle, W.L., 1968. The vegetation and soils of Chishinga Ranch, Luapula Province, Zambia. Kirkia, 7: 73-102. Burt&Davy, J. and Hoyle, A.C., 1936. Nyasaland Protectorate: Checklists of the Forest Trees and Shrubs of the British Empire. Forestry Institute, Oxford, pp. 15-20. Campbell, B.M., 1987. The use of wild fruit in Zimbabwe. Econ. Bot., 41: 375-385. Carr, W.R., 1957. Notes on some Southern Rhodesia indigenous fruits with particular reference to their ascorbic acid content. Food Res., 22: 590-593. Chidumayo, E.N., 1987. Woodland structure, destruction and conservation in Copperbelt area of Zambia. Biol. Conserv., 40: 89-100. Coote, H.C., Luhanga, J.M. and Lowore, J.D., 1993. Community use and management ofmdigenous forests in Malawi: the case of Chemba village forest area. Forestry Research Report No. 93006, pp. l-28. Drummond, R.B., 1981. The Common Trees of the Central Watershed Woodlands of Zimbabwe. Department of Natural Resources, Causeway, pp. 112-113. Ernst, W., 1975. Variation in mineral content of leaves of trees in miombo woodland in southern central Africa. J. Ecol., 63: 801-807. Fanshawe, D.B., 1968a. The vegetation of Zambia. Zambia Ministry of Rural Development, Forest Research Bulletin No. 7, pp. l-67. Fanshawe, D.B., 1968b. Fifty common trees of Zambia. Forest Department Bulletin No. 5, pp. l-96. Fanshawe, D.B., 1972. Useful Trees of Zambia for the Agriculturist. Government Printer, Lusaka, 26 pp. 116 M.R. Ngulube et al. /Forest Ecology Food and Agriculture Organization (FAO), 1983. Food and fruitbearing forest species, 1. Examples from eastern Africa. FAO For. Pap., 44/l: 1-172. Food and Agriculture Organization (FAO), 1990. Utilization of tropical foods: fruits and leaves. FAO Food Nutr. Pap., 47/7: l-32. Fox, F.W. and Young, M.E.N., 1982. Food from the Veld: Edible Wild Plants of Southern Africa. Delta, Craighal, pp. 196-197. Goldsmith, B. and Carter, D.T., 1981. The indigenous timbers of Zimbabwe. Zimbabwe Bull. For. Res., 9: 1-331. Grundy, LM., Campbell, B.M., Balebereho, S., Cunliffe, R., Tafangenyasha, C., Fergusson, R. and Parry, D., 1993. Availability and use of trees in Mutanda resettlement area, Zimbabwe. For. Ecoi. Manage., 56: 243-266. Hans, AS., 1980. Cytogenetics of Uapaca kirkiana Baill. Zambia National Council for Scientific Research, Tree Improvement Research Centre Paper No. 14, pp. 1-5. Hans, A.S., 1981. Uapaca kirkiana Muell. Arg. (Euphorbiacea). Musuku (Local name). Zambia National Council for Scientific Research, Tree Improvement Research Centre Technical Report, 48 pp. Hans, A.S., Parker, E.J. and Kumar, V., 1978. Potentialities of Zambian forest fruit trees. Paper presented to the 8th World Forestry Congress, Jakarta, 16-28 October, 6 pp. Hans, A.S. and Mwamba, C.K, 1982. Spatial relationships between male and female treesof Uapacu kirkiuna Muell. Arg. Zambia National Council for Scientific Research, Tree Improvement Research Centre Technical Report, 13 pp. Hogberg, P., 1982. Mycorrhizal associations in some woodland and forest trees and shmbsin Tanzania. New Phytol., 92: 407-415. Hogberg, P. and Piearce, G.D., 1986. Mycorrhizas in Zambian trees in relation to host taxonomy, vegetation type and succession patterns. J. Ecol., 74: 775-785. Hyde, W.F. and Seve, J.E., 1993. The economic role of wood products in tropical deforestation; the severe example of Malawi. For. Ecol. Manage., 57: 283-300. Jachmann, H., 1989. Food selection by elephants in the miombo biome in relation to leaf chemistry. Biochem. Syst. Ecol., 17: 15-24. Kikula, I.S., 1986. The influence of fire on the composition of miombo woodland of southwest Tanzania. Oikos, 46: 317-324. Kwesiga, F. and Chisumpa, S.M., 1992. Multipurpose trees of eastern Zambia: an ethnobotanical survey of their use in farming systems. International Centre for Research in Agroforestry. AFRENA Rep., 49: l-65. Lawton, R.M., 1964. The ecology of the Marquisia acuminata (Gilg.) R.E. Fr. evergreen forests and the related chipya vegetation types of north-eastern Rhodesia. J. Ecol., 52: 467479. Lawton, R.M., 1978. A study of the dynamic ecology of Zambian vegetation. J. Ecol., 66: 175-198. Lawton, R.M., 1980. Browse in miombo woodland. In: H.N. Le Hou&rou (Editor), Browse in Africa, the Current State of Knowledge. International Livestock Centre for Africa, Addis Ababa, pp. 25-31. and Management 77 (1995) 107-l I7 Lee, R.F., 1971. A preliminary annotated list of Malawi forest insects. Department of Forestry and Game. Malawi For. Res. Inst. Rec., 40: 1-132. Lovett, J., 1993. Uapaca kirkiana Muell. Arg. Family Euphorbiacea/Uapacacea. Miombo, 10: 3-4. Maghembe, J.A. and Seyani, J.H., 1991. Multipurpose trees used by smallholder farmers in Malawi: results of an ethnobotanical survey. International Centre for Research in Agroforestry. AFRENA Rep., 42: l-30. Maghembe, J.A., Kwesiga, F., Ngulube, M., Prins. H. and Malaya. F.M., 1994. Domestication potential of indigenous fruit trees of the miombo woodlands of southern Africa. In: R.R.B. Leaky and A.C. Newman (Editors), Tropicat Trees: Potential for Domestication and Rebuilding of Forest Resources. HMSO, London, pp. 220-229. Makuku, S.J., 1993. Community approaches in managing common property forest resources: the case of Norumedzo Community in Bikita, Zimbabwe. For. Trees People New&, 22: 18-23. Malaisse, F.P. and Parent, G., 1985. Edible wild vegetable products in the Zambezian woodland area: a nutritional and ecological approach. Bcol. Food Nutr., 18: 43-82. McGregor, J. 1994. Woodland pattern and structure in a peasant farming area of Zimbabwe: ecological determinants and present and past use. For. Ecol. Manage., 63: 97-133. Morris, B., 1987. Common Mushrooms of Malawi. FUNGIFLORA, Oslo, pp. 108. Mwamba, C.K., 1983. Ecology and distribution of Zambian wild fruit trees in relation to soil fertility: status of representative areas. M.Sc. Thesis, University of Ghent, 102 pp. Mwamba, C.K., 1989a. An outlook on the role of indigenous frnit trees in agroforestry. Paper presented at the First Agroforestry Workshop, 16-19 April 1989, Lusaka, Zambia, 13 pp. Mwamba, C.K., 1989b. Natural variation in fruits of Uapaca kirkiana in Zambia. For. Ecol. Manage., 26: 299-303. Mwamba, C.K., Zgambo, Y. and Chongo, G., 1992. Effect of seedling source on post-planting growth of Unpaca kirkiana Muell. Arg. South Afr. For. J., 161: 35-40. Palgrave, K.C., 1981. Trees of Southern Africa, 2nd edn. Strnik, Cape Town, 959 pp. Palgrave, 0.H and Palgrave, K.C., 1957. Trees of Central Africa. Rhodesia and Nyasaland, National Publications Trust. University Press, Glasgow, 466 pp. Pardy, A.A., 1951. Notes on indigenous trees and shrubs of Southern Rhodesia. Rhod. Agric. J., 48: 261-266. Parker, E.J., 1978. Causes of damage to Zambian wild fruit trees. Zambian J. Sci. Technol., 3: 74-83. Pegler, D.N., 1977. A preliminary agaric flora of Fast Africa. Kew Bull. Addit. Ser., 6: l-615. Pegler, D.N and Piearce, D., 1980. The edible mushrooms of Zambia. Kew Bull., 35: 475-491. Peters, C.R., O’Brien, E.M. and Drwnmond, R.B., 1992. Edible Wild Plants of Sub-saharan Africa. Royal Botanical Gardens, Kew. Piearce, G.D. and Gumbo, D.J. @ditors), 1993. The Ecology and Management of Indigenous Forests in Southern Africa. Pro- M.R. Ngulube et al/Forest Ecology ceedings of an International Symposium, Victoria Falls, Zimbabwe, 27-29 July 1992. The Forestry Commision, Harare, Zimbabwe (in collaboration with SAREC). Phipps, J.B. and Goodier, R., 1962. A preliminary account of the plant ecology of the Chimanimani mountains. J. Ecol., SO: 291-319. Pielou, EC., 1952. Notes on the vegetation of Rukwa Rift Valley, Tanganyika. J. Ecol., 40: 383-392. Pullinger, J.S. and Kitchin, A.M., 1982. Trees of Malawi. Blantyre Print, Blantyre, 229 pp. Rattray, J.M., 1961. Vegetation types of Southern Rhodesia. Kirkia, 2: 68-93. Rees, W.A., 1974. Preliminary studies into bush utilization by cattle in Zambia. J. Appl. Ecol., 11: 207-214. Saka, J.D.K., Msonthi, J.D. and Sambo, E Y., 1992. Dry matter acidity and ascorbic acid contents of edible wild fruits growing in Malawi. Trop. Sci., 32: 217-221. Saka, J.D.K. and Msonthi, J.D., 1994. Nutritional value of sixteen edible wild fruits growing in Malawi. For. Ecol. Manage., 64: 245-248. Savory, B.M., 1963. Site quality and tree root morphology in Northern Rhodesia. Rhod. J. Agric. Res., 1: 55-64. Seyani, J.H., 1991. Uapaca kirkiana: underutilized multipurpose tree species in Malawi worth some development. Paper presented at the International Workshop on the Biodiversity of Traditional and Underutilized Crops, 12-15 June 1991, Valletta, Malta, 10 pp. Shewry, P.R., Woolhouse, H.W. and Thompson, K., 1979. Relationships of vegetation to copper and cobalt in the copper clearings of Haut Shaba, Zaire. Bot. J. Linn. Sot., 79: l-36. and Management 77 (1995) 107-117 117 Shorter, C., 1989. An Introduction to the Common Trees of Malawi. The Wildlife Society of Malawi, Liiongwe, 115 pp. Storm, J.H., 1979. Know Your Trees: Some of the Common Trees Found in Zambia. Zambia Forestry Department, Ndola, 380 PP. Stromgaard, P., 1985. Biomass, growth and burning of woodlands in a shifting cultivation area of south central Africa. For. Ecol. Manage., 12: 163-178. Stromgaard, P., 1986. Early secondary succession on abandoned shifting cultivators’ plots in the miombo of south central Africa. Biotropica, 18: 97-106. Suti, N.A. and Kaputo, M.T., 1977. Identification of free sugars in musuku fruit (Uapaca kirkiana). Zambia J. Sci. Technol., 2: 23-25. Thoen, D. and Ba, A.M., 1989. Ectomycorrhizas and putative ectomycorrhizal fungi of Afielia africana Sm. and Uapaca guineensis Mull. Arg. in southern Senegal. New Phytol., 113: 549-559. Trapnell, C.G., 1959. Ecological results of woodland burning experiments in Northern Rhodesia. J. Ecol., 47: 129-168. Walker, B.H., 1980. A review of browse and its role in livestock production in southern Africa. In: H.N. Le Houerou (Editor), Browse in Africa, the Current State of Knowledge. Intemationanl Livestock Centre for Africa, Addis Ababa, pp. 8-12. Watson, J.P., 1964. A soil catena on granite in Southern RhodeSian. Rhod. J. Sci., 15: 238-257. White, F., 1983. The vegetation of Africa. UNESCO Nat. Resour. Res., 20: l-356. Willianson, J., 1975. Useful Plants of Malawi, revised edn. Montfort Press, Limbe, pp. 336.