SEARCH

SEARCH BY CITATION

Keywords:

  • Sinarundinaria alpina;
  • East African mountains;
  • forest disturbance by herbivores;
  • tropical montane forest;
  • vegetation zones

Abstract

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

Kilimanjaro has a large variety of forest types over an altitudinal range of 3000 m containing over 1200 vascular plant species. Montane Ocotea forests occur on the wet southern slope. Cassipourea and Juniperus forests grow on the dry northern slope. Subalpine Erica forests at 4100 m represent the highest elevation cloud forests in Africa. In contrast to this enormous biodiversity, the degree of endemism is low. However, forest relicts in the deepest valleys of the cultivated lower areas suggest that a rich forest flora inhabited Mt Kilimanjaro in the past, with restricted-range species otherwise only known from the Eastern Arc mountains. The low degree of endemism on Kilimanjaro may result from destruction of lower altitude forest rather than the relatively young age of the mountain. Another feature of the forests of Kilimanjaro is the absence of a bamboo zone, which occurs on all other tall mountains in East Africa with a similarly high rainfall. Sinarundinaria alpina stands are favoured by elephants and buffaloes. On Kilimanjaro these megaherbivores occur on the northern slopes, where it is too dry for a large bamboo zone to develop. They are excluded from the wet southern slope forests by topography and humans, who have cultivated the foothills for at least 2000 years. This interplay of biotic and abiotic factors could explain not only the lack of a bamboo zone on Kilimanjaro but also offers possible explanations for the patterns of diversity and endemism. Kilimanjaro's forests can therefore serve as a striking example of the large and long-lasting influence of both animals and humans on the African landscape.

Résumé

Le Kilimandjaro possède, sur un gradient d'altitude de 3000 mètres, une grande variété de types forestiers qui comprennent plus de 1200 espèces de plantes vasculaires. Les forêts de montagne àOcotea se trouvent sur le flanc sud, humide. Les forêts àCassipourea et àJuniperus poussent sur le versant nord, plus sec. Les forêts subalpines àErica situées à 4100 m représentent les plus hautes forêts néphéliphiles d'Afrique. Face à cette formidable biodiversité, le degré d'endémisme est faible. Cependant, des reliquats de forêts dans les plus profondes vallées des zones cultivées en basse altitude suggèrent qu'il y avait une riche flore forestière sur le Kilimandjaro dans des temps anciens, avec des espèces à la distribution restreinte que l'on ne connaît que sur les monts Eastern Arc. Le faible taux d'endémisme du Kilimandjaro pourrait être la conséquence de la destruction de la forêt de basse altitude plutôt que de l’âge relativement jeune de la montagne. Une caractéristique des forêts du Kilimandjaro est l'absence de zone de bambous, que l'on retrouve sur toutes les autres hautes montagnes d'Afrique de l'Est qui ont des fortes chutes de pluie comparables. Les tiges d'Sinarundinaria alpina sont appréciées par les éléphants et les buffles. Sur le Kilimandjaro, on trouve ces gros herbivores sur les flancs nord, trop secs pour qu'une grande zone de bambous puisse se développer. Ils sont exclus des forêts humides des flancs sud par la topographie et par les hommes qui cultivent le pied de la montagne depuis plus de 2000 ans. Ce mélange de facteurs biotiques et abiotiques pourrait expliquer non seulement l'absence d'une zone de bambous sur le Kilimandjaro, mais aussi, de plusieurs façons, les schémas de la diversité et de l'endémisme. Les forêts du Kilimandjaro peuvent servir d'exemple parfait de l'influence énorme et durable qu'ont eue les animaux et les hommes sur le paysage africain.


Introduction

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

Kilimanjaro offers a huge variety of forest types ranging from dry succulent forests on the foothills at 800 m to luxuriant montane rainforests with a wealth of epiphytes and ferns to the highest subalpine cloud forests of Africa at 4100 m. As one of the most famous mountains of the world, the flora and vegetation of Kilimanjaro are considered to be well investigated. However, a detailed overview over the forests of the whole mountain is still lacking. Until recently there were only more general and quite old descriptions of the forests (Volkens, 1897; Meyer, 1900; Jaeger, 1909; Geilinger, 1930; Schlieben, 1937), short notes (Greenway, 1965; Morris, 1970; Mwasaga, 1991; Lovett & Pócs, 1993) or studies restricted to small areas (Grimshaw, 1996) as well as floristic check-lists and keys (Steele, 1966a,b; Bigger, 1968; Lovett & Uronu, 1994). More detailed descriptions of the forest vegetation were presented between 2001 and 2005 (Hemp, 2001, 2002, 2005a). These studies have revealed that the southern slope forests are much richer in species than expected. In particular, ferns are very abundant because of high precipitation exceeding that of other East African tall mountains. Moreover, new observations of small gorge forest relicts inside the cultivated coffee–banana belt suggest that forests now replaced by cultivation might have been even richer, possibly comparable with the Eastern Arc mountains whose endemism has been related to age and ecoclimatic stability (Rodgers & Homewood, 1982; Lovett, 1988; Iversen, 1991; Fjeldsået al., 1997).

In addition to a huge altitudinal gradient, a high biodiversity and a low degree of endemism, another typical feature of the forests of Kilimanjaro is the absence of a bamboo zone. This is one of the great biogeographical mysteries of East Africa (Grimshaw, 1999). Bamboo (Sinarundinaria alpina (K. Schum.) C. S. Chao and Renvoize) occurs from 2400 to 3000 m, with isolated occurrences in favoured places between 1630 and 3200 m, over extensive areas of nearly all East African mountains, e.g. covering 65,000 ha on the Aberdares, 51,000 ha on the Mau escarpment and 39,000 ha on Mt Kenya (Hedberg, 1951; White, 1983). This prevalence led Hedberg (1951) to define a ‘bamboo zone’ as a characteristic feature of the montane forest belt on East African mountains. However, on Kilimanjaro Sinarundinaria alpina is inconspicuous and rare, but not completely missing (Volkens, 1897; Moreau, 1944; Greenway, 1965), while it forms a bamboo zone on the nearby volcano Mt Meru only 40 km away. The lack of a bamboo zone on Kilimanjaro was recognized by the first visiting scientists at the end of the 19th century (e.g. Volkens, 1897; Uhlig, 1904) but reasons were considered obscure (Greenway, 1965; White, 1983) or referred to a supposed drier climate (e.g. Hedberg, 1951; Hastenrath, 1973; Lind & Morrison, 1974; Shugart et al., 2001). More recently it has become clear that rainfall in the central southern forest zone is high (Hemp, 2001) and so this explanation becomes unlikely.

The aim of the present review was to give a phytosociological overview of the forest vegetation of Kilimanjaro with particular emphasis on the dynamic processes, which may explain the missing bamboo zone.

Forest zonation

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

Based on significant floristic discontinuities in the composition of the forests, the following forest zones can be distinguished on Kilimanjaro (Hemp, 2005a): a colline (=lowland) zone; submontane, lower and middle montane forest zones; an upper montane forest zone; and a subalpine forest zone. This terminology is similar to that used by Lovett (1993). Forests of the middle, upper montane and subalpine zone can be defined as ‘cloud forests’ on the basis of fog water input and structure (e.g. richness of epiphytes). Figure 1 shows the altitudinal zones and the main vegetation types viewed from the western side of the mountain demonstrating the contrast between the dry northern and wet southern slopes. The zonation is significantly correlated with temperature and soil acidity; and rainfall is of particular importance for zonation of epiphytes. Other key factors are humidity (influenced by stable cloud condensation belts) and minimum temperature (in particular, the occurrence of frost above 2700 m elevation; Hemp, 2005a).

image

Figure 1.  Schematic north–south profile showing the western slope of Mt Kilimanjaro (Shira, Kibo, Mawenzi) with main altitudinal zones and vegetation types. 1: colline (savanna) zone; 2: submontane zone with Croton–Calodendrum forest; a: coffee-banana plantations in the submontane and lower montane zone on the southern slope; b: lower montane gorge forests on the southern slope; 3: lower montane zone with Cassipourea forests on the northern slope and Agarista–Syzygium–Ocotea forests on the southern slope; 4: middle montane zone with Cassipourea forests on the northern slope and Ocotea forests on the southern slope; 5: upper montane (cloud forest) zone with Juniperus forests on the northern slope and Podocarpus-Ocotea forests on the southern slope; 6: upper montane (cloud forest) zone with Juniperus forests on the northern slope and Podocarpus forests on the southern slope; 7: subalpine heathlands (Erica bush); 8: lower alpine zone with Helichrysum cushion vegetation; 9: upper alpine and nival zone, mainly bare of vegetation (adapted from Hemp, 2005a)

Download figure to PowerPoint

Forest types

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

Natural forests of the submontane-subalpine zone cover an area of 1020 km2 (Table 1). Most of these forests are protected as forest reserve, 138 km2 lie above the reserve boundary at 2700 m inside the National Park. Owing to a huge altitudinal range of over 3000 m and the strong climatic contrast of the slopes, there is a large variety of different forest types. Table 2 presents species composition together with the structural and ecological parameters (e.g. height and cover of the vegetation layers, pH, precipitation) of the forests of Kilimanjaro based on over 600, 0.1 ha, forest plots (nomenclature follows Flora of Tropical East Africa: FTEA, 1952–2005; Beentje, 1994; Lovett et al., 2006; for details see Hemp, 2001, 2005a). In total, over 1200 vascular plant species were recorded, but only species with high frequency (except Sinarundinaria alpina) were included. Figure 2 shows the distribution of the main vegetation types on Kilimanjaro.

Table 1.   Area of the forest types inside the forest reserve and national park
Vegetation typeCommunity no. (Table 2)Area in the year 2000 (km2)
Remnants of subalpine Erica trimera forest21<1
Upper montane Erica excelsa forest1932
Upper montane Hagenia forest20107
Upper montane Podocarpus forest1760
Lower–upper montane Ocotea forest11, 12, 14, 15220
 Middle and upper montane Ocotea forest14, 15120
 Lower montane Ocotea forest11, 12100
Potential Ocotea forest (Ocotea stands over-exploited) 110
Upper montane Juniperus forest1840
Lower–middle montane Cassipourea forest6, 7, 9282
 Lower montane Cassipourea forest745
 Middle montane Cassipourea forest6162
 Lower montane Cassipourea–Ocotea forest975
Submontane Croton–Calodendrum forest (west and north)4, 545
 Submontane Croton–Calodendrum forest (west)427
 Submontane Croton–Calodendrum forest (north)518
Olea regeneration stages in Croton and Cassipourea forests 41
Lower–upper montane riverine forest13, 1667
Lower montane gorge and riverine forests815
Natural forest 1020
Clearing, meadow 36
Forest plantation 160
 Planted with trees 70
 Not planted with trees 90
Potential forest area 1216
Forest reserve 1078
Forest inside the national park 138
Table 2.   Percentage degree (constancy) table of the forest vegetation at Mt. KilimanjaroThumbnail image of Thumbnail image of Thumbnail image of Thumbnail image of Thumbnail image of
image

Figure 2.  Vegetation map of Kilimanjaro, based on the evaluation of about 1400 vegetation plots following the method of Braun-Blanquet (1964) and a supervised classification of Landsat ETM images taken on 29 January and 21 February 2000 (source: USGS/UNEP-GRID, Sioux Falls, SD, U.S.A.) using the software IDRISI 3.2 (Clark Labs, Worcester, MA, U.S.A.). Digital elevation model by Christian Lambrechts and Janet Akinyi Ong‘injo, Nairobi, based on toposheets at scale1/1:50,000. Red rectangle: location of the time series of Fig. 7

Download figure to PowerPoint

Very dry forests occur on parasitic volcanic vents in the colline savanna on the south-eastern slope. In the tree layer, succulents such as Euphorbia quinquecostata, E. candelabrum (Euphorbiaceae) and deciduous species such as Commiphora baluensis (Burseraceae) and Haplocoelum foliolosum (Sapindaceae) dominate (community 1).

In the cultivated submontane and lower montane zone of the southern slopes, the forest is restricted to deep valleys and gorges. These forests, although of very small extent (15 km2), are of great biogeographical and palaeobotanical importance. They differ completely in species composition and structure from the forests of higher altitudes, resembling instead the diverse wet montane forest of the Pare and Usambara mountains. Typical trees are Ekebergia capensis (Meliaceae, up to 60 m tall), Heinsenia diervilleoides and Hallea rubrostipulata (both Rubiaceae), Newtonia buchananii (Mimosaceae), Leptonychia usambarensis (Sterculiaceae), Strombosia scheffleri (Loganiaceae), Dasylepis integra (Flacourtiaceae), Garcinia sp. nov. (Clusiaceae) and Polyscias stuhlmannii (Araliaceae) (Fig. 3, communities 3 and 8).

image

Figure 3.  Profile (45 × 5 m) of a lower montane gorge forest at 1500 m a.s.l. on the southern slope (community 8; Table 2). Aa, Pouteria (=Aningeria) cf. altissima (death); Cb, Casearia battiscombei; Cp, Chassalia parvifolia; Ec, Ekebergia capensis; Ga, Garcinia sp. nov.; Hd, Heinsenia diervilleoides; Lb, Landolphia buchananii; Lu, Leptonychia usambarensis; Sg, Syzygium guineense; Ss, Strombosia scheffleri; Uh, Urera hypselodendron. Mean canopy height 40 m with some emergent trees (Ekebergia capensis) reaching heights of over 60 m. In the upper part of the profile lianas and shrubs form dense thickets under a gap in the tree canopy

Download figure to PowerPoint

In the lower and middle montane forests on the wet southern slopes, the dominant tree species is camphor (Ocotea usambarensis; Lauraceae). Camphor forests cover an area of about 220 km2. In the lower areas (1800–2200 m), Ocotea usambarensis is associated with Agarista salicifolia (Ericaceae), Macaranga capensis var. kilimandscharica (Euphorbiaceae), Syzygium guineense (Myrtaceae) and Polyscias fulva (communities 11 and 12). The middle montane zone (2200–2500 m) is the main habitat of camphor where pure stands exist. Here humidity reaches its maximum, indicated by the wealth of epiphytes and ferns, in particular filmy ferns and tree ferns (Hemp, 2001, 2002, 2006a) (community 14, Fig. 4). In the drier south-western part of the camphor belt, Ocotea forests with the tree Faurea wentzeliana (Proteaceae) exist at this altitude (community 10, not indicated in Fig. 2). In gorges and along streams Cornus volkensii (Cornaceae) is an important constituent of the tree layer (communities 13 and 16). In the upper montane zone, the gymnosperm Podocarpus latifolius (Podocarpaceae) starts to prevail, replacing Ocotea with increasing altitude between 2500 and 2800 m (community 15). Above, between 2800 and 3200 m, Podocarpus latifolius, Hagenia abyssinica (Rosaceae) and Prunus africana (Rosaceae) form the tree canopy (communities 17 and 20). Monodominant stands of Erica excelsa (Ericaceae) (community 19) also play an important role in this zone, replacing Podocarpus and Hagenia forests after fire (Hemp & Beck, 2001; Hemp, 2005b), forming the actual upper closed forest line at 3200 m. However, small remnants and burnt forests indicate that the upper closed forest line reached up to 3850 m recently (1997); and remnants of subalpine Erica trimera forests (community 21) mark the former and potential upper closed forest line at above 4000 m (Hemp, 2005b), representing today the highest elevation forests in Africa. The forests of the northern and western slopes are completely different in species composition and structure owing to lower precipitation. On the western slopes below 1600 m, and on the northern slopes below 2000 m, the relatively dry submontane forest is dominated by Olea europaea ssp. cuspidata (Oleaceae), Croton megalocarpus (Euphorbiaceae), Calodendrum capense (Rutaceae) and Diospyros abyssinica (Ebenaceae) (communities 4 and 5). The lower and middle montane forest types (1600–2500 m) on the eastern, northern and western slopes extending over 282 km2 are characterized by Cassipourea malosana (Rhizophoraceae), Vepris simplicifolia (Rutaceae), Fagaropsis angolensis (Rutaceae) and Olea capensis (communities 6 and 7). Community 9 represents an intermediate Cassipourea forest type on the south-western slope with species of Ocotea forests and the Flacourtiaceae Dasylepis integra in the tree layer. Similar forest types occur in the nearby Pare mountains. Between 2500 and 3100 m Juniperus procera (Juniperaceae), Podocarpus latifolius and Hagenia abyssinica are dominant tree species (community 18).

image

Figure 4.  Profile (a) (60 × 5 m) and ground plan (b) (60 × 10 m, edged in thick lines: area of the profile) of a middle montane Ocotea forest at 2200 m a.s.l. on the southern slope (community 14, Table 2), rich in vascular epiphytes and tree ferns. At, Aphloia theiformis; Co, Canthium oligocarpum; Cp, Chassalia parvifolia; Da, Dracaena afromontana; Es, Embelia schimperi; Gs, Galiniera saxifraga; Im, Ilex mitis; Lk, Lasianthus kilimandscharicus; Ou, Ocotea usambarensis; Pa, Pavetta abyssinica; Pc, Psychotria cyathicalyx; Pf, Psychotria fractinervata; Rm, Myrsine melanophloeos; Sm, Schefflera myriantha; Xm, Xymalos monospora. Crosses mark dead tree fern trunks. Mean tree height about 30 m with an upper canopy of 40 m built up by Ocotea usambarensis. The tree fern Cyathea manniana is a typical companion, playing an important role in the forest regeneration as is apparent from the pure Cyathea stand in the right part of the profile, where a large branch of the Ocotea on the right had fallen several years before causing a gap in the tree canopy

Download figure to PowerPoint

Forest dynamics and disturbance

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

During the last 100 years the forests of Kilimanjaro have experienced major changes in their extent and species composition.

Fire

Fire distribution on Kilimanjaro follows the precipitation regime. Regular fires occur every year in the colline savanna zone and in the upper montane and subalpine zone and to a lesser degree in the submontane and lower montane forest zone (Hemp, 2005c). As a result of increasingly drier climate, precipitation has decreased on Kilimanjaro by 30% in recent years, and under higher anthropogenic impact, fires have played an increasingly destructive role in the forests of Kilimanjaro during the last 100 years and in particular over the last three decades (Hemp, 2005b). During this time, Kilimanjaro has lost about 300 km2 of high altitude forests and the upper closed forest line was lowered by 900 m because of fire. As these forests have an important function for fog water collection, this has an impact on the water balance of the whole mountain (Hemp, 2005b). Most of these fires are anthropogenic, but natural fires do occur as well and old charcoal horizons in the soil suggest that fires have been occurring over a very long span of time, dating back several 10,000 years (Hemp & Zech, unpublished data), but would have been less frequent than today.

Fire causes sharp discontinuities in the composition and structure of the tall (20–30 m canopy) upper montane Hagenia–Podocarpus forests at 2800–3000 m. The giant heather Erica excelsa becomes dominant at this altitude forming dense monospecific stands of about 10 m height (Fig. 5), consisting of multi- and single-stemmed trees of apparently similar age, suggesting simultaneous sprouting after a fire (Hemp & Beck, 2001). During long periods of dry climate with recurrent fires, the Erica forest boundary moves downslope and advances upslope during wet periods. The presence of Erica enhances fire risk, as even fresh Erica wood burns well, which in turn prevents the Podocarpus forest from re-establishing (Fig. 6). At high fire frequency, the closed Erica excelsa forest degrades into open bushland of c. 1.5 m height dominated by E. trimera and E. arborea at elevations of between 3200 and 4000 m (the potential treeline). A high frequency of fires even destroys this bush, resulting in Helichrysum cushion vegetation, which is the climatic climax vegetation at altitudes above 4000 m. These dynamic fluctuations are shown in Fig. 7. Owing to the open canopy of the Erica forest, the microclimate changes as is obvious from the composition of the herb layer: montane forest species disappear and light-demanding species of the alpine flora such as representatives of the genera Helichrysum or Senecio become dominant, which cannot successfully compete in shady forests (Community 19, Table 2).

image

Figure 5.  Profile (16 × 2 m) of an upper montane Erica excelsa forest at 3000 m a.s.l. (community 19; Table 2). A, Anthospermum usambarense; As, Agarista salicifolia; P, Pittosporum sp.; trees and shrubs not labelled are Erica excelsa, crosses mark dead trunks. The multi-stemmed growth form of Erica indicates resprouting from stumps after fire (adapted from Hemp & Beck, 2001)

Download figure to PowerPoint

image

Figure 6.  Regeneration of the canopy in upper montane and subalpine forests. 1: Podocarpus forest, 2: Erica excelsa forest with montane forest species, 3: Pure Erica excelsa forest. Replacement of Podocarpus forest by an Erica forest is related to intensity and frequency of fire (black arrow). The higher the percentage of Erica, the higher is the risk of another fire and the more difficult is re-establishment of a Podocarpus forest (white arrows). As, Agarista salicifolia; Ee, Erica excelsa; Im, Ilex mitis; Pl, Podocarpus latifolius; Rm, Myrsine melanophloeos; Sv, Schefflera volkensii (adapted from Hemp & Beck, 2001)

Download figure to PowerPoint

image

Figure 7.  Land cover changes on North Kilimanjaro (location see Fig. 2) shown in a time series based on different Landsat images (source: USGS/UNEP-GRID, Sioux Falls, SD, U.S.A.). In 1976, upper montane Erica forests covered large areas. After fires these forests were replaced in 1984 by Erica bush and grasslands and Helichrysum cushion vegetation moved down-slope into areas of burnt Erica bush. At this time, logging of East African Cedar (Juniperus procera) had started in the montane forest belt and in 1993 most of the Cedar forest was cut, indicated in red. After the ban on commercial logging the forest started to regenerate in 2000 as did the Erica bush. However, compared with the situation in 1976, (sub-)alpine vegetation and the closed forest line in this area has moved down-slope by several hundred meters. (Photo by C. Lambrechts, adapted from Hemp, 2005b)

Download figure to PowerPoint

Burning in low altitude forests changes the species composition and the structure as well. About 10% of the tree species in the drier submontane Croton–Calodendrum forests of the western and northern slopes were found to be deciduous. This habit is an adaptation to dry seasons, and also possibly to recurring fires. In the same forest type, distinct fire-induced Olea europaea ssp. cuspidata dominance stages are quite common, covering 41 km2 (Table 1, Fig. 2). Similar to Olea, the fire-resistant trees Agarista salicifolia and Morella salicifolia (both with a thick, corky bark) are distributed in the fire-influenced forests of the lower montane zone (community 11) and the upper montane zone (communities 17–20; Hemp, 2005c).

Logging

In addition to fires devastating the forest there are other threats to this ecologically important vegetation zone. The results of an aerial survey (Lambrechts et al., 2002, Fig. 8) in combination with a ground survey revealed that the forests of Mt Kilimanjaro are heavily impacted by logging of indigenous trees in most areas below 2500 m elevation. In particular, the moist Ocotea forests that cover most of the southern slopes are undergoing serious destruction caused by the intensive illegal logging of camphor trees. On the eastern slopes, this overexploitation has resulted in forests free of mature Ocotea but still with the same structure and otherwise the same species composition. These ‘potential montane Ocotea forests’ (Fig. 2) cover an area of 110 km2. This means that one-third of the actual camphor zone is already depleted and the remaining part is being heavily modified. The situation is similar for the second most targeted tree species, Juniperus procera. Only 40 km2 of cedar forest of a potential area of about 120 km2 on the whole northern slope, today covered by Hagenia and Podocarpus forest, is left. However, most of the cedar forests were felled by legal sawmills up to the 1980s. Therefore, only a few recent illegal felling activities were observed on the northern slope during the aerial survey (Fig. 8), during which nearly 8000 cut trees were counted. These numbers approach the number of trees lost by heavy legal felling activities during World War II and the early 1980s (Hemp et al., in press).

image

Figure 8.  Illegal logging of indigenous trees in the forests of Kilimanjaro, documented during an aerial survey in 2001. Altogether nearly 8000 cut trees were counted during the flights (from Lambrechts et al., 2002). Figure reproduced with the kind permission of Unep, Nairobi.

Download figure to PowerPoint

The most common plants in larger (>0.5 ha) man-made forest clearings are the ferns Pteridium aquilinum (as an indicator for recurrent fires) and Hypolepis sparsisora (Hemp, 2001). Both may climb and thus completely cover the vegetation, and impede forest regeneration for several decades. At altitudes above 2100 m, the fern Histiopteris incisa is a typical species of natural and man-made clearings. Other typical widespread plants on clearings are Plectranthus albus, Thalictrum rhynchocarpum, Pycnostachys meyeri and Rubus steudneri. Smaller clearings with a more humid forest-like microclimate harbour the spectacular Lobelia giberroa as a characteristic constituent and here the tree fern Cyathea manniana forms a distinct regeneration stage inside the wet Ocotea forests (Fig. 4). Natural gaps in these forests are mainly caused by treefalls after heavy rains, when the epiphyte load becomes too heavy. Strong down winds in the evening, in particular on the lower slopes, sometimes cause large clearings as well. On the western and northern slope, large herbivores (elephants and buffaloes) play an important role in forest destruction; here, Urtica massaica is a typical plant in clearings. Large clearings on this side of the mountain are partly due to large herbivores activities and partly to fires (Wood, 1965a) and former sawmills (Fig. 2).

Overall forest loss and its impact

In addition to the losses of about 300 km2 of upper montane and subalpine forests from fire since 1880, losses on account of clear cutting of lower elevation forests amount to 450 km2 since 1929, bringing the total loss to c. 750 km2. Thus Kilimanjaro has lost about 50% of its former forest cover (Hemp et al., in press). Deforestation on mountain foothills raises mean cloud condensation level resulting in a gradual shrinking of the cloud zone. A similar effect is caused by global warming and drying of the air (Bruijnzeel, 2001). In addition to changes in the water balance of the mountain, loss of cloud cover may have added to the observed general decreasing trend in precipitation during the last century.

Forest plantations

About 15% of the indigenous forests have been converted into forest plantation areas on the north-western and northern slopes since 1950 (Fig. 2), mainly using fast-growing exotic tree species, such as pine (Pinus patula) and cypress (Cupressus lusitanica) (Wood, 1965b). These forest plantations are usually established by allowing local farmers to inter-crop annual agricultural crops (on Mt Kilimanjaro mainly potatoes, carrots and cabbage) with tree seedlings for the first years, an agro-forestry practice which is commonly called ‘Shamba system’ or ‘Taungya system’. The system, however, has not worked well as evidenced by the findings of the aerial survey undertaken in 2001 (Lambrechts et al., 2002). Over 50% of the ‘Shamba system’ areas were not planted with tree seedlings (Fig. 2, Table 1). In addition, some sixteen villages were found in these forest plantation areas.

Diversity and endemism

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

1223 plant species were recorded in the forest plots on Kilimanjaro (Hemp, 2005a). Seventeen per cent of the forest flora (206 species) are trees, 40% (485 species) occurred in the shrub layer, 15% (183 species) are lianas, 14% (172 species) epiphytes and 70% (858 species) were found in the herb layer. Many taxa occurred in different layers, so the sum of the percentages exceeds 100. These 1223 forest species represent about 40% of the vascular flora on Mt Kilimanjaro, which is about 3000 taxa (A. Hemp, unpubl. data). Therefore, the forest belt is the most important habitat for biodiversity on Mt Kilimanjaro, followed by grasslands and ruderal vegetation (Fig. 9).

image

Figure 9.  Vascular plant species richness in the main vegetation formations (as presented with constancy tables by Hemp, 2001) of Mt Kilimanjaro, based on the evaluation of about 1400 vegetation plots

Download figure to PowerPoint

The number of plant taxa is similar to the one given by Iversen (1991) for the Usambara mountains, one of the most famous biodiversity hotspots in Africa, which belong to the Eastern Arc mountains, a chain of ancient crystalline mountains in East Africa. The supposed lower biodiversity and the actual lower degree of endemism of Kilimanjaro are commonly explained by the higher age and greater ecoclimatic stability of the Eastern Arc mountains as suggested by e.g. Rodgers & Homewood (1982), Lovett (1988), Iversen (1991) or Fjeldsået al. (1997). However, in the cultivated submontane and lower montane zone of the southern slopes of Mt Kilimanjaro, the deepest valleys harbour forest relicts (community 8, Table 2) that are of great biogeographical and palaeobotanical importance. Many of the species in these forests were previously not known from Kilimanjaro and some of them were believed to be endemic to the Eastern Arc mountains: Polyscias stuhlmannii, Syzygium micklethwaitii (=S. sclerophyllum; Myrtaceae), Psychotria crassipetala, P. alsophila (Rubiaceae), Millettia oblata (Fabaceae), Impatiens serpens (Balsaminaceae), Dasylepis integra and Leptonychia usambarensis. Others are even endemic or nearly endemic to these small gorges: Garcinia sp. nov., a huge tree of 40 m (see Fig. 3), Rinorea sp. nov (Violaceae), a shrub species, Euphorbia sp. aff. obovalifolia, a tree of over 30 m, the recently described Lellingeria paucipinnata, a lithophytic fern on boulders in streams (Hemp, 1997, 2006a) and Brucea macrocarpa (Simaroubaceae). The occurrence of such species suggests a rich forest flora inhabited the lower areas of the southern slopes of Mt Kilimanjaro in former times. As humans have continuously cultivated the lower slopes of Mt Kilimanjaro for at least the last 2000 years (Schmidt, 1989), it can be assumed that many forest species were extirpated together with the forest cover. Thus, the lower degree of endemism of Kilimanjaro can be explained by the wide destruction of the lower montane forest rather than the younger age of the mountain. Occurrence of several fern relicts in these forests (Adiantum reniforme, Trichomanes radicans, T. rigidum, Pteris tripartita, Lomariopsis warneckei, Coniogramme africana, Asplenium inaequilaterale, Sticherus flagellaris; Hemp, 2002) leads to the same conclusion. This is corroborated by the fact that forest species such as insects of the group Saltatoria (grasshoppers, locusts and katydids), which are affected less by forest devastation, have similar numbers of endemic forest species in the submontane and montane zone on Mt Kilimanjaro (including Mt Meru) and the East Usambara mountains. Endemic grasshopper species such as Ixalidium sjöstedti, Parepistaurus deses and Altiusambilla modicicrus have coped with the habitat change from forest to plantations (Hemp & Hemp, 2003; C. Hemp, 2005).

In addition to high rainfall and a large number of habitats, the high biodiversity of Kilimanjaro's forests is at least partly because of the lack of a bamboo zone. In place of monospecific bamboo stands with low species numbers, highly diverse upper montane Ocotea and Podocarpus forests cover large areas, which support high numbers of epiphytes and pteridophytes (Hemp, 2002; Figs 10 and 11). In particular, epiphytes suffer from lack of complex stratification, sparse ramification and from the smooth bark of bamboo.

image

Figure 10.  Bamboo forest on the wet south-eastern slope of Mt Kenya at 2750 m. Because of structural parameters such monodominant stands are low in species numbers in general and of epiphytes and woody plants in particular. In this site, only eleven species (no epiphytes, no pteridophytes and only one woody plant) were found

Download figure to PowerPoint

image

Figure 11. Podocarpus forest on the wet southern slope of Mt Kilimanjaro, similar altitude (2750 m) and climate conditions as in Fig. 10. In this structurally complex forest relevé 46 species (beside others, twelve epiphytes, seventeen pteridophytes and twelve woody plants) occurred

Download figure to PowerPoint

Potential reasons for the lack of a bamboo zone on Kilimanjaro

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

The following section reviews potential hypotheses for the missing bamboo zone on Kilimanjaro and offers a new explanation.

Climate

Because of lack of data, rainfall distribution on Mt Kilimanjaro was unclear until recently (Hemp, 2001). My own measurements, which include the first rainfall data from the high precipitation area inside the southern forest belt, show that annual rainfall on the central southern slope increases to about 1900 mm at 1400 m and to about 2700 mm at 2200 m in the lower part of the forest belt, thus exceeding the precipitation on other East African high mountains. At higher elevation, precipitation declines, reaching 80% of the maximum at 2400 m, 70% at 2700 m, 50% near the upper forest border at 3000 m and only 20% at 4000 m (Hemp, 2005a; Fig. 12). These data are in contrast to the more general earlier reports that Mt Kilimanjaro is drier than other East African high mountains (Hedberg, 1951; Hastenrath, 1973; Lind & Morrison, 1974; Richter, 1996). Species richness and abundance of epiphytes, filmy ferns, tree ferns and ferns are obvious signs of high humidity (Hemp, 2001, 2002, 2006a). In most areas of the southern slope, rainfall far exceeds the 1250 mm a year which is the lower limit for Sinarundinaria alpina (Dale, 1940; Lind & Morrison, 1974). One small stand of Sinarundinaria included in forest community 2 (Table 2) received slightly less than this limit (1200 mm year−1), which is possibly compensated by a comparatively wet riverine situation. On the other hand, there is no evidence that Sinarundinaria alpina could be excluded by high precipitation and if so, all areas above 2600 m receive <2000 mm, which is similar to the climatic situation of other bamboo stands such as those on Mt Elgon (Masiga, Esegu & Mwambu, 2001). Temperature has also been cited as a limiting factor, with bamboo forests confined to areas with a mean annual temperature of 10–15.9°C (Winiger, 1979; Schmitt, 1991). On Kilimanjaro Sinarundinaria was even found in areas with a mean annual temperature of 20°C, such as in the above-mentioned riverine forest. The required temperature range (10–20°C) prevails on the southern slope between 1000 and 2900 m (Table 2), including the largest part of the forest belt.

image

Figure 12.  Rainfall (data from Hemp et al., in press) and distribution of large herbivores on Kilimanjaro. Areas of high precipitation (≥1300 mm) providing suitable abiotic growing conditions and occurrence of megaherbivores favouring growth and propagation of Sinarundinaria alpina are mutually exclusive, explaining the lack of a bamboo belt on Kilimanjaro

Download figure to PowerPoint

Geological and soil constraints

Although there is no closed bamboo zone on Kilimanjaro, this does not mean that Sinarundinaria alpina is missing completely. The species was encountered during the field work several times in small stands in riverine habitats in the cultivated submontane and colline zone of the southern slope (Fig. 13, Table 2). Here, it may be partly syanthropic, although some stands were already reported 100 years ago by Volkens (1897) and Widenmann (1899). However, there is at least one natural patch of 6 ha in the forests of the northern slope (Moreau, 1940; Grimshaw, 1999). Grimshaw (1999) suggested that geological constraints could exclude Sinarundinaria alpina from the southern slope, as the bamboo stand on the northern slope occurs on phonolite (alkaline lava), which is missing on the southern slope. This explanation is not very convincing as phonolite occurs in vast areas of the southern slope as well (Downie & Wilkinson, 1972; A. Hemp, pers. obs.). Furthermore, Sinarundinaria alpina grows well on nonvolcanic mountains such as the Udzungwa, Ulugurus, Nguru (Lovett, 1994) and the Ruwenzori mountains. On the southern slope, Sinarundinaria alpina thrives very well on nonphonolitic bedrocks, as long as conditions are open: a 0.5-m piece of the rhizome, planted by the author at 1400 m and receiving about 2000 mm of annual precipitation, produced about 50 shoots of 8–10 m after 5 years. The large range of the measured pH values on the southern slope (between about pH 7 and pH 3; Table 2) makes it likely that Sinarundinaria alpina could find a niche on this part of the mountain. Other soil characteristics (texture, structure and depth) were similar to the bamboo stands investigated by Masiga et al. (2001) on Mt Elgon with a high clay loam texture and generally deep profiles. Nutrient availability and soil moisture of the prevailing volcanic Andosols on Kilimanjaro should be similar to those of the same soil types on other volcanoes with a bamboo belt; however, these parameters were not measured during this study and need further investigation.

image

Figure 13.  Vegetation cover, occurrence of elephants and buffaloes (edged in black) and records of Sinarundinaria alpina (black dots) on Mt Kilimanjaro; record of bamboo on the northern slope according to Moreau (1944) and Grimshaw (1999). Note the exclusive distribution of large herbivores in the montane forests (green) and the Chagga home gardens of the densely populated submontane and lower montane coffee–banana belt (brown), through which no crossing for big animals is possible. Very deep and steep gorges on the south-western slope (Kikafu river system below Shira, indicated by arrows) and on the north-eastern slope (Great and Little Barranco below Mawenzi) restrict large herbivores from migrating inside the forest belt into the forests of the southern slope

Download figure to PowerPoint

Successional stages and vegetation history of Sinarundinaria alpinaon Kilimanjaro

Lovett (1994) suggests that the erratic flowering and subsequent die-back of large stands of Sinarundinaria alpina may account for its absence from Kilimanjaro, in other words Kilimanjaro may only be in a successional stage without bamboo. However, the cyclic sequences of Sinarundinaria alpina stands – besides lasting normally not several hundred years but decades – include a distinct stage with Sambucus africana (Agnew 1965), a companion of Sinarundinaria, which significantly is missing on Kilimanjaro as well. Furthermore, rhizomes of Sinarundinaria alpina are able to persist for several decades in a stage of dormancy without sprouting (E. Beck, pers. comm.). However, in fifteen soil profiles (varying from 1 to >3 m depth) along an altitudinal transect between 1400 and 3200 m on the central southern slope, no remnants of rhizomes were found by the author. Pollen analyses would be the best method to bring light onto the vegetational history of Kilimanjaro; here, further work is needed to detect a possible ancient bamboo belt.

Sinarundinaria and site disturbance

Sinarundinaria alpina has reached the mountain and is growing successfully, and so, migratory, climatic, soil and geological constraints and possible successional dynamics can be ruled out as the factors responsible for the missing bamboo zone on Kilimanjaro. Observations on other East African mountains showed that the occurrence of bamboo is linked to a special type of disturbance: the activity of large herbivores. Buffaloes and elephants are living in large numbers in the bamboo zones of Mt Kenya, Mt Meru and the Aberdares. In all investigated plots with Sinarundinaria, obvious signs (paths and droppings) of the activity of buffaloes and elephants were found, which bend and pull up old bamboo shoots and dig the soil. This propagates bamboo from fallen culms and fragmentized parts of the rhizomes, enhancing vegetative propagation (Agnew, 1985). This suggestion is corroborated by Banana & Tweheyo (2001), who observed that bamboo is being replaced by hardwood trees in Echuya forest in Uganda after elephants and buffaloes had become extinct by 1960 as a result of hunting. There is considerable evidence that Sinarundinaria alpina is a light-demanding pioneer species that benefits generally from disturbance, not only the influence by megaherbivores, but also especially from fires (Lebrun, 1960; Glover & Trump, 1970, cited in White, 1983; Masiga et al., 2001) or human activities (Hamilton & Perrott, 1981; Marchant & Taylor, 1998). On Mt Kilimanjaro, large herbivores occur only on the western and northern slope (Fig. 13), whereas they are missing on the southern slope. This may be owing to several factors. The Kitendeni corridor which connects the elephant and buffalo population in the forests with the game population of the Amboseli National Park in Kenya is on the drier northern slopes (Grimshaw & Foley, 1991; Blanc et al., 2003). On the wetter southern and eastern slopes, an upward migration of large herbivores is now no longer possible through the densely populated submontane coffee–banana belt, which covers an area of about 1000 km2 (Hemp, 2006b; see Fig. 13). Even the surrounding former savanna areas are now cultivated and covered by human settlements and not populated by large herbivores since the 1960s. According to ethnographic studies by Widenmann (1899) 50,000–60,000 Chagga people were estimated to have lived on Mt Kilimanjaro in 1895. In 2002, the census counted 1,053,204 people (National Bureau of Statistics, Central Census Office, 2003). The human population has multiplied 20 times since 1895.

From studies on Mt Kenya (Vanleeuwe & Lambrechts, 1999) it is known that elephants climb slopes only up to a steepness of about 30°. On the south-western and north-eastern slopes of Kilimanjaro, very deep (up to several 100 m) and very steep (>30°) valleys exist (Figs 13 and 14), which reach high up into the alpine zone. These deep gorges prevent large herbivores migrating from the northern side of the mountain to the southern. Combined with human occupation of the wetter slopes, this means that the southern and south-eastern montane forests of Mt Kilimanjaro are no longer accessible for buffaloes and elephants. From early descriptions (e.g. Widenmann, 1899; Volkens, 1897; Jaeger, 1909), when the savanna on the southern foothills was still intact and not yet settled by humans, it is known that elephants lived in the forests there. However, Volkens, who intensively explored Kilimanjaro's landscape between 1893 and 1895, did not see a single elephant and similarly Widenmann stated that there were few in the forests of Kilimanjaro compared with the adjacent Mt Meru (where a bamboo zone exists). Both the authors reported that the Chagga people hunted elephants and even at this time the human population impeded the elephants from migrating inside the large forest block between the deep inaccessible gorges of Kikafu and Weru-Weru rivers on the southern slope as observed by Jaeger (1909) (see Figs 13 and 14). As the southern slopes of Kilimanjaro have been continuously populated by humans for at least 2000 years (Schmidt, 1989), population density of large herbivores in this area was therefore probably comparatively low for a very long time.

image

Figure 14.  The very deep gorges of the Kikafu river system are inaccessible for large herbivores. These gorges reach from the cultivated submontane zone far into the alpine zone

Download figure to PowerPoint

The actual influence of elephants and buffaloes on forest structure and composition was observed during the ground studies and aerial survey of the threats to Kilimanjaro's forests in the year 2001 (Lambrechts et al., 2002). Large herbivores can change huge forest areas creating a mosaic of clearings, open forest stands and close forest patches thereby creating ideal conditions for the light-demanding bamboo. The ecological reason for the lack of a bamboo zone on the northern side of the mountain is clearly the low precipitation, rather than the lack of herbivores, as the rainfall is less than about 1100 mm year−1 (Fig. 12), below the critical amount of 1250 mm. This is similar to the situation on Mt Kenya, where the bamboo zone is restricted to the wet south-eastern slope but is absent from the drier northern slope. Similar coincidences between the occurrence of megaherbivores, bamboo zones and climate are obvious on many other mountains in East Africa as well (Fig. 15).

image

Figure 15.  Occurrence of bamboo zones, megaherbivores and areas of high annual precipitation (>1250 mm) on several mountains in East Africa. A bamboo zone occurs only on wet slopes, which are inhabited by elephants and buffaloes

Download figure to PowerPoint

Based on these observations, bamboo seems to be favoured by the disturbance caused by big animals and not endangered by them as suggested by Agnew (1985). This author observed extensive trampling and grazing of young shoots of Sinarundinaria in the Aberdares and was concerned about the future of bamboo. As long as population densities of large herbivores in the montane forests of East Africa are as high as they are today, it is probable that the bamboo zones will even increase but not shrink, a suggestion which is corroborated by the direct observations of Banana & Tweheyo (2001) in Uganda. The montane forests of Kilimanjaro's southern slope appear to be climatically and edaphically suitable for Sinarundinaria alpina. However, the biotic habitat factors, especially site preparation by large herbivores, are today found only on the northern slopes, which are too dry.

This interplay of biotic and abiotic factors possibly not only explains the lack of a bamboo zone on Kilimanjaro but also on the adjacent Pare and Usambara mountains, which have a similar settlement history and steep slopes; and perhaps the general distribution of bamboo zones in East Africa (Fig. 15). Furthermore, land use history also offers possible explanations for levels of diversity and endemism. Kilimanjaro's forests can therefore serve as a striking example of the large and long-lasting biotic influence of both animals and humans on the African landscape.

Acknowledgements

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References

I gratefully acknowledge grants by the Deutsche Forschungsgemeinschaft and the Tanzanian Commission for Science and Technology for permitting research. For the support in getting permits, I owe gratitude to the Chief Park Wardens of Kilimanjaro National Park, Mr Lomi Ole Moirana and Mr Nyamakumbati Mafuru, to the Catchment Forest officers and to my counterpart Mr Jacob Mushi (Tanzania Association of Foresters), Moshi. I further thank the keepers of the East African Herbarium, Nairobi (Kenya) Dr Siro Masinde and Kew Herbarium, England, Prof. Dr Owens for permission to study their collections, Quentin Luke, Simon Mathenge (both Nairobi) and Dr Bernard Verdcourt (Kew) for help in identifying difficult species, Christian Lambrechts (Nairobi) for assistance in GIS analyses and Dr Jon Lovett and Dr Rob Marchant (both York) for valuable comments on the manuscript.

References

  1. Top of page
  2. AbstractRésumé
  3. Introduction
  4. Forest zonation
  5. Forest types
  6. Forest dynamics and disturbance
  7. Diversity and endemism
  8. Potential reasons for the lack of a bamboo zone on Kilimanjaro
  9. Acknowledgements
  10. References
  • Agnew, A.D.Q. (1985) Cyclic sequences of vegetation in the plant communities in the Aberdare Mountains, Kenya. J. E. Afr. Nat. Hist. Soc. Nat. Mus. 75, 112.
  • Banana, A.Y. & Tweheyo, M. (2001) The ecological changes of Echuya afromontane bamboo forest, Uganda. Afr. J. Ecol. 39, 366373.
  • Beentje, J. (1994) Kenya Trees, Shrubs and Lianas. National Museums of Kenya, Nairobi.
  • Bigger, M. (1968). A Check List of the Flora of Kilimanjaro. Coffee Research Station, Lyamungu.
  • Blanc, J.J., Thouless, C.R., Hart, J.A., Dublin, H.T., Douglas-Hamilton, I., Craig, C.G. & Barnes, R.F.W. (2003) African Elephant Status Report 2002: an Update from the African Elephant Database. IUCN/SSC African Elephant Specialist Group. IUCN, Gland, Switzerland, Cambridge, U.K.
  • Braun-Blanquet, J. (1964) Pflanzensoziologie. Springer, Wien.
  • Bruijnzeel, L.A. (2001) Hydrology of tropical montane cloud forests: a reassessment. Land Use Water Resour. Res. 1, 1.11.18 (http://www.luwrr.com).
  • Dale, I.R. (1940) The forest types of Mount Elgon. J. E. Afr. Nat. Hist. Soc. Nat. Mus. 36, 7482.
  • Downie, C & Wilkinson, P. (1972) The Geology of Kilimanjaro. Geological Survey of Tanzania and Department of Geography, Sheffield University, Sheffield, U.K.
  • Fjeldså, J., Ehrlich, D., Lambin, E. & Prins, E. (1997) Are biodiversity ‘hotspots’ correlated with current ecoclimatic stability? A pilot study using the Noaa-AVHRR remote sensing data. Biodivers. Conserv. 6, 401422.
  • FTEA (19522005) Flora of Tropical East Africa. Royal Botanic Garden, Kew.
  • Geilinger, W. (1930) Der Kilimandjaro. Sein Land und seine Menschen. Verlag Hans Huber, Bern, Berlin.
  • Glover, P.E. & Trump, E.C. (1970) An Ecological Survey of the Narok District of Kenya Masailand, Pt 2. Vegetation. Kenya National Parks, Nairobi.
  • Greenway, P.J. (1965) The vegetation and flora of Mount Kilimanjaro. Tangan. Notes Rec. 64, 97107.
  • Grimshaw, J.M. (1996) Aspects of the ecology and biogeography of the forest of the northern slope of Mt Kilimanjaro, Tanzania. Thesis, Oxford University, Oxford.
  • Grimshaw, J.M. (1999) The afromontane bamboo, Yushania alpina, on Kilimanjaro. J. E. Afr. Nat. Hist. 88, 7983.
  • Grimshaw, J.M. & Foley, C.A.H. (1991) Kilimanjaro Elephant Project 1990 Final Report. Friends of Conservation, London.
  • Hamilton, A.C. & Perrott, R.A. (1981) A study of altitudinal zonation in the montane forest belt of Mt. Elgon, Kenya/Uganda. Vegetatio 45, 107125.
  • Hastenrath, S. (1973) Observations on the periglacial morphology of Mts. Kenya and Kilimanjaro, East Africa. Z. Geomorphol. Neue Folge 16, 161179.
  • Hedberg, O. (1951) Vegetation belts of the East African mountains. Svensk Bot. Tidskrift 45, 140202.
  • Hemp, A. (1997) New fern records for Kilimanjaro. J. East Afr. Nat. Hist. 86, 3742.
  • Hemp, A. (2001) Ecology of the pteridophytes on the southern slopes of Mt. Kilimanjaro. Part II: Habitat selection. Plant Biol. 3, 493523.
  • Hemp, A. (2002) Ecology of the pteridophytes on the southern slopes of Mt. Kilimanjaro. Part I: Altitudinal distribution. Plant Ecol. 159, 211239.
  • Hemp, A. (2005a) Continuum or zonation? Altitudinal gradients in the forest vegetation of Mt. Kilimanjaro. Plant Ecol. 184(1), 2742.
  • Hemp, A. (2005b) Climate change-driven forest fires marginalize the impact of ice cap wasting on Kilimanjaro. Global Change Biol. 11, 10131023.
  • Hemp, A. (2005c) The impact of fire on diversity, structure and composition of Mt. Kilimanjaro's vegetation. In: Land Use Change and Mountain Biodiversity (Eds E.Spehn, M.Liberman and C.Körner). CRC Press, Boca Raton, FL.
  • Hemp, A. (2006a) Ecology and altitudinal zonation of pteridophytes on Mt. Kilimanjaro. In: African Plants: Biodiversity, Ecology, Phytogeography and Taxonomy (Eds S. A.Ghazanfar and H.Beentje). Royal Botanic Gardens, Kew.
  • Hemp, A. (2006b) The banana forests of Kilimanjaro: biodiversity and conservation of the Chagga home gardens. Biodiv. Conserv. 15(4), 11931217.
  • Hemp, A. & Beck, E. (2001) Erica excelsa as a component of Mt. Kilimanjaro's forests. Phytocoenologia 31, 449475.
  • Hemp, A., Lambrechts, C. & Hemp, C. (in press). Global Trends and Africa. The Case of Mt. Kilimanjaro. UNEP, Nairobi.
  • Hemp, C. (2005) The Chagga home gardens – relict areas for endemic Saltatoria species (Insecta: Orthoptera) on Mount Kilimanjaro. Biol. Conserv. 125, 203210.
  • Hemp, C. & Hemp, A. (2003) Saltatoria coenoses of high-altitude grasslands on Mt. Kilimanjaro, Tanzania (Orthoptera: Saltatoria). Ecotropica 9, 7197.
  • Iversen, S.T. (1991) The Usambara Mountains, NE Tanzania: phytogeography of the vascular plant flora. Symb. Bot. Ups. XXIX, 1234.
  • Jaeger, F. (1909) Forschungen in den Hochregionen des Kilimandscharo. Mitt. Dtsch. Schutzgebieten 22, 113146; 161–196.
  • Lambrechts, C., Woodley, B., Hemp, A., Hemp, C. & Nnyiti, P. (2002) Aerial Survey of the Threats to Mt. Kilimanjaro Forests. UNDP, Dar es Salaam.
  • Lebrun, J. (1960) Sur une methode de délimitation des horizons et étages de végétation des montagnes du Congo oriental. Bull. Jardin Bot. l’État à Bruxelles 30, 7594.
  • Lind, E.M. & Morrison, M.E.S. (1974) East African Vegetation. Longman, London.
  • Lovett, J.C. (1988) Endemism and affinities of the Tanzanian montane forest flora. Monogr. Syst. Missouri Bot. Gard. 25, 591598.
  • Lovett, J.C. (1993) Eastern Arc moist forest flora. In: Biogeography and Ecology of the Rain Forests of Eastern Africa (Eds J. C.Lovett and S. K.Wasser). Cambridge University Press, Cambridge.
  • Lovett, J.C. (1994) Notes on moist forest bamboos and bambusoid grasses in eastern Tanzania. E. A. N. H. S. Bull. 24, 25.
  • Lovett, J.C. & Pócs, T. (1993) Assessment of the Condition of the Catchment Forest Reserves. Catchment Forestry Report 93.3. Forest and Beekeeping Division, Ministry of Tourism, Natural Resources and Environment, Dar es Salaam.
  • Lovett, J.C. & Uronu, L.O.N. (1994) Field Guide to the Forest Trees of Kilimanjaro. Botanical Museum, Copenhagen.
  • Lovett, J.C., Ruffo, C.K., Gereau, R.E. & Taplin, J.R.D. (2006) Field Guide to the Moist Forest Trees of Tanzania. The Society for Environmental Exploration, London and Dar es Salaam.
  • Marchant, R.A. & Taylor, D.M. (1998) A late Holocene record of montane forest dynamics from south-western Uganda. The Holocene 8, 375381.
  • Masiga, W.C., Esegu, J.F.O. & Mwambu, G.C.M. (2001) Arundinaria alpina in Mount Elgon National Park, Uganda. Bamboo Sci. Cult. 15, 813.
  • Meyer, H. (1900) Der Kilimandscharo. Reisen und Studien. Reimer, Berlin.
  • Moreau, R.E. (1944) Kilimanjaro and Mount Kenya: some comparisons, with special reference to the mammals and birds; and with a note on Mount Meru. Tanganyika Notes Rec. 18, 2859.
  • Morris, B. (1970) The zonal vegetation of Kilimanjaro. Afr. Wildl. 24, 157166.
  • Mwasaga, B.C. (1991) The natural forest of Mount Kilimanjaro. In: The Conservation of Mount Kilimanjaro (Ed. W. D.Newmark). IUCN, Gland, Switzerland and Cambridge, U.K.
  • National Bureau of Statistics, Central Census Office (2003) 2002 Population and Housing Census. General Report. The United Republic of Tanzania, Dar es Salaam.
  • Richter, M. (1996) Klimatologische und pflanzenmorphologische Vertikalgradienten in Hochgebirgen. Erdkunde 50, 205237.
  • Rodgers, W.A. & Homewood, K.M. (1982). Species richness and endemism in the Usambara mountain forests, Tanzania. Biol. J. Linn. Soc. 18, 197242.
  • Schlieben, H.J. (1937) Aus der Pflanzenwelt des Kilimanjaro. Gartenflora 86, 135138, 163–165, 186–188, 207–211.
  • Schmidt, P.R. (1989) 8. Early exploitation and settlement in the Usambara Mountains. In: Forest Conservation in the East Usambara Mountains Tanzania (Eds A. C.Hamilton and R.Bensted-Smith). IUCN, Gland, Switzerland and Cambridge, U.K.
  • Schmitt, K. (1991) The Vegetation of the Aberdare National Park Kenya. Universitätsverlag Wagner, Innsbruck.
  • Shugart, H.H., French, N.H.F., Kasischke, E.S., Slawski, J.J., Dull, C.W., Shuchman, R.A. & Mwangi, J. (2001) Detection of vegetation change using reconnaissance imagery. Global Change Biol. 7, 247252.
  • Steele, R. C. (1966a), A checklist of the trees and shrubs of the south Kilimanjaro forest, Part 1. Tangan. Notes Rec. 65, 67.
  • Steele, R.C. (1966b) A check-list of the trees and shrubs of the south Kilimanjaro forests, Part 2. Tangan. Notes Rec. 66, 183186.
  • Uhlig, C. (1904) Vom Kilimandscharo zum Meru. Zeitschr. Ges. Erdk. Berlin 9, 627718.
  • Vanleeuwe, H. & Lambrechts, C. (1999) Human activities on Mount Kenya from an elephant's perspective. Pachyderm 27, 6973.
  • Volkens, G. (1897) Der Kilimandscharo. Darstellung der allgemeineren Ergebnisse eines fünfzehnmonatigen Aufenthalts im Dschaggalande. Reimer, Berlin.
  • White, F. (1983) The vegetation of Africa. Natural Resources Research, UNESCO, Paris.
  • Widenmann, A. (1899) Die Kilimandscharo-Bevölkerung. Anthropologisches und Ethnographisches aus dem Dschaggalande. Petermanns Geogr. Mitt. 129, 1104.
  • Winiger, M. (1979) Bodentemperaturen und Niederschlag als Indikatoren einer klimatisch-ökologischen Gliederung tropischer Gebirgsräume. Methodische Aspekte und Anwendbarkeit diskutiert am Beispiel des Mt. Kenya (Ostafrika). Geomethodica 4, 121150.
  • Wood, P.J. (1965a) The forest glades of West Kilimanjaro. Tangan. Notes Rec. 64, 108111.
  • Wood, P.J. (1965b) A note on forestry on Kilimanjaro. Tangan. Notes Rec. 64, 111114.