Content uploaded by Hemp Andreas
Author content
All content in this area was uploaded by Hemp Andreas on Apr 14, 2019
Content may be subject to copyright.
Climate change and its impact on the forests
of Kilimanjaro
Andreas Hemp*
Ecological Botanical Gardens, University of Bayreuth, Universita¨tsstr. 30, 95440 Bayreuth, Germany
Abstract
Cloud forests are of great importance in the hydrological
functioning of watersheds in subhumid East Africa. How-
ever, the montane forests of Mt. Kilimanjaro are heavily
threatened by global change impacts. Based on an evalu-
ation of over 1500 vegetation plots and interpretation of
satellite imagery from 1976 and 2000, land-cover changes
on Kilimanjaro were evaluated and their impact on the
water balance estimated. While the vanishing glaciers of
Kilimanjaro attract broad interest, the associated increase
of frequency and intensity of fires on the slopes of Kili-
manjaro is less conspicuous but ecologically far more sig-
nificant. These climate change-induced fires have lead to
changes in species composition and structure of the forests
and to a downward shift of the upper forest line by several
hundred metres. During the last 70 years, Kilimanjaro has
lost nearly one-third of its forest cover, in the upper areas
caused by fire, on the lower forest border mainly caused by
clearing. The loss of 150 km
2
of cloud forest – the most
effective source in the upper montane and subalpine fog
interception zone – caused by fire during the last three
decades means a considerable reduction in water yield. In
contrast to common belief, global warming does not nec-
essarily cause upward migration of plants and animals. On
Kilimanjaro the opposite trend is under way, with conse-
quences more harmful than those due to the loss of the
showy ice cap of Africa’s highest mountain.
Key words: cloud forest, forest fires, global change
Introduction
The shrinking ice cap of Kilimanjaro is a well-known
phenomenon since over 100 years (Meyer, 1900; Jaeger,
1909; Klute, 1920; Thompson et al., 2002). The globally
synchronous fluctuations of mountain glaciers in modem
times lead to the assumption that their causes are also of a
global scale (Kaser, 1999), partly referred to a recent
warming of the tropical middle troposphere (Hastenrath &
Kruss, 1992; Diaz & Graham, 1996; Kaser, 1999; Gafffen
et al., 2000; Thompson, 2000; Irion, 2001) or to a pro-
nounced decrease of precipitation and air humidity (Kaser
et al., 2004). During the last 120 years, annual precipi-
tation on Mount Kilimanjaro (Tanzania) has decreased by
600–1200 mm (Hemp, 2005a) while since 1976 tem-
peratures have increased drastically (Altmann et al.,
2002). Both these climatic changes do not only cause the
glaciers on the mountain to retreat but also fires to become
more aggressive and devastating on the higher slopes of
the mountain, resulting in a potentially much greater
impact on the overall Kilimanjaro ecosystem than that of
the melting glaciers.
Although most of these fires are lit by the carelessness of
humans (e.g. honey collectors or poachers: Hemp & Beck,
2001), they would not be so devastating had the climate
not become drier. Whilst the lower and middle montane
forests on Kilimanjaro are heavily threatened by illegal
logging (Lambrechts et al., 2002), the drier upper montane
and subalpine zones in particular are strongly influenced
by fire. Fires change species composition and structure of
the forests (Hemp, 2005b) impacting the Kilimanjaro
ecosystem to a far greater extent than the melting of its
glaciers. The aim of this study is to give an overview about
these climate change impacts on the forests of Kilimanjaro
and to assess the fire-caused land cover changes with their
respective impact on the water balance of Kilimanjaro.
Study area, materials and methods
Study area
Mt. Kilimanjaro is located 300 km south of the equator in
Tanzania on the border with Kenya between 245¢and
*Correspondence: E-mail: andreas.hemp@uni-bayreuth.de
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10 3
325¢S and 370¢and 3743¢E. It represents an eroded relic
of an ancient volcano with three peaks (Shira, Mawenzi
and Kibo) rising from the savanna plains at 700-m eleva-
tion to a snow- and ice-clad summit of 5895-m altitude. Its
diameter from northwest to southeast is about 90 km.
Due to its equatorial location, two distinct rainy seasons
occur on Kilimanjaro: the long rains from March to May,
and the short rains around November. According to the
climate classification system of Ko¨ppen and Troll ⁄Pfaffen
(in Mu
¨ller, 1983), Mt. Kilimanjaro belongs to the zone of a
seasonal dry tropical climate. However, rainfall and tem-
perature vary with altitude and exposure to the dominant
wind from the Indian Ocean. The northern slopes, on the
lee side of the mountain, receive much less annual rainfall
than the southern slopes. Mean annual temperature
decreases linearly upslope with a lapse rate of 0.56C per
100 m (Hemp, 2006a) starting with 23.4C at the foothills
in Moshi (813 m; Walter, Harnickell & Mueller-Dombois,
1975), and decreasing to )7.1C at the top of Kibo
(Thompson et al., 2002). Annual precipitation reaches its
maximum in the mid-montane zone of the southern slope,
where c. 3000 mm was recorded at 2200 m. 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, 2001, 2006a).
Several bioclimatic belts can be distinguished along the
slopes of Mt. Kilimanjaro (Fig. 1). A dry and hot colline
savanna zone surrounds the mountain base between 700
and 1000 m a.s.1. (mostly farmland; some intact savanna
vegetation left around Lake Chala and Ngare Nairobi).
The submontane and lower montane zone between
1000 and 1800 m has been converted to coffee–banana
plantations. Montane forests cover an area of about
1000 km
2
on Mt. Kilimanjaro. In the lower and middle
parts of the southern slope, these forests are characterized
by Ocotea usambarensis. The cloud forest zone is dominated
by Podocarpus latifolius,Hagenia abyssinica and Erica excelsa.
On the drier northern slope, the lower forest zone is
dominated by Croton-Calodendrum forests, mid-altitudes are
dominated by Cassipourea-forests, and Juniperus character-
izes the higher altitudes. Above c. 3100 m these forests
have been replaced by Erica bush in recent decades (Erica
arborea and Erica trimera,Protea caffra and Euryops dacry-
dioides).Around 3900m elevation the Erica bush grades
into Helichrysum cushion vegetation with Helichrysum
Fig 1 Land use and vegetation on Mt. Kilimanjaro. Dotted lines represent transects
4Andreas Hemp
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
newii and Helichrysum citrispinum reaching c. 4500 m.
Higher altitudes are poorly vegetated. For a more detailed
description of these vegetation types, see Hemp (2001,
2006b,c) and Hemp & Hemp (2003).
Vegetation analysis
Between 1996 and 2005, over 1500 plots were examined
along 34 elevational transects using the method of Braun-
Blanquet (1964) (Fig. 1). Based on these plant species
inventories and a classification of a Landsat ETM image
from 2000 (idrisi 3.2 software, Clark Labs, Worcester, MK,
USA), a vegetation map was produced (Hemp, 2006c),
which was compared with a Landsat MSS image from 1976
to evaluate changes in vegetation cover between these two
years.
Climate analysis
Since 1997, rainfall and temperature were recorded across
five elevational gradients (for details see Hemp, 2006a).
Long-term rainfall trends were analysed using data of the
Fig 2 Annual precipitation at (a) Moshi
Meteorological Station, 830 m a.s.l., (b)
Kilema Mission, 1430 m a.s.l. and (c)
Kibosho Mission 1430 m a.s.l. on the
southern slope of Kilimanjaro. Data
source: Tanzania Meteorological Agency,
Kibosho and Kilema Mission
Climate change impacts on Kilimanjaro 5
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
three oldest and permanent weather stations on the
mountain (provided by the Tanzania Meterological Agen-
cy) using linear regressions (Fig. 2).
Water balance
Based on the vegetation map, the altitudinal forest zona-
tion and rainfall data, the Kilimanjaro forest belt was
divided into ten eco-climatic zones. For each of these zones,
annual rain water input, evapotranspiration and fog water
input were estimated (for details see Hemp, 2005a).
Results
Evaluation of long-term rainfall data
Figure 2 shows the mean annual precipitation from three
different stations located on the southern slope of Kili-
manjaro. Using a linear regression, rainfall of the three
stations decreased by 34%, 27% and 39% (i.e. by 392, 532
and 986 mm per annum ) for the periods 1902–2004,
1911–2004 and 1922–2004 respectively.
Elevational gradient in plant species richness
As the altitudinal zonation on the wet southern and dry
northern slope differs significantly (cf. Hemp, 2006a), the
following explanations focus on the southern slope with its
larger altitudinal gradient. Numbers of 2000 vascular
plants species (including 143 pteridophytes) at 100-m
elevation intervals on the southern slope of Mt. Kilimanj-
aro are shown in Fig. 3.
Vascular plants reach a maximum of over 900 species at
1000 m in the mosaic-type interface between the colline
savanna and the banana–coffee plantations of the sub-
montane zone, supporting the well-known phenomenon
that plant species numbers are peaking in moderately
cultivated or disturbed areas and not in natural, com-
pletely untouched areas. Remarkably, there is a second
peak at 2600 m at the upper border of the montane zone.
It is this altitude where fires start to become important on
Mt. Kilimanjaro, creating a mosaic of different fire-induced
successional stages of forest, shrub and tussock grassland
stands. By numerical evaluation of the releve
´s, several
plant associations were revealed (Hemp, 2002, partly
published in Hemp, 2001, 2006c). The numbers of these
plant associations in different altitudes show a clear cor-
relation with the species numbers (Fig. 3), suggesting that
the number of habitats, the beta diversity (as indicated by
the number of plant communities) is a major controlling
factor of species richness on Mt. Kilimanjaro. This means
that high community (beta) diversity in the fire-influenced
areas of the upper montane zone (twelve communities)
leads to a higher species (alpha) diversity (311 species) as
compared to the closed undisturbed forest at lower alti-
tudes with 233 species in seven communities and the
monotonous Erica bush at higher altitudes with about 50
species in four to five communities.
In particular, the distribution of pteridophytes illustrates
these patterns very well. Because of their species richness
and continuous distribution, this vascular plant group is
an excellent tool for recognizing the altitudinal zonation of
tropical mountains (Hemp, 2001). In Zambia, geophytic
and deciduous fern species indicate dry conditions with
recurring fires (Kornas
´, 1978, 1985). Similar observations
were made on Kilimanjaro with such ferns occurring in
colline ⁄submontane and subalpine grasslands (Hemp,
2001). Geophytes and deciduous ferns are of lower
Fig 3 Altitudinal changes of species
numbers of pteridophytes and of all occur-
ring vascular plants on the southern slope
of Mt. Kilimanjaro in relation to number of
plant associations (habitats), based on the
evaluation of 1270 plots with about 2000
species. The second diversity peak is lo-
cated in the fire-influenced altitudes in the
upper montane and subalpine zone
6Andreas Hemp
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
importance in the mid altitude forests than in the fire-
affected lower and higher forest areas. This is reflected by
the bimodal discontinuous distribution pattern of decidu-
ous, fire tolerating fern species such as the geophyte
Pteridium aquilinum and the hemicryptophyte Dryopteris
pentheri which have a distribution gap in the central forest
zone (Hemp, 2001, 2005b).
Influence of fire on low altitude forests
Burning in low altitude forests changes species composi-
tion and structure. About 10% of the tree species in the
drier submontane Croton-Calodendrum forests of the wes-
tern 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 km
2
(Hemp, 2006c). 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 and
the upper montane zone.
Impact of fire on the upper vegetation zones
Induced by the elevational drop of precipitation above the
major cloud zone, fire causes a natural sharp discontinuity
in composition and structure of the tall (20–30 m high)
subalpine Hagenia-Podocarpus forests at 2800–3000 m
a.s.l. The giant heather Erica excelsa becomes dominant at
this altitude forming dense monospecific stands of about
10 m height (Hemp & Beck, 2001). The occurrence of an
Erica forest is an obvious fire sign. During long periods of
dry climate with recurrent fires, the Erica forest boundary
moves downslope and advances upslope during wet peri-
ods. The presence of Erica enhances the fire risk, as even
fresh Erica wood burns well, which in turn prevents the
Podocarpus forest from re-establishing. At high fire fre-
quency, the closed Erica excelsa forest degrades into open
bushland of c. 1.5 m height dominated by Erica trimera
and Erica arborea between 3200 and 4000 m a.s.l. (the
potential treeline). Continuously, high frequency of fires
even destroys this bush, resulting in Helichrysum cushion
vegetation, which is the climatic climax vegetation at
altitudes above 4000 m.
The comparison of 1976 and 2000 Landsat images
reveals enormous changes in the upper vegetation zones of
Kilimanjaro during the last 24 years (Table 1). In 1976,
the Erica trimera bush – currently depressed below 3800 m
– reached 4100 m, in part forming a continuous belt in
areas which have become covered by Helichrysum cushion
vegetation since then. That Helichrysum shrub is too sparse
and low in biomass to fuel the spreading of fires and
increased its area more than three times between 1976
and 2000. In 1976, closed Erica forests covered nearly six
times the area of today (187 instead of 32 km
2
), extending
up to 3800 m in many places. This means nearly 15% of
Kilimanjaro’s forest cover was destroyed by fire since 1976
and was replaced by Erica bush which extended its total
area by 50 km
2
(mainly downslope).
Impacts of increasing fires and melting glaciers on the water
balance
The forests of Kilimanjaro above 1300 m a.s.l. m receive
nearly 1600 million m
3
water annually, 95% by rainfall
and c. 5% by fog interception (‘Rain’ and ‘Fog’, respectively
in Table 2). About 500 million m
3
of water (31%) perco-
lates into the groundwater or into streams.
If one assumes that fog precipitation is close to zero once
the forest is destroyed, the loss of 150 km
2
of upper
montane and subalpine forests since 1976 corresponds to
an estimated loss of 20 million m
3
of fog water deposition
per year. This is more than a quarter of the estimated
annual fog water yield of the whole forest belt or equiva-
lent to the annual water demand of the one million
Table 1 Landcover changes 1976–2000
Vegetation type 1976 (ha) 2000 (ha) Change (%)
Montane forest 106,624 97,410 )9
Subalpine Erica
cloud forest
18,657 3212 )83
Erica bush 20,166 25,675 27
Helichrysum cushion
vegetation
6928 21,755 214
Grassland 9039 4396 )51
Table 2 Hydrometrical data of the forest belt on Mt. Kilimanjaro
(million m
3
)
Water input Water output
Rain fog Evapotranspiration
Groundwater
and streams
1.509 73 1085 497
95% 5% 69% 31%
Climate change impacts on Kilimanjaro 7
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
inhabitants on Kilimanjaro (according to numbers given
by United Republic of Tanzania and CES, 2002). In con-
trast, the average annual water output of the 2.6 km
2
of
glaciers can be estimated at only 1 million m
3
(5%) (Hemp,
2005a).
Discussion
Evidence for climate change linked with fire activity and glacier
retreat
Long-term rainfall data from three different stations
located on the southern slope of Kilimanjaro show a clear
decrease in annual precipitation during the last century
(Fig. 2).
Temperature data, since 1976, from Amboseli on the
northern foothills of Kilimanjaro reveal a drastic increase of
mean daily maximum temperature at a rate of over 2 K per
decade (Altmann et al., 2002). Increases were greatest
during the hottest months, February and March. According
to the climate trends in the Kilimanjaro area presented by
Hay et al. (2002) for 1941–1995, temperatures rose
between 1951 and 1960, and between 1981 and 1995, but
were stable or decreased slightly during the remaining
intervals.
Prior to 1900, only few sporadic rainfall data are
available from Kilimanjaro. Hence, the historical fluctua-
tions must be reconstructed using proxy indicators, such
as lake levels. Over the past millennium, equatorial east
Africa has alternated between contrasting climate condi-
tions, with significantly drier climate than today during
the ‘Medieval Warm Period’ (c. 1000–1270 ad) and a
relatively wet climate during the so-called ‘Little Ice Age’
(c. 1270–1850 ad; Verschuren, Laird & Cumming, 2000).
Enhanced solar radiation due to diminished cloud cover,
accompanied by reduced precipitation during the last two
decades of the 19th century caused a drop of lake levels
and glacier recession (Hastenrath, 1984, 2001; Nicholson,
2000;Verschuren et al., 2000; Nicholson & Yin, 2001).
The drastic drop of the water level of Lake Victoria be-
tween 1880 and 1900 had been attributed to the reduc-
tion in annual precipitation by about 150–200 mm
(Hastenrath, 1984).
During the last century, mean annual precipitation on
Kilimanjaro decreased by about 400–1000 mm (Fig. 2).
When compared with the situation before 1880, the drop
amounts to 600–1200 mm per annum. Consistent with
the pronounced decrease of precipitation at the end of the
19th century, fires in subalpine forests have been reported
already by the first Europeans on the mountain (e.g.
Meyer, 1890, 1900; Volkens, 1897; Jaeger, 1909) and in
parallel the glaciers started to recede (Kaser et al., 2004)
but then entered a few decades of more stable conditions
(Kaser, 1999). Since 1976, the warming effect became
more significant (Altmann et al., 2002) and also led to a
second phase of rapid melting and strongly enhanced fire
activity.
Impact of fire on the upper tree-line
Fire not only transforms land cover, but also maintains
certain land cover types (Eva & Lambin, 2000). On Kili-
manjaro, fire influences species diversity and vegetation
structure in the different altitudinal zones in various ways
(Hemp, 2001, 2002, 2005b; Hemp & Beck, 2001) where
in the upper forest and shrub zone fires play a destructive
role. From field observations and historical descriptions
(Jaeger, 1909; Klute, 1920), it can be assumed that dif-
ferent tall subalpine forest types extended up to over
4000 m in many areas of Kilimanjaro at the beginning of
the last century. This is corroborated by the existence of
forest relicts at 4000 m that represent the highest cloud
forests in Africa (Hemp, 2006a). The fires of the years
1996 and 1997 destroyed vast areas of old Erica forest,
which became replaced by low stature Erica bush and
moved the upper closed forest line downslope by about
800 m of altitude. Thus, fire effects overrun potential
positive effects of climatic changes in the upper forest belt.
Fire and movements of vegetation belts
Pollen diagrams of East African mountains show that the
treeline has never been stable. They indicate a dry climate
between 25000 and 12500 bp (Lind & Morrison, 1974)
with charcoal horizons in today’s lower montane forests
on Kilimanjaro, suggesting an extent of ericaceous forest
1000 m below its current lower limit (Hemp & Beck,
2001).
Alpine vegetation has been found to migrate upslope due
to rising temperatures and enrichment of atmospheric CO
2
and nitrogen in temperate zone mountains (in the case of
pioneer species) (Grabherr & Pauli, 1994), or remain more
or less stable (in the case of late successional communities)
(Ko¨rner, 2003). In contrast, alpine vegetation on Kili-
manjaro migrates downslope, replacing subalpine forests.
This is due to a typical feature of high mountains in Africa:
8Andreas Hemp
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
the vast ericaceous belt. This plant formation is very
inflammable and becomes more susceptible to fire in a
warmer climate.
In addition to the losses of 150 km
2
of upper montane
and subalpine forests from fire since 1976, losses due to
clear cutting of lower elevation forests amount to 450 km
2
since 1929, bringing the total loss to c. 600 km
2
. Thus
Kilimanjaro has lost about one-third of its former forest
cover (Hemp, Lambrechts & Hemp, in press). This sum
does not account for the massive logging inside the still
existing forest belt as documented during an aerial survey
in 2001, when about 8000 newly cut trunks had been
counted (Lambrechts et al., 2002).
Impact on the water balance
Cloud forests are of great importance for watersheds in East
Africa (Po
´cs, 1976). In addition to the function of filtering
and storing water, the upper montane and subalpine cloud
forests have a high potential of collecting cloud water (cf. e.g.
Cavelier & Goldstein, 1989; Juvik & Nullet, 1993; Cavelier,
Solis & Jaramillo, 1996; Bruijnzeel, 2001). Fog interception
increases with altitude, and so does its contribution to water
yielding on Kilimanjaro (Hemp, 2005a). Thus, the loss of
cloud forests due to climate induced fires as well as the loss of
montane forest due to clearing, causes a considerable
reduction and enhanced variability of water yields of the
Kilimanjaro catchments. Deforestation on mountain foot-
hills raises mean cloud condensation level that results 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.
These trends affect over one million people living on the
mountain, by far exceeding the hydrological consequences
of the loss of the glaciers.
Conclusions
Global warming combined with increasing levels of carbon
dioxide and nitrogen is expected to lead to an upslope shift of
the various vegetation zones. On Mt. Kilimanjaro, however,
climate change induced fires had an overriding impact
leading to a downslope shift of the (sub-)alpine vegetation
belts within few decades. During the last 80 years, Kili-
manjaro has lost nearly a third of its forest cover. Compared
with these landscape changes, the hydrological significance
of the melting of the glaciers is almost negligible. However,
the disappearance of the glaciers is an alarming indicator of
the substantial changes in the Kilimanjaro environment. At
current rates of incidence, fires will have extinct most of the
Kilimanjaro’s high altitude forests within the next few
years, and with this, the mountain will have lost its most
effective water source in the fog interception zone. With its
glaciers, Kilimanjaro will lose a part of its beauty and an
important archive of palaeoclimatic records (Thompson,
2000), with its forests, it loses its major ecosystem service to
a water demanding society.
Acknowledgements
Field work was carried out with financial support by the
Deutsche Forschungsgemeinschaft and permission of
Tanzanian Commission for Science and Technology. For
further support, I owe gratitude to the Chief Park Wardens
of Kilimanjaro National Park, Lomi Ole Moirana and
Nyamakumbati Mafuru, to my research counterpart Jacob
Mushi, Moshi and to Christian Lambrechts, Nairobi. I am
grateful to Prof. Frederick Kayanja, Kampala for comments
on the manuscript.
Conflicts of interest
The author declares no conflicts of interest.
References
Altmann, J., Alberts, S.C., Altmann, S.A. & Roy, S.B. (2002)
Dramatic change in local climate patterns in the Amboseli
basin, Kenya. Afr. J. Ecol. 40, 248–251.
Braun-Blanquet, J. (1964) Pflanzensoziologie. Springer, Wien.
Bruijnzeel, L.A. (2001) Hydrology of tropical montane cloud
forests: a reassessment. Land Use and Water Resour. Res. 1,
1.1–1.18 Available at: http://www.luwrr.com(last accessed on
5 August 2008).
Cavelier,J.&Goldstein, G. (1989) Mist and fog interception in
elfin cloud forests in Columbia and Venezuela. J. Trop. Ecol. 5,
309–322.
Cavelier, J., Solis,D.&Jaramillo, M.A. (1996) Fog interception in
montane forests across the Central Cordillera of Panama
´.J. Trop.
Ecol. 12, 357–369.
Diaz, H.F. & Graham, N.E. (1996) Recent changes in tropical
freezing heights and the role of sea surface temperature. Nature
383, 152–155.
Eva,H.&Lambin, E.F. (2000) Fires and land-cover change in the
tropics: a remote sensing analysis at the landscape scale. J.
Biogeogr. 27, 765–776.
Climate change impacts on Kilimanjaro 9
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
Gaffen, D.J., Santer, B.D., Boyle, J.S., Christy, J.R., Graham, N.E. &
Ross, R.J. (2000) Multidecadal changes in the vertical
temperature structure of the tropical troposphere. Science 287,
1242–1245.
Grabherr,G.&Pauli, H. (1994) Climate effects on mountain
plants. Nature 369, 448.
Hastenrath, S. (1984) The Glaciers of Equatorial East Africa. Solid
Earth Sciences Library, Reidel, Dordrecht, Boston, Lancaster.
Hastenrath, S. (2001) Variations of East African climate during
the past two centuries. Clim. Change 50, 209–217.
Hastenrath,S.&Kruss, P.D. (1992) The dramatic retreat of
Mount Kenya’s glaciers between 1963 and 1987: greenhouse
forcing. Ann. Glaciol. 16, 127–133.
Hay, S.I., Cox, J., Rogers, D.J., Randolphs, S.E., Stern, D.I., Shanks,
G.D., Myers, M.F. & Snow, R.W. (2002) Climate change and the
resurgence of malaria in the East African highlands. Nature
415, 905–909.
Hemp, A. (2001) Ecology of the pteridophytes on the southern
slopes of Mt. Kilimanjaro. Part II: Habitat selection. Plant Biol. 3,
493–523.
Hemp, A. (2002) Ecology of the pteridophytes on the southern
slopes of Mt. Kilimanjaro. Part I: Altitudinal distribution. Plant
Ecol. 159, 211–239.
Hemp, A. (2005a) Climate change driven forest fires marginalizes
the ice cap wasting on Mt. Kilimanjaro. Glob. Change Biol. 11,
1013–1023.
Hemp, A. (2005b) The impact of fire on diversity, structure and
composition of Mt. Kilimanjaro’s vegetation. In: Land Use
Changes and Mountain Biodiversity (Eds E. Spehn,M.Liberman
and C. Ko
¨rner). CRC Press LLC, Boca Raton, FL, U.S.A.
Hemp, A. (2006a) Continuum or zonation? Altitudinal gradients in
the forest vegetation of Mt. Kilimanjaro. Plant Ecol. 184, 27–42.
Hemp, A. (2006b) The banana forests of Kilimanjaro: biodiversity
and conservation of the Chagga home gardens. Biodivers. Con-
serv. 15, 1193–1217.
Hemp, A. (2006c) Vegetation of Kilimanjaro: hidden endemics and
missing bamboo. Afr. J. Ecol. 44, 305–328.
Hemp,A.&Beck, E. (2001) Erica excelsa as a fire-tolerating com-
ponent of Kilimanjaro’s forests. Phytocoenologia 31, 449–475.
Hemp,C.&Hemp, A. (2003) Saltatoria coenoses of high-altitude
grasslands on Mt. Kilimanjaro, Tanzania (Orthoptera: Saltato-
ria). Ecotropica 9, 71–97.
Hemp, A., Lambrechts,C.&Hemp, C. (in press) Global Trends &
Africa. The Case of Mt. Kilimanjaro. UNEP, Nairobi.
Irion, R. (2001) The melting snows of KiIimanjaro. Science 291,
1690–1691.
Jaeger, F. (1909) Forschungen in den Hochregionen des Kil-
imandscharo. Mitt. Dtsch. Schutzgebieten 22, 113–146; 161–196.
Juvik, J.O. & Nullet, D. (1993) Relationships between rainfall,
cloud-water interception, and canopy throughfall in a Hawaian
montane forest. In: Tropical Montane Cloud Forests. Proceedings of
an International Symposium at San Juan, Puerto Rico, 31 May–5
June 1993 (Eds L. S. Hamilton,J.O.Juvik and F. N. Scatena).
Springer, Berlin, Germany.
Kaser, G. (1999) A review of the modern fluctuations of tropical
glaciers. Glob. Planet. Change 22, 93–103.
Kaser, G., Hardy, D.R., Mo
¨lg, T., Bradley, R.S. & Hyera, T.M.
(2004) Modern glacier retreat on Kilimanjaro as evidence of
climate change: observations and facts. Int. J. Climatol. 24,
329–339.
Klute, F. (1920) Ergebnisse der Forschungen am Kilimandscharo
1912. Reimer, Berlin.
Kornas
´, J. (1978) Fire-resistance in the pteridophytes of Zambia.
Fern Gaz. 11, 373–384.
Kornas
´, J. (1985) Adaptive strategies of African pteridophytes to
extreme environments. Proc. R. Soc. Edinb. 86B, 391–396.
Ko
¨rner, C. (2003) Alpine Plant Life. Functional Plant Ecology of High
Mountain Ecosystems. Springer, Berlin, Heidelberg.
Lambrechts, C., Woodley, B., Hemp, A., Hemp,C.&Nnyiti,P.
(2002) Aerial Survey of the Threats to Mt. Kilimanjaro Forests.
UNDP, Dar es Salaam.
Lind, E.M. & Morrison, M.E.S. (1974) East African Vegetation.
Longman, London.
Meyer, H. (1890) Ostafrikanische Gletscherfahrten. Duncker &
Humblot, Leipzig.
Meyer, H. (1900) Der Kilimandscharo. Reisen und Studien. Reimer,
Berlin.
Mu
¨ller, M.J. (1983) Handbuch Ausgewa¨hlter Klimastationen der Erde.
Bodenerosion der Universita¨t Trier Mertesdorf (Ruwertal), Trier.
Nicholson, S.E. (2000) The nature of rainfall variability over
Africa on time scales of decade to millenia. Glob. Planet. Change
26, 137–158.
Nicholson, S.E. & Yin, X. (2001) Rainfall conditions in Equatorial
East Africa during the nineteenth century as inferred from the
record of Lake Victoria. Clim. Change 48, 387–398.
Po
´cs, T. (1976) The role of the epiphytic vegetat in the water
balance and humus production of the rain forests of the Ulug-
uru Mountains, East Africa. Boissiera 2, 499–503.
Thompson, L.G. (2000) Ice core evidence for climate change in the
tropics: implications for our future. Quat. Sci. Rev. 19, 19–35.
Thompson, L.G., Mosley-Thompson, E., Davis, M.E. et al. (2002)
Kilimanjaro ice core records: evidence of holocence climate
change in tropical Africa. Science 298, 589–593.
United Republic of Tanzania, Ministry of Water and CES, Con-
sulting Engineers Salzgitte GmbH (2002) Hydrometic Report of
the Hai District Water Supply Phase IV (2002). United Republic of
Tanzania, Ministry of Water and CES, Consulting Engineers
Salzgitte GmbH, Dar es Salaam.
Verschuren, D., Laird, K.R. & Cumming, B.F. (2000) Rainfall and
drought in equatorial east Africa during the past 1,100 years.
Nature 403, 410–414.
Volkens, G. (1897) Der Kilimandscharo. Darstellung der Allgemein-
eren Ergebnisse Eines Funfzehnmonatigen Aufenthalts im
Dschaggalande. Reimer, Berlin.
Walter, H., Harnickell,E.&Mueller-Dombois, D. (1975) Climate-
diagram Maps of the Individual Continents and the Ecological Cli-
matic Regions of the Earth. Supplement to the Vegetation
Monographs. Springer, Berlin.
10 Andreas Hemp
2009 The Author. Journal compilation 2009 Blackwell Publishing Ltd, Afr. J. Ecol.,47 (Suppl. 1), 3–10
