Atlas

Mt. Elgon
of
 Atlas
Mt. Elgon
of
Copyright © 2015 by the African Collaborative Centre for Earth System Science
and the IUCN Eastern and Southern Africa Regional Programme

This publication may be reproduced in whole or in part and in any form for educa-
tional or non-profit purposes without special permission from the copyright holder,
provided acknowledgment of the source is made. No use of this publication may
be made for resale or for any other commercial purpose whatsoever without the
prior permission in writing from the African Collaborative Centre for Earth System
Science and the IUCN Eastern and Southern Africa Regional Programme.

Copyright for certain images remains with original copyright holders as indicated in
the Acknowledgments.

Disclaimer

The colors, boundaries, denominations, and classifications in this report do not

imply, on the part of United States Agency for International Development, the
International Union for Conservation of Nature, the African Collaborative Centre
for Earth System Science or the Lake Victoria Basin Commission any judgment on
the legal or other status of any territory, or any endorsement or acceptance of any
boundary. Maps are not to be used for navigation.

Data Access and Questions

For access to the GIS files and maps referred to in this report or for any questions
and/or clarifications, please write to ACCESS at [email protected].

Recommended citation

ACCESS (2015) Atlas of Mt. Elgon. Prepared by African Collaborative Centre for
Earth System Science (ACCESS). Authors: Thomas J. Ballatore and Lydia Olaka.
Edited by Eric Odada and Daniel Olago. Published by USAID/IUCN. 54p.
Contents

Preface...........................................6

Overview Maps
Mt. Elgon in East Africa............... 10
Satellite....................................... 12
Physical........................................ 14
Political........................................ 16

Thematic Maps
Geology.......................................20
Soil............................................... 22
Topography..................................24
Vegetation...................................26
Surface Water.............................. 28
Temperature................................ 30
Precipitation................................ 32
Protected Areas........................... 34
Population.................................... 36
Roads and Fires............................ 38

Special Topics
Land Cover Change......................42
Climate Change............................44
Landslide Risk..............................46

End Matter
Acknowledgments.......................48
 Preface
M ount Elgon is a transboundary
ecosystem shared between Ken-
ya and Uganda. With the peak of the
the montane environment, leading
to increased risks of hazards such as
landslides, water shortages, water pol-
mountain at 4,320 m above sea level, lution, and biodiversity loss. Climate
it is the seventh highest mountain in change projections indicate that the
Africa. It is an important watershed montane ecosystem will experience
in the eastern Africa region, providing further stress in the long term through
ecosystems goods and services directly changes in temperature and precipita-
to those who live on its slopes, as well tion, and indicate that adaptation and
as to others far afield, including in mitigation actions need to be initiated
the Nile basin, primarily through its now in order to build the resilience of
river network. Some of its rivers feed, both the natural and human commu-
for example, the River Nile, through nities and their associated natural and
the Kyoga aquatic system and Lake socio-economic systems, respectively,
Victoria. The mountain boasts of so as to sustain benefits from the nat-
a National Park both in Kenya and ural systems and the well-being of the
Uganda, and is an important biodiver- people even under long-term climate
sity resource. Studies so far conducted change related stresses.
on the mountain ecosystems indicate
that it is undergoing changes that are The Atlas of Mt. Elgon has been
likely due to complex climate-land produced to provide easy to assimilate
interactions and climate change. Com- yet powerful evidence of the changes
munities have reported higher tem- that are taking place on the mountain
peratures and more erratic, variable as a consequence of global warming,
and intense rainfall accompanied by climate change and other actions that
changes in the onset and cessation of may be attributed to human activities.
rainy seasons, as well as changes in the The Atlas draws from information and
distribution of rains within the season. data that were produced during the
In addition, increasing human pressure USAID/IUCN project on “Imple-
due to growing settlements, farms, and menting a Resilience Framework to
use of timber resources is degrading Support Climate Change Adaptation
in the Mount Elgon Region of the human activities; providing an evi-
Lake Victoria Basin” carried out from dence base to support actions geared
2012 to 2015. This project was imple- towards preparedness for a changing
mented by IUCN in collaboration future in the Mount Elgon region
with the African Collaborative Centre including the adoption of adaptation
for Earth System Science based at strategies which mainstream ecosys-
the University of Nairobi, the Lake tem services, economic diversification,
Victoria Basin Commission and the adaptive management, and learning in
Global Water Partnership in Eastern water and land management; influenc-
Africa, with financial support from the ing regional policy frameworks to
United States Agency for Internation- better utilise systems approaches for
al Development (USAID). building climate resilience and inte-
grating these approaches across
It is our hope that this Atlas conveys sectors and into poverty reduction
information that can be put to good strategies and national development
use in many different ways, including: plans; and, enhancing learning from
providing scientific evidence to policy local to regional levels through better
makers, practitioners, and communi- access to information, networking,
ties on the changes taking place as a capacity building and leadership
consequence of climate change and development.

Mr. Luther Bois Anukur Prof. Eric Odada

Regional Director, IUCN African Collaborative Centre
Eastern and Southern Africa, for Earth System Science,
Regional office University of Nairobi
Overview
Maps
Mt. Elgon in East Africa............... 10
Satellite....................................... 12
Physical........................................ 14
Political........................................ 16
Mount Elgon is colossal by al-
most any measure. At its highest,
Wagagai Peak (4,321 m), it rises far
above anything in the surrounding
terrain. And while it may be lower
than the better known Kilimanjaro,
the shield of Mt. Elgon is an as-
tounding ~75 km in breadth—mak-
ing it approximately 10 times wider
than Mt. Everest! The following
overview maps set Mt. Elgon in its
regional context and put on display
the rich variety of physical and
political features one would expect
of this beautiful, sleeping giant.
10 Overview Maps

Juba
dary
.Boun
min
S O A N
Ad

U T H S U D

4°N

le
Ni
ite
Wh
Gulu
D E M O C R A T I C
R E P U B L I C

2°N
o f
U G A N D A
C O N G O
t
er

lb
L.A
L . Ky o g a
MOUNT
4321m+
Mbale
Kitale
Vic
tori
a
Nile
ELGON
RI

Jinja
ZO

a
Kampala Yal
EN
RW

0°N
Kisumu

d
ar
dw
E

Mbarara

L ake
Vi ctori a

Goma

R W A N D A
Kiv

2°S Kigali
u

Kagera

Mwanza

B U R U N D I
Bujumbura

si
ya
T A N Z A N I A L.E
Ta

4°S
ng
any
ika

30°E 32°E 34°E

 Mt. Elgon in East Africa 11

E T A
H I O P I

4°N

Lake
Tur ka n a

A
I
L 2°N

A
M
O

We
b i Jub
S

ba
K E N Y A

M T. K E N YA 0°N
AB

+
5199m Tana
ER

Nakuru
DA

Garissa
RE RANGE

Nairobi

2°S

Lamu

K Galena
IL +
5895m
IM
AN
JA
RO
I N D I A N
O C E A N

4°S
Mombasa
Equirectangular Projection – wgs84
Scale 1:4,000,000 1 cm = 40 km
0 20
2.5 40
5 80
10 120
15 160
20

km
36°E 38°E 40°E
 12 Overview Maps

1°30´N

1°15´N

1°N

This satellite image of Mt. Elgon

and its surroundings is a composite
of several Landsat 7 scenes
centered on 16 March 2015. It
shows the great diversity of land
cover types in the region ranging
from the afromontane rainforests
and alpine vegetation high on
the mountain, to the surrounding
cropland and urban areas at lower
0°45´E elevations, along with deforested
zones at mid-elevations. Seasonal
biomass burning on the mountain
can be seen as orange fires with
light, smoky plumes.

34°E 34°15´E 34°30´E

 Satellite 13

DA
AN
UG A
NY
KE

1°30´N

1°15´N

1°N

0°15´N

Equirectangular Projection – wgs84

Scale 1:400,000 1 cm = 4 km
0 2
2.5 4
5 8
10 12
15 16
20

km
34°45´E 35°E
14 Overview Maps L. Opeta

L. Bisina

Lake Nabyoto Kerim

he
IN

bo
LA

rom
P
OD
1°30´N FLO
NAL
SEASO

Olape

ri
Ata
Sip
ole i
eng
Olel

Sis
i
Sim
u
Kaiyape
Tutum Cave

1°15´N
M O U N T
Kaw
anoni S ir
onk Caldera Rim
o

Ug
and
a Pas
s
Masaba (4166m)
+
(Jackson’s Summit)
tala
ma Wagagai(4321m) +
Na

Bududa
Kubuna
Crater
Wanale
Ridge

1°N la fwa
Mana
ta
ma
Na

fwa
M an a

o
jar
mi
Na Em Ka
kw er
iga iki
t

Elevation
4,321 m ha k ha
Lwak

wa
Mata
Malakisi

Mala
wa
o

1,021 m L i g ag a

34°E 34°15´E 34°30´E

 Physical 15

Kerim
DA
AN
UG A
NY
KE

Ch
es i
ra
1°30´N

S ua m
ep
p lek
Ka

e
ob
pk
Che
)k
ree
(G
im
r
Ke

am
Su

Koi
nya
nga
o
B uk w

1°15´N

Suam

Hot
Springs O
rge m
Go on
Suam

a
+

r
sa
Koitobos (4222m)

p
Ka
+
Lower Elgon
(4302m)
Kimothon
Chepnyalil Cave
Endebess
Ngawarisha Cave Bluff

MacKingeny Cave
Kitum Cave

Kabe
w yan 1°N
Sa
bw

E L G O N
an i

et
eg

am
ig
No Suw
er w
a
us
tb
Koi

Ziw
a
Nz
oia

oia
Nz
Little
Ki
bu
k

0°15´N
Kip
sangw
a

i
i
Nzo

Mi
si kh
u

Equirectangular Projection – wgs84

Sergoi
Scale 1:400,000 1 cm = 4 km
Olemsong
ai 0 2
2.5 4
5 8
10 12
15 16
20

km
34°45´E 35°E
16 Overview Maps N A K A P I R I P I R I T
R A
O
G

K W
N

1°30´N
Kumi
B U L A M B U L I

K U M I

Kapchorwa

B U K E D E A K
Cheptui A
P
Bukedea Chebonet C
H
O
R

W
A
1°15´N
Sironko
Buluganya
P A L Buhugu
L I S M T.
S I
A R Budadiri ELG ON
O NP
N
Kabwangasi Nakaloke K O

Iki Iki
Mbale
A
B U D A K A D
Bushika U
Bugeme D
B U
Naboa

Budaka M B A L E Bulucheke
Bududa
1°N Bungokho

Nabiganda

A
J
E A
L W
T A A F
B U M A N Bumbo
Busolwe
Emerikit Bugobero
Busumbu
C H E P
T A I S

0°45´E

S I R I S I A
T O R O R O
Tororo

AMAGORO
Malaba
Malaba KAND UYI
34°E 34°15´E 34°30´E
 Political 17

A
A M U D A T B
I
E L
E E 1°30´N
N H
C

A
C H E PA R E R I A

K
Kaburon

K A P E N G U R I A

Bukwo

B U K W O
Makutano
1°15´N

KA
K W A N Z A PC
HE
RO
P

Endebess

Kitale
C H E R A N G A N I
1°N

MT.
ELG O N
TRANS NZOIA WEST

Kiminini

Y M O I B E N
ON Moi’s Bridge

O KW
PS
KA
Matunda

K I MI L IL I
Kimilili Kamukuywa
T O N G A R E N I

0°15´N
Natiri
Chwele Corner
S O Y
I
R
A
W E B U Y E L U G Equirectangular Projection – wgs84
Scale 1:400,000 1 cm = 4 km
Lugulu
0 2
2.5 4
5 8
10 12
15 16
20
Turbo
km
34°45´E 35°E
Webuye
‑

Thematic
Maps
Geology.......................................20
Soil............................................... 22
Topography..................................24
Vegetation...................................26
Surface Water.............................. 28
Temperature................................ 30
Precipitation................................ 32
Protected Areas........................... 34
Population.................................... 36
Roads and Fires............................ 38
Each map in this section explores
a single theme in depth. All maps
are presented at the same scale and
alignment. We encourage readers
to “flip” the pages back and forth
to enable comparison of the various
themes. The maps start with the
foundational themes of geology,
soils and topography before
moving on to vegetation and then
surface water, temperature and
precipitation. The maps end with a
look at the human-nature themes
of protected areas, population, and
fires and roads.
20 Thematic Maps

Geology ABOUT THE MAP >>

This map shows the geology of Mt. Elgon and the surrounding area. The
Elgon shield, composed of undifferenciated volcanics covers most of the
mountain proper and extends north through Mt. Kadam. Notably, the land-
slide-prone area of Bududa on the southwest of the mountain has granulite,
migmatite, gneiss and some undifferenciated agglomerates.

M
t. Elgon is an extinct volcano occurring at the
border of Uganda and Kenya, formed during
the Miocene age (22- 12 million years). The
rocks on the mountain are volcanic comprised mostly
of tuffs, ashes phonolites and nephelinites which are
collectively classified as undifferentiated volcanics.
Alkaline intrusives also occur in the region near Bududa.
The Mount Elgon volcanic succession is underlain by
Precambrian granitoid batholiths, and older hornblende
gneisses, schists and granulites of the Samia Series
(Davies, 1952, 1957) that occur also on the lower reaches
of Elgon, which are mostly metamorphics such as gran-
ullites and gneisses and migmates around Tororo and
Sironko in Uganda, while in Kenya there are intrusive
igneous rocks to the south and metamorphics in the
western border.
MASABA PEAK Also known as Jackson's Summit, Masaba (4,166m) rises
above the surrounding terrain of the crater rim. The name Masaba was
The mountain peak is at 4,321 m (Wagagai peak locat- once the generally used name for Elgon on the Ugandan side (Hancock and
ed in Uganda). The topography is characteristically Soundy, 1929). Looking west, Wanale Cliff near Mbale can be seen in the
a low, convex structure. The landscape is made up of distance. Photo by Gabriel Gersch.
cliffs, rocks and ash (Davenport et al., 1996). In the
caldera there are tiny lakes and moraine ridges, which
are remnants of the glaciations which occurred here
during the Pleistocene era (about 1.5 million years ago).
However, there are extensive signs of glaciations on the
upper reaches of Mt. Elgon, with the last glaciers having
disappeared perhaps 10,000 years ago (Miriam, 1995).
Moraines are present both within Mt. Elgon’s crater and
on the outer slopes of the mountain. On the northern
slopes, moraines occur as low as 3,350 m. One terminal
moraine feature within the crater rim is some 150 m high
and 3 km long (Miriam, 1995).

Glaciation and weathering activities coupled with other

factors gave rise to the various physical features that
characterize the mountain such as the Caldera, Endebess ~10cm
Bluff and the Elephant Platform. Within Mt Elgon, the
sodium rich rocks alter to form mineral rich salts that ELEPHANT TUSK MARKS IN KITUM CAVE Geophagy is common in ele-
are important for the wild animals such as elephants that phants, sometimes leading to a receding cliff. On Mt. Elgon, cooled lava
flows which are impervious to water act as (1) an umbrella protecting the
mine salt from these areas resulting in an abundance of soluble salts in the volcanic agglomerate underneath and (2) a structurally
caves. sound roof enabling the receding rock-face to keep going back into the
mountain. The result is a number of elephant-excavated caves of which Ki-
tum is currently the deepest (Redmond, 2015). Photo by Donald McFarlane.
LEGEND: CODES AND TYPES
AR Migmatic gneiss, banded biotite and orthogneiss, Ngn Nyanzian, polymetamorphic
ultrabasic intrusives Ni Banded quartzite and iron formation
As Flaser gneiss, mylonite q Quartzite, manganiferous metachert
BC Granulite, gneiss and migmatite, undifferenciated Sy Carbonatit (C), alkaline intrusives (Sy), foyaite (F),
C Carbonatit (C), alkaline intrusives (Sy), kimberlite (Kt)
foyaite (F), kimberlite (Kt) t Terrestrial, lacustrine, fluviatile deposits
Do Dolerite tat Undifferenciated agglomerate, tuff, ash with lava layers
gn Gneissose granite and granodiorite, mylonitic tN Nephelinite, melanephelinite, partly limburgite, augite
granitoid gneiss tP Phonolite, partly trachyphonolite, phonolitic ignimbrite
GpK Granite, granodiorite : post-Kavirondian Tu Undifferenciated pelitic metamorphics, surbodinate
K Undifferentiated metasediments crystalline limestone with graphite, psammitic
Kgn Kavirondian, polymetamorphic intercalations, para-amphibolite, magmatics
Ks Sandstone, graywacke Tuq Quartzite, mucovite quartzite
mm Biotite migmatite, biotite-hornblende migmatite tV Volcanics, undifferenciated
N Undiff. metasediments with B.I.F. and metavolcanics U Ultrabasite, serpentine, pyroxenite
NA Andesite, dacite W Granulite, pyroxene gneiss, chomockite
NB Basalt xA Metabasite, amphibolite
 W tV
W
Geology 21
mm
W C
q BC mm
BC q

Lake
Swamp
W
BC
q
W A
Swamp A ND
UG A
NY
KE
W
q

Sy
BC mm

Kween tN
xA
Kapchorwa
xA

xA
U
As
xA xA

xA
Bukwo
U
xA
U
tV
Sironko

As tN
tat BC

tN tN

tN
q
Mbale Endebess
tN
As
mm

tat tN tN Kitale
tat

Sy

NA
t t tV
t BC

Tuq

gn
C
Sy
Tu
Tuq gn
Tu
Kgn
Tuq
t
Kimilili
Ngn
Kgn
Ngn gn

Kgn tP
K
Tororo
Ngn mm
C
Ngn
gn

C Ks
gn
Malaba Tu
GpK Webuye
NA
Ngn
Bungoma
Equirectangular Projection – wgs84
Scale 1:500,000
Tu 1 cm = 5 km
0 2.5 5 10 15 20

NB km N
K
K NB
22 Thematic Maps

Soils ABOUT THE MAP >>

This map shows the soils of Mt. Elgon and the surrounding area with
hatched areas representing soils of either poor or very poor drainage. The
colors are an attempt to reflect the actual color of each soil type although
that is not possible as each soil is heterogenous and conditions are variable.

S
oils in the Mount Elgon region show a great variety
due to geological and geomorphological processes.
The dominant soil types are Nitisols (well weath-
ered acidic and reddish-brown to brown in colour) and
Andesols (high in organic matter and derived from
volcanic rock weathering), as well as Acrisols, Cambisols
and Ferralsols. Humic nitisols are the most dominant
occurring between 2100 and 3000 m above sea level, in
the forest belt. These are typically brown to red-brown
loams that overlie a deep red clay loam subsoil.

Another dominant soil on the mountain is Humic Cam-

bisols, occurring above 3000 m, and are characteristical-
ly very dark in colour. At higher elevations, above 3500
m, terric histosols occur in swamps and peatlands.

In the lower elevation areas below 2,000 m, a variety of HIGHLY ERODIBLE SOILS This is a photo of the aftermath of a landslide
that occurred in Bududa, Uganda. Failure of soils on steep, concave, defor-
soil types occur. Notably, Eutric vertisols occur in the ested slopes after soaking rains is a perennial problem on Mt. Elgon and is
north and northwest of Elgon around rivers. discussed at length in the section on Landslide Risk (pp. 46-47). Photo by
Thomas Ballatore.
Under natural conditions, soils of the region are deep
and support a varied tropical forest. Cleared of forest
cover, the soils support a highly productive agriculture
for limited periods, and remain susceptible to landslides
especially on steeper cleared slopes during rainy periods.

The Vertisols of some areas are characterised by high

clay content (exceeding 41%) and a high plasticity; such
properties qualify the soils as ‘problem soils’ that are
susceptible to landslides (Mugagga et al., 2011).

RED ROCKS AND SOIL This photograph shows the common red POORLY DRAINED SOILS NEAR OUTLET OF CRATER This is where Kenya
color of rocks and soil found in the Mt. Elgon region. Photograph and Uganda meet at the start of the Suam river. The area is home to a
by Lydia Olaka. number of hot springs and supports abundant alpine moorland vegeta-
tion. Photo by Matthias Gehringer.

LEGEND: CODES AND TYPES

ACf Ferric Acrisols Fl Fluvisols NTr Rhodic Nitisols
Ach Haplic Acrisols FLe Eutric Fluvisols NTu Humic Nitisols
ACp Plinthic Acrisols FRh Haplic Ferralsols PHh Haplic Phaeozems
ALh Haplic Alisols FRr Rhodic Ferralsols PHl Luvic Phaeozems
ALu Humic Alisols FRu Humic Ferralsols PLd Dystric Planosols
ARl Luvic Arenosols GLe Eutric Gleysols PTa Albic Plinthosols
ARo Ferralic Arenosols GLm Mollic Gleysols PTe Eutric Plinthosols
CMd Dystric Cambisols HSs Terric Histosols RGd Dystric Regosols
CMe Eutric Cambisols LVx Chromic Luvisols UR Urban, mining, etc.
CMu Humic Cambisols LXh Haplic Lixisols VRe Eutric Vertisols
CMx Chromic Cambisols NTh Haplic Nitisols WR Water bodies
 ALh

ALu Soils 23
LVx
ARl NTh

Fl
WR
WR VRe

ARl
CMx

Fl
Ach

ALh
PHh

PTe FLe

Kween
VRe Kapchorwa
LVx

NTu
PHh
FRr
Bukwo LXh FLe

FRh
Sironko
CMe CMe

VRe
FRu

PTa NTh
VRe
HSs

Mbale Endebess FRu GLe

CMu
VRe NTr VRe
PHh
Kitale
UR
CMu
NTh NTh FRu
PTa GLm
CMu FRh

VRe FRu GLm

GLm
NTu
GLm
PHl NTu

NTr GLm
NTh FRr
PTa

Kimilili GLm

PTa FRh GLm

NTr GLm
ACp CMx
FRr FRr
GLe
Tororo Ach
FRh Ach
ACp Ach ARo FRh
GLe
PHl GLe
PLd
Malaba RGd
ACp
GLe
GLe
Webuye Ach
GLe GLe
FRr
Bungoma GLe Ach
GLe Equirectangular Projection – wgs84
ACp ACp Scale
ARo1:500,000 1 cm = 5 km
GLe Ach
FRr
0 2.5 5 10 15 20
PLd GLe
GLe ACf
FRr km
GLe FRr
GLe GLe LXh
Ach
24 Thematic Maps

Topography ABOUT THE MAP >>

This map shows the topography of Mt. Elgon and the surrounding region
derived from the recently released high resolution (~30 m) SRTM1 digital
elevation model (NASA). Contours placed at 500 m intervals enclose hypso-
metric tints increasing in brightness with elevation. The shield nature of the
volcano is clearly evident, as is the prominent flank vent in Bududa.

M
aasai call it Ol Doinyo Ilgoon—breast moun-
tain. From the pictures presented here, the
reader can perhaps imagine why Mt. Elgon
has gotten this moniker. Whether seen from above (as
in the map to the right) or from the side (as seen in the
figures below), Mt. Elgon does bear a resemblance to the
a mother's life-giving breast.

Although there is no generally agreed definition of where

the mountain starts, the altitude generally ranges from
approximately 1,500 m up to the highest peaks of 4,321
m (Wagagai) in Uganda and 4,301 m (Lower Elgon) in
Kenya. The summit is a massive, 8-km wide caldera—a
flat-bottomed depression formed as a result of past
volcanic activities and home to a rich flora of moorland
vegetation.

The characteristic landscape of Mt. Elgon with gen-

tle and steep slopes is a result of the series of volcanic
eruptions and subsequent weathering processes. Steeper
slopes are mostly found on the western and south-west-
ern part of the mountain in Uganda. The mountain is
also characterized by steep valleys and gorges cutting
down slope through which rivers flow, perhaps the most
dramatic of which is the Suam Gorge through which the
crater drains and the Suam River forms, marking the
northeastern border between Kenya and Uganda.
SLOPE >>
In this figure, slope increases from green through yellow to red.
Areas of widespread steep slope occur in the Bududa District of
Uganda—an area prone to landslides. The flat plains surrounding
the mountain support agriculture and grazing.

4000

3500
Elevation (m)

3000

2500

2000

1500

1000
0   10   20   30   40   50   60   70   80   90   100  km

TOPOGRAPHIC PROFILE The above figure is a topographic profile of Mt. Elgon as drawn from Mbale, Uganda (0 km) to Kitale, Kenya (92 km)—two
cities on opposite sides of the mountain. Note the generally higher slopes and rate of slope change on the Ugandan side (due mainly to topography
of the Bududa crater and the Wanale Cliffs) and the gentler slopes on the Kenyan side, as well as the higher elevation of Kitale than Mbale.

MT. ELGON SEEN FROM AFAR Mt. Elgon is so large and is a sight so far removed from most people's common experience that it can look decep-
tively small when seen by the unaccustomed eye. The mountain does not greatly protrude above the terrain like Kilimanjaro or even Mt. Kenya, but
after hours or days of traveling in the area, its formidable breadth becomes evident. Black line added to highlight the ridge from this view point.
 Surface
Topography
Water 25

A
A ND
UG A
NY
KE

1500

2000
Kapchorwa Kween

2500

Bukwo

3000

Sironko

3500

4000

Wagagai +
4321m
+ Endebess
Mbale Lower Elgon

3500
Kitale
2000

1500

3000

2500

2000
Kimilili

Tororo

Malaba
1500

Webuye

Bungoma
Equirectangular Projection – wgs84
Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20

km
26 Thematic Maps

Vegetation ABOUT THE MAP >>

This map shows both current vegetation (density indicated by shade of
green) and potential vegetation (overlaid polygons) which give an idea of
how Mt. Elgon's vegetation might look without human influence. Of partic-
ular note is the previously much larger extent of the afromontane forest.

T
he vegetation of Mt. Elgon is divided into dis-
tinct belts, which reflect altitudinally controlled
climatic zones commonly associated with large
mountain massifs. The major altitudinal vegetation belts
from base to top of the mountain are montane rainforest
(moist lower montane rainforest (1500-2450 m and dry
lower montane rainforest 2000-3050 m), bamboo forest
(2450-3050 m), upper montane forest (dominated by
Hagenia and Hypericum, 3050-3300 m), Ericaceaous belt
(also called moorland, 3300-3550 m) and the afroalpine
zone (above 3550 m) (Hedberg 1951, Hitimana et al.,
2004; Mugagga et al., 2012; Scott, 1994).

The moist lower montane forest is ecologically the most

diverse in tree species and is the preferred habitat for
Elgon Teak (Olea capensis), a highly exploited commercial
timber species (Marshall & Jenkins, 1994; Hitimana et FOREST CONVERTED TO FARMLAND As discussed in the section on Land
Cover Change (pp. 42-43), large areas of forest have been cut down and
al., 2004). Above 3500 m, swamps and moorland vege- converted to farmland. This area on the southwest Kenyan side of the bor-
tation dominate (Mugagga et al., 2012) including many der shows cultivation on steep slopes with the occasional remaining tree.
rare species such as the giant lobelia plant (Lobelia telekii, Photograph by Thomas Ballatore.
shown on p. 34) and giant groundsels (Dendrosenecio
elgonensis, below left) along with more common grasses
which are abundant in parts of the caldera (below right).
The natural rainforest vegetation of the lower slopes has
been cleared and replaced with agricultural fields and
plantations of potatoes, coffee and banana. (Podocarpus
forest in north)

APLINE VEGETATION NEAR THE SUMMIT Giant groundsels (Den- GRASS IN THE SUMMIT CRATER The caldera is an area of abundant rainfall
drosenecio elgonensis) are one of the striking species of plant and rich soils which support a moorland ecosystem comprised of mires,
that occurs at high elevations. Photograph by Arjen de Ruiter. various alpine plants and grasslands. Photograph by Gabriel Gersch.

LEGEND: CODES AND TYPES

A Afroalpine vegetation R Riverine wooded vegetation
B Afromontane bamboo S Somalia-Masai semi-desert grassland and shrubland
Bd Somalia-Masai Acacia-Commiphora deciduous bushland Wb Butyrospermum wooded grassland
and thicket Wcd Dry combretum wooded grassland
Bd/Wcd Transitional zone Somalia-Masai Acacia-Commiphora Wcd/wd Transition zone of dry Combretum wooded grassland
deciduous bushland and thicket and Dry Combretum and edaphic wooded grassland on drainage-impeded or
wooded grassland seasonally flooded soils
Bds Acacia-Commiphora stunted bushland Wcm Moist Combretum wooded grassland
Be Evergreen and semi-evergreen bushland and thicket WdK Acacia tortilis wooded grassland and woodland
E Montane Ericaceous belt We Upland Acacia wooded grassland
Fa Afromontane rain forest X Freshwater swamp
Fb Afromontane undifferentiated forest g Edaphic grassland on drainage-impeded or seasonally
Fd Single-dominant Hagenia abyssinica forest flooded soils
Ff Lake Victoria transitional rain forest w Water bodies
Fh Afromontane dry transitional forest wd Edaphic wooded grassland on drainage-impeded or
Fi Lake Victoria drier peripheral semi-evergreen seasonally flooded soils
Guineo-Congolian rain forest wd/R Edaphic wooded grassland on drainage-impeded or
P Palm wooded grassland seasonally flooded soils or riverine wooded vegetation
 Wcd
27
wd/R
E Vegetation
Wcd Be
Wb
wd Wcd Fb We
Wcd
X
wd

g
We Wcd
w Wcd
Wcd/wd Wcd/wd
wd Wcd Bd
Wcd
Wcd/wd
Wcd
Bd/Wcd Bds

wd
wd Wcd Wcd
Bd
Wcd
X

Wb Bd/Wcd

Fb
R
Wcd/wd Wcd
Wb
Kween Be
g
Be Kapchorwa Bd

Fa Fb

wd
Be
Wcm
Wcm
Wb Wb Bukwo
Wcm Fd Wcd

Sironko
E Fb
g B
Fd
Wcm Wcd
B
Wcd
Wcm
Fa wd
Be
Fb
B Endebess
wd Mbale A
Fa
Be
Wcm
Kitale
Fb
E
Fd
Fi
Fb
Wcd Fb
Wcm Fi B Fb wd

B B

Fa Fb

wd
g

Kimilili wd

Wcm
wd
Wcm
wd
Wcm
wd Wcm
Tororo

Wcm Wcm
wd
wd Webuye Wcm
wd
wd Bungoma
Fi Equirectangular Projection – wgs84
wd/Fi/Wcm mix wd Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20
wd
wd Wcm wd
Wcm km

Ff Fh
Fi
28 Thematic Maps

Surface Water ABOUT THE MAP >>

This map shows the surface waters of the Mt. Elgon region. In general,
the river network exhibits a radial drainage pattern as expected from the
topography of a shield volcano. In the flatter portions of the mountain's
lowlands, flooding is a hazard especially in the areas marked in dotted blue.

A
bundant rainfall nourishes vegetation not only
on Mt Elgon but, in the form of streamflow
and runoff, downstream areas as far away as
Egypt. Mt. Elgon is major water tower for both Uganda
and Kenya with three main river catchments flowing
into lakes Kyoga, Turkana, and Victoria. The Kyoga
system of lakes is fed by North Bugisu and Sebei streams
comprising of Nabongo, Muyembe, Simu, Nalugugu,
Siroko, Simu, Muyembe, Sipi, Cheptui, Atari, Nyenye
and Kerim and the perennial rivers on the south western
quadrant of Mt. Elgon comprising of the Malawa,
Manafwa, Namatala. The rivers that feed into Lake
Turkana consists of perennial rivers of the north eastern
quadrant including the Suam, while the rivers that feed
into Lake Victoria basin consists of some tributaries of
the river Nzoia and the trans-boundary rivers known as PRISTINE MOUNTAIN STREAM. A small stream flowing through a ver-
dant part of the north side of Mt. Elgon. Photograph by Gabriel Gersch.
Sio-Malaba-Malakisi.

There are a number of small shallow lakes in the summit MT. ELGON WATER TOWER
region including Jackson’s Pool is located at 3,943 m Elgon serves as a key water
tower for portions of three
and Lake Kimilili located at 4,150 m in a glacial cirque
major African lake basins:
(Street Perrott et al. 2007). Moreover, hot springs Kyoga, Turkana and Victo-
such as Maji ya Moto are found on the Kenyan side of ria. Flow to Turkana ends in
the mountain within the caldera. The mountain is thus the lake; flow to Kyoga and
Victoria makes its way to the
an important source of water for the population that
Mediterranean through the
directly depends on it for domestic and agricultural uses Nile River. Note: the red box
and for the sustenance of the three major lakes Victoria, below is the same extent as
Turkana and Kyoga. However, in recent years, land use the map on the next page.
(and possibly climate) change has contributed to higher Lake
silt loads in downstream rivers and damaging floods in Turkana

low-lying areas with poor drainage.

L. Kyoga

Lake
Victoria

SILTATION IN DEFORESTED AREA Stream in Cheptais after sud-

den rains showing high silt load. Photo by Thomas Ballatore.

JACKSON'S POOL Located near the peak of Masaba (Jackson's Summit), Jackson's Pool is one of the highest lakes in Africa (at 3,943m). It is sur-
rounded by alpine vegetation such as numerous giant groundsels reflected in the still waters. Photograph by Arjen de Ruiter.
 Surface Water 29

Lake
Opeta

Ker
im A
A ND
UG A
NY
KE
Seasonally
Flooded

Kween
Kapchorwa

am
Kerim Su

Sip
Si

i i
s

m Bukwo
Si

u
o
B ukw

Sironko

Sir Suam
onk
o

Jackson's Pool

tala
Nama

Mbale Lake Kimilili

Endebess
Kimothon

Kitale

Kabew
yan

f wa
M an a

Kimilili
ha
ha k Kib
Lwak uk

Tororo

Malaba i
akis
Mal

a ba Webuye
al
M
Bungoma
Equirectangular Projection – wgs84
Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20
a
oi km
Nz
30 Thematic Maps

Precipitation ABOUT THE MAP >>

This map shows the annual average precipitation in mm for the Mt. Elgon
region. The data is taken from the WorldClim dataset and represents the
average for the period 1950-2000. Higher rainfall on the western and
southern slopes is evident as is lower rainfall in the northeast.

R
ain is the main form of precipitation falling on
Mt. Elgon. It is highly influenced by the oro-
graphic effect of the mountain and the prox-
imity of Lake Victoria. Moisture-laden winds from the
southwest and the enormous expanse of Lake Victoria
bring rain to the region, resulting in higher rainfall 2015 Jan 03 2015 Jan 19 2015 Feb 04 2015 Feb 20

on the western and southern slopes compared to the

eastern and northern slopes with the consequence that
the Uganda portion receives more rain than the Kenyan
portion on average.
2015 Mar 08 2015 Mar 24 2014 Apr 06 2014 Apr 22
On the southern and western slopes, rain falls through-
out the year up to 1500-2500 mm per year, though it
tends to be drier between July to August and December
to February (the map on the opposite page shows annual
average rainfall). On the eastern and southern slopes,
December to March is the drier period, with precipi- 2014 May 08 2014 May 24 2014 Jun 09 2014 Jun 25

tation influenced by the moist southwesterly and dry

northeasterly winds.

Annual rainfall on Mt. Elgon has a bimodal pattern and

a duration that varies from area to area depending on 2014 Jul 11 2014 Jul 27 2014 Aug 12 2014 Aug 28
the altitude and topography, with the wettest months
occurring from March to October (van Heist, 1994).
The forest zone receives the maximum rainfall (Synott,
1968) and is important in the mountain’s role as a water
catchment for several million people (van Heist, 1994).
2014 Sep 13 2014 Sep 29 2014 Oct 15 2014 Oct 31

Areas to the north, and especially the northwest, which

include the districts of Nakapiripirit and Amudat
in Uganda and Kacheliba and Kapenguria in Kenya,
received much lower rainfall, in some cases reaching
semi-arid conditions. 2014 Nov 16 2014 Dec 02 2014 Dec 18

Possible changes in precipitation amounts and patterns MT. ELGON CLOUDS THROUGHOUT THE YEAR Landsat images for one
year showing persistent cloud cover on parts of mountain. Clouds are
due to climate change are an area of intense interest and a persistent feature Mt. Elgon, especially on the southwestern slopes.
are discussed in the section on Climate Change on pp. While this is due to orthographic effects of the mountain which are typ-
44-45. Anecdotal evidence suggests that the pattern of ically already present by the time of satellite overpass (approx. 11 a.m.),
rainfall has become less predictable and severe events it makes the task of the remote sensing analyst trying to find cloud-free
images to perform land use/land cover classifications quite challenging.
have become more common.

180
160
Monthly Precipitation (mm)

140
120
100
Mbale
80
60
Kitale
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12
Month
INTRA-ANNUAL VARIATION IN PRECIPITATION Kitale and Mbale. Data from the WorldClim database for precipitation from 1950-2000 show the typ-
ical bimodal pattern seen at both Mbale and Kitale. Note that even in the dry season (December to February), rainfall does occur. The values presented
here are averages and do not reflect the sometimes wet and sometimes dry years that occur.
 Precipitation 31

900

1000

1100 A
A ND
UG A
NY
KE

1200

1300

1400
Kween

1500
Kapchorwa

1600

Bukwo

Sironko
1700

1800

Mbale Endebess

2000 1900 Kitale

Kimilili

Tororo

Malaba

Webuye

Bungoma
Equirectangular Projection – wgs84
Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20

km
32 Thematic Maps

Temperature ABOUT THE MAP >>

This map shows the annual average temperature in °C for the Mt. Elgon
region. The data is taken from the WorldClim dataset and represents the
average for the period 1950-2000. The dependence of temperature on alti-
tude is striking and in contrast with the precipitation patterns seen on p.31.

S
urface temperatures in the Mt. Elgon region
depend on time of day and season, but are mostly
dependent on altitude. The temperature gradient
seen in the figures to the right is not something that
occurs anywhere else in the region except in the areas of
Mt. Kenya and Kilamanjaro. This has profound effects
on the biota and for the management of the mountain's
resources.

One can imagine that as temperature increases due to

climate change (see pp. 44-45), this role of the mountain
as an oasis and refuge from heat will become even more
important. In other areas, where species would have to
migrate hundreds or thousands of kilometers latitudinal-
ly to find climates similar to their “natural” range, barri-
ers to movement are likely to be high given the distance.
For species around Mt. Elgon, the required movement
in terms of distance is much lower, although it would
involve a altitude change, e.g. moving up 100 m on Elgon
may have the same effect as travelling hundreds or more
km in any direction. Nevertheless, land use pressures on
the lower edge of the forest are significant and would
play a key role in affecting any possible changes in range.

There are interesting differences in the intra-annual

temperature regimes of the two major cities on each side
of the mountain, Mbale (Uganda) and Kitale (Kenya).
Mbale lies at a lower elevation (approx. 1130 m) than
Kitale (approx. 1900 m). This is the main cause of the
prominent absolute difference in average monthly
temperature as shown below. On the other hand, the
average diurnal range of temperature is greater in Kitale, LAND SURFACE TEMPERATURE Thermal Landsat image for 16 Mar 2015.
with an annual average high of 27.2 °C and low of 10.1 In this non-optical, thermal emission image, the hottest surface areas are
°C, whereas in Mbale comparable figures are 29.2 °C and shown in red and coolest ones in white. The temperature gradient shown
16.6 °C, respectively. This difference in range, 17.1 vs. on the opposite page is evident but there are notable “warmer” areas
near the top of the mountain in areas of low vegetation cover, especially
12.6 is significant and has affects on the flora and types of on southern facing slopes. The coolest areas correspond with dense forest
crops raised in each area. canopy, especially on the northerly facing slopes. The fires noted on pp.
12-13 on upper part of mountain are remarkably clear. It is also interesting
to note the generally lower surface temperature of Lake Opeta and other
water-logged areas in the northwest.

25
24
Monthly Ave. Temperature (°C)

23
22 Mbale
21
20
19
18
Kitale
17
16
15
1 2 3 4 5 6 7 8 9 10 11 12
Month
INTRA-ANNUAL VARIATION IN TEMPERATURE Kitale and Mbale. This figure shows the average monthly temperature for an average calendar year
between 1950-2000 (based on WorldClim data). The general seasonal trend of highest temperatures of lowest temperatures from approx. July to
September is evident and shared by both towns. The general coolness of Kitale relative to Mbale is due mainly to Kitale's higher elevation.
 Temperature 33

DA
AN
UG A
NY
KE

23

22
21
20 Kapchorwa Kween

19
18

17

16

Bukwo
15
14
13
Sironko
12

11
10
9
8
7

Mbale Endebess

Kitale

Kimilili

Tororo

Malaba

Webuye

Bungoma Equirectangular Projection – wgs84

Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20

km
34 Thematic Maps

Protected Areas ABOUT THE MAP >>

This map shows the major protected areas in the greater Mt. Elgon region.
Communities living around the protected areas are highly dependent on ac-
cess to various resources within the protected areas adjacent to which they
live and most protected areas allow some access (Petursson, 2013).

M
t. Elgon is a protected area in both Uganda
and Kenya. On the Uganda side, Mount Elgon
National Park covers the entire protected
area while on the Kenya side there are three protected
areas: Chepkitale National Reserve, Mount Elgon Forest
Reserve and Mount Elgon National Park.

The Mt Elgon Forest Reserve in Kenya was gazetted un-

der Legal Gazettement Notice No.44 of 30th April 1932
under the management of the then Forest Department
and initially covered an area of 107,821 hectares. After
the annexation of the National Park and the Chepkitale
National Reserve, the Forest Reserve now covers an
area of 73,705 ha and is currently administered by eight
gazetted forest stations namely Saboti, Mt. Elgon (Kip-
togot), Sosio, Suam, Kimothon in Trans-Nzoia County, ZEBRA WITH FOAL Mt. Elgon NP, ANTELOPE Mt. Elgon NP, Kenya.
and Cheptais, Kaberua and Kaboiywa forest stations in Kenya. Photograph by Lydia Olaka. Photograph by Josep M. Gracia
Bungoma County (Tokro et al., 2014).

The Mt Elgon Nature Reserve was annexed under Legal

Notice No 111 of 25th October 1967 and gazetted as a
National Park under Legal Gazettment Notice No. 112
of 5th April 1968 of an area 16,916 ha. The Kenyan part
of Mt Elgon National Park is under statutory man-
agement and administration of Kenya Wildlife Service
(KWS) and the Park Headquarters is situated at Chorlim.

The Chepkitale National Reserve was recently annexed

from the Forest Reserve under Legal Gazettment Notice
No. 88 of June 2000 and is under the jurisdiction of the
County Government of Bungoma (CGB). It covers an
area of 17,200 ha. The land use categories constitute a
total land cover of 107,821 hectares.

Mt. Elgon National Park in Uganda was gazetted in 1993

having been initially gazetted as Central Forest Reserve
in 1938. The change in status from a FR to NP was in
recognition of its water catchment, biological, cultural, GIANT LOBELIA PLANT (Lobelia telekii) Mt. Elgon National Park, Uganda,
near the summit. The giant lobelia is one of the rare forms of vegetation
historical and other values in line with the Uganda Wild-
found in the alpine zone of Mt. Elgon that attracts visitor from around the
life Act (Cap 200). world to the mountain. Photograph by Gabriel Gersch.

ELEPHANTS IN KITUM CAVE Shown here is an amazing example of elephant culture, passed down through generations, of salt mining that results in cre-
ation of caves (speleogenesis) with a dark zone. Mt. Elgon is the only place in the world where such activity is known to take place. The evidence suggests
that caves such as Kitum cave shown here are not places into which elephants simply wander in search of salt, but rather elephant salt mines­—the result
of an unbroken line of instruction dating back hundreds of generations (Bowell et al. 1996, Redmond 1982, 1992, 2015). Photograph by Ian Redmond.
 Protected Areas 35

Kadam
Forest Reserve

Amudat Community
Lake Pian Upe Wildlife Reserve Wildlife Management Area
Opeta

Seasonally
Flooded

Kween
Kapchorwa

Bukwo

Mount Elgon National Park

Sironko

Mount Elgon
Forest Reserve

Mbale Endebess

Chepkitale
National Reserve Mount Elgon
National Park Kitale
Mbale FR

Kitalale
Forest Reserve
Mount Elgon Forest Reserve

Sekhendu Forest Reserve

Kimilili

Tororo Forest Reserve

Tororo
Turbo FR
Lugari FR
Malaba

Webuye

Bungoma
Equirectangular Projection – wgs84
Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20

km
36 Thematic Maps

Population
ABOUT THE MAP >>
This map shows the various administrative areas in the Mt. Elgon region
along with population density as expressed in persons per hectare at a res-
olution of 100m produced by WorldPop through the use of adjusting census
figures with axillary information such as lights at night, land use and so on.

F
ossil evidence suggests that our species arose on
the plains of East Africa millions of years ago. Un-
doubtedly, long before recorded history, the first
Ka r a m o jo n g
human laid eyes upon Mt. Elgon. Perhaps these ancient
humans even made use of resources on the mountain—
but we may never know as the area is unfortunately not
conducive to fossil formation. But what is certain is that Po kot
at no other time in history has the population of humans
been greater than now. This anthropogenic pressure is Kup s a b i n y
undoubtedly changing the face of the mountain. Te s o

Population in the area is spread over two countries

(Kenya and Uganda) and is ethnically and linguistically
diverse and growing. The map to the right shows the nu-
merous linguistic groups in the area. Within the Mount
Elgon region there are five main ethnic groups, the re Masaba
Gwe

Bagisu and Sabiny live in Mbale and Kapchorwa districs

of Uganda and in Kenya the Sabaots inhabit the higher
S a b aot
elevations of the mountain while the lower elevations
le
Nyo

are inhabited by the Bukusu, other Luhya communities,

Pokots and Kalenjins.
Bu k u su -
L u hya -
Population growth has been rapid. According to the Ka le nj in
World Bank, growth rates in both countries have been
in the range of 2.6~3.8% per annum since 1960 with the
latest figures in 2014 of 2.65% for Kenya and 3.31% for
Te s o
Uganda (World Bank, 2015). The rates of growth in the B uk us u
fertile areas surrounding Mt. Elgon have even exceeded Adhola

these high figures.

Kalenjin
Luhya
Related to growing populations is perhaps the most Luhya

significant land-use change trend: the expansion of LANGUAGES Africa is known for its abundance of linguistic groups and the
agricultural fields from lower and gentle slopes to Mt. Elgon region is no exception. While English is an official language of
critically steep ones. The high population density and both Kenya and Uganda and is widely spoken in the major towns around
land shortage in places such as Bududa compel people to Mt. Elgon, the importance of local languages becomes more important in
more rural areas. Populations are fluid and many people are multilingual:
cultivate on unstable slopes that are prone to landslide. the areas shown above should be taken as general indications of prevalence
Such hazards have been reported since the beginning of of a given language and not as strict lines on a map. It is interesting to note
the 20th Century in certain places (Knappen et al, 2006) that for some languages, such as Pokot and Teso, the populations essential-
and loss of life continues to be a common occurrence. ly span the border between Kenya and Uganda, demonstrating in part the
close ties those dependent on the mountain have with each other.

UGANDA (2014 CENSUS) KENYA (2009 CENSUS)

District Population District Population District Population District Population

Amudat 111,758 Kumi 258,073 Amagoro 117,947 Kwanza 236,218
Bududa 211,683 Kween 95,623 Amakura 137,924 Lugari 292,151
Bukedea 188,918 Manafwa 352,864 Cheptais* 172,377 Malava 205,166
Bukwo 89,253 Mbale 492,804 Kacheliba 156,011 Nambale 327,852
Bulumbuli 177,322 Nakapiripirit 169,691 Kanduyi 408,598 Sirisia 243,535
Kapchorwa 104,580 Sironko 246,636 Kapenguria 181,063 Tongareni* 320,300
Katakwi 165,553 Tororo 526,378 Kapsokwony* 172,377 Trans Nzoia West 387,366
Kimilili* 320,300 Webuye 230,253

POPULATION OF DISTRICTS The latest census data for the districts shown in the map on the right are presented above. The Ugandan census fig-
ures are from the recently completed 2014 census are still considered provisional figures. For Kenya, reconciling the 2009 census figures with the
districts on the map on the right was challenging given re-districting and re-naming that has occurred since. See notes on p. 50, especially for the
information on Cheptais and Kapsokwony (previously Mt. Elgon district) and Kimilili and Tongareni (previously Bungoma North).
 Population 37

K ATA K W I Population
(persons/ha)
N A K A P I R I P I R I T <1
1-4
5-9
10 - 24
K U M I 25 - 49
50 - 99
A M U D A T > 100

A
B
K W E E N I

L
BULAMBULI

E
H
C
A
K
Kween
Kapchorwa
BUKEDEA

KAPCHORA
KAPENGURIA
Bukwo

B U K W O

Sironko

MT.
ELGON
SIRONKO NP
K W A N Z A

Mbale Endebess

B U D U D A
M B A L E Kitale

MT. ELGON
TRANS
NZOIA
WEST
M A N A F W A Y
ON
O KW
PS
KA
C H E P T A I S
KIMILILI
TONGARENI
Kimilili

T O R O R O S I R I S I A

Tororo

AMAGORO
LUGARI
WEBUYE
Malaba

Webuye

Bungoma
Equirectangular Projection – wgs84
AMAKURA Scale 1:500,000 1 cm = 5 km
K A N D U Y I 0 2.5 5 10 15 20
NAMBALE MOSOP
M A L AVA km
38 Thematic Maps

Roads and Fires

ABOUT THE MAP >>
This map shows a landscape marked by human activities—criss-crossed by
roads and lit up by fires. As one would expect, the density of roads is well
correlated with population density (pp. 36-37); however fires tend to occur
in areas of low habitation implying they are related to rural activities.

H
umans change the natural environment in
myriad ways including through conversion of
land to agriculture, construction of buildings,
and deforestation. Two other prevalent signs of human
impact—roads and fires—are examined here.

The figure on the opposite page shows the network of

trunk, primary, secondary, tertiary, and other types of
roads in the Mt. Elgon region. The majority of trunk
roads have been tarmacked and help facilitate transport
in the region and beyond. Most secondary, tertiary and
other roads, however, are unpaved and, especially in the
rainy season, hard to pass. Much transport of agricultur-
al products from individual fields takes place by bicycle
and other non-motorized means.

The fires which are so common in the dry season (see UNPAVED ROAD The majority of the roads shown in the map to the right
figure below) are another obvious sign of the human are similar to the one shown above (lower slopes on Kenyan side). Rains
impact on Mt. Elgon. Perennial bush fires cause great often make roads impassable to most vehicles. Much of the transport of
agricultural products is done by bicycle. Photo by Lydia Olaka.
damage especially on fragile moorlands, bamboo and the
Hagenia zone of the afro-montane zone where there is
low resilience. Fires are often set deliberately by honey
harvesters, poachers, and others. Accidental fires also
originate from the nearby farmlands during the process
of farm preparation.

The fires shown in the map on the right occur mainly in

two general areas: at higher elevations on the mountain
and on the northern plains. Notably, these areas quite
isolated from large population centers and the fires are
likely set by individuals or small groups. Needless to
say, it is difficult for the authorities to monitor such a
widespread area, especially considering the near absence
of roads in the fire-prevalent areas.

FIRES FROM SPACE The images shown here are Landsat 7 ETM+ images
from 16 March 2015. The image on the left shows a “psuedo-natural” color
view (using bands 7-5-3) with fires appearing as bright orange-red pixels
or blackened recently burned areas. The image on the right uses the ETM+
band 1 which is useful for seeing smoke—plumes of which are abundantly
present on this day in this otherwise cloudless image.

FIRES IN THE MT. ELGON REGION (MODIS)

1200
THERMAL ANOMOLIES

1000

800

600

400

200

0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

FIRES THROUGHOUT THE YEAR NASA’s MODIS Active Fire Mapping programme monitors thermal anomalies from space on a daily basis. The data
shown here are from April 2014 to March 2015 (the axis is organized along the traditional Jan-Dec scale, however). Note the virtual absence of fires
detected from mid-April through mid-October due to absence of land-clearing activities during the rainy seasons (p. 30).
 Roads and Fires 39

Thermal Anomolies
(Fire Radiative Power)
4 - 49
50 - 100
> 100+

Roads
Trunk/Primary
Secondary
Teriary
Other

Kween
Kapchorwa

Bukwo

Sironko

Mbale Endebess

Kitale

Kimilili

Tororo

Malaba

Webuye

Bungoma Equirectangular Projection – wgs84

Scale 1:500,000 1 cm = 5 km
0 2.5 5 10 15 20

km
Special
Topics
Land Cover Change......................42
Climate Change............................44
Landslide Risk..............................46
Having explored overview and in-
dividual thematic topics in previous
sections, attention now turns to
integrated themes. Here, you will
find maps on Land Cover Change,
Climate Change, and Landslide
Risk—all of great importance to
the people living on and near Mt.
Elgon. Each topic tells a complex
story: forest has been lost but
restoration efforts are underway;
climate is changing but precisely
how is not clear; and landslides are
an ongoing threat but prevention
programmes are gathering pace.
42 Special Topics

Land Cover Change

T
he Mt. Elgon region has seen dramatic changes in land Landsat 1 MSS sensor used for the 1973 image. The change
cover over recent decades. In particular, there has been in red area from a “well-rounded” shape in 1973 to a much
a marked reduction in forest cover due to clearing more jagged appearance shows the numerous excisions and
of land for agricultural production. A continuous record of encroachments over the years. Less clear at this scale but quite
satellite observations since the early 1970s to present allows important are the areas that have been reforested through
mapping of this land use/land cover (LULC) change. numerous conservation efforts on the mountain (See ACCESS
2014 for details.)
Below, two color-infrared images taken from 1973 and 2013, re-
spectively, are presented. Vegetation (in this case, mainly trees) Corresponding maps of tree loss and current (2013) tree cover
is shown in red, as per convention dictated by the nature of the are shown at the bottom. The decrease in forest cover from

COLOR INFRARED IMAGE (1973) COLOR INFRARED IMAGE (2013)

34°20'E 34°30'E 34°40'E 34°20'E 34°30'E 34°40'E

1°20'N
1°20'N

1°20'N
1°20'N

1°10'N

1°10'N
1°10'N

1°10'N
1°N

1°N
1°N

1°N
0°50'N
0°50'N

0°50'N

0°50'N
34°20'E 34°30'E 34°40'E 34°20'E 34°30'E 34°40'E

TREE LOSS 1973-2013 TREE COVER 2013

34°20'E 34°30'E 34°40'E 34°20'E 34°30'E 34°40'E

1°20'N
1°20'N

1°20'N
1°20'N
1°10'N

1°10'N
1°10'N

1°10'N
1°N

1°N
1°N

1°N
0°50'N

0°50'N

0°50'N
0°50'N

34°20'E 34°30'E 34°40'E 34°20'E 34°30'E 34°40'E

 Land Cover Change 43

1973 to 2013 is dramatic and is a crucial input to the landslide tea and maize farming, and a lowland zone where beans, yams
risk analyses that follow in this atlas. and onions are grown. Arabica coffee is traditionally the major
cash crop in south western Uganda. The staple food for Uganda
The results shown in the figure on the opposite page were de- is bananas while in Kenya the staple is maize. Much of the cul-
rived through a process of object-based image analysis in which tivation takes place on steep slopes ranging between 36° and 58°
groups of pixels were assigned to one of the four given classes (Mugagga et al., 2011).
through a segmentation algorithm. The overall accuracy was
found to be 92.1% when compared with 200 randomly spatially Subsistence small-scale agriculture is common, however, some
distributed known points. places such Trans-Nzoia have large areas under large-scale
farming. Large-scale farming of maize is found especially in the
In general, land in the Mount Elgon region can be divided Endebess-Kitale plain. Most of the large-scale farms belong to
between the protected areas (National Parks) and farmland. the Agricultural Development Corporation (ADC). On the
Land use in the farmlands is itself divided between two topo- Uganda side, agroforestry is a common practice with coffee,
graphic zones: an upland zone characterised by intensive coffee, banana and maize systems.

LAND COVER AND ITS CHANGE ON MT. ELGON (1973-2013)

34°20'E 34°30'E 34°40'E

Bare/Deforested
Grass or Tea
Forest
High Mountain Vegetation

1°20'N
1°20'N

1°10'N
1°10'N

U G A N DA
K E N YA
1°N
1°N

0°50'N
0°50'N

0 2.5 5 10 km

34°20'E 34°30'E 34°40'E

44 Special Topics

Climate Change Change in

A
Annual Ave.
necdotal evidence from people living on Mt. precipitation) might change by 2050 relative to the Temperature
(2050)(°C)
Elgon suggests that climate has been and is baseline (1950-2000) average. Analysis is based on 0.8 - 1.1
changing. In particular, many locals have five General Circulation Models (GCMs) used in the 1.1 - 1.4
noted that precipitation patterns are no longer as Fifth Assessment Report of the Intergovernmental 1.4 - 1.7
predictable as they once were, and when rain does Panel on Climate Change (IPCC) from the Coupled 1.7 - 2.0
come, it often falls with greater intensity (see Cohen Model Intercomparison Project-Phase 5 (CMIP5) 2.0 - 2.3
2014). of the World Climate Research Programme. These 2.3 - 2.6
five GCMs were chosen from a larger set of down- 2.6 - 2.9
The maps below present outputs of a selection of the scaled (2.5 arc-minute, ~5 km resolution) GCMs 2.9 - 3.2
latest climate models to demonstrate how climate available from WorldClim because they are reported 3.2 - 3.5
(i.e. annual average temperature and annual average to perform better than others in predicting the El 3.5 - 3.8

GENERAL CIRCULATION MODEL

CCSM4 GISS-E2-R HadGEM2-ES MRI-CGCM3 NORESM1-M
RCP 2.6
REPRESENTATIVE CONCENTRATION PATHWAY

RCP 4.5
RCP 6
RCP 8.5
 Climate Change 45

Niño-Southern Oscillation (ENSO) dynamics so in spots up to 3.8°C by 2050 for the worst case (RCP Change in
important for East Africa (Bellenger, et al. 2014). 8.5). On the other hand, if strong global efforts are put Annual Ave.
in place (RCP 2.6), and if CCSM4 or NORESM1-M Precipiation
Results for Temperature (p.44) and Precipitation are accurate, then temperature rise might only be (2050)(mm/yr)
< -300
(p.45) for the Mt. Elgon area (note: each box is the ~1°C. For precipitation change, the picture is less clear.
-299 - -200
same extent as the thematic maps–the mountain CCSM4 predicts large increases of up to 500mm by -199 - -100
shading is slightly visible) from these five models 2050 under RCP 8.5, whereas other models show little -99 - 0
are given below for each of the four Representative change or even drier conditions. Ideally, we would 0 - 99
Concentration Pathways (RCPs) used by the IPCC like to know how the distribution of rainfall between 100 - 199
to predict future radiative forcing. Overall, all models seasons, or the intensity of given events, or the change 200 - 299
and all scenarios show an increase in temperature but in probability of droughts or floods might change; 300 - 399

not agree on how much or where. For example, the however, current models do not seem able to reliably 400 - 499

HadGEM M2-ES model runs “hottest” with increases predict for this important region. > 500

GENERAL CIRCULATION MODEL

CCSM4 GISS-E2-R HadGEM2-ES MRI-CGCM3 NORESM1-M
RCP 2.6
REPRESENTATIVE CONCENTRATION PATHWAY

RCP 4.5
RCP 6
RCP 8.5
46 Special Topics

Landslide Risk

O
ne of the marked features of Mt. Elgon is the people SLOPE FACTOR
living on steep, deforested slopes on weak terrain.
0 (<20% and >70%)
Landslides are common, sometimes taking tens to 1 (20-30% and 60-70%)
hundreds of lives. Much academic and field work has been 2 (30-40%)
done on characterizing the conditions that lead to slope 5 (40-60%)

failure. In particular, Knapen et al. (2006), Claessens et al.

(2007), and Mugagga et al. (2012) have used field studies at
landslide sites to develop a list of causative factors.

In the work presented here, information from the literature

was used to assign values to different attributes for the main
factors contributing to landslide risk. The values shown in
the five figures on this page roughly reflect the weight that
a given factor has based on the information provided in the
sources noted.

Overall, five key factors are apparent from the literature

(1) Slope Factor (derived from SRTM3-NASA), (2) Plan
Curvature Factor (derived from SRTM3-NASA), (3) Aspect

PLAN CURVATURE FACTOR ASPECT FACTOR

0 (convex) 1 (other)
1 (flat) 2 (between 300-45o)
5 (concave)

PRECIPITATION FACTOR LAND COVER FACTOR

1 (<1600 mm/yr) 0 (Natural Trees)

2 (>1600 mm/yr) 1 (High Mountain Veg.)
5 (Altered Land)
 Landslide risk 47

Factor (derived from SRTM3-NASA), (4) Precipitation Factor Sustained rainfall events that saturate and weaken slopes are
(based on WorldClim), and (5) Land Cover Factor (based on also identified as a key landslide risk factor. While the data
classification of Landsat images). Full documentation of the available was for annual average precipitation and not for
methods used can be found in ACCESS (2014). specific rainfall events, information from Claessens et al. (2006)
was used to reclassify the precipitation data into two risk class-
The section on slope gradient in Figure 7 of Knapen et al. es.
(2006) was used to reclassify slope into four classes represent-
ing the relative risks of a given slope based on field data from Deforested land was shown to also be a critical factor in slope
98 recent landslides. The relative risks represent the fact that stability. Information in Mugagga et al. (2012) was used to
landslides are more likely to occur in moderate to steep slopes reclassify the land cover raster into three classes.
but unlikely in extreme slopes or flat areas.
To determine Total Cumulative Landslide Risk, the above five
Plan curvature is the “slope of the slope” in the downward risk factors were summed to create a “sum of risk factors” map
direction and is highlighted by Knapen et al. (2006) as a risk ranging from 0 (no risk) to 19 (highest risk). For slopes <20% or
factor. Concave slopes show a markedly higher likelihood of >70%, the total sum of risk factors is set to zero.
landslide than convex slopes.
It is important to note that the calculations given here are based
The direction a slope faces also has an effect on the likelihood on general risk factors and not physical models of given sites;
of a landslide. In particular, for Mt. Elgon, Knapen et al. (2006) therefore, results should be taken as indicative of possible risk
show that north facing slopes have a higher risk. and not of actual, definite risk.

TOTAL CUMULATIVE LANDSLIDE RISK

34°15'E 34°30'E 34°45'E 35°E

Cumulative
SUM OF ALL RISK FACTORS Risk
0
3
4
5
6
1°30'N

1°30'N
7
8
Bulambuli 9

Kween 10
11
12
13
Kapchorwa 14
15
16
Bukwo 17
1°15'N

1°15'N
18
19

Mt. Elgon NP
Sironko
Kwanza

Mbale Muncipality
Bududa
Mbale
1°N

1°N

Mt. Elgon
Trans Nzoia West

Manafwa

Cheptais
0°45'N

0°45'N

34°15'E 34°30'E 34°45'E 35°E

48 End Matter

Acknowledgments
Authors Thomas J. Ballatore and Lydia Olaka. Edited from auxiliary images. Ocean bathymetry
by Eric Odada and Daniel Olago. Published from General Bathymetric Chart of the
Thomas J. Ballatore and Lydia Olaka by USAID/IUCN. 45p. Available at: http:// Oceans-GEBCO (2014), a 30 arc-second global
cmsdata.iucn.org/downloads/tech_annex_re- grid of underwater elevations.
Editors port_web_version.pdf
Boundaries: Global Adminitrative Areas Data-
Dan Olago and Eric Odada For any questions and/or clarifications, please set GADM (2015) modified with auxiliary data
write to ACCESS at [email protected]. taken from satellite images and rivers derived
Reviewers from SRTM1 and SRTM3 as well as other local
FRONT MATTER sources.
A validation workshop was convened in
Kisumu, Kenya on 16-17 April 2015. The FRONT COVER Place and Feature Names: National Geo-
authors and editors would like to express p. 1 spatial-Intelligence Agency GEOnet Names
their gratitude to the participants as listed Server (NGA-GNS) (2015) and local stakehold-
below, noting that all remaining errors are the Photograph credit: Arjen de Ruiter ©2010. er input.
responsibility of the authors and editors: Location: Tutum Cave, Uganda.
SATELLITE MAP
To Mr. Godfrey Wekesa, Director Environ- INSIDE COVER pp. 12-13
ment, Trans Nzoia County, and Mr. Jackson pp. 2-3
Kibor, Warden, Mt. Elgon National Park from The satellite image presented here is a
the Sabwani Catchment, Trans Nzoia County. Data: Natural Earth, Tom Patterson, US Na- composite of four recent, cloud-free Landsat
tional Park Service. 7 ETM+ (NASA) images from 2015. The
To Mr. David Tsolobi, Community Develop- spatial extent is large and requires two
ment Officer, Bududa, Ms. Marion Namono, TABLE OF CONTENTS scenes (Path 170/Row 59 and Path 170/
District Environment Officer, Bududa, and Mr. pp. 4-5 Row 60) for full coverage. Furthermore,
Musamali Michael, District Natural Resources the Landsat 7 ETM+ sensor does not have
Officer, Bududa, from the Manafwa Catch- Photograph credit: Lydia Olaka ©2015. Loca- complete coverage within a scene due to
ment, Bududa District. tion: Sipi Falls, Uganda. the Scan Line Corrector failure. Two scene
from very close in time (28 Feb 2015 and
To Mr. Rono Benson, District Forest Officer, PREFACE 16 Mar 2015) were merged using a local
Bukwo, and Ms Chemutai Olive, District pp. 5-6 histogram matching algorithm to produce
Environment Officer, Bukwo, from the Suam this apparently “seemless” image. The four
Catchment, Bukwo District. Background photograph credit: Gabriel scenes used were LE71700592015059SG100,
Gersch ©2015. http://www.gabriel-gersch. LE71700592015075SG100,
To Ms Bisikwa Sarah, District Natural com/ LE71700602015059SG100, and
Resources Officer, Manafwa, Mr. Weyusya LE71700602015075SG100. All were down-
Joseph, Community Development Officer, Headshot photograph credits: Odada by AC- loaded from http://glovis.usgs.gov/ The bands
Manafwa, and Mr. Ronaldo Wasukira, Phys- CESS ©2015 and Anukur by IUCN ©2015. used in this composite are 7, 5, and 3.
ical Planner, Manafwa, from the Manafwa
Catchment, Manafwa District. The composite satellite image was made
OVERVIEW MAPS slightly transparent and overlaid on a shading
To Mr. Bintoora K. Adonia from the Uganda derived from SRTM1 (NASA) to enhance
OVERVIEW MAPS
Wildlife Authority. appearance.
pp. 8-9
To Ms. Christine Omuombo, the University of International Boundary: Global Administrative
Terrain in the background image is based
Nairobi and ACCESS. Areas Dataset GADM (2015) modified with
on shading generated from SRTM1 eleva-
auxiliary data taken from satellite images and
tion data (NASA). No-data voids were filled
To Mr. James Omoding, Programme Officer, other local sources.
through spatial interpolation and the use of
IUCN, Calvin Odur, Project Assistant/Driver,
auxiliary datasets.
and John P. Owino, Programme Officer, IUCN PHYSICAL MAP
from the IUCN. pp. 14-15
MT. ELGON IN EAST AFRICA
pp. 10-11
And to Mukambi Titus, Musawa David, Barna- Topography: Shading generated from SRTM1
bas Makosi, G.M. Wasike, and Weboya Sirati. elevation data (NASA). No-data voids were
Topography: Shading generated from SRTM3
filled through spatial interpolation and the
elevation data (NASA). No-data voids were
Thanks are also due to Dr. Rene Dommain for use of auxiliary datasets.
filled through spatial interpolation and the
reviewing the text.
use of auxiliary datasets.
Surface Water: Lakes from SRTM Water Body
Note on Geospatial Data Dataset (SWBD-NASA). Rivers derived from
Vegetation: Derived from Hansen/UMD/
SRTM1 (NASA) with groundtruthing from
Google/USGS/NASA (2013) originally from
The following sections provide a summary auxiliary images.
Hansen, M. C., P. V. Potapov, R. Moore, M.
view of data sources and, where applicable,
Hancher, S. A. Turubanova, A. Tyukavina, D.
copyrights, for the maps and images con- International Boundary: Global Administrative
Thau, S. V. Stehman, S. J. Goetz, T. R. Love-
tained in this atlas. For those wanting a more Areas Dataset GADM (2015) modified with
land, A. Kommareddy, A. Egorov, L. Chini,
detailed description of the methods used auxiliary data taken from satellite images and
C. O. Justice, and J. R. G. Townshend. 2013.
in this work, please refer to the following other local sources.
“High-Resolution Global Maps of 21st-Cen-
publication:
tury Forest Cover Change.” Science 342 (15
Place and Feature Names: National Geo-
November): 850–53. Data available on-line
ACCESS (2014) Technical Annex on Geospa- spatial-Intelligence Agency GEOnet Names
from: http://earthenginepartners.appspot.
tial Data and Methods for the USAID/IUCN Server (NGA-GNS) (2015) and local stakehold-
com/science-2013-global-forest.
Project on Implementing a Resilience Frame- er input.
work for Climate Change Adaptation in the
Surface Water: Lakes from SRTM Water
Mt. Elgon Region of the Lake Victoria Basin.
Body Dataset (SWBD-NASA). Rivers derived
Prepared by African Collaborative Centre for
from SRTM3 (NASA) with groundtruthing
Earth System Science (ACCESS). Authors:
 Acknowledgments 49

POLITICAL MAP IIASA, Laxenburg, Austria. C. O. Justice, and J. R. G. Townshend. 2013.

pp. 16-17 “High-Resolution Global Maps of 21st-Cen-
Topography: Shading generated from SRTM1 tury Forest Cover Change.” Science 342 (15
Topography: Shading generated from SRTM1 elevation data (NASA). No-data voids were November): 850–53. Data available on-line
elevation data (NASA). No-data voids were filled through spatial interpolation and the from: http://earthenginepartners.appspot.
filled through spatial interpolation and the use of auxiliary datasets. com/science-2013-global-forest.
use of auxiliary datasets.
Place and Feature Names: National Geo- Place and Feature Names: National Geo-
Surface Water: Rivers derived from SRTM1 spatial-Intelligence Agency GEOnet Names spatial-Intelligence Agency GEOnet Names
(NASA) with groundtruthing from auxiliary Server (NGA-GNS) (2015) and local stakehold- Server (NGA-GNS) (2015) and local stakehold-
images. er input. er input.

Boundaries: Global Administrative Areas TOPOGRAPHY SURFACE WATER

Dataset GADM (2015) modified with auxiliary pp. 24-25 pp. 28-29
data taken from satellite images and rivers
derived from SRTM1 and SRTM3 as well as Figure on Slope: Data derived from SRTM1 Photograph of Mountain Stream by Gabriel
other local sources. (NASA). No-data voids were filled through Gersch ©2015. http://www.gabriel-gersch.
spatial interpolation and the use of auxiliary com/
Place and Feature Names: National Geo- datasets. Projection GCS-WGS1984. Z-factor
spatial-Intelligence Agency GEOnet Names of 0.00000898 used for calculation. Figure on Mt. Elgon as a Water Tower: Based
Server (NGA-GNS) (2015) and local stakehold- on hydrological analysis of SRTM3 (NASA)
er input. Figure on Topographic Profile. Data derived data. Waterbodies from SRTM Water Body
from SRTM1 (NASA). No-data voids were Dataset (SWBD-NASA).
THEMATIC MAPS filled through spatial interpolation and the
use of auxiliary datasets. A line was drawn Photograph of Jackson's Pool by Arjen de
THEMATIC MAPS from the approximate middle of Kitale across Ruiter ©2011.
pp. 18-19 to the middle of Mbale.
Topography: Shading generated from SRTM1
Terrain in the background image is based Photograph of Mt. Elgon. Thomas Ballatore elevation data (NASA). No-data voids were
on shading generated from SRTM1 eleva- ©2015. filled through spatial interpolation and the
tion data (NASA). No-data voids were filled use of auxiliary datasets.
through spatial interpolation and the use of Topography: Shading generated from SRTM1
auxiliary datasets. elevation data (NASA). No-data voids were Surface Water: Lakes from SRTM Water Body
filled through spatial interpolation and the Dataset (SWBD-NASA). Rivers derived from
GEOLOGY use of auxiliary datasets. SRTM1 (NASA) with groundtruthing from
pp. 20-21 auxiliary images. Seasonally flooded areas de-
International Boundary: Global Administrative rived from swampy and seasonally flooded ar-
Photograph of summit of Mt. Elgon by Gabri- Areas Dataset GADM (2015) modified with eas in VECEA by van Breugel P, Kindt R, Lillesø
el Gersch ©2015. http://www.gabriel-gersch. auxiliary data taken from satellite images and JPB, Bingham M, Demissew S, Dudley C,
com/ other local sources. Friis I, Gachathi F, Kalema J, Mbago F, Moshi
HN, Mulumba, J, Namaganda M, Ndangalasi
Photograph of elephant tusking marks in Place and Feature Names: National Geo- HJ, Ruffo CK, Védaste M, Jamnadass R and
Kitum Cave by Donald McFarlane ©2015. spatial-Intelligence Agency GEOnet Names Graudal L (2012) Potential Natural Vegetation
Server (NGA-GNS) (2015) and local stakehold- Map of Eastern Africa (Ethiopia, Kenya, Mala-
Geology: Digitized from the African Karten- er input. wi, Rwanda, Tanzania, Uganda and Zambia).
werk Geological map of East Africa pro- Version 1.1. Forest and Landscape (Denmark)
duced by U. Freitag and others (2001) at VEGETATION and World Agroforestry Centre (ICRAF). Avail-
1:1,000,000 scale. pp. 26-27 able at: http:vegetationmap4africa.org

Topography: Shading generated from SRTM1 Photograph of Alpine Vegetation Near the International Boundary: Global Administrative
elevation data (NASA). No-data voids were Summit by Arjen de Ruiter ©2011. Areas Dataset GADM (2015) modified with
filled through spatial interpolation and the auxiliary data taken from satellite images and
use of auxiliary datasets. Photograph of Forest Converted to Farmland other local sources.
by Thomas Ballatore ©2015.
International Boundary: Global Administrative Place and Feature Names: National Geo-
Areas Dataset GADM (2015) modified with Photograph of Grass in the Summit Crater by spatial-Intelligence Agency GEOnet Names
auxiliary data taken from satellite images and Gabriel Gersch ©2015. http://www.gabri- Server (NGA-GNS) (2015) and local stakehold-
other local sources. el-gersch.com/ er input.

Place and Feature Names: National Geo- Vegetation: Potential vegetation types PRECIPITATION
spatial-Intelligence Agency GEOnet Names derived from VECEA by van Breugel P, pp. 29-30
Server (NGA-GNS) (2015) and local stakehold- Kindt R, Lillesø JPB, Bingham M, Demissew
er input. S, Dudley C, Friis I, Gachathi F, Kalema J, Images of Mt. Elgon Clouds Throughout the
Mbago F, Moshi HN, Mulumba, J, Namagan- Year: All available Landsat 8 images from
SOILS da M, Ndangalasi HJ, Ruffo CK, Védaste M, April 2014 to March 2015 downloaded from
pp. 22-23 Jamnadass R and Graudal L (2012) Potential http://glovis.usgs.gov/ and cropped to a
Natural Vegetation Map of Eastern Africa uniform extent.
Photograph of Bududa landslide by Thomas (Ethiopia, Kenya, Malawi, Rwanda, Tanzania,
Ballatore ©2015. Uganda and Zambia). Version 1.1. Forest and Figure on Intra-annual Variation in Precipi-
Landscape (Denmark) and World Agroforestry tation: Data for Kitale and Mbale taken from
Photograph of outlet of crater by Matthias Centre (ICRAF). Available at: http:vegetation- closest pixels in WorldClim 2.5 arc-minute
Gehringer ©2015. map4africa.org dataset for the period 1950-2000 (http://bio-
geo.ucdavis.edu/data/climate/worldclim/1_4/
Soils: Based on data from Harmonized World Tree Cover: Derived from Hansen/UMD/ grid/cur/prec_2-5m_esri.zip).
Soil Database ©2008-2012 FAO, IIASA, ISRIC, Google/USGS/NASA (2013) originally from
ISSCAS, JRC. Citation: FAO/IIASA/ISRIC/ Hansen, M. C., P. V. Potapov, R. Moore, M. Precipitation: Mean Annual Precipitation for
ISSCAS/JRC, 2012. Harmonized World Soil Hancher, S. A. Turubanova, A. Tyukavina, D. the period 1950-2000 from WorldClim 30
Database (version 1.2). FAO, Rome, Italy and Thau, S. V. Stehman, S. J. Goetz, T. R. Love- arc-second dataset (http://www.worldclim.
land, A. Kommareddy, A. Egorov, L. Chini, org/tiles.php?Zone=37) described in Hijmans,
50 End Matter

R.J., S.E. Cameron, J.L. Parra, P.G. Jones ©2015. (as of 2015). The corresponding new/old
and A. Jarvis, 2005. Very high resolution names are as follows: Amagoro/Teso North,
interpolated climate surfaces for global land Photograph of Giant Lobelia by Gabriel Amakura/Teso South, Cheptais/Mt. Elgon,
areas. International Journal of Climatology Gersch ©2015. http://www.gabriel-gersch. Kacheliba/Pokot North, Kanduyi/Bungoma
25: 1965-1978. com/ South, Kapenguria/West Pokot, Kapsokwony/
Mt. Elgon, Kimilili/Bungoma North, Kwanza/
Topography: Shading generated from SRTM1 Photograph of Elephants by Ian Redmond Kwanza, Lugari/Lugari, Malava/Kakamega
elevation data (NASA). No-data voids were ©2015. North, Nambale/Busia, Sirisia/Bungoma
filled through spatial interpolation and the West, Tongareni/Bungoma North, Trans
use of auxiliary datasets. Protected Areas: A high resolution image of Nzoia West/Trans Nzoia West, and Webuye/
Figure 1 in Sassen et al. 2013 (Sassen, M, D Bungoma East. Major boundary changes
International Boundary: Global Administrative Sheil, KE Giller, CJF ter Braak (2013) Complex include the following: Mt. Elgon was divided
Areas Dataset GADM (2015) modified with contexts and dynamic drivers: Understand- into Cheptais and Kapsokwony, Bungoma
auxiliary data taken from satellite images and ing four decades of forest loss and recovery North was divided into Kimilili and Tongareni.
other local sources. in an East African protected area. Biological Furthermore, Kacheliba is only part of the
Conservation 159: 257-268) was scanned previous Pokot North so the figure in the table
Place and Feature Names: National Geo- and georeferenced with a series of tie points is overestimated.
spatial-Intelligence Agency GEOnet Names linking locations on the figure with known
Server (NGA-GNS) (2015) and local stakehold- locations for the five protected areas on the Population: Population density grids at 100 m
er input. mountain. Other protected areas were taken resolution from WorldPop data as described in
from the World Database on Protected Areas Tatem AJ, Gething PW, Bhatt S, Weiss D and
TEMPERATURE (WDPA) at http://www.protectedplanet.net/ Pezzulo C (2013) Pilot high resolution poverty
pp. 31-32 maps, University of Southampton/Oxford.
Topography: Shading generated from SRTM1 Available at http://www.worldpop.org.uk/
Thermal Image: The satellite image presented elevation data (NASA). No-data voids were data/data_sources/
here is based on data from Band 6 (thermal filled through spatial interpolation and the
emissions) of the Landsat 7 ETM+ sensor and use of auxiliary datasets. Topography: Shading generated from SRTM1
is depicted using a “temperature” color ramp. elevation data (NASA). No-data voids were
The image is a composite of four recent, Surface Water: Lakes from SRTM Water Body filled through spatial interpolation and the
cloud-free Landsat 7 ETM+ (NASA) images Dataset (SWBD-NASA). Rivers derived from use of auxiliary datasets.
from 2015. The spatial extent is large and SRTM1 (NASA) with groundtruthing from
requires two scenes (Path 170/Row 59 and auxillary images. Seasonally flooded areas Administrative Boundaries: Global Adminis-
Path 170/Row 60) for full coverage. Fur- derived from classification of Landsat 8 image trative Areas Dataset GADM (2015) modified
thermore, the Landsat 7 ETM+ sensor does LC81700592014256LGN00 of 14 September with auxiliary data taken from satellite imag-
not have complete coverage within a scene 2014 downloaded from http://glovis.usgs. es and other local sources.
due to the Scan Line Corrector failure. Two gov/
scene from very close in time (28 Feb 2015 Place and Feature Names: National Geo-
and 16 Mar 2015) were merged using a local International Boundary: Global Administrative spatial-Intelligence Agency GEOnet Names
histogram matching algorithm to produce Areas Dataset GADM (2015) modified with Server (NGA-GNS) (2015) and local stakehold-
this apparently “seemless” image. The four auxiliary data taken from satellite images and er input.
scenes used were LE71700592015059SG100, other local sources.
LE71700592015075SG100, ROADS AND FIRES
LE71700602015059SG100, and Place and Feature Names: National Geo- pp. 37-38
LE71700602015075SG100. All were down- spatial-Intelligence Agency GEOnet Names
loaded from http://glovis.usgs.gov/ Server (NGA-GNS) (2015) and local stakehold- Photograph of Unpaved Road by Lydia Olaka
er input. ©2015.
Figure on Intra-annual Variation in Mean
Temperature: Data for Kitale and Mbale taken POPULATION Images of Fires from Space from Land-
from closest pixels in WorldClim 2.5 arc-min- pp. 35-36 sat 7 ETM+ image on 16 Mar 2015
ute dataset for the period 1950-2000 (http:// (LE71700592015075SG100 downloaded from
biogeo.ucdavis.edu/data/climate/world- Figure on Languages based on data from http://glovis.usgs.gov/). The bands used in
clim/1_4/grid/cur/tmean_2-5m_esri.zip). Global Mapping International's World Lan- the left-hand image are 7, 5, and 3, whereas
guage Mapping System ©GMI 2015 (http:// the right-hand image is based solely on Band
Temperature: Mean Annual Temperature for www.worldgeodatasets.com/language/). 1.
the period 1950-2000 from WorldClim 30
arc-second dataset (http://www.worldclim. Table with Uganda Census data based on Na- Figure on Fires throughout the Year based on
org/tiles.php?Zone=37) described in Hijmans, tional Population and Housing Census 2014: data set that was provided by LANCE FIRMS
R.J., S.E. Cameron, J.L. Parra, P.G. Jones Provisional Results, November 2014, Revised operated by NASA/GSFC/ESDIS with funding
and A. Jarvis, 2005. Very high resolution Edition, Uganda Bureau of Statistics, available provided by NASA/HQ. NASA Near Real-Time
interpolated climate surfaces for global land at http://www.ubos.org/onlinefiles/uploads/ and MCD14DL MODIS Active Fire Detections
areas. International Journal of Climatology ubos/NPHC/NPHC%202014%20PROVISION- (TXT format). Data set. Available on-line
25: 1965-1978. AL%20RESULTS%20REPORT.pdf https://earthdata.nasa.gov/active-fire-data-
tab-content-6
International Boundary: Global Administrative Table with Kenya Census data based on
Areas Dataset GADM (2015) modified with The 2009 Kenya Population and Housing Roads: Road data downloaded from Open-
auxiliary data taken from satellite images and Census, Volume 1C:Population Distribution StreetMap on 31 Mar 2015. Available at
other local sources. by Age, Sex and Administrative Units, Kenya https://www.openstreetmap.org/
National Bureau of Statistics, August, 2010,
Place and Feature Names: National Geo- available at http://www.knbs.or.ke/index. Fires: Data on thermal anomolies from 01
spatial-Intelligence Agency GEOnet Names php?option=com_phocadownload&view=cat- Jan 2014 to 27 Mar 2015 provided by LANCE
Server (NGA-GNS) (2015) and local stakehold- egory&id=109:population-and-housing-cen- FIRMS operated by NASA/GSFC/ESDIS with
er input. sus-2009&Itemid=599. funding provided by NASA/HQ. NASA Near
Real-Time and MCD14DL MODIS Active Fire
PROTECTED AREAS Note on Kenyan district names and bound- Detections (TXT format). Data set. Available
pp. 33-34 aries: After the 2009 census, major changes on-line https://earthdata.nasa.gov/active-
were made in Kenyan district names and, in fire-data-tab-content-6
Photograph of Zebras by Lydia Olaka ©2015. some cases, the boundaries themselves. The
names given in the table on p. 36 and map on Topography: Shading generated from SRTM1
Photograph of Antelope by Josep M. Gracia p. 37 are the latest available official names elevation data (NASA). No-data voids were
 Acknowledgments 51

filled through spatial interpolation and the use LANDSLIDE RISK 113, pp. 63-79.
of auxiliary datasets. pp. 46-47
Claessens L., A. Knapen, M.G. Kitutu, J. Po-
Place and Feature Names: National Geospa- Maps on Slope Factor, Plan Curvature, and esen, J.A. Deckers (2007) Modelling landslide
tial-Intelligence Agency GEOnet Names Server Aspect based on analysis done using SRTM3 hazard, soil redistribution and sediment yield
(NGA-GNS) (2015) and local stakeholder input. (NASA) data. of landslides on the Ugandan footslopes of
Mount Elgon. Geomorphology 90: 23-35.
SPECIAL TOPICS Map on Precipitation Factor based on : World-
Clim 30 arc-second dataset (http://www. Cohen, S. (2014) Reaching the Peak: A docu-
SPECIAL TOPICS worldclim.org/tiles.php?Zone=37) described mentary filmed, edited and produced by Stuart
pp. 40-41 in Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Cohen for the Uganda Wildlife Authority
Jones and A. Jarvis, 2005. Very high resolution UWA, in association with the IUCN. Available
Terrain in the background image is based on interpolated climate surfaces for global land at https://www.youtube.com/watch?v=LoY-
shading generated from SRTM1 elevation data areas. International Journal of Climatology 25: IHtKwfRY.
(NASA). No-data voids were filled through 1965-1978.
spatial interpolation and the use of auxiliary Davenport, T., P. Howard, and C. Dickenson
datasets. Map on Land Cover Factor based on anal- (1996) Mount Elgon National Park Biodiversity
ysis described on pp. 42-43 of this atlas Report. Forest Department, Kampala, Uganda.
LAND COVER CHANGE which made use of the following satellite
pp. 42-43 images: Landsat 1 MSS on 01 February 1973 Davies, K. A. (1952) The building of Mount
(LM1182057007303290), Landsat 7 ETM+ on Elgon. Geological Survey of Uganda, Memoir
Satellite Images: The image from 1973 05 January 2013 (LE71700592013005ASN00) VII Geological Survey Department, Entebbe.
is Landsat 1 MSS on 01 February 1973 and Landsat 7 ETM+ on 21 January 2013
(LM1182057007303290) and the 2013 images (LE71700592013021ASN00) all downloaded Davies, K. A. (1957) The Building of Mount
are from Landsat 7 ETM+ on 05 January 2013 from http://glovis.usgs.gov/ Elgon (East Africa). The geological survey of
(LE71700592013005ASN00) and 21 Janu- Uganda, Entebbe.

ary 2013 (LE71700592013021ASN00) all Map on Sum of All Risk Factors based on data
downloaded from http://glovis.usgs.gov/. For presented on p. 46 and described in detail in Hancock, G.L.R. and W.W. Soundy (1929)
complete details on image preparation and ACCESS (2014) Technical Annex on Geospatial Notes on the fauna and flora of Northern Bugi-
processing, please see ACCESS (2014) Tech- Data and Methods for the USAID/IUCN Project shu and Masaba (Elgon). Journal of The East
nical Annex on Geospatial Data and Methods on Implementing a Resilience Framework for Africa and Uganda Natural History Society:
for the USAID/IUCN Project on Implementing Climate Change Adaptation in the Mt. Elgon 165-183.
a Resilience Framework for Climate Change Region of the Lake Victoria Basin. Prepared
Adaptation in the Mt. Elgon Region of the by African Collaborative Centre for Earth Hedberg, O. (1951) Vegetation belts of the
Lake Victoria Basin. Prepared by African System Science (ACCESS). Authors: Thomas East-African mountains. Svensk Botanisk
Collaborative Centre for Earth System Science J. Ballatore and Lydia Olaka. Edited by Eric Tidskrift, 45, 141-202.
(ACCESS). Authors: Thomas J. Ballatore and Odada and Daniel Olago. Published by USAID/
Lydia Olaka. Edited by Eric Odada and Daniel IUCN. 45p. Hitimana, J., J. Legilisho Kiyiapi, and J. Thairu
Olago. Published by USAID/IUCN. 45p. Njunge (2004) Forest structure character-
Topography: Shading generated from SRTM3 istics in disturbed and undisturbed sites of
Topography: Shading generated from SRTM1 elevation data (NASA). No-data voids were Mt. Elgon Moist Lower Montane Forest,
elevation data (NASA). No-data voids were filled through spatial interpolation and the use western Kenya. Forest Ecology and Man-
filled through spatial interpolation and the use of auxiliary datasets. agement 194(1-3):269-291. http://dx.doi.
of auxiliary datasets. org/10.1016/j.foreco.2004.02.025
BACK MATTER
Administrative Boundaries: Global Adminis- Knapen A., M.G. Kitutu, J. Poesen, W. Breu-
trative Areas Dataset GADM (2015) modified BACK COVER gelmans, J. Deckers, A. Muwanga (2006)
with auxiliary data taken from satellite images p. 54 Landslides in a densely populated county at
and other local sources. the footsteps of Mount Elgon (Uganda): Char-
Background photograph credit: Gabriel Gersch acteristics and causal factors. Geomorphology
CLIMATE CHANGE ©2015. http://www.gabriel-gersch.com/ 73: 149-165.
pp. 44-45
Marshall, N.T. and Jenkins, M. (1994) Hard
LITERATURE CITED Times for Hardwood: Indigenous Timber
Temperature and Precipitation: Projections
for 2050 (period 2040-2059) derived from and the Timber Trade in Kenya. A TRAFFIC
ACCESS (2014) Technical Annex on Geospatial
Coupled Model Intercomparison Project-Phase Network Report East ⁄ Southern Africa,
Data and Methods for the USAID/IUCN Project
5 (CMIP5) of the World Climate Research Pro- Cambridge.
on Implementing a Resilience Framework for
gramme at http://cmip-pcmdi.llnl.gov/cmip5/ Climate Change Adaptation in the Mt. Elgon
used in the Intergovernmental Panel on Cli- Mugagga, F., Kakembo, V., Buyinza, M. (2011)
Region of the Lake Victoria Basin. Prepared
mate Change (IPCC) Fifth Assessment Report. A characterization of the physical properties
by African Collaborative Centre for Earth
Downscaled 2.5 arc-minute grids downloaded of soil and the implications for landslide occur-
System Science (ACCESS). Authors: Thomas
from WorldClim at http://www.worldclim.org/ rence on the slopes of Mount Elgon, Eastern
J. Ballatore and Lydia Olaka. Edited by Eric
cmip5_2.5m. Calculation of change to 2050 Uganda. Natural Hazards. doi:10.1007/
Odada and Daniel Olago. Published by USAID/
used the 2.5 arc-minute “baseline” 1950-2000 s11069-011-9896-3.
IUCN. 45p. (Available at: http://cmsdata.iucn.
data from WorldClim available at http://bio- org/downloads/tech_annex_report_web_ver-
geo.ucdavis.edu/data/climate/worldclim/1_4/ Mugagga, F., V. Kakembo, M. Buyinza (2012)
sion.pdf)
grid/cur/bio_2-5m_esri.zip. WorldClim is Land use changes on the slopes of Mount
described in Hijmans, R.J., S.E. Cameron, J.L. Elgon and the implications for the occurrence
Bellenger, H., E. Guilyardi, J. Leloup, M.
Parra, P.G. Jones and A. Jarvis, 2005. Very of landslides. Catena 90: 39-46.
Lengaigne, and J. Vialard (2014) ENSO repre-
high resolution interpolated climate surfaces sentation in climate models: from CMIP3 to
for global land areas. International Journal of Petursson, J. G., P. Vedeld, and A. Vatn (2013)
CMIP5. Climate Dynamics 42(7): 1999-2018.
Climatology 25: 1965-1978. Going transboundary? An institutional analysis
of transboundary protected area manage-
Bowell, R.J., A. Warren, and I. Redmond (1996)
Topography: Shading generated from SRTM1 ment challenges at Mt Elgon, East Africa.
Formation of cave salts and utilization by
elevation data (NASA). No-data voids were Ecology and Society 18(4): 28. http://dx.doi.
elephants in the Mount Elgon region, Kenya.
filled through spatial interpolation and the use org/10.5751/ES-05729-180428
In: Appleton, J.D., R. Fuge, and G.J.H. McCall
of auxiliary datasets. (Eds) Environmental Geochemistry and Health.
Redmond, I. (1982) Salt-mining elephants of
Geological Society Special Publication No.
Mount Elgon. Wildlife 24: 288-293.
52 End Matter

Redmond, I. (1992) Erosion by elephants. In:

Shoshani, J. (ed.) Elephants, majestic crea-
tures of the wild. RD Press, Oxford, 128-130.

Redmond, Ian. Personal Communication on 02

July 2015.

Sassen, M., D. Sheil, K.E. Giller, C.J.F. ter

Braak (2013) Complex contexts and dynamic
drivers: Understanding four decades of forest
loss and recovery in an East African protected
area. Biological Conservation 159: 257-268.

Scott, P. (1994) An assessment of natural

resource use by communities from Mt. Elgon
National Park. UNDP/Technical Report No. 15,
1994.

Street-Perrott, F.A., P.A. Barker, D.L. Swain,

K.J. Ficken, M.J. Wooller, D.O. Olago, Y-S.
Huang (2007) Late Quaternary changes in
ecosystems and carbon cycling on Mt. Kenya,
East Africa: a landscape-ecological perspec-
tive based on multi-proxy lake-sediment
influxes. Quaternary Science Reviews 26(13-
14):1838-1860.

Synnott, T.J. (1968) Working plan for Mount

Elgon central forest reserve. 1st revision. Pe-
riod 1968-1978. Uganda forest department,
Entebbe.

Tokro G.O., Bintoora A., Wanyama. F., Na-

kidde A., Nyadru S, Diisi. J, Odhiambo C.N,
Ochieng’ C.A, Wagura.P, Galima S. (2014)
Development of Trans-Boundary Ecotourism
for Mt Elgon Ecosystem, a GIS Approach
Draft regional report, 2014.

Van Heist, M. (1994) Accompanying report

with the Land unit map of Mount Elgon
National Park. Mount Elgon conservation and
development project, Mbale, Uganda.

World Bank (2015) Population Growth Annual

Percent. Available at http://data.worldbank.
org/indicator/SP.POP.GROW?cid=GPD_2