Content uploaded by Anson W Mackay
Author content
All content in this area was uploaded by Anson W Mackay on Nov 02, 2020
Content may be subject to copyright.
- 1 -
Environmental Change
Research Centre
Research Report No.113
Climate Change and
the Aquatic E
cosystems of the
Rwenzori Mountains, Uganda
Final
Report
to the Royal Geographical Society
R.G. Taylor, N.L. Rose, A.W. Mackay,
V. Panizzo, L. Mileham, I. Ssemmanda,
C. Tindimugaya, B. Nakileza, A. Muwanga
& J. Hau
2007
ISSN: 1366-7300 Environmental Change Research Centre
University College London
Pearson Building, Gower St
London, WC1E 6BT
- 1 -
Climate Change and the Aquatic Ecosystems
of the Rwenzori Mountains, Uganda
Final Report to the Royal Geographical Society
R. G. Taylor1, N.L. Rose1, A.W. Mackay1,
V. Panizzo1, L. Mileham1, I. Ssemmanda2,
C. Tindimugaya3, B. Nakileza4 A. Muwanga2
& J. Hau1
2007
1. Environmental Change Research Centre
University College London
Pearson Building, Gower St.
London, WC1E 6BT
United Kingdom
2. Department of Geology
Makerere University
P.O. Box 7062, Kampala
Uganda
3. Water Resources Management Department
Directorate of Water Development
P.O. Box 19, Entebbe
Uganda
4. Department of Geography
Makerere University
P.O. Box 7062, Kampala
Uganda
- 1 -
TABLE OF CONTENTS
1 Introduction.........................................................................................................................................7
1.1 Project rationale and objectives ...................................................................................................7
1.2 Rwenzori Mountains National Park..............................................................................................9
1.3 Project planning .........................................................................................................................10
1.4 Project team...............................................................................................................................11
1.4.1 June 2003 expedition.............................................................................................................................................. 11
1.4.2 January-February 2005 expedition.......................................................................................................................... 12
1.4.3 Project advisory panel.............................................................................................................................................13
1.5 Report layout..............................................................................................................................13
2 Recent glacial recession in the Rwenzori Mountains and rising air temperatures ...................14
2.1 Introduction ................................................................................................................................14
2.2 Field and satellite mapping ........................................................................................................14
2.3 Results .......................................................................................................................................17
2.3.1 Field surveys of glacial extent................................................................................................................................. 17
2.3.2 Mapping glacial cover by remote sensing................................................................................................................ 18
2.4 Meteorological trends.................................................................................................................21
2.5 Conclusions................................................................................................................................24
3 Climatological implications of glacial recession in the Rwenzori Mountains............................25
3.1 Introduction ................................................................................................................................25
3.2 Relationship between Ta at the surface and mid troposphere...................................................25
3.3 Trends in atmospheric humidity .................................................................................................26
3.4 Conclusions................................................................................................................................29
4 Hydrological implications of glacial recession .............................................................................30
4.1 Introduction ................................................................................................................................30
4.2 Hydrology of the Rwenzori Mountains .......................................................................................31
4.2.1 Quaternary glaciations............................................................................................................................................31
4.2.2 Modern glacial recession.........................................................................................................................................32
4.2.3 Glacial meltwater drainage...................................................................................................................................... 34
4.2.4 Precipitation............................................................................................................................................................35
4.3 Methodology...............................................................................................................................37
4.4 Longitudinal trends in alpine riverflow........................................................................................39
4.4.1 River Mubuku.......................................................................................................................................................... 39
4.4.2 River Nyamagasani.................................................................................................................................................40
4.4.3 River Semliki........................................................................................................................................................... 40
4.5 Altitudinal trends in riverflow from spot measurements .............................................................41
4.6 Meteorological trends over the last century...............................................................................42
4.7 Hydrological trends over the two centuries................................................................................43
4.8 Concluding Discussion...............................................................................................................44
5 Recent changes in aquatic productivity of Lake Bujuku..............................................................46
5.1 Introduction ................................................................................................................................46
5.2 Study site ...................................................................................................................................46
5.3 Methodology...............................................................................................................................48
5.3.1 Field Methods .........................................................................................................................................................48
5.3.2 Laboratory methods................................................................................................................................................49
5.4 Results .......................................................................................................................................51
5.4.1 Contemporary diatom ecology.................................................................................................................................51
5.4.2 Chronology ............................................................................................................................................................. 52
5.4.3 Organic geochemistry............................................................................................................................................. 52
5.4.4 % TOC.................................................................................................................................................................... 52
5.4.5 C/N ratios................................................................................................................................................................54
5.4.6 Stable isotopes of carbon........................................................................................................................................ 54
5.4.7 Diatoms ..................................................................................................................................................................54
5.4.8 Pollen......................................................................................................................................................................56
5.5 Discussion..................................................................................................................................56
5.5.1 Palaeolimnological reconstruction over period of glacial retreat...............................................................................59
5.6 Conclusions................................................................................................................................61
6 Lake-sediment archives of atmospherically deposited pollutants in the Rwenzori Mountains62
6.1 Introduction ................................................................................................................................62
6.1.1 Atmospheric pollution..............................................................................................................................................62
6.1.2 Trace metals........................................................................................................................................................... 62
6.1.3 Fly ash.................................................................................................................................................................... 63
6.2 Study areas................................................................................................................................64
6.3 Methodology...............................................................................................................................65
- 2
6.3.1 Sediment-core collection and extraction.................................................................................................................. 65
6.3.2 Lithostratigraphy .....................................................................................................................................................65
6.3.3 Sediment-core dating.............................................................................................................................................. 65
6.3.4 X-ray fluorescence (XRF)........................................................................................................................................ 65
6.3.5 Atomic Absorption Spectrometry (AAS)...................................................................................................................66
6.3.6 SCP analysis...........................................................................................................................................................66
6.3.7 Interpretative techniques.........................................................................................................................................66
6.4 Results .......................................................................................................................................67
6.4.1 Percentage loss on ignition, dry weight and wet density.......................................................................................... 67
6.4.2 Age - depth profiles of sediment cores.................................................................................................................... 68
6.4.3 Sediment-core trace-element profiles......................................................................................................................68
6.4.4 SCPs ......................................................................................................................................................................72
6.5 Discussion and analysis.............................................................................................................73
6.5.1 Application of Hilton’s model................................................................................................................................... 73
6.5.2 Assessment of anthropogenic origin of trace metals................................................................................................75
6.5.3 Mercury profiles ......................................................................................................................................................76
6.5.4 Trace metal flux profiles.......................................................................................................................................... 78
6.5.5 Trace-metals concentrations in moss...................................................................................................................... 78
6.5.6 Possible sources of trace metals.............................................................................................................................82
6.6 Conclusions................................................................................................................................84
7 Overall project findings....................................................................................................................85
7.1 magnitude of current glacial recession ......................................................................................85
7.2 impact of glacial recession on alpine riverflow...........................................................................85
7.3 recent environmental change from observational datasets and sediment-core archives..........85
8 Dissemination & knowledge transfer .............................................................................................87
8.1 stakeholder meetings.................................................................................................................87
8.2 scientific publications .................................................................................................................87
8.3 conference presentations and speaking invitations...................................................................88
8.4 popular press .............................................................................................................................88
8.5 student dissertations ..................................................................................................................89
9 Acknowledgements..........................................................................................................................89
10 REFERENCES ...................................................................................................................................90
- 3
LIST OF FIGURES
Figure 1. The Central Rwenzori Massif showing alpine lakes, wetlands, streams and the extent of
glacial snow cover in 1955 and 1990.
Figure 2. A conceptual representation of the uplift of the Rwenzori horst tilting upwards from the ESE
to WNW with approximate ages drawn from Taylor and Howard (1998)).
Figure 3. LandSat7 (ETM+) optical satellite image of western Uganda on January 31, 2003 featuring
the Rwenzori Mountains to the north and Lake Edward to the south.
Figure 4. (a) Map of Uganda showing the location of the Rwenzori Mountains and meteorological
stations. (b) The Central Rwenzori Massif showing indicator glaciers (Elena and Speke) and
the extent of glacial cover in 1955 and 1990 (Kaser and Osmaston, 2002). (c) Changes in the
terminus and areal extent of the Elena Glacier from 1906 to 2005. (d) Subsample of the
NDSI-classified LandSat7 ETM+ satellite image on January 31, 2003 for Mount Speke
superimposed on mapped glacier extents in 1906, 1955 and 1990 (Kaser and Osmaston,
2002).
Figure 5. Photograph of Rwenzori Mountaineering Service guide, Baluku Josephat, leading the
research team on a trek across the Stanley Plateau (30 January 2005). Alexandra and
Margherita peaks on Mount Stanley occupy the background.
Figure 6. The changing rate of terminus retreat of the Speke Glacier from 1958 to 2003 is shown in the
inset of photograph (a). Field survey of terminus retreat in 2003, relative to the 1993 marker
(Talks, 1993), is also shown in photograph (a). Photograph (a) is panoramic image taken by
Andrea Fischer on 22 June 2003 whereas the panoramic photograph on 29 January 1990 (b)
is from Kaser and Noggler (1991).
Figure 7. (a) Terrestrial photograph of the Speke Glacier from Mount Stanley in 2005 showing the
steep slopes below the glacier's current and former terminus. (b) NDSI-classified Landsat7
image from 31 January 2003.
Figure 8. A conceptual, cross-sectional representation of changes in the profiles of ‘indicator’ valley
glaciers (shaded): (a) Elena and (b) Speke in the Rwenzori Mountains from 1991 to 2005.
Surface profiles are drawn relative to an air-photograph survey of glacier cover in 1955.
Figure 9. Plot of changes in glacial areal extent on the Central Rwenzori Massif since 1906.
Figure 10. Standardised anomalies in annual mean maximum (solid circles) and mean minimum (open
circles) air temperature observed at meteorological stations in western Uganda and annual
mean temperature from gridded CRU2 climate data over the 20th century.
Figure 11. Longitudinal trends in annual precipitation, plotted as deviations from the mean, for stations
at meteorological stations in western Uganda over the 20th century.
Figure 12. Time series of monthly air temperature anomalies in the lower and middle troposphere from
homogenised radiosonde datasets (HadAT2) at (a) 700hPa and (b) 500hPa for the most
proximate grid cell (35ºE, -2.5ºS) to the Rwenzori Mountains. Bold lines in (a) and (b)
represent the 12-month running mean.
Figure 13. Time series of mean annual anomalies in (a) vapour pressure (1901 to 1995) and (b)
precipitation (1901 to 1998) from gridded CRU TS 2.0 climate data (New et al., 2002) for the
most proximate grid cell (29.5ºE, +0.5ºN) to the Rwenzori Mountains.
Figure 14. Level of Lake Victoria at Jinja (33.2ºE, 0.2ºN) from 1800 to 2005 based on historical
evidence from 1800 to 1896 (Nicholson and Yin, 2001) and monthly observations from 1896
to 2005.
Figure 15. Topographic map of the Rwenzori Mountains showing the main river channels draining alpine
areas and the location of river-gauging and local meteorological stations.
Figure 16. Map of the drainage network for the Rwenzori Mountains. The international border divides
the Republic of Uganda to the east and the Democratic Republic of Congo to the west.
Figure 17. Profile of equilibrium-line-altitudes (ELAs) estimated for former glaciers in the Rwenzori
Mountains (redrawn from Osmaston and Kaser, 2001).
Figure 18. Map of glacial extent and drainage in the Central Rwenzori Massif. Glaciers occur on Mounts
Speke (north), Stanley (west) and Baker (southwest).
Figure 19. Annual distribution in precipitation recorded at Kilembe, Uganda (0º13’N, 30º00’E) from 1949
to 1996.
Figure 20. Plot of trends in observed precipitation and stream discharge with altitude in the Rwenzori
Mountains. The main vegetation zones (ecotones) are indicated for reference.
Figure 21. Photograph of the confluence of the River Mubuku and its main tributary the River Bujuku at
sampling site no. 4.
Figure 22. Longitudinal trends in deviations from mean riverflow observed at the base of the Rwenzori
Mountains from 1952 to 1978.
Figure 23. Plot of changes in glacial cover on the Rwenzori Mountains and Kilimanjaro since 1906.
Figure 24. Regional hydrological trends in lake levels and river discharge since 1800.
Figure 25. Photograph of Lake Bujuku from the Speke Glacier (June 22, 2003).
- 4
Figure 26. Digitised image of Lake Bujuku bathymetry, inflows, outflow and bog region. Numbers refer
to contemporary diatom sampling sites (Table 3) and X refers to core location of Buju3.
Figure 27. (a) 137Cs apex acts as a stratigraphic marker for 1963; (b) plot of age versus depth for Buju3,
based on the simple model of 210Pb dating. Error bars shown are calculated using the 2 SD
error on the gradient of the linear regression fit of ln210Pb excess vs. depth.
Figure 28. Age (depth) profiles of bulk organic carbon analyses: 13C (‰), C/N and % TOC. Zones are
defined by diatom data.
Figure 29. Diatom stratigraphy as percentage abundances for Buju3. Valve concentrations, diatom flux
rates and PCA axes scores 1 and 2 are also shown.
Figure 30. Pollen stratigraphy from Buju3 as percentage abundances. Results are grouped according to
their family for ease of interpretation and displayed with PCA axes scores 1 and 2. Zones
applied are derived from constrained cluster analysis carried out on diatom data to aid
interpretation.
Figure 31. Spheroidal carbonaceous particles (SCP) under an electron microscope.
Figure 32. Map of the location of Lakes Mahoma, Bujuku and Kitandara in the Rwenzori Mountains.
Figure 33. Loss on ignition (LOI), dry weight and wet density profiles for each analysed sediment core
(BUJU3, KITA3, MAHO1).
Figure 34. The age-depth profile of sediments cores.
Figure 35. Depth (age) profiles of trace metal concentrations at Lower Kitandara Lake (KITA2).
Figure 36. Depth (age) profiles of trace metal concentrations at Lake Bujuku (BUJU1).
Figure 37. Depth (age) profiles of trace metal concentrations at Lake Mahoma (MAHO1).
Figure 38. Element concentration correlations and Pearson’s correlation coefficient (r).
Figure 39. (a) Hg profiles; (b) anthropogenic Hg profiles; and (c) Correlation of background Hg and
%LOI.
Figure 40. Depth (age) profiles of trace-metal fluxes for Lower Lake Kitandara.
Figure 41. Depth (age) profiles of trace-metal fluxes for Lake Bujuku.
Figure 42. Depth (age) profiles of trace-metal fluxes for Lake Mahoma.
LIST OF TABLES
Table 1. Areal extent of glacial cover on the Central Rwenzori Massif.
Table 2. Mean, mean minimum, and mean maximum relative humidity (%) from measurements available over
two continuous periods, 1967 to 1974 and 1991 to 2000, at Kasese. Data derive from daily
measurements recorded at 9AM and 3PM. The standard deviation in each mean value is given in
parentheses.
Table 3. Summary of diatom assemblage compositions from contemporary sampling at eight sites around the
shore of Lake Bujuku. Only the most abundant species, and their relative abundances, at each of the
sites is displayed.
Table 4. Elemental concentrations in moss sampled near Lake Bujuku.
Cover photo: Speke Glacier bounded by steep scarps within the Rwenzori Mountains National
Park (Uganda) in June 2003. Distinctive Afroalpine vegetation, Tree Senecio
(Dendrosenecio adnivalis) and Everlasting flower (Helichrysum stuhlmannii),
occupy the foreground. All photos © ECRC.
- 5
EXECUTIVE SUMMARY
The Rwenzori Mountains are home to one of the last remaining tropical icefields
outside of the Andes. Over the last century, equatorial icefields of the East
African highlands have been steadily shrinking but the precise climate tropical
alpine glaciers remain unclear. More than a decade had passed since the last
detailed measurements of glacial cover were made in the Rwenzori Mountains.
Recent evidence from Kilimanjaro suggests that its icecap will disappear entirely
by the year 2020(1). The Rwenzori glaciers contribute meltwater flows to aquatic
ecosystems of the Rwenzori Mountains National Park, a Word Heritage Site
featuring spectacular, rare Afroalpine flora and fauna, and are headwaters of the
River Nile. With the overall aim of assessing the impact of recent climate change
on alpine aquatic ecosystems of the Rwenzori Mountains, a collaborative,
international research team led by the University College London (United
Kingdom) and Makerere University (Uganda), and involving the Institut für
Geographie from the University of Innsbruck (Austria) and Water Resources
Management Department (Uganda) was assembled in order to pursue three
primary scientific objectives:
• to assess the magnitude of current glacial recession;
• to assess the impact of glacial recession on alpine riverflow; and
• to assess recent environmental change from observational datasets and
available, environmental archives stored in lake sediment and glacial ice.
The research was supported by grants from the Royal Geographical Society
(Ralph Brown Award 2003), The Royal Society, University of London (Central
Research Fund, Convocation Trust), University College London (Dean’s Travel
Fund, Department of Geography, The Friends Trust), Quaternary Research
Association, and Earth and Space Foundation and Rwenzori Beverage Company
Limited (Uganda). Institutional support was also provided by the Water
Resources Management Department (Uganda), Makerere University, and
Uganda Wildlife Authority.
Field research was primarily conducted in June 2003. The international research
team comprising 7 researchers and 6 students were supported by 70 porters,
guides and rangers making this the largest scientific expedition in the Rwenzori
Mountains for half a century. A small follow-up expedition mapping glacial cover
and measuring stream discharges was executed in January 2005.
(1) Thompson et al., 2002. Science, Vol. 298, pp. 589-593.
- 6
The project made significant progress toward each of its objectives and led to
several important findings detailed below.
1. The areal extent of glaciers in the Rwenzori Mountains in 2003 deduced
from satellite data and field surveys is 1.0±0.3 km2. Glaciers are receding
at a rate of ~0.7 km2 per decade, consistent with a linear trend observed
over the last century. Assuming present trends continue, glaciers are
predicted to disappear within the next two decades.
2. Deglaciation over the last century is associated rising air temperatures.
Insufficient data exist to quantify the link between changes in climate
variables and glacial mass but there is currently greater evidence of trends
of increasing air temperature than decreasing humidity to explain recent
glacial recession in the Rwenzori Mountains. Changes in humidity and
radiative fluxes associated with rising air temperatures are expected to
have contributed to observed glacial recession.
3. Glacial recession will have a minimal impact on alpine riverflow. Glacial
meltwater flows, based on fluxes during the dry season, constitute a very
small proportion (<0.5%) of the total river discharge realised at the base of
the Rwenzori Mountains.
4. The lake ecology and flora of Afroalpine areas of the Central Rwenzori
Massif have, based on sediment-core archives of diatom and pollen, not
undergone significant changes over the period of deglaciation. There is,
however, evidence of a recent decline in epiphytic habitats and
concomitant increase in algal productivity.
5. Atmospherically deposited mercury, consistent with global trends and
emissions, is detected in lake sediment within alpine environments (i.e.,
Heath-moss Forest zone at 3000 mamsl and Afroalpine zone at 4000
mamsl) in the Rwenzori Mountains. Trace-metal contamination via
atmospheric deposition from more localised sources is observed in the
Heath-moss Forest zone from the mid-1950s and coincides with onset of
copper mining downslope.
Apart from the scientific outputs from this research, the project initiated a new
scientific collaboration in palaeolimnology between the Environmental Change
Research Centre, UCL and Department of Geology, Makerere University which
included a subsequent visit to the UCL by Dr. Ssemmanda in 2004, funded by
the Royal Society, to support palynological research examining Late Holocene
changes in alpine vegetation of the Rwenzori Mountains. On-going research
includes: (i) further analyses of environmental change based on sediment-core
archives from Lakes Kitandara and Mahoma; (ii) a stable-isotope study of
meltwater contributions and the origin of rainy season precipitation; and (iii) an
altitudinal assessment of stream fluxes during the rainy season.
- 7
1 Introduction
1.1 Project rationale and objectives
Alpine aquatic ecosystems in the tropics are in a period of transition. Tropical
alpine wetlands, lakes and streams that are supported by glacial meltwater flows,
are affected by rapid recession of glaciers in response to global climate change
(Thompson, 2000). Glaciers form important reservoirs of fresh water that store
seasonal inputs of precipitation and release meltwaters during drier periods.
They consequently serve the vital ecological function of regulating alpine
streamflow and water levels in lakes and wetlands.
On the Rwenzori Mountains (0º10’ to 0º30’N, 29º50’ to 30º00’E), which straddle
the border between the Republic of Uganda and the Democratic Republic of
Congo, alpine wetlands, lakes and streams are supplied, in part, by snowfields
that occur primarily on three mountains, Mount Stanley, Mount Speke and Mount
Baker (Fig. 1). Glacial meltwater contribute to alpine riverflow upon which
downstream BaKonzo and BaAmba communities rely for year-round water
supplies and hydro-electric power generation. These aquatic environments,
headwaters of the Nile, are home to a diverse range of flora and fauna, many of
which are endemic and a few of which are listed in the IUCN Red List of
Threatened Animals (Busulwa, 1998; Schmitt, 1998). Spectacular vascular plants
include Giant Heather (Erica spp.), Giant Lobelia (Lobelia wollastonii) and Tree
Senecio (Dendrosenecio adnivalis). Indigenous fauna include small mammals
(e.g., Rwenzori Otter-shrew), fish (e.g., Varicurhinus rwenzori) and large
mammals such as the Rwenzori Leopard.
By 1990, glaciers on the Rwenzori Mountains had receded to about 40% of their
extent recorded in 1955 (Fig. 1) and less than one quarter of that measured by
the Duke of Abruzzi in 1906 (Kaser and Osmaston, 2002). Apart from an
expedition in 1993 (Talks, 1993), more recent assessments of glacial extent have
been hindered by civil unrest in Uganda and conflict in the Democratic Republic
of Congo (1998-1999). During the 20th century, glacial recession on the
Rwenzori Mountains mirrored recession on Kilimanjaro in Tanzania (Hastenrath
and Greischar, 1997) where Thompson et al. (2002) predict glaciers will
disappear entirely by 2020. With continuing political stability in Uganda, a
research project in the Rwenzori Mountains was proposed with the overall aim:
• to assess the impact of recent climate change on alpine aquatic
ecosystems of the Rwenzori Mountains.
- 8
Three scientific objectives were identified in order to meet the project's overall
aim:
1. to assess the magnitude of current glacial recession;
2. to assess the impact of glacial recession on alpine riverflow; and
3. to assess recent environmental change from observational datasets
and available, environmental archives stored in lake sediment and
glacial ice.
Figure 1. The Central Rwenzori Massif showing alpine lakes, wetlands, streams and the extent of
glacial snow cover in 1955 and 1990 (redrawn and adapted from Osmaston and Kaser, 2001).
- 9
1.2 Rwenzori Mountains National Park
The Rwenzori Mountains lie within the western arm of the East African Rift
System and comprise an uplifted block (horst) of Precambrian crystalline rocks
including gneisses, amphibolites, migmatites, and granites. The Rwenzori horst
was tilted up in a ESE to WNW direction 4 km above the surrounding peneplain
known as the ‘African Surface’ in the Late Pliocene (Figs. 2) (Osmaston, 1989;
Taylor and Howard, 1998). Occupying an area of 3000 km2, the horst is bounded
to the north and south by grabens that feature Lakes Albert and Edward
respectively (Figs. 1 and 3). A maximum elevation of 5109 metres above mean
sea level (mamsl) is currently reached by Margherita Peak on Mount Stanley.
The name, “Rwenzori” is a colonial-era corruption of ”Rwenzururu” meaning
place of snow (“rwe nzururu”). Although the mountain range is only 40km from
the equator, glaciers currently occur on three mountains: Stanley, Speke and
Baker, due to a combination of cold temperatures and abundant precipitation
ranging from 2000 to 2700 mm⋅a-1 (Kaser and Osmaston, 2002). Glaciers also lie
at the centre of the traditional belief system of the BaKonzo who have long lived
in the foothills of the Rwenzori Mountains (Alnaes, 1998). Snow/ice, "Nzururu", is
the 'father' of the BaKonzo deities, "Kitasamba" and “Nyabibuya” who are
responsible for human life, its continuity and its welfare. The area now occupied
by the Rwenzori Mountains above an approximate elevation of 1700 mamsl in
Uganda (996 km2) was gazetted as a national park in 1991. Rwenzori Mountains
National Park was subsequently made a World Heritage Site in 1994.
Figure 2. A conceptual representation of the uplift from the Late Miocene to present of the
Rwenzori horst tilting upwards from the ESE to WNW (adapted from Osmaston (1989) with
approximate ages drawn from Taylor and Howard (1998)). Vertical exaggeration is x 4.
- 10
Figure 3. LandSat7 (ETM+) optical satellite image of western Uganda on January 31, 2003
featuring the Rwenzori Mountains to the north and Lake Edward to the south. The latter is
connected to Lake George in the east by the Kazinga Channel. Areas in white represent cloud
cover. Image accessed from the United States Geological Survey (http://edcdaac.usgs.gov and
http://glovis.usgs.gov).
1.3 Project planning
A joint expedition between University College London, UK (UCL) and Makerere
University, Uganda (MUK) was first proposed during research visits by Richard
Taylor (UCL) to Uganda in 2001 (December) and 2002 (July-August). The broad
aim and scientific objectives of the project were established during the latter visit
in 2002 and through correspondence with members of the expedition team
including primarily Andrew Muwanga, Bob Nakileza and Immaculate
Ssemmanda (MUK), Callist Tindimugaya (Water Resources Management
- 11
Department), Neil Rose and Anson Mackay (UCL), and Andrea Fischer
(University of Innsbruck). Extensive consultation via electronic mail was also
made with members of the expedition advisory panel including most notably
Henry Osmaston, Daniel Livingstone, Georg Kaser and Deo Lubega. The
programme of research in the Rwenzori Mountains National Park was
subsequently approved by the Uganda National Council for Science and
Technology (File No. EC 583) and Uganda Wildlife Authority.
Expedition planning began in February 2003 following confirmation that the
proposed research had received a grant from the Royal Geographical Society
(co-winners of the Society's 2003 Ralph Brown Award). In April (2003),
researchers from Ohio State University including Professor Lonnie Thompson
were forced to pull out of the expedition as the American State Department
advised American citizens against travel to East Africa due to the threat of
terrorist activities associated with the Anglo-American invasion of Iraq. Collection
of an environmental archive in glacial ice was consequently abandoned. In this
same month, logistical details including a project timetable, equipment needs and
division of responsibilities were resolved during a visit by Richard Taylor to
Uganda. Food and necessary equipment were subsequently purchased in
Uganda and United Kingdom and greatly facilitated by the work of Nelson Kisaka
(Makerere University).
1.4 Project team
1.4.1 June 2003 expedition
The project features close collaboration between Ugandan scientists and
students from Makerere University and Water Resources Management
Department, and their counterparts from University College London (UK) and
Institut für Geographie (Austria). To develop new scientists, six students from
Makerere University and University College London participated in this
expedition. Overall, the team possesses expertise in key fields of hydrology,
alpine glaciology, palaeolimnology, palynology, ecology, hydrogeochemistry and
geomorphology. Critical logistical support was provided by the Rwenzori
Mountaineering Service (guides, cooks, porters) and Uganda Wildlife Authority
(project liaison, rangers). Members of the expedition team are listed below.
Department of Geography, University College London (UCL)
Dr. Richard Taylor (co-expedition leader, hydrologist, geochemist)
Dr. Anson Mackay (palaeolimnologist, diatomist)
Dr. Neil Rose (palaeolimnologist, environmental geochemist)
Lucinda Mileham (student)
Virginia Panizzo (student)
Adinah Shackleton (student)
- 12
Departments of Geography and Geology, Makerere University
Dr. Andrew Muwanga (co-expedition leader, hydrologist, geochemist)
Dr. Immaculate Ssemmanda (palynologist)
Dr. Bob Nakileza (alpine geomorphologist)
Nelson Kisaka (student)
Alex Mbonimba (student)
Allen Ndyanabo (student)
Water Resources Management Development (WRMD)
Callist Tindimugaya (isotope geochemist)
Institut für Geographie, University of Innsbruck
Dr. Andrea Fischer (glaciologist)
Uganda Wildlife Authority (UWA)
Aggrey Rwetsiba (project liaison, Kampala)
Baluku Salevano (project liaison, Nyakalengija)
Sinairi Koffi (ranger)
Michael Mugabe (ranger)
Rwenzori Mountaineering Service (RMS)
Joel Nzenge (guide), Patrick Bwabu (guide), Azalia Mawano (guide), Nason
Rwaburara (cook), Erifaza Salongo (cook), porters: Xavier Bonabama, George
Bakluku, Lazarus Limbali, Batrumao Muthende, Monday Nzwenge, Kabugho
Yodesi, Alfred Bwambale Waka-waka, Sipiriano Bwambale, Lazaro Bwambale,
Zephania Kibwana, Erifaza Wakibanahi, Richard Kithamuliko, Mitusera
Bakamwegha, Isaac Sibyaleghana, John Mukirania, Henry Bwambale, Baluku F.
Nyoro, Masereka Muthabuli, Luka Katalikawi, Clovice Masereka, Thembo
Isebbani, Josephine Muthabali, Peter Kijuma, Sunday Munyanyika, Zalimon Kule
Bagenda, Amon Marayi, Asasio Katalyaburo, Azalia Syabugha, Alice Masika,
Dominico Marayi, Koroneri Limbali, Geoffrey K. Bagheni, Tobius Bakamwegha,
Andrew Masumbuko, Moni Marahi, Wilson Muwunza, Zaverio Mbakania, Ezera
Kabukobi, Erisa Muhongya, Bruno Limbali, Naton Mbaju, Sudrak Muthabali,
Joshia Matheka, Paul Bukokoli, Erinerico Bakamwegha, Stephen Kindenga,
James Marayi, Jowasi Kule, Koroneri Byalemene, Sarah Basolene, Simon
Muhindo, James Kathikayima, Ernest Bwambale, Paul Baluku, Musumba
Walholire, Mirikiodi Kibaya, Magadalena Biira, Sereka Kathikayima, Tito Sunday,
Robert Syamulsaryira, Joseph Thembo, Isaac Kule, Charles Baluku, Girison
Solongo, Julius Mbusa, Robert Muhindo, Richard Mattre, Bruno Bwambale
1.4.2 January-February 2005 expedition
Department of Geography, University College London (UCL)
Dr. Richard Taylor (expedition leader, hydrologist, geochemist)
Lucinda Mileham (student)
- 13
Uganda Wildlife Authority (UWA)
Aggrey Rwetsiba (project liaison, Kampala)
Guma Nelson (project liaison, Nyakalengija)
Rwenzori Mountaineering Service (RMS)
Bwambale Jales (guide), Baluku Josephat (guide), Nason Rwaburara (cook),
porters: Cypriano Bwambale, Iseban Gideon, Baluku Simon, Kule Kikumbwa,
Zablan Wilson, Nyamambisi Misaki, Bwambale Zikalia, Baluku Dimiano, Monday
spay, Tsomwa Erineo, Thahimba Erinerico,
1.4.3 Project advisory panel
Henry Osmaston (Cumbria, UK)
Georg Kaser (University of Innsbruck, Austria)
Daniel Livingstone (Duke University, USA)
Lonnie Thompson (Ohio State University, USA)
Deo Lubega (Kampala, Uganda)
Joel Okonga (Water Resources Management Department, Uganda)
1.5 Report layout
The report is grouped into six chapters. Chapters 2 and 3 detail surveys of glacial
extent and changes in climate deduced from meteorological observations over
the last century. Substantial portions of both of these chapters have been
published in the journal, Geophysical Research Letters (Taylor et al., 2006a;
2006b). Chapter 4 discusses the impact of glacial recession on alpine riverflow
and investigates trends in hydrological observations coincident with the recent
period of deglaciation. Chapter 5 assesses environmental changes in the
Afroalpine region of the Rwenzori Mountains deduced from environmental
proxies, primarily diatom and pollen, in sediment-core archives from Lake Bujuku
(Fig. 1). A slightly revised version of chapter 5 is expected to appear in a
forthcoming volume of the Journal of Paleolimnology (Panizzo et al., in review).
Chapter 6 reviews evidence of pollution from atmospherically deposited trace
metals and fly ash recorded in the sediment of alpine lakes ranging in elevation
from 3000 to 4000 mamsl. This chapter is a revised version of the M.Sc.
dissertation of Jenny Hau (2005) entitled, Sediment record of atmospherically
deposited pollutants in Rwenzori Mountain lakes, Uganda.
- 14
2 Recent glacial recession in the Rwenzori Mountains
and rising air temperatures
2.1 Introduction
Tropical alpine glaciers serve as highly sensitive indicators of tropical climate
(Wagnon et al., 1999; Francou et al., 2003) that are particularly valuable in areas
where meteorological records are scarce. In the East African Highlands, glaciers
have been shrinking over much of the 20th century (Hastenrath and Kruss, 1992;
Kaser and Noggler, 1996; Hastenrath and Greischar, 1997; Kaser and
Osmaston, 2002; Thompson et al., 2002). Mapping of glacial extent in the
Rwenzori Mountains (Fig. 4a), was, however, last conducted more than a decade
ago (Kaser and Noggler, 1991; Talks, 1993). There is also uncertainty regarding
the nature of climate change in these highlands (Hay et al., 2002; Patz et al.,
2002; Kaser et al., 2004).
The first survey of glaciers in the Rwenzori Mountains was conducted in 1906 by
the Duke of the Abruzzi (1907) when the glacial cover over the entire range was
estimated to be 7.5 km² (Kaser and Noggler, 1996) and the lowest altitude of
glaciation is thought to have reached 4400 metres above mean sea level
(mamsl) (Osmaston, 1989). Scientific surveys carried out in the 1950s (Menzies,
1951; Bergstrøm, 1955; Whittow et al., 1963) and early 1990s (Kaser and
Noggler, 1991; 1996; Talks, 1993) indicate a continuing trend of glacial recession
though a brief episode of terminal advance was observed in the early 1960s
(Whittow et al., 1963; Temple, 1967). This coincides with a period of anomalously
high precipitation when the levels of Lake Victoria rose by 2.5 m (Kite, 1981).
2.2 Field and satellite mapping
Recent glacial recession was assessed by field mapping of the terminal positions
of previously monitored 'indicator' glaciers, Elena and Speke (Fig. 4b), and
quantitative interpretations of snow and ice cover using optical spaceborne
imagery (LandSat5, LandSat7). Field surveys of glacial termini on Mounts Speke
and Stanley using a handheld global positioning system (GPS) were conducted
in June 2003 and January 2005 (Fig. 5). These included an assessment of the
distance between observed termini and the positions of terminus markers set in
1958 (Whittow et al., 1963) and 1993 (Talks, 1993).
- 15
Figure 4. (a) Map of Uganda showing the location of the Rwenzori Mountains and meteorological
stations. (b) The Central Rwenzori Massif showing indicator glaciers (Elena and Speke) and the
extent of glacial cover in 1955 and 1990 (Kaser and Osmaston, 2002). Topographic contours
(interval: 305 m) are in metres above mean sea level (originally in feet). (c) Changes in the
terminus and areal extent of the Elena Glacier from 1906 to 2005. (d) Subsample of the NDSI-
classified LandSat7 ETM+ satellite image on January 31, 2003 for Mount Speke superimposed
on mapped glacier extents in 1906, 1955 and 1990 (Kaser and Osmaston, 2002).
Changes in the areal extent of glaciers were assessed using two geometrically
corrected LandSat5 (TM) and one systematically corrected LandSat7 (ETM+)
optical satellite image accessed from the United States Geological Survey
(http://edcdaac.usgs.gov and http://glovis.usgs.gov). Three LandSat images with
views of the still glacierised summits unobstructed by clouds were identified in
1987 (7 August), 1995 (17 January) and 2003 (31 January). In the inner tropics
where diurnal variations in mean air temperature (~8ºC) significantly exceed
seasonal variations (~2ºC) (Paffen, 1967 cited in Kaser and Osmaston, 2002),
ablation on glacial tongues occurs throughout the year. Snow falling below glacial
termini in the Rwenzori Mountains is subject to rapid melting through daily
ablation during daylight hours and accumulation is confined to high glacial areas
during periods of heavy precipitation (rainy seasons). The areal extent of snow
and ice inferred from satellite imagery was, therefore, considered to represent
glacial cover.
- 16
Figure 5. Photograph of Rwenzori Mountaineering Service guide, Baluku Josephat, leading the
research team on a trek across the Stanley Plateau (30 January 2005). Alexandra (left) and
Margherita (right) peaks on Mount Stanley occupy the background.
LandSat images were subsampled to create specific images of each mountain in
the Central Rwenzori Massif (Fig. 4b). Areal extent of snow and ice was
determined using supervised classification (SC) on a false-color composite of
bands 2 (visible (green), 0.52-0.60 µm), 4 (near infrared, 0.75-0.90 µm) and 5
(mid-infrared, 1.55-1.75 µm) as this best represents snow (Vogel, 2002). To test
the accuracy of estimates derived from SC, the deduced areal extent of glaciers
was also calculated using the Normalised Difference Snow Index (NDSI) (eq. 1).
The NDSI contrasts the brightness of snow in band 2 with its low reflectivity in
band 5. For the most recent (2003) image, field surveys assisted in supervision
of classification of glacial cover and confirming the applied NDSI threshold
distinguishing glacial cover from rock.
5
2
52
band
band
bandband
NDSI +
−
= (eq. 1)
- 17
2.3 Results
2.3.1 Field surveys of glacial extent
Field surveys of the terminal positions of the Speke and Elena Glaciers
demonstrate a continuation of the overall trend of recession observed between
1906 and 1990. The terminus of the Elena glacier has retreated by ~400 m since
1906 and 140 m ± 17 m since 1990 (Fig. 4c). Terminal retreat on the Speke
glacier is more rapid, ~600 m since 1906 and 311 m since 1993 (Fig. 4d). The
contrasting rates are considered to result primarily from differences in the supply
of ice to these valley glaciers as a result of their elevation and bed morphology
(Fig. 4b). Hypsographic curves prepared by Kaser and Osmaston (2002) show
that between 15 and 20% of the surface area of Mount Stanley (i.e. Stanley
Plateau) resides above the highest peak on Mount Speke (Vittorio Emanuele), an
elevation of approximately 4900 mamsl. As the equilibrium line altitude (ELA) for
each mountain has risen since the last glacial maximum (Mahoma Stage) and
particularly over the last century (Osmaston and Kaser, 2001), the supply of ice
to Elena Glacier from the Stanley plateau would have continued for longer than
that to the Speke Glacier. Mölg et al. (2003) contend that the ELA for Mount
Stanley was 4900mamsl in 1955 and conclude that no effective accumulation
has occurred on Mount Speke and neighbouring Mount Baker (Fig. 4b) during
the 20th century. The impact of ice supply on rates of terminal retreat (i.e., glacial
recession) is considered in detail below.
The Speke Glacier is the largest single valley glacier in the Rwenzori snowfield
(Fig. 6) and its terminal position has, in a relative sense, been monitored quite
frequently. Repeated measurements of the terminal position of the Speke Glacier
in relation to a marker established in 1958 (Fig. 6c) show that rates of terminal
retreat have increased from 3 m per annum (1962-1977) to 31 m per annum
(1993-2003). Similar exponential retreat of glacial termini has been observed for
the Qori Kalis Glacier (Thompson, 2000), the largest outlet of the Quelccaya ice
cap in Peru (13°56'S, 70°50'W). As linear rates of recession have been used to
predict the lifetime of tropical glaciers (Thompson et al., 2002), these
observations appear to suggest that the loss of tropical glaciers may be even
more rapid. These trends can, however, be misleading. 'Glacier retreat', the term
used to describe the current behaviour of the termini of many glaciers worldwide,
reflects a rise in the equilibrium-line-altitude (ELA)1. The ELA varies in response
to climatic changes, primarily temperature and precipitation. A rise in ELA results
in a reduction of total net mass balance and, thus, a progressive reduction in the
volume of a glacier. It is this last parameter (volume) that is a true measure of the
condition of a glacier and the way in which it is being affected by climate change.
1 The equilibrium-line-altitude represents the elevation at which glacial accumulation is balanced
by ablation. In the wet inner tropics that include the Rwenzori Mountains, high accumulation (as a
result of a high precipitation) tends to depress the ELA just below the 0ºC elevation (Kaser and
Osmaston, 2002).
- 18
Depending on the morphology of the glacier and its bed, the rate of retreat of the
terminus may directly reflect the rate of volume loss. This is unlikely, except in
uniformly sloping valleys, and the Speke Glacier shows this discrepancy very
clearly.
Prior to the 1950s, the terminus of the Speke Glacier lay adjacent to the edge of
a steep cliff (Figs. 7 and 8) from which avalanches of snow shed down a steep
rock face. Substantial volume loss subsequently occurred by thinning since,
before any significant retreat of the terminus could take place, it had to thin down
to the level of the underlying rock. Retreat was consequently negligible. Retreat
then took place horizontally across the basin, now occupied by a pool (Lake
Ruhandika). Although the glacier still sloped down steeply, it was still thick so
that retreat occurred at a modest rate. To the northeast of Lake Ruhandika, the
bed slopes steeply at approximately the same gradient as did the glacier surface.
There is abundant photographic evidence of thinning of the whole Speke Glacier
(e.g., Kaser and Noggler, 1996), right up to its highest limits. The ice lying on this
slope thinned fairly uniformly until at a critical thickness it became effectively
stagnant, cut off from supply due to the convex-concave slope profile and subject
to rapid melting. This would result in a rapid retreat of the terminus to the change
in gradient at the top of the slope, where the present thin terminus can be seen.
Thus, the bed and glacier morphologies are able to explain, in part, the
exponentially accelerating rate of terminus retreat that has been observed and
the fact that retreat has occurred more rapidly than Elena Glacier on Mount
Stanley.
2.3.2 Mapping glacial cover by remote sensing
Analyses of LandSat imagery using supervised classification (SC) and NDSI
identify a ~50% decrease in the total area of glaciers from 1987 (2.01±0.11 km²)
to 2003 (0.96±0.34 km²). Broad agreement exists between estimates of glacial
cover (<12% difference) derived from each method (Table 1). The results of the
NDSI-classified, LandSat image from 1987 are consistent with historical data
derived from aerial and terrestrial photography (Kaser and Noggler, 1996; Kaser
and Osmaston, 2002). On all three of the still glacierised mountains in the
Central Rwenzori Massif, the estimated areal extent of glaciers in 1987 is less
than the area in 1955 but greater than that assessed for 1990 (Table 1). The
estimated areal extent of glaciers on Mount Baker derived from LandSat data in
1995 is slightly larger (0.08 km²) but within analytical error of the area estimated
from terrestrial photographs in 1990. Outlines of glacial extent on Mount Speke
mapped by Kaser and Osmaston (2002) for 1906, 1955 and 1990 are overlain on
a subsample of the NDSI-classified LandSat7 image from 2003 in Fig. 4d. High
reflectance areas (pixels), classified as glacial cover, clearly demonstrate the
loss of glacial cover at lower elevations since 1990.
- 19
Table 1. Areal extent of glacial cover (km2) on the Central Rwenzori Massif (Fig. 4b). Estimated
errors derive from the classification and geometry of pixels.
Year Method
Baker
(km2) Speke
(km2) Stanley
(km2) Total
(km2)
1906a
1.47 2.18 2.85 6.50
1955a
0.62 1.31 1.88 3.81
1987b
SC
NDSI 0.43±0.13
0.38±0.04 0.65±0.18
0.63±0.02 1.03±0.25
1.00±0.05 2.11±0.56
2.01±0.11
1990a
0.12±0.01 0.56±0.06 1.00±0.10 1.68±0.17
1995b
SC
NDSI 0.21±0.06
0.20±0.09 0.45±0.11
0.44±0.17 0.69±0.15
0.86±0.10 1.35±0.32
1.50±0.36
2003b
SC
NDSI 0.16±0.05
0.11±0.03 0.40±0.08
0.35±0.11 0.53±0.09
0.50±0.20 1.09±0.22
0.96±0.34
a: (Kaser and Osmaston, 2002); b: this work
Figure 6. The changing rate of terminus retreat of the Speke Glacier from 1958 to 2003 is shown
in the inset of photograph (a). Field survey of terminus retreat in 2003, relative to the 1993 marker
(Talks, 1993), is also shown in photograph (a). Photograph (a) is panoramic image taken by
Andrea Fischer on 22 June 2003 whereas the panoramic photograph on 29 January 1990 (b) is
from Kaser and Noggler (1991).
- 20
Figure 7. (a) Terrestrial photograph of the Speke Glacier from Mount Stanley in 2005 showing
the steep slopes below the glacier's current and former terminus (photo: Alan Kilian). (b) NDSI-
classified Landsat7 image from 31 January 2003.
Figure 8. A conceptual, cross-sectional representation of changes in the profiles of ‘indicator’
valley glaciers (shaded): (a) Elena and (b) Speke in the Rwenzori Mountains from 1991 to 2005
(adapted from Kaser and Osmaston, 2002). Surface profiles are drawn relative to an air-
photograph survey of glacier cover in 1955. Glacial thickness and bed topography are not known
precisely. Vertical exaggeration is x 2.
- 21
The recent field surveys and satellite mapping of glacial cover on the Central
Rwenzori Massif show continued rapid reduction in glacial extent (Fig. 9) since
the last field measurements were taken in the early 1990s (Kaser and Noggler,
1991; Talks, 1993; Kaser and Noggler, 1996; Kaser and Osmaston, 2002).
Measurements over the last century indicate a relatively steady rate of decline in
glacial extent though linearity (r2 = 0.999) in the rate of recession may derive, in
part, from the limited number and uneven distribution of observations. Advance
of the Speke Glacier was, for instance, detected during a period (1961 to 1962)
of frequent monitoring and anomalously high precipitation (Temple, 1967).
Glaciers in the Rwenzori Mountains are, nevertheless, expected to disappear
within the next two decades if deglaciation continues to follow the observed linear
trend (Fig. 9).
Figure 9. Plot of changes in glacial areal extent on the Central Rwenzori Massif since 1906. Data
from 1906, 1955 and 1990 derive from Kaser and Osmaston (2002). Remaining observations
derive from NDSI classification of Landsat 5 (TM) and Landsat 7 (ETM+) scenes (Table 1).
2.4 Meteorological trends
The absence of continuous and proximate meteorological observations in the
Rwenzori Mountains prevents direct analysis of the climatic factors driving
observed glacial recession. Previous studies of glacial dynamics in the East
African Highlands (e.g. Kruss and Hastenrath, 1987; Kaser and Noggler, 1991;
Mölg et al., 2003; Kaser et al., 2004) contend that recession over the 20th
century arises principally from an abrupt decrease in humidity at the end of the
19th century (ca. 1880). Decreased humidity increases the exposure of glaciers
to solar radiation through reduced cloud cover. An associated decline in
precipitation lowers accumulation and increases absorption of radiation due to
the lower albedo of ice, relative to snow (Mölg et al., 2003; Kaser et al., 2004).
As a result, the rate of glacier net mass loss consequently rises.
- 22
Hastenrath (2001) citing studies on Mount Kenya (Kruss, 1983; Hastenrath and
Kruss, 1992), posits that glacial recession in the East African Highlands beyond
the early decades of the 20th century has been promoted by a warming trend
that has increased atmospheric humidity. This inhibits sublimation and permits
more of the sun’s energy to melt glacial ice due to a saving of latent heat. This
has been well demonstrated on the Zongo and Chaccaltaya glaciers in Bolivia
where higher melt rates in the wet season (i.e. period of increased humidity)
result from reduced sublimation (Wagnon et al., 1999; Francou et al., 2003).
Though differential recession of glaciers in response to variations in solar
incidence has been proposed for the Rwenzori Mountains (Mölg et al., 2003), the
spatially uniform loss of glacial cover in the Rwenzori Mountains at lower
elevations over the last decade strongly suggests increased air temperature is
the main driver.
Terrestrial observations of air temperature (Ta) are consistent with a warming
trend indicated by recent glacial recession. Daily records of maximum and
minimum air temperature at meteorological stations east of the Rwenzori range
between latitudes of 1º41'N and 1º15'S in Uganda (Fig. 4a) show significant (at
confidence intervals of 99% or greater) and consistent trends toward increased
air temperatures of ~0.5ºC per decade since the last period of glacial advance in
the early 1960s (Fig. 10). These data contain, however, significant gaps and are
limited in duration. Gridded CRU2 climate data (New et al., 2002) for the grid cell
closest to the Rwenzori Mountains (0º50'N, 29º30'E) also demonstrate a small
but significant rise in mean surface temperature of 0.15ºC per decade from 1960
to 1998 that is consistent with a regional warming trend of the same magnitude
determined by Patz et al. (2002). Because of the thermal homogeneity of the
troposphere in the inner tropics (Kaser and Osmaston, 2002; Oerlemans, 2005),
the recent (post-1960) rise in air temperature observed at stations between 960
and 1869 mamsl is also expected to occur in areas of glacial cover between
4800 and 5100 mamsl. This assumption is discussed in detail in section 3.2.
The possibility that recent glacial recession arises from a reduction in
precipitation is unsupported by station data in western Uganda over the 20th
century. Records of annual precipitation demonstrate considerable interannual
variability (Fig. 11) which is thought largely to be controlled synoptically by sea
surfaces temperatures of the Indian Ocean via the El Niño Southern Oscillation
(Ogallo, 1988) and Indian Ocean Dipole (Conway et al., 2007). Mölg et al. (2006)
cite historical evidence in support of a reduction in humidity in East Africa
beginning around 1880 (Nicholson and Yin, 2001). The evidence is anecdotal
and it remains unclear whether the posited change follows a sustained period of
greater humidity or a brief anomaly (see sections 3.3 and 4.6). Meteorological
records in Uganda begin at the end of the 19th century and are, thus, unable to
investigate whether a posited warming trend starting in the 19th century
(Oerlemans, 2005) also contributed to the onset of deglaciation in the East
African Highlands.
- 23
Figure 10. Standardised anomalies in annual mean maximum (solid circles) and mean minimum
(open circles) air temperature observed at meteorological stations in western Uganda and annual
mean temperature from gridded CRU2 climate data (New et al., 2002) over the 20th century.
- 24
Figure 11. Longitudinal trends in annual precipitation, plotted as deviations from the mean, for
stations at meteorological stations in western Uganda over the 20th century.
2.5 Conclusions
Recent field mapping and analysis of Landsat imagery confirm a rapid decline in
the areal extent of glaciers on the Central Rwenzori Massif that is consistent with
an overall recessionary trend over the 20th century. Glacial cover on the three
remaining glacierised summits (Mounts Stanley, Speke and Baker) has
decreased from 2.01±0.56 km2 in 1987 to 0.96±0.34 km² in 2003 and is expected
to disappear within the next two decades. Field surveys highlight the importance
of bed morphology in determining the rates of glacial termini retreat. Climate
inferences attributed solely to rates of termini advance or retreat must therefore
be regarded with caution. Increased air temperature suggested by the spatially
uniform nature of recent loss of glacial cover at lower elevations, is supported by
station data in western Uganda and gridded climate data sets. The observed rise
in air temperatures over the last four decades is also consistent with warming
trends predicted in the tropical troposphere from climate model simulations that
incorporate historical increases in greenhouse gases (Santer et al., 2005).
- 25
3 Climatological implications of glacial recession in the
Rwenzori Mountains
3.1 Introduction
Debate persists as to the extent to which recent glacial recession observed in
tropical highlands is driven primarily by changes in air temperature (e.g. Bradley
et al., 2006; Thompson et al., 2006) and atmospheric humidity (e.g. Kaser et al.,
2004; Mölg and Hardy, 2004). Uncertainty has also been expressed in the
relationship between temperature trends at the surface and higher elevations in
the tropical free troposphere (e.g. Christy et al., 2003; Christy and Norris, 2004;
Douglass et al., 2004; Fu et al., 2004; Tett and Thorne, 2004) where alpine
glaciers reside. Although the surface energy balance and mass balance are best
able to describe the relationship between climate parameters and glacier change
(e.g. Wagnon et al., 1999; Mölg and Hardy, 2004), measurements that would
form the basis of a glacier mass balance model of the Rwenzori Mountains do
not exist. Although a definitive, quantitative understanding of the climate
variables responsible for glacier mass losses in the Rwenzori Mountains remains
elusive, we show that trends of increasing air temperature are better supported
by currently available evidence than decreasing humidity.
3.2 Relationship between Ta at the surface and mid troposphere
The validity of the assumption that Ta trends observed in gridded CRU TS 2.0
climate data sets (New et al., 2002) and at meteorological stations between 960
and 1869 mamsl, reflect Ta trends in the middle troposphere where glaciers in
the Rwenzori Mountains occur, is open to question. Significant uncertainty
persists in temperature data for the tropical troposphere whether these derive
from satellite-borne Microwave Sounding Unit (MSU) observations or in situ
measurements using radiosondes, particularly in data-poor regions like East
Africa. Indeed, linear Ta trends in the tropical troposphere can vary significantly
based simply upon choice of start and end date as is the case in the paper by
Gaffen et al. (2000) using MSU data in which at 500hPa a cooling trend is
detected between 1979 and 1997 but an overall warming trend occurs between
1960 and 1997. Nevertheless, recent studies that employ diurnal corrections to
MSU observations between 1979 and 2003 (Mears and Wentz, 2005) and
homogenized radiosonde datasets (HadAT2) between 1958 and 2002 (Thorne et
al., 2005), show that the middle troposphere warmed at a similar or slightly
greater rate to the surface in the tropics (Fu and Johanson, 2005; Santer et al.,
2005), consistent with the sign and (within error) magnitude of Ta trends (+0.13ºC
- 26
per decade) at the surface from climate model (HadCRU2v) predictions (Jones
and Moberg, 2003).
Upper air temperature records from gridded HadAT2 radiosonde data (Thorne et
al., 2005) for the most proximate (and only) grid cell to the Rwenzori Mountains
show consistent warming trends in the lower and middle troposphere (700hPa,
500hPa) from 1958 to 2005 (Fig. 12). These warming trends coincide with
increased Ta trends at the surface over the second half of the 20th century that
have been detected in gridded (homogenized) CRU TS 2.0 datasets (New et al.,
2002) at four locations in the East African Highlands by Pascual et al. (2006) and
the Rwenzori Mountains (Fig. 10). The increased incidence of malaria observed
in the East African Highlands is considered, in part, to arise from rising air
temperatures (Pascual et al., 2006) as mosquitoes are able to colonise
environments at elevations that previously restricted mosquito populations and,
hence, malaria transmission by temperature (i.e., <15ºC). Local records for
Kasese District at the base of the Rwenzori Mountain show over a 6-year period
that malaria incidence have risen at a rate of over 24 000 cases per year from
172 992 cases initially recorded in 1998. A peak of 457 601 cases was then
recorded in 2004. Other factors, apart from a rise in air temperature, such as the
conversion of forest cover to cropland, increased migration, and improved
reporting procedures are also considered to have contributed significantly to the
observed rise in the incidence of malaria.
A comparison of temperature trends from surface observations at high elevations
and free troposphere (radiosonde measurements) indicates more rapid warming
of alpine surfaces than the free troposphere (Pepin and Seidel, 2005) though this
discrepancy is reduced for mountain peaks and may stem from a systematic
cooling bias arising from daytime heating of the radiosonde sensors (Sherwood
et al., 2005). Analyses of station data in the tropical Andes (Vuille and Bradley,
2000) and on the Tibetan Plateau (Liu and Chen, 2000) show that Ta trends
between 1000 and 5000 masl remain constant in sign (i.e. increasing Ta) but can
vary in magnitude (+0.1 to +0.3ºC per decade). It is worth noting that a step-wise
increase in Ta during the 1970s, noted globally at the surface (Jones and
Moberg, 2003) and in the troposphere (Thorne et al., 2005) as well as in the
tropical Andes (Vuille and Bradley, 2000), is also observed at the surface in CRU
TS 2.0 datasets in the East African Highlands (Fig. 1 in Pascual et al., 2006) and
station data in western Uganda (Fig. 10).
3.3 Trends in atmospheric humidity
Mölg et al. (2006) employ NCEP reanalysis data to indicate a trend of decreasing
specific humidity in the mid-troposphere (600hPa) from 1948 to 2005. Quite apart
from the time-dependent biases in all NCEP data, the reliability of the specific
humidity data is particularly questionable as NCEP humidity is a statistically
derived parameter. The ability of NCEP humidity data to represent interannual
- 27
Figure 12. Time series of monthly air temperature anomalies in the lower and middle troposphere
from homogenised radiosonde datasets (HadAT2) (Thorne et al., 2005) at (a) 700hPa and (b)
500hPa for the most proximate grid cell (35ºE, -2.5ºS) to the Rwenzori Mountains. Bold lines in
(a) and (b) represent the 12-month running mean.
Figure 13. Time series of mean annual anomalies in (a) vapour pressure (1901 to 1995) and (b)
precipitation (1901 to 1998) from gridded CRU TS 2.0 climate data (New et al., 2002) for the most
proximate grid cell (29.5ºE, +0.5ºN) to the Rwenzori Mountains.
- 28
precipitation anomalies associated with the dominant modes of climate variability
in equatorial east Africa does not bear on the reliability of these datasets for trend
analyses. Radiosonde-derived humidity from 1965 to 1984 (Hense et al., 1988)
cited in support of NCEP specific humidity trends from 1948 to 2005, are in fact
uncorrected; systemic dry biases have been carefully removed from more recent
corrected datasets (Guichard et al., 2000). A decline in humidity over the 20th
century is, furthermore, unsupported by surface CRU TS 2.0 precipitation and
vapour pressure datasets (Fig. 13).
Mölg et al. additionally argue that observed glacial recession in the East African
Highlands over the last century originates from a drastic reduction in moisture in
the late 19th century. This drop in moisture, based on historical evidence of the
levels of Lake Victoria and other East African lakes (Nicholson and Yin, 2001), is
actually the descending limb of a temporary (less than two decades) high lake
stand (Fig. 14). Lake levels, a remote and indirect proxy of regional humidity, are
variable during the 19th century prior to their peak in 1880 but comparable to
lake levels throughout the 20th century. A modern comparison to the 19th
century event is the 2.3 m rise in the level of Lake Victoria between October 1961
and May 1964 (Fig. 14). The implied increase in humidity associated with this
lake-level rise coincided with a very brief (1 year) and very marginal advance (3
to 5 m) in the terminal positions of valley glaciers in the Rwenzori Mountains
(Temple, 1967). The humidity hypothesis proposed by Mölg et al. (2006)
contends that (1) termination of a brief period of accumulation due to enhanced
precipitation around 1880 led to continued glacial retreat into the latter half of the
20th century and (2) a trend of decreasing humidity, supported only by NCEP
reanalysis data for which trend analysis is inappropriate, has driven glacial
recession since 1970. Even ignoring concerns regarding this evidence, the
argument that these climate events are responsible for the expected demise of
small, fast-responding glaciers that have persisted for at least 5000 years
(Thompson et al., 2006) is improbable.
Figure 14. Level of Lake Victoria at Jinja (33.2ºE, 0.2ºN) from 1800 to 2005 based on historical
evidence from 1800 to 1896 (Nicholson and Yin, 2001) and monthly observations from 1896 to
2005.
- 29
3.4 Conclusions
Both increasing air temperature and reduced air humidity remain plausible and
likely related hypotheses to explain recent glacial recession in the Rwenzori
Mountains of East Africa. There are insufficient data to represent the complex
interactions of radiant energy and heat at the glacier’s surface and thus quantify
the link between changes in climate variables and glacial mass in the Rwenzori
Mountains. There is, however, currently greater evidence of trends of increasing
air temperature than decreasing humidity to explain deglaciation in the Rwenzori
Mountains. This conclusion does not preclude, however, the likelihood that
changes in humidity and radiative fluxes associated with rising air temperatures,
have also contributed to observed glacial recession.
- 30
4 Hydrological implications of glacial recession
4.1 Introduction
The hydrological implications of recent deglaciation in the Rwenzori Mountains
outlined in Chapter 2 are unclear. Shrinking icefields not only affect glacial
meltwater discharges but also signal a net deficit between accumulation and
ablation. Gasse (2002) and Thompson et al. (2002) have expressed concern
over the direct impact of glacial recession on alpine riverflow and, thus,
freshwater resources in the East African Highlands particularly during dry (low-
flow) periods. In the Rwenzori Mountains, communities rely upon alpine riverflow
for year-round water supplies and generation of hydro-electric power. Little is
known, however, of the characteristics of alpine riverflow in the highland areas
due to a near absence of glacial meltwater and stream discharge measurements.
Current monitoring of riverflow at the base of the Rwenzori Mountains (Fig. 15) is
restricted to the River Nyamagasani though historical records exist for Rivers
Mubuku, Semliki, Rwimi, and Rukoki that drain the alpine areas of the Rwenzori
Mountains.
Tropical alpine glaciers are highly sensitive indicators of climate change [Kaser,
1999) but, as discussed in Chapter 3, the precise climate signals indicated by
shrinking icefields in the East African Highlands that include Mount Kenya,
Kilimanjaro and the Rwenzori Mountains, are the subject of much debate [e.g.
Thompson et al., 2002; Kaser et al., 2004; Chapter 2). Hastenrath [2001)
contends that glacial recession on Mount Kenya was initiated by a sharp
reduction in humidity in the late 19th century but sustained beyond the early
decades of the 20th century by a warming trend that has increased humidity. As
observed in the tropical Andes [Wagnon et al., 1999), a rise in humidity inhibits
sublimation and permits more of the sun’s energy to melt glacial ice due to a
saving of latent heat. For the Rwenzori Mountains, glacial recession over the
latter half of the 20th century coincides with rising air temperatures observed
regionally both at the base of the mountain range from station records and in the
mid-troposphere from radiosonde measurements (Chapters 2 and 3). In contrast,
Mölg et al. (2003) argue that recent glacial recession in the Rwenzori Mountains
has resulted from a reduction in atmospheric humidity and associated cloud
cover that has led to more rapid recession of glaciers on east-facing slopes. This
hypothesis draws upon observations on Mount Kenya by Baker (1967) who
noted that glaciers on northwest facing slopes are protected from morning sun by
the peaks and from afternoon sun by cloud.
In this chapter, we assess both the impact of glacial recession on alpine riverflow
in the Rwenzori Mountains and hydrometeorological trends in western Uganda,
indicated by station data and historical evidence, over the period of observed
deglaciation.
- 31
Figure 15. Topographic map of the Rwenzori Mountains showing the main river channels
draining alpine areas and the location of river-gauging and local meteorological stations.
4.2 Hydrology of the Rwenzori Mountains
4.2.1 Quaternary glaciations
Glaciers have played a major role in shaping the upper slopes of river basins that
drain the Rwenzori Mountains (Fig. 16). Alternating cycles of glacial and fluvial
erosion, extensively reviewed by Osmaston (1989), have produced glaciated
surfaces and deeply incised valleys. Moraine evidence reveals three major
Pleistocene glaciations: the Katabarua stage, ca. 300 000 years before present
(BP) (glacial cover ~500 km²); Rwimi Stage, ca. 100 000 years BP (glacial cover
~300 km²); and Mahoma stage, ca. 15 000 to 20 000 years BP (glacial cover
~260 km²) (Kaser and Osmaston, 2002). The Katabarua stage occurred on
previously undissected land allowing a large plateau ice cap to develop, creating
the most extensive glaciers in East Africa (Osmaston and Kaser, 2001). The
upward tilt of the Rwenzori horst in a WNW direction (Fig. 2) caused glaciers to
develop more extensively on ESE-facing slopes and occupy a large proportion
- 32
(~430 km2) most of what is now demarcated as the Rwenzori Mountains National
Park in Uganda.
The incision of drainage during the long cycle of erosion that followed the
Katabarua stage meant that subsequent glaciations were significantly smaller
and occurred as valley glaciers (Osmaston, 1989). Evidence of the Mahoma
Stage is most abundant on the mountain with large moraines in all of the main
valleys. Ice tongues at this time extended to 2000 mamsl and pollen evidence
suggests that temperatures were 4 to 6 ºC colder than present (Kaser and
Osmaston, 2002). Caution is, however, required in drawing simple associations
from the past between the elevation of glacial cover and temperature. As
highlighted by Osmaston and Kaser (2001), ELAs of former glaciers on the
Rwenzori Mountains, estimated using the area-height-accumulation method, are
consistently asymmetrical from east to west (Fig. 17). This asymmetry indicates
that local climatic conditions are favourable for the formation of glaciers at lower
elevations in the east than the west. Similarly, the lower limit of moraines on
east-facing slopes is lower than west-facing slopes due, in part, to larger
catchment areas in the east. According to Osmaston (1989), this combination of
larger catchments and more favourable climatic conditions on the east-facing
slopes of the Rwenzori Mountains enabled glaciers to reach elevations far lower
than anywhere else in the East African Highlands. Following the Mahoma stage,
alpine glaciers in the Rwenzori Mountains have, however, been in retreat apart
from two short intervals of advance ca. 10,000 years BP and 100 to 700 years
BP, known respectively as the Omurubaho (70 km²) and Lac Gris (10 km²)
stages (Fig. 15).
4.2.2 Modern glacial recession
Modern glacial recession in the Rwenzori Mountains is considered to have
started during the late 19th century (ca. 1880) following an abrupt reduction in
precipitation at the end of the Little Ice Age (LIA) (Kruss, 1983; Hastenrath, 2001;
Nicholson and Yin, 2001; Mölg et al., 2003; Kaser et al., 2004). The first survey
of glacial extent in the Rwenzori Mountains was conducted in 1906 by the Duke
of the Abruzzi (1907). At this time, Kaser and Noggler (1996) estimate that
glaciers covered a total area of approximately 7.5 km2 using maps prepared by
De Filippi (1909) and photographs taken by Vittorio Sella from the pioneering
Duke of the Abruzzi expedition. Glaciers were primarily restricted to Mounts
Stanley, Speke and Baker of the Central Rwenzori Massif (Fig. 3b) but also
included small glaciers with a total area of 1.0 km2 on Mounts Gessi, Emin and
Luigi de Savoia (Fig. 16). Glaciers largely existed above 4400 mamsl though
some valley glaciers (e.g. Elena, Speke) extended a few hundred metres lower in
elevation (Osmaston, 1989). As reviewed in Chapter 2, field research carried out
in the 1950s (Menzies, 1951; Bergstrøm, 1955; Whittow et al., 1963), early 1990s
(Kaser and Noggler, 1991; 1996; Talks, 1993) and most recently (Chapter 2)
confirm a pattern of continued recession over the last century (Fig. 9). The
terminal positions of some valley glaciers advanced, however, during the early
- 33
1960s. This brief period coincides with both frequent measurements of glacial
termini positions (Temple, 1967) and anomalously high precipitation across East
Africa when the level of Lake Victoria and the River Nile in Uganda rose by 2.5 m
(Lamb, 1966; Kite, 1981).
Figure 16. Map of the drainage network for the Rwenzori Mountains. The international border
divides the Republic of Uganda to the east and the Democratic Republic of Congo to the west.
Redrawn from the Fort Portal (1:250 000) sheet, NA-36-I3 (Lands and Surveys Department
Uganda, 1961).
- 34
Figure 17. Profile of equilibrium-line-altitudes (ELAs) estimated for former glaciers in the
Rwenzori Mountains (redrawn from Osmaston and Kaser, 2001). Note: vertical exaggeration is x
4.7.
4.2.3 Glacial meltwater drainage
Over the last century, glaciers and their meltwater flows have formed headwaters
of primarily three rivers in the Rwenzori Mountains: the River Mubuku that flows
in an easterly direction in the Republic of Uganda and Rivers Butawu and
Lusilube which drain toward the west in the Democratic Republic of Congo (Figs.
16 & 18). Under Lac Gris and Omurubaho glaciations during the Holocene,
glaciers and their meltwater flows would also have contributed to Rivers Ruanoli
and Nyamagasani. Ultimately, all of these rivers supply the River Semliki which
discharges into Lake Albert, a source of the White Nile. As such, the drainage
confirms the prescient claim of Claudius Ptolemy who wrote in 150 A.D. "... the
Mountains of the Moon, whose snows feed the lakes, sources of the Nile." (cited
in Osmaston, 2006). The contribution of the River Mubuku to the flow of the River
Semliki occurs via a more circuitous route that includes Lake George, the
Kazinga Channel, and Lake Edward (Fig. 3). The River Nyamugasani discharges
directly into Lake Edward.
Of the glaciers that presently remain in the Central Rwenzori Massif (Fig. 18),
most of the largest including East Stanley, Speke, Vittorio Emanuele, and
Margherita glaciers form headwaters of the River Mubuku. The East Stanley,
Margherita and Speke glaciers sustain meltwater streams that flow into Lake
Bujuku whereas the meltwaters from the Vittorio Emanuele glacier discharge to
the east into a northern tributary (Bukurungu) of the River Mubuku (Fig. 16).
Meltwater discharges from a comparatively smaller group of glaciers drain
westward into headwater basins of the River Semliki. From the Edward, Savoia
and Elena glaciers, meltwaters discharge into the Kitandara Lakes that
subsequently drain westward as part of the River Butawu basin (Fig. 16).
- 35
Meltwater discharges from Moebius, West Elena, West Stanley and Alexandra
glaciers also supply the River Butawu. Meltwater discharges from the now
fragmented Grant Glacier on Mount Speke form headwaters of the River
Lusilube.
Figure 18. Map of glacial extent and drainage in the Central Rwenzori Massif. Glaciers occur on
Mounts Speke (north), Stanley (west) and Baker (southwest). Redrawn and adapted from
Osmaston and Kaser (2001).
4.2.4 Precipitation
Precipitation in the Rwenzori Mountains occurs primarily during two pronounced
seasons from March to May and August to November (Fig. 19). The bimodal
pattern results from the regional movement of air masses associated with the
inter-tropical convergence zone (ITCZ). Unlike typical monsoon climates that are
derived from a reversal of wind currents from the northeast in January to the
- 36
southwest in July, a north-south reversal in east Africa causes the heavy rains to
occur in April and October.
Apart from the seasonal control on precipitation exerted by movement of the
ITCZ, there is a strong orographic effect on local precipitation. Mean annual
precipitation from 1964 to 1995 recorded at Kilembe (Fig. 14) at an elevation of
1370 mamsl is 1540 mm⋅a-1 whereas this flux drops to 890 mm⋅a-1 just 11km
away but 410 m lower in elevation at Kasese Airport (960 mamsl). The only
sustained measurements of precipitation within alpine areas of the Rwenzori
Mountains were collected by Osmaston (2006). Mean annual precipitation at four
locations from 1951 to 1954 similarly show pronounced variations with altitude
(Fig. 20). From the base of the mountains around 1250 mamsl, precipitation was
observed to increase with rising elevation from 1150 mm⋅a-1 to a maximum
annual precipitation of 2600 mm⋅a-1 recorded at 3290 mamsl in the Heath-moss
forest zone. Above this, precipitation decreased to 2000 mm⋅a-1 at Lake Bujuku in
the Afroalpine zone (Fig. 18) within the Central Rwenzori Massif.
Based on very limited data that are available from the west side of the Rwenzori
in the Democratic Republic of Congo, Osmaston (2006) suggests that
precipitation is lower on west side of the Rwenzori Mountains. Reduced
precipitation would explain the higher elevation of ELAs from contemporaneous
glaciations on the west-facing slopes (Fig. 17) and may reflect a ‘rain shadow’
assuming that the dominant source of precipitation to the Rwenzori Mountains
derives from the east (Indian Ocean). Fluctuations in sea surface temperatures of
the Indian Ocean are thought to account largely for the observed interannual
variability in East African precipitation (section 2.4). The source of precipitation to
alpine areas of the Rwenzori Mountains remains, however, complicated by the
fact that air currents of a different origin and direction than those at low
elevations are frequently observed in the mid-troposphere (Whittow, 1960).
Figure 19. Annual distribution in precipitation recorded at Kilembe, Uganda (0º13’N, 30º00’E)
from 1949 to 1996.
- 37
Figure 20. Plot of trends in observed precipitation and stream discharge with altitude in the
Rwenzori Mountains. The main vegetation zones (ecotones) are indicated for reference. Data
derive from this study and Osmaston (1989).
4.3 Methodology
The impact of glacial recession on alpine river riverflow was investigated through
both an analysis of trends in river discharge for basins receiving meltwater
discharges and an assessment of altitudinal variations in streamflow from spot
measurements. Apart from a few spot measurements of glacial meltwater
discharges summarised by Temple (1967), no measurements of alpine
streamflow have previously been collected at elevations above the river gauges
- 38
at the base of the Rwenzori Mountains (Fig. 15). Spot measurements of
streamflow conducted under this research project have, to date, been restricted
to dry seasons in June 2003 and January 2005. On-going research is assessing
how streamflow varies with elevation during the rainy season and apply stable-
isotope tracers of water movement to supplement physical measurements.
Accurate measurement of streamflow is hindered by the highly heterogeneous
nature of alpine stream reaches that feature large ice-rafted boulders (Fig. 21).
As a result, it was necessary to employ dilution gauging (Okunishi et al., 1992)
rather than the more common, velocity-area method. Analysis of potential
hydrometeorological trends associated with glacial recession over the last
century included records of precipitation and humidity from station observations
at the base of the mountain and regional hydrological records of lake levels and
river discharge. Trends in regional precipitation over the last two centuries were
investigated using both observational datasets and historical evidence.
Figure 21. Photograph of the confluence of the River Mubuku and its main tributary the River
Bujuku at sampling site no. 4 (Fig. 16). The widespread occurrence of ice-rafted boulders
produces complex stream reaches necessitating the use of dilution gauging to measure alpine
streamflow.
- 39
4.4 Longitudinal trends in alpine riverflow
Discharge records for the rivers Mubuku, Nyamagasani, and Semliki (Fig. 15) are
plotted in Figure 22. Each river is, or was recently, supplied in part by glacial
meltwater discharges. Records of river discharge have not been found for the
Rivers Butawu and Lusilube in the Democratic Republic of Congo and simply
may not exist. An unpublished report on hydrological measurements in the
Rwenzori Mountains by Kayondo (1967), cited by Temple (1967), has not been
located.
4.4.1 River Mubuku
The flow of the River Mubuku has for several decades been diverted for
generation of hydro-electric power (HEP) at an intake in Nyakalengija (Fig. 16).
Diverted flow is directed via a conveyor to the Ibanda HEP generating facility and
returned to the River Mubuku via one of its tributaries, the River Isha. Sluices and
drains at the Nyakalengija diversion permit peak flows to continue along the river
channel. A time lag in the response of the river gauge downstream (river gauge
site no. 16 in Fig. 16) to precipitation is expected as a result of the longer and
more circuitous route taken by diverted flow. The virtual elimination at the
beginning of the calendar year 1966 of a difference between maximum and
minimum flows casts doubt on the reliability of discharge records available from
1966 to 1971 and limits analysis of these records to a period from 1954 to 1965
inclusive. A second diversion of the River Mubuku for HEP generation by the
Kasese Cobalt Company Limited (KCCL) occurs at Bugoye, downstream of the
confluence of the River Mubuku and its tributary, the River Isha, but upstream of
gauging station no. 16 (Fig. 16). Records of river discharge begin in early 2000
but feature several gaps. At Bugoye, diverted flow is conveyed to the KCCL HEP
generating facility east of the River Mubuku downstream and returned to the
outlet of the River Mubuku (Figure 16) in the floodplain east of the Rwenzori
horst that includes Lake George (Fig. 2).
The River Mubuku has a catchment area of 256 km2 with a mean discharge from
1954 to 1965 (no totals for 1955, 1959 and 1960) of 14 m3⋅s-1 or a depth
equivalent over the catchment of 1730 mm⋅a-1. Although this period includes the
anomalously wet period of 1961 to 1963, the depth equivalent of mean annual
catchment riverflow is exceptionally high (1730 mm⋅a-1), three to five times that of
other basins (e.g., Nyamagasani) during the same period and of a comparative
size with similar evaporative channel losses. Indeed, this observation may have
led many to presume a significant contribution of glacial meltwaters to riverflow.
However, as altitudinal trends in riverflow indicate (section 4.5), the anomalously
high, depth equivalent of catchment riverflow arises primarily from the fact that
the catchment features very high, orographic precipitation that ranges from 2 to 3
m per year (Fig. 20), the highest annual flux in Uganda. Records of river
discharge (Fig. 22) show considerable interannual variability but are too limited in
- 40
duration to permit trend analyses. The anomalously wet conditions of the early
1960s are well represented in the discharge of the River Mubuku and Rivers
Nyamagasani and Semliki. As a result, the mean discharge for this period may
be slightly in excess of the long-term mean. For comparison, the mean discharge
of the River Mubuku for the slightly smaller catchment gauged by KCCL was 12
m3⋅s-1 over the calendar year 2000.
4.4.2 River Nyamagasani
The River Nyamagasani has a catchment area of 507 km2 and is the only river
draining the Rwenzori Mountains (in Uganda) that is currently monitored (Fig.
15). Available records of discharge from 1955 to 1978 are presented in Figure
22. A long gap in observations begins in 1979 and lasts until November 1998. An
automated recording gauge then operated at this site for more than two years
before being destroyed by flooding in 2001. The station was rehabilitated on
September 4, 2001. Similar to the River Mubuku, discharge records (Fig. 22) are
too limited in duration to permit trend analyses. In a qualitative sense, there is no
evidence of a clear decrease or increase in river discharge. Positive deviations in
mean annual discharge in the calendar years, 1964 and 1978, are consistent
with anomalously high precipitation recorded locally at Kilembe and Kasese
Airport (Fig. 15). No significant temporal trends (at a 95% confidence interval) in
precipitation are evident over a 50-year period (1949 to 1998) at Kilembe and 38-
year period (1964 to 2001) at Kasese Airport.
4.4.3 River Semliki
The River Semliki has a catchment area of 23 621 km2 and is the net recipient of
the discharges from all rivers and lakes draining the Rwenzori Mountains and the
surrounding region (Figs. 15 and 16). A continuous record of river discharge is
available over a 26-year period from 1952 to 1977 (Fig. 22). Although direct
inspection of the gauge for the River Semliki has not yet been undertaken, all
river discharge records in Uganda have recently been subjected to a quality
control analysis by the Water Resources Management Department (Uganda).
The anomalously wet conditions from 1962 to 1964, clearly represented in the
discharge record, complicate the assessment of any underlying trends in river
discharge. Indeed, a statistically significant trend at a 95% confidence interval is
not observed (p = 0.077). Overall, the observations point to a weak trend (r =
0.35) of increasing riverflow (1.3% or 2.7 mm⋅a-1 per annum) over the 27-year
period.
- 41
Figure 22. Longitudinal trends in deviations from mean riverflow observed at the base of the
Rwenzori Mountains from 1952 to 1978. Missing data for the Rivers Mubuku and Nyamagasani
reflect incomplete discharge records for these calendar years. Calculated errors derive from
uncertainty in the derivation of the rating curve. Data compiled by the Water Resources
Management Department of Uganda.
4.5 Altitudinal trends in riverflow from spot measurements
Altitudinal trends in the flow of the River Mubuku basin, deduced from spot
measurements collected during the season in January 2005, are plotted in Figure
20. These data show that contemporary glacial meltwater discharges aggregated
at Lake Bujuku represent a tiny fraction (<0.2 m3⋅s-1) of the mean river discharge
(14 m3⋅s-1) recorded from 1954 to 1971 at the base of the Rwenzori Mountains
(Fig. 15). The altitudinal profile of river discharge highlights, furthermore, the
relative increase in river discharge that occurs below the glaciers (>4500 mamsl)
- 42
in the Heath-moss and Montane forest zones between 2000 and 4000 mamsl as
the catchment area expands and where precipitation is highest. The minimal
influence of glacial meltwaters on alpine riverflow in the East African Highlands is
contrary to speculation (Gasse, 2002; Thompson et al., 2002) but consistent with
a previous suggestion by Temple (1967) and evidence from Kilimanjaro (Kaser et
al., 2004). The conclusion that recent glacial recession has had a negligible
impact on alpine riverflow is sensible in light of the fact that less than 0.5%
(1.0±0.3 km2) of the River Mubuku basin (256 km2) is covered by glaciers. On-
going research seeks to confirm these deductions by taking measurements
during the rainy season in April 2007.
4.6 Meteorological trends over the last century
As glacial recession does not constitute a direct threat to the magnitude of alpine
riverflow, research then focused on the possibility that receding glaciers signal a
reduction in alpine precipitation, potentially lowering riverflow, and/or humidity.
The former would reduce glacial accumulation whereas the latter may reflect a
decrease in cloud cover that increases the exposure of glaciers to solar radiation
that enhances ablation. Records of river discharge for the Mubuku basin are,
however, inadequate to permit analysis of longitudinal trends and thus represent
potential changes in alpine precipitation. Analysis of the only long-term records of
precipitation in western Uganda which is restricted to stations in lowland areas
<2000 mamsl (Fig. 10), highlights significant interannual variability discussed in
section 2.4. Significant downward trends in precipitation are not realised at any of
the four stations. Small but significant trends of increasing precipitation (at a 95%
confidence interval) are observed at two stations (Fort Portal, Mbarara).
Daily measurements of relative humidity are available over two continuous
periods from 1967 to 1974 and 1991 to 2000 at the base of the Rwenzori
Mountains in Kasese (Fig. 15). Although these records are too brief to permit an
analysis of temporal trends, a comparison of mean values of relative humidity as
well as mean minimum and mean maximum relative humidity for these two
periods fails to support the decline in relative humidity (Table 2) proposed by
Mölg et al. (2006) based on reanalysis datasets. Marginal increases in relative
humidity are uniformly observed.
A key limitation in the relationship between lowland and alpine precipitation
presumed in the above analyses is that the source of precipitation remains the
same. As noted by Whittow (1960), air currents of a different origin and direction
than those at low elevations are frequently observed in the mid-troposphere in
the Rwenzori Mountains. Current research seeks to trace the origin of
precipitation in lowland and alpine areas on both sides of the Rwenzori
Mountains using the stable isotopes of O and H.
Table 2. Mean, mean minimum, and mean maximum relative humidity (%) from measurements
available over two continuous periods, 1967 to 1974 and 1991 to 2000, at Kasese (Fig. 15). Data
- 43
derive from daily measurements recorded at 9AM and 3PM. The standard deviation in each mean
value is given in parentheses.
relative humidity 1967 – 1974 1991 – 2000
mean - 9AM 80 (6) 84 (5)
mean minimum – 9AM 63 (8) 66 (9)
mean maximum – 9AM 94 (4) 97 (3)
mean – 3PM 52 (7) 54 (7)
mean minimum – 3PM 34 (8) 37 (7)
mean maximum – 3PM 82 (10) 86 (10)
4.7 Hydrological trends over the two centuries
According to Mölg et al. (2006), glaciers in the East African Highlands are relicts
from more humid climatic conditions during the 19th century. Although this
hypothesis might explain the monotonic decline in glacial cover in East African
Highlands over the 20th century (Fig. 23), there is currently no compelling
evidence (e.g. measurements of precipitation or humidity during the 19th century)
to support it. As recognised by Kaser and Osmaston (2002), the apparent
linearity in rate of decline in glacial cover over the 20th century may stem from the
limited number of measurements. More frequent observations of glacial cover on
Kilimanjaro toward the end of the 20th century reveal, for instance, a slowing the
rate of loss of glacial cover (Fig. 23) and raise the possibility that glaciers there
may persist up to ~2040 rather than ~2020 suggested by Thompson et al.
(2002).
Figure 23. Observed changes in glacial cover on the Rwenzori Mountains and Kilimanjaro over
the 20th century. Data for Kilimanjaro are given in Thompson et al. (2002). Rwenzori data from
1906, 1955 and 1990 derive from Kaser and Osmaston (2002). Remaining Rwenzori data derive
from NDSI classification of two Landsat 5 (TM) and one Landsat 7 (ETM+) scenes (Table 1).
- 44
Changes in the lake levels, an indirect and imperfect proxy of regional humidity,
are often cited (e.g. Mölg et al., 2006) in support of the assertion of that a
dramatic reduction in humidity in East Africa occurred towards the end of the 19th
century (ca. 1880). Nevertheless, historical evidence of changes in the level of
Lake Victoria (Fig. 24), speculative as the data may be, indicate that for all but
two decades of the 19th century, the level of Lake Victoria was lower than, or
commensurate to, its level throughout the 20th century. A very similar rise and fall
in the level of Lake Tanganyika during the latter half of the 19th century is also
suggested from historical evidence (Nicholson and Yin, 2001). During the 20th
century, lake levels and the discharge of major river systems draining East Africa
(i.e., Rivers Nile and Congo) exhibit remarkable consistency in their response to
years or seasons of anomalously high precipitation (e.g. 1917, 1961-1964, 1969,
1979). These records are, furthermore, consistent with observations from
meteorological stations in western Uganda in that they show strong interannual
variability but provide no evidence of a decline in precipitation or by speculative
inference, atmospheric humidity, over the 20th century.
4.8 Concluding Discussion
Meltwater flows from glaciers in the Rwenzori Mountains do not contribute
significantly (> 0.5%) to alpine riverflow. This conclusion, based on stream fluxes
measured during the dry season, is consistent with suggestions from Temple
(1967) and Osmaston (2006), and a similar assertion by Kaser et al. (2004)
based on observations on Kilimanjaro. There is, furthermore, no evidence of a
decline in precipitation over the 20th century in lowland station records from
western Uganda and regional hydrological records. Based on historical evidence
of changes in the levels of Lake Victoria and Tanganyika over the 19th century
(Fig. 24), the dramatic shift in atmospheric humidity purported to have initiated
glacial recession in the late 19th century is actually the descending limb of a brief,
high lake stand which lasted for less than two decades. Lake levels do not signal
a sustained shift from a more humid climate to a subsequent, less humid one.
Insufficient data exist to verify local reports in Uganda (Kasese District) of an
increased frequency and magnitude of flood events at the base of the Rwenzori
Mountains. Changes in land use and land pressures outside of the Rwenzori
Mountains National Park (i.e., below 1700 mamsl) in the River Mubuku basin
may account, in part, for the perceived increase in flood risk.
- 45
Figure 24. Regional hydrological trends in lake levels and river discharge since 1800. Data derive
from Nicholson and Yin (2001), Water Resources Management Department (Entebbe, Uganda),
Hurst and Phillips The Nile Basin, Supplements 1931-1950, and Vörösmarty et al. (1998).
- 46
5 Recent changes in aquatic productivity of Lake
Bujuku
5.1 Introduction
Alpine environments are hotspots of biodiversity and considered among the most
sensitive to climatic changes occurring on a global scale (Diaz et al. 2003).
Recent warming observed in tropical alpine regions (e.g. Liu and Chen, 2000;
Vuille and Bradley, 2000), has been linked to increased incidence of malaria in
the East African Highlands (Pascual et al. 2006) and rapid retreat of glaciers
(Bradley et al. 2006; Thompson et al. 2006; Taylor et al. 2006a). Because
tropical alpine glaciers are highly sensitive to changes in climate, they serve as
important indicators of environmental change in these regions (Kaser et al. 2004)
where meteorological observations are often limited. Rapid glacial retreat over
the 20th century has been observed on all three ice-fields in the East African
Highlands including Kilimanjaro (e.g. Thompson et al. 2002), Mount Kenya (e.g.
Hastenrath and Kruss, 1992) and the Rwenzori Mountains (Chapter 2, Fig. 23).
Environmental changes associated with the recession of alpine glaciers in East
Africa remain unclear. In contrast to long-term monitoring records (including
documentary evidence) from alpine regions in temperate latitudes (e.g. Koining
et al. 1998, 2002; Lotter and Bigler, 2000; Psenner and Schmidt, 1992), there is
a paucity of similar datasets in tropical alpine regions of Africa (Verburg 2003;
Verschuren 2003). In light of rapid tropical glacier recession, well-resolved, high
quality palaeolimnological records from African lakes are especially important
(ibid.).
This chapter investigates the recent palaeolimnological record from Lake Bujuku,
a high altitude lake adjacent to the ice-fields in the Rwenzori Mountains (Fig. 18).
Reconstructions focus on the recent past (ca. 150 years) when the glaciers here
have been in significant retreat, and for which we are able to produce a robust
chronology using radiometric 210Pb and 137Cs analyses. We adopt a multiproxy
approach, based on algal production within the lake and vegetation changes in
the immediate catchment (diatoms, pollen and organic carbon isotope
composition, including TOC, C:N and 13C. The aim of this paper is to investigate
the sensitivity of remote, tropical, alpine lakes in documenting environmental and
to thereby provide a case study and palaeoecological context for 20th century
climatic change observed in East Africa.
5.2 Study site
Lake Bujuku (0º22’N and 29º54’E) lies at an altitude of 3960 metres above sea
level (alpine) within the Rwenzori Mountains National Park (RMNP). The lake’s
catchment features steep, scree slopes in its upper reaches (Fig. 25) and is
partly fed by meltwater flows from the East Stanley and Speke glaciers. The lake
- 47
is surrounded on three sides by the last remaining mountains that support ice-
fields (Mount Baker, Stanley and Speke) (Fig. 18). Rainfall in the order of 2000
mm⋅a-1 is controlled by the seasonal displacement of the Inter Tropical
Convergence Zone [ITCZ] which predominantly falls during the region’s wet
seasons (March to May and August to November) (Osmaston, 2006). The lake is
located within the Afro-alpine zone that is characterised by vegetation dominated
by Lobelia wollastonii (giant lobelia), Dendroscenecio advinalis (tree senecio)
and Erica spp. (including giant heather). At the edges of the lake, sedges and
grasses dominate. Both lake inflows and outflow pass through an extensive
Carex runssoroensis bog, with the outflow feeding through the cascading system
of bogs (Fig. 18). Diurnal temperatures change from a mean minimum to mean
maximum of -1ºC to 10ºC which contributes to the polymictic nature Lake
Bujuku’s water column (Temple, 1967). Due to its tropical location, the site
receives near constant annual solar radiation (Livingstone, 1967).
Figure 25. Photograph of Lake Bujuku from the Speke Glacier (June 22, 2003).
- 48
5.3 Methodology
5.3.1 Field Methods
A bathymetry of Lake Bujuku was created by making a number of traverses
across the basin while recording water depth with a Plastimo echosounder and
simultaneous location using a hand-held e-Trex Summit GPS. Data were
interpolated using ArcView. Although a maximum depth of 14.5 m was recorded
in the lake, a more uniformly flat region at a depth of 13.5 m (Fig. 26) was
selected for coring. Four cores (Buju1-4) were collected using a Glew gravity
corer (Glew, 1991). The longest sequence retrieved was 40.5 cm (Buju3) and is
used in this study for all analyses. Extrusion was conducted in the field at a
resolution of every 2.5 mm (from 0 to 10 cm) and thereafter every 5 mm. An
assessment of the contemporary characteristics of the lake was also made.
Eight diatom samples were collected from littoral regions of the lake, including
epilithic, epipsammic and epiphytic habitats (Fig. 26). Core site water
temperature readings gave a spot reading at the time of coring of 8.1ºC ± 2.1ºC,
a pH of 7.2 ± 0.3 units and a Secchi-disc depth reading gave an estimate of light
penetration of 4.5 m.
Figure 26. Digitised image of Lake Bujuku bathymetry, inflows, outflow and bog region. Numbers
refer to contemporary diatom sampling sites (Table 3) and X refers to core location of Buju3.
- 49
5.3.2 Laboratory methods
5.3.2.1 Chronology
Sediment samples were dated using 210Pb and 137Cs by non-destructive gamma
spectrometry (Appleby and Oldfield, 1992) at the Centre for Environmental
Research, University of Sussex. Twelve core sub-samples were counted for at
least 8 hours on a Canberra well-type ultra-low background HPGe gamma ray
spectrometer to determine the activities of 137Cs, 210Pb and other gamma
emitters. Sediment accumulation rates were determined using the ‘simple model’
of 210Pb dating (e.g. Robbins, 1978), where sedimentation rate is given by the
slope of the least squares fit for the natural log of the 210Pb excess activity versus
depth.
5.3.2.2 % DW and LOI analyses
Eighty samples (every 5 mm) were analysed for wet densities (WD), percentage
dry weight (%DW) at 105ºC and percentage loss-on-ignition at 550oC (% LOI) as
an estimate of organic carbon. Calculations followed procedures outlined by
Charles et al. (1994) and Bengtsson and Enell (1986).
5.3.2.3 Organic geochemistry
C:N ratios were calculated to examine the relative importance of autochthonous
and allochthonous sources of organic material within Lake Bujuku sediments
whereas the stable isotope ratios of carbon (13C/12C) were analysed to trace the
dominant plant source of carbon in the lake (Talbot and Laerdal, 2000). Bulk
carbon samples at a resolution of every 5 mm were prepared by placing 2 g of
wet sediment overnight in 50 ml of 5% hydrochloric acid to remove carbonates.
Subsequently, they were washed using deionised water through Whatman No.
41 filter papers and dried in air at 40ºC. Once ground into a fine powder, 13C/12C
analyses were performed by combustion using a Carlo Erba NA1500 (series 1)
on-line to a VG TripleTrap and Optima dual-inlet mass spectrometer, with 13C
values calculated to the VPDB scale using a within run laboratory standard
calibrated against NBS19 and NBS22. C/N ratios were determined by reference
to an Acetanilide standard, and replicate analyses of well-mixed samples indicate
a precision of + <0.1 ‰ (1 S.D.).
5.3.2.4 Diatom analyses
Procedures for diatom analysis followed those outlined by Battarbee (1986)
involving digestion of 0.1 g of wet sediment in 30% H2O2 and HCl before
mounting on slides with Naphrax (Renberg, 1990). Resolution followed every
5mm between depths 0 and 1 cm, every 10 mm between 1 and 20 cm and every
20 mm between depths 20 and 40 cm. A known concentration of microspheres
was added to samples in order to calculate diatom concentrations. Diatoms were
- 50
counted using oil immersion phase contrast Leica Axiostar light microscopy at
x1000 magnification. At least 300 valves were counted for each sample, revised
to a maximum of 500 valves when dominating Fragilaria sp. represented > 70 %
of sample assemblages. Taxa were identified according to a range of published
papers and books including Krammer and Lange-Bertalot (1986, 1988, 1991a,
1991b), Camburn and Charles (2000) and those with tropical African taxonomy
(Gasse, 1986). The resultant stratigraphy, created using C2 v1.4 (Juggins, 2004)
and displayed on the constructed time scale, was divided into zones using
stratigraphically constrained cluster analysis by incremental sum of squares
(CONISS) using the programme ZONE Version 1.2 (Juggins, 1992).
5.3.2.5 Pollen analysis
Approximately 0.05 g of freeze dried sediment was weighed and tablets
containing a known number of Lycopodium spores were added to calculate
pollen concentrations (Stockmarr, 1972). The samples were further processed
using standard methods (Faegri and Iversen, 1975). Resolution followed every 5
mm between depths 0 and 1 cm, every 10 mm between 1 and 4 cm and every 20
mm between depths 4 and 40 cm. Between 500 and 1000 terrestrial pollen
grains and spores were counted using a Zeiss D-7082 microscope (identified at
x1000 under oil immersion). The total terrestrial plant pollen and spore sum was
used for pollen calculation of all terrestrial taxa. Aquatic pollen is also expressed
relative to this terrestrial sum. Species were retained when they were > 5%
abundance and grouped according to their ecological affinities. C2 v1.4 was
used to construct the stratigraphy and diatom zonation was applied to aid
interpretation.
5.3.2.6 Multivariate Analyses
Statistical analyses were carried out on diatom and pollen taxa present in
samples with an abundance of > 2%. Unconstrained (detrended correspondence
analysis - DCA) and principal components analysis (PCA) ordinations were
carried out on Buju3 diatom and pollen data using the software package
CANOCO 4.5 (ter Braak and Šmilauer, 2002). The principal gradients of floristic
variation within the core were assessed prior to ordination analysis using DCA,
with detrending by segments, non-linear scaling of axes, square-root
transformation of species data and down-weighting of rare species. A first axis
gradient of 1.067 SDs was obtained for diatom data and 0.621 for pollen, thus
core data were subsequently analysed using the linear ordination technique of
PCA (ibid.). Axes 1 and 2 of the diatom data were further analysed with broken-
stick analysis to test their significance (Jollifer, 1986).
- 51
5.4 Results
5.4.1 Contemporary diatom ecology
Fragilaria pinnata, F. construens and A. minutissima are represented in littoral
samples (Table 3). However, T. flocculosa dominates 61% of the total valve
abundance at site 6 and is present at all the sites apart from the epipsammic
sample (site 8). Littoral habitats are more diverse than suggested by the surface
sediment sample from Buju3 (Table 3). For example, neither Tabellaria
flocculosa nor Synedra spp. are represented in Buju3 yet they represent a large
proportion of the species abundances in the contemporary assemblages.
Synedra tenera and S. linearis also appear in the contemporary sites, not
represented in Buju3, and account for 58% of assemblage abundance at site 5.
The representivity of modern assemblages is addressed more fully in the
discussion.
Table 3. Summary of diatom assemblage compositions from contemporary sampling at eight
sites around the shore of Lake Bujuku. Only the most abundant species, and their relative
abundances, at each of the sites is displayed.
Site 1 Achnanthes
minutissima (70%)
epilithic Synedra linearis (12%)
Tabellaria flocculosa
(8%)
Site 2 Fragilaria capucina
(13%)
epiphytic F. pinnata (7%)
S. linearis (20%)
S. tenera (14%)
T. flocculosa (35%)
Site 3 A. minutissima (14%)
epilithic Brachysira brebissonii
(5%)
F. contruens (5%)
F. pinnata (31%)
T. flocculosa (11%)
Site 4 A. minutissima (14%)
epiphytic B. brebissonii (6%)
F. pinnata (7%)
S. tenera (18%)
T. flocculosa (22%)
Site 5 F. capucina (5%)
epiphytic Gomphonema
parvulum (5%)
S. linearis (26%)
S. tenera (32%)
T. flocculosa (18%)
- 52
Site 6 Eunotia bilunaris (4%)
epiphytic F. capucina (4%)
S. linearis (8%)
S. tenera (8%)
T. flocculosa (61%)
Site 7 E. bidens (9%)
epilithic F. pinnata (12%)
S. linearis (7%)
S. tenera (16%)
T. flocculosa (16%)
Site 8 A. minutissima (7%)
epispammic
E. bidens (7%)
F. capucina (12%)
F. contruens (7%)
F. pinnata (21%)
G. parvulum (26%)
5.4.2 Chronology
210Pb activity in Buju3 shows a broadly exponential decline with depth (Fig. 27).
The 210Pb derived sediment accumulation rate, based on a constant rate of
sedimentation, was 2.9 mm⋅a-1 (2 SD range = 2.5 – 3.4 mm⋅a-1). 137Cs shows a
clear subsurface maximum in activity at 10 cm in Buju3 that is most likely a
stratigraphic marker for 1963, the year of peak global 137Cs fallout from
atmospheric nuclear weapons testing (Fig. 27a). This therefore suggests an
accumulation rate of 2.6 mm⋅a-1 from the base to the uppermost sediments. Both
are in good agreement therefore and suggest that Buju3 spans ca. 140 years
based on a continuous rate of deposition, dating back to 1860 ± 20 years (Figure
27b).
5.4.3 Organic geochemistry
For explanatory purposes, the four diatom zones have also been applied to the
organic geochemistry and lower resolution pollen stratigraphies (Fig. 28). C2 v.
1.4 was also adopted to create stratigraphies for organic geochemistry (Juggins,
2004).
5.4.4 % TOC
In the case of Buju3 there has been a gradual increase with time in %TOC, from
ca. 6% at the base of the core to ca. 13% at the surface. This increase has been
predominantly steady (apart from a more rapid increase in values between c.
1895 and 1910 of ca. 3%) and suggests a gradual increase in productivity in the
sequence, especially from ca. 1930 onwards.
- 53
Figure 27. (a) 137Cs apex acts as a stratigraphic marker for 1963; (b) plot of age versus depth for
Buju3, based on the simple model of 210Pb dating. Error bars shown are calculated using the 2
SD error on the gradient of the linear regression fit of ln210Pb excess vs. depth.
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
Years BP
ZONE I
ZONE II
ZONE III
ZONE IV
-28.0
-27.0
-26.0
-25.0
-24.0
-23.0
d13C
12.0
13.0
14.0
15.0
16.0
17.0
C/N
0
5
10
15
TOC
Figure 28. Age (depth) profiles of bulk organic carbon analyses: 13C (‰), C/N and % TOC.
Zones are defined by diatom data.
137Cs vs. depth, core Buju 3
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
0 10 20 30 40
Activity (Bq/kg)
Depth (cm)
Age vs. depth graph, Buju3 core
-35
-30
-25
-20
-15
-10
-5
0
0 50 100 150 200
Age (years)
Depth (cm)
- 54
5.4.5 C/N ratios
C/N ratios for the Buju3 sequence remain between 12 and 17 indicating a
predominantly submerged macrophyte and/or aquatic algal source for carbon in
Lake Bujuku (Meyers and Teranes, 2001). C/N ratios increase to 16.5 during the
% TOC peak in zone III between ca. 1895 and 1910. This peak, although
consistent with a signature of submerged and aquatic vegetation, may also
represent a mixed aquatic allochthonous source event in the record here. After
ca. 1975 there is a shift in C/N values suggesting an increase in algal dominance
in the lake (12.4).
5.4.6 Stable isotopes of carbon
13C shows a gradual lowering in values over the period of reconstruction from -
25‰ to -27.5‰. C/N ratios suggest that this carbon source is autochthonous
(algal predominantly) and corroborates increasing % TOC values in Buju3.
Assuming no significant plant community change within the lake (and significant
changes in the inwash of terrestrial material can be discounted by the C/N ratio
data) then one interpretation is that this represents an increase in productivity at
Bujuku.
5.4.7 Diatoms
Throughout the diatom record three species dominate, F. construens and A.
minutissima and F. pinnata, with the latter being the most abundant (Fig. 29). In
zone IV F. pinnata accounts for > 40% of total species abundances although
during this zone F. construens is also shown to increase to between c. 15-20%
abundance. These increases in the dominance of F. construens are concurrent
with reductions in the relative abundance of F. pinnata. The increasing trend
seen from zone IV by F. construens continues throughout the record until zone I.
PCA axis 2 displays the shift between these Fragilaria species. Diatom
concentrations and flux rates decline in number from the base of the core ca.
1860 from 19 x 105 (valves g per wet weight) and 4.9 x 105 g⋅cm-1⋅a-1
respectively. After that, they begin to increase at ca. 1880 and towards zone III
once again.
Flux rates and concentrations show a double peak in values at ca. 1900 (2.3 x
105 g⋅cm-1⋅a-1 and 8.9 x 105 valves g per wet weight respectively) and 1930 (2.0
x 105 g⋅cm-1⋅a-1 and 7.7 x 105 valves g per wet weight respectively) with a fall in
numbers c. 1920. By ca. 1960 there is also an increase in F. capucina, which by
1970 declines once again. By 1975 and towards the transition with zone I, this
species increases once more reaching 10%. The dominance of F. pinnata
increases after c. 1975 once more (accounting for >60%), with an increase of
23% in the space of 3 years. This change is reflected by PCA axis 1 scores
shifting to positive values. F. pinnata values then begin to decline after ca. 1990
while an increase in F. construens (up to 20%) and A. minutissima, the former
- 55
Figure 29. Diatom stratigraphy as percentage abundances for Buju3. Valve concentrations,
diatom flux rates and PCA axes scores 1 and 2 are also shown. Zones applied are derived from
constrained cluster analysis.
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
Years BP
Zone II
Zone III
Zone IV
0 20
Achnanthes minutissima
0
A. thermalis
0
Brachysira brebissonii
0
Cymbella minuta
0 20 40
Fragilaria construens
0
F. brevistriata
0 20
F. capucina
0
F. elliptica
0 20 40 60 80
F. pinnata
0
Navicula crytocephala
0
N. halophila
0 4 8 12 16 20
Valve concentrations x 10 5 (valves g-1 wet weight)
0.0 1.0 2.0 3.0 4.0 5.0
Diatom flux rates x 10 5 (valves g/cm/yr)
-2.0 -1.0 0.0 1.0 2.0 3.0
PCA axis 1 scores
-2.0 -1.0 0.0 1.0 2.0 3.0
PCA axis 2 scores
- 56
recovering from its lowest abundances in the profile, is seen. Once again valve
concentrations increase within the profile at c. 1990 (11.3 x 105 valves g per wet
weight), as do valve accumulations 2.9 x 105 (valves g⋅cm-1⋅a-1) although these
values drop slightly towards the present day.
5.4.8 Pollen
A total of 89 pollen and spore types were identified (Fig. 30), although here we
have classified the taxa into the following groups: non-local trees, local trees,
herbs, aquatics and spores (mainly from undifferentiated ferns and Sphagna).
The pollen group stratigraphy shows little variation during zone IV, with
indications of non-local and local tree pollen decreasing after ca. 1880, while
herbs remain dominant (between 34 and 44%). Aquatic pollen shows an
increase in numbers concomitant with the fall in diatom flux rates, after which
they begin to increase towards zone III.
Pollen results show that aquatic pollen shows that greatest change during zone
III (from ca. 5 to 20 %), with PCA axis 2 scores also reflecting this increasing
trend. This is most noted after ca. 1900. Herbs show little change until ca. 1920
after which values begin to fall to < 30% and a temporary increase is seen in
local terrestrial pollen. However, during zone III there is a gradual decline in
aquatic pollen that can be seen after ca. 1950 concomitant with the fall in diatom
concentrations. This is again reflected by the pollen PCA axis 2 scores. After
ca. 1980, aquatic pollen shows small, fluctuations towards the present day
concomitant with variations in F. construens and F. pinnata valve abundances.
5.5 Discussion
Despite evidence that glaciers in the RMNP have undergone rapid retreat since
their first documented measurement in 1906 (Kaser and Osmaston, 2002),
catchment and limnological changes recorded in Lake Bujuku have been more
subtle. Central to our interpretations are the limnological controls on diatom
species representation and abundance within the sedimentary record, and these
are addressed first here.
The limnology of Lake Bujuku has major consequences for diatom presence and
distribution in the lake, related to the lake’s location and altitude. As highlighted
above, Lake Bujuku is polymictic and subject to diurnal temperature fluctuations
that are more pronounced than annual changes. Temperature readings
conducted by Livingstone (1967) found that the difference between top and
bottom waters was insufficient to maintain stratification, whereas exposure to
strong winds (due to minimal geostrophic influence in the tropics) will also be
responsible for the limited development of thermal stratification in this lake
(Talling and Lemoalle, 1998).
- 57
Figure 30. Pollen stratigraphy from Buju3 as percentage abundances. Results are grouped
according to their family for ease of interpretation and displayed with PCA axes scores 1 and 2.
Zones applied are derived from constrained cluster analysis carried out on diatom data to aid
interpretation.
186 0
187 0
188 0
189 0
190 0
191 0
192 0
193 0
194 0
195 0
196 0
197 0
198 0
199 0
200 0
Yeasr BP
Zone I
Zone II
Zone III
Zone IV
0 10 20 30 40 50
Non Local Trees
0 10 20 30 40 50
Local Trees
0 10 20 30 40 50
Herbs
0 10 20 30 40 50
Aquatic
0 10 20 30 40 50
Spores
-2.0 -1.0 0.0 1.0 2.0 3.0
PCA axis 1 scores
-3.0 -1.0 1.0 3.0
PCA axis 2 scores
Sediment Depth (cm)
- 58
Adaptive Fragilaria sp. can take advantage of dynamic water columns. Continual
turnover allows the recycling of oxygen and nutrients from the profundal zone,
which was confirmed by Richardson (1968) who demonstrated that high amounts
of oxygen occurred at the bottom of Lake Bujuku. A further ecological preference
associated with oxygen charged bottom waters is associated with the species A.
minutissima, which is regarded as the best indicator of oxygen rich conditions
(Gasse, 1986). Evidence for the polymictic nature of Lake Bujuku can account for
the presence of these i) non-planktonic species (with requirements of high
oxygen values) and ii) tychoplanktonic species (readily adaptive to dynamic
thermal characteristics and to both benthic and planktonic environments (Barker
et al., 1994) within the sediment profile.
The contemporary ecology of Lake Bujuku shows the presence of many of the
abundant species within the stratigraphy and particularly the surface sediments
of Buju3 (Table 3). Although common taxa include A. minutissima, F. pinnata, F.
construens and F. capucina, several abundant species in the littoral habitats
(Tabellaria flocculosa, Synedra tenera and S. linearis) are poorly represented
within the sub-surface samples. It is possible that these taxa are not present in
the sub-surface layers as a result of ecological change in recent years. However,
it is also possible that these species are under-represented in the stratigraphy
because of i) taphonomic processes such as dissolution, which can preferentially
impact finely silicified taxa such as Synedra (Ryves et al. 2001) and ii) core
location with respect to the basin morphology of the lake and relative size of the
littoral shelf in comparison to lake volume (Jones and Flower 1986). It is
important to appreciate that taphonomic processes, determining the formation of
fossil diatoms, are often site specific and are issues overlooked in the field
(Cameron, 1995).
Taxa, such as T. flocculosa, S. linearis, and S. tenera dominate macrophyte
habitats in littoral margins (Werner, 1977). Small chain forming Fragilaria species
which dominate Buju3, (including Fragilaria pinnata) are regarded as R-strategist,
benthic taxa, meaning they are able to reproduce at a comparatively higher rate
than other taxa when environmental conditions rapidly change (Lotter and Bigler,
2000). As a result, they often dominate dynamic environments. Moreover, they
have often been classified as tychoplanktonic (Sayer, 2001), showing the
adaptive nature of the genus that takes advantage of both benthic and planktonic
environments (when nutrients are more limiting; Barker et al. 1994).
Unfortunately, little is documented on the individual ecologies of Fragilaria
species. Moreover, due to its cosmopolitan nature, a multiplicity of evidence
exists regarding their optimal ecology, thereby complicating palaeolimnological
interpretations when they dominate sequences. Nevertheless, F. pinnata appear
to tolerate harsher conditions than other species, and is often regarded as a
disturbance indicator, found in benthic environments growing on sediments (Barr,
pers. comm.). Other species such as F. capucina require high amounts of light
- 59
penetration for growth (Sweets pers. commun.). Palaeoenvironmental
interpretations below are based in part on these observations.
The photic depth of Lake Bujuku is argued to be a significant factor accounting
for the dominance of benthic taxa in Buju3. Although Secchi-disc depth readings
obtained at the core site define the photic zone to be 4.5 m deep, Cole (1975)
suggests that the photic zone of a lake may be greater than this. Here, diffusive
light remains in the water column which is scattered by particulate matter,
causing deeper light. In the case of oligotrophic lakes such as Lake Bujuku,
factors of 2.7 to 3.0 times Secchi-disc depths have often been found to be a
better estimated depth of the photic zone (Davis and Brinson, 1980). As a result,
the photic depth of Lake Bujuku could reach in excess of 13 m, signifying that a
large proportion of the water column and therefore lake basin is within the photic
zone. Furthermore, Kinzie et al. (1998) argue that high ultraviolet radiation (UVR)
irradiance incident on tropical mountain ecosystems, has a greater negative
effect on photosynthetic aquatic communities than at a lower elevation. The
amount of UVR exposure can be up to 20% higher than at sea level of
comparable latitude (ibid.). Such exposure may be one reason why the
stratigraphy of Buju3 displays predominantly benthic, R-strategist taxa, as
phytoplankton cells in frequently circulating waters cannot spend extended
periods of time exposed to UVR. Indeed, Vinebrooke and Leavitt (1996) state
that the periphytic species A. minutissima in alpine lakes is sensitive to increases
in UVR due to its inability to migrate upon its substrates for refuge, compared
with other periphytic species. Such decreases in this species, towards the latter
part of the 1990s at Bujuku, may suggest that UVR has indeed been increasing
in this region.
5.5.1 Palaeolimnological reconstruction over period of glacial retreat
Over the period of observed glacial retreat since 1906 in the Rwenzori
Mountains, the diatom record of Lake Bujuku contains only relatively subtle
responses to global warming and glacier recession in comparison to its higher
latitude alpine counterparts (e.g. Koining et al. 1998, 2002; Lotter and Bigler,
2000; Psenner and Schmidt, 1992). Meyers and Teranes (2001) highlights that
C/N ratios can be used to distinguish between algal and higher plant carbon
sources in lake sediment sequences. In particular, as in this case, levels between
10 and 20 are typical of a mixed source of lacustrine algae (< 10 – 12) and
submergent/floating aquatic macrophytes (Tyson, 1995). Over the whole period,
organic carbon isotope analysis demonstrates a gradual, but distinct, increase in
productivity in the lake as a whole (i.e. a shift to more negative 13C values and
increasing % TOC). Leng et al. (2006) highlight that such periods of increased
productivity are explained due to the preferential uptake of 12C by aquatic plants
during photosynthesis; as this increases, the carbon pool in the water is enriched
in 13C. C/N ratios indicate that productivity has been largely dominated by
aquatic macrophytes, until ca. the last 30 years, when ratios highlight a shift in
algal productivity dominating the ecosystem. This change, although not reflected
- 60
in the aquatic pollen record, is accompanied by the largest shift in diatom
assemblage composition in the core where F. construens abundances show a
marked decline at this time, to be replaced by F. pinnata.
The possible habitat changes associated with this shift are not able to be
discussed here. Indeed, it is difficult to elucidate whether the species change
after 1975 is due to inherent within-lake variability or a more direct response to
environmental change in the region, as the glaciers at this time undergo a
continuous recessional trend (Taylor et al. 2006a). In relation to taphonomy and
representation issues at Bujuku, littoral coring and macrophyte analyses would
be valuable in order to investigate macrophyte and submergent community
change at this time, allowing for a more comprehensive investigation of the
changes observed here.
It is evident therefore that both catchment changes (as evidenced by the pollen
record) and diatoms, which have continued to be dominated by benthic Fragilaria
species, may simply be insensitive to the relatively subtle changes that may have
occurred in temperature, humidity and solar radiation which have driven glacial
recession. Such changes (e.g. small change in temperature requiring ice to melt)
may not have readily affected catchment vegetation and the limnology of Lake
Bujuku. At this stage, we cannot rule out issues of taphonomy and representivity
dampening any potential impacts from glacial recession on aquatic responses in
the lake, but we are as yet unable to quantify these. For example, during the
early 1960s, unusually high precipitation levels led to a 2.5 m rise in the level of
Lake Victoria (Kite, 1981) and a small advance in the terminal positions of
glaciers in the Rwenzori Mountains (Temple, 1967), yet even these extreme
events are not documented in the palaeolimnological record from Lake Bujuku.
Moreover, although there is a distinct increase in TOC in the profile
(accompanied by a small increase in C/N ratio) between c. 1890 - 1910, these
changes are again not reflected in either the diatom or pollen records.
There is unambiguous evidence of rapid glacial mass loss in east Africa and
other alpine regions of the tropics yet the precise nature of climatic change and
its historical context over the past two centuries remain poorly resolved. This
emphasises the need for high resolution, multiproxy palaeoecological studies to
resolve this. Interestingly, it would appear that at least the pollen records at
Lake Bujuku do not record the dramatic recession of the East Stanley and Speke
glaciers, despite the lake’s geography (high, remote catchment, partially fed by
retreating glaciers), although organic carbon isotope analysis suggests an
increase in lake productivity, possibly in response to warming over the last 150
years.
- 61
5.6 Conclusions
Analysis of diatom and pollen assemblages in Buju3 have shown that, unlike the
glacial archives (Chapter 2), the Afroalpine lake ecology and flora have not
undergone dramatic changes over the last century and a half. This is most likely
due to the limnology of Lake Bujuku and highly adaptive nature of its dominating
diatom species (Fragilaria sp.). Changes in diatom species after ca. 1975
suggest, however, that there has been a perturbation lasting over 25 years
superimposed upon the general trend of increasing lake productivity. These
trends in diatom flora later begin to reverse in the last few years of the record.
Although reasons for such an ecological change are unclear at the moment, the
responses indicate a shift to a decline in epiphytic habitats and a concomitant
increase in algal productivity. Overall, the palaeoenvironmental record shows
only subtle and gradual changes in response to observed glacial retreat.
- 62
6 Lake-sediment archives of atmospherically deposited
pollutants in the Rwenzori Mountains
6.1 Introduction
6.1.1 Atmospheric pollution
Atmospherically transported pollutants have reached all parts of the globe and
nowhere can now be classed as truly pristine. Both the Arctic and Antarctic are
known to have been impacted by trace metals (e.g., Boutron 1982; Hermanson
1993; Wolff and Suttie 1994; Wolff et al. 1999), persistent organic pollutants
(POPs) (e.g., Macdonald et al. 2000; Borghini et al. 2005) and fossil-fuel derived
particulates (e.g., Murphey and Hogan 1992; Rose et al. 2004) and the record
stored in lake sediments and ice cores demonstrates that fluxes have increased
over recent decades. More volatile contaminants (e.g. some POPs; Hg) are
preferentially deposited in colder regions and this effects polar zones by
movement latitudinally (Wania and Mackay 1996) as well as areas that are colder
by virtue of their altitude (Grimalt et al. 2001). As a consequence, these
contaminants have been shown to have bioaccumulated to significant levels in
the sediments (e.g., Fernández et al. 1999; 2000) and ecosystems (e.g., Vives et
al. 2004a, b) of remote, high altitude European and Arctic lakes.
Mountain lakes are therefore particularly relevant to assessments of atmospheric
deposition because they are not directly (i.e. within the catchment) affected by
human activities and all contamination enters via the atmosphere. In this respect
they can often act as ‘early warning indicators’ for lower altitude environments or
for sites directly affected by human activities. Despite growing interest in
atmospheric contaminants in mountain lakes, very little work has been
undertaken in the tropical mountain lakes of central Africa. This study employs a
palaeolimnological approach to investigate atmospherically deposited trace
metals and fly-ash in three remote mountain lakes in the Rwenzori Mountains of
Uganda. The aims of this study were:
• to determine the temporal trends of pollutant deposition of three mountain
lakes and identify direction and rates of change; and
• to assess possible sources of any contamination.
6.1.2 Trace metals
Trace-metal background concentrations in sediments are mainly driven by the
weathering of bedrock. Worldwide emission inventories have been developed by
Pacyna (1986) assessing both natural and anthropogenic sources. It is worth
noting that soil-derived dust accounts for over 50% of the total Cu, Mn and V
- 63
emissions, as well as for 20 to 30% of the Cu, Mo, Ni, Pb, Sb and Zn released
annually to the atmosphere. Biogeochemical processes within lakes and their
catchments may also affect trace metal background concentrations in sediments.
Soils consist of heterogeneous mixtures of different organic and inorganic-
mineral substances, crystalline and clay minerals, oxides and hydroxides of Fe,
Mn and Al, and other solid components as well as a variety of soluble
substances. Increases in the transport of catchment-derived material to a lake
can lead to enhanced levels of trace metals. Hence, it is often difficult to
differentiate between naturally-derived and anthropogenic trace metals. It is
consequently important to consider methods of identifying possible natural and
human sources when analysing lake sediment records.
Anthropogenic sources of trace metals to the atmosphere include a wide range of
industrial activities including, electricity generation, metallurgical industries,
mining and smelting, chemical and electronic industries, waste disposal,
transport and agricultural practices. Once emitted to the atmosphere, these
contaminants can travel very long distances prior to deposition. In the case of
Hg, the volatile nature of the element means that it has an atmospheric life-time
of ca. 1 year and sources of contamination to even the remotest of regions must
be considered on an hemispherical or global scale.
6.1.3 Fly ash
Fly-ash particles are produced by the combustion of fossil-fuels at industrial
temperatures and at a high rate of heating (Rose, 2001). This process produces
spheroidal carbonaceous particles (SCPs), which are mainly composed of
elemental carbon, and inorganic ash spheres (IASs) which are formed from the
mineral component of the original fuel (Fig. 31). SCPs were used in this study
because they have no natural sources and are not produced from the burning of
wood, biomass or charcoal. SCPs are, therefore, an unambiguous indicator of
deposition from industrial combustion of fossil fuels. Due to their composition,
SCPs are chemically robust and can easily be extracted from lake sediments
(Rose, 1994).
Figure 31. Spheroidal carbonaceous particles (SCP) under an electron microscope (Rose, 2004).
- 64
Most studies on SCPs have been focused within Europe where the temporal
record stored in lake sediments faithfully reflects historical patterns in fossil-fuel
combustion. Other studies have been undertaken in lakes from China, Siberia,
the USA, and in circum-polar Arctic regions. However, no studies of this nature
have been undertaken in central Africa.
6.2 Study areas
Sediment cores were taken from three alpine lakes in the Rwenzori Mountains:
Lower Kitandara Lake, Lake Bujuku and Lake Mahoma (Fig. 32). Kitandara Lake
is situated at 0º21’N, 29º53’E and was formed during the most recent
Omurubaho Glaciation (Osmaston et al., 1998), 6890±100 years B.P
(Livingstone, 1966). The altitude of the lake is about 3990 mamsl and most of its
catchment area is covered by alpine vegetation. The lake is shallow, with a
maximum depth of 9m. Lake Bujuku is situated at 0º22’N, 29º54’E at about 3920
mamsl. The lake formed following the Omurubaho Glaciation about 2960±60
years B.P. Damming of the lake appears to have occurred as a result of
landslides from the slopes of Mount Baker (Fig. 14). The lake has a maximum
depth of 13 m. Lake Mahoma, located at an altitude of 2690 mamsl has the
lowest elevation of the three. It is situated at 0º21’N, 29º58’E, and forms a
headwater catchment of the River Mahoma, a tributary of the River Mubuku. The
lake formed during the waning phase of the Lake Mahoma Glaciation and is
characterised by large converging lateral moraines extending up to three miles
long and 150 m high. Carbon dating from this lake implies an age of 15 to 20
thousand years. No lower altitude glacial lake appears to exist in equatorial
Africa. Vegetation around the lake is dominated by Arundinaria forest but there
are also many scattered heather trees. The lake has a maximum depth of 25 m.
The water is described as deeply stained, with bubbles of gas rising continually
to its surface (Livingstone, 1962). The lake is stratified and only the upper 10 m
contains more than 1 ppm of oxygen.
Figure 32. Map of the location of Lakes Mahoma, Bujuku and Kitandara in the Rwenzori
Mountains (reproduced from Livingstone, 1966).
- 65
6.3 Methodology
6.3.1 Sediment-core collection and extraction
Sediment cores were taken from a representative part of the deep-water area of
each lake from an inflatable boat using a gravity corer (Glew, 1991). Bathymetric
surveys of each lake were conducted using an echosounder and hand-held GPS.
All cores were extruded on site at 0.25 cm intervals (in the upper 10cm), 0.5 cm
(10 cm to the base of the core). Due to time limitations at the site, core MAHO1
from Lake Mahoma was extruded in 1 cm from 20 – 45cm. All sediment samples
were placed in Whirlpak bags to avoid contamination during transfer and to
prevent loss of water prior to lithostratigraphic analysis.
6.3.2 Lithostratigraphy
Lithostratigraphic measurements were undertaken using standard techniques.
Dry weight (DW) was measured by drying sub-samples in a furnace at 105ºC for
at least 12 hours. The same sediment was then placed in the furnace at 550ºC
for 2 hours to give the percentage loss of ignition (LOI) as an estimate for organic
matter content. Sediment wet density (WD) was measured using 2 cm3 capacity
brass phials.
6.3.3 Sediment-core dating
Sediment cores KITA3 and MAHO3 were dated by Dr. Handong Yang, University
College London by analysing for 210Pb, 226Ra, 137Cs and 241Am by direct gamma
assay in the Bloomsbury Environment Institute Facility using an ORTEC HPGe
GWL series well-type coaxial low background intrinsic germanium detector. 210Pb
was determined via its gamma emissions at 46.5keV, and 226Ra by the 295 keV
and 352 keV gamma rays emitted by its daughter isotope 214Pb following three
weeks storage in sealed containers to allow radioactive equilibration. Cesium-
137 and 241Am were measured by their emissions at 662 keV and 59.5 keV
respectively. The absolute efficiencies of the detector were determined using
calibrated sources and sediment samples of known activity. Corrections were
made for the effect of self absorption of low energy gamma rays within the
sample. BUJU3 was dated by Dr. Andy Cundy, University of Sussex. Core sub-
samples were counted for at least 8 hours on a Canberra well-type ultra-low
background HPGe gamma ray spectrometer to determine the activities of 137Cs,
210Pb and other gamma emitters.
6.3.4 X-ray fluorescence (XRF)
X-ray fluorescence (XRF) analyses were carried out in Department of
Geography, University of Liverpool. The Metorex Xmet920 X-ray fluorescence
spectrometer uses 55Fe and 109Cd isotope X-ray sources and is run by the
software DECONV. Reference Buffalo River sediment (NIST2704), stream
sediment (GBW7309) and pond sediment (NIE2) were used to evaluate the
- 66
accuracy of the measurements. The technique produces reasonable accuracy for
Si, K, Ca, Ti, Fe, Rb and Zr and was chosen because it is able to analyse over
70 samples per day. It is also a rapid and non-destructive method. It provides
background geochemical data and acts as an exploratory tool prior to more
detailed and time consuming trace element analyses.
6.3.5 Atomic Absorption Spectrometry (AAS)
Samples were prepared for Atomic Absorption Spectrometry (AAS) using a
standard nitric acid digestion. AAS analysis was carried out in the Department of
Geography, University of Liverpool using a Unicam 939 AAS instrument to
measure Cd, Co, Cu, Ni Pb and Zn. Hg analysis was undertaken by cold vapour
atomic absorption spectrometry (CV-AAS) by Dr. Handong Yang using the same
instrument.
6.3.6 SCP analysis
Sediment samples were prepared for SCP analysis using the standard approach
described by Rose (1994). This involves a sequential mineral acid digestion to
remove unwanted fractions of the sediment resulting in a suspension of mainly
carbonaceous material in water. A known sub-sample of this suspension is then
evaporated onto a cover-slip and mounted on a microscope slide. The number of
SCPs are then counted using a light microscope at x400 magnification. The
concentration of SCPs can then be expressed in units of ‘number of SCPs per
gram dry mass of sediment (g⋅DM-1).
6.3.7 Interpretative techniques
Unlike SCPs, trace metals in sediment cores have both natural and
anthropogenic sources. It is therefore necessary to apply techniques to metals
data to assess the amount of element contributed by human activities. A number
of methods have been used in various studies (Norton and Kahl, 1991; Kamman
and Engstrom, 2002; Roulet et al., 2002; Pacyna, 1998; Bruland et al., 1974;
Bilali et al., 2002) and most of them are based on enrichment factors or ratios of
the element of interest to a ‘natural element’. Due to differences in geochemistry,
different elements may be needed to act as passive tracers at the sites (Yang
and Rose, 2005). The method devised by Hilton et al. (1985) is the most
appropriate. Using this method, the trace-metal concentration is regressed
against a major cation. Assuming that the trace element is associated with
natural contributions, the scatter plot should be linear. However, if there is an
additional enhancement due to anthropogenic pollution a change in slope gives
an indication of the change in source.
- 67
6.4 Results
6.4.1 Percentage loss on ignition, dry weight and wet density
Figure 33 shows the percentage loss-on-ignition (LOI), dry weight and wet
density in the analysed cores. The percentage LOI gives a crude measure of the
organic content of the sediment. The percentage organic matter for KITA2 and
BUJU1 shows a general decreasing trend as depth increases, with the former
having a slightly higher organic content. MAHO1 is the most organic-rich
sediment core and its percentage LOI fluctuates around 80 % throughout the
profile. Its lowest organic content of 71% at 41 cm is higher then the highest
organic content for the other two lakes. In general, percentage LOI values often
show an inverse relationship with percentage dry weight. This phenomenon is
more obvious for BUJU1 but less so for KITA2 and MAHO1.
KITA2 % LOI
0
5
10
15
20
25
30
35
40
45
0 10 20 30 40 50 60
BU JU 1 %LOI
0
5
10
15
20
25
30
0 10 20 30 40
MAHO1 %LOI
0
5
10
15
20
25
30
35
40
45
50
0 20 40 60 80 100
KITA2 % dry w eight and w et de nsity
0
5
10
15
20
25
30
35
40
45
0 10 20 30
Depth (cm)
%DW
%WD
BUJU1 %Dry Weight and %Wet Dens ity
0
5
10
15
20
25
30
0 10 20 30 40 50
Dep th ( cm)
%DW
%WD
MAHO1 %Dry Weight and Wet Density
0
5
10
15
20
25
30
35
40
45
50
0 5 10 15
D e p th (c m )
%DW
%WD
Figure 33. Loss on ignition (LOI), dry weight and wet density profiles for each analysed sediment
core (BUJU3, KITA3, MAHO1).
- 68
6.4.2 Age - depth profiles of sediment cores
Details of the dating of BUJU3 core are presented in Chapter 5 (Section 5.4). A
constant rate of supply model for 210Pb (Appleby 2001) is applied to KITA3
whereas a constant initial concentration (CIC) model (Appleby 2001) is applied to
MAHO1. Dating results are unusual for KITA3, as the estimated 1963 level by
137Cs is higher than the level dated by 210Pb. This can be caused by the loss of
surface sediments during sampling or by a small degree of sediment mixing.
However, the similarity of the LOI profiles of KITA3 and the other KITA cores
suggests that the possibility of loss of the top sediment is low. It is also unusual
that the dry mass density in the uppermost sediments is the highest. 137Cs dates
suggest the post-1963 sedimentation rate to be 0.011 g⋅cm-2⋅a-1 which agrees
reasonably well with the rate of 0.0114 ± 0.0025 g⋅cm-2⋅a-1 for the entire core as
calculated using CRS model for 210Pb.
Figure 34 shows that the age-depth profiles for the Bujuku cores are relatively
linear suggesting little change in sediment accumulation rate through time. The
KITA and MAHO profiles suggest an increase in sediment accumulation rate in
more recent times. The earliest dates obtained for KITA3, BUJU3 and MAHO3
are 1849, 1865 and 1836 respectively. BUJU1 is a shorter core and can only be
dated back as far as 1927.
6.4.3 Sediment-core trace-element profiles
Within a sediment core, the distribution of heavy metals such as Cd, Co, Cu, Pb,
Zn and Ni often exhibits correlations with the amount of organic matter or
coincidental elements such as Ti, Al, Fe and Mn. With the addition of
anthropogenic derived heavy metals deposited in lake systems, such correlations
may breakdown. The comparison can, therefore, help to differentiate natural from
anthropogenic contributions and geochemical processes.
6.4.3.1 Lower Kitandara Lake
Cadmium is the only trace metal that shows an increase towards the surface of
the core (Fig. 35). With the exception of the uppermost sample, Ni concentration
decreases towards the surface whereas Zn, Cu, Co and Pb show no obvious
trends. Cu and Zn display similar trends, with corresponding peaks and troughs.
A distinctive peak appears at 10cm on the Pb profile. Apart from Cd, the trace
metals shows opposite trends to that of LOI.
- 69
KITA2 and KITA3 Depth vs age
0
2
4
6
8
10
12
14
16
18
20
1845
1865
1885
1905
1925
1945
1965
1985
2005
Depth (cm)
KITA2
KITA3
BUJU1 and BUJU3 Depth vs Age
0
5
10
15
20
25
30
35
40
45
1845
1865
1885
1905
1925
1945
1965
1985
2005
Depth (cm)
BUJU1
BUJU3
MAHO1 and MAHO3 Depth vs Age
0
5
10
15
20
25
30
1825
1845
1865
1885
1905
1925
1945
1965
1985
2005
Depth (cm)
MAHO1
MAHO3
Figure 34. The age-depth profile of sediments cores.
.
- 70
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.005 0.01 0.015 0.02
Cd Concentration (mg/g)
Depth (cm)
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.02 0.04 0.06
Co concentration (mg/g)
Depth (cm)
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.2 0.4 0.6
Cu concentration (mg/g)
Depth (cm)
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.01 0.02 0.03
Ni concentration (mg/g)
Depth (cm)
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.2 0.4 0.6
Zn concentration (mg/g)
Depth (cm)
Kitandara
0
5
10
15
20
25
30
35
40
45
0 0.05 0.1 0.15
Pb concentration (mg/g)
Depth (cm)
Figure 35. Depth (age) profiles of trace metal concentrations at Lower Kitandara Lake (KITA2).
6.4.3.2 Lake Bujuku
Like KITA2, the BUJU1 Cd profile differs from other trace-metal profiles (Fig. 36).
Copper, like LOI, shows a slight increase towards the top of the core, whereas
Co, Ni, Zn and Pb concentrations fluctuate. Zn and Pb concentrations show
some similarity in their trends. An exceptional decline in Pb concentrations
occurs at 18 cm and does not correspond with any trace metals. The trace metal
profiles do not appear to correspond to the trends in geochemical data.
- 71
Bujuku
0
5
10
15
20
25
30
0 0.001 0.002 0.003 0.004
Cd concentration (mg/g)
Depth (cm)
Bujuku
0
5
10
15
20
25
30
0 0.005 0.01 0.015 0.02
Co concentration (mg/g)
Depth (cm)
Bujuku
0
5
10
15
20
25
30
0 0.05 0.1
Cu concentration (mg/g)
Depth (cm)
Bujuku
0
5
10
15
20
25
30
0 0.01 0.02 0.03 0.04
Ni concentration (mg/g)
Depth (cm)
Bujuku
0
5
10
15
20
25
30
0 0.1 0.2 0.3 0.4
Zn concentration (mg/g)
Depth (cm)
Bujuku
0
5
10
15
20
25
30
0 0.05 0.1 0.15
Pb concentration (mg/g)
Depth (cm)
Figure 36. Depth (age) profiles of trace metal concentrations at Lake Bujuku (BUJU1).
6.4.3.3 Lake Mahoma
Trace-metal concentrations in Figure 37 show unexpected patterns. Extremely
high concentrations of Cu, Ni and Pb are detected in the shallowest (most recent)
sample. The Ni concentration in this uppermost sample is ten times higher than
the rest of the core (137 g⋅g-1) and has been removed from Figure 37 in order to
represent temporal trends clearly. For Ni, Zn and Pb, high concentrations are
observed but not at similar depths (ages). They do not appear to be comparable
with peaks in the geochemical data. There is an increase in the concentration of
Co, Ni, Pb and most obviously Cu, towards the most recent (shallow) sediments
from a depth of approximately 16 cm.
- 72
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.002 0.004 0.006
Cd concentration (mg/g)
Depth (cm)
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.05 0.1
Cu concentration (mg/g)
Depth (cm)
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.001 0.002 0.003 0.004
Co concentration (mg/g)
Depth (cm)
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.005 0.01 0.015
Ni concentration (mg/g)
Depth (cm)
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.2 0.4 0.6 0.8
Zn concentration (mg/g)
Depth (cm)
Mahoma
0
5
10
15
20
25
30
35
40
45
50
0 0.02 0.04 0.06
Pb concentration (mg/g)
Depth (cm)
Figure 37. Depth (age) profiles of trace metal concentrations at Lake Mahoma (MAHO1).
6.4.4 SCPs
The concentrations of SCPs in the sediment cores were extremely low. Black,
‘disc-like’, particles were, however, detected throughout BUJU1. Unlike usual
SCPs, these particles appear to be flat, have similar sizes and concentrations at
each depth. This kind of particle has been found previously in sediments from
Lakes Wandakara and Kasenda in Uganda but at present, it has not been
possible to identify their nature and origin.
- 73
6.5 Discussion and analysis
6.5.1 Application of Hilton’s model
Hilton’s model (see 6.3.7) was applied to the trace-metal data in order to identify
anthropogenic sources of metals in the three sediment cores. Figure 38 shows
the relationship between trace metals with other elements as well as percentage
LOI.
6.5.1.1 Lower Kitandara Lake
The model shows that for KITA2, Co, Cu, Ni and Zn all have high negative
correlations with LOI. These elements do not appear to be controlled by the
amount of organic matter in the sediment. Cu and Ni both show high correlation
with K, Fe and Mn. This suggests a high affinity of Cu and Ni for mineral material
containing these three elements. In many soils, the oxides and hydroxides of Fe
and Mn play the most significant role in the distribution and behaviour of trace
elements. The hydroxides occur in soils as a coating on soil particles, as filling in
cracks and veins, and as concretions of nodules. They also have high sorption
capacity for trace metals (Kabata-Pendias and Sadurski, 2004). Fe and Mn
oxides co-precipitate and adsorb cations including Co, Cu, Mn, Mo, Ni, V and Zn
(Alloway, 1995). In this core, Co is likely to be associated with minerals rich in Mn
and Fe as it shows a high correlation with these elements.
Zn and Pb have a weak correlation with all geochemical elements as well as LOI.
However, the Zn profile shares many similarities with those of Co, Cu and Ni
suggesting that Zn derives from the same source. The Pb profile is similar but the
low correlation likely stems from anomalously high concentrations observed at
depths of between 7 and 12 cm. The origin of high Pb concentrations remains
uncertain but is unlikely to be a result of contamination due to the stable rising
and falling limb. Muwanga (1997) suggested that Pb dissolves at pH values
below 4.5 and above 6. If it was a result of diagenesis or change in pH causing
the mobility of the Pb, such changes should be seen in other elements. Cd is the
only metal that shows an increase towards the surface. Although the model
shows a low correlation with LOI, a Cd/LOI ratio gives a very consistent value,
suggesting that the occurrence of this trace element is closely related to the
presence of organic matter, and therefore the increase is likely to be due to a
natural geological contribution.
6.5.1.2 Lake Bujuku
The application of Hilton’s model to BUJU1 shows that Cu shows a high positive
correlation with percentage LOI, suggesting the increase is due to the
contribution of organic matter. Cu in solid phase is often adsorbed by soil
components in the following order: organics > Fe/Mn oxides > clays (Baker,
- 74
r = 0.7603 KITA2
0
10
20
30
40
50
60
70
0 100 200 30 0 400
Co ( g / g )
r = 0.7226 KITA2
0
1
2
3
4
5
6
0 20 40 60 80
Co ( g / g )
r = 0.5109 KITA2
60
70
80
90
100
110
120
130
140
150
0 20 40 60 80
Co ( g / g )
r = 0.5938 KITA2
0
5
10
15
20
25
30
35
40
0 100 20 0 300 400
Cu ( g / g )
r = 0.4780 KITA2
0
10
20
30
40
50
60
70
0 100 200 300 400
Cu ( g / g )
r = 0.5422 KITA2
0
1
2
3
4
5
6
0 100 200 300 4 00
Cu ( g / g )
r = 0.8284 KITA2
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100
Ni ( g / g)
r = 0.8836 KITA2
0
10
20
30
40
50
60
70
0 20 40 60 80 100
Ni ( g / g)
r = 0.8783 KITA2
0
1
2
3
4
5
6
0 20 40 60 80 100
Ni ( g / g )
r = 0.8464 BUJU1
0
5
10
15
20
25
30
0 20 40 60 80
Cu ( g / g )
r = 0.8592 BUJU2
0
5
10
15
20
25
30
0 0.1 0.2 0.3 0.4 0.5
Cd ( g / g )
Figure 38. Element concentration correlations and Pearson’s correlation coefficient (r).
- 75
1995). It has been documented that up to 98% of total Cu is associated with the
organic fraction in soils (Flemming and Trevors, 1989). The lack of a correlation
with mineral parameters in this case can be attributed to the greater natural
contribution of organic matter to their adsorption by sediments. Cd also produces
a high overall correlation with the percentage of LOI. At greater sediment depths
(13 to 26 cm), a negative correlation occurs. This is due to the extremely low
concentrations at that depth. No significant correlation between other trace
metals and geochemical data or LOI was obtained.
6.5.1.3 Lake Mahoma
Hilton’s model does not appear to apply to MAHO1 as no correlation can be
identified between trace metals and other geochemical data or LOI. It is possible
that the high organic content (mean of 79%) in the sediment means that a very
low percentage of mineral is present to provide sufficient adsorption sites for the
metals. At the same time, the organic fraction is high so adsorption sites are
unlikely to reach saturation. The increase in Co, Cu, Ni and Pb is attributed to
anthropogenic sources. Comparisons are hindered by the fact that dissolution
and sorption may affect some geochemical elements but not all (Alloway, 1995).
6.5.2 Assessment of anthropogenic origin of trace metals
The assessment of background and anthropogenic impact of trace metals in
sediment cores is essential but often difficult. The ‘background’ usually refers to
the natural contribution. In this study, it is more appropriate to describe this as a
‘baseline’. Metal contamination in sediments can be traced back as far as the
Bronze Age (2500BC) (Kabata-Pendias and Sadurski, 2004). The 150-year
period covered by sediment core archives includes considerable atmospheric
pollution. Baseline concentrations in the sediment cores are unlikely to be solely
natural.
Application of the Hilton model to trace-metal profiles suggests that Lake
Kitandara and Lake Bujuku, both located at higher latitudes, are relatively
unpolluted. No obvious increase in metals due to anthropogenic influence can be
identified. In contrast, the Lake Mahoma profiles (Fig. 37) reveal increased
concentrations of Co, Cu, Ni and Pb in the most recent (shallow) sediments. Co
and Pb show an increase from around 13 cm depth, which corresponds to a date
between 1957 and 1962. Increased concentrations of Cu and Ni begin earlier in
the 1940s (16 cm). For Co, Ni and Pb, the surface sediment is highly enriched.
Calculation shows that the contaminated sediments of Co, Cu, Ni and Pb are 3.5,
1.4, 2.3 and 3.3 times higher than the baseline, respectively. The Pb and Co
contamination levels remain relatively constant prior to the surface peak,
whereas for Cu, the highest concentration level reaches 32 g⋅g-1 at 5 cm and
corresponds to a period between 1985 and 1989. A massive drop in Ni
concentration is recorded at 6 and 8 cm, representing periods in the 1970s and
1980s.
- 76
6.5.3 Mercury profiles
Mercury vapour (Hg0) has a long atmospheric residence time and mercury
contamination of lacustrine environments is ubiquitous (Engstrom and Swain,
1997). Hg analysis for the three Rwenzori lake sediment cores shows increasing
concentrations for all profiles starting at different depths (Fig. 39a). In KITA2, Hg
values fluctuate between 70 to 90 ng⋅g-1 from the deep sediment to 18cm, where
the concentration increases and eventually exceeds 200 ng⋅g-1 in the top few
centimetres. Three major stages can be seen in the BUJU1 profile, (i) 18-22 cm,
(ii) 8 to 17 cm, and (iii) 0 to 7 cm. Each of these stages shows a progressively
higher Hg value than the previous one. MAHO1 shows a gradual increase from
the base of the core (Fig 39a) until a very significant increase is observed at 22
cm depth to above 200 ng⋅g-1 where it remains up to 10 cm when concentrations
decline to the surface.
Hilton’s model suggests a high positive relationship between MAHO1 Hg and
percentage LOI at 19 cm and below (Fig 33). This correlation is significantly
lower above this depth. Using this relationship it is possible to identify the
anthropogenically induced Hg concentration above this depth (Fig 39b). The
Hg:LOI ratio for KITA2 is consistent below 17 cm, indicating Hg is contributed by
organic matter at lower depths. The amount of Hg pollution is estimated by the
linear relationship between background Hg and percentage LOI. The same
method is used for BUJU1, where sediment between 17 and 26 cm provides a
baseline.
Anthropogenically induced Hg contamination seems to have begun in the ca.
1860s in Lake Kitandara. It then reaches a peak in the 1920s, followed by a
decline. The trend increases almost continuously since 1950s. In BUJU1, the
pollution signal starts in ca. 1950s. No pollution above the background value is
detected at 7.5 cm, but this is followed by a rapid increase towards the most
recent sediments. MAHO1 Hg pollution signal started in ca. 1930s. A reduction
occurs around 1971. A modern baseline ratio suggests an enrichment of Hg by
2.1, 2.2 and 1.5 for KITA2, BUJU1 and MAHO1 respectively. This is in
reasonable agreement with the enrichment figure of 3 ± 1 (D. Engstrom, pers.
commun.) for remote lakes in other areas of the world including Arctic Canada
and Europe and New Zealand. The influence of Hg on this part of the Rwenzori
Mountains is more widespread then other trace metals since it is detected in all
three lakes. It is logical to suggest that the Hg originates from long-distance
transport. It is unlikely to be local as contamination from other metals has not
been detected in Lake Kitandara and Bujuku.
- 77
a)
0
5
10
15
20
25
30
35
40
45
0 50 100 150 200 250
KITA Hg (ng/g)
BUJU1 Hg (ng/g )
0
5
10
15
20
25
30
0 50 100 150 200
MAHO Hg (ng/ g)
0
5
10
15
20
25
30
35
40
45
50
0 50 100 150 200 250 300
b)
KITA2 Ant hropog enic induced Hg (ng/g)
0
2
4
6
8
10
12
14
16
18
20
0 50 100 150
BUJU1 Anthropo genic induced Hg (ng/g )
0
2
4
6
8
10
12
14
16
18
0 20 40 60 80
MA HO1 A nthropog enic induced Hg (ng/g)
0
5
10
15
20
25
0 2 0 40 60 80 100 120 140
c)
KITA2
y = 1.4158x + 30.868
0
20
40
60
80
100
120
0 10 20 30 40 50
% LOI
BUJU1
y = 4.5245x
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
%LOI
MA HO1
y = 7.7471x - 490.55
0
50
100
150
200
250
70 75 80 85 90
%LOI
Figure 39. (a) Hg profiles; (b) anthropogenic Hg profiles; and (c) Correlation of background Hg
and %LOI.
- 78
6.5.4 Trace metal flux profiles
Metal concentration profiles are strongly influenced by sedimentation rates.
Concentrations of elemental constituents are accentuated under periods of
reduced sedimentation, and vice versa (Kamman and Engstrom, 2002).
Conversion of concentrations to flux rates normalises this covariance, and
permits a better comparison between lakes. Figures 40 to 42 show the trace
metal flux profiles.
Trace-metal fluxes for Lower Kitandara Lake (Fig. 40) show increasing deposition
of Cu, Zn and Pb, and to a greater extent Cd and Hg. The change in Cu, Zn and
Pb can be attributed to an increase in sedimentation rate. However, the
enrichment factor of Cd and Hg in the upper layers is significantly higher than the
sedimentation rate. Cd influx increases from 14 cm upwards. The influx Hg
above this depth is 3.7 times higher than the baseline while the difference in
sedimentation rate is less then 2. However, as seen earlier, this increase is very
likely to be a result of an increase in organic matter. This agrees with the Hg:LOI
ratio data and suggests a breakdown in the influence on Hg by LOI above 17 cm
and supports the anthropogenic origin of Hg above that depth.
For Lake Bujuku, the profiles reveal a decrease in flux towards the surface. This
is due to the decrease in sedimentation rate towards the uppermost sediments
and can again be explained by the amount of organic matter present. The flux of
trace metals in Lake Mahoma increases particularly in the shallowest (most
recent) sediments from 7 cm. To a certain extent, this can also be attributed to a
rapid increase in sedimentation rate. However, for Co, Cu, Ni, Pb and Hg, the
proportional increase is too big to be solely attributable to the change in
sedimentation rate. The upper sediment enrichments are 7 times higher than the
bottom sediment for Co and Hg, and 3, 9 and 8 times higher for Cu, Ni and Pb
respectively. This supports the hypothesis that the metals are probably
anthropogenically enhanced in these upper layers.
6.5.5 Trace-metals concentrations in moss
The use of certain terrestrial moss species to monitor the deposition of metals is
now well established in Europe and elsewhere. Briefly, these mosses obtain
everything they require from atmospheric deposition and so do not uptake any
nutrients or minerals through a root system which is used only to secure the plant
to its substrate. As a consequence, the metal content of the moss tissue reflects
the metal deposited onto the moss and its analysis provides the opportunity to
study the transfer of trace metals from the atmosphere virtually independent of
the underlying mineral material or organic matter.
- 79
KITA2 Cd ( g /cm2/yr)
0
5
10
15
20
25
30
35
40
45
0 0.005 0.01 0. 015 0.02 0.025 0.03
KITA2 Co ( g/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
0 0.1 0.2 0.3 0.4 0. 5 0. 6
KITA2 Cu ( g/cm2/yr )
0
5
10
15
20
25
30
35
40
45
0 0.5 1 1.5 2 2 .5 3 3.5 4
KITA2 Ni ( g/cm2/ yr)
0
5
10
15
20
25
30
35
40
45
0 0.2 0. 4 0.6 0. 8 1
KITA2 Zn ( g/cm2/yr)
0
5
10
15
20
25
30
35
40
45
0 1 2 3 4 5
KITA2 Pb ( g/cm2/ yr)
0
5
10
15
20
25
30
35
40
45
0 0.05 0.1 0.15 0.2 0.2 5 0.3 0. 35
KITA2 Hg (ng/cm2/ yr)
0
5
10
15
20
25
30
35
40
45
0 0.5 1 1.5 2 2.5 3 3.5
Figure 40. Depth (age) profiles of trace-metal fluxes for Lower Lake Kitandara.
- 80
BUJU1 Cd ( g/cm2/yr )
0
5
10
15
20
25
30
0 0.01 0.02 0.03 0.04
BUJU1 Co ( g/cm2/ yr)
0
5
10
15
20
25
30
0 1 2 3 4 5
BUJU1 Cu ( g /cm2/yr)
0
5
10
15
20
25
30
0 1 2 3 4 5 6
BUJU1 Ni ( g/cm2/yr)
0
5
10
15
20
25
30
0 1 2 3 4 5
BUJU1 Zn ( g /cm2/yr)
0
5
10
15
20
25
30
0 5 10 15 20 25
BUJU1 Pb ( g/cm2/yr)
0
5
10
15
20
25
30
0 10 20 30 40 50 60
BUJU1 Hg (ng/cm2/yr)
0
5
10
15
20
25
30
0 5 10 15
Figure 41. Depth (age) profiles of trace-metal fluxes for Lake Bujuku.
- 81
MA HO1 Cd (ug/cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
M AHO Co (ug/cm2/yr )
0
5
10
15
20
25
30
35
40
45
50
0 0.02 0.04 0.06 0.08 0.1
MA HO1 Cu (ug/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.1 0.2 0.3 0.4 0.5
MA HO1 Ni (ug/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.05 0.1 0.15 0.2 0.25
MA HO1 Zn (ug/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.5 1 1.5 2 2.5 3
MA HO1 Pb (ug/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.05 0.1 0 .15 0.2 0.25 0.3
MA HO1 Hg flux (ng/ cm2/yr)
0
5
10
15
20
25
30
35
40
45
50
0 0.5 1 1.5 2 2.5 3
Figure 42. Depth (age) profiles of trace-metal fluxes for Lake Mahoma.
- 82
Two moss species are typically used for this in Europe, Hylocomium splendens
and Pleurozium schreberi, based on their widespread distribution. Fortuitously,
Hylocomium splendens was found near Lake Bujuku and opportunistically
sampled. Species identification was confirmed by experts at the Natural History
Museum, London. The Bujuku moss sample analysis resulted in metal
concentrations shown in Table 4. Hg and Cu are relatively low, similar to values
found in relatively unpolluted areas of Great Britain. Hg concentration is
comparable to that of Lochnagar, a remote mountain lake in Scotland. Pb, Zn
and Cd are in the middle of the range for British sites. This further supports that
the area around Lake Bujuku is relatively free from atmospheric pollution.
Table 4. Elemental concentrations in moss sampled near Lake Bujuku.
Hg ng⋅g-1 Cu g⋅g-1 Pb g⋅g-1 Zn g⋅g-1 Cd g⋅g-1
30 2.4 4.3 37 0.46
6.5.6 Possible sources of trace metals
Extremely low concentrations of SCPs detected in the sediment of the Rwenzori
lakes may occur for two reasons. First, the source that produced trace metals
detected in Lake Mahoma does not involve the combustion of fossil fuels.
Electricity in Uganda is generated almost entirely by hydroelectric means and
most of this is generated from a single hydroelectric station at Owen Falls.
Another possibility is that the source of metal contamination is at a distance
sufficient to preclude the transport of SCPs to the sites. This suggestion is
consistent with the conclusion that increased Hg in the most recent (shallow)
sediments arise from global sources.
Copper was discovered in the Rwenzori Mountains in 1906 and since then an
extensive exploration programme regarding the potential for mining has been
undertaken (Muwanga, 1997). Therefore, a possible source of Co, Cu, Pb and Ni
detected in MAHO1 is local mining in the area. The production of copper in
Kilembe, which is in close proximity to the study area (Figure 15) commenced in
1956 when the railway to Kasese was completed. Nearly 16 million tonnes of ore
averaging 1.95% Cu had been mined by 1978 when production ceased. After
processing, the material was sent to Jinja about 400 km away in Eastern Uganda
for smelting (Figure 15a). The increase in Co and Pb seems to correspond with
the period when the copper production commenced whereas the increase in Cu
and Ni pre-dates this time and may stem from local, traditional smelting of
copper. Considering the distance from the smelter, it is uncertain whether metal
pollution detected in Lake Mahoma is a result of this smelting.
According to Baker and Senft (1995), major sources of pollution such as
smelters, usually yield the highest concentrations in soils within 1 to 3 km of the
stack with concentrations decreasing exponentially with distance. For the Cu-Ni
complex in Sudbury (Canada), most of the emitted Cu is deposited within 32 km
- 83
but soils in areas closer than 7.5 km frequently contain well over 1000 mg⋅kg-1.
Copper smelters also release large quantities of As and Zn but no increase in Zn
in MAHO1 was detected.
Atmospheric inputs of Cu to soils from both rain and dry deposition vary
considerably according to the proximity of industrial emissions containing Cu and
the type and quantity of wind-blown dust. Hence, the atmosphere is an important
medium for the transmission of pollutant Cu to remote areas of the Earth and
analysis of moss samples and polar ice cores reveal a substantial increase in
airborne Cu at locations far from emission sources (Baker and Senft, 1995).
As the increase in Co, Cu, Pb and Ni continues to the present day and Cu
production ended in 1978, metals in Lake Mahoma may derive from local mine
deposits and tailings. According to Muwanga (1997), wastes produced by the
mining operation were disposed of in Kilembe valley. These mine wastes have
been exposed to strong tropical weathering which leads to the oxidation of
residual sulphides resulting in mobilisation and migration of heavy metals into the
surrounding terrestrial and aquatic environments. Elevated levels of Cu, Co, Ni
and Zn are enriched in soils, water and stream sediments in the mine area and
around the waste disposal sites. Cu toxicity is also detected in vegetation in the
Rwenzori National Park (Edroma, 1974) associated with copper concentrate
leakages along pipe lines. It is possible that the trace metals detected in Lake
Mahoma are associated with physical processes during mining, when the metals
are exposed and transported by aeolian processes into lake catchments. The
fact that the pollution signal is not detected in the other two lakes may then be
explained by their greater distance and higher altitude from the source. If this
were the case, then the exposed tailings may continue to provide a source of
metals to the lakes long after production has ceased and provide a mechanism
for contamination into the future. Further research is required to better
understand the most likely causes.
- 84
6.6 Conclusions
Among the three remote, high mountain lakes in the Rwenzori Mountains under
study only Lake Mahoma, the lowest in altitude, displays signs of anthropogenic,
atmospheric deposition of trace metals. The flux of SCPs to the lake is below
detection limits so it has not been possible to conclude whether emission sources
involve the combustion of fossil fuels. Enrichment of the sediment profiles in
mercury is consistent with (1.5 to 2.2) with that of remote environments around
the world (3 ± 1). The observed increase in atmospherically deposited Co and Pb
at Lake Mahoma began between the mid-1950s and early-1960s and coincides
with the production of copper in the Kilembe mine. However, initial contamination
from anthropogenic Cu and Ni pre-dates this period. The sources of continued
pollution are unclear but likely involve the re-mobilisation of trace metals in mine
tailings in the Kilembe valley and atmospheric transport from remote (non-local)
sources.
- 85
7 Overall project findings
The project has been considerable progress toward realising its scientific
objectives. Each is discussed below.
7.1 magnitude of current glacial recession
Field mapping and analysis of Landsat imagery show that glacial cover on the
three remaining glacierised summits (Mounts Stanley, Speke and Baker) has
decreased from 2.01±0.11 km2 in 1987 to 0.96±0.34 km² in 2003. These
determinations confirm a continued, steady decline in the areal extent of glaciers
in the Rwenzori Mountains of ~0.7 km2 per decade over the last century.
Assuming present trends continue, glaciers in the Rwenzori Mountains are
expected to disappear within the next two decades.
7.2 impact of glacial recession on alpine riverflow
Meltwater flows from glaciers in the Rwenzori Mountains do not contribute
significantly (> 0.5%) to alpine riverflow. This conclusion, based on stream fluxes
measured during the dry season, is consistent with previous assertions by
Temple (1967) and Osmaston (2006). The possibility that receding glaciers
signal a decline in precipitation over the 20th century is supported neither by
meteorological records from lowland stations in western Uganda nor regional
hydrological records. Insufficient data exist to verify local reports in Uganda
(Kasese District) of an increased frequency and magnitude of flood events at the
base of the Rwenzori Mountains. Changes in land use and land pressures
outside of the Rwenzori Mountains National Park (i.e., below 1700 mamsl) in the
River Mubuku basin may explain, in part, the perceived increase in flood risk
rather than significant changes in the hydrology of alpine riverflow.
7.3 recent environmental change from observational datasets
and sediment-core archives
Glacial recession is associated with increased air temperatures not only
suggested by the spatially uniform nature of recent loss of glacial cover but also
indicated by meteorological observations at the surface (station data, gridded
climate datasets) and mid-troposphere (radiosonde measurements) in western
Uganda. There are, however, insufficient data to represent the complex
interactions of radiant energy and heat at the glacier’s surface and thus quantify
the link between changes in climate variables and glacial mass in the Rwenzori
- 86
Mountains. Although increasing air temperature and reduced air humidity remain
plausible and likely related hypotheses to explain recent glacial recession in the
Rwenzori Mountains, there is currently greater evidence of trends of increasing
air temperature than decreasing humidity to explain deglaciation in the Rwenzori
Mountains. This conclusion does not preclude, however, the likelihood that
changes in humidity and radiative fluxes associated with rising air temperatures,
have also contributed to observed glacial recession.
Despite a steady reduction in the areal extent of glaciers, the sediment-core
archive from Lake Bujuku indicates that the lake ecology and flora of Afroalpine
areas of the Central Rwenzori Massif have not undergone dramatic changes over
the last century and a half. This is most due to the highly adaptive nature of its
dominating species (e.g., Fragilaria sp.). Changes in the diatom assemblage
suggest, however, that a perturbation lasting over 25 years is superimposed
upon the general trend of increasing lake productivity. These trends in diatom
flora later begin to reverse in the last few years of the record. Overall, the
responses indicate a recent decline in epiphytic habitats and concomitant
increase in algal productivity.
Atmospherically deposited mercury, consistent with global trends and emissions,
is detected in lake sediment within the Heath-moss Forest zone (3000 mamsl)
and Afroalpine zone (4000 mamsl). Trace-metal contamination via atmospheric
deposition from more localised sources is evident from the mid-1950s and
coincides with onset of copper mining downslope at Kilembe (1700 mamsl).
Some anthropogenic loading of Cu and Ni appears to pre-date operation of the
Kilembe mine and may stem from localised, small-scale traditional copper
smelting practices.
- 87
8 Dissemination & knowledge transfer
The dissemination of research findings and transfer of knowledge gained through
this research project to local stakeholders and wider scientific community have
been achieved through a variety of mechanisms including stakeholder meetings,
scientific journals, popular press, conference presentations, and speaking
invitations. Four student dissertations at the B.Sc. and M.Sc. levels have also
been completed as part of this research.
8.1 stakeholder meetings
Focused meetings were held to discuss the findings of research and solicit input
from local stakeholders including Uganda Wildlife Authority, Rwenzori
Mountaineering Service and Water Resources Management Department.
On 24 January 2005, discussions were held at a meeting room of the Rwenzori
Mountaineering Service in Nyakalengija (Kasese) with the Uganda Wildlife
Authority, RMS and local BaKonzo elders pertaining to (i) current glacial extent
and the rate of glacial recession, (2) the implications of glacial recession on the
safe movement of visitors and their support teams from UWA and RMS, (3) the
factors, both local and regional, driving glacial recession, (4) the impact of glacial
recession and related climate change on the Afroalpine environment and alpine
rvierflow, and the impact of glacial recession on the traditional cultural beliefs of
the BaKonzo.
On 17 July 2006, discussions were held with the Water Resources Management
Department in Entebbe pertaining to (i) current glacial extent and the rate of
glacial recession, (2) the factors, both local and regional, driving glacial
recession, and (3) the impact of glacial recession and related climate change on
alpine rvierflow.
8.2 scientific publications
Panizzo, V.N., Mackay, A.W., Ssemmanda, I., Taylor, R.G., Rose, N. and Leng,
M., in review. Recent changes in aquatic productivity in a remote, tropical
alpine lake in the Rwenzori Mountain National Park, Uganda, associated with
glacier recession since the 1860s. Journal of Paleolimnology
Taylor, R.G., Mileham, L., Tindimugaya, C., Majugu, A., Nakileza, R., Muwanga,
A., 2006. Reply to Comment by Mölg et al. on Recent deglaciation in the
Rwenzori Mountains of East Africa due to rising air temperatures. Geophysical
Research Letters Vol. 33, L20405, doi:10.1029/2006GL027606
Taylor, R.G., Mileham, L., Tindimugaya, C., Majugu, A., Nakileza, R., Muwanga,
A., 2006. Recent glacial recession in the Rwenzori Mountains of East Africa
- 88
due to rising air temperature. Geophysical Research Letters, 33, L10402,
doi:10.1029/2006GL025962.
8.3 conference presentations and speaking invitations
Taylor, R.G., Muwanga, A., Fischer, A. and Tindimugaya, C., 2004. Glacial
recession in the Rwenzori Mountains of Uganda: assessing its implications for
regional climate and consequences for local livelihoods. EGU Congress
(Vienna, Austria)
Taylor, R.G., 2004. Glacier recession on the Rwenzori Mountains: recent
hydrological observations and future research. Department of Biology,
University of Victoria, Canada.
Taylor, R.G., Mileham, L., Muwanga, A., Fischer, A. and Tindimugaya, C., 2004.
Glacial recession in the Rwenzori Mountains of Uganda: assessing its
implications for regional climate and consequences for local livelihoods. In:
Trees, Rain and Politics in Africa: dynamics and politics of climatic and
environmental change, unpublished conference proceedings (Oxford, UK),
September 29th - October 1st, 2004.
8.4 popular press
"Glaciers in Africa expected to disappear" appeared in: BBC News, Nature,
Science, Die Spiegel, Pravda, Washington Post, Malaysia Sun, New Kerala,
USA Today, FOX News, Forbes, SciDevNet, ABC News, CBS News, MSNBC
News, LA Times, Chinese News Service, Chinese Broadcast Service, The
Calgary Sun, IOL South Africa, Journal of the Turkish Weekly
"African ice caps will soon disappear due to global warming" also appeared in
European Commission DG Environment News Service, Science for Policy,
Issue 25 (June 8, 2005)
Taylor, R., 2006. Nights with Bryan Crump on National Radio, Radio New
Zealand. Snows of Uganda. June 26th, 2006.
Taylor, R., 2006. Quirks and Quarks, Canadian Broadcasting Corporation.
Tropical glaciers. May 20, 2006.
Taylor, R. 2006. BBC Science in Action. Snows of Uganda. May 19, 2006
Johnston, I., 2006. Climate threat to glaciers on Mountains of the Moon. The
Scotsman, p. 26, May 17th, 2006
Rose, J., 2006. Bara hälften kvar av Afrikas glaciärer. Forskning & Framstag
(Swedish popular magainse), p. 11, September 2006.
Virto, E.C., 2006. Afrikako glaziarrak: iragharritako galera? Elhuyar Zietzia eta
Teknia (Basque popular magasine) Vol. 33, p. 9
Taylor, R.G., 2005. The son of the snow is angry - loss of glaciers threatens
indigenous culture. The New Internationalist Vol. 378, p. 6.
Amodeo, C., 2003. African glaciers in retreat. Geographical Magasine, pp. 10-11,
December 2003.
- 89
Mugisha, M., 2003. Adventure up the Rwenzori. The New Vision Sunday
Magazine (Uganda's National Newspaper), pp. 6-13, October 5th, 2003.
Sample, I., 2003. "Return to the Mountains of the Moon" The Guardian (Life
Section), pp. 8-9, October 2nd, 2003.
Taylor, R.G., 2003. Rwenzori glaciers disappearing. The New Vision (Uganda's
National Newspaper), October 28th, 2003
Taylor, R.G. 2003. Will Rwenzori's snow vanish? The New Vision (Uganda's
National Newspaper), June 16th, 2003
8.5 student dissertations
Jenny Hau, 2005. Sediment record of atmospherically deposited pollutants in
Rwenzori Mountain lakes, Uganda. Unpublished M.Sc. Dissertation, Department
of Geography, University College London.
Lucinda Mileham, 2004. An investigation into the rate, causes and extent of
glacial retreat on the Speke and Elena Glaciers on the Rwenzori mountains,
Western Uganda*. Unpublished B.Sc. Dissertation, Department of Geography,
University College London.
Virginia Panizzo, 2004. A 150 year palaeolimnological investigation of a remote,
tropical, alpine lake**. Unpublished B.Sc. Dissertation, Department of
Geography, University College London.
Adinah Shackleton, 2004. Assessing the impact of glacial recession on Alpine
runoff in the Rwenzori central Massif, Uganda. Unpublished B.Sc. Dissertation,
Department of Geography, University College London.
* Awarded Best Undergraduate Dissertation Prize in hydrological science by the British
Hydrological Society.
** Awarded Best Undergraduate Dissertation Prize in Quaternary science by the Quaternary
Research Association.
9 Acknowledgements
The research was supported by grants from the Royal Geographical Society
(Ralph Brown Award 2003), The Royal Society (Taylor, Ssemmanda, Mackay),
University of London (Central Research Fund, Convocation Trust), University
College London (Dean’s Travel Fund, Department of Geography, The Friends
Trust), Quaternary Research Association (Mackay), and Earth and Space
Foundation (Mileham) and Rwenzori Beverage Company Limited (Uganda).
Institutional support was also provided by the Uganda National Council for
Science and Technology (Permit No. EC 583), Water Resources Management
Department (Uganda), Makerere University, and Uganda Wildlife Authority.
- 90
10 REFERENCES
Abruzzi, Duke of the (1907), The snows of the Nile, Geog. J., 29, 121-147.
Alloway, B.J. (1995) The origins of heavy metals in soils. In Alloway, B.J. (ed)
Heavy metals in soils, Blackie Academic and Professional, London.
Alnaes, K., 1998. The snow as the centre of the Konzo universe. In: H.
Osmaston, J. Tukahirwa, C. Basalirwa, and J. Nyakaana (Eds.), The
Rwenzori Mountains National Park, Uganda. Proceedings of the Rwenzori
Conference, Department of Geography, Makerere University, pp. 288-299.
Appleby PG and Oldfield F 1992 Application of 210Pb to sedimentation studies. In:
Ivanovich M & Harmon RS (eds.), Uranium series disequilibrium OUP, 731-
778.
Baker, B.H. (1967) Geology of Mount Kenya. Geological Survey of Kenya Report
No. 79, p. 78.
Baker, D.E. and Senft, J.P. (1995) Copper. In Alloway, B.J. (ed) Heavy metals in
soils, Blackie Academic and Professional, London.
Barker PA, Roberts N, Lamb HF, van der Kars S, and Benkaddour A 1994,
Interpretation of Holocene lake-level change from diatom assemblages in
Lake Sidi Ali, Middle Atlas, Morocco, Journal of Paleolimnology, 12, 223-234
Battarbee RW 1986 Diatom Analysis, Chpt. 26. In Berglund BE (ed.) Handbook
of Holocene Palaeoecology and Palaeohydrology, John Wiley and sons:
Chichester
Bengtsson L and Enell M, 1986. Chemical analysis, pp. 423–451. In Berglund B.
(ed.), Handbook of Holocene palaeoecology and palaeohydrology. J. Wiley,
Sons, Chichester
Bengtsson, L., S. Hagemann and K.I. Hodges (2004), Can climate trends be
calculated from reanalysis data? J. Geophys. Res., 109, D11111,
doi:10.1029/2004JD004536.
Bergstrøm, E. (1955), British Ruwenzori Expedition, 1952: Glaciological
observations – preliminary report, J. Glaciol., 2, 468-473.
Bilali, L.E., and Rasmussen, P.E., Hall, G.E.M. and Fortin, D. (2002) Role of
sediment composition in trace metal distribution in lake sediments, Applied
Geochemistry 17, 1171-1181.
Borghini F., Grimalt J.O., Sanchez-Hernandez J.C. and Bargagli R. 2005.
Organochlorine pollutants in soils and mosses from Victoria Land (Antarctica).
Chemosphere. 58: 271-278.
Boutron C.F. 1982. Atmospheric trace metals in the snow layers deposited at the
South Pole from 1928 to 1977. Atmos. Environ.16: 2451-2459.
Bradley RS, Vuille M, Diaz HF and Vergara W, 2006 Threats to water supplies in
the Tropical Andes, Science, 312, 1755-1756.
Bruland, K.W., Bertine, K., Koide, M. and Goldberg, E.D. (1974) History of metal
pollution in Southern California coastal zone, Environmental Science and
Technology, 8(5), 425-432.
Busulwa, H., 1998. The river fishery of the Rwenzori Mountains, Uganda. In: H.
Osmaston, J. Tukahirwa, C. Basalirwa, and J. Nyakaana (Eds.), The
- 91
Rwenzori Mountains National Park, Uganda. Proceedings of the Rwenzori
Conference, Department of Geography, Makerere University, pp. 130-134.
Camburn KE and Charles DF, 2000, Diatoms of low alkalinity lakes in the
Northeastern United States, The Academy of Natural Sciences of
Philadelphia Special Publication, 18: pp. 152
Cameron NG 1995 The representation of diatom communities by fossil
assemblages in a small acid lake, Journal of Paleolimnology, 14, 185-223
Charles DF, Smol JP, and Engstrom DR 1994, Palaeolimnological approaches to
biological monitoring, pp. 161-251. In Loeb SL and Spacie A (eds), Biological
Monitoring of Aquatic Systems, CRC Press, Florida
Christy, J.R. and W.B. Norris (2004), What may we conclude about global
tropospheric temperature trends?, Geophys. Res. Lett., 31, L06211,
doi:10.1029/2003GL019361.
Christy, J.R. et al. (2003) Error estimates of version 5.0 of MSU-ASMU bulk
atmospheric temperatures. J. Atmos. Oceanic Technol., 20, 613-629.
Cole G.A. (1975) Textbook of limnology, Mosby Company: St.Louis, pp.283.
Conway, D., Hanson, C.E., Doherty, R. and Persechino, A. (2007) GCM
simulations of the Indian Ocean dipole influence on East African rainfall:
present and future. Geophysical Research Letters, 34.
Davis GJ and Brinson MM 1980 Responses of submersed vascular plant
communities to environmental change, Aqua Biontic Report-for U.S.
Department of the Interior, East Carolina University; Greenville
Diaz HF, Grosjean M, and Graumlich L 2003 Climate variability and change in
high elevation regions: Past, present and future, Climatic Change,59, 1-4
Douglass, D.H., B.D. Pearson and S.F. Singer (2004), Altitude dependence of
atmospheric temperature trends: climate models versus observation,
Geophys. Res. Lett., 31, L13208, doi:10.1029/2004GL020103.
Edroma, E.L. (1974) Copper pollution in Rwenzori National Park, Uganda, The
Journal of Applied Ecology 11, 1043-1056.
Engstrom, D. and Swain, E. (1997) Recent declines in atmospheric Mercury
deposition in the upper Midwest, Environmental Science and Technology 31,
960-967.
Fægri K and Iversen J 1975, Textbook of Pollen Analysis. Munksgaard,
Copenhagen, 3rd ed., 295 pp
Fernández P., Vilanova R.M. and Grimalt J.O. 1999. Sediment fluxes of
polycyclic aromatic hydrocarbons in European high altitude mountain lakes.
Environ. Sci. Technol. 33: 3716-3722.
Fernández P., Vilanova R.M., Martinez C., Appleby P. and Grimalt J.O. 2000.
The historical record of atmospheric pyrolytic pollution over Europe registered
in the sedimentary PAH from remote mountain lakes. Environ. Sci. Technol.
34: 1906-1913.
Filippi, F. de, 1909. Der Ruwenzori. Briockhause (Leipzig), p. 471.
Flemming, C.A. and Trevors, J.T. (1989) Copper toxicity and chemistry in the
environment: a review, Water, Air and Soil Pollution 44, 143-148.
Francou, B., Vuille, M., Wagnon, P., Mendoza, J., and Sicart, J. (2003), Tropical
climate change recorded by a glacier in the central Andes during the last
- 92
decades of the twentieth century: Chacaltaya, Bolivia, 16°S, J. Geophys.
Res., 108, D5, 4154.
Fu, Q. et al. (2004), Contribution of stratospheric cooling to satellite-inferred
tropospheric temperature trends, Nature, 429, 55-58.
Fu, Q. and C.M. Johanson (2005), Satellite-derived vertical dependence of
tropical tropospheric temperature trends, Geophys. Res. Lett., 32, L10703,
doi:10.1029/2004GL022266.
Gaffen, D.J. et al. (2000), Multidecadal changes in the vertical temperature
structure of the tropical troposphere, Science, 297, 1242-1245.
Gasse F 1986 East African diatoms: taxonomy, ecological distribution, Cramer:
Stuttgart, pp.201
Gasse F 2002 Palaeoclimate: Kilimanjaro’s secrets revealed, Science, 298,
5593, 548-551
Glew, J.R. (1991) Miniature gravity corer for recovering short sediment cores,
Journal of Paleolimnology 5, 285-287.
Grimalt J.O., Fernández P., Berdie L., Vilanova R.M., Catalan J., Psenner R.,
Hofer R., Appleby P.G., Rosseland B.O., Lien L., Massabuau J.C. and
Battarbee R.W. 2001. Selective trapping of organochlorine compounds in
mountain lakes of temperate areas. Environ. Sci. Technol. 35: 2690 – 2697
Guichard, F., D. Parsons and E. Miller (2000), Thermodynamic and radiative
impact of the correction of sounding humidity bias in the tropics, J. Clim., 13,
3615-3624.
Hastenrath, S. and Kruss, P.D. (1992), The dramatic retreat of Mount Kenya’s
glaciers between 1963 and 1987: Greenhouse forcing, Ann. Glaciol., 16, 127-
133.
Hastenrath, S. and Greischar, L. (1997), Glacier recession on Kilimanjaro, 1912-
1989, J. Glaciol., 43, 455-449.
Hastenrath, S. (2001), Variations of East African climate during the past two
centuries, Climatic Change, 50, 209-217.
Hay, S.I., Cox, J., Rogers, D.J., Randolph, S.E., Stern, D.I., Shanks, D., Myers,
M.F. and Snow, R.W. (2002), Climate change and the resurgence of malaria
in the East African highlands, Nature, 415, 905-909.
Hense, A., P. Krahe and H. Flohn (1988), Recent fluctuations of tropospheric
temperature and water vapour content in the tropics, Meteorol. Atmopsh.
Phys., 38, 215-227.
Hermanson M.H. 1993. Historical accumulation of atmospherically derived
pollutant trace metals in the Arctic as measured in dated sediment cores.
Wat. Sci. Technol. 28: 33-41.
Hilton, J., Davison, W. and Ochsenbein, U. (1985) A mathematical model for
analysis of sediment core data: Implication for enrichment factor calculations
and trace-metal transport mechanisms, Chemical Geology 48, 281-291.
Intergovernmental Panel on Climate Change (IPCC) (2001), Climate Change
2001: The Scientific Basis: Contribution of Working Group I to the Third
Assessment Report of the Intergovernmental Panel on Climate Change,
edited by J.T. Houghton et al., 881 pp., Cambridge Univ. Press, Cambridge
Jollifer IT 1986 Principal Components Analysis, Springer-Verlag: NY, pp. 271
- 93
Jones, P.D. and A. Moberg (2003), Hemispheric and large-scale surface air
temperature variations: An extensive revision and an update to 2001, J. Clim.,
16, 206-223.
Jones V and Flower R 1986 Spatial and temporal veriability in periphytic diatom
communities: palaeoecological significance in an acidified lake, pp. 87-94. In
Smol JP et al. (eds), Diatoms and Lake Acidity, Dr W. Junk Publishers:
Dordrecht
Juggins S 1992 Zone v 1.2 unpublished computer programme, Environmental
Change Research Centre: University College London
Juggins S 2004, C2 Data Analysis 1.4.2 unpublished computer programme,
Environmental Change Research Centre: University College London
Kabata-Pendias, A. and Sadurski, W. (2004) Trace elements and compounds in
soil. In Merian, E., Anke, M., Ihnat, M. and Stoeppler, M. (eds) Elements and
their compounds in the environment volume 1: General aspects, Wiley,
Weinheim.
Kalnay, E. et al. (1996), The NCEP/NCAR 40-year reanalysis project, Bull. Am.
Meteorol. Soc., 77, 437-471.
Kamman, N.C. and Engstrom, D.R. (2002) Historical and present fluxes of
mercury to Vermont and New Hampshire lakes inferred from 210Pb dated
sediment cores, Atmospheric Environment 36, 1599-1609.
Kaser, G., 1999. A review of the modern fluctuations of tropical glaciers. Global
and Planetary Change, Vol. 22, pp. 93-103.
Kaser, G. and Noggler, B. (1991), Observations of Speke Glacier, Ruwenzori
Range, Uganda, J. Glaciol., 37, 313-318.
Kaser, G. and Noggler, B. (1996), Glacier fluctuations in the Ruwenzori Range
(East Africa) during the 20th century – a preliminary report, Z.Gletscherk.
Glazialgeol., 32, 109-117.
Kaser, G. and Osmaston, H. (2002), Tropical Glaciers, p. 207, Cambridge
University Press, Cambridge.
Kaser, G., Hardy, D.R., Mölg, T., Bradley, R.S. and Hyera, T.M. (2004), Modern
glacier retreat on Kilimanjaro as evidence of climate change: observations
and facts, Int. J. Climatol., 24, 329-339.
Kinzie RA, Banaszak AT and Lesser MP 1998 Effects of ultraviolet radiation on
primary productivity in a high altitude tropical lake, Hydrobiologia, 385, 23-32
Kite, G.W. (1981), Recent changes in level of Lake Victoria, Hydrol. Sci. Bull., 26,
233-243.
Koining KA, Schmidt R, Sammaruga-Wögrath, S, Tessadri R, Psenner R,1998,
Climate change as the primary cause for pH shifts in a high arctic lake, Water
Air and Soil Pollution, 104, 167– 180
Koining KA, Kamenik C, Schmidt R, Augusti-Panareda A, Appleby P, Lami A,
Prazakova M, Rose N, Schnell OA, Tessardi R, Thompson R, and Psenner R,
2002 Environmental changes in an alpine lake over the last two centuries-the
influence of air temperature on biological parameters, Journal of
Paleolimnology, 28, 147-160
Krammer K and Lange-Bertalot H 1986 Bacillariophycaea. I. Teil. Naviclaceae. In
Süsswasserflora von Mitteleuropa, Band 2/1, pp. 876
- 94
Krammer K and Lange-Bertalot H 1988 Bacillariophycaea 2. Teil. Bacillariacaea,
Epithemiacaea, Surirellacaea. In Süsswasserflora von Mitteleuropa, Band 2/2
Krammer K and Lange-Bertalot H (1991a) Bacillariophycaea 3. Teil. Zentrische
Diatomeen, Diatoma, Meridion, Asterionella, Tabellaria, Fragilaria, Eunotia
und Verwandte, Peronia und Actinella, In Süsswasserflora von Mitteleuropa,
Band 2/4
Krammer K and Lange-Bertalot H (1991b) Bacillariophycaea 4. Teil, Achnanthes,
Navicula, Gomphonema, Kritische Nachtraege, Literatur, In Süsswasserflora
von Mitteleuropa
Kruss, P.D. (1983), Climate change in East Africa: A numerical simulation from
100 years of terminus record at the Lewis Glacier, Mount Kenya,
Z.Gletscherk. Glazialgeol., 19, 43-60.
Kruss, P.D and Hastenrath, S. (1987), The role of radiation geometry in the
climate response of Mount Kenya’s Glaciers, part 1: Horizontal reference
surfaces, Int. J. Climatol., 7, 493-505.
Lamb, H.H. (1966) Climate in the 1960s. The Geographical Journal 132, 183–
212.
Leng, M.J., Lamb, A.L., Heaton, T.H.E., Marshall, J.D., Wolfe, B.B., Jones, M.D.,
Holmes, J.A., Arrowsmith, C., Chapter 4., “Isotopes in lake sediments”, Leng,
M.J., (Ed.), 2006, Isotopes in palaeoenvironmental research, Springer:
Netherlands, pp. 307.
Liu, X. and B. Chen (2000), Climatic warming in the Tibetan Plateau during
recent decades, Int. J. Climatol., 20, 1729-1742.
Livingstone, D.A. (1962) Age of deglaciation in the Ruwenzori Range, Uganda.
Nature 194, 859-860.
Livingstone, D.A. (1967) Postglacial vegetation of the Ruwenzori Mountains in
equatorial Africa, Ecological Monographs 37(1), 25-52.
Lotter AF and Bigler C 2000 Do diatoms in the Swiss Alps reflect the length if ice-
cover?, Aquatic Sciences, 62, 125-141
MacDonald R.W. et al., 2000. Contaminants in the Canadian Arctic: 5 years of
progress in understanding sources, occurrence and pathways. Sci. Tot.
Environ. 254: 93-234.
Mears, C.A. and F.J. Wentz (2005), The effect of diurnal correction satellite-
derived lower tropospheric temperature, Science, 309, 1548-1551.
Menzies, I. (1951), Some observations on the glaciology of the Ruwenzori
Range, J. Glaciol., 1, 511-512.
Meyers PA and Teranes JL, 2001 Sediment Organic Matter, pp. 239-270. In:
W.M. Last and J.P. Smol (eds.) Tracking environmental changes using lake
sediment - Vol. 2: Physical and geochemical methods.
Mölg, T., Georges, C. and Kaser, G. 2003. The contribution of increased
incoming shortwave radiation to the retreat on the Rwenzori glaciers, east
Africa, during the 20th Century. Int. J. Climatol., 23, 291-303.
Mölg, T. and Hardy, D.R. (2004), Ablation and associated energy balance of a
horizontal glacier surface on Kilimanjaro, J. Geophys. Res., 109, D16104,
doi:10.1029/2003JD004338.
- 95
Mölg T, Rott H, Kaser G, Fischer A and Cullen NJ, 2006, Comment on “Recent
glacial recession in the Rwenzori Mountains of East Africa due to rising air
temperature” by Taylor RG, Mileham LJ, Tindimugaya C, Majugu A,
Muwanga A, and Nakileza N, Geophys. Res. Lett., 33, L20404,
Murphey B.B. and Hogan A.W. 1992. Meteorological transport of continental soot
to Antarctica? Geophys. Res. Lett. 19: 33-36.
Muwanga, A (1997) Environmetal impacts of Copper mining at Kilembe, Uganda.
Unpublished thesis.
New, M.G., Lister, D., Hulme, M. and Makin, I. (2002), A high-resolution data set
of surface climate for terrestrial land areas, Clim. Res., 21, 1-25.
Nicholson, S.E. and Yin, X. (2001), Rainfall conditions in equatorial East Africa
during the nineteenth century as inferred from the record of Lake Victoria,
Climatic Change, 478, 387-398.
Norton, S.A. and Kahl, J.S. (1991) Progress in understanding the chemical
stratigraphy of metals in lake sediments in relation to acidic precipitation,
Hydrobiologia 214, 77-84.
Oerlemans, J. (2005), Extracting a climate signal from 169 glacier records,
Science, 308, 675-677.
Ogallo, L. J. (1988). Relationships between seasonal rainfall in East Africa and
the Southern Oscillation. Journal of Climatology, 8, 31–43.
Okunishi, K., Saito, T. and Yoshida, T., 1992. Accuracy of stream gauging by
dilution methods. Journal of Hydrology, 137, 231-243.
Osmaston, H. (1989), Glaciers, glaciations and equilibrium line altitudes on the
Rwenzori, in Quaternary and Environmental Research on East African
Mountains, edited by W.C. Mahaney, pp. 31-104, Balkema, Rotterdam.
Osmaston, H. (1998) The influence of the Quaternary history and glaciations of
the Rwenzori on the present landscape and ecology. In: H. Osmaston, J.
Tukahirwa, C. Basalirwa, and J. Nyakaana (Eds.), The Rwenzori Mountains
National Park, Uganda. Proceedings of the Rwenzori Conference,
Department of Geography, Makerere University, pp. 49-65.
Osmaston, H.A., 2006. Guide to the Rwenzori. The Rwenzori Trust (Ulverston).
Osmaston, H. and Kaser, G., 2001. Rwenzori Mountains National Park, Uganda
and Parc National des Virungas, Democratic Republic of Congo Glaciers and
Glaciations. WWF (ISBN: 0-95180394-8).
Pacyna, J.M. (1998) Source inventories for atmospheric trace metals. In
Harrison, R.M. and Grieken, R.V. (eds) Atmospheric particles, Chichester,
John Wiley and Sons.
Panizzo, V.N., Mackay, A.W., Ssemmanda, I., Taylor, R.G., Rose, N. and Leng,
M. (in review) Recent changes in aquatic productivity in a remote, tropical
alpine lake in the Rwenzori Mountain National Park, Uganda, associated with
glacier recession. Journal of Paleolimnology
Pascual, M. et al. (2006), Malaria resurgence in the East African Highlands:
Temperature trends revisited, P. Natl. Acad. Sci. USA, 103, 5829-5834.
Patz, J.A., Hulme, M., Rosenzweig, C., Mitchell, T.D., Goldberg, R.A., Githeko,
A.K., Lele, S., McMichael, A.J., LeSueur, D. (2002), Regional warming and
malaria resurgence, Nature, 420, 627-628.
- 96
Pepin, N.C. and D.J. Seidel (2005), A global comparison of surface and free-air
temperatures at high elevations, J. Geophys. Res., 110, D03104,
doi:10.1029/2004JD005047.
Psenner R and Schmidt R 1992 Climate-driven pH control of remote alpine lakes
and effects of acid deposition, Nature, 356, 781-783
Renberg I 1990 A procedure for preparing large sets of diatom slides from
sediment cores, Journal of Paleolimnology, 4, 87 – 90
Richardson JL 1968 Diatoms and lake typology in East and Central Africa, Int.
Revue Ges. Hydrobiol., 53: 2, 299-338
Robbins JA 1978 Geochemical and geophysical applications of radioactive lead,
pp. 285–393. In Nriagu JO (ed.) Biogeochemistry of Lead in the
Environment. Elsevier Scientific, Amsterdam
Rose, N.L. (1994) A note on further refinements to a procedure for the extraction
of carbonaceous fly-ash particles from sediments, Journal of Paleolimnology
11, 201-204.
Rose, N. (1995) Dating of recent lake sediments in the UK and Ireland using
SCP concentration profiles, Holocene 5, 328-335
Rose, N.L. (2001) Fly-ash particles. In Last, W.M. and Smal, J.P (eds) Tracking
environmental change using lake sediments volume 2: Physical and
geochemical methods, London, Kluwer Academic Publishers.
Rose, N.L. (2004) CARBYNET:
www.geog.ucl.ac.uk/ecrc/carbynet2/SCP_lake.htm
Rose N.L., Rose C.L., Boyle J.F and Appleby P.G. 2004. Lake sediment
evidence for local and remote sources of atmospherically deposited pollutants
on Svalbard. J. Paleolim. 31: 499 – 513
Roulet, M., Lucotte, M., Canuel, R., Farella, N., Courcelles, M., Guimaraes,
J.R.D., Merrgler, D. and Amorim, M. (2000) Increase in mercury
contamination recorded in lacustrine sediments following deforestation in the
central Amazon, Chemical Geology 165, 243-266.
Ryves DB, Juggins S, Fritz SC and Battarbee RW 2001, Experimental diatom
dissolution and the quantification of microfossil preservation in sediments,
Palaeogeography, Palaeoclimatology, Palaeoecology, 172, 99-113
Santer, B.D. et al. (2005), Amplification of surface temperature trends and
variability in the tropical atmosphere, Science, 309, 1551-1556.
Sayer CD 2001, Problems with the application of diatom-total phosphorous
transfer functions: examples from a shallow English lake, Freshwater Biology,
46, 743-757.
Schmitt, K., 1998. The biodiversity of the Rwenzori Mountains. In: H. Osmaston,
J. Tukahirwa, C. Basalirwa, and J. Nyakaana (Eds.), The Rwenzori Mountains
National Park, Uganda. Proceedings of the Rwenzori Conference,
Department of Geography, Makerere University, pp. 91-102.
Sherwood, S.C., J.R. Lanzante and C. L. Meyer (2005), Radiosonde daytime
biases and late-20th century warming, Science, 309, 1556-159.
Stockmarr J 1972 Tablets with spores used in absolute pollen analysis, Pollen
Spores, 13, 615-621
- 97
Talbot MR and Laerdal T 2000 The late Pleistocene-Holocene palaeolimnology
of Lake Victoria, East Africa, based upon elemental and isotopic analyses of
sedimentary organic matter, Journal of Paleolimnology, 23, 141 – 164
Talks, A. (1993), East African Hot Ice 1993, Internal Report - Sir Roger
Manwood's School, Kent (UK), copies available from the Royal Geographical
Society (UK).
Talling JF and Lemoalle J 1998 Ecological dynamics of tropical inland waters,
Cambridge University Press: UK, pp. 441
Taylor RG and Howard KWF 1998 Post-Palaeozoic evolution of weathered
landsurfaces in Uganda by tectonically controlled cycles of deep weathering
and stripping, Geomorphology, 25, 173-192
Taylor RG, Mileham LJ, Tindimugaya C, Majugu A, Muwanga A, and Nakileza N
2006a Recent recession in the Rwenzori Mountains of East Africa due to
rising air temperature, Geophysical Research Letters, 33, 10402
Taylor RG, Mileham LJ, Tindimugaya C, Majugu A, Muwanga A, and Nakileza N,
2006b, Reply to comment by T. Mölg et al. on “Recent glacial recession in the
Rwenzori Mountains of East Africa due to rising air temperature”, Geophys.
Res. Lett., Vol. 33, No. 20, L20405
Temple, P.H. (1967), Further observations on the glaciers of the Ruwenzori,
Geogr. Ann. A, 50, 136-150.
ter Braak CJF and Šmilauer P 2002, CANOCO Reference manual and
CanoDraw for Windows User’s Guide: Software for Canonical Community
Ordination (version 4.5), Microcomputer Power (Ithaca, NY, USA), pp.500
Tett, S. and P. Thorne (2004), Tropospheric temperature series from satellites,
Nature, 432, doi:10.1038/nature03208.
Thompson LG 2000 Ice core evidence for climate change in the Tropics:
implications for our future, Quaternary Science Reviews, 19, 19-35
Thompson, L.G., Mosley-Thompson, E., Davis, M.E., Henderson, K.A., Brecher,
H.H., Zagorodnov, V.S., Mashiotta, T.A., Lin, P-N., Mikhalenko, V.N., Hardy,
D.R. and Beer, J. (2002), Kilimanjaro ice core records: evidence of Holocene
climate change in tropical Africa, Science, 298, 589-593.
Thompson, L.G. et al. (2006), Abrupt tropical climate change: Past and present,
P. Natl. Acad. Sci. USA, 103, 10536-10543.
Thorne, P.W. et al. (2005), Revisiting radiosonde upper air temperatures from
1958 to 2002, J. Geophys. Res., 110, D18105, doi:10.1029/2004JD005753.
Tyson R V 1995 Sedimentary Organic Matter: Organic Facies and Palynofacies.
Chapman and Hall, London.
Verburg P, Hecky RE and Kling H. 2003 Ecological consequences of a century of
warming in Lake Tanganyika, Science, 301, 505-507
Verschuren D 2003, Lake-based climate reconstruction in Africa: progress and
challenges, Hydrobiologia, 500, 315-330.
Vinebrooke, R.D., and Leavitt, P.R., 19936, Effects of ultraviolet radiation on
periphyton in an alpine lake, Limnologogy, Oceanography, 41, (5), 1035-
1040.
Vives I., Grimalt J.O., Catalan J., Rosseland B.O. and Battarbee R.W. 2004a.
Influence of altitude and age in the accumulation of organochlorine
- 98
compounds in fish from high mountain lakes. Environ. Sci. Technol. 38: 690 –
698.
Vives I., Grimalt J.O., Lacorte S., Guillamón M., Barceló D, and Rosseland B.O.
2004b. Polybromodiphenyl ether flame retardants in fish from lakes in
European high mountains and Greenland. Environ. Sci. Technol. 38: 2338-
2344.
Vogel, S.W. (2002), Usage of high-resolution Landsat7 band 8 for single band
snow cover classification, Ann. Glaciol., 34, 53-57.
Vörösmarty, C.J., B. Fekete, and B.A. Tucker. 1998. River Discharge Database,
Version 1.1 (RivDIS v1.0 supplement). Available through the Institute for the
Study of Earth, Oceans, and Space / University of New Hampshire, Durham
NH (USA).
Vuille, M. and R.S. Bradley (2000), Mean annual temperature trends and their
vertical structure in the tropical Andes, Geophys. Res. Lett., 27, 3885-3888.
Wagnon, P., Ribstein, P., Francou, B., and Poyaud, B. (1999), Annual cycle of
energy balance of Zongo Glacier, Cordillera Real, Bolivia, J. Geophys. Res.,
104, D4:3907-3923.
Wania F. and Mackay D. 1996. Tracking the distribution of persistent organic
pollutants. Environ. Sci. Technol. 30: 390A – 396A
Werner D (ed.) 1977, The Biology of Diatoms, Botanical Monographs: vol.13,
Blackwell Science Publications: Oxford, p. 498.
Whittow, J.B. (1960) Some observations on the snowfall of the Ruwenzori.
Journal of Glaciology, 3, 765-772.
Whittow, J.B., Shepherd, A., Goldthorpe, J.E., and Temple, P.H. (1963),
Observations on the Glaciers of the Rwenzori, J. Glaciol., 4, 581-615.
Whittow, J.B. (1966) The landforms of the central Ruwenzori, East Africa, The
Geographical Journal 132(1), 32-42.
Wolff E.W. and Suttie E.D. 1994. Antarctic snow record of southern hemisphere
lead pollution. Geophys. Res. Lett. 21: 781-784.
Wolff E.W., Suttie E.D. and Peel DA. 1999. Antarctic snow record of cadmium,
copper and zinc content during the twentieth century. Atmos. Environ.33:
1535-1541.
Yang, H. & Rose, N. (2005) Trace element pollution records in some UK lake
sediments, their history, influence factors and regional differences.
Environment International 31, 63-75.
