This paper presents the first bathymetric map for the approximately 17 km by 2 km alkaline Lake Bogoria situated in the eastern Rift Valley of Kenya. Longitudinal and transverse cross sections of the lake are also provided. Northern, central and southern basins of the lake had maximum depths of 5.9 m, 10.2 m and 8.4 m respectively. Average depth was 5.68 m and volume was calculated to be 164 × 106 m3.
BATHYMETRY OF LAKE BOGORIA, KENYAPhil HickleyEnvironment Agency, Arthur Drive, Hoo Farm Industrial Estate, Worcester Road,Kidderminster, DY11 7RA, U.K.firstname.lastname@example.orgRosalind R. BoarCentre for Ecology, Evolution and Conservation (CEEC), Environmental Sciences,University of East Anglia, Norwich NR4 7TJ, U.K.R.Boar@uea.ac.ukKenneth M. MavutiDepartment of Zoology, University of Nairobi, P.O. Box 30197, Nairobi, Kenyakmavuti@uonbi.ac.keABSTRACTThis paper presents the first bathymetric map for the approximately 17 km by 2 km alkaline Lake Bogoria situated in the eastern Rift Valley of Kenya. Longitudinal and transverse cross sections of the lake are also provided. Northern, central and southern basins of the lake had maximum depths of 5.9 m, 10.2 m and 8.4 m respectively. Average depth was 5.68 m and volume was calculated to be 164 ×10⟨sup⟩6⟨/sup⟩ m⟨sup⟩3⟨/sup⟩.INTRODUCTIONLake Bogoria is situated at 0º 16' N and 36º 06' E, 240 km north of Nairobi in the floor of the eastern Rift valley of Kenya, and has an elevation of approximately 980 m (a.s.l.). It is a closed-basin alkaline saline lake with many boiling springs and fumaroles present along its shores (Cioni et al., 1992). Principal freshwater inflows comprise two perennial, spring-fed streams at the southern end of the lake and the seasonal Waseges-Sandai River to the north. The lake forms part of the Lake Bogoria National Reserve which not only is a popular tourist destination but also provides resources such as grazing and access to medicinal plants for the local community (Jones, 2000). The Reserve is a proposed World Heritage Site, an Integrated Management Plan is in preparation and a community based Wetlands Conservation Project has been initiated (Wetlands International, 2003).Lake Bogoria is one of the lakes upon which the lesser flamingo (Phoeniconaias minor Gray) depends for its life strategy, making it an important aquatic resource. The lake usually supports dense blooms of the cyanobacterium Spirulina platensis (Nordst.) Geitl., principal food of the lesser flamingo. Up to 1,074,000 flamingos, 98.8% of the central Rift population, have been counted on Lake Bogoria at any one time. (Owino et al., 2001).In 1988 the International Lake Environment Committee Foundation (ILEC), in cooperation with the United Nations Environment Programme (UNEP), started a data collection project to gather basic information on natural and artificial lakes and its World Lakes Database (ILEC, 2001) now holds data on more than 500 lakes in 73 countries. An attempt is being made by ILEC to update the information on lakes covered by the database and additional information has been requested. For Lake Bogoria the currently available information on bathymetry is limited to figures for mean and maximum depth of 5.4 m and 10 m respectively.One of the most diagnostic physical features of a lake is its depth with knowledge of such being an important tool for the limnologist. Accordingly, the purpose of this paper is to place previously unavailable information on the bathymetry of Lake Bogoria into the domain of future researchers. At the present time, ongoing work includes: organic inputs (authors), primary production (Mills, pers. comm.), secondary production (Otieno, pers. comm.), flamingo ecology (Childress, pers. comm.), climate change indicators (Verschuren, pers. comm.).METHODSLake Bogoria was surveyed during the period 16-26 August 2002 using a Lowrance X-15A chart recording echo-sounder with a 20º transducer beam as used previously by Hickley et al. (2002) to determine the depth of Lake Naivasha. The soundings were carried out in calm conditions from a small boat driven at constant speed along transect lines (figure 1), the positions of which were monitored with a Garmin 12 hand held GPS receiver running 4.59 software. To optimise the yield of the survey in terms of accuracy, the recommended system of tracks crossing the main track was used (Håkanson, 1981). For each paper echo-recording chart, the positions along the transect line at which the depths of 1 m, 2 m, etc. and maximum depth occurred were converted by interpolation to UTM grid readings. The mE and mN values were then plotted as x:y co-ordinates on a plan outline of the lake. Mathematical interpolation not being necessary (Håkanson, 1981), these spot depths were joined using best fit by eye to form contour lines. The lake outline was derived from Survey of Kenya 1:50,000 scale maps made in 1973 and based on aerial photographs taken during January - February 1969. To produce depth profiles for each transect line, actual depths to the nearest 0.1 ft (3 cm) were read at 10 m intervals from the echo-recording charts. Morphometrical values were determined according to Håkanson (1981).Figure 1. Locations of echo-sounding transects used to produce depth profiles and bathymetric map for Lake Bogoria. eastings (mE) and northings (mN) are GPS co-ordinates within the UTM grid zone 37N.RESULTSA summary of morphometrical data for Lake Bogoria is given in table 1. A bathymetric map is presented in figure 2 and shows the lake to comprise northern, central and southern basins, the northern being the shallowest. Bed profiles are shown in figure 3 (longitudinal), in figure 4 (transverse) and, as an example of where the bed appeared noticeably undulating, in figure 5. The hypsographic curve and associated depth distributions are given in figure 6, indicating the lake form to be Lme (Linear meso (Håkanson,1977)). By way of a benchmark for the shoreline at the time of the echosounder survey, the waters edge was 45.6 m distant from the highest-spouting geyser (GPS reading: 0028028 mE, 0175457 mN) which is located about halfway along the western shore in the area referred to as the Hot Springs.Table 1. Morphometrical data for Lake Bogoria, Kenya, August 2002.ParameterUnitValueMaximum lengthkm17.25Maximum widthkm3.75Mean widthkm1.68Maximum depthm10.2Mean depthm5.68Modal depthm7.8Median depth D⟨sub⟩50⟨/sub⟩m6.491⟨sup⟩st⟨/sup⟩ quartile depth D⟨sub⟩25⟨/sub⟩m4.813⟨sup⟩rd⟨/sup⟩ quartile depth D⟨sub⟩75⟨/sub⟩m8.16Shoreline lengthkm44.0Lake areakm⟨sup⟩2⟨/sup⟩28.95Volumem⟨sup⟩3⟨/sup⟩ ×10⟨sup⟩6⟨/sup⟩164.4DISCUSSIONSome of the morphometrics of three basins, such as the steep gradients on the eastern side, are likely to be symptomatic of the lake being situated in a region of notable faulting and volcanic activity (Young et al., 1991; Grimaud et al., 1994). The strong variations in depth which occurred in places (figure 5), and which were up to 2 m vertical per 100 m horizontal, are not an artefact of the echo-sounding technique. The calm conditions in which the survey was carried out, the use of a wide (20º) transducer beam and the occurrence ofFigure 2. Bathymetric map of Lake Bogoria (August 2002) showing 1 m contour lines together with maximum depths for northern, central and southern basins. The shaded zone of undulating bed is an area where the lake bottom was noticeably more uneven than elsewhere. The frame shows latitude and longitude co-ordinates. Also marked are representative mE and mN intersections of the UTM grid as used in figure 1.Figure 3. South to North depth profile of Lake Bogoria (August 2002). The dotted line within the lake outline marks the position of the profile that was derived from transect lines SN1, SN3, SN5 and SN7 (figure1).smooth sections of lake bed on the continuously recorded echograms for all transects confirm the reality of the undulations. These undulations are less likely to represent structural geology than they are to reflect long term patterns in sedimentation caused by currents generated by underwater, geothermal springs and wave action. Recent local faulting, however, has made a noticeable contribution to bed morphology, with the leading fault being N-S on the eastern shore. Transects WE 6, 7 & 8 (figure 4) show good examples on their eastern sides of a consistent break in slope.The accuracy of the volume estimate for subsequent limnological calculations will depend upon the amount of annual variation in lake level that is likely to occur. Although Lake Bogoria is the only alkaline lake within the Rift Valley that has relatively negligible water level fluctuations (Wetlands International, 2003), a consequence of step-faulting within the region is a poor supply of water from a small catchment area (Onywere et al., 1996). Accordingly, lake levels are likely to be linked closely to local rainfall patterns. Indeed, the seasonal Wasengas-Sandai river was noted to have zero flow at both the time of the August 2002 survey and the same time the previous year.Twenty-five years (1977-2001) of daily precipitation records from a rain gauge located within the Lake Bogoria National Reserve (0° 21' N, 36° 04' E) have been collated by LaVigne and Ashley (2001). Annual average precipitation over the period was 708.3 mm, distributed bimodally throughout the year. The wettest year was 1977 (1053.4 mm) and the driest was 1984 (243.1 mm). These local rainfall patterns are likely to have been influenced by the effects in East Africa of El Niño and La Niña events (Ogallo, 1987; Nicholson, 1996; Nicholson & Selato, 2000) but is also complicated by the topography of the Rift Valley which widens in a triangular shape northwards with escarpments and mountain ranges rising from its widened floor. Evapotranspiration is high at c. 2500 mm/yr. Some variation in lake level was observed in the three years prior to the bathymetric survey. This was very noticeable at the northern end of the lake where the gradient was slight but much less soFigure 4. West to East depth profiles of lake Bogoria (August 2002). The dotted lines within the lake outline mark the positions of the profiles where WE1-WE16 are transect lines as shown in figure 1.elsewhere. Aerial photographs taken in September 1999 (plate 1) and August 2001 (plate 2), when precipitation in the preceding 3 months was 227 mm and 87 mm respectively, show that the northern lake edge was 280 m further receded in the drier of the two years. At the time of the echosounder survey in August 2002, the lake was even shallower (plate 3) withFigure 5. Examples of bottom profiles of Lake Bogoria in the zone of undulating bed. Positions of the profiles are marked on the selected portion of the bathymetric map as solid lines A, B and C.Figure 6. (a) Percentage hypsographic curve and (b) Percentage of lake area underlain by each depth zone for lake Bogoria (August 2002).Plate 1. Aerial view across the northern end of lake Bogoria looking towards the eastern shore in September 1999. Top left is a partially inundated belt of riparian trees along the Waseges-Sandai river. The solid and open circles mark positions on the shore that are also marked in Plates 2 and 3. The pale dots scattered all over the water surface are lesser flamingos. (Photograph by P. Hickley)Plate 2. Aerial view across the northern end of lake Bogoria looking towards the eastern shore in August 2001. The solid and open circles mark positions on the shore that are also marked in Plates 1 and 3. The dotted lines show the approximate position of the water's edge in September 1999 (Plate 1). (Photograph by P. Hickley)Plate 3. Aerial view across the northern end of lake Bogoria looking towards the eastern shore in August 2002. The solid and open circles mark positions on the shore that are also marked in Plates 1 and 2. The dotted line shows the approximate position of the water's edge in August 2001 (Plate 2). (Photograph by P. Hickley)the northern lake edge being 190 m further down the shore than in August 2001, 470 m further than September 1999. On that part of the western shore with the shallowest gradient, about midway along, the shoreline receded approximately 50 m in the same three-year period. Although seemingly dramatic to the casual observer, the impact of these changes in level on lake area was a reduction of less than 5%, this because, in contrast with the above, the position of the shoreline around the rest of the lake with its steeper gradients receded only 5-10 m. Moreover, and importantly, from checks made against GPS readings the position of the lake edge at the time of the echosounder survey aligned closely to the outline used for the bathymetric map.The difference in lake level observed during the period September 1999 to August 2002 was calculated to represent a 7.4×10⟨sup⟩6⟨/sup⟩ m⟨sup⟩3⟨/sup⟩ (4.5%) reduction in lake volume. Direct implications for lake ecology of any such reduction in volume include changes in the behaviour and distribution of flamingo with indirect impacts on ecology being linked to primary production by S. platensis through changes in lake chemistry and mixing of the water column (Harper et al., 2003). Also, exposure of the shallow gradient on the western shore creates opportunity for colonization by the dominant shoreline grass, Sporobolus spicatus Kunth although simultaneous drying of seasonal freshwater inflows would presumably reduce the contribution to primary production on the lake shore by the important sedge-dominated (Cyperus rubicundus Vahl and C. laevigatus L.) community. On the much steeper eastern shoreline where the plant community is dominated by Acacia and Commiphora species on unstable slopes, changes in water level will have comparatively small implications for riparian primary production.Some of the alkaline lakes of the Rift Valley are known to have periods of significant reduction in size. Lake Nakuru, for example, dried out twice during the 1990s (Nasirwa, 2000) although such an event is much less likely to happen to Lake Bogoria since it has been the most stable of the alkaline lakes within Kenya (Wetlands International, 2003). Notwithstanding the likelihood, when Lake Nakuru was being studied in the 1970s, Lake Bogoria was reported by Vareschi (1978) to be 12 m deep, 2 m deeper than now.The outcome of the above echo-sounding survey has been to provide the first satisfactory description of the bathymetry of Lake Bogoria, Kenya. Such information should be of value to subsequent limnological and ecosystem research.ACKNOWLEDGEMENTSEcho-sounding was carried out as part of the "Lakes of the Rift Valley" project sponsored by the Earthwatch Institute. The authors are grateful to the Office of the President of the Government of Kenya for research permission and to the Earthwatch volunteers for their assistance in the field.REFERENCESCioni, R., G. Fanelli, M. Guidi, J.K. Kinyariro & L. Marini (1992). Lake Bogoria hot springs (Kenya): Geochemical features and geothermal implications. Journal of Vulcanology and Geothermal Research 50: 231-246.Grimaud, P., J.-P. Richert, J. Rolet, J.-J. Tiercelin, J-P. Xavier, C.K. Morley, C. Coussement, S.W. Karanja, R.W. Renaut, G. Guerin. (1994). Fault geometry and extension mechanisms in the central Kenya Rift, East Africa: A 3D remote sensing approach. Bulletin des Centre de Recherches Exploration-Production Elf-Aquitaine 18: 59-92.Håkanson, L. (1977). On lake form, lake volume and lake hypsographic survey. Geografiska Annaler 59: 1-29.Håkanson, L. (1981). A Manual of Lake Morphometry. Springer-Verlag, Berlin, Heidelberg & New York.Harper, D.M., R.B. Childress, M.M. Harper, R.R. Boar, P. Hickley, S.C. Mills, N. Otieno, A. Drane, E. Vareschi, O. Nasirwa, W.E. Mwatha, J.P.E.C. Darlington & X. Escuté-Gasulla (2003). Aquatic Biodiversity and Saline Lakes: Lake Bogoria, National Reserve, Kenya. Hydrobiologia 500: 259-276.Hickley, P., R.G. Bailey, D.M. Harper, R. Kundu, S.M. Muchiri, R. North & A. Taylor (2002). The status and future of the Lake Naivasha fishery, Kenya. Hydrobiologia 488: 181-190.ILEC (2001). World Lakes Database. International Lake Environment Committee Foundation website, http://www.ilec.or.jp/database/database.htmlJones, T. (ed.) (2000). Meeting the challenge: WWF's Work in Africa and Madagascar. World Wide Fund for Nature, Gland.LaVigne, M. & G. Ashley (2001). Climatology and Rainfall Patterns: Lake Bogoria National Reserve (1976-2001) Rutgers University, Department of Geology, New Brunswick.Nasirwa, O. (2000). Conservation status of flamingos in Kenya. Waterbirds 23 Suppl.1: 47-51.Nicholson, S.E. (1996). A review of climate dynamics and climate variability in eastern Africa. In T.C. Johnson & E.O. Odada (eds.), The limnology, climatology and paleoclimatology of the East African Lakes. Gordon & Breach Publishing Group, New York. Pp. 25-56.Nicholson, S.E. & J.C. Selato (2000). The influence of La Niña on African rainfall. International Journal of Climatology 20: 1761-1776.Ogallo, L. (1987). Relationships between seasonal rainfall in East Africa and the southern oscillation. Journal of Climatology 7: 1-13.Onywere, S.M., S.J. Gaciri & N. Opiyo-Akech (1996). Structural analysis of the drainage basin of Kenyan Rift Valley lakes (Aberdare detachment system) using geographical information system (GIS): Understanding the causes of the changes in the water levels of the lakes. In I.O. Nyambok & D.W. Ichangi (eds.), Geology for development within a sustainable environment. Geological Society of Kenya , Nairobi. Pp. 677-708.Owino, A.O., J.O. Oyugi, O. Nasirwa & L.A. Bennun (2001). Patterns of variation in waterbird numbers on four Rift Valley lakes in Kenya, 1991-1999. Hydrobiologia 458: 45-53.Vareschi, E. (1978). The ecology of Lake Nakuru (Kenya) I: Abundance and feeding of the lesser flamingo. Oecologia 32: 11-35.Wetlands International (2003). Ramsar Sites Database. Wetlands International website, http://www.wetlands.org/RDB/Ramsar_Dir/Kenya/KE003D02.htmYoung, P.A.V., P.K.H. Maguire, N.d'A. Laffoley & J.R. Evans (1991). Implications for the distribution of seismicity near Lake Bogoria in Kenya Rift. Geophysical Journal International 105: 665-674.