This article was downloaded by: [Leabharlann Choláiste na Tríonóide/Trinity College Library & IReL] On: 19 August 2015, At: 04:07 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG Journal of Maps Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tjom20 Geology and stratigraphy of the south- eastern Lake Edward basin (Petroleum Exploration Area 4B), Albertine Rift Valley, Uganda Christopher J. Nicholasa, Ian R. Newthb, Dozith Abeinomugishac, Wilson M. Tumushabec & Lauben Twinomujunid a Department of Geology, School of Natural Sciences, Trinity College, University of Dublin, Dublin 2, Republic of Ireland b Count Geophysics Ltd., 7 Hatton Garden, London EC1N 8AD, UK Click for updates c Petroleum Exploration and Production Department (PEPD), Ministry of Energy and Mineral Development, P.O. Box 9, Entebbe, Republic of Uganda d Department of Geological Survey and Mines of Uganda (DGSM), Plot 21-29, Johnstone Road, P.O. Box 9, Entebbe, Uganda Published online: 06 Feb 2015. To cite this article: Christopher J. Nicholas, Ian R. Newth, Dozith Abeinomugisha, Wilson M. Tumushabe & Lauben Twinomujuni (2015): Geology and stratigraphy of the south-eastern Lake Edward basin (Petroleum Exploration Area 4B), Albertine Rift Valley, Uganda, Journal of Maps, DOI: 10.1080/17445647.2015.1010616 To link to this article: http://dx.doi.org/10.1080/17445647.2015.1010616 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms- and-conditions 5 1 0 2 st u g u A 9 1 7 0 4: 0 at ] L e R I & y r a r b Li e g e oll C y nit ri T e/ d ói n o rí T a n e st ái ol h C n n a rl a h b a e L [ y b d e d a o nl w o D JournalofMaps,2015 http://dx.doi.org/10.1080/17445647.2015.1010616 SCIENCE 5 Geology and stratigraphy of the south-eastern Lake Edward basin 1 0 (Petroleum Exploration Area 4B), Albertine Rift Valley, Uganda 2 ust Christopher J. Nicholasa∗, Ian R. Newthb, Dozith Abeinomugishac, Wilson M.Tumushabec g Au and Lauben Twinomujunid 9 1 7 aDepartmentofGeology,SchoolofNaturalSciences,TrinityCollege,UniversityofDublin,Dublin2, 4:0 RepublicofIreland;bCountGeophysicsLtd.,7HattonGarden,LondonEC1N8AD,UK;cPetroleum 0 ExplorationandProductionDepartment(PEPD),MinistryofEnergyandMineralDevelopment,P.O. at Box9,Entebbe,RepublicofUganda;dDepartmentofGeologicalSurveyandMinesofUganda(DGSM), L] Plot21-29,JohnstoneRoad,P.O.Box9,Entebbe,Uganda e R I (Received24July2013;resubmitted3December2014;accepted19December2014) & y r TheLake EdwardbasinlieswithintheAlbertineRiftValleyofUgandaandtheDemocratic a br Republic of Congo which forms the northern end of the western arm of the East African Li Rift System. It is a frontier petroleum prospective area, which, at the outset of this study, ge hadnoexplorationwellsdrilledwithinitoranydeepreflectionseismicsurveys.Therehave Colle btheeenonssohmoereprriefvt-ifiolulsssetduidmieesntisnotrheesbtaabsilnis,hbedutanownoerkparboldeucsterdatiagrgaepohliocgifcraalmmewaporskubfodrivtihdeimng. y Between 2007 and 2010, Dominion Uganda Ltd., in collaboration with Trinity College nit Dublin and the Petroleum Exploration and Production Department of the Ministry of ri Energy, Uganda, undertook a geological mapping survey of the south-eastern onshore part T e/ of the basin, known as petroleum ‘Exploration Area 4B’ (EA4B). Five rift sediment d ói formations were identified and mapped across the area to produce a new geological map of n EA4B. Palynological analyses suggest that all exposed rift sediments are (Late to Mid) o rí Pleistocene–Holocene.EA4Bisdominatedbyanorth-easttosouth-westtrendingfaultzone T a whichunderwentsignificantextension withinthelast130,000yearstoproduceatrough,or n sub-basin, to the south-east against the rift margin. This trough subsequently filled, initially e st with ponded swamp clays, followed by coarse fluvial and alluvial clastics. There is field ái evidencefor minorinversionand‘pop-up’structures alongsomefootwallcrests,suggesting hol that the neotectonic phase is compressional or transpressional, and this has caused stream C rejuvenationandincision. n n a Keywords: East African Rift System; Lake Edward; structure and stratigraphy; petroleum arl exploration h b a e L [ y b 1. Introduction d e d The Albertine Rift Valley forms the northern of three sectors which comprise the Western Arm a o of the Miocene to Recent East African Rift System, with the national border between Uganda nl w and the Democratic Republic of Congo (DRC) passing along the centre of the valley. The o D ∗Correspondingauthor.Email:[email protected] #2015ChristopherJ.Nicholas 2 C.J.Nicholas et al. Albertine Rift is structurally segmented into a series of asymmetric graben depocentres, or domains, each separated by NW–SE or W–E trending basement highs or ‘Accommodation Zones’ (AZs), manyof which are also associated with volcanics (Rubondo, 2005). Forthe pur- posesofpetroleumexploration,eachoftheriftbasinsontheUgandansideoftheborderhasbeen subdivided into blocks, or ‘Exploration Areas’ (EA). EA 4 covers the Ugandan Lake Edward– 5 Lake George basins at the southern end of this northern rift sector (Figure 1). Lakes Edward 1 0 2 st u g u A 9 1 7 0 4: 0 at ] L e R I & y r a r b Li e g e oll C y nit ri T e/ d ói n o rí T a n e st ái ol h C n n a rl a h b a e L [ y b d e d a o nl w o D Figure1. GeographicalpositionandmainfeaturesofEA4B.ThefigureshowstheextentofEA4Bduring theinitialphasesofpetroleumexplorationbyDominionUgandaLtd.,between2007and2010.Alsoshown arethetwovolcanicprovincesinthebasin,thelocationofdatedDuralimestones,andthepositionofthe Kazinga–Kikarara–KirurumaFaultZoneandBwambaratrough. Journal of Maps 3 andGeorgeareseparatedfromtheLakeAlbertbasinbyacomplexcombinationofanAZmarked by the alkali basalt Katwe tuffs, and the Rwenzori Mountains, which are flanked along their eastern margin by further sporadic volcanics including rare carbonatite lavas. Exploration Area 4 was originally split into two further EAs, one to the north (EA4A) and the other to the south (EA4B). EA4B included the onshore rift sediments to the south-east of Lake Edward and a 5 large proportion of the Ugandan side of the lake (Figure 1). At the time that the present study 01 was undertaken, no petroleum exploration wells had previously been drilled in the Lakes st 2 Edward and George basins, and only a shallow penetrating seismic reflection survey (depth≤ u 50m) had been carried out on Lake Edward (see below). Consequently, EA4 to the south of g u A the Rwenzori Mountains remained a virgin frontier area unexplored for hydrocarbons. 9 EA4Bisboundedtothesouth-eastbythefaultededgeoftheriftvalley,whichexposesbase- 1 7 mentschistosegneissesalongtheriftshoulder.Tothesouth-west,itsonshoreboundaryfollows 0 4: thenationalborderwiththeDRCalongtheIshashaRiver,andfollowsitwestwardsasthispasses 0 at offshore and diagonally across the centre of the lake. To the north-east, EA4B adjoins EA4A L] within an elevated plateau region covered by the dense Maramagambo rain forest. Within e R these confines, onshore EA4B physiography and vegetation effectively subdivide it into two I & halves; to the north-west towards the lakeside, the area is flat with open savannah grasslands y of the Queen Elizabeth National Park and Kigezi Game Reserve, which are home to many r a br large African mammals such as elephants, hippos, water buffalo, various antelope, and most Li famously prides of tree-climbing lions. The south-eastern half of the area, however, consists of e g gradually raised and subsequently incised topography covered with dense vegetation in parts e oll and is cultivated for coffee,tea, bananas, and papaya. C y InitialgeologicalmappingintheregionwasundertakenbytheGeologicalSurveyofUganda nit in the late 1950s, mainly to identify mineral resources in the Proterozoic basement (Geological Tri Survey of Uganda, 1961). Although the Late Pleistocene–Holocene Katwe and Bunyaruguru e/ volcanics were identified, the remaining rift-fill sediments were left undifferentiated. However, d ói some lineations on aerial photographs were recognised as potential faults trending NE–SW n río across the onshore area. Subsequently, a period of academic research in conjunction with some T a preliminary geological field mapping and combined with a prior gravity survey, conducted by e n the Petroleum Exploration and Production Department (PEPD) of the Ministry of Energy and st Minerals Development, Uganda between 1960 and 2000 in the Lake Edward–Lake George ái ol region, identified up to approximately 13 informal lithological formations in various litho- and h C chrono-stratigraphic schemes (see Bishop, 1969; Byakagaba, 1997; Musisi, 1991; Pickford, nn Senut, & Hadoto, 1993; Senut & Pickford, 1994) (Figure 2). These were typically assigned an a rl age range of anywhere between Late Miocene and Holocene. However, such informal schemes a h didnotclearlyidentifyordefineformationalboundingsurfacesorinternalfaciescharacteristics. b a e Consequently,thelateralcorrelationoftheseunitswas,andhasremained,impossible.Inaddition, L [ the primary biostratigraphic constraints used in these schemes relied on correlation between a y b combination of gastropod and mammalian faunal provinces in the Albertine Rift and similar d e assemblages amongst dated tuffs in Ethiopia. This correlation is in itself problematic, as it d a o does not allow for palaeo-biogeographical differences across the East African region at this nl w time, and assumes that the species concernedappeared simultaneouslyin both areas. o D The International Decade for the East African Lakes (IDEAL) marked a period of renewed research on Lake Edward. A limited, shallow-penetration reflection seismic survey (depth ≤ 50m) was carried out by Syracuse University in 1996, accompanied by some shallow piston coring on Ugandan Lake Edward. This was subsequently followed by a more extensive IDEAL Lake Edward shallow seismic survey in 2003, the results of which appeared in a series of publications (Laerdal, 2000; McGlue, Scholz, Karp, Ongodia, & Lezzar, 2006; Russell & Johnson, 2005; Russell, Johnson, Kelts, Laerdal, & Talbot, 2003). High-resolution 14C dating 4 C.J.Nicholas et al. 5 1 0 2 st u g u A 9 1 7 0 4: 0 at ] L e R I & y r a r b Li e g e oll C y nit ri T e/ d ói n o rí T a n e st ái ol h C n n a arl Figure2. LithostratigraphicschemesproducedinstudiesoftheLakeEdward–LakeGeorgebasinbetween h b 1960and2000.Avarietyofinformalformationswererecognisedatindividualexposuresacrossthearea, a e ranginginagefromLateMiocenetoHolocene. L [ y b d e of plant material within short cored intervals of lake sediment, combined with sequence strati- d oa graphic interpretation of seismic sections, led to a new chrono- and event-stratigraphy for the wnl easternLakeEdwardareaandtherecognitionofaneotectonicframework.Thisnewstratigraphic Do scheme presented a much younger chrono-stratigraphy for eastern Lake Edward than those studies undertaken prior to 2000, mainly due to the accuracy of absolute 14C age dating (Russell & Johnson, 2005; Russell et al., 2003). Identification of arid intervals, wet phases, andtheircorrelationwithanoscillatingEastAfricanclimateduringglacial/interglacialepisodes indicate that the near-surface sediments of present-day Lake Edward are latest Pleistocene– Holocene in age ((cid:4)20,000 years BP to the present day) (Laerdal, 2000; McGlue et al., 2006; Russell& Johnson,2005; Russell et al., 2003) (Figure 3). Journal of Maps 5 5 1 0 2 st u g u A 9 1 7 0 4: 0 at ] L e R I & y r a r b Li e g e oll C y nit ri T e/ d ói n o rí T a n e st ái ol h C n n a rl a h b a e L [ y b d e d a o nl w o D Figure 3. Stratigraphic summary chart of shallow cored sediments in eastern Lake Edward based on IDEALstudies,correlatedagainstagedatesfromKatweandBunyarugurutuffsandpreviousageattempts onsporadicoccurrencesofhotspringtufalimestoneswithinthebasin.Colouredstarsindicatewhereunits identifiedinFigure2wouldplotinthisstratigraphicscheme.AlthoughtheLakeEdwardcoredsedimentsare likelytobeyoungerthanmostonshoreexposures,theKikyereswampclaysmaycorrespondtooneormore ofthewetphasesindicatedduringtheHolocene. 6 C.J.Nicholas et al. UponsigningaProductionSharingAgreement forEA4Bwith theUgandanGovernment in 2007, Dominion Uganda Ltd. (a subsidiary of Dominion Petroleum Ltd.) began an extensive geology and geophysics field research programme in collaboration with Trinity College Dublin andPEPD.Oneoftheprincipalresultsofthisnewphaseofgeologicalfieldsurveyingwasageo- logical map sheet of onshore EA4B, presented here at 1:66,666 scale (1.5cm¼ 1km) (Main 5 Map). This map not only subdivides the stratigraphy of rift-fill sediments for the first time but 01 also identifies relatively recent normal faults which have controlled the style and architecture 2 st of sedimentation in the basin. In addition, as this map records the first petroleum exploration u in this basin, it is also something of a historical document, in that additional features, such as g u A the course of the original seismic lines through the bush, have been shown, as well as sites 9 where (cid:4)80mdeepupholesweredrilledtohelp processtheseismic data, andwherelithofacies 1 7 were recorded in samples collected from the bottom of drilled 9m shot holes (Main Map). 0 4: 0 at ] 2. Mappingstrategy and attempts to establish thestratigraphy in EA4B L e R The Dominion field programme began with an initial field reconnaissance of the EA4 Lakes I & EdwardandGeorgearea,undertakeninSeptember2007.Aspartofthisreconnaissance,attempts y were made to revisit the key localities at which informal formations had been described pre- r a br viously by Musisi (1991) and Byakagaba (1997). However, problems were encountered not Li onlyinrelocatingsomeofthese‘type’localitiesbutalsoinmakinganykindoflithostratigraphic e g correlationbetweenisolated,short,loggedsections.Inessence,thecruxoftheproblemlayinthe e oll factthatexposedrift-fillsedimentsinEA4Bwereoriginallydepositedindominantlyfluvialand/ C y or alluvial depositional environments by north-westerly prograding complexes from the SE rift nit margin. As such, a change in sediment can be expected to occur over very short distances both Tri laterallyanddistallydown-dipfromsourcearea,aswellasestablishingcomplexstackingpatterns e/ in what canbe thin stratigraphic intervals of only a few metres in thickness. d ói In order to circumvent this problem, yet continue mapping until the stratigraphy could be n río resolved,alithofaciesapproachwasadopted.Usingthismethod,ateachexposure,thecomponent T a lithofacies were identified and described separately, followed by an interpretation of their litho- e n facies associations and depositional environments. In the field, five broad lithofacies types st were recognised (‘Lithofacies A–E’; see ‘Key to Intraformational Lithofacies’ on Main Map), ái ol each associated with a particular suite of depositional environments related to factors such as h C water energy and distance from the south-eastern rift margin scarp. Each characteristic nn lithofacies was assigned a specific colour code for the map representing each of the five main a rl lithofacies(see‘KeytoIntraformationalLithofacies’onMainMap),andallindividualexposures a h visitedareshowndelineatedandcoloured,accompaniedbytheoriginallocalitycodewhichcor- b a e respondstoitsdescriptioninfieldnotebooks.Exposureofrift-fillsedimentsinEA4Bisgenerally L [ poorcomparedwith,forinstance,equivalentareasoftheAlbertineRiftinUgandanLakeAlbert y b to the north. Thus, many of the smaller exposures simply consisted of one lithofacies type. d e However, larger cliff exposures, most of which have been visited by previous authors and d a o logged as formation ‘stratotypes’, were a combination of stacked lithofacies. This approach nl w proved to be far more flexible and realistic in describing precisely what was observed in the o D field. Eventually, multiple sampling from hand augering along road sections meant that each lithofaciescouldbesubdividedfurther(forinstance,DI,DII,DIII),toreflectmorespecificdiffer- ences in local depositional environments leading to sedimentation of beds with roughly similar physical properties, and this further division is shown in the ‘Key to Road Section Lithologies’ which accompanies the road section correlation panel at the bottom of the Main Map. Samples fromclayintervalsoflithofacies‘D’and‘E’weretakenfromavarietyofstratigraphicintervals androadsections, andprocessedbyRPSLtd.(UK) following themethod ofShaw,Logan,and Journal of Maps 7 Weston (2008) for their palynological assemblages. Biostratigraphic analysis of these samples indicated that all exposures were Pleistocene–Holocene in age (i.e. ≤2.59Ma). Correlation with the IDEAL 14C results from shallow lake-floor sediments just offshore on Lake Edward (Figure 3) would make this age range more likely to be Middle or Late Pleistocene–Holocene (i.e. ≤0.78Ma). However, without further age constraints, a more detailed synthesis with 5 shallow lake bottom data remains problematic. 01 Whilstgeologicalsurveyingwasinprogress,Dominionconductedbothanonshoreandoff- 2 st shore deep seismic reflection survey within EA4B. As a necessary precursor to shooting the u seismic lines, shot holes for dynamite were drilled along lines at 25m intervals, and up to a g u A depth of about 9m. The drillers were asked to take a sediment sample from the base of each 9 hole.Samplecoveragewasunfortunatelynotcomplete,butyieldedfurtherinformationonlitho- 1 7 faciesvariationacrosstheareainplaceswherenoexposurewaspresent.Despitetheseadditional 0 4: data,attheendofthisprocess,therecognitionofformationalstratigraphicunitswasstillunclear, 0 at withnodistinctpatternemergingfromlithofaciesdistributionsacrossthearea.However,whatthe L] exposure and shot hole lithofacies data amply demonstrated by their colour distribution on the e R accompanyinggeologicalmapwasthatthecommentabove,regardingrapidfluvialandalluvial I & lithofacies changes both laterally and over time, holds true for the rift-fill sediments in Lake y Edward. In essence, this lithofacies approach views the stratigraphy at too high a resolution r a br perhaps for distinct sedimentary packages to become clear. Consequently, following on from Li this lithofacies mapping, a slightly different approach was needed. This involved attempting e g lateral correlation across the area by bundling characteristic lithofacies packages together with e oll the recognition or inference of under- and overlying significant bounding surfaces. This takes C y the original goal of mapping lithostratigraphic formations towards the establishment of a nit sequence stratigraphic framework of distinct chronostratigraphic sedimentary packages or Tri units, separated by timesurfaces. e/ ThroughouttheEA4Bregion,ageneralobservationwhichholdstrueisthattheriftsediment d ói strataareinalmostallcasesveryshallowlydippingtothehorizontal(≤108),reflectingtheirorig- n río inal angle of repose during deposition in their respective fluvial environments. Making the T a assumption that over large lateral distances, localised differences in angles of dip and direction e n broadlycanceleachotherouttogiveanapproximately horizontal regionalangleofdiptosedi- st mentary packages means that altitude (height above sea level in metres) would correspond ái ol directly to true stratigraphic height in a sedimentary section. In reality, of course, where strata h C are dipping gently,this equivalence of difference in altitudes to true stratigraphic thickness will nn be a slight over-exaggeration. Consequently, key road or track sections were identified which a rl traversed areas of stratigraphic complexity and with the maximum altitude variation. A hand a h auger was used to drill below the soil along stretches of no exposure and recover sediment b a e samples, generally at 20 m vertical intervals along each section, and incorporating any data L [ from roadside exposures. Using the assumption of horizontal dip gives a maximum thickness y b in each road section equivalent to the difference in altitude. Despite the obvious shortcomings d e of these assumptions, it should be remembered that there are very few thick exposures in d a o EA4B, and that short of continuous coring in boreholes, this method remains the best way to nl w stacktogetherorcorrelatebetweenroadsectionsandfinallybuildalithostratigraphicframework o D for EA4B. This framework is shownacross the lower half of the Main Map. 3. Coupled structural and stratigraphic framework in EA4B Seismic sections combined with aero-gravity and magnetic data demonstrate that the Lake Edwardbasinisahighlyasymmetricgrabenorasymmetrichalf-graben,withthemajorcontrol- ling fault bounding the western edge of the lake in the DRC. It became clear early on in the 8 C.J.Nicholas et al. geologicalmappingthatthesurfacetopographyofEA4Breflectedthisunderlyingstructureofthe region. Linear or slightly sinuous, low, asymmetric hills are common with pronounced scarp slopes along one side and a flat, gentle dip-slope on the other. The base of scarp slopes thus mark the surface expression of underlying faults and the scarp face approximates to the fault plane. This interpretation was confirmed by observing fault drag on beds exposed along scarps 5 at some localities. Given that all exposed sediments in EA4B are Mid–Late Pleistocene– 01 Holoceneinage,faultswhichhaveproducedascarpatthesurfacemusthavehadcomparatively 2 st recent movementwithin tens of thousands of years. u Mapping revealed the presence of a (cid:4)7km wideNE–SW trending, east-facing extensional g u A fault zone exposed across the centre of onshore EA4B. Within this zone, particularly well- 9 developed exposed fault geometries are seen at Kazinga, Kikarara, and Kiruruma. The major 1 7 faults seen at Kikarara and Kiruruma trend towards each other and the likelihood is that they 0 4: are in fact the same fault which has had its scarp breached and eroded in the centre by the 0 at Ntungu River. This scarp-breaching appears common in many of the faults seen in EA4B and L] is also accompanied by recurved, eroded fault scarp surface terminations. Some minor west- e R facing antithetic reverse faults can be seen, and these have a curving trace at the surface which I & bifurcates from the main fault scarp. This is particularly well developed along the Kazinga y Fault at Katookye Gate (UTM 802300 9935300 on the Main Map). Here, where antithetic r a br faultsjointhemainfaultateitherend,theareainthecentrecanbeobservedtobeanupstanding Li hillorraised‘pod’onthefootwallcrestwithanaccompanyingpronounced‘shoulder’onthedip- e g slope.Suchraised‘pop’-uphillsinthecentralareaoffootwallblocksonmajorfaultssuggestthat e oll the neotectonic phase in EA4B iscompressional or transpressional. C y ThepresenceoftheKazinga–Kikarara–KirurumaFaultZone(KKK-FZ),andthecompara- nit tivelyrecentextensionalmovementacrossit,hashadaprofoundeffectonthesedimentaryrift-fill Tri ofthisportionoftheLakeEdwardbasin.Thelow-energy,distalfluvialfloodbasinplainandmar- e/ ginallacustrinedepositsofthecomparativelyolderKisenyiFormationformthefootwallofthese d ói majorfaultsandisexposedintheirscarps.However,tothesouth-eastofthisfaultzone,thesedi- n río mentary fill has altogether a different style and architecture, reflecting rapid deposition from T a higherenergyfluvial and alluvial systems. Although these sediments are clearly more proximal e n to the rift margin, and therefore would be expected to be coarser clastics, the change from st theseintotheKisenyiFormationisabruptacrossthefaultzoneandnotgradational.Thissuggests ái ol thatextensionalmovementalongtheKKK-FZhascreatedasub-basin,ortrough,trendingNE– h C SWthroughthedistrictofBwambaraandalongthesouth-easternriftmargin(henceitisreferred nn tohereasthe‘Bwambaratrough’)(Figure1).Thishassubsequentlybecomefilledwithrelatively a rl immature clastics shed directly from the exposed basement to the south-east. Seismic interpret- a h ation indicates that there is a gently dipping shallow unconformable surface present which b a e separates a gently dipping near-surface sedimentary package from underlying tilted fault block L [ reflectors in the onshore part of the Lake Edward basin. This surface marks the onset of y b KKK-FZ extension and subsidence of the Bwambara trough, and as a significant bounding d e surface,itseparates the Kisenyi Formation from overlying troughfill (Figure 4). d a o TheBwambaratroughcontainsthreerecognisablestratigraphicunits.Thelowermostofthese, nl w the Kayonza Formation, is a fluvial unit with well-developed channel conglomerate lenses and o D thick, laterally persistent clay intervals. These clays represent formation of extensive ponded floodbasinswampsdevelopedduringinitialsubsidenceinthetroughasreorganisationofdrainage pooledriversystemstothesouth-eastofKKK-FZscarps.Inaddition,itseemslikelythatthisalso representsaperiodof deposition duringawetclimatic period,suchaswouldbeexperienced in the equatorial tropics of East Africa during northern and southern hemisphere interglacials (for instance, see Trauth, Deino, & Strecker, 2001) (Figure 4(b)). In Section IXa, at the Kiruruma River Quarry close to the base of the Kayonza unit, a thin, (cid:4)15cm thick, hard, white ‘clay’