structureandgeologicalhistoryof thecongobasin ... · structureandgeologicalhistoryof thecongobasin:...

Structure and geological historyof the Congo Basin:an integrated interpretation of gravity, magnetic andreflection seismic dataE. Kadima,n D. Delvaux,w S. N. Sebagenzi,n L. Tackw and S. M. KabeyaznService de Ge¤ ophysiqueApplique¤ e, University of Lubumbashi, DR, CongowRoyalMuseum for CentralAfrica,Tervuren, BelgiumzCNE, Kinshasa, DR, Congo

ABSTRACT

The stratigraphic, paleogeographic and tectonic evolution of the intracratonic Congo Basin inCentral Africa has been revised on the basis of an integrated interpretation of gravity, magnetic andre£ection seismic data, together with a literature review of papers sometimes old and di⁄cult toaccess, map compilation and partial reexamination of outcrop and core samples stored in the RoyalMuseum for Central Africa (RMCA).The Congo Basin has a long and complex evolution starting inthe Neoproterozoic and governed by the interplay of tectonic and climatic factors, in a variety ofdepositional environments.This multidisciplinary study involving 2Dgravity and magnetic modelingas additional constraints for the interpretation of seismic pro¢les appears to be a powerful tool toinvestigate sedimentary basins where seismic data alone may be di⁄cult to interpret.The tectonicdeformations detected in the Congo Basin after the1970^1984 hydrocarbon exploration campaign inthe Democratic Republic of Congo (DRC) have been attributed to crustal contraction and basementuplift at the center of the basin, following a transpressional inversion of earlier graben structures.Two-dimensional gravity and magnetic models run along key seismic lines suggest the presence ofevaporite sequences in some of the deeper units of the stratigraphic succession, in the lateralcontinuity with those observed in theMbandaka andGilson exploration wells.The poorly de¢nedseismic facies that led to the previous basement uplift interpretation of the crystalline basement isshown to correspond to salt-rich formations that have been tectonically de-stabilized.These featuresmay be related to vertical salt-tectonics connected to the near/far- ¢eld e¡ects of the late Pan-Africanand the Permo-Triassic compressive tectonic events that a¡ected this African part of Gondwana.

INTRODUCTION

Cuvette Centrale (Central Basin), Cuvette Congolaise,Zaire Interior Basin and Congo Basin are all terms usedin the literature to denominate the broad and long-livedintracratonic depression of Central Africawhere sedimentaccumulation, tectonic inversion and erosion occurred in along history since the Neoproterozoic. This basin is lo-cated in the centre of the African Plate, covering most oftheDemocratic Republic of Congo (DRC, formerly Zaire),the People’s Republic of Congo and the Central AfricanRepublic (Veatch, 1935; Cahen, 1954; Lepersonne, 1977;Daly et al., 1992; Giresse, 2005).

The major problem in depicting the geological historyof this broad structure is that outcrops, wells and seismiclines provide a picture too scattered and with a poorly

de¢ned time frame to interpret the past geological historyof this huge area as that of a single and coherent basin.De-pending on the location and time of deposition, the Neo-proterozoic^Early Phanerozoic sediments might berelated to failed rift basin, foreland basins, or platform cov-er as in theTaoudeni basin over theWest AfricanCraton ortheRiphean-Vendian over theSiberianPlatform. Since theLate Carboniferous and up to theTriassic, they are mainlyrelated to the Karoo period in the Gondwana continent,with glaciations, rifting and tectonic inversions (de Wit &Ransome, 1992; Burke et al., 2003; Catuneanu et al., 2005).During theCretaceous andCenozoic, a succession of con-tinental deposits separated by peneplanation surfaces andtotaling a maximum thickness of 1000m emplaced in aslow subsidence process (Giresse, 2005). Since the Ceno-zoic, the basin acquired the form of an internal sag basin,from which the term ‘Cuvette Centrale’ has been derivedby the ¢rst explorers. It de¢nes the hydrographic basin ofthe Congo River and is surrounded by topographic highsinterpreted as swells (Burke et al., 2003; Burke & Gunnell,

EAGE

Correspondence: D. Delvaux, Royal Museum for Central Africa,B-3080 Tervuren, Belgium. E-mail: [email protected]

BasinResearch (2011) 23, 499–527, doi: 10.1111/j.1365-2117.2011.00500.x

r 2011Muse¤ e royal de l’Afrique centrale / Koninklijk Museum voorMidden-AfricaBasin Researchr 2011Blackwell Publishing Ltd, European Association of Geoscientists & Engineers and International Association of Sedimentologists 499

mailto:[email protected]

mailto:[email protected]

2008); successively the Adamlia, Nil-Congo, East African,North-Zambian, Bie and Mayombe swells (Fig. 1). Othersag basins also appeared during the same period more tothe south (Kalahari and Okavango basins; Haddon &McCarthy, 2005). During the Miocene, broad £ank upliftof the Atlantic rifted margin closed the basin, stimulatingthe formation of gorges and rapids across the MayombeMountains and maintaining the basin £oor to a present-day altitude of 290m asl (Se¤ ranne & Anka, 2005).

The two most widely used terms to depict this broad,long-lived and multievent intracratonic sedimentary de-pression are ‘Congo Basin’ and ‘Cuvette centrale’. Bothterms appeared almost simultaneously. As reviewed byCornet (1894), A.Wauters recognized in1885 that the Con-go River and its tributaries were forming an hydrologicalbasin in the form of a large ‘cuve’ (bowl) with his centralpart systematically lower in altitude than its periphery.This geographic approach was con¢rmed by E. Dupontin 1888 (Dupont, 1889), which ¢rst speci¢ed its geologicalsigni¢cance. Cornet (1894) de¢ned the ‘Congo Basin’ as ageological entity corresponding to all the nonmeta-morphic sedimentary formations overlying the crystallinebasement, di¡erentiating it from the hydrological basin.He rapidly recognized its possible petroleum potential(Cornet,1911).The termCongoBasin became internation-ally used since the memoir of Veatch (1935). In parallel,‘Cuvette centrale’, which refer to the present-day bowlshaped depression, became equally used (Evrard, 1957;Chorowicz etal.,1990;Daly etal.,1991,1992).The geologicalsigni¢cance of these terms also varied with time. Theywere initially de¢ned before the development of the platetectonic concepts and the present understanding of thedynamics of sedimentary basin formation.While Cornet(1894,1911) considered all the sediments covering the crys-talline basement as part of the basin, only the Phanerozoic(Carboniferous toRecent) sedimentswere attributed to thebasin by Veatch (1935), Cahen (1954) and Lepersonne(1974a, 1974b, 1977).These authors attributed the Neopro-terozoic to early Paleozoic sediments overlying the crystal-line basement as part of the platform cover. Integrating theplate tectonic concepts, Chorowicz et al. (1990) and Dalyetal. (1991,1992) reincorporated the pre-Phanerozoic sedi-mentary succession into the Congo Basin concept. As it isoutside the scope of thiswork to revise these terms and de-¢nitions, we will preferably use ‘Congo Basin’, referring tothe entire sedimentary succession above the crystallinebasement butwithout the particular morphological mean-ing associated to the term‘Cuvette centrale’.

Owing to a scarcity of exposure, the geology of the Con-go Basin is poorly known. Although the Congo Basin is acontinent-scale feature (1.2Mkm2, i.e. 10% of the area ofthe Africa continent), recent reviews on sedimentary ba-sins in Africa have sometimes totally ignored it (Selley,1997) or have restricted their focus on its recent historysince the Late Carboniferous times (Lepersonne, 1977) oreven theMesozoic times (Giresse, 2005).

The ¢rst explorationwork of theCongoBasin combinedsurface mapping, geophysical surveys and drilling by the

‘Syndicat pour l’e¤ tude ge¤ ologique et minie' re de la Cuvettecongolaise’ for de¢ning its stratigraphy and structure. In1952^1956, the ‘Socie¤ te¤ de Recherche Minie' re en Afrique’(REMINA) drilled two c. 2000m deep stratigraphic wells,fully cored (Samba and Dekese; Fig. 2).The cores are pre-served in the RoyalMuseum for Central Africa,Tervuren,Belgium, constituting a unique geologic data base.The re-sults of this comprehensive exploratory work have beenpublished among others in Cahen etal. (1959,1960) andEv-rard (1957, 1960).The seismic refraction and re£ection re-sults, calibrated with the Samba and Dekese wells andcorrelated to outcrop data allowed to sketch the geologicalstructure of the basin (Evrard, 1960).The results con¢rmthe earlier subdivision of the basin stratigraphy in threemajor sediment packages, aMesozoic to Recent cover witha maximum thickness of 1200m, Late Carboniferous toPermian sediments with an apparent thickness reaching1000m in the Dekese well (not corrected for the average401 tilting and the numerous slip planes observed on thecore), and earlier sediments several thousands of meterthick.

From1970 to1984, a second phase of oil exploration con-ducted by SHELL,TEXACO and the Japan National OilCompany (JNOC) led to the acquisition of additionalaeromagnetic, gravity data and 2900 km of seismic re£ec-tion pro¢les, and geochemical samples (JNOC, 1984).These campaigns were followed by the drilling of two c.4500m deep exploration wells (Mbandaka and Gilson) byESSO Zaire. A ¢rst compilation of the available data wasconducted by Exploration Consultant Limited (ECL,1988) and the resulting interpretation of the basin evolu-tionwas outlined byLawrence &Makazu (1988).The aver-age thickness of the sediments above the crystallinebasement in the central part of the basin is now estimatedat 5^6 km, locally reaching 8^9 km.The tectonic evolutionmodel was re¢ned byDaly et al. (1991, 1992) who suggestedthat the basin was initiated in a failed rift setting and thatthe tectonic features observed on the available seismiclines are related to the Early and Late Paleozoic^EarlyMe-sozoic compressional events, reactivating the rift struc-tures. They proposed that the observed geometries canbe explained by a process involving contraction and upris-ing of the lithospheric crust as far- ¢eld e¡ects of the Capefolding event in South Africa.The stratigraphy, structureand petroleum potential of the Congo Basin have been re-cently reviewed byKadima (2007).

The structure of the Congo Basin has been interpretedby ECL (1988), Lawrence &Makazu (1988), and Daly et al.(1992) as a series of sub-basins separated by basementhighs (Fig. 2). In the center of the basin, the narrowWNW^ESEpresumedKiri high is shown to be connectedto the broadLokoniaHigh.Following our integrated inter-pretation of gravity, magnetic and re£ection seismic data,we will present an alternative interpretation with the pre-sumed crystalline basement of the Kiri High is insteadcomposed of salt-rich sediments.

After an updated description of the geological sett-ing (including the well data) and a review of tectono-

r 2011Muse¤ e royal de l’Afrique centrale / Koninklijk Museum voorMidden-AfricaBasin Researchr 2011Blackwell Publishing Ltd, European Association of Geoscientists & Engineers and International Association of Sedimentologists500

E. Kadimaet al.

stratigraphic constraints, this paper proposes a new inter-pretation of the structure and deformations of the CongoBasin using a multidisciplinary approach integrating well,outcrop and geophysical data with modeling results in or-der to clarify the structure and deformations of the CongoBasin.

GEOLOGICAL SETTING

The Congo Basin covers about 1.2� 106 km2 and is one ofthe largest intracratonic basins in the world. It is under-lain by a thick (200 � 30 km) lithosphere and coincideswith a region of pronounced long-wavelength gravityanomaly (Crosby et al., 2010). Cenozoic to Recent sedi-ments occupy the center of the basin while the larger partof it contains Mesozoic sediments. The Paleozoic andNeoproterozoic sediments outcrop at the periphery ofthe basin and have been explored by drilling in the centralpart (Fig. 2). They overly a metamorphic to crystallinebasement that was initially considered as forming a largecratonic mass: the Congo craton (Cahen, 1963, 1982). Con-tinental-scale tomography however indicates no pro-nounced lithospheric keel under the central part of thebasin, in contrast with thick lithosphere beneath the ex-posed Archean surrounding blocks (Pasyanos & Nyblade,2007). The Congo craton is now considered as composed

of several Archean nuclei (Angola-Kasai, NE Congo,Chailu/Gabon, Bangweulu) that are welded together as aresult of Paleoproterozoic collision orogeny (De Waele etal., 2008; Kanda Nkula et al., 2009). The presence of tec-tonic dislocations a¡ecting the sediments of the CongoBasin (Daly et al., 1991, 1992; this paper) and its currentseismic activity (Ayele, 2002; Delvaux & Barth, 2010) showthat the crystalline basement does not behave fully as a ri-gid cratonic mass.

The Congo basin appears to evolve in a succession ofdi¡erent geodynamic processes, rather than under a singleprocess operating continuously over 800Ma. Initiation ofthe basin in the Neoproterozoic could be at least partlycaused by intracratonic extension, and the subsequentlong evolution is likely in£uenced by subsidence relatedto the cooling of stretched lithosphere, despite several per-iods of basin inversion. Seismic high shear-velocity struc-tures that extend to 250km depth have been noted beneaththe Congo craton (Ritsema & van Heijst, 2000). TheGRACE satellite free-air gravity ¢eld (Adam, 2002; Cros-by et al., 2010) shows a spectacular long-wavelength low of� 30mGal over most of the Congo Basin and � 40mGalin its center, con¢rming the existence of a large low-den-sity sedimentary in¢ll, at least partly compensated byhigh-density material in the lower crust and/or uppermantle but which, according to Braitenberg & Ebbing(2009) cannot be explained fully by a crustal thinning

0

EasternRift

WesternRift

10 15 20 25 30 35 40Longitude East (°)

10 15 20 25 30 35 40

–10

–5

0

5

10

Lat

itu

de

(°)

–10

–5

5

10

Kinshasa

Kananga

Kisangani

Kindu

Bangui

Mbandaka

0 km 400 km

TanzanianCraton

Congo Basin / Cuvette Centrale

North Zambian Swell

Adamlia Swell

Nil-Congo Swell

East A

frican S

well

Bie Swell

Gulf of G

uineaM

ayumbe Sw

ell

Ind

ian O

cean

Ethi

opia

n Rift

Elevation(m a.s.l.)

Congo River

Ubangi River

Kasai River

Sangha R

.

Lukenie R.

Fig.1. Oro-hydrography of theCongoBasin inCentralAfrica. It forms a huge sub-circular depression encircled by relativelyhigh reliefzones and drained by the Congo River and his tributaries (Gtopo altimetric data color-coded in function of elevation). Location of themajor cities are indicated. Rectangle shows the area covered by Fig. 2.


Structure and geological history of the Congo Basin

mechanism and therefore as due to concealed rift basins.Considering the coincidence of the anomalous topo-graphic depression of the Congo Basin with this large ne-gative free-air gravity anomaly and a large positive upper-mantle shear-wave velocity anomaly, Downey & Gurnis(2009) explain this hugeCenozoic depression by the actionof a downward dynamic force on the lithosphere that theyrelate to a high-density object within the lithosphere.Others relate the observed present-day gravity anomaly toa late phase of basin subsidence in response to a downwel-ling mantle plume (Hartley & Allen, 1994; Forte et al.,2010). Both Downey & Gurnis (2009) and Crosby et al.(2010) place the source of the gravity anomaly below thebase of the lithosphere, the ¢rst suggesting the anomalousmass could be made of eclogite, the second that it could begenerated by a convective drawdown caused by chemicaldensity depletion of the lithospheric mantle and relatedto the African Superswell of Nyblade & Robinson (1994).Recently, Kadima et al. (2010, 2011) mapped a NW^SEelongated positive and narrow residual gravity anomaly re-lated to a post-rift event before basin subsidence, alignedwith the Sembe-Ouesso and the Mbuji-Maji rift se-quences that outcrop respectively to the NW and the SE

of the Congo basin.This supports the Neoproterozoic riftorigin of the basin as earlier postulated byDaly etal. (1992).

STRATIGRAPHIC EVOLUTION

As the Congo Basin lies for its larger partwithin theDRC,the knowledge of its stratigraphy also depended on theprogressive geological exploration conducted in thiscountry. Before the ¢rst subsurface exploration periodwith the geophysical acquisitions and drillings, its strati-graphy was inferred from the succession of formationsoutcropping at its periphery, assuming that theywere con-tinuous over large distances (Veatch,1935;Cahen,1954;Ca-hen & Lepersonne, 1954). The sedimentary succession ofthe area was ¢rst recognized by two stratigraphic wells(Samba and Dekese wells) drilled in 1955 and 1956 by RE-MINA that terminated in red arkoses of LateNeoprotero-zoic to Early Paleozoic age.The litho-stratigraphic scale ofCahen et al. (1959, 1960) was later re¢ned by Lepersonne(1974a, 1974b, 1977), taking into account the results of var-ious studies on the Samba andDekese drill holes and out-crop data.

30252015Longitude East (°)

30252015

–5

0

5

Lat

itu

de

(°)

–5

0

5

0 km 400 km

Seismic lines

Faults

Boreholes

Archean

Paleo- & Meso-Proterozoic

Tonian

Ediacaran-Cryogenian

Early Paleozoic

Carb.-Perm.

Triassic

Cretaceous

Jurassic

Tertiary

Quaternary

Fig. 2. Simpli¢ed geological map of the Congo Basin, compiled fromvarious published map.The stratigraphic units have beenassembled into major sequences (LateNeoproterozoic, Paleozoic andCenozoic), corresponding respectively to seismic units A, B andCas de¢ned here from the seismic andwell data. Bold labels de¢ne the Archean cratonic blocs, the stratigraphic Groups and the internalstructural element of the Congo Basin. Deep faults interpreted from the seismic pro¢les are shown as dashed lines. Faults that haveevidence for activity during thePaleozoic-Mesozoic are shown as continuous lines.Location of the 4wells: (S: Samba ^TDat 2038m,D:Dekese ^TD at1826m,M:Mbandaka ^TD at 4350m,G: Gilson ^TD at 4536m), and the seismic pro¢les acquired between1970 and1984 by Shell,Texaco and JNOC.Major morphotectonic provinces are indicated. Luk: Lukuga coal-bearing basin.


E. Kadimaet al.

Two exploration wells (Mbandaka and Gilson) were la-ter drilled in 1981 by ESSO Zaire, and bottomed in theNeoproterozoic (ESSO Za|« re SARL, 1981a, 1981b). Com-bining surface geology with well information, Daly et al.(1992) presented a single composite stratigraphic columnin order to illustrate their views of the tectono-strati-graphic evolution of the basin. This column is howevertoo generalized to be useful here as it interpolates variousstratigraphic units which are discontinuously observedwithin the c. 1200 km wide Congo Basin and which areloosely correlated. In addition, time control is largely lack-ing for the pre-Karoo deposits and some units have re-stricted lateral extension, lateral facies change or may bemissing in place.

An updated stratigraphy of the Congo Basin is pre-sented hereafter, integrating the knowledge gained fromthe 4 wells and presenting possible correlations with out-crop data (Fig. 3a and b). As limited seismic and well dataare available for understanding the stratigraphic and tec-tonic evolution of the deeper part of the Congo Basin, alarge part of the knowledge of the Neoproterozoic andPhanerozoic has been gained from the study of outcrop-ping units. The Late Paleozoic to Recent sediments arebetter represented in the wells and paleontologicallydated, making the correlations with the outcrops moresecure.

Neoproterozoic to early Paleozoic

The stratigraphy of the Neoproterozoic to Early Paleozoicsediments which have been considered in the past to formthe basement of the Congo Basin, is complex, sometimespoorly known andwith few age constraints.The sedimen-tary successions vary from one region to another, as-sembled in Groups and Supergroups (Fig. 3a) ascompiled byCahen (1954);Lepersonne (1974a, b,1977),Al-varez (1995), Cailteux et al. (2005).

The Neoproterozoic to early Paleozoic sediments out-crop in several large regions surrounding theCongoBasin,each with its own characteristics (Cahen, 1954; Leper-sonne, 1974a, b, 1977; Alvarez, 1995 and Cailteux et al.(2005). For the Cryogenian, Ediacaran and early Paleozoicperiods, they are assembled in three major regionally dis-tinct Groups (Figs 2 and 3a, c):

� to the SW, the West Congolian Group as part of theWest Congo Supergroup in the West-Congo orogenicbelt, resting over Paleoproterozoic to Archean and theNoqui granite dated at 999 � 7Ma (Tack et al., 2001;Frimmel et al., 2006).

� to theNandNE, the mostly tabularLindi Supergroup(Verbeek, 1970;Thibaut, 1983);

� to the SE, theKatanga Supergroup in theLu¢lian arc/Katangan orogenic belt (Robert, 1956; Franc� ois, 1974;Cailteux et al., 2005; Batumike et al., 2006), restingabove a Paleoproterozoic magmatic arc sequence andthe Nchanga granite intrusion dated at 883 � 10Ma(Master et al., 2005) and extending up to �535Ma(John et al., 2004; Armstrong et al., 2005).

In addition, sediments of the Itombwe Supergroup in theKivu region in eastern DRC are considered as time-equi-valent to the former groups (Cahen et al., 1979;Walemba &Master, 2005).

Older low-grade to nonmetamorphic poorly dated sedi-ments of pre-Sturtian age and attributed to the EarlyCryogenian T̂onian are assembled into:

� theMayumbian and Zadinian Groups inWest-Congo(lower part of theWest Congo Supergroup), composedof volcanic and volcaniclastic, and dated respectively910^920Ma and 920^1000Ma (Frimmel et al., 2006);

� theMbuji-Mayi (fomer Bushimay) Supergroup in theKasai, southeast of the Congo Basin, preliminary esti-mated at 900^1050Ma (Raucq, 1957; Cahen et al., 1984)but whose time-equivalence with the Roan Group inKatanga cannot be excluded (Cahen, 1970 and Ber-trand-Sarfati, 1972);

� The poorly de¢ned Liki-Bembian Supergroup withdominantly siliciclastics and basic volcanics, outcrop-ing on the north-eastern side of the Congo basin (Figs2 and 3c) in the Liki-Bembe basin (Lepersonne,1974b;Poidevin, 1985) and in the Sembe-Ouesso basin alongthe Sangha andDja rivers (Ge¤ rard,1958;Vicat & Vellu-tini, 1987; Poidevin, 1985).

TheWest-Congo andKatanga Supergroups have been re-spectively involved in theWest Congo and Lu¢lian (or Ka-tangan/Zambezi) orogenic accretionary belts at thewestern and south-eastern margin of the Congo Basin(WestCongo belt: Alvarez etal.,1995;Tack etal., 2001;Kan-da-Nkula et al., 2004; Frimmel et al., 2006; Pedrosa-Soareset al., 2008; Lu¢lian arc: Porada & Berhorst, 2000; Arm-strong et al., 2005; Cailteux et al., 2005; Rainaud et al.,2005). In their foreland, these deposits remained relativelytabular but their uppermost sequences composed of con-tinental red beds (Inkisi in theWest-CongolianGroup andPlateaux/Biano in the Katanga Supergroup, also referredto as the Redbeds), overlie unconformably the earlierfolded sequences (Tack et al., 2008). At both locations, thisunconformity likely corresponds to the paroxysm of thePan-African orogeny, estimated at �550Ma from the co-incidence of the paleomagnetic poles of the di¡erentblocks composing Central Gondwana (Tohver et al., 2006).The timing of the Lu¢lian orogeny in Katanga has beenrecently constrained as spanning �600 to 512Ma (Arm-strong et al., 2005), with a possible ¢nal stage of collisionin the Lu¢lian arc at �530Ma (John et al., 2004; Rainaudet al., 2005). These Redbeds sequences are therefore postNeoproterozic in age, and can be bracketed between theEarly to Middle Paleozoic.They are overlain by the KarooGroup with its base (a glacial diamictite) attributed to the( �320Ma) Late Carboniferous (Cahen et al., 1960) on thebasis of spore determination and correlation with similardeposits of the Lukuga basin in the DRC.We do not con-sider here that the presence of acritarchs that have been at-tributed to the EarlyCambrian in the sequence underlyingthe Redbeds in the Mbandaka well (Daly et al. 1992) is



(a)

(b)


E. Kadimaet al.

convincing enough to suggest an Ordovician to Devonianage for the Redbeds, but this cannot be excluded.

Away from the deformed West-Congolian and the Ka-tangan foreland regions, and in the basin margin (Lindiregion), the Redbeds are seen overlying in an apparentconcordance the subtabular older sequences (Lepersonne,1974a, b;Tack et al., 2008).Therefore, the Pan-African un-conformity can be di⁄cult to identify.

The tabular forelands of both Pan African belts extendfrom the outer rim towards the inner part of theCongoBa-sin and, like for the Lindi Supergroup, dip gently under-neath the Karoo to Recent sequences over long distances.

These deposits thus display a centripetal structural andsedimentary polarity. Attempts at lithostratigraphic corre-lations have been ongoing for decades, mainly based on si-milarities in lithology (e.g. the glaciogenic diamictites; e.g.Cahen, 1963, 1982). However, age di¡erences of c. 200Mymay exist between similar rock types from di¡erent re-gions (e.g. the association of stromatolites and silici¢edoolitic rocks in the Katanga- and West Congo Groups;Kanda-Nkula et al., 2004).This fact indicates that similardepositional paleo-environments may have existed dia-chronously during the Neoproterozoic in various placesof the Congo Basin. Therefore, strict lithostratigraphic

353025201510

353025201510(c)

–10

–5

0

5

Lat

itu

de

(°)

–10

–5

0

5

S

D

M

G

0 km 400 km

Seismic faults

Boreholes

Pan-Africanfault lines

Archeancores

Paleo-proterozoic

Meso-protorozoic

Tonian

Ediacaran -Cryogenian

EarlyPaleozoic

Carb.-Recentsediments

Chailu-GabonBlock

North East Congo Block

Kasai Block

Lindian

Belt

West Congo

Oubanguides Belt

UbendeBelt

UgandaBlock

AngolaBlock

TanzaniaBlock

BangweuluBlock

East-A

fricanM

ozam

biq

ue

Belt

Irum

ide

Belt

Kibar

aBel

t

Karag

we

- Ankol

ebe

ltKarag

we

- Ank

ole

belt

Sang

haAul

acog

en

ArcRoan Basin

Concealedfailed

rift

basin

Fig. 3. continued

Fig. 3. (a) Lithostratigraphic synthesis for theNeoproterozoic^EarlyPaleozoic period.Compiled after various authors (see text).Dottedlines betweenBanalia, Alolo andGalamboge formations: stratigraphic transition by recurrences. (b) Lithostratigraphic columns for theCongo Basin established using data from the 4 wells in the central part of the basin and outcrops on its NE margin (Lindi-Ubangi andKisangani^Kindu region), comparedwith theWest-Congo andKatanga stratigraphy. Compiled after various authors (see text).De¢nition of the seismostratigraphic units and unconformities. (c) Tectonic setting of the Neoproterozoic basins of present-dayCentral Africa, compiled from the1 : 2M geological map of the Zaire (Lepersonne, 1974a) and the1 : 4MmapGeology andMajor OreDeposits of Africa (Milesi et al., 2006).The highlighted continental mass, often surrounded by deformation zones, corresponds to thelarge Congo craton, formed byArchean cores welded together by Paleoproterozoic belts.To be complete, it should include also the Sao-Francisco in block in Brazil, now drifted apart by the opening of the Atlantic.The Karagwe-Ankole and Kibara belts represent aMesoproterozoic intracratonic tectono-magmatic event (Tack et al., 2010). In the Lu¢lian Arc, fragments of the RoanGroup (earlyCryogenian) have been involved in the Late Pan-African deformation and could be eventually be time-equivalent to the upper part oftheMbuji-Maji Supergroup. Autochtonous remnants of the RoanGroup (Roan Basin) are located in the internal part of the Lu¢lianArc.



correlations among the three Supergroups will only bepossible, provided enough well-constrained age determi-nations of critical rock types have been obtained in eachseparate lithostratigraphic column.

Despite these limitations, a tentative long-distance cor-relation and age calibration is proposed inFig. 3a, based onthe recognition of a series of major tectono-stratigraphicevents that are supposedly coeval in time for the three ma-jor sedimentary sequences outcropping along the rim ofthe Congo Basin: i.e. the West Congolian Group and theLindi and Katanga Supergroups. As it can be hazardousto perform correlations on the basis of lithological faciesonly, the major glacial events have been used instead, ac-cepting the ‘Snowball Earth’ theory following which theCryogenian glaciations are global and thus contempora-neous (Ho¡man & Schrag, 2002). Other authors arguehowever that the glacial deposits are related to tectonicallycontrol uplifted block and are therefore not necessary con-temporary (Eyeles & Januszczak, 2004). In absence of bet-ter age constraints, we tentatively accept that these glacialevents are contemporary at the scale ofCentral Africa.TheCryogenian part of theNeoproterozoic is characterized byseveral major glacial events, with the Sturtian (750^710Ma) and the Marionan (650^635Ma) as the most pro-minent ones (Ho¡man & Schrag, 2002; Ho¡man et al,2004; Ho¡man & Li, 2009).

InWest Congo as in the Katanga region, diamictites as-sociated to the Sturtian and Marionan glaciations havebeen identi¢ed respectively asTillite Infe¤ rieure andGrandConglomerat for the Sturtian event and as Tillite Supe¤ r-ieure andPetit Conglomerat for theMarinoan period. Fol-lowing Master et al. (2005), the Petit Conglomerat and itsoverlying cap carbonate (Calcaire rose)might be correlatedwith the Ghaub diamictite in the Damara orogen, dated at650^635Ma by Ho¡man et al. (2004). In the Lindi Super-group, the Akwokwo tillite also corresponds to a glacialdiamictite (Delpomdor&Tack, 2007). As only one diamic-tite level is known in the Lindian, it can be correlatedeither to the Sturtian or to the Marinoan event. A 730^755Ma depositional age has been estimated by Poidevin(2007) on the basis of Sr-geochronometry, suggesting itcould correspond to the Sturtian. We favour however aMarionan age because, on the basis of detailed ¢eld de-scriptions of Verbeek and the rock archive stored in theRoyalMuseum forCentralAfrica, itwas found that theAk-wokwo tillite appears in direct associationwith typical ‘capcarbonates’ (Tait etal., 2010).This succession of glacial dia-mictites surmounted by pink coloured dolomites (‘cap car-bonates’) is characteristic for the Marionan glaciation(Ho¡man et al., 2004; Shields, 2006).

As discussed above, the tectonic unconformity relatedto the paroxysm of the Pan-African orogeny (550Ma) sepa-rates the uppermost deposits (Inkisi inWest Congo, Aru-wimi within the basin and Plateaux/Biano in the Katanga)from the lower ones.

Using these tectono-stratigraphic events as markers,the West Congolian Group is subdivided into the Sansik-wa, Haut-Shiloango, Schisto-Calcaire^Mpioka and Inkisi

Subgroups (Fig. 3a). Stratigraphically below the Sturtiandiamictites, the Sansikwa Subgroup inWest Congo couldbe equivalent to the Roan Group in Katanga but has noequivalent in the Lindi^Ubangi region.The upper part ofthe Mbuji-Maji Supergroup could eventually also beequivalent to them. Between the Sturtian and the Mario-nan diamicties, the Haut-Shiloango Subgroup of WestCongo is time-equivalent to the Ituri Group in the Lindiregion and to theNgubaGroup inKatanga, but is missingin the Ubangi region. The Sansikwa Subgroup in WestCongo and the Roan Group of Katanga are both cappedby a Sturtian glacial diamictite, respectively the LowerDiamictite/Tillite and the Grand Conglomerat. Above (orstarting with) theMarinoan glacial diamictites, the Schis-to-Calcaire of West Congo correlates with the LokomaGroup of the Lindian and the lower and middle Sub-groups of the Kundelungu Ku1 (Kalue) andKU2 (Kibuo).

Above the Pan-African unconformity, the c. 1000mthick tabular £uvio-deltaic red arkoses (Redbeds) can befound, de¢ned as the Inkisi Subgroup inWest Congo (Al-varez etal.,1995), theAruwimiGroup in theLindi^Ubangiregion (Verbeek, 1970) and the Upper Kundelungu (Ku3)in the foreland of the Katanga belt. For the latter, it hasbeen described as ‘Plateaux’ Subgroup (Lepersonne,1974a, 1974b; Dumont et al., 1997; Dumont & Hanon,2000; Cailteux et al., 2005; Batumike et al., 2006) and re-cently renamed ‘Biano’ Subgroup by Batumike et al.(2007) and Haest et al. (2007). Ar-Ar dating of detriticalmuscovites form this unit by Master et al. (2005) gave amaximum age of 573Ma (terminal Edicarian), supportingthe idea that it has been deposited in the foreland of theLu¢lian orogen. A similar detrical zircon age of 565Mawas obtained for the Inkisi formation of West Congo(Frimmel et al., 2006). The depositional environment ofthe Redbeds are largely continental, lacustrine to £uvio-deltaic semi-arid, with lacustrine to lagunar at the centerof the basin (Alolo shales), in a very large platform-like de-pression spanning thewholeCongoBasin and grading intoforeland basins on its rim.

In theLindi Supergroup, theAruwimi red arkoses over-lie the Lokoma Group with a weak discordance or trans-gressive contact.Within the Aruwimi itself, the �100mthick dark Alolo shales and sandstones represent a lacus-trine to lagoon interval between the continental Galam-boge (�100m thick) and Banalia ( �1000m thick)arkoses of respectively semi-arid £uviatile to aeolian andbraided river facies. As the transition between these for-mations is gradual, it seems justi¢ed to associate thesethree formations to theAruwimiGroup rather than corre-lating only the Banalia red arkoses with the Inkisi and Pla-teaux/Biano units as proposed byTait et al. (2010).

The sedimentary successions in the Liki-Bembe andthe Sembe-Ouesso basins are remarkably similar and havebeen parallelized by Ge¤ rard (1958) and Poidevin (1985).They comprise quartzites, slates and conglomerate at thebase, and calcareous shales and limestones with levels ofblack shales. In the Sembe-Ouesso basin, these series arecapped by the Dja tillite (Cryogenian) and a¡ected by two


E. Kadimaet al.

episodes of dolerite intrusions (Vicat&Vellutini,1987;Poi-devin, 1985).The ¢rst period of intrusion occurred beforethe Dja tillite and corresponds to intraplate magmatismwhile the second one post-dates the tillite and is of oceanica⁄nity. Vicat & Vellutini (1987) interpret this basin as afailed rift, later involved in compressive tectonics.

Evidence for Neoproterozoic evaporites

During the Neoproterozoic, several basins framing theCongo craton and also inside the craton, formed in an ex-tensional (rift) context and developed partly in restrictedcoastal plain, and rift environment.

In theWest-CongolianGroup of Bas-Congo, the Schis-to-Calcaire Subgroup has been deposited in a shallow car-bonate shelf environment rich in algal mats, withstromatolites, limestones and dolomudstones with cherts,oolites and sulphates, (Delpomdor et al., 2008). The pre-sence of early diagenetic sulphate suggests subtidal re-stricted lagoons evolving into a sabhka environment. InSouth Gabon, a shallow subtidal to supartidal evaporiticenvironment is also evidenced for the Schisto-Calcaireby pseudomorphosed relicts of gypsum and anhydrite(Pre¤ at et al., 2010).The carbonates precipitated from sul-phate-rich hypersaline waters and were subsequently do-lomitized and silici¢ed during diagenesis.

In the Mbuji-Mayi basin, similar rocks are found in arift-like setting, with stromatolites and dolomudstoneswith cherts and oolites (Raucq, 1957). Recognition ofabundant pseudomorphosed relicts of sulphates allowsDelpomdor (personal communication) to propose a simi-lar environment as for the Schisto-Calcaire. The carbo-naceous formations of the Mbuji-Maji region contain thelargest variety of stromatolites that can be compared withthe northern border of the Taoudeni basin (Bertrand-Sarfati, 1972).

In theKatangaSupergroup, theRoan carbonated-pelitedominated sediments form the basal Subgroup of the Ka-tanga Supergroup, below the Grand Conglomerat whichrepresents the Sturtian glaciation (750^710Ma). As com-piled by Jackson et al. (2003), several lines of evidence in-cluding pseudomorphs of gypsum and anhydrite, chloriteinclusions in ores and saline springs, point to a sabhka en-vironment.

The stromatolites from the West-Congo and Lindi re-gions are restricted to a small number of species andforms, with a possible correlation of the Lenda limestonesof the IturiGroup (Lindi Supergroup)with the upper partof theMbuji-Maji Supergroup (Bertrand-Sarfati, 1972).

In theCongoBasin itself, well and geophysical evidencealso suggest the presence of deep-lying evaporites in theNeoproterozoic sedimentary section (this paper).

Karoo Group

The Karoo Group in the DRC (Figs 2 and 3b) comprisesthe Lukuga formation and overlying Haute Lueki Forma-tion (Lepersonne, 1977; Cahen & Lepersonne, 1978).

The Lukuga formation is preserved in tectonic grabensrelated to the East African rift, in the Lukuga basin alongthe border of LakeTanganyika (Fourmarier, 1914), in theLuena basin in theUpemba depression ofCentralKatanga(Cambier, 1930), inU- shaped glacial valleys on the easternmargin of the Congo Basin (Boutako¡, 1948) as well as inthe Dekese sub-basin of the Congo Basin (Cahen et al.,1960). Its base displays typical diamictites and varval clayof theGondwana glaciation, equivalent to the ‘DwykaCon-glomerates’diamictites of the Karoo type locality in SouthAfrica, attributed to the Late Carboniferous ( �320Ma).The remaining part of the Lukuga formation is repre-sented by post-glacial claystones and sandstones (Dekesesub-basin) and coal-bearing measures (Lukuga andLuenagrabens), correlated to the Ecca Subgroup of South Africa(Daly et al., 1991; Catuneanu et al., 2005; Johnson et al.,2006).

TheHaute Lueki formation is present along the easternmargin of the Congo Basin, overlying the Lukuga forma-tion with a slight unconformity (Fourmarier, 1914; Leper-sonne, 1977). The base of the Haute Lueki formation hasbeen paleontologically dated as EarlyTrias, equivalent tothe Beaufort Subgroup of South Africa (Johnson et al.,2006).

Jurassic to Cenozoic continental deposits

The formations of the Karoo Group are capped succes-sively by upper Jurassic Stanleyville formation andCretac-eous and Cenozoic continental deposits (Figs 2 and 3b;Lepersonne, 1974a, b, 1977; Giresse, 2005). A stratigraphichiatus exists therefore between the Early Triassic HauteLueki and the Stanleyville formations, encompassing atleast the Early Jurassic (Fig. 3b).

The Stanleyville formation occurs as an alternatingsuccession of bituminous shales and limestones (maxi-mum 200m) deposited in a lacustrine basin in the regionsouth of Kisangani (formerly Stanleyville) and extendingalong the Congo River in the direction of Kindu until 51S(Passau, 1923; Lombard, 1960). It has been traversed over323m in the Samba well where it is dominantly composedof calcareous sandstones and shales, with only thin bitu-minous levels (Cahen et al., 1959). A thin (0^20m thick)lenticular formation of Upper Jurassic, possibly Kimmer-idgian age has been also found in wells underKinshasa and Brazzaville on the western rim of the basin(Egorov &Lombard,1962). In the Sambawell, the Stanley-ville formation has been dated as upper Jurassic (Kimmer-idgian) on the basis of Ostracods (Greko¡, 1957) andfossil ¢shes (De Saint-Seine, 1955). Palynological investi-gations by Stough (1965) have re¢ned its age to Bajocian-Oxfordian.

Sediments from the Cretaceous period are well repre-sented, covering a large part of the Congo Basin (Leper-sonne, 1974a, 1974b). They are exposed in the marginalparts of the present-day ‘Cuvette’ and traversed by all thestratigraphic and exploration wells. They consist of threeformations, successively the Loia, Bokungu and Kwango,



totalling a maximum of 750m and covered by a maximumof 200^300mCenozoic sediments.These Cretaceous-Ter-tiary sediments are mostly continental (Giresse, 2005).Dating on the basis of ostracod (Greco¡, 1957), fossil ¢sh(Casier,1961) and pollen (Boulouard&Calandara,1963) al-lows to attribute the Loia formation to the Aptian^Neoco-mian, the Bokungu formation to the Albian, and theKwango formation to the Upper Cretaceous (Turonian toMaastrichtian, possiblyCenomanian at base).They are ap-proximately equivalent, respectively to the Lukunga, Ma-vuma and Bulu-Zambi formation in the coastal basin(Cahen, 1954).

The Tertiary comprises the Paleogene ‘Gre' s Poly-morphes’ and the Neogene ‘Sables Ocres’ formations (Ca-hen, 1954; Lepersonne, 1977; Giresse, 2005), while theQuaternary is represented by super¢cial deposits.

TECTONIC DEFORMATIONS

The succession of deformation observed on the seismiclines led Lawrence & Makazu (1988) and Daly et al. (1991,1992) to postulate that the initial basin subsidence was in-itiated by a Late Proterozoic failed rift and subsequentthermal relaxation.This rifting phase could be the conse-quence of theRodinia breakup,which subdivided theCon-go craton alongNW^SE andNE^SW fractures (Ho¡man,1999). As postulated by the same authors, the compres-sional inversions could be related to the far- ¢eld e¡ects oftwo tectonic events causing the assembly of theGondwanacontinent: the Early Paleozoic Pan-African event (Kenne-dy, 1964; Kr˛ner et al., 2001), and the Permo-TriassicGondwanide Andean margin type oblique convergingevent at the southern margin of Gondwana (Daly et al.,1991; Visser & Praekelt, 1996; Delvaux, 2001a; Newtonet al., 2006).

Pan-African early rifting stage

In the Congo Basin, a poorly de¢ned early rifting stage issuggested by Lawrence & Makazu (1988) and Daly et al.(1991, 1992) to have controlled the deposition of the 210mthick carbonate-evaporite sequence, before the depositionof the 700m thick clastic sequence of the Schisto-CalcaireSubgroup.

A network of Late Neoproterozoic intracratonic tec-tonic depressions developing within the Congo cratonhave been proposed to control the Neoproterozoic carbo-nate sedimentation in Central Africa (Alvarez, 1995). TheNE-trending Sangha aulacogen which presently controlsthe lower course of the Congo river between Mbandakaand the Atlantic ocean could be a dominant extensionalstructure, linking three belts of NW-trending basins, onecentered on the outcropping region of the Lindian northof Kisangani, one centered on the Upper Sangha river inthe People’s Republic of Congo, the Lukenie river and con-nected to theMbuji-Maji and Roan basins, and one corre-sponding to theWest-Congo belt (Fig. 3c).They might not

be all contemporaneous, but they are believed to controlthe depositional histories in the early stage of developmentof the future Congo Basin.These basins apparently devel-oped as failed rifts (aulacogens) between the Archeanblocks of theCongo craton,without reaching the full ocea-nic stage.They could have been connected to the large riftsystem that possibly separated the Congo craton from theKalahari craton, leading to the formation of a passive con-tinental margin on the southern side of the Congo craton(Porada, 1989; Porada & Berhorst, 2000; Burke et al., 2003;John et al., 2004). In the region of the future Lu¢lian Arc,the WNW-trending Roan rift basin created the adequateenvironment for the deposition of the evaporate-rich car-bonates (Porada & Berhorst, 2000; Jackson et al., 2003).These have been later involved in salt tectonics duringthe Lu¢lian orogeny as proposed by John et al. (2004). Si-milarly, the West-Congolian basin, and possibly also theMbuji-Maji basin also created favorable conditions forcarbonate shelf development and evaporate deposition asreviewed above.

The gravity modelling and subsidence analysis of Kadi-ma et al. (2010, 2011) further suggest the presence of a largefossil rift concealed under the central part of the CongoBasin and aligned in a NW^SE direction with theSembe-Ouesso and the Mbuji-Maji failed rift basins(Fig. 3c).This Pan-African rifting stage, not only governedthe sedimentation but also created permanently weakzones within the Congo craton, probably by reactivatingthe fabric of the Eburnean belts that sealed the cratoniccores. This is supported by the signi¢cant seismicity re-corded within the Congo Basin (international cataloguesand Ayele, 2002).

Pan-African near-field collisional stage

The Pan-African cycle terminates with the ¢nal amalga-mation of continental blocs into the Gondwana continent(Collins & Pisarevsky, 2005; Tohver et al., 2006). Thisoccurred with a series of collisional events that a¡ecteddi¡erent parts of the continent. In Central Africa, thePan-African deformations surrounding the Congo Cratonand a¡ecting the Neoproterozoic deposits of the CongoBasin are known mainly in the southeastern (Katanga)and western (Bas Congo) regions of the DemocraticRepublic of Congo:

� in the Copperbelt of Katanga and Northern Zambia,the Pan-African event is responsible for the Lu¢lianArc which is located approximately1000 km away fromthe center of the Cuvette. It is characterized by a com-pressional regime inducing curved folds, thrusted ter-ranes and shear zones, with its paroxysm at �550Ma(Cailteux & Kampunzu, 1995; Porada & Berhorst,2000; Batumike et al., 2006).

� in the Bas Congo, the Pan-African event is also re-sponsible for theWestCongo belt (Fig. 2), which paral-lels the Atlantic coast betweenGabon andAngola, andis located 200^400 km west of the Congo Basin. Theintensity of folding and thrusting within the West


E. Kadimaet al.

Congo belt decreases fromwest to east (Boudzoumou& Trompette, 1988; Tack et al., 2001; Pedrosa-Soareset al., 2008).

To theNE and SE of the Cuvette, outcroppingNeoproter-ozoic to early Paleozoic sediments are respectively knownas the Lindi and Mbuji-Maji Supergroups (Fig. 3c). The¢rst is weakly folded with few synclines or anticlinesbounded by high-angle faults whereas the second forms agentle NW^SE trending syncline a¡ected toward thesoutheast by Pan-African NNE^SSW-oriented folds(Cahen et al., 1984).

The lower age limit for the paroxysm of the Pan-Africanorogeny surrounding the Congo Craton is at c. 550Ma,with the ¢nal collision at �530Ma with later manifesta-tions at c. 520Ma., at the vicinity of the 542MaNeoproter-ozoic^Cambrian transition (Porada & Berhorst, 2000;John et al., 2004; Armstrong et al., 2005; Rainaud et al.,2005).

Gondwanide far-field activemargin stage

Some of the compressional structures observed on theseismic pro¢les have been attributed by Daly et al. (1991)to theLate PaleozoicÊarlyMesozoic deformation.Withinthe Lindi Supergroup near Kisangani, Sluys (1945) andVerbeek (1970) describe a narrow WNWÊSE belt of in-tense deformation involving folding, brittle shearing andmetasomatism that a¡ect the whole Lindi Supergroup (in-cluding the Aruwimi Redbeds) but which pre-dates theUpper Jurassic and Cretaceous sediments. In the Lukugacoal ¢eld near Kalemie along the Congolese side of LakeTanganyika,Cahen andLepersonne (1978) describe an un-conformity between the Permian sediments of the LukugaGroup and the overlyingTriassic sediments of the HauteLueki Group, thus constraining the age during the Per-mo-Triassic transition. Further to the northeast, strike-slip deformations due to NNE^SSW to N^S tectoniccompression and a¡ecting the Permian coal-bearing Kar-oo rocks have been noted in the Ubende belt inTanzania,particularly in the Namwele-Mkolomo coal ¢eld (Delvauxet al., 1998; Delvaux, 2001b).

These localized deformations scattered over a wide ter-ritory are coeval with the Late PermianÊarlyTriassic de-velopment of the Cape Fold Belt of South Africa (H�lbichetal.,1983;LeRoux,1995;Newton etal., 2006;Tankard etal.,2009).This belt was part of the Gondwanide passive mar-gin orogen that framed the southern margin of Gondwanaand was related to the Paleo-Paci¢c subduction plate dip-ping beneath Gondwana during Late Carboniferous tomid-Triassic times (Ziegler, 1993; Visser & Praekelt, 1996;Trouw & De Wit, 1999; Delvaux, 2001a; Giresse, 2005).The contractional deformations identi¢ed in the seismicsections of the Congo Basin are also related to the far- ¢elddeformation resulting from the distant (roughly 2500 kmaway) Cape folding event (Daly et al., 1991, 1992). At thesame time, an extensional tectonic regime prevailed alongthe Tethyan northern margin of Gondwana (W˛pfner,1994; Ring, 1995; Delvaux, 2001a; Catuneanu et al., 2005).

WELL DATA

The two fully cored stratigraphic wells were drilled, in thelocalities of Samba (2.039m; Cahen et al., 1959) andDekese(1.856m; Cahen et al., 1960) (Fig. 2), reaching the Redbedsat respectively 1.167 and 1.677m deep but without traver-sing them entirely.

The exploration wells of Mbandaka (4350m; Esso Za|« reS.A.R.L., 1981a) and Gilson (4563m; Esso Za|« re S.A.R.L.,1981b) have been cored only over a few meters. Cuttingswere sampled every 10m but they are no longer accessible(probably lost in the DRC during the political troubles).The Mbandaka well was stopped in massive salt depositswith anhydrite interbeds after encountering stromatoliticcarbonates while the Gilson well was stopped in massivedolomite. None of the four wells reached the crystallinebasement, evidenced by the seismic re£ection pro¢lesand tentatively assigned to the Paleoproterozoic as re-viewed above.

Only the Jurassic to Recent sediments encountered intheDekese (Couches A^C) and Samba (Couches1^5) wellshave been accurately dated paleontologically as reviewedabove. Sediments attributed to the Karoo and correlatedwith the Lukuga formation have been recognized in theDekese well but are absent in the Samba well. The HauteLueki formation, outcropping along the eastern marginof the basin, has not been recognized in these stratigraphicwells. In the lower part of both wells, the red arkoses ofCouches H (Dekese) and Couches 6 (Samba) have beencorrelated to the Aruwimi Group.

The stratigraphic information related to theGilson andMbandaka exploration wells are unfortunately sparse, asthey are provided mainly by cuttings that cannot be acces-sible anymore for further study.The upper sediments areattributed to the Cretaceous to Recent, but the correlationof the deeper deposits is more problematic.The presenceof the Stanleyville and Haute Lueki formations is sug-gested in the well Geological Completion Reports (EssoZa|« re S.A.R.L., 1981a, 1981b) but is not mentioned expli-citly in the E.C.L (1988) report nor in Lawrence &Makazu(1988). Belowwhat Lawrence &Makazu (1988) name as the‘Base Jurassic unconformity’, they identify an Early toMiddle Paleozoic Upper Zaire Sequence (UZS) separatedby the ‘Pan-African Unconformity’ from the Lower ZaireSequence (LZS) which itself overlies the Zaire Carbo-nate/Evaporite Sequence (ZC/ES). The UZS includes in-di¡erently both the Aruwimi and Lukuga Subgroups. Inthe E.C.L. report, however, the Lukuga formation (domi-nantly conglomeratic) is identi¢ed above successively the‘Infra-Cambrian’ Schisto-Greseux, the Schisto-Calcaireand the Carbonate-Evaporite sequences. The Carbonate-Evaporite sequence was not originally di¡erentiated fromthe rest of the Schisto-Calcaire in the well GeologicalCompletion Reports. This subdivision of the Schisto-Calcaire was proposed by Daly et al. (1992) by correlationwith the outcropping sections of the Lindian (Verbeek,1970), referring the upper part (calcareous to dolomiticsilciclastics) to the Lokoma Group and the lower part



(‘Carbonate-evaporite sequence’), to the Ituri Group, se-parated by a weak unconformity (Fig. 3b).

An important result of the Gilson and Mbandaka wellsis the discovery of the presence of evaporites and salt intheir deeper parts. In order to prevent possible confusion,we use the term evaporite for designing all the primaryminerals formed by direct precipitation from brines (sul-fates and halides) and salt as in salt tectonics, for low-den-sity and deformable halite.

The lower part of the Schisto-Greseux sequence inboth the Gilson and Mbandaka wells contains thin anhy-drite layers, mixedwith clay in the Gilsonwell (Esso Za|« reS.A.R.L., 1981b), and with traces of gypsum in the Mban-daka well (Esso Za|« re S.A.R.L., 1981a).

In the Mbandaka well Geological Completion Report(Esso Za|« re S.A.R.L., 1981a), the upper part of the Schis-to-Calcaire is described as composed of dark-grey calcar-eous to silty shales grading to limestone, the middle partcontains argillaceous, dolomitic and stromatolitic lime-stone, and the lower part is composed of massive salt withanhydrite interbeds, down to the base of the well. Thecomposition of this salt is not mentioned but in petroleum

exploration, salt is generally assimilated as halite. In theGilson well Geological Completion Report (Esso Za|« reS.A.R.L.,1981b), theSchisto-Calcaire is described as an al-ternating sequence of sandstone, siltstone, shale and dolo-mite but neither salt nor anhydrite is reported.

SEISMIC REFLECTION DATA

A total of 2900 km of seismic re£ection lines have been re-corded in an area of �0.5Mkm2 between the Congo andKasai rivers, both along rivers (lines R) and roads (lines L)in 1974^1976 by Shell, Texaco and Esso. The locations ofthe lines are indicated on Figs 2, 7 and 8 and selected partsof the lines are shown in Fig.9 (L50), 10 (L59), 11 (R13), 12(R15), 13 (R5).

Using seismic re£ection lines calibrated with the fourwells and correlated tentatively with outcrop data fromthe margin of the basin, three major seismo-stratigraphicunits can be de¢ned within the Congo Basin (Units A, B,C; Figs 4 and 5). They are resting over the seismic base-ment and are separated by prominent seismic re£ectors

Fig.4. Calibration of seismic lineL52 with the stratigraphic log ofSambawell. Seismic units areshownwith reference to Fig. 3.Onlythe fragment of the line at thevicinity of the well is shown.

Fig. 5. Calibration of seismic lineR5 with the sonic diagraphy ofMbandakawell. Seismic units areshownwith reference to Fig. 3.Onlythe fragment of the line at thevicinity of the well is shown.


E. Kadimaet al.

corresponding to major regional unconformities (U0^U3).The nature of the seismic basement underneath thebasin is unknown, but it is reasonable to infer that it ismainly of crystalline and metamorphic nature, and ofpre-Neoproterozoic age. Although four stratigraphic hia-tuses are shown on the litho-stratigraphic column ofDalyet al. (1992), only two major regional unconformities areobserved on the majority of the seismic lines, and onemore local and deeper (Fig. 3b). Lawrence & Makazu(1988) refer the lower one (U1) as separating the Ituri fromthe Lokoma Groups, the intermediate one (U2) as the‘Pan-African unconformity’at the Neoproterozoic^Paleo-zoic transition, and the upper one (U3) as the ‘Base Juras-sic unconformity’. The latter corresponds to thecombination of the ‘Late Paleozoic deformation’ of Daly etal. (1992) and a regional hiatus identi¢ed over most of thebasin that spans from the end of the Permian, up to thebase of the upper Jurassic and which corresponds to theGondwana peneplaine of King (1963).

These unconformities are clearly marked on the Pro¢leSonicVelocity (PSV) run in theMbandakawell (Figs. 4 and5) and subdivide the stratigraphic sequence of the CongoBasin into three seismo-stratigraphic units, from baseto top:

� UnitA overlies the crystalline basement (U0 unconfor-mity) and represents Neoproterozoic sedimentaryrocks. In the Mbandaka and Gilson wells (Lawrence&Makazu, 1988; Daly et al., 1992), it comprises a 210mthick carbonate-evaporite sequence at the base of thewells, followed by a 700m thick siliciclastic sequence.It is deformed by the Pan-African orogeny and trun-cated by the seismic re£ector corresponding to the U2unconformity. A chaotic seismic facies hampering theidenti¢cation of the basement is locally observed, butit might at least partly be the result of processing. Onthe seismic pro¢les, the two sequences can be di¡eren-tiated on the basis of their seismic facies, allowing theseismic unit A to be subdivided itself into sub-unitsA1 (at the base) and A2 (at the top), respectively corre-sponding to the carbonate-evaporite sequence and theclastic sequence. Compared with the Lindi Super-group lithofacies de¢ned by Verbeek (1970), thesesub-units can be tentatively correlated respectively tothe Ituri and theLokomaGroups, separated by a slightunconformity (U1) (Fig. 3).

� UnitB lies above theU2unconformity and is boundedat its top by the U3 unconformity. It represents withsome hiatus the Paleozoic sedimentary sequence. Se-diments representing this period crop out mainly tothe east of the Congo Basin and are known as the Red-beds (in particular, the ‘Banalia arkoses’) of the Aruwi-mi Group (undated upper part of the LindiSupergroup; Tack et al., 2008) and the Lukuga forma-tion (Carboniferous-Permian). In the Dekese well(Fig. 3b) where it has been proposed by Cahen et al.(1960), the transition between the ‘Couches G’ (tillitesand varval clay of the Late Carboniferous Lukuga for-

mation) and the Redbeds of the ‘Couches H’ appearsrelatively subtle. Re-examination of the cores storedwithin the RMCA shows that this transition is markedby a facies change, but without particular discontinu-ity in sedimentation. No marked discontinuity can beseen and the contact clearly shows that fragments ofred sandstones have been reworked and re-depositedwithin the diamictite at the base of the Lukuga se-quence.The bedding has the same 40 to 451 inclinationat both sides of this contact.Therefore, questions canbe raisedwhether the red arkoses of the Aruwumi for-mation progressively grade into the glacial sedimentsand are thus at least partlyMiddle Paleozoic in age, orif they are separated by an important stratigraphic hia-tus spanning a large part of the Middle Paleozoic. Inthe seismic sections, these two units cannot be clearlydi¡erentiated and do not appear to be separated by amarked re£ector.They are therefore treated as a singleseismo-stratigraphic unit (B).

TheHaute Lueki formation (lowerTriassic) which overliesthe Lukuga formation is known in outcrop only on theeastern margin of the basin and within the Lukuga basinnear LakeTanganyika, but has not been recognized in theSamba andDekesewells.Both formations are suspected intheMbandaka andGilsonwells, butwithout paleontologi-cal con¢rmation (Esso Za|« re SARL, 1981a, 1981b). There-fore, the position of the Haute Lueki formation in termsof the seismic units is di⁄cult to de¢ne, but for reasons ex-plained below, we favor the idea that this formation post-dates the Late Paleozoic (Permo-Triassic) deformationevent.

� Unit C overlies the U3 unconformity and correspondstoMesozoic and/orCenozoic sediments.We include inthis unit the Haute Lueki formation that we considerto post-date the Permo-Triassic tectonic deformationevent.With reference to the succession known in theoutcrops surrounding the basin (Lepersonne, 1977),this unit comprises, from base to top: the MesozoicHauteLueki, Stanleyville, Loia, Bokungu andKwangoformations, and the Tertiary Gre' s Polymorphes andOchre Sands. Unit C is almost £at and un-deformedas shown on the seismic lines.

In summary, according to this threefold seismo-strati-graphic subdivision and accepting that the Haute Luekiformation postdates the Permo-Triassic deformationevent, Unit Awould be equivalent to the Ituri andLokomagroups (Ediacaran and Cryogenian), and possibly also theeventual prolongation of the Mbuji-Mayi and Liki-Bem-bian/Sembe-Ouesso Supergroups (Tonian) under the ba-sin, Unit B would be include the Aruwimi Group andLukuga formation (Paleozoic) andUnit C, the Stanleyvilleformation up to the super¢cial deposits (Meso-Cenozoic)(Fig. 3b).



POTENTIAL FIELDS

In order to test the hypothesis of Daly et al. (1992) that thestructures in the central part of the Congo Basin are re-lated to crustal contraction and uplift of a basement block(Kiri High), we reinterpreted the seismic pro¢les takinginto account additional constraints from potential ¢elds(gravity and magnetism).

TheMbandaka andGilsonwells have evidence that thecarbonate-evaporite sub-unit A1 locally contains massivesalt in addition to anhydrite (ESSO Za|« re SARL, 1981a,1981b). The models constructed hereafter are designed toinvestigate the lateral extent of the presence of salt at thebase of the basin ¢ll and check the existence of the pro-posed Kiri High.With gravity data, we expect to identifylow-density sediments such as deposits rich in salt andshale and to di¡erentiate them from denser crystallinerocks. Salt-rich formations usually have lower magneticsusceptibilities than shale-rich formations. Hence, onpole-reduced magnetic pro¢les, salt-rich formations aretypically associated with negative magnetic anomalieswhile shale-rich formations are typically associated withpositive magnetic anomalies.

Gravitymap

The ¢rst gravity measurements in the Congo Basin wereacquired by the ‘Syndicat pour l’e¤ tude ge¤ ologique etminie' re de la Cuvette congolaise’ (the Syndicate) be-tween 1952 and 1956. A total of 6000 gravity measure-ments were made at an average interval of 5 km alongthe rivers and existing roads. Based on these data and

using a reduction density value of 2.67 gm cm� 3 the ¢rstmap of the Bouguer anomaly was published by Jones etal.(1960).

The database used in the present study has beengraciously provided by the ‘Bureau Gravime¤ trique Inter-national’ (BGI) of Toulouse (France). It contains dataacquired by the Syndicate, the Lamont Doherty Geologi-cal Observatory, the ORSTOM, and the Leopoldville(now Kinshasa) Meteorological Service. A thorough andsystematic comparison of 4500 duplicate gravity measure-ments was made on the various sets of data.This leads tothe computation of a root mean square error of� 0.67mgal, which is suitable for a regional study and al-lows the combined use of the di¡erent datasets.

TheBouguer anomaly map obtainedwhich has been es-tablished from a 30� 30 km grid interpolated from therawdata by minimum curvature (0.25 tension factor, 20 ex-tension ratio of surface iteration set, 0.1convergence limit),shows Bouguer anomaly values ranging between �120and � 20mgals (Fig. 6). This map is dominated by NW^SE trending anomalies with a wide gravity low centeredon the Lokonia High separated from a gravity high on itsSW side by a steep gradient, and by a rapid succession ofgravity highs and lows along a SW^NE line from the Inon-go High to the Busira Basin.

Magnetic susceptibilitymap

In 1984 a 30 000 km aeromagnetic survey was acquired byHunting Services Inc. on behalf of JNOC using a CessnaTitan 404 airplane over the central part of the Cuvette,from 21N to 21S and from 18 to 261E, covering

Fig. 6. Bouguer anomaly map ofthe Congo Basin region as overlayover the shaded-relief GtopoDEMwith the modeled seismic lines.Bouguer density: 2.67 g cm� 3;4mgal spacing; step grid: 30 km.


E. Kadimaet al.

380 000 km2. The N^S £ight and E^W tie lines wererespectively £own 18 and 55 km apart, while the altitudewas ¢xed at a constant barometric elevation of 750m abovemain sea level.The magnetic data acquiredwere processedby Nikko Exploration and Development Co to stripthe noise and to perform the required correction beforeproducing a Total Field Magnetic map at a 1 : 1000 000scale.

We used there already corrected and ¢ltered data by di-gitizing the map produced by Nikko and compiled a5 � 5 km LCT grid (Mercator WGS84). The resultingmap (Fig. 7) shows a general gradient from the NE to theSW, with sharp variations along WNW-trending lineswhich might correspond to fault lines.

In order to get the shape anomalies to what would beobserved at the pole where the inducing ¢eld is vertical,pole reductionwas applied to the total magnetic ¢eld.Notethat the horizontal gradient reduction to pole introducesE^W striping that should not be over-interpreted.

A ¢rst horizontal gradient was derived from the result-ing map (Fig. 8), showing that the magnitude of the hori-zontal gradient de¢nes magnetization boundaries andthat three major magnetic zones can be individualized:

� a northeastern zone with NW^SE and E^W trendingof high magnetic gradient which corresponds to theLindi Supergroup, possibly covered in part by recentsediments;

Fig.7. Total aeromagnetic map ofthe northern part of the CongoBasin (20 nTspacing; 5 km gridstep).

Fig. 8. Horizontal gradient of thepole-reduced aeromagnetic mapfor the central part of the CongoBasin.1nTkm�1 spacing; stepgrid: 5 km.Note that the horizontalgradient reduction to poleintroduces E^W striping thatshould not be over-interpreted.



� a central zone dominated by a low magnetic gradientcorresponding to the region of deeply buried base-ment under the Busira Basin;

� a southwestern zone characterized by a moderatemagnetic gradient, with a NW-SE magnetic trend inthe area where the top of basement is less deep, fromthe Lokonia High to the Lokoro Basin.

Potential field modeling

To better image the subsurface structure of the basin aswell as the deformation taking into account the geome-tries interpreted from the seismic lines, several potential¢eld models were run and tested.

The quality and validity of the interpretation of gravitydata depends on the knowledge of the lateral variation inthe density of the rocks. In this study, density values usedin the various models were determined from Compen-sated Neutron Density Logs run in the Mbandaka andGilson wells, completed by density measured from out-crop samples collected from the di¡erent stratigraphicunits. The density values obtained are summarized inTable1.

Magnetic susceptibility was measured by JNOC (1984)on rock samples collected from the Kisangani, Kananga,and the coastal bulge regions, fromCenozoic to Paleopro-

terozoic in age. The determined magnetic susceptibilityvalues are listed in theTable 2.

The gravity andmagnetic datawere modeledwith the 2-D seismic, Gravity and Magnetic Earth Modeling Systemusing a computer algorithm based on theTalwani methodwith 2-D polygonal bodies (Talwani et al., 1959).The poly-gons were generated from seismic horizons and relateddensities and/or susceptibility were added in the program.Five 2Dmodels have been constructed along selected seis-mic lines where post- sedimentary deformations are de-tected. The density and susceptibility values used in themodels are summarized inTable 3.

� Model 1 (Fig. 9) is based on the N-S oriented seismicline L50 located in the centre of the Cuvette (Figs 2and 6^8). This line illustrates the deformation of theapproximately 2000m thick Unit A, and a later defor-mation a¡ecting both units A and B.The seismic pro-¢le (Fig. 9a) shows in the central part of the basalsedimentary sequence (Unit A), a poorly de¢ned seis-mic zone marked by a chaotic seismic facies overlying atransparent seismic facies.Within this chaotic seismicfacies, structures resembling to folds, contractionalfaults can be identi¢ed below a toplap seismic interfaceseparating it from the overlying Unit B and whichde¢nes the unconformity U2.The two units have beenlater a¡ected by a second deformation, probably relatedto the Permo-Triassic event. This second event pro-

Table1. Density values measured and estimated for gravity model

Seismic unit Lithology

Average density (gm cm� 3)

Measured Estimatedin models(1) (2) (3) (4)

Unit C Clastics, Siltstones, rarely Shales 2.09 � 0.18 2.35 � 0.1 2.13 � 0.21 2.25 � 0.05Unit B Clastics, Siltstones,Tillites, rarely Shales 2.47 � 0.05 2.51 � 0.1 2.45 � 0.05Unit A Clastics, Shales, Limestone, Salt, Anhydrite 2.57 2.68 2.54 2.42 � 0.22 2.55 � 0.05Crystallinebasement

Undi¡erentiated Crystalline rocks 2.73 2.69 � 0.16 2.70 � 0.03

(1) CompensatedNeutronDensity Log fromGilson (Esso Za|« re SARL,1981a), (2) Compensated NeutronDensity Log fromMbandaka (Esso Za|« reSARL, 1981b); (3) Samples measurements after JNOC (1984); (4) Samples measurements after Unocal (1986).

Table 2. Magnetic susceptibility ( � 10� 6 e.m.u.) measurements from the study area (JNOC,1984)

Rock typeNumber ofsamples

Minimal value� 10� 6 e.m.u.

Maximal value� 10� 6 e.m.u.

Average� 10� 6 e.m.u.

Sandstone 33 9 143 49Limestone 6 42 82 60Silt 4 25 50 36Quartzite 9 6 87 34Granite 10 50 1118 470Granodiorite 6 28 2202 1094Schist 13 51 2195 695Shale 7 36 75 53

(4): Number of samples.


E. Kadimaet al.

Table3. Densities r (g cm� 3) and magnetic susceptibility ( � 10� 6 e.m.u.) values used in the models for the di¡erent units

Units

Models 1a and1b Model 2 Model 3 Model 4 Model 5

g cm� 3 � 10� 6 e.m.u. g cm� 3 g cm� 3 � 10� 6 e.m.u. g cm� 3 � 10� 6 e.m.u. g cm� 3 � 10� 6 e.m.u.

C 2.3 48 2.3 2.3 48 2.3 48 2.3 48B 2.46 56 2.54 2.37 103 2.35 87 2.46 56AA2 2.58 59 2.53 102 2.54 104 2.58 59A1 2.45 0.2-0 2.47 0.2-0

Seismicbasement

2.67 792 2.67/2.77 2.67 648 2.67 590 2.67/2.7 792

(c)(b)

(a)

Fig.9. (a) Interpreted entire seismic lineL50 (N^S, in the center of theCuvette, Fig. 2 and6^8) illustratingmajor unconformities, faultsand a poorly de¢ned seismic zone marked by a chaotic seismic facies overlying a transparent seismic facies in the center of the pro¢le. Inthe light of the results of gravity and magnetic modeling, these characteristic seismic facies and deformations are interpreted as acollapsed salt structure (see discussions).TheGondwana and Pan-African unconformities separate the seismo-stratigraphic units A, Band C as de¢ned in Fig. 3. (b)Model 1a (seismic line L50): 2D potential ¢eld models along the same line, case when a basement high isconsidered. Note a mismatch between measured and modeled gravity curves for this case. Polygons: 1, Unit C; 2, Unit B; 3, Unit A; 4,Basement. Densities and susceptibilities are given inTable 3. (c)Model1b (seismic line L50): 2D potential ¢elds models showing a goodmatch between modeled and measured curves with a non uplift basement hypothesis. Note the higher frequency and amplitudemagnetic anomaly related to the induced magnetization in the poor seismic de¢nition zone. Polygons: 1, Unit C; 2, Unit B; 3, Unit A; 4,Basement. Densities and susceptibilities are given inTable 3.



ducedweak deformations inUnit B,mainly in apparentreactivation of the underlying dislocation zones. Itmight also be responsible for the transparent seismicfacies seen on the northern end of the seismic pro¢le.

This poorly de¢ned seismic zone in the center of thepro¢le has been interpreted byDaly etal. (1992) as an up-

lifted basement linking the Lokonia andKiri Highs andbounded laterally by high-angle reverse faults, gradingat depth towards low-angle thrusts.To test this hypoth-esis, we modeled the gravity along this seismic line(Model 1a, Fig. 9b) using a typical crystalline basementdensity (r5 2.67) for the transparent part of Unit A.This con¢guration gives a clear divergence between the

(a)

(b)

150100500Distance (km)

–10

–5

0

Dep

th (

km)

SW4 3

2

1


–100

–80

–60

–40

mG

al Observed gravity

Calculated gravity

NE

Fig.10. (a) Interpreted entire seismic lineL59 (NE^SW, in the region of theDekesewell, Fig. 2 and6^8) illustratingmajor unconformitiesand anticlinal £exures possibly rooted on salt pillows.TheGondwana and Pan-African unconformities separate the seismo-stratigraphicunits A, B and C as de¢ned in Fig. 3. (b)Model 2 (seismic line L59): 2D gravity models showing a good match between modeled andmeasured curves obtainedwith a lateral variation of density in the basementwhich suggested variation in velocity. Polygons: 1, Unit C; 2,Units B and A; 3, Basement1 (2.67 g cm� 3); 4, Basement 2 (2.77g cm� 3). Densities and susceptibilities are given inTable 3.

Fig.11. (a) Interpreted entire seismic line R13 (WNW^ESE, Fig. 2 and 6^8) illustrating major unconformities, faults and seismictransparent zones interpreted as salt pillows within the seismic unit A1 (Carbonate-evaporite sequence).The Gondwana and Pan-African unconformities separate the seismo-stratigraphic units A, B and C as de¢ned in Fig. 3. An additional discontinuity separatingsub-unitsA1andA2 is shown.The top of the crystalline basement is inferred from the gravity andmagnetic modeling results. (b)Model3 (seismic line R13): 2D potential ¢elds models along the line. Note a negative potential ¢eld’s signature and a good match betweenmodeled and measured curves obtained by considering a low density and magnetic lithology contrasting with surrounding andoverlying sediments. Polygons:1,UnitC; 2,UnitB; 3, sub-UnitA2; 4, sub-UnitA1; 5, Basement.Densities and susceptibilities are giveninTable 3.


E. Kadimaet al.

observed andmodeled gravity curves (Fig.9b).Agood ¢tbetween the observed and modeled gravity (Fig. 9c) isonly obtained when a unique density value of

2.55g cm�1 is allocated to Unit A (Table 3) on the entireline.This strongly suggests that no major lateral litholo-gical change occurs withinUnit A along this line L50.

(a)

(b)


–10

–5

0

Dep

th (

km)

SENW

5

43

21


–105

–103

–101

mG

al


–1.2

–0.7

–0.2

0.3

nT

/km

Observed gravity

Calculated gravity

Observed magnetics

Calculated magnetics



The magnetic modeling shows in the central part of lineL50 a considerable divergence between the observedgradient and the modeled curves, except at its northernend (Model 1b, Fig. 9c). This magnetic mismatch con-trasts with the excellent ¢t achieved in gravity modelingandwill be discussed later.

� Model 2 (Fig. 10) is constructed on the NE^SW lineL59 in the region of the Dekese well in the southernmargin of the Cuvette (Fig. 2), outside the area coveredby the aeromagnetic survey (Fig. 8), but characterizedby a high gravity gradient (Fig. 6). Anticlinal folds af-fecting the Unit A, and possibly underlain by salt pil-lows (see discussion hereafter) are clearly observed onthis line (Fig.10a). In the related model, the ¢t betweenobserved and modeled gravity curves is achieved witha gradual lateral change in the density of the basement,parallelizing the variation observed in the seismicstacked velocities (Fig. 10b).This change in basementdensity suggests the presence of a heterogeneous base-ment or a dense basic magmatic intrusion in thesouthern part of the pro¢le.

� Models 3 and 4 (Figs 11 and 12) are located along twoWNW^ESE seismic lines, respectively R13 in the Lo-konia High and R15 in the Lomami Basin (Figs 2 and6^8). Line R13 (Fig.11a) shows in its central part an an-ticline within the seismo-stratigraphic Unit A, whichdisrupted the continuity of this unit and reached theU3 unconformity. It disturbed also the overlying UnitB. Smaller anticlines can be seen on both sides of theline, and positive £ower structures are rooted on allthese anticlines. A similar but less pronounced anti-cline can be clearly observed on the pro¢le R15 (Fig.12a) where it a¡ects both Units A and B. Convergingre£ectors at the top of the structure and toplap geome-tries highlight the U3 unconformity surface.On both lines, Unit B and the underlying Pan-Africanunconformity appear passively involved in the defor-mation of theUnitA.On lineR-13, deformation locallya¡ects theU3 unconformity while on line R15, this un-conformity is undisturbed.In order to obtain a good match between the observedand modeled curves for the anticline for these twomodels (Figs 11b and 12b), a layer of lower density(r5 2.45^2.47) and susceptibility value (m5 0.2^0 � 10� 6 e.m.u.) has to be introduced above the base-ment (r5 2.67; m5 648� 10� 6 e.m.u.) and below alayer of relative high density (r5 2.53 and 2.54) andsusceptibility (m5 102� 10� 6 e.m.u.). This producesa local inverted trend compared with the expectedand normal trend of increasing values of density andsusceptibility with depth. The modeled low-densitylayer (r5 2.45^2.47) is intermediate between the aver-age density values for sediments used for theUnit A inother models (2.55) and the classical density value ofevaporite or salt minerals (2.1^2.3). The low modeledsusceptibility values suggest that this low-density layercontains a signi¢cant proportion of salt. According to

calibrations and correlations, this layer would beequivalent to the carbonate-evaporite sequencefound at the bottom of theMbandaka andGilsonwells(sub-unit A1). The intermediate density and low sus-ceptibility values used to model it are consistentwith the idea that this layer is not (or no more)composed of pure salt. It could either have been de-posited as a mixture of sediments and impure salt,or as massive salt deposits which are now partlydissolved, or any combination of both possibilities.The overlying, higher density layer would correspondto the clastic sequence found in the same wells(sub-unit A2).

� Model 5 (Fig.13) is constructed along theNE-SW seis-mic line R5 shot on the Congo River, and passingthrough theMbandakawell (Figs 2 and6^8). It displaysat its southwestern extremity a poorly de¢ned seismiczone below the U3 unconformity (Fig.13a).This zone,identi¢ed as the ‘Kiri High’ by Daly et al. (1992) ismarked by a positive gravity anomaly (Fig. 6). On thisline, both Unit A and Unit B are deformed, but thesestructures are sealed byUnitC above theU3unconfor-mity.Within the poorly de¢ned seismic zone, weak butrelatively well de¢ned sub-parallel re£ectors can beseen on more than 2 sTWT(TwoWay travelTimes) be-low the upper unconformity.To achieve a fair ¢t between the observed and themodeled pro¢les (Fig. 13b), a gradual lateral change inbasement density along the trend of the pro¢le has tobe considered. In such a case, the basement is located3000m below the U3 unconformity, i.e. at more than4000m depth, corresponding approximately to thebottom of the Mbandaka well (Fig. 13a). This modelshows also a gravity high as in model 2 interpreted asre£ecting a lateral change in basement density.The mismatch between the observed and modeledmagnetic values onModel 5 (Fig.13b) reveals a possiblesimilarity with the deformation style on line L50(Model 1).

DISCUSSION

Combining seismic, gravity and magnetic 2D models runfor the Congo Basin allows individualizing vertical inter-faces as well as lateral variations in the subsurface densityand magnetic susceptibility of the basement and the sedi-mentary cover.

Our models show that deformation in the basin isnot entirely due to locally uplifted basement as earliersuggested. They indicate that lateral variations of densityand magnetic susceptibility values in various compart-ments of the basement may also in part account for thedeformation and potential ¢eld signatures in the region.Based on the results of this study, we propose an alterna-tive solution to the basement uplift hypothesis of Dalyet al. (1992) and provide a possible explanation for the ori-


E. Kadimaet al.

0

(a)

(b)

5 10 15 20 25 30Distance along line (km)

Seismic line R15

–3

–2

–1

0T

WT

(se

c)

WNW ESE


–10

–5

0

Dep

th (

km)

WNW ESE5

4

3

2

1


–86

–84

–82

–80

mG

al


–1

0

1

nT

/km

Observed magnetics


Observed gravity

Calculated gravity

Fig.12. (a) Interpreted entire seismic line R15 (WNW-ESE, Fig. 2 and 6^8) illustrating major unconformities and possible salt pillowsunder large anticlines.TheGondwana andPan-African unconformities separate the seismo-stratigraphic unitsA,B andC as de¢ned inFig. 3. An additional discontinuity separating sub-units A1and A2 is shown. (b)Model 4 (seismic line R15): 2D potential ¢eld modelsalong the line showing a good match between modeled and measured curves with a non uplift basement hypothesis. Note a typicalmagnetic negative signature related to the presence of salt. Polygons: 1, Unit C; 2, Unit B; 3, sub-Unit A2; 4, sub-Unit A1; 5, Basement.Densities and susceptibilities are given inTable 3.



gin of some seismic structures observed in the Congo Ba-sin (i.e. anticlines, seismically transparent zones, andzones of chaotic seismic facies).The proposed alternativedoes not contradict but rather complements the horstand graben structure of the deep basin as shown by theaeromagnetic grid and the interpretation of the seismiclines.

Evidence for the presence of a deep lyingevaporitic formation

InModels 3 and4 (linesR13 andR15,Fig.8),we have shownthat a local inversion in the normal trend of both densityand susceptibility increase with depth is needed to obtaina good match between the observed and modeled curves.


–10

–5

0

Dep

th (

km)

5 4

321

SW NE


–70

–60

–50

mG

al


–10

–5

(a)

(b)

nT

/km

Observed magnetics


Observed gravity

Calculated gravity


E. Kadimaet al.

This has been explained by the possible presence at thebase of the sedimentary column of a salt-rich low densityand susceptibility layer within the carbonate-evaporite se-quence (sub-unit A1).

The presence of a deep lying carbonate-evaporite for-mation in the Congo Basin has been recognized by theEsso Zaire exploration campaign (ECL, 1988), but Daly etal. (1992) did not considered the possible contribution ofsalt for explaining the seismic structures in lines L50 andR5 interpreted to be related to basement uplift.These lineswhich cross the presumed Kiri High (Figs 2 and 8) arecharacterized by the presence of poorly de¢ned seismiczones with a transparent seismic facies (Figs 9a and 13a),attributed to an uplifted basement by these authors. Forthese lines, our 2D gravity and magnetic models 1 and 5show an important mismatch if we assume high densityand susceptibility values to the sub-unit A1 as would bethe case with a basement uplift. Instead, a close match be-tween the observed and computed curves in the two mod-els is obtained with crossover or nil zones, i.e., anintermediate salt (evaporite) and sediment density, insteadof an uplifted crystalline basement.

These arguments, although not de¢nitive by themselvesfor the presence of salt in the basal sedimentary series ofthe Congo Basin, are in line with the suspected develop-ment of failed rift basins during theNeoproterozoicwithinthe Congo craton. Moreover restricted marine (lagoon) tosabhka environments with deposition of stromatoliticlimestones and evaporites has recently been identi¢ed inthe adjacentWest-Congo, Nguba andMbuji-Mayi basins.

Evidence for salt movements and collapsedsalt structures

In line L50 (Fig. 9a, Model 1), the transparent seismic fa-cies appears as chaotic, while in line R-5 (Fig. 13a, Model5), it displays weak seismic horizontal re£ectors. In thiscase, these poorly de¢ned seismic zones might correspondto a ‘collapsed salt structure’ due to the truncation of thesalt wall structure, fossilizing past salt movements.Although the poorly de¢ned seismic facies might re£ectsthe limits of the seismic data processing, the charac-teristics of the U2 unconformity surface and the relatedtoplap seismic facies seen on these lines sustain theseconclusions.

In complement to the discussion of models 1-5 above,and in the light of the results of gravity and magneticmodeling, we provide a new interpretation for the struc-tures observed on the corresponding seismic lines L50and R5, assuming that they are related to the presence ofsalt in the lower part of the sedimentary succession (sub-unit A1). At the center of line L50 (Fig.9a), the characteris-tic seismic facies and deformations observed can be con-sidered as a collapsed structure resulting from saltdissolution. The large anticlinal folds observed in lineL59 (Fig. 10a), are interpreted as being rooted by salt pil-lows. Similarly, the weak transparent seismic facies ob-served at the base of the anticlines on lines R13 and R15(Figs11a and12a), are considered also as collapsed salt pil-lows.On theNE side of theMbandakawell (line 5,Fig.13a),a gentle anticline could be also rooted by a collapsedsalt pillow.

The features observed along these seismic lines typi-cally characterize gravity-driven tectonic deformations ofsedimentary formations rich in salt as evidenced in manyother basins (Wu et al., 1990; Vendeville & Nilsen, 1995;Vendeville & Rowan, 2002; Zouaghi et al., 2005; Rowan &Vendeville, 2006). In particular, Rowan & Vendeville(2006)modeled the shortening reactivation of existing saltdiapirs and related salt-withdrawal minibasins and com-pared the structures with natural examples.They evidencethat the minibasins remain largely undeformed and thediapirs localize the contractional strain. Short diapirs ap-pear to form the core of anticlinal folds and tall diapirs aresqueezed.The heterogeneous deformation that results be-tween the minibasins, salt-rooted anticlinal folds andsqueezed diapirs is accommodated by variably orientedcontractional and even extensional structures.With refer-ence to these examples, we propose that the observed de-formations can be due to salt movement caused bycompressional basin inversion.

Beyond all structural features mentioned above, geo-physical signatures could be used to constrain gravity-dri-ven tectonic features related to the two types of structuresas postulated by Prieto (1996). In gravity modeling, suchcollapsed salt structure corresponds to ‘crossover or nilzone’ where salt and sediment generating a zero densitycontrast are observed like in Prieto (1996). Our models 1and 5 do not require a low density value to match theobserved curves for these zones but the corresponding

Fig.13. (a) Interpreted entire seismic line R5 (NE^SW, running along the Congo River on the western side of the basin, Fig. 2 and 6^8)illustrating the major units and unconformities.The faults forming a £ower structure close to theMbandaka well separate thenortheastern part of the pro¢le withwell expressed parallel re£ectors corresponding to Units A and B, from a poorly de¢ned seismiczone to the southwest.This facies, which Daly et al. (1992) relates to a basement uplift (the Kiri High), is interpreted here as beingassociated to the presence of salt in the sediments. A close examination re£ects the presence ofwell-de¢ned but relatively weakre£ectors in continuity with those, better de¢ned of Unit B northeast of the £ower structure. Using gravity and magnetic modelingconstraints, the top of the crystalline basement is de¢ned under the poorly de¢ned seismic facies in continuity with the northeasternpart of the pro¢le.TheMbandakawell stopped at 4350m deep in massive halite salt. (b)Model 5 (seismic line R5): 2D potential ¢eld’smodels showing a good match between modeled and measured curves obtainedwith a lateral variation of potential ¢eld’s properties inthe basement. Note a higher frequency and amplitude magnetic anomaly related to the induced magnetization in the zone of poorseismic de¢nition zone. Polygons: 1, Unit C; 2, Unit B; 3, Unit A; 4, Basement1 (2.67 g cm� 3); 5, Basement 2 (2.70 g cm� 3). Densitiesand susceptibilities are given inTable 3.



magnetic models do not ¢t (Figs 9c and13b).The observedmagnetic curves along these poorly de¢ned seismic zonesare characterized by a short wavelength and high ampli-tude magnetic signal. The short wavelength of the mag-netic ¢eld anomaly is indicative for a shallow source.Thehigh magnetic amplitude of this anomaly can be explainedby the presence of a secondary induced magnetization likefor example, a chemical remnant magnetization. Suchphenomena can be produced by the truncation of saltwallsstructure and the ¢lling of dissolved zones by unconsoli-dated sediments.

Salt tectonics and tectonic inversion

The implication of salt in the tectonic development of theCongo Basin has been inferred to explain the deformationstyle observed on the seismic pro¢les (ECL, 1988). It hasbeen further suggested that salt had a passive role in con-trolling the geometry of sometimes complex zones ofstructural inversion during compressional tectonics.Salt tectonics in the Congo Basin was proposed by Dalyet al. (1992) as a consequence of tectonically inducedbasement uplift. Catuneanu et al. (2005) rejected thispossibility considering that the contraction and crustaluplift hypothesis was based on an overestimation of thephysical strength of the crust, but they did not considerthat far- ¢eld deformation is generally localized, nor theevidence provided by Guiraud & Bosworth (1997) andGuiraud et al. (2005) and the Asian example that far- ¢eldstresses are a dominant factor controlling deformation inthe cratonic interiors.

The relatively poor processing quality of the seismicsections and their large spacing makes di⁄cult to con¢rmthe structural interpretations of Daly et al. (1992). A denserand 3D seismic networkwould be needed to prove unequi-vocally the presence of strike-slip and reverse faults in thebasin using seismic data alone.However, tectonic fracturesare e¡ectively present in the cores of the Dekese well pre-served in the archives of the MRAC (Cahen et al, 1960).They occur in the Lukuga unit (Couches E-G) that arefor a large part inclined at �301 from the horizontal, andlocally subvertical to overturned.These fractures are nu-merous, sometimeswith a density of several bymeter, bothreactivating the bedding planes and cutting obliquely to it.They occur as shining surface with nicely impressed sliplines and clear slip sense indicators.Their investigation iscurrently in progress by the second author (D. Delvaux),and their tectonic character as reverse to strike-slip faultscan be con¢rmed.

The present modeling shows that deformation in thisbasin is typical of inversion processes related to salt tec-tonics. The vertical salt movement caused the formationof £exures, anticlines, and contractional faults. Such com-plex zones can also be responsible for the observed chaoticseismic facies. In function of the regional geodynamicevolution deduced from the analysis and modeling of theseismic lines, these vertical movements within the carbo-nate-evaporite formation could have been stimulated by

the two major tectonic events that a¡ected the AfricanPlate (Pan-African and Permo-Triassic). Although theseeventswere apparently of low intensity in theCongoBasin,the localization of the far- ¢eld tectonic stress on existingdiscontinuities was probably su⁄cient to have initiated orenhanced the salt movements preferentially above reacti-vated basement structures.

Early Paleozoic salt movement ¢rst started in depocen-ters like the Dekese depression, and in the present centerof the Cuvette, where thick clastic sequences comparableto the Lokoma Group of the Lindi Supergroup overlaythe Ituri-type carbonate-evaporite deposits.The topogra-phy of the basement, sediment loading and transtensionalto compressional regimes related to the Pan-African oro-geny in£uenced vertical salt movement during the EarlyPaleozoic.Toplap terminations of the seismic re£ectors inthe upper part of Unit A (Figs 9a and10a)mark theU2 un-conformity and fossilize the truncation of the underlyingstructure. A second period of salt movement occurredduring the Permo-Triassic tectonic reactivation, preferen-tially in the areawhere units A andB are the thickest. Suchremobilization rejuvenated the Pan-African structuresand caused the development of strike-slip faults andthrusts while folds or anticlines are more speci¢callyfound in the eastern region. Onlap/toplap geometries ofthe seismic re£ectors within units B and C highlight theU3 unconformity and fossilize the truncation of the Per-mo-Triassic salt movements (Figs 11a and12a).

This interpretation can be paralleled with the model ofJackson et al. (2003) that infers large-scale salt tectonicsduring the Lu¢lian compressional tectonics in the Katan-ganCopperbelt. After deposition of theNgubaGroup, saltdestabilization began to form diapirs that have beensqueezed during the Lu¢lian deformation, causingextrusion, megabreccia formation and nape emplacementwhile the salt progressively disappeared by dissolution.In the Congo Basin, salt dissolution during the longgeological history could also be expected, but not comple-tely, as shown by the presence of salt at the base ofthe Mbandaka well and by the potential ¢eld signatureevidenced here.

The regional analysis of the structures related to thislast deformation event shows a west to east decrease in in-tensity, corroborating observations made by Tack et al.(2001) in theWest-Congo fold belt (Fig. 2).To the west, onthe prominent Kiri structure (line R5, Fig. 13a), intensefaulting is developed and the chaotic seismic facies pre-sent directly below the U3 unconformity indicates verticalmovement. To the east (line R13, Fig. 11a), salt movementcaused the anticlinal deformation of both units A and B.In this case, the top of the anticlinal structures reachesthe base of the U3 unconformity, and this creates a nega-tive gravity and magnetic anomaly on the observed pro¢le(Model 3, Fig.11b). Meanwhile, in the northeastern region(line R15, Fig.12a), an anticline is principally developed inUnit A. In this case, the low density carbonate-evaporitelayer is so deep (6^8 km) that it is not in£uencing the ob-served gravity ¢elds.


E. Kadimaet al.

Heterogeneous and deformable basement

Both Models 2 (line L59) and 5 (line R5) need a laterallyheterogeneous basement in order to ¢t the gravity data.The alignment of the central portions of these lines alonga NW trend parallel to the major structures interpretedfrom the seismic survey (Fig. 2) suggest the presence ofNW-SE oriented zone of basement discontinuity beneaththe basin.This couldwell have been originated during theNeoproterozoic failed rifting, itself reactivating the in-ferred Paleoproterozoic mobile belts that weld togetherthe Archean nuclei of the Congo craton.

CONCLUSIONS

Revision of the stratigraphic, paleogeographic and tec-tonic constraints allow updating and re¢ning the evolutionmodel of the Congo Basin since the Neoproterozoic.

The Congo Basin evolved largely in an intraplate con-text, but under the geodynamic in£uence of tectonicevents a¡ecting directly the Congo craton in the Neopro-terozoic, its margin in theNeoproterozoic^Paleozoic tran-sition and the margin of Gondwana during the Paleozoic^Mesozoic transition. It has a discontinuous evolution andacquired its present morphology of a ‘Cuvette’during theMesozoic^Cenozoic period. The lower sedimentary se-quence at the center of the basin is poorly known, only itsupper part is recognized in the Gilson andMbandaka ex-plorationwells.With reference to adjacentNeoproterozoicbasins surrounding theCongoBasin (West-Congo,Uban-gi-Lindi, Katanga and Mbuji-Mayi regions), a laguna tosabhka depositional environment in failed rift basins is in-ferred for this sequence, with stromatolitic limestone andevaporites.

Integration of potential ¢eld data as additional con-straints in the interpretation of the seismic pro¢les allowsa more precise understanding of the deep sedimentarystructures present in the Congo Basin in Central Africa.It highlights the probable importance of salt in the tec-tonic development of this basin in response to the far- ¢elde¡ects of thePan-African events that presided to the amal-gamation of the Gondwana continent and to the far- ¢elddeformation induced by distant the Cape Fold Belt inSouth Africa during the Permo-Triassic transition as partof the Gondwanide convergent passive margin.

In particular, the seismic interpretation and strati-graphic calibration using the Samba and Mbandaka wellscon¢rm the presence of two prominent regional unconfor-mities (U2 and U3) representing the Pan-African andGondwanide events that separate the sedimentary succes-sion of the Congo Basin into 3major seismo-stratigraphicunits (A, B and C), corresponding broadly to the (Late)Neoproterozoic, Paleozoic andMeso-Cenozoic.The care-ful re-examination of the seismic pro¢les together withintegration of gravity and magnetic modeling suggest thelocally signi¢cant presence of salt within the lower part oftheNeoproterozoic sequence (carbonate-evaporite forma-

tion, sub-unit A1). The deformations observed along theseismic lines can be locally related to vertical movementof the Neoproterozoic salt, triggered by the Pan-AfricanandGondwanide tectonic events.

As a consequence, deformations within the Congo Ba-sin can, at least locally, be explainedwithout invoking onlythe previous hypothesis of uplifted basement (Daly et al.,1992). For instance, the narrowWNW^ESE trending KiriHigh has been interpreted as an uplifted basement basedon the assumption that the poorly de¢ned seismic faciescorresponded to the crystalline basement. Using the po-tential ¢eld constraints, we have shown that it can be moresatisfactorily explained by the presence of salt-rich forma-tions.The chaotic aspect of the seismic facies and the re-lated tectonic structures are explained by the mobilizationof the carbonate-evaporite formation during the Pan-African and Permo-Triassic reactivations. We are alsoaware thatwith thiswork, we reached the limit of the inter-pretation of the seismic pro¢les in their present form andthat they certainly need a modern reprocessing.

ACKNOWLEDGEMENTS

We thank the CNE (Commission Nationale de l’Energie)of the Democratic Republic of Congo for the seismic andmagnetic data and BGI (Bureau Gravimetrique Interna-tional) at Toulouse-France for the release of gravity data.The CRGM (Centre de Recherches et de Ge¤ ologie Mini-e' re) at Kinshasa is thanked for providing access to unpub-lished reports. K. Kadima gratefully acknowledges G.Lamy,A.Price, F.Lucazeau for the discussion on 2-D seis-mic, gravity and magnetic earth modeling and interpreta-tion and for their critical and constructive reviews of themanuscript which signi¢cantly improved its content. E.Kadima and S. Sebagenzi are supported by the FrameAgreement program MRAC-DGCD ‘S1_RDC_Geodyn_UNILU’ and D. Delvaux is supported by the Belgian FederalSPP Science Policy, Action 1 program. This paper greatlybene¢ted from the reviews of R. Durheim, C. Ebinger and ananonymous, which are greatly thanked.

REFERENCES

Adam, D. (2002) Gravity measurement: amazing grace. Nature,416, 10^11.

Alvarez,P. (1995)Evidence for aNeoproterozoic carbonate rampon the northern edge of the Central African Craton: relationswith Late Proterozoic intracratonic troughs. Geol. Rund., 84,636^648.

Alvarez, Ph.,Maurin, J.-C.&Vicat, J.-P. (1995) La Formationde l’Inkisi (SupergroupeOuest-congolien) en Afrique centrale(Congo et Bas-Za|« re): un delta d’a“ ge Pale¤ ozo|« que comblant unbassin en extension. J. Afr. Earth Sci., 20, 119^131.

Ayele,A. (2002) Active compressional tectonics in central Africaand implications for plate tectonic models: evidence from faultmechanism studies of the 1998 earthquakes in the Congo Ba-sin. J. Afr. Earth Sci., 35, 45^60.



Armstrong,R.A.,Master,S.&Robb,L.J. (2005)Geochronol-ogy of the NchangaGranite, and constraints on the maximumage of the Katanga Supergroup, Zambian Copperbelt. J. Afr.Earth Sci., 42, 32^40.

Batumike, M.J., Cailteux, J.L.H. & Kampunzu, A.B. (2007)Lithostratigraphy, basin development, base metal deposits,and regional correlations of the Neoproterozoic Nguba andKundelungu rock successions, central African Copperbelt.GondwanaRes, 11, 432^447.

Batumike,M.J.,Kampunzu,A.B.&Cailteux, J. (2006) Petrol-ogy and geochemistry of the Neoproterozoic Nguba andKun-delungu Groups, Katangan Supergroup, southeast Congo:implications for provenance, paleoweathering and geotectonicsetting. J. Afr. Earth Sci., 44, 97^115.

Bertrand-Sarfati, J. (1972) Stromatolites columnaires de cer-taines formations carbone¤ te¤ es du Pre¤ cambrien supe¤ rieur dubassin congolais (Bushimay, Lindien, Oest-Congolien). Ann.Mus. Roy. Congo belge, Tervuren (Belgique), se¤ rie in-8, Sci. Ge¤ ol.,74, 45pp.

Boudzoumou, F. & Trompette, R. (1988) La cha|“ ne panafri-caine ouest-congolienne au Congo (Afrique Equatoriale): unsocle polycyclique charrie¤ sur un domaine snbautochtoneforme¤ par l’aulacoge' ne du Mayombe et le bassin du l’Ouest-Congo.Bull. Soc. Ge¤ ol. France, 8(6), 889^896.

Boulouard, C. & Calandara, F. (1963) Etude palynologiquede quelques sondages de la Re¤ publique du Congo (Congo ex-belge). S.N.P.A., Direction Exploration & Production, Unpublishedreport, 6p. & ann.

Boutakoff, N. (1948). Les formations glaciaires et post gla-ciaires fossilife' res d’a“ ge permo-carbonife' re (Karoo infe¤ rieur)de la re¤ gion de Walikale (Kivu, Congo belge).Mem. Inst. Geol.Univ. Louvain, 9, 124pp.

Braitenberg, C. & Ebbing, J. (2009) The GRACE-satellitegravity and geoid ¢elds in analysing large-scale, cratonic or in-tracratonic basins.Geophys. Prospect., 57(4), 559^571.

Burke,K.&Gunnell,Y. (2008) TheAfrican erosion surface: acontinental- scale synthesis of geomorphology, tectonics, andenvironmental change over the past 180 million years. Geol.Soc. Am.Mem., 201, 1^66.

Burke,K.,MacGregor,D.&Cameron,N. (2003)Africa’s pet-roleum systems: four tectonic Aces in the past 600 millionyears. In:PetroleumGeology of Africa:NewThemes andDevelopingTechnologies (Ed. byT. Arthur, D. MacGregor & N. Cameron),Geol. Soc. Lond. Spec. Publ., 207, 21^60.

Cahen,L. (1954)Ge¤ ologie duCongo belge.Vaillant-Caramanne,Lie' ge, 577pp.

Cahen,L. (1963)Glaciations anciennes et de¤ rive des continents.Ann. Soc. Ge¤ ol. Belgique, 86, 19^84.

Cahen,L. (1970) Igneous activity andmineralization episodes inthe evolution of the Kibaride and Katangide orogenic belts ofCentral Africa. In: AfricanMagmatism and Tectonics (Ed. byT.N.Cli¡ord & I.G. Gass), pp.97^117. Oliver and Boyd.

Cahen, L. (1982) Geochronological correlation of the late Pre-cambrian sequences on and around the stable zones of Equa-torial Africa. Precambrian Res., 18, 73^86.

Cahen,L., Ferrand, J.J.,Haarsma,M.J.F.,Lepersonne, J. &Vebeek,Th. (1959) Description du Sondage de Samba. Ann.Mus. Roy. Congo belge, Tervuren (Belgique), se¤ rie in-8, Sci. Ge¤ ol.,29, 210pp.

Cahen,L., Ferrand, J.J.,Haarsma,M.J.F.,Lepersonne, J. &Vebeek,Th. (1960) Description du Sondage de Dekese. Ann.Mus. Roy. Congo belge,Tervuren (Belgique), se¤ rie in-8, Sci. ge¤ ol., 34,115pp.

Cahen, L., Ledent, D. & Villeneuve, M. (1979) Existenced’une cha|“ ne plisse¤ e Prote¤ rozo|« que supe¤ rieur auKivu oriental(Za|« re). Donne¤ es ge¤ ochronologiques relatives au Supergroupede l’Itombwe.Bull. Soc. belge Ge¤ ol., 88, 71^83.

Cahen, L. & Lepersonne, J. (1954) Etat actuel des connais-sances relatives aux se¤ ries me¤ sozoiques et l’inte¤ rieur du Con-go.Bull. Soc. Belge. Ge¤ ol., 77, 20^37.

Cahen,L.&Lepersonne, J. (1978) Synthe' se des connaissancesrelatives au Groupe (anciennement Se¤ rie) de la Lukuga (Per-mien du Zaire). Ann. Mus. Roy. Congo belge, Tervuren (Belgique),se¤ rie in-8, Sci. ge¤ ol., 82, 115^152.

Cahen, L., Snelling, N.J., Delhal, J., Vail, J.R., Bon-

homme, M. & Ledent, D. (1984) Geochronology and Evolutionof Africa. Clarendon, Oxford, 512pp.

Cailteux, J. & Kampunzu, A.B. (1995) The Katangan tectonicbreccias in the Shaba province (Zaire) and their genetic signif-icance. Ann. Mus. Roy. Afr. Cent., Tervuren (Belgique), se¤ rie in-8,Sci. ge¤ ol., 101, 63^76.

Cailteux, J.,Kampunzu,A.B.,Lerouge,C.,Kaputo,A.K. &Milesi, J.P. (2005) Genesis of sediment-hosted stratiformcopper-cobalt deposits, central African Copperbelt. J. Afr.Earth Sci, 42, 134^158.

Cambier, R. (1930) Contribution a' l’e¤ tude des bassins houillersde la Luena. Ann. Serv. Des Mines, Comite¤ Spe¤ cial du Katanga, 1,3^54.

Casier,E. (1961)Mate¤ riaux pour la faune ichtyologique e¤ ocre¤ ta-cique du Congo. Ann. Mus. Roy. Congo belge,Tervuren (Belgique),se¤ rie in-8, Sci. ge¤ ol., 39, 96pp.

Catuneanu, O.,Wopfner, H., Eriksson, P.G., Cairncross,B., Rubidge, B.S., Smith R, M.H. & Hancox, P.J. (2005)The Karoo basins of south-central Africa. J. Afr. Earth Sci.,43, 211^253.

Chorowicz, J., Le Fournier, J. & Makazu, M.M. (1990) LaCuvette Centrale du Za|« re: un bassin initie¤ au Prote¤ rozo|« quesupe¤ rieur. Contribution de l’analyse du re¤ seau hydrographi-que.C. R. Acad. Sci. Paris, 311, II, 349^356.

Collins,A.S.&Pisarevsky,S.A. (2005) Amalgamating easternGondwana: The evolution of the Circum-Indian Orogens.Earth Science Reviews, 71, 229^270.

Cornet, J. (1894) Les formations post-primaires du Bassin duCongo.Ann. Soc. Geol. Belg., 21(1893^1894), 251^258.

Cornet, J. (1911) Sur la possibilite¤ de l’existence de gisementsde pe¤ trole au Congo. Ann. Soc. Geol. Belg., 36(1910^1911),9^15.

Crosby, A.G., Fishwick, S. &White,N. (2010) Structure andevolution of the intracaraonic Congo Basin. Geochem., Geo-phys., Geosyst., 11, Q06010; doi: 10.1029/2009GC003014.

Daly,M.C.,Lawrence,S.R.,Diemu-Thiband,K.&Matoua-

na,B. (1992) Tectonic evolution of the Cuvette Centrale, Zaire.J. Geol. Soc. Lond., 149, 539^546.

Daly,M.C., Lawrence, S.R.,Kimun’a,D. & Binga,M. (1991)LatePaleozoic deformation in central Africa: a result of distantcollision?Nature, 350, 605^607.

Delpomdor, F., Pre¤ at, A. & Tack, L. (2008) Paleoenviron-ments of Neoproterozoic carbonates in part of the Schisto-Calcaire subgroup,West Congo Belt, Democartic Republic ofCongo (DRC): Lithostratigraphy, sedimentology and diagen-esis. 22nd Colloquium of African Geology, November 4^6, 2008,Hammamet ^Tunisia, Abstract book, p.190.

Delpomdor, F. & Tack, L. (2007) Etude des de¤ po“ ts diamicti-tiques du Ne¤ oprote¤ rozo|« que supe¤ rieur en RDC et au SW duRwanda. Royal Museum for Central Africa, Tervuren, Bel-gium, unpublished report, 46p.


E. Kadimaet al.

10.1029/2009GC003014

Delvaux, D. (2001a) Tectonic and palaeostress evolution of theTanganyika-Rukwa-Malawi rift segment, East African RiftSystem. In: Peri-Tethys Memoir 6: PeriTethyan Rift/Wrench Basinsand Passive Margins (Ed. by P.A. Ziegler, W. Cavazza, A.H.F.Robertson & S. Crasquin-Soleau), Me¤ m. Mus. National Hist.Nat., Paris 186, 545^567.

Delvaux,D. (2001b) Karoo rifting in westernTanzania: precur-sor of Gondwana breakup? In: Contributions to Geology and Pa-leontology of Gondwana. In honour of Prof. Dr. HelmutWopfner, pp.111^125, Cologne.

Delvaux,D.&Barth,A. (2010) African stress pattern from for-mal inversion of focal mechanism data. Implications for riftingdynamics.Tectonophysics, 482, 105^128.

Delvaux, D., Kervyn, R.,Vittori, E., Kajara, R.S.A. & Ki-

lembe, E. (1998) Late quaternary tectonic activity and lake le-vel £uctuation in the Rukwa rift basin, East Africa. J. Afr. EarthSci., 26, 397^421.

De Sainte Seine, P. (1955) Poissons fossiles de l’e¤ tage de Stan-leyville (Congo belge). Premie' re partie: la faune des argiliteset schistes bitumineux.Ann.Mus. Roy. Afrique centrale,Tervuren(Belgique), se¤ rie in-8, Sci. ge¤ ol., 14, 125p.

DeWaele,B., Johnson,S.P.&Pisarevsky,S.A. (2008) Palaeo-proterozoic to Neoproterozoic growth and evolution of theeastern Congo Craton: its role in the Rodinia puzzle. Precam-brian Res., 160, 127^141.

deWit,M.J.&Ransome, I.G.D., Editors (1992) InversionTectonics ofthe Cape Fold Belt, Karoo and Cretaceous Basins of Southern Africa.Balkema, Rotterdam, 269pp.

Downey,N.J. &Gurnis,M. (2009) Instantaneous dynamics ofthe cratonic Congo Basin. J. Geophys. Res., B114(B6), B06401.Doi: 2008JB006066R.

Dumont, P. & Hanon, M. (2000) La stratigraphie du Kunde-lungu supe¤ rieur au Plateau des Kundelungu. Royal Museumfor Central Africa,Tervuren, Belgium, unpublished report, 21pp.

Dumont, P., Hanon, M. & Ngoyi, L.K. (1997) La structurege¤ ologique du plateau des Biano (Shaba, Za|« re).Mus.Roy. Afri-que centrale,Tervuren (Belgique), De¤ p. Ge¤ ol-Min., Rapp ann, 1995^1996, 169^182.

Dupont,E. (1889)Lettres sur le Congo.Re¤ cit d’unvoyage scien-ti¢que entre l’embouchure du Fleuve et la con£uent du Kasa|« .Ed. by C.R. Paris). 3 (pp. 398^403 724p. Summary Report in:Bull.Soc. belge Ge¤ ologie.

ECL. (1988)Hydrocarbon Potential of Cuvette Centrale (Repub-lic of Zaire). Exploration Consultants Limited, CelluleTechni-que Pe¤ trolie' re, Pe¤ trozaire, unpublished report, 41pp.1¢gures,tables, appendices and enclosures.

Egorov, A. & Lombard, A.L. (1962) Pre¤ sence des couches deStanleyville dans le sous-sol de Le¤ poldville, Re¤ publique duCongo. Note pre¤ liminaire.Ann. Soc. Geol. Belg., 85, 103^109.

Esso Za|« re SARL. (1981a) Geological Completion Report.Mbandaka-1. Unpublished internal report.

Esso Za|« re SARL. (1981b) Geological Completion Report. Gil-son-1. Unpublished internal report.

Evrard, P. (1957) Les recherches ge¤ ophysiques et ge¤ ologiques etles travaux de sondage dans la Cuvette congolaise. Acad. Roy.Sci. Coloniale., Sc.Techn. Bruxelles,VII(1), 62pp.

Evrard, P. (1960) Sismique. (Re¤ sultat scienti¢que des missionsdu Syndicat pour l’e¤ tude ge¤ ologique et minie' re de Cuvettecongolaise). Ann. Mus. Roy. Afrique centrale, Tervuren (Belgique),se¤ rie in-8, Sci. ge¤ ol., 33, 87p.

Eyeles, N. & Januszczak, N. (2004) ‘Zipper-rift’: a tectonicmodel for the Neoproterozoic glaciations during the breakupof Rodinia after 750Ma. Earth Sci. Rev., 65, 1^73.

Forte,A.M.,Que¤ re¤ ,S.,Moucha,R.,Simmons,N.A.,Grand,S.P., Mitrovica, J.X. & Rowley, D.B. (2010) Joint seismic-geodynamic-minral physical modelling of African geodyna-mics: a reconciliation of deep-mantle convection with surfacegeophysical constraints. Earth Planet. Sci. Lett., 295, 329^341.

Fourmarier,P. (1914) Le bassin charbonnier d’a“ ge Permo-Trias-sique de la Lukuga. Ann. Soc. Ge¤ ol. Belg., Pub. rel. Congo belge,41(1913^1914), C77^C227.

Franc� ois,A. (1974) Stratigraphie, tectonique et mine¤ ralisationsdans l’Arc Cuprife' re du Shaba (Re¤ p. Du Za|« re). In: Gisementsstratiformes et provinces cuprife' res (Ed. by P. Bartholome¤ ), pp. 79^101. Centenaire de la Socie¤ te¤ Ge¤ ologique de Belgique, Lie' ge.

Frimmel, H.,Tack, L., Basei, M., Nutman, A. & Boven, A.(2006) Provenance and chemostratigraphy of the Neoprotero-zoic West Congolian Group in the Democratic Republic ofCongo. J. Afr. Earth Sci., 46, 221^239.

Ge¤ rard, G. (1958) Carte ge¤ ologique de l’Afrique e¤ quatorialefranc� aise au 1/2.000.000, Note explicative. Gouvernementge¤ ne¤ ral de l’Afrique e¤ quatoriale franc� aise,Direction desMineset de la Ge¤ ologie, Paris, 198pp.

Giresse, P. (2005)Mesozoic-Cenozoic history of the Congo Ba-sin. J. Afr. Earth Sci., 43, 301^315.

Grekoff, N. (1957) Ostracodes du basin du Congo. I. Jurassiquesupe¤ rieur etCre¤ tace¤ infe¤ rieur du nord du basin.Ann.Mus.RoyalAfrique centrale,Tervuren (Belgique), se¤ rie in-8, Sci. ge¤ ol., 19, 97p.

Guiraud, R. & Bosworth,W. (1997) Senonian basin inversionand rejuvenation of rifting in Africa and Arabia: synthesisand implications for plate- scale tectonics.Tectonophysics, 282,39^82.

Guiraud, R., Bosworth,W.,Thierry, J. & Delplanque, A.(2005) Phanerozoic geological evolution of Northern and cen-tral Africa: an overview. J. Afr. Earth Sci., 43, 83^143.

Haddon, I.G. & McCarthy,T.S (2005) The Mesozoic-Ceno-zoic interior sag basins of Central Africa: the Late-Cretac-eous-Cenozoic Kalahari and Okavango basins. J. Afr. EarthSci., 43, 316^333.

Haest,M.,Muchez, Ph.,Dewaele, S.,Franey N&Tyler,R.(2007) Structural Control on the Dikulushi Cu-Ag Deposit,Katanga, Democratic Republic of Congo. Econ. Geol., 102,1321^1333.

H�lbich, I.W., Fitch, F.J. & Miller, J.A. (1983) Dating theCape orogeny.Geol. Soc. S. Africa, Spec. Pub., 12, 149^164.

Hartley, R. & Allen, P.A. (1994) Interior cratonic basins ofAfrica: relation to continental berakup and role of mantle con-vection.Basin Res., 6, 95^113.

Hoffman, K.H., Condon, D.J., Bowring, S.A. & Crowley,J.L. (2004)U-Pb zircon dates from theNeoproterozoicGhaubformation, Namibia: constraints on the Marinoan glaciation.Geology, 32(9), 817^820.

Hoffman,P.F. (1999)The breakup ofRodinia, birth ofGondwa-na, true ploar wander and the snowball Earth. J. Afr. Earth Sci.,28, 17^34.

Hoffman, P.F. & Li, Z.-X. (2009) A palaeogeographic contextfor the Neoproterozoic glaciation. Paleogeogr. Palaeoclim. Pa-laeoecol., 277, 158^172.

Hoffman, P.F. & Schrag, D.P. (2002) The snowball earth hy-pothesis: testing the limits of global change. Terra Nova, 14,129^155.

JNOC. (1984) Rapport des investigations ge¤ ophysiques et ge¤ olo-giques dans la Cuvette centrale de l’ Re¤ publique du Za|« re.Japan National Oil Corporation, Report for Departmentof Mines and Energy, Government of Zaire, Unpublished,205pp.



2008JB006066R

Jackson, M.P.A.,Warin, O.N.,Woad, G.M. & Hudec, M.R.

(2003) Neoproterozoic allochtonous salt tectonics during theLu¢lian orogeny in the Katangan Copperbelt, central Africa.Geol. Soc. Am. Bull., 115(3), 314^330.

John,T., Schenk,V., Mezger, K. & Tembo, F. (2004) Timingand PTevolution ofwhiteschist metamorphism in the Lu¢lianarc-Zambezi belt orogen (Zambia): implications for the assem-bly of Gondwana. J. Geol., 112, 71^90.

Johnson, M.R., van Vuuren, C.J., Visser, J., Cole, D., deV.Wickens, H., Christie, A., Roberts, D. & Brandl, G.

(2006) Sedimentary rocks of the Karoo supergroup. In: TheGeology of South Africa (Ed. byM.R. Johnson, C.R. Anhaeusser& R.J. Thomas), pp. 461^499. Geol Soc. S. Africa, Johannes-burg, and Council for Geoscience, Pretoria.

Jones, L.,Mathieu, P.L. & Strenger,H. (1960) Gravime¤ trie:Les re¤ sultats scienti¢ques des missions du syndicat pourl’e¤ tude ge¤ ologique et minie' re de la Cuvette Congolaise et tra-vaux connexes. Ann. Mus. Roy. Congo belge, Tervuren (Belgique),se¤ rie in-8, Sci. ge¤ ol., 36, 46pp.

Kadima, E. (2007). Sismostratigraphie, structure ge¤ ologique etprospectivite¤ pe¤ trolie' re du bassin de la Cuvette congolaise(RDCongo).Master thesis, University of Lubumbashi, 126pp.

Kadima, E., Sebagenzi, S. & Lucazeau, F. (2010) Analysis ofProterozoic rifting and subsequent subsidence of the CentralCongo Basin. Geophysical Research Abstract, 12, EGU2010-11438-1.

Kadima, E.K., Sebagenzi, S.M.N. & Lucazeau, F. (2011) AProterozoic-rift origin for the structure and the evolution ofthe cratonic Congo basin. Earth Planet. Sci. Lett., In press.DOI: 10.1016/j.epsl.2011.01.037.

Kanda-Nkula,V.,Tack, L., Fernandez-Alonso,M., Fran-ceschi,G.&Frimmel,H. (2004).The PanAfricanWest Con-go and Katanga thrust and fold belts and their forelanddomains: similarities and di¡erences in Neoproterozoic basinevolution and mineralization. 20th Colloquium on AfricanGeology, Orle¤ ans, June 2004, Abstracts volume, p. 227.

Kennedy,W.Q. (1964) The Structural Di¡erentiation of Africain the Pan-African (1/� 500m.y.) tectonic episode. ResearchInstitute for African Geology, LeedsUniv. 8th Ann. Rep., 48^49.

King, L.C. (1963) South African Scenery. Oliver and Boyd, Edin-burgh, London, 308pp.

Kr˛ner, A., Collins, A.S., Hegner, E.,Willner, A.P.,Mu-

hongo,S.&Kehelpannala,K.V.W. (2001)The East AfricanOrogen: New Zircon and Nd Ages and implications for Rodi-nia and Gondwana supercontinent formation and dispersal.GondwanaRes., 4, 179^181.

Lawrence, S. & Makazu, M.M. (1988) Zaire’s Central basin:prospectivity outlook.Oil Gas J., 86(38), 105^108.

Lepersonne, J. (1974a) Carte ge¤ ologique du Za|« re au 1/2.000.000. Re¤ publique du Za|« re, Direction de la Ge¤ ologie, Kinshasa&Mus. Roy. Afr. Centr.,Tervuren.

Lepersonne, J. (1974b)Notice explicative de laCarte ge¤ ologiquedu Za|« re au1/2.000.000.,Re¤ publique duZa|« re, Direction de laGe¤ o-logie, Kinshasa&Mus. Roy. Afr. Centr.,Tervuren.

Lepersonne, J. (1977) Structure ge¤ ologique du bassin inte¤ -rieur du Za|« re. Bull. Acad. Roy. Belg., Cl. Sci., 5e se¤ rie, 63(12),941^965.

Le Roux, J.P. (1995)Heartbeat of a mountain: diagnosing the ageof depositional events in the Karoo (Gondwana) Basin fromthe pulse of the Cape Orogen.Geol. Rund., 84, 626^635.

Lombard, A.L. (1960) L’extension me¤ ridionale des couches deStanleyville, Congo belge (Note pre¤ liminaire). Bull. Soc. Ge¤ ol.Belg., 69, 23^27.

Master, S., Rainaud, C., Armstrong, R.A., Phillips, D.& Robb, L.J. (2005) Provenance ages of the Neoprotero-zoic Katanga Supergroup (Central African Copperbelt),with implications for basin evolution. J. Afr. Earth Sci., 42, 41^60.

Milesi, J.P.&Toteu,S.F., etal (2006)An overview of the geologyand major ore deposits of Central Africa: explanatory note forthe 1:4,000,000 map (Geology and major ore deposits of Cen-tral Africa. J. Afr. Earth Sci., 44, 571^595.

Newton, A.R., Shone, R.W. & Booth, P.W. (2006) The Capefold belt. In:The Geology of South Africa (Ed. by M.R. Johnson,C.R. Anhaeusser & R.J. Thomas), pp. 521^530. GeolSoc. S. Africa, Johannesburg, and Council for Geoscience,Pretoria.

Nyblade, A.A. & Robinson, S.W. (1994) The African super-swell.Geophys. Res. Lett., 21, 765^768.

Passau,G. (1923) La ge¤ ologie du bassin des schistes bitumineuxde Stanleyville (Congo belge).Ann. Soc. ge¤ ol. Belg., Publ rel. Con-go belge, 19(1921^1922), C91^C243.

Pasyanos,M.E.&Nyblade,A.A. (2007) A top to bottom litho-spheric study of Africa and Arabia.Tectonophysics, 444, 27^44.

Pedrosa-Soares, A.C., Alkmin, F.F.,Tack, L., Noce, C.M.,Babinski, M., Silva, L.C. & Martins-Neto, M.A. (2008)Similarities and di¡erences between the Brazilian andAfrican counterparts of the Neoproterozoic Arac� uai-WestCongo Orogen. In: West Gondwana: Pre-Cenozoic CorrelationsAcross the South Atlantic Region (Ed. by R.J. Pankhurst, R.A.J.Trouw, B.B. Brito Neves & M.J. De Wit), Geol. Soc. Lond. Spec.Publ., 294, 153^172.

Poidevin, J.-L. (1985) Le Prote¤ rozo|« que supe¤ rieur de la Re¤ pub-lique Centrafricaine. Ann. Mus. royal Afrique centrale, Tervuren(Belgique), se¤ rie in-8, Sci. ge¤ ol., 91, 75p.

Poidevin, J.-L. (2007) Stratigraphie isotopique du strontium etdatations des formations carbonate¤ es et glacioge¤ niques ne¤ o -prote¤ rozo|« ques du Nord et de l’Ouest du craton du Congo. C.R. Geosci., 339, 259^273.

Porada,H. (1989) Pan-African rifting and orogenesis in south-ern to Equatorial Africa and Eastern Brazil. Precambrian Res.,44, 103^136.

Porada,H.&Berhorst,V. (2000)Towards a new understandingof the Neoproterozoic-Early Paleozoic Lu¢lian and northernZambezi belts inZambia and theDemocraticRepublic ofCon-go. J. Afr. Earth Sci., 30, 727^771.

Pre¤ at, A., Kolo, K., Piran, J.-P. & Delpomdor, F. (2010) Aperitidal evaporite environment in the Neoproterozoic ofSouth Gabon (Schisto-Calaire Subgroup, Nyanga Basin). Pre-cambrian Res., 177(3^4), 253^265.

Prieto, C. (1996) Gravity/magnetic signatures of various geolo-gic models ^ an exercise in pattern recognition. IGC FootnoteSeries, 4(4), 1^24.

Rainaud, C., Master, S., Armstrong, R.A., Phillips, D. &Robb, L.J. (2005) Monazite U-Pb dating and 40Ar-39Ar ther-mochronology of metamorphic events in the Central AfricanCopperbelt during the Pan-African Lu¢lian Orogeny. J. Afr.Earth Sci., 42, 183^199.

Raucq,P. (1957) Contribution a' la connaissance duSyste' me de laBushimay. Ann. Mus. Roy. Afr. Cent.,Tervuren (Belgique), se¤ rie in-8, Sci. ge¤ ol., 18, 427p.

Ring, J.P. (1995)Tectonic and lithological constraints on the evo-lution of the Karoo graben of northern Malawi (East Africa).Geol. Rund., 84, 607^625.

Ritsema, J. & van Heijst,H.J. (2000) New seismic model of theupper mantle benearth Africa.Geology, 28, 63^66.


E. Kadimaet al.

Robert,M. (1956) Ge¤ ologie et Ge¤ ographie du Katanga, y com-pris l’e¤ tude des ressources et de la mise envaleur.Ed.M.Hayez,Bruxelles, 621pp.

Rowan, M.G. & Vendeville, B.C. (2006) Foldbelts with earlysalt withdrawal and diapirism: physical model and examplefrom the northern Gulf of Mexico and the £inders ranges,Australia.Mar. Petrol. Geol., 23, 871^891.

Selley, R.C. (1997) African basins. In: Sedimentary Basins of theWorld, 3 (Ed. by K.J. Hsˇ) Elsevier, Amsterdam. 394pp.

Se¤ ranne, M. & Anka, Z. (2005) South Atlantic continentalmargins of Africa: a comparison of the tectonic vsv climate in-terplay on the evolution of equatorialwest Africa andSWAfri-ca margins. J. Afr. Earth Sci., 43, 283^300.

Shields,G.A. (2006) Snowball earth is dead! Long live snowballearth!. Episodes, 29(4), 287^288.

Sluys, M. (1945) La ge¤ ologie de l’Ituri. Le groupe de la Lindi.Bull. Serv. Ge¤ ol. Congo belge et Rwanda-Urundi (Le¤ opoldville), 1,95^182.

Stough, J.B. (1965) Paleozoic andMesozoic palynomorphs fromthe Republic of Congo. Esso, Stratigraphic and StructuralGeologyDivision, Unpublished report, 4pp.

Tack, L., Delvaux, D., Kadima, E., Delpomdor, F.,Tahon,A., Dumont, P., Hanon, M., Fernandez-Alonso, M.,Baudet, D., Dewaele, S., Cibambula, E., Kanda Nkula,V.&Mpiana,Ch. (2008)The1.000m thick Redbeds sequenceof the Congo River Basin (CRB): a generally overlooked testi-mony in Central Africa of post-Gondwana amalgamation(550Ma) and pre-Karoo break-up (320Ma). 22nd ColloquiumAfrican Geology (CAG22), November 04-06, 2008, Hamma-met,Tunisia, 86^88.

Tack, L.,Wingate,M.T.D., De Waele, B.,Meert, J., Belou-sova, E., Griffin, B., Tahon, A. & Fernandez-Alonso,M. (2010)The1375Ma‘‘Kibaran event’’ inCentral Africa: pro-minent emplacement of bimodal magmatism under exten-sional regime. Prec. Res, in press, doi: 10.1016/j.precamres.2010.02.022.

Tack, L., Wingate, M.T.D., Lie¤ geois, J.-P., Fernandez-

Alonso, M. & Deblond, A. (2001) Early Neoproterozoicmagmatism (100^910Ma) of the Zadinian and MayumbianGroups (Bas-Congo): onset of Rodinia rifting at the westernedge of the Congo craton. Prec. Res., 110, 277^306.

Tait, J.,Delpomdor,F.,Pre¤ at,A.,Tack,L.,Straathof,G.&Nkula, V.K. (2010) Neoproterozoic sequences of the WestCongo and Lindi/Ubangi supergroups in the Congo Craton,Central Africa. In:TheGeological Record ofNeoproterozoic Glacia-tions (Ed. by E. Arnaud, G.P. Halverson &G.A. Shields-Zhon),Geol. Soc. Lond,Mem., in press.

Talwani, M.,Worzel, J.-L. & Landisman, M. (1959) Rapidgravity computations for two dimensional bodieswith applica-tion to theMendocino fracture zone. J.Geophys.Res., 64, 49^59.

Tankard,A.,Welsink,H.,Aukes, P.,Newton,R.& Settler,E. (2009) Tectonic evolution of the Cape and Karoo basins ofSouth Africa.Marine Petrol. Geol., 26, 1379^1412.

Thibaut, P.-M. (1983) Synthe' se des travaux de la conventionZa|« re ^ BRGM1969^1980, tome1,Ge¤ ologie etMine¤ ralisations

de la re¤ gion de l’Equateur.De¤ partement desMines de la re¤ publiquedu Za|« re, 163pp.

Tohver, E., D’Agrella-Filho, M.S. & Trinidade, R.L.F.(2006) Paleomagnetic record of Africa and South America forthe 1200-500Ma interval, and evaluation of Rodinia andGondwana assemblies. Precambrian Res., 147, 193^222.

Trouw,R.A.&DeWit,M.J. (1999) Relation between theGond-wanide Orogen and contemporaneous intracratonic deforma-tion. J. Afr. Earth Sci., 28, 203^213.

Unocal. (1986) Density andMagnetic Susceptibility in samplesfrom the Congo Basin, Zaire. Internal report Unocal Scienceand Technology Division.

Veatch,A.C. (1935) Evolution of the Congo Basin.Geol. Soc. Am.Mem., 3, 184pp.

Vendeville,B.C.&Nilsen,K.T. (1995) Episodic growth of saltdiapirs driven by horizontal shortening. In: Salt, Sediment, andHydrocarbons (Ed. by C.J. Travis, H. Harrison, M.R. Hudec,B.C.Vendeville, F.J. Peel & B.F. Perkins), pp. 285^295. SPEMGulf Coast Section16th annual Research Foundation Conference.

Vendeville, B.C. & Rowan, M.G. (2002) 3-D kinematics ofminibasins and salt ridges remobilized by late contraction:physical models and seismic example (SE Mississipi Canyon,Gulf of Mexico). AAPGAnnual Meeting O⁄cial Program, A182-A183.

Verbeek,T. (1970) Ge¤ ologie et lithologie du Lindien (Pre¤ cam-brien Supe¤ rieur du nord de la Re¤ publique De¤ mocratique duCongo). Ann. Mus. Roy. Afr. Cent.,Tervuren (Belgique), se¤ rie in-8,Sci. ge¤ ol., 66, 311pp.

Vicat, J.-P.&Vellutini,P.-J. (1987) Sur la nature et la signi¢ca-tion des dole¤ rites du bassin pre¤ cambrien de Sembe-Ouesso(Re¤ publique du Congo). Prec. Res., 37, 57^69.

Visser, J.N.J.& Praekelt,H.E. (1996) Subduction, mega-shearsystems and the late Paleozoic basin development in the Afri-can segment of Gondwana.Geol. Rund., 85, 632^646.

Walemba, K. & Master, S. (2005) Neoproterozoic diamictitesfrom the Itombwe Synclinorium, Kivu province, democraticRepublic of Congo: Palaeoclimatic signi¢cance and regionalcorrelations. J. Afr. Earth Sci., 42, 200^210.

W˛pfner,H. (1994) The Malagasy Rift, a chasm in theTethyanmargin of Gondwana. J. SE Asian Earth Sci., 9, 451^461.

Wu, S.A., Bally, A.W. & Cramez, C. (1990) Allochtonous salt,structure and stratigraphy of the northeasternGulf ofMexico,Part II: structure.Mar. Petrol. Geol., 7, 334^3370.

Ziegler, P.A. (1993) Plate-moving mechanisms: their relativeimportance. J. Geol. Soc. Lond., 150, 927^940.

Zouaghi,T., Be¤ dir,M.& Inoubli,M.H. (2005) 2D Seismic in-terpretation of strike-slip faulting, salt tectonics, and Cretac-eous unconformities, Atlas Mountains, central Tunisia. J. Afr.Earth Sci., 43, 464^486.

Manuscript received 21 April 2010; Manuscript accepted8 December 2010.



10.1016/j.precamres.2010.02.022

10.1016/j.precamres.2010.02.022

structureandgeologicalhistoryof thecongobasin ... · structureandgeologicalhistoryof thecongobasin:...

Documents