Sedimentary patterns and geometries of the Bahamian outer carbonate ramp (MioceneLower Pliocene, Great Bahama Bank)

Sedimentary patterns and geometries of the Bahamian outercarbonate ramp (Miocene±Lower Pliocene, Great Bahama Bank)

CHRISTIAN BETZLER*, JOHN J. G. REIJMER , KARIN BERNETà , GREGOR P. EBERLIà andFLAVIO S. ANSELMETTI§*Geologisch-PalaÈontologisches Institut, Senckenberganlage 32±34, 60054 Frankfurt am Main, Germany(E-mail: [email protected]) Geomar, Wischhofstr. 1±3, 24148 Kiel, Germany (E-mail: [email protected])àRSMAS-MGG, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149,USA (E-mail: [email protected], [email protected])§Geologisches Institut ETHZ, Sonneggstrasse 5, 8092 ZuÈrich, Switzerland (E-Mail: ¯[email protected])

ABSTRACT

Core, logging and high-resolution seismic data from ODP Leg 166 were used to

analyse deposits of the Neogene (Miocene±Lower Pliocene) Bahamian outer

carbonate ramp. Ramp sediments are cyclic alternations of light- and dark-grey

wackestones/packstones with interbedded calciturbidite packages and minor

slumps. Cyclicity was driven by high-frequency sea-level changes. Light-grey

layers containing shallow-water bioclasts were formed when the ramp exported

material, whereas the dark-grey layers are dominantly pelagic. Calciturbidites

are arranged into mounded lobes with feeder channels. Internal bedding of the

lobes shows a north-directed shingling as a result of the asymmetrical growth of

these bodies. Calciturbidite packages occur below and above sequence

boundaries, indicating that turbidite shedding occurred during third-order sea-

level highstands and lowstands. Highstand turbidites contain shallow-water

components, such as green algal debris and epiphytic foraminifera, whereas

lowstand turbidites are dominated by abraded bioclastic detritus. Gravity ¯ow

depocentres shifted from an outer ramp position during the early Miocene to a

basin ¯oor setting during the late Miocene to early Pliocene. This change was

triggered by an intensi®cation of the strength of bottom currents during the

Tortonian, which was also responsible for shaping the convex morphology of

the outer ramp. The Miocene and Lower Pliocene of the leeward ¯ank of Great

Bahama Bank provides an example of the poorly known depositional setting of

the outer part of distally steepened carbonate ramps. The contrast between its

sedimentary patterns and the well-known Upper Pliocene±Quaternary slope

facies associations of the ¯at-topped Great Bahama Bank shows the strong

control that the morphology of a carbonate platform exerts on the depositional

architecture of the adjacent slope and base-of-slope successions.

Keywords Bahama carbonate platform, carbonate ramp, cyclicity, Neogene,sea-level changes, turbidites.

INTRODUCTION

Distally steepened ramps, in contrast to homo-clinal ramps, may bear major amounts of gravity¯ow deposits in the outer ramp facies (Burchette& Wright, 1992). However, little is known about

depositional geometries and sea-level-controlledstacking patterns of such ramps. Detailed coredata, logging data, excellent biostratigraphic con-trol and high-resolution seismic pro®les wereacquired before and during ODP Leg 166 for theNeogene outer ramp deposits of Great Bahama

Sedimentology (1999) 46, 1127±1143

Ó 1999 International Association of Sedimentologists 1127

Bank; these provide an ideal opportunity forinvestigating such a distally steepened rampsetting.

The aim is to reconstruct lateral and verticalfacies changes of this outer ramp and to show itssedimentary evolution. High-resolution seismicpro®les have enabled us to map the lateralextension and geometries of gravity ¯ow depos-its in detail. The signatures of third-order andhigher order sea-level ¯uctuations are discussed,and an overview of the sedimentary regime inthis largely uninvestigated depositional setting isprovided.

GEOLOGICAL SETTING

The Bahamas archipelago with Great BahamaBank consists of several carbonate platforms. TheCenozoic progradation of the leeward ¯ank ofGreat Bahama Bank was of the order of 27 km(Austin et al., 1986; Eberli & Ginsburg, 1987,1989). The platform growth occurred in pulsesduring sea-level highstands; each pulse resultedin an unconformity-bounded depositional se-quence, the boundaries of which were generatedduring sea-level lowstands. Material, which nowforms the slope of the bank, was largely providedby shallow-water carbonate particles produced onthe bank top. This sediment export pattern isintimately linked to the con®guration of GreatBahama Bank as a ¯at-topped platform duringLate Pliocene and Quaternary times (Eberli &Ginsburg, 1987, 1989).

Beach & Ginsburg (1980), Schlager & Ginsburg(1981) and Reijmer et al. (1992) have pointed out

that this recent con®guration and most of therelated processes acting on the Bahama carbonateplatform can only be projected back to the LatePliocene. Based on sediment compositional chan-ges in boreholes across the Great Bahama Bank,they postulated a Pliocene change in platformgeometry. During older stages of platform growth,the Bahama Bank had a different geometry,which, according to the above authors, wassimilar to a reef-rimmed atoll. The change ingeometry has been veri®ed by seismic data (Eberli& Ginsburg, 1987, 1989), which show a Plioceneturnover from a distally steepened ramp to a ¯at-topped platform.

The turnover in platform geometry is recordedin slope sediments at the Clino borehole (Fig. 1),located in the inner platform where sea-level-driven sedimentary cycles show only minorcompositional changes in the Pliocene rampstage. Bioclastic packstones characterize bothlowstand and highstand deposits. The ¯at-toppedplatform exported peloidal micrites during high-stands and coarse-grained sparites with calcitegrains, algae and carbonate lithoclasts duringlowstands (Kenter et al., 1999; Westphal et al.,1999). Two other boreholes also provide a limitedrecord from the ramp deposits (Fig. 1): ODP Site626 in the Straits of Florida (Austin et al., 1986;Schlager et al., 1988) and the Great Isaac Well onwestern Great Bahama Bank (Schlager et al.,1988). In the Straits of Florida drill site, Miocenesediments are winnowed carbonate sands andcarbonate debris ¯ows/calciturbidites. The GreatIsaac Well recovered deep-water slope anddebris-apron deposits of the platform. DuringODP Leg 166, seven sites were drilled at the

Fig. 1. (A) Location of ODP Leg 166drill sites on the leeward ¯ankof Great Bahama Bank. Position of100 m isobath delimiting areas ofshallow-water carbonates is shown.Square indicates position of ODPLeg 166 Site map (B). (B) Position ofboreholes from ODP Leg 166 andseismic lines studied.

1128 C. Betzler et al.

Ó 1999 International Association of Sedimentologists, Sedimentology, 46, 1127±1143

leeward ¯ank of Great Bahama Bank (Eberli et al.,1997a). Five sites (1003±1007) in the Straits ofFlorida (Figs 1 and 2) are analysed in this study.

METHODS

Descriptions of the lithologies rely on coredescriptions presented in Eberli et al. (1997a),as well as post-cruise thin-section, smear slide,scanning electron microscopy, X-ray powderdiffraction, carbonate content and geophysicallog data analyses. Pelagic carbonates are termedooze or chalks. For lithologies dominated byshallow-water components, the Dunham (1962)classi®cation was applied. In addition, threeclasses are differentiated: unlithi®ed, partiallylithi®ed and lithi®ed.

Large-scale geometries and seismic responsesof analysed lithologies were obtained from multi-channel, high-resolution seismic data (45 cubicinch GI-airgun, 50±500 Hz) shot on R/V Lone Starin co-operation with Rice University (Eberli et al.,1997a). The biostratigraphy of the presenteddeposits is extensively discussed in the sitereports of the individual drill holes by Eberliet al. (1997a).

LITHOFACIES ANDLITHOSTRATIGRAPHY

A preliminary delimitation of sediment types andsedimentary units in the drill holes was present-ed by Eberli et al. (1997a). Here, post-cruiseanalysis of thin sections and geophysical logsare incorporated. As such, the limits of somesedimentary units (Fig. 3) are shifted with respectto the initial descriptions, as additional calci-turbidite packages were identi®ed in the logs.

The sedimentary successions at the outer rampSites 1003, 1004, 1005 and 1007 are dominated bythe input of platform-derived components,whereas basinal Site 1006 shows more pelagic-dominated sedimentation. In the following,sedimentary units of the outer ramp sites arediscussed. For detailed descriptions of lithofa-cies, the reader is referred to Eberli et al. (1997a).Three large-scale sedimentary units could bedistinguished: (1) Oligocene to lowermostPliocene; (2) lower Pliocene; and (3) upperPliocene to Pleistocene.

Oligocene to lowermost Pliocene

A major element of the Oligocene to lowermostPliocene sedimentary unit 1 is an alternationof light-grey and dark-grey wackestones and

Fig. 2. Seismic pro®le for Line 106 and Western Line (above) and interpreted cross-section (below). Letters refer todepositional sequences (modi®ed after Eberli et al., 1997b).

Neogene Bahamian ramp 1129


packstones (Fig. 4). The light-grey wackestones/packstones are uncompacted throughout thesuccession and contain shallow-water allochems(such as Amphistegina, Cibicides and red algal

debris) and planktonic foraminifera. The dark-grey wackestones/packstones contain clay. AtSites 1003 and 1005, siliciclastic contents in thesesediments are between 2% and 10%, at Site 1007

Fig. 3. Synthesis of lithological successions at ODP Sites 1003±1007. Numbers refer to sedimentary units, letters tosequence boundaries. Vp: velocity log (km s±1); c ray: gamma ray log (CPS units). Age assignments after Eberli et al.(1997a).



between 10% and 20%. Carbonate componentsare ®ne-grained bioclasts, as well as planktonicand benthic foraminifera. No shallow-waterbioclasts occur in the dark-grey intervals. In mostcases, the rhythmic change between lithologies isgradational and symmetric (Fig. 4). However,there are also strongly asymmetric cycles, withinthe upper part of which an intensely bioturbatedhorizon occurs. Such layers are sharply overlainby an interval of light-grey wackestones/pack-stones that grade upwards into dark-greywackestones/packstones. The transition from the

dark-grey to the light-grey intervals in these casesis again gradational, as in the symmetric cycles.

Intercalated in these deposits are calci-turbidites and some slump horizons (Fig. 3).The turbiditic packstones to ¯oatstones containshallow-water bioclasts and planktonic foramini-fera; an overview of representative compositionalchanges in the calciturbidites is provided inFig. 5. Slumped deposits occur as isolated inter-vals. The slumps consist of contorted light-greyand dark-grey wackestones/packstones andcalciturbidites. A subdivision of sedimentary unit1 into different subunits (Fig. 3) relies on theoccurrence of hardgrounds/®rmgrounds, on theposition of calciturbidites and on coarsening-upward trends of the light-grey wackestones/packstones (Eberli et al., 1997a).

Lower Pliocene

At Sites 1003 and 1005, lower Pliocene sedimen-tary unit 2 consists of an interval with poorlydifferentiated unlithi®ed to partially lithi®edmudstones to wackestones (Fig. 3) withplanktonic and benthic foraminifera, calcareousnannoplankton, minor diatoms and radiolarians.Sediments in general contain more carbonatemud than the deposits of the underlying unit.At Site 1007 (Fig. 3), the corresponding interval isooze and chalk.

Upper Pliocene to Pleistocene

The upper Pliocene to Pleistocene sedimentaryunit 3 is a succession of unlithi®ed to partiallylithi®ed peloidal mudstones and wackestoneswith calciturbidites, slumps, ooze and chalk.Calciturbidites consist of unlithi®ed to partiallylithi®ed packstones to ¯oatstones with shallow-water bioclasts, blackened components and lit-hoclasts. The composition of slumped horizons ispolymict, with interbeds of wackestones, pack-stones and ¯oatstones.

SEDIMENT GEOMETRIES

Large-scale depositional geometries of the sedi-ments were analysed in seismic pro®les obtainedfrom the Leg 166 site survey and older industrialseismic data imaging the entire Bahama Transect(Western Line; Fig. 1). Sequence stratigraphicanalyses on these data were performed by Eberli& Ginsburg (1987, 1989) and Eberli et al.(1997a,b). Most of the seismic sequences were

Fig. 4. Core photograph showing alternation of light-grey and dark-grey wackestones/packstones (Site 1003,1078á9±1088á5 m below sea ¯oor; core 1003C-66R).



identi®ed below the Great Bahama Bank.Erosional truncations and onlap geometries de®nethe sequence boundaries. These geometries showthat the boundaries were produced during sea-level lowstands. Sequence boundaries are termed`A' to `Q', the overlying respective seismicsequences `a' to `q' (Eberli et al., 1997a,b). Calcu-lated depth positions of individual sequenceboundaries were presented in Eberli et al.(1997a); Fig. 3 shows the positions of these geo-physically de®ned limits in the individual holes.

In order to correlate depositional geometries, asobserved in the seismic lines, with the facies inthe drilled holes, the time±depth conversioncurves calculated from check-shot survey data(Eberli et al., 1997a) were used. These ensuredthat the position and lateral extension of indivi-dual sedimentological units could be tracedprecisely in the seismic pro®les. Here, emphasislies on turbidite and slump deposits. Theseinterpretations provide a minimum estimate ofthe occurrence of these facies in the Leg 166transect, because only the geometries in thelateral continuation of the facies drilled in theindividual holes were mapped. Using this meth-od, we are also aware that the seismic pro®lesonly image large-scale geometrical features (Eber-li et al., 1994; Sta¯eu & Schlager, 1995; Anselm-etti et al., 1997). Resolution of the seismic lines isin the order of 5±12 m.

To provide an overview of the changes in the¯ank dip of the platform, slope angles werecalculated (Table 1, Fig. 6). These angles refer to

the dip of the sequence boundaries. As Line 106crosses the slope of Great Bahama Bank at anangle of »48° to the strike of the slope, depths ofsequence boundaries were projected onto a lineperpendicular to the ¯ank before calculatingindividual angles. As such, these angles are truedip values.

Geometry of depositional sequences

Miocene and Pliocene slope angles reached amaximum value of »4° (Fig. 6). The distal part ofthe Early and Middle Miocene ramp (sequencesp to k) had a concave geometry, which changed toa convex surface during the Late Miocene andEarly Pliocene (sequences i to f ). During the LatePliocene, the turnover from a ramp geometry toa ¯at-topped platform occurred (Fig. 6). GreatBahama Bank aggraded during the formation ofsequences p±n (Figs 2 and 6). The overlyingsigmoidal sequences m±g are arranged in aprograding to of¯aping pattern. The uppersequences f±a ®nally record the increased steep-ening of the strongly prograding margin of GreatBahama Bank (Fig. 6).

An overview of the depositional geometries inLine 106 along the transect between Sites 1005and 1006 is provided in Fig. 7. The sedimentationalong the ¯ank of Great Bahama Bank in this lineis re¯ected in a stack of wedge-shaped deposi-tional sequences in the proximal part of the drill-hole transect. Distally, between Sites 1006 and1007, these prograding clinoforms inter®nger

Fig. 5. Thin-section point-countingdata (300 points) from samples ofupper Miocene sequence k at Site1003. Note that shallow-water indi-cators (green and red algae, intra-clasts) only occur in the uppermostsamples of the depositionalsequence. K and I are sequenceboundaries. Benth. hyal., hyalinebenthic foraminifera; benth. mil.,miliolid benthic foraminifera;benth. aggl., agglutinated benthicforaminifera.



with continuous re¯ections, which, in part, dis-play an onlap or downlap con®guration againstthe prograding margin deposits. According toEberli et al. (1997a), these deposits are driftdeposits, which are arranged into drift wedgesin some sequences (e.g. in sequence f ).

Sequences p to i are characterized by a slightbasinward decrease in thickness. In the slope partof these sequences, downlapping re¯ectionsoccur above the different sequence boundaries.The angle of slope increases gently upsectionthroughout this package (Table 1). Sequence hshows a similar trend of thickness distribution tothe foregoing sequences, but it dips with aslightly increased angle (2á6°). Sequence g hasan opposite trend in thickness distribution andshows an increase basinwards.

Sequence f has a complex internal geometry.Similar to sequence g, it thickens between Sites1007 and 1006. But in contrast to sequence g, inwhich some re¯ections can be traced from thedrift deposits to the distal part of the progradingbank deposits, major sequence-bounding andintrasequential unconformities occur. The uncon-formity at the lower bounding sequence boundaryis not visible in Line 106, but it is well displayedin Line 108 (Fig. 9). The boundary is character-ized by deeply incised canyons, which areoriented downslope. Locally, most of the under-lying sequence g was eroded. The erosion at the

upper sequence boundary (E), resulting in theformation of deeply incised canyons, is welldisplayed in Line 106 (Figs 7 and 8), in Line108 (Fig. 9) and in Line 102 (Fig. 10).

One of the intrasequential unconformities isshown in Fig. 7 between Sites 1006 and 1007. It ischaracterized by an onlap con®guration of re¯ec-tors of a slopeward-thinning body onto the distalpart of a prograding lower slope wedge. Anothersequence-internal unconformity is shown inFig. 8, a detailed view of Line 106 between Sites1005 and 1007. This unconformity is character-ized by erosional truncation of re¯ectors anddeeply incised canyons. The position of this levelin the boreholes corresponds to intervals withinsedimentary unit 2 (Sites 1003, 1005 and 1007) inwhich clear downhole changes in the velocitylogs are present (Fig. 3); at Site 1007, this boun-dary is represented by a hardground. Thus, thisunconformity may represent an additionalsequence boundary, which we term E2.

Sequence e is very thin on the slope part of thetransect. A view along strike (Line 108, Fig. 9)shows that it is not continuous along the marginof Great Bahama Bank. The sequence merely ®llsthe incised canyons of sequence boundary E.Between Sites 1007 and 1006, deposits ofsequence e ®ll-in a slope-parallel depression withan erosive base at the toe-of-slope. This results ina convex-upward geometry, which inverts the

Table 1. True dip angles of sequence boundaries along the Bahama Transect.

Site 1007 Site 1003 Site 1004 Site 1005

Sequenceboundary

Depth(mbsf)

Angle(°)

Depth(mbsf)

Angle(°)

Depth(mbsf)

Angle(°)

Depth(mbsf)

Angle(°)

A ND ND 8 3á2 15 3á2 20 NDB ND ND 25 1á3 65 2á2 90 NDC 35 1á4 100 0á7 150 1á7 185 NDD 210 3á3 145 1á3 185 1á4 225 NDE 210 2á8 175 1á4 ND ND 255 NDF 310 2á3 315 1á2 ND ND 400 NDG 365 2á6 350 1á4 ND ND 430 NDH 420 2á6 400 1á2 ND ND 485 NDI 490 1á9 520 2á7 ND ND 550 NDK 600 1á4 670 ND ND ND ND NDL 670 1á4 740 ND ND ND ND NDM 810 0á8 915 ND ND ND ND NDN 900 0á6 1025 ND ND ND ND NDO 960 0á3 1105 ND ND ND ND NDP 1020 0á3 1165 ND ND ND ND NDSea ¯oor (mbsl) 650á3 2á4 481á4 3á5 418á9 3á4 350á7 46á8

mbsf, metres below sea ¯oor; mbsl, metres below sea level.ND, not determined.



pre-existing relief (Fig. 6). In the median part ofthis body, re¯ections show a ¯at, concavedepression.

Sequence d is again very thin on the slope partof the transect. Between Sites 1003 and 1007, thesequence thickens, but it thins out towards Site1006. In the lower part of the sequence, driftdeposits onlap the basal sequence boundary. At

Site 1007, slope angles through sequences f to dincrease from 2á3° to 3á3° (Table 1). At Site 1003,values are more uniform (1á2° and 1á4°). Theupper sequences c±a record the increased steep-ening of the strongly prograding margin of GreatBahama Bank. Sequence a is not seismicallyresolvable at Site 1007, because it thins belowseismic resolution.

Fig. 6. (A) Evolution of slope angles of the leeward ¯ank of Great Bahama Bank. Note that the depositional relief isprojected on a line normal to the Holocene platform edge. Dashed lines indicate where the calculation of angles isbased on seismic data alone. Depth conversions in these cases were performed for individual boundaries (inparentheses) using data from Eberli et al. (1997a,b). (B) Location of turbidite and drift depocentres along the ¯ank.



Facies geometries

Figure 8 shows a detailed view of the slope partof Line 106, with the facies geometries of rede-posited sediments such as turbidites and slumps.The positions of these deposits in the individuallithological successions at the different sites arepresented in Fig. 3. The interpretation of theseismic line was not performed below re¯ector P,as the sea-bottom multiple crosses the line at thislevel and partly masks the primary re¯ectionpattern. The sea-bottom multiple also inhibits adetailed correlation of the lower part of Site 1003with Site 1005.

Much of the Miocene turbidite packages in theanalysed seismic lines have mounded lobatedepositional geometries. A good example of thisarrangement is provided in sequence m, SW ofSite 1007 (Figs 8±10). At this site, sequencem forms the lower part of sedimentary unit 1,subunit 3, with a lower turbidite-poor and anupper turbidite-dominated interval. The turbiditesuccession can be subdivided into three indivi-dual bodies. On a seismic scale, these bodiespinch out updip from Site 1007. The lower twobodies are up to 1 km wide and 20±30 m thick,mounded structures. Re¯ections in thesemounded zones are discontinuous. In the central

Fig. 7. Seismic Line 106 (above) and interpretation (below). The grey stippled area in the interpretation shows theposition of drift deposits; the white arrow points to an intrasequence unconformity in sequence f (see text forexplanation).



part of the mounds, ¯at to slightly concavere¯ections occur. Geometries of re¯ections indi-cate that individual turbidite bodies prograde. For

example, in sequence k, SW of Site 1003, adownlapping occurs within a turbidite package,indicating a progradational pattern (Fig. 8).



In addition to down- and updip terminations,wedging out of the turbidite bodies also occursparallel to strike (Figs 9 and 10). Figure 10 showsthat, 1±1á5 km S of Site 1007, the two lowerbodies in sequence m have shingled to sigmoidalinternal geometries. The same depositionalgeometries also occur in the turbidite systems ofsequences l and k. The dominant direction oflateral accretion of all of these bodies is towardsthe north; only very minor south-directed pack-age-internal dips occur.

Pliocene and Pleistocene depositional systemswith turbidites are restricted to the uppermostpart of the successions in sequences c±a. Thegeometries of these deposits, as re¯ected in theseismic lines, are those of steeply inclined

Fig. 9. Seismic Line 108 (above)and interpretation (below). See textfor discussion.

Fig. 8. Proximal (slope) part of Seismic Line 106 withline drawing showing interpretation of geometries ofturbidite and slump packages. Note the presence ofinterpreted slumps and associated faults. Similar faultswere observed by Austin et al. (1988) and Harwood &Towers (1988) in other seismic lines of the western¯ank of Great Bahama Bank. These faults probablyrepresent detachment surfaces produced by downslopemovements of sediment. Such slope adjustment pro-cesses were also described from the Miocene of thenorthern ¯ank of Little Bahama Bank (Harwood &Towers, 1988). The arrows in the line drawing point tofeatures discussed in the text. CI, canyon incision insequence f; PT, prograding turbidite package insequence k; ML, mounded lobes in sequence m; CR,convoluted re¯ections in slump of sequence n.



prograding wedges (Fig. 8), which are represen-tative of the recent leeward ¯ank of Great BahamaBank (Wilber et al., 1990). Along strike (Fig. 9),these turbidite packages display a pronouncedlateral continuity. No major package-internalstructures occur in the seismic line. South of Site1003, some discontinuous, slightly inclined tocurved re¯ectors may indicate the position ofvery shallow, incised channel complexes.

Slump deposits occur in different positionsalong the transect of drill holes (Fig. 8). Theslumps of sequences n and i are clearly related tothe occurrence of turbidite depositional systems.In sequence n at Site 1007, the interval of theseismic line that corresponds to the slump ofsedimentary unit 1, subunit 1 (Fig. 8), appears asa zone with curved and convoluted re¯ections.The same type of geometrical expression of

Fig. 10. Seismic Line 102 (above) and interpretation (below). Arrow indicates shingled internal geometry of aturbidite body. See text for discussion.



slumped intervals is in sequence d at Site 1007and in sequences d and c at Sites 1003±1005. Theslumps in the basal part of sequence i at Site 1007(sedimentary unit 1, subunit 4; Fig. 3) show adifferent geometry (Fig. 8). This interval is char-acterized by sigmoidal, downlapping re¯ections.

Re¯ections of `background' deposits are notlaterally traceable over greater distances. Theseismic image of these sediments is characterizedby strong re¯ections with a non-continuous,partly chaotic seismic facies. This re¯ectionpattern is caused by small-scale faults (Fig. 8).These faults probably represent detachment sur-faces produced by downslope movements ofsediment. Such slope adjustment processes havealso been described from the Miocene of thenorthern ¯ank of Little Bahama Bank (Harwood &Towers, 1988).

INTERPRETATION AND DISCUSSION

`Background' sedimentation

According to Eberli et al. (1997a), the Mioceneand lower Pliocene alternation of light-grey anddark-grey wackestones/packstones re¯ects sea-level changes with a frequency of 20 kyr (Bernet& Eberli, 1999). As such, the cycles can be treatedas high-frequency sequences (Mitchum & vanWagoner, 1991), as genetic sequences in the senseof Homewood et al. (1992) or as elementarysequences (Pasquier & Strasser, 1997). However,the exact translation of the signal of sea-level¯uctuations into the cycles still needs discussion.Sea-level changes in a ramp setting may triggeronly minor differences in the character of low-stand and highstand sediments (Burchette &Wright, 1992), because facies belts can shift

down- and upramp hand-in-hand with sea-level.Potential areas for shallow-water carbonateproduction may not be reduced signi®cantlyduring lowstands.

This pattern should also apply to the Bahamianramp. Taking the slope angle values of Table 1and the ramp pro®les of Fig. 6, the zone ofshallow-water carbonate production during theformation of Miocene sequences p±i occupiedapproximately the inner 500 m of the ramppro®le (<2° dip). This value is obtained if oneassumes that major shallow-water carbonate pro-duction is within the upper 15 m of the watercolumn, according to the exponential carbonateproduction curve of Bosscher & Schlager (1992).Assuming 10 m of sea-level fall for the high-frequency cycles (e.g. Mitchum & van Wagoner,1991; Pomar & Ward, 1995), the shoreline wouldshift about 400 m basinwards, thus not greatlyreducing the shallow-water carbonate productionarea. This is in strong contrast to the Miocenecyclic record observed at the outer ramp Sites1003, 1005 and 1007, where compositionalchanges point towards a major shutdown ofshallow-water carbonate particle export from theinner ramp during the formation of the dark-greylayers. It is concluded that one can apply differ-ent scenarios for cycle formation, as discussedpreviously for similar Jurassic limestone±marl-stone cycles by Pittet & Strasser (1998).

One possible model for cycle generation in theouter ramp would be that shallow-water compo-nents are exported during sea-level highstands(Fig. 11a). During falling sea level and low-stands, the ramp export system would have beenshut down, and siliciclastic input would havebeen intensi®ed. This model applies the prin-ciple of highstand shedding (sensu Schlageret al., 1994) to the ramp depositional setting.

Fig. 11. Two different models for sea-level control on the formation of light-grey/dark-grey wackestone/packstonecycles. In (a), the dark-grey intervals (shaded) form during falling sea level and lowstands. In (b), the dark-greyintervals form during rising sea level. See text for discussion. LSD, lowstand deposits; TSD, transgressive deposits;HSD, highstand deposits; LHSD, late highstand deposits.



However, as discussed above, it requires anadditional factor to suppress shallow-water car-bonate production during lowstand conditions,such as a change in water quality producingrelated changes in shallow-water faunal and¯oral composition.

Alternatively, the dark-grey layers may haveformed as condensed intervals during sea-levelrises (Fig. 11b), when the shallow-water carbonateproduction area of the inner ramp stepped back.With decreasing rates of sea-level rise, carbonateproduction caught up (sensu Kendall & Schlager,1981), and export of sediment commenced, butit may also have persisted during periods ofrelative sea-level fall, as the production beltshifts downramp. This model would largelyfollow the classical carbonate ramp sequencestratigraphic concept described by Handford &Loucks (1993).

Both the above models could account for thesymmetric or asymmetric nature of the cycles.Symmetric cycles would re¯ect gradational shut-down or turning on of shallow-water carbonateproduction. In the highstand-shedding model,hardgrounds of asymmetric cycles would havebeen generated if the high-frequency sea-levellowering was superimposed on a lower frequencysea-level stillstand or even lowering. Symmetriccycles would develop if the high-frequency cycleis superimposed onto a lower frequency sea-level rise. If the dark-grey layers formed duringthe transgressive part of the high-frequencycycles, the sharp contact would re¯ect a rapidbackstepping of the shallow-water carbonateproduction area as a result of the superimpo-sition of the high-frequency and lower order sea-level rises. It is not possible to differentiatebetween the two scenarios based on availabledata. Information on the inner ramp depositionalareas is needed in order to analyse potentialchanges in the shallow-water areas during sea-level ¯uctuations.

The occurrence of the light-grey/dark-greyhigh-frequency cycles is intrinsically linked tothe ramp con®guration. These cycles disappearupsection during the formation of sequence f(Fig. 3); Fig. 6 shows that the ramp pro®le steep-ened and shortened considerably in this lowerPliocene interval. Disappearance of the cyclesgoes hand-in-hand with the onset of sedimenta-tion of the poorly differentiated muddy sedimentsof sedimentary unit 2. The classical highstandshedding cycles (e.g. Droxler et al., 1988) ®nallycharacterize the upper Pliocene and Quaternaryperiplatform sediments (Eberli et al., 1997a).

Calciturbidites

Lowstand and highstand turbidites

Turbiditic packages occur on top and belowsequence boundaries (Fig. 8), indicating thatcalciturbidites were shed during sea-level low-stands and highstands. This is corroborated bythe trends in sediment composition within onedepositional sequence. Seventy-®ve thin sectionstaken between sequence boundaries K and I atSite 1003 were analysed quantitatively (Fig. 5).This sequence contains a lower and an upperturbidite depositional system (Figs 3, 5 and 8).The input of shallow-water components in thelower part of the sequence is negligible, but itreaches values of around 20% within the upperpart. Shallow-water allochems are green algae,red algae and intraclasts. This compositionaltrend correlates with a textural trend; turbiditesin the upper part of the sequence are more grainsupported than in the lower part, as re¯ected byan upwards increase in the abundance of inter-granular cement (Fig. 5).

A model for turbidite deposition during a third-order sea-level ¯uctuation would be as follows:lowering of sea-level, re¯ected in the geophysi-cally de®ned sequence boundaries, producesredeposition along the distally steepened ramp,mainly as a response to basinward shift of baselevel (lowstand turbidites). With progressive¯ooding of the ramp during rising sea level,turbidite shedding is suppressed (sequences p, n,m and i) or reduced (sequences l and k). Thisreduction could re¯ect a general decrease inshallow-water material input in response toincreasing accommodation space in the shallowpart of the ramp. The upper packages in thesequences represent highstand shedding of calci-turbidites as a result of a slowing of the rate ofincrease in accommodation space on the platformtop during sea-level highstands. These trends aresimilar to those in siliciclastic turbidite systems,described by Mutti (1985). Overall, it is linked tothe availability of accommodation space on theinner ramp.

Geometry of turbidite packages

Some of the mounded lobes in the analysedtransect are similar to the system of coalescinglobes developing at the toe-of-slope and basin¯oor of the Tongue of the Ocean (Schlager &Chermak, 1979). Other lobes are laterally discon-nected. The ¯at to convex-down re¯ections in the



central part of the lobes indicate the positions ofturbidite feeder systems (channelled complexes).Thus, Miocene turbidite depositional systemswere fed by point sources. This is in contrast toQuaternary turbidites along the leeward ¯ank ofGreat Bahama Bank, which are supplied by linesources (Eberli, 1991).

Maximum shedding of turbidites in the ana-lysed part of the Great Bahama Bank ¯ankcorresponds to the prograding±of¯apping plat-form-growth episode of the concave ramp-shapedsequences m±i. Within this package, the amountof turbiditic bank-top shedding decreasedupwards, and turbidite packages step backupramp. This evolution directly re¯ects thedecrease in shallow-water carbonate productionspace on the ramp within the progradational toof¯apping platform-growth episode. An overprintand intensi®cation of this trend may re¯ect a`memory effect' of the carbonate ramp. Carbonateproduced in excess on the inner ramp is exportedand redeposited along the ramp pro®le. At leastin the proximal part, this material is prone toreworking during basinward shifts in base level.Thus, lowstand turbidites are an admixture ofmaterial formed during the previous highstandand contemporary formed allochems. With pro-gressive reduction in the area of shallow-watercarbonate production (see above), the materialavailable for redeposition during each lowstandof the long-term sea-level fall becomes less.

Formation of the convex ramp geometry duringthe late Tortonian was accompanied by a shift ofdepocentres, re¯ected by the build-up of largedrift wedges on the basin ¯oor (Fig. 7). Althoughnorth-directed bottom currents were activethroughout the Early and Middle Miocene, as isdemonstrated by the shingled geometry of theturbidite lobes in the underlying sequences forexample, we propose that the geometrical changewas linked to an intensi®cation of bottom cur-rents. Stronger currents would account for theconvex slope shape, a feature also known fromHolocene Great Bahama Bank, where it is attrib-uted to current activity (Mullins & Neumann,1979; Mullins, 1983). The currents would alsoaccumulate the drift wedges, which would consistof material that bypassed the convex outer ramp.

A further change in the bottom current systemoccurred during the formation of the lowerPliocene sequence f. Whereas underlyingsequences are characterized by stratal continuitybetween the deep ramp and the basinal driftdeposits, drift deposits onlap the deep rampdeposits in this sequence. Taking into account

the basinal erosion along sequence boundary E, itseems appropriate to postulate that these changeswere again driven by a strengthening or reorgani-zation of bottom currents.

Slumps

Slumps are a minor component in the analysedtransect. The lower slumps in sequences p2 and qhave not been analysed in further detail, as theyare situated below the sea-bottom multiple in theseismic line. In the upper part of the succession,slumps were probably related to instabilities ofthe platform edge and slope resulting fromsediment overloading, a process described fromExuma Sound by Crevello & Schlager (1980).Slumps in sequences o and n (Site 1007) fallbroadly within the early Middle Miocene interval,in which major sediment gravity ¯ow events areknown from West Florida and Little Bahama Bank(Great Abaco Member; Fulthorpe & Melillo,1988). These regional events are interpreted astectonically triggered slope failures. However, thegravity ¯ows recorded in the Straits of Florida atODP Site 626 in middle Miocene planktonicforaminiferal zones N 11 and N 12 (Fulthorpe &Melillo, 1988) do not have equivalents in theBahamas transect.

CONCLUSIONS

Deposits of the Miocene to lower PlioceneBahamian carbonate ramp consist of an alterna-tion of light-grey and dark-grey wackestones/packstones with intercalations of turbidite pack-ages and slumps. Light-grey wackestones/pack-stones record phases of export of shallow-watercomponents, whereas the ramp did not exportmaterial during deposition of the dark-greylayers, which contain up to 20% clay.

The cyclic alternations of both lithologies aresymmetric and asymmetric. Cycles re¯ect high-frequency (20 kyr) sea-level changes and thusform genetic sequences. Two models are possibleto explain cycle generation. The ®rst one appliesthe principle of highstand shedding and corre-lates the dark-grey layers to sea-level lowstands.The second assumes that the dark-grey layersformed as condensed sections during sea-levelrises. Basinward shoreline shifts during high-frequency sea-level falls were around 400 malong the inner ramp, which had an inclinationof around 2°. As such, the shallow-watercarbonate production area was not reduced



signi®cantly during sea-level lowstands. Theseobservations favour the second interpretation.The dark-grey/light-grey cyclicity is linked tothe Miocene ramp geometry. It disappearedduring the Early Pliocene, as the ramp began tochange into a ¯at-topped platform.

Correlation of core, log and seismic datashows that the calciturbidites are arranged intomounded lobate bodies with feeder channels.Some of the lobe systems are coalescent; othersare laterally disconnected, indicating supply viapoint sources. Turbidite packages of the outerramp were deposited during sea-level highstandsand lowstands. Highstand and lowstand turbid-ites have different compositions. Highstandturbidites are characterized by the occurrence ofshallow-water carbonate particles includinggreen algal debris and epiphytic foraminifera,whereas lowstand turbidites are dominated byabraded bioclasts.

Depocentres of gravity ¯ows change from anouter ramp position during the Early and MiddleMiocene to a basin ¯oor position during the LateMiocene and Early Pliocene. This evolution wastriggered by the strengthening of the bottomcurrent system in the Santaren Channel and theStraits of Florida.

ACKNOWLEDGEMENTS

Discussions with Don McNeill on Bahamas geol-ogy are gratefully acknowledged. Andre Strasserkindly went through the manuscript and gavevaluable comments; Wolfgang Schlager providedvery valuable comments on processes acting alongthe slope of Great Bahama Bank. We also want tothank the crew of the JOIDES Resolution for theircontinuous support on board ship. This researchwas funded by the Deutsche Forschungsgemein-schaft (projects Be 1272/5 and /6 to C.B. and Re1051/4 to J.R.) and the National Science Founda-tion (166F000330 to K.B., 166F000333 to G.E. andOCE-9314586 to G.E. and F.A.). Thanks are alsoextended to Gilbert Camoin and Bob Garrison forconstructive reviews, and to Ian Jarvis for the veryhelpful comments and his editorial efforts.

REFERENCES

Anselmetti, F.S., Eberli, G.P. and Bernoulli, D. (1997)Seismic modeling of a carbonate platform margin(Montagna della Maiella, Italy): variations in seismicfacies and implications for sequence stratigraphy. In:

Carbonate Seismology (Eds I. Palaz and K.J. Marfurt),SEG Geophys. Dev. Ser., 6, 373±406.

Austin, J.A., Schlager, W., Palmer, A.A. et al. (1986)Proceedings ODP, Initial Report, 101. College Station.

Austin, J.A., Ewing, J.I., Ladd, J.W., Mullins, H.T. andSheridan, R.E. (1988) Seismic stratigraphic implica-tions of ODP Leg 101 Site surveys. In: ProceedingsODP, Scienti®c Research (Eds J.A. Austin, W. Sch-lager et al.), 101, 391±424.

Beach, D.K. and Ginsburg, R.N. (1980) Facies succes-sion of Pliocene±Pleistocene carbonates, northwest-ern Great Bahama Bank. Am. Assoc. Petrol. Geol.,Bull., 64, 1634±1642.

Bernet, K. and Eberli, G.P. (1999) Spectral analysis onresistivity data along the western Great Bahama Bank(Leg 166, Site 1003). In: Proceedings ODP, Scienti®cResearch (Eds G.P. Eberli, P. Swart, et al.), 16611 (inpress).

Bosscher, H. and Schlager, W. (1992) Computer simu-lation of reef growth. Sedimentology, 39, 503±512.

Burchette, T.P. and Wright, V.P. (1992) Carbonate rampdepositional systems. Sedim. Geol., 79, 3±57.

Crevello, P.D. and Schlager, W. (1980) Carbonate debrissheets and turbidites, Exuma Sound, Bahamas.J. Sedim. Petrol., 50, 1121±1147.

Droxler, A.W., Bruce, C.H., Sager, W.W. and Watkins,D.H. (1988) Pliocene-Pleistocene variations in ara-gonite content and planktonic oxygen-isotope recordin Bahamian periplatform ooze, Hole 633A. In:Proceedings ODP, Scienti®c Research (Eds J.A. Aus-tin, W. Schlager et al.), 101, 221±244.

Dunham, R.J. (1962) Classi®cation of carbonate rocksaccording to depositional texture. In: Classi®cationof Carbonate Rocks (Ed. W.E. Ham), Am. Assoc.Petrol. Geol., Mem., 1, 108±121.

Eberli, G.P. (1991) Calcareous turbidites and theirrelationship to sea-level ¯uctuations and tectonism.In: Cyclic and Event Strati®cation (Eds G. Einsele, W.Ricken and A. Seilacher), pp. 340±359. Springer-Verlag, Berlin.

Eberli, G.P. and Ginsburg, R.N. (1987) Segmentationand coalescence of Cenozoic carbonate platforms,northwestern Great Bahama Bank. Geology, 15,75±79.

Eberli, G.P. and Ginsburg, R.N. (1989) Cenozoic pro-gradation of northwestern Great Bahama Bank, arecord of lateral platform growth and sea-level ¯uc-tuations. In: Controls on Platform and BasinDevelopment (Eds P.D. Crevello, J.L. Wilson, J.F. Sargand J.F. Read), Soc. econ. Paleont. Mineral., Spec.Publ., 44, 339±351.

Eberli, G.P., Kendall, C.G.S.C., Moore, P., Whittle, G.L.and Cannon, R. (1994) Testing a seismic interpreta-tion of Great Bahama Bank with a computer simula-tion. Am. Assoc. Petrol. Geol., Bull., 78, 981±1004.

Eberli, G.P., Swart, P.K., Malone, M.J. et al. (1997a)Proceedings ODP, Initial Report, 166, College Station.

Eberli, G.P., Swart, P.K., McNeill, D.F., Kenter, J.A.M.,Anselmetti, F.S., Melim, L.A. and Ginsburg, R.N.(1997b) A synopsis of the Bahamas Drilling Project:results from two deep core borings drilled on theGreat Bahama Bank. In: Proceedings ODP, Initial



Report (Eds G.P. Eberli, P.K. Swart, M.J. Maloneet al.), 166, 23±41.

Fulthorpe, C.S. and Melillo, A.J. (1988) Middle Mio-cene carbonate gravity ¯ows in the Straits of Floridaat Site 626. In: Proceedings ODP, Scienti®c Research(Eds J.A. Austin, W. Schlager et al.), 101, 179±191.

Handford, C.R. and Loucks, R.G. (1993) Carbonatedepositional sequences and systems tracts ± res-ponses of carbonate platforms to relative sea-levelchanges. In: Carbonate Sequence Stratigraphy (EdsR.G. Loucks and J.F. Sarg), Am. Assoc. Petrol. Geol.,Mem., 57, 3±41.

Harwood, G.M. and Towers, P.A. (1988) Seismic sedi-mentologic interpretation of a carbonate slope, northmargin of Little Bahama Bank. In: Proceedings ODP,Scienti®c Research (Eds J.A. Austin, W. Schlageret al.), 101, 263±277.

Homewood, P., Guillocheau, F., Eschard, R. and Cross,T.A. (1992) CorreÂlations haute reÂsolution et strati-graphie geÂneÂtique: une deÂmarche inteÂgreÂe. Bull.Centre Rech. Expl. Prod., Elf-Aquitaine, 16, 357±381.

Kendall, C.G.S.C. and Schlager, W. (1981) Carbonatesand relative changes in sea level. Marine Geol., 44,181±212.

Kenter, J.A.M., Ginsburg, R.N. and Troelsta, S.R. (1999)The Western Great Bahama Bank: Sea-level-drivensedimentation patterns on the slope and margin. In:Soc. Econ. Paleont. Mineral., Contr. Sedimentology(Ed. R.N. Ginsburg) (in press22 ).

Mitchum, R.M. and van Wagoner, J.C. (1991) High-frequency sequences and their stacking patterns:sequence-stratigraphic evidence of high-frequencyeustatic cycles. Sedim. Geol., 70, 131±160.

Mullins, H.T. (1983) Modern carbonate slopes andbasins of the Bahamas. In: Platform Margins andDeep Water Carbonates (Eds H.E. Cook, A.C. Hineand H.T. Mullins), Soc. econ. Paleont. Mineral.,Short Course, 12, 4/1±4/138.

Mullins, H.T. and Neumann, A.C. (1979) Deep car-bonate bank margin structure and sedimentation inthe Northern Bahamas. In: Geology of ContinentalSlopes (Eds L.J. Doyle and O.H. Pilkey), Soc. econ.Paleont. Mineral., Spec. Publ., 27, 165±192.

Mutti, E. (1985) Turbidite systems and their relation todepositional sequences. In: Provenance of Arenites(Ed. G. Zuffa), pp. 65±93. Reidel, Dordrecht.

Pasquier, J.-B. and Strasser, A. (1997) Platform-to-basincorrelation by high-resolution sequence stratigraphyand cyclostratigraphy (Berriasian, Switzerland andFrance). Sedimentology, 44, 1071±1092.

Pittet, B. and Strasser, A. (1998) Depositionalsequences in deep-shelf environments formedthrough carbonate-mud import from the shallowplatform (Late Oxfordian, German Swabian Alb andeastern Swiss Jura). Eclogae geol. Helv., 91, 149±169.

Pomar, L. and Ward, W.C. (1995) Sea-level changes,carbonate production and platform architecture: theLlucmajor Platform, Mallorca, Spain. In: SequenceStratigraphy and Depositional Response to Eustatic,Tectonic and Climatic Forcing (Ed. B. Haq), pp. 87±112. Kluwer Academic Press, Dordrecht.

Reijmer, J.J.G., Schlager, W., Bosscher, H., Beets, C.J.and McNeill, D.F. (1992) Pliocene/Pleistocene plat-form facies transition recorded in calciturbidites(Exuma Sound, Bahamas). Sedim. Geol., 78, 171±179.

Schlager, W., Bourgeois, F., Mackenzie, G. and Smit, J.(1988) Boreholes at Great Isaac and Site 626 and thehistory of the Florida Straits. In: Proceedings ODP,Scienti®c Research (Eds J.A. Austin, W. Schlageret al.), 101, 425±427.

Schlager, W. and Chermak, A. (1979) Sediment facies ofplatform-basin transition, Tongue of the Ocean,Bahamas. In: Geology of Continental Slopes (Eds L.J.Doyle and O.H. Pilkey), Soc. econ. Paleont. Mineral.,Spec. Publ., 27, 193±208.

Schlager, W. and Ginsburg, R.N. (1981) Bahama car-bonate platforms ± the deep and the past. MarineGeol., 44, 1±24.

Schlager, W., Reijmer, J.J.G. and Droxler, A.W. (1994)Highstand shedding of carbonate platforms. J. Sedim.Res., B64, 270±281.

Sta¯eu, J. and Schlager, W. (1995) Pseudo-unconfor-mities in seismic models of large outcrops. Geol.Rundsch, 84, 761±769.

Westphal, H., Reijmer, J.J.G. and Head, M.J. (1999)Input and diagenesis on a carbonate slope (Bahamas):Response to morphologic evolution and sea-level¯uctuations. In: Advances in Carbonate SequenceStratigraphy ± Applications to Reservoirs, Outcropsand Models (Eds P.M. Harris, A.H. Saller, T. Simoand R. Handford), Spec. Publ. Soc. econ. Paleont.Miner33 (in press).

Wilber, R.J., Milliman, J.D. and Halley, R.B. (1990)Accumulation of bank-top sediment on the westernslope of Great Bahama Bank: rapid progradation of acarbonate megabank. Geology, 18, 970±974.

Manuscript received 8 April 1998;revision accepted 12 April 1999.



Sedimentary patterns and geometries of the Bahamian outer carbonate ramp (MioceneLower Pliocene, Great Bahama Bank)

Documents