Abstract: Two cores drilled on the western margin of Great Bahama Banks penetrated a total of 1131 m of Quaternary to Late Neogene platform and periplatform carbonate sediments. These carbonates show a wide range of diagenetic textures ranging from largely unaltered primary carbonates to texturally mature limestones and dolostones. Although some of these carbonates have been diagenetically altered under freshwater conditions and show classic geochemical and petrographic indicators of such, other rocks exhibit similar petrographic features yet have never experienced meteoric influences. These features include micrite envelopes, moldic porosity, blocky spar cementation, and aragonite neomorphism The core Unda (453.8 m) includes three successions of shallow-water platform to reef sediments that alternate with deeper shelf deposits of silt to coarse sand. Clino, the more distal core, penetrated 677.3 m. It contains an upper reef to platform (21.6-140 m) overlying a thick package of deep forereef to upper and lower slope sediments (140-677.3 m). Three diagenetic zones were identified on the basis of similar diagenetic fabrics. Diagenetic zone 1 (Unda 0-108.08 m; Clino 0-152.71 m) is characterized by minor early marine diagenesis overprinted by numerous episodes of subaerial exposure with caliches, large-scale dissolution, and blocky spar cementation. Diagenetic zone II includes the lower reef and platform in Unda (292.82360.28 m) and has the most complex alteration in these cores. Minor early marine diagenesis was followed by several episodes of meteoric diagenesis, Most of the early fabrics, however, were destroyed during pervasive dolomitization in the burial environment. Diagenetic zone III includes deeper shelf to slope facies in both cores (Unda 108.08-292.82 m; 360.28-452.94 m; Clino 152.71-677.27 m) and shows only marine to marine-burial diagenesis. The exact fabrics within these intervals varies by lithology. Peloidal packstones to grainstones show minimal alteration (primarily compaction, minor dolomitization, and some recrystallization). Skeletal grainstones show two different styles of modification. Most commonly the grainstones are characterized by nearly complete dissolution of aragonitic components, minor cementation, and 10% to 20% early burial dolomite. Grainstones interbedded with the tight peloidal sediments, on the other hand, show nearly complete blocky spar cementation along with neomorphism of aragonitic skeletal grains (peloids were dissolved). Several marine hardgrounds with penecontemporaneous dolomite and/or phosphate and blackening also occur within the deeper-water facies. Each diagenetic zone in Clino and Unda is characterized by fabrics developed in one or more diagenetic environment. The larger-scale controls, therefore, are those that govern the diagenetic environment. For these young sediments, the most important are the depositional system (i.e., reef and platform versus deeper margin to slope) and the sea-level history. As a result, the diagenetic zones largely coincide with the depositional successions and seismic sequences identified in these cores. This study shows that the end product of marine to marine-burial diagenesis appears very similar to the end product of alteration in the phreatic meteoric environment i.e., a limestone composed of LMC and minor dolomite with micrite envelopes, moldic porosity, blocky spar cementation, and aragonite neornorphism. On this basis, a reevaluation is needed of the criteria for recognizing meteoric diagenesis in ancient carbonate sequences. Without clear physical evidence of subaerial exposure (e.g., caliche horizons or vadosecements) or chemical evidence of meteoric fluids (e.g., negative δ 18 O) great care is needed to identify meteoric diagenesis, In addition, the large-scale (> 100 m) sea-level lowstands of the latest Pleistocene did not result in significant phreatic meteoric diagenesis, perhaps because the fresh-water lens is too far below the recharge zone.

Abstract: During the Neogene, western Great Bahama Bank prograded more than 25 km into the Straits of Florida. This progradation occurred in pulses seen on seismic lines as distinct seismic sequences. Two core holes (Unda and Clino) were drilled through the proximal parts of these prograding carbonate sequences to test several fundamental questions in carbonate sedimentology, diagenesis, and sequence stratigraphy. The sequence strarigraphic objectives of the drilling were to (1) provide information on facies within the seismic sequences, (2) test if predicted facies changes coincide with seismic sequence boundaries and inferred systems tracts, (3) determine the ages of sequence boundaries, and (4) test whether the prograding sequences were sea-level controlled, and if so, provide information about the timing and amplitude of these relative changes in the later part of the Neogene. The cores at sites Unda and Clino penetrated eight seismic sequences that have a Mid/Late Miocene to Recent age. The sedimentary successions in the cores display a repetitive pattern of changing facies that can be related to changes in sea level. Sea-level falls are indicated on the slope by either hardgrounds, firmgrounds, or coarser-grained beds, whereas caliche horizons or karst surfaces document exposure on the platform. Aggrading shallow-water carbonates and thick fine-grained slope deposits are interpreted to be deposits of sea-level highstands. These lithologic indications of changing sea level coincide with the interpretation of the seismic sequence architecture. Seismic sequence boundaries correlate to horizons and/or sedimentary packages indicative of sea-level falls, whereas prograding pulses coincide with the thick slope sections deposited during sea-level highstands. In addition to the facies there is a strong correlation between diagenesis and the seismic sequences. This relationship can be seen in both the paragenetic sequences and the variations in mineralogy. In core Unda, all paragenetic boundaries coincide with sequence boundaries. In the slope section of Clino, mineralogical changes occur across diagenetic horizons, e.g., hardgrounds, which also coincide with seismic sequence boundaries. These surface-bounded changes in diagenesis give strong indications that changes in sea level alter the diagenetic potential of the carbonates within the sequence stratigraphic pattern. Petrophysical analyses document that the combination of changing facies and diagenesis across sequence and systems-tract boundaries produces the necessary impedance contrasts for seismic reflections along stratigraphic surfaces. The ages of the eight seismic sequence boundaries were determined using a combination of biostratigraphy, Sr-isolope stratigraphy, and magnetostratigraphy. The ages of the sequence boundaries are the same in both drill sites and do not cross seismic reflections, documenting that seismic reflections separate sedimentary packages of different ages. This age consistency confirms one of the basic assumptions of sequence stratigraphy, i.e., that seismic reflections are time lines and have chronostratigraphic significance. The duration for the deposition of the individual sequences is between 0.5 and approximately 1 Myr, indicating that the sequences are produced by third-order sea-level fluctuations. The sequence architecture reflects a long-term pattern of sea-level changes that is in good agreement with the known global sea-level history. A general lowering of sea level at the end of the Middle Miocene resulted in progradation of the platform and the development of a margin at drill site Unda. At about 8.9 Ma progradation was interrupted and the margin stepped back because of a sea-level rise. The lowering of sea level in the latest Miocene and the subsequent sea-level lowstand that lasted during most of the Messinian is recorded in Unda by the development of a reef on the former slope deposits and recognized on the seismic line as a major erosional unconformity at the margin. Backstepping of the platform margin in the early Pliocene indicates a rapid flooding, which is also a major transgression on the global chart. The high-amplitude sea-level changes in the Pleistocene are recorded as laterally stacked sequences with increasingly higher angles in the foresets.

Abstract: The Sr-isotope composition of pore fluids and co-occurring sediments (carbonates, phosphorites, and dolomites) have been measured from two cores retrieved from western margin of the Great Bahama Bank. The Sr-isotope compositions of the pore fluids, corrected for possible contamination by surface waters using tritium, indicate that the formation fluids have 87 Sr/ 86 Sr ratios less radiogenic than modern sea water, but in all instances considerably more radiogenic than the co-occurring sediments and rocks. Assuming that similar conditions of fluid movement have been prevalent since deposition, the Sr-isotopic composition of the sediments has been used to constrain the age of deposition and diagenesis. The 87 Sr/ 86 Sr ratios of the solid components can be treated in two ways. First, if it is assumed that the chronostratigraphy determined from biostratigraphic and magnetostratigraphic methods is correct, then the Sr-isotopic composition can be used to place constraints on the timing of diagenesis. Second, if the diagenetic alteration occurred relatively soon after deposition, then the Sr-isotope composition can constrain the age of deposition. The data from the sediments show that the 87 Sr/ 86 Sr ratios are in most instances close to the 87 Sr/ 86 Sr ratio calculated from the chronostratigraphic age. The only substantial disagreement between the Sr-isotope age and the biostratigraphic/magnetostratigraphic age occurs between 300 and 500 mbsf in Clino. This could indicate either recrystalli-zation after deposition by more radiogenic solutions or slight uncertainties in the chronostratigraphy in this interval.

Abstract: Two continuous core borings 8.5 km apart on the leeward margin of Great Bahama Bank provided a special opportunity to interpret the depositional history during the Late Pliocene and Pleistocene. The upper ca. 200 m of the borings presented in this paper show seaward progradation of the margin with overall shallowing. At the base of this succession are skeletal grainstones and packstones that accumulated on the proximal slope of the margin. These are succeeded by thick intervals of reefal and coral-bearing deposits that in turn are capped by nonskeletal grainstones similar to the modern sediments in the interior of the Bahama Banks. Reefal and coral-bearing limestones produce about half of the leeward progradation. In both borings, the major episodes of coral growth began on unconsolidated sediments and show indications of upward shallowing according to the interpretations of Budd and Manfrino (this volume). In Clino, the seaward boring, the major episode of reef growth was terminated by deposition of fine-grained sediment mixed with coral floatstone; in Unda, the bankward boring, reef demise was the result of subaerial exposure. In Unda, the post-reefal deposits, termed the layered cap, are largely a succession of packstones and grainstones of peloids and skeletal debris, which are interpreted as platform-top accumulations similar to those of the Holocene. Thirteen discontinuity horizons, which are distinct breaks in sedimentation and have numerous features indicative of subaerial exposure, are used to divide this interval into sea -level flooding events (rising sea level and highstands) and periods of exposure (falling sea level and lowstands) during the Pleistocene and Late Pliocene. In Clino, above the major episodes of reef development, corals and skeletal grainstones and packstones continued to be the major component of aggradation. Nonskeletal packstones and grainstones from the platform interior are present only in the uppermost 20 m. Ten discontinuity horizons, which indicate falling sea level and sea-level lowstands, are identified in the Late Pleistocene. Two magnetostratigraphic datums provide the primary dating for this upper succession (McNeill et al., this volume). Three factors show that reef development occurred first in Unda during a period of relative rise of sea level: (1) correlating the top of the Olduvai between the two borings, (2) coral assemblages, and (3) interpretations of depositional environments. Coral assemblages in Unda include Pliocene species, whereas those in Ciino are largely Pleistocene to recent (Budd and Manfrino, this volume). A drop in sea level produced a downshift of major reef growth of some tens of meters (and as much as 100 m) to Clino. Another relative rise of sea level allowed for reefal aggradation. This offlap succession of reefal deposits confirms the pre-drilling model of margin evolution developed from seismic stratigraphy ( Eberli and Ginsburg, 1989 ). The downshift in reef development during the Late Pliocene have been coincident with the onset of Northern Hemisphere ice accumulation as inferred from the deep-sea isotope record. These results confirm and expand on the role of reefs and sea-level fluctuations in the architecture of Great Bahama Bank. Reefs make a major contribution to leeward progradation, a circumstance that adds support to considering the Neogene Bank an atoll-Like structure. It is estimated that the margin shifted seaward some 10 km in approximately two million years. This episodic progradation was controlled by a hierarchy of sea-level fluctuations during the Late Pliocene and Pleistocene. Major episodes of sea-level rise allowed for aggrading reef development; higher-frequency cycles of sea-level fluctuation in the Pleistocene are recorded in the alternations of subaerial exposure and accumulation of coral-bearing limestones in Clino and the nonskeletal grainstones and packstones of Unda. The ten alternations in the Upper Pleistocene (above the Brunhes/Matnyama boundary) interval matches the number of sea-level oscillations (glacial and interglacial stages) inferred from the oxygen isotope record in the deep sea.

Abstract: Species of reef corals are identified within lithologic units in the two Bahamas Drilling Project cores (Unda and Clino); species richness, frequencies of different colony shapes, and occurrences of indicator species are used to interpret reef environments. The results suggest that twelve Pliocene-Pleistocene coral-bearing units in the upper portions of Unda (26.2-180.8 m) were deposited under generally shallowing-upward conditions, ranging from intermediate (10-20 m) or deeper (> 20 m) forereef environments, to shallow (0-10 m) forereef environments, to shallow (0-10 m) and intermediate-depth (10-20 m) platform environments, in contrast, five Upper Miocene coral-bearing units in the lower portions of Unda (298.1-376.7 m) were deposited under generally deeper conditions, including (1) a small intermediate-depth (10-20 m) or deeper (20 to > 30 m) reefal unit and (2) intermediate-depth (10-20 m) to deeper (20 to > 30 m) forereef environments with large numbers of plate-shaped corals. Like the upper portions of Unda, eleven Pliocene-Pleistocene coral-bearing units in Clino (26.2-197.5 m) formed under shallowing-upward conditions, ranging from intermediate-depth (10-20 m) and deep (20 to > 30 m) forereef environments, to shallow (< 10 m) forereef environments, to shallow (< 10 m) or intermediate (10-20 m) platform environments. However, middle forereef deposits in the upper portions of Unda contain diverse assemblages dominated by Stylophora , whereas middle forereef deposits in Clino are dominated by Acropora . These interpretations generally correspond well with those based on lithology, and indicate that coral assemblages provide unique and important criteria for interpreting Late Cenozoic reefal deposits.

Abstract: The downhole logs included in the pocket of this volume include the geophysical, geochemical, lithologic and petrophysical logs of borings Clino and Unda. Data displayed include mineralogy, stable isotopes, recovery, lithology, depositional fades, trace fossils, seismic sequences, porosity, and permeability. Geophysical logs include gamma ray, caliper, density, neutron, sonic velocity, and neutron porosity. These logs are plotted to the same scale to allow direct comparison of the data from the two borings. The upper Pliocene and Pleistocene reefs and platform-top deposits are readily recognizable on geophysical logs where they show higher amplitudes of density and velocity than the forereef and slope deposits. Some of the peaks coincide with subaerial exposure surfaces and their associated cemented intervals. It is not however possible to distinguish reef from platform top using logs alone. The Miocene reef facies in Unda stands out from surrounding slope facies with higher density and sonic velocity and much greater variability. The fine-grained slope facies show consistent low density and sonic velocity broken by interruptions or hardgrounds. The background values gradually increase with depth. The interruptions of platform-derived sediment are characterized by sonic velocity and density peaks of less than a meter to a few meters in thickness alternating with low velocity/density beds of similar thickness. Hardgrounds are identified by sharp gamma peaks and usually also have density and sonic velocity peaks either at or just below the hardground.

Abstract: Permeability was measured on cores from two wells (Clino and Unda) from the western margin of Great Bahama Bank. Permeability varies from < 1 md to > 8000 md and shows no simple correlation with porosity. Pore types, however, are important in determining permeability. Macropore types include intergranular, intragranular, intercrystalline, moldic, and vuggy to cavernous. Micropore types include both primary and secondary intercrystalline pores. Moldic pores occur as either isolated molds or connected into an open pore network. On the basis of the most abundant pore type, samples with intercrystalline porosity show the highest permeability (median = 1041 md), followed by those with connected-moldic pores (median = 529 md), and then by those with intergranular pores (median = 123 md). Moldic pores are generally considered not to contribute to permeability because they are often isolated from other pores. Many of these samples, however, have moldic pores that are well connected in a continuous pore network, As a result, these molds contribute to, and even control, the permeability. The permeability distribution in the cores can be related to the original lithology and to diagenesis, because both control the pore types present. Meteoric diagenesis of shallow-water facies leads to generally low matrix permeability (median = 9 md), but the permeability is controlled by vuggy to cavernous porosity that we cannot measure. Deeper-water facies were altered in the marine burial environment, and are characterized by either moldic porosity with minimal cementation, or neomorphism with nearly complete cementation. Peloidal sediments compacted early and show little diagenetic alteration. The poorly cemented skeletal intervals have moderate to high permeability (range 0.1-8445 md) generally controlled by the degree of connectivity between moldic pores. The peloidal and well-cemented skeletal intervals have low permeability (range 0.1-198 md), reflecting the controlling role of microporosity. Permeability in the dolomitized portion of Unda, the more proximal core, is high (median = 1274 md) due to the predominance of intercrystalline porosity. On the basis of this study, the most important factor controlling permeability in these young carbonates is the specific pore type rather than the amount of porosity. In addition, the principal pore type can be related to the diagenesis and, to a lesser degree, the depositional facies. Although intercrystalline porosity from dolomitization and vuggy porosity from meteoric diagenesis have long been known to control permeability where they occur, this study also documents the importance of the connected-moldic porosity that forms during marine burial diagenesis.

Abstract: Independent analyses to date prograding carbonate platform-margin sediments have enabled development of an integrated chronostratigraphy of the west margin of the Great Bahama Bank (GBB). The chronostratigraphy permits interpretation of periods of deposition, nondeposition, and erosion on the platform and margin. The integrated chronostratigraphy is based on planktic foraminiferal biostratigraphy, calcareous nannofossil biostratigraphy, strontium-isotopestratigraphy, and magnetostratigraphy. Critical to interpreting each type of stratigraphy is a clear understanding of slope dynamics and the depositional system. The biostratigraphic data, in particular, are affected by the extreme dilution of key microfossils by platform-derived sediments during margin progradation. Because of this factor, the highest abundance of microfossils is restricted to thin units of pelagic sediment, deposited during temporary intervals when platform sediment supply was greatly reduced. However, these selective microfossil concentrations are also more likely to yield premature last-appearance datums and delayed first-appearance datums than the diluted intervals, because they represent short periods of time. Despite these problems, the microfossils provide critical age control. The detailed chronostratigraphy allows interpretation of a well-constrained platform-margin evolution. Core Unda, the more landward location, penetrated the oldest sediments (earliest late Miocene). The more seaward core Clino, although deeper, bottomed in younger sediments (latest Miocene). Three major progradational episodes were delineated using seismic stratigraphy, lithostratigraphy, and information on depositional age. Progradation occurred during the late Miocene, late early Pliocene, and latest Pliocene. In the Pliocene shelf and ramp setting margin progradation began during a highstand, but also occurred in a forced-regression-type situation during a fall in sea level. Rapid reef progradation occurred near the end of the Pliocene and early Pleistocene, when the platform had aggraded to a nearly horizontal surface. The transformation from a ramp-type platform topography to a flat-topped platform culminated in the middle Pleistocene. Age constraints across the west margin of GBB indicate that the seismic reflectors that constitute sequence boundaries are synchronous, within our age resolution. As expected in a slope setting, the sequence boundaries represent condensed time periods of both erosion and nondeposition. Downslope plankton-rich onlapping units are correlated to upslope marine hardgrounds and are thought to represent periods of falling sea level. Subsequent high-resolution dating of additional basinal and shallow-platform borings will provide a rich archive of paleoceanographic changes and will serve as a direct link between the deep-ocean and shallow-platform accumulations.