Site 1005 is the most proximal of the five sites that constitute the Bahamas Transect. Site 1005 is located in 352 m of water, approximately 1150 m from the modern platform edge on the upper slope, 30 shot points SW of the crossing of seismic lines 106 and 107. It is positioned on the thickest portion of the prograding Neogene sequences seen on the seismic line. The target depth of 700 m was designed to penetrate seven seismic sequences. The sea level objectives of Site 1005 were (1) to date precisely the sequence boundaries; (2) to determine the facies within the different systems tracts, especially the nature of the onlapping units that were interpreted as lowstand deposits; and (3) retrieve a high-resolution record of climate and sea-level fluctuations for the Quaternary and late Pliocene. This site served also as the proximal site for the fluid-flow transect. A logging suite was collected to provide a continuous record of the sedimentary succession. In addition, a vertical seismic profile (VSP) was shot for an accurate time-depth conversion of the seismic reflections, so that the cores could be correlated precisely to the seismic data.
A 700-m-thick upper middle Miocene to Holocene section was recovered at Site 1005. The sedimentary succession consists of a periplatform sedimentary section of mixed pelagic and bank-derived carbonates, with a carbonate content of 72-99%. The section is comprised of unlithified to partially lithified wackestones and slightly coarser grained intervals consisting of packstones and grainstones. Compositional variations document an alternating pattern of bank flooding, concomitant shedding to the slope with periods of exposed banks, and a shutdown of shallow-water carbonate production and largely pelagic sedimentation that is recorded in alternating high and low sedimentation rates. The pulses of bank-derived material coincide with the prograding pulses observed on the seismic data that were interpreted to be seismic sequences. This indicates that sea-level changes exceed the major control on the development of the seismic sequences. Foraminifer and nannofossil biostratigraphy yielded precise ages of the lithologic units and seismic sequence boundaries. One of the seismic sequence boundaries coincides with a hiatus that probably straddles the Miocene/Pliocene boundary. Three other boundaries correlate to changes in sedimentation rates, whereas a biostratigraphic hiatus was not detected with the remaining boundaries. The biostratigraphic data from Site 1005 confirm the ages that were carried along seismic reflections from Sites 1003 to 1005. This consistency gives a first indication that the seismic sequence boundaries along the Bahamas Transect are indeed time lines. The cores, in combination with the logging data, provide the facies information in the seismic sequences and systems tracts. Fewer mass-gravity flow deposits are present than at Site 1003, indicating most of the platform-derived turbidites bypassed the upper slope. The thick Quaternary section is characterized by sedimentation cycles with all the characteristics of the periplatform aragonite cycles found in the basin surrounding the Bahamian platforms. These cycles are interpreted to have formed as a result of the high-frequency Quaternary sea-level changes.
As observed at Sites 1003 and 1004, pore-water chemistry profiles at Site 1005 display normal seawater concentrations of all constituents in the upper 45 mbsf, indicating rapid exchange with the water column. Below this zone, organic matter remineralization reactions dominate the diagenetic changes. These reactions are a result of the high in situ abundance of organic material trapped within these marginal sediments during rapid Pliocene-Pleistocene platform shedding events, and provide the fuel that drives the early carbonate alteration reactions in these shallow sediments. Below 400 mbsf, increased salinity suggests the presence of deep-seated saline fluids. Positive salinity anomalies between 50 and 150 mbsf suggest that some of the fluid is derived from the platform surface.
The sedimentary succession of Site 1005 can be subdivided into four units. Although these sediments were deposited in a proximal position on the slopes of GBB, they are generally fine-grained. Their main components are aragonite mud, platform-derived skeletal debris, and pelagic foraminifers and nannofossil ooze.
The basal boundary of Unit I (0-261.7 mbsf; late Pliocene to Holocene) is placed at the onset of peloids and aragonite mud. Variations in abundance of these two components were used to further subdivide the unit. Only a few turbidite layers are preserved as a result of heavy bioturbation that obliterated many primary sedimentary structures. The sediments consist of unlithified to partially lithified mudstones, peloidal wackestones, and packstones. Components in the silt and sand fraction are peloids, benthic foraminifers, Halimeda, and other skeletal debris. Unit IA shows a cyclic-sedimentation pattern with aragonite intervals alternating with more pelagic intervals. Turbidites are concentrated in the aragonite intervals. Subunit IB is partially dolomitized and has a similar composition, but it does not display the pronounced cyclicity of Subunit IA. Subunit IC is also slightly dolomitized and consists mostly of partially lithified mudstones and wackestones with thin layers of clay- to silt-sized laminated intervals. The silt-rich layers are enriched in pelagic and benthic foraminifers, and pellets.
Unit II (261.7-360.9 mbsf, early Pliocene) is a homogeneous section of partially lithified and dolomitized mudstone to wackestones. The constituents are predominantly pelagic foraminifers, with minor amounts of benthic foraminifers, lithoclasts, and unidentifiable bioclasts.
In Unit III (360.9-534.3 mbsf, early Pliocene to late Miocene), aragonite needles or pellets are no longer observed. The succession of partially dolomitized foraminifer wackestones displays a cyclic pattern. Gray to light-gray, well-cemented biowackestones alternate with less cemented gray to olive-gray wackestones with compacted burrows. The boundary between the two lithologies is generally sharp. The average thickness of the compacted intervals in this unit is approximately 44 cm, but is thickest in the bottom portion of the interval.
In the middle to upper Miocene Unit IV (534.3-700 mbsf), a cyclicity similar to Unit III is observed, however, the compacted interval of the cycles is nearly 80 cm thick. The top of Unit IV is marked by a series of thin-bedded turbidites that are intercalated in the cycles. The cycles are interpreted as being formed by variations in the amount of platform-derived material from the slope. The gray, well-cemented intervals have a higher abundance of platform-derived clasts, whereas the darker layers contain more planktonic foraminifers and have a higher organic content.
The cores recovered at Site 1005 document that the upper slope environment of the prograding margin of GBB is dominated by fine-grained sediments. The variations in these mudstones and wackestones are subtle, but nevertheless give a record of high-frequency cyclicity of platform shedding as a result of high-frequency sea-level changes. The observed trend of increasing thickness of the gray lithologies of the cycles within Units III and IV indicate an increase of progradation in these units.
Although Site 1005 was only 1.5 km from the modern platform edge of GBB on the upper slope, biostratigraphic dating was possible, and there was good agreement between the nannofossil and planktonic foraminiferal biostratigraphic events. Biostratigraphy indicates continuous sedimentation from the Pleistocene to the lower Pliocene. A major hiatus occurs across the Miocene/Pliocene boundary that spans from 5.6 to 8.6 Ma. A hiatus of similar duration was also found further downslope at Site 1003. In the previously drilled shallow water platform sites Unda and Clino, the hiatus is not as extensive as more of the upper Miocene section is preserved. At Site 1005, the faunal assemblages below this unconformity are early late to late middle Miocene in age.
The benthic foraminiferal assemblages indicate an upper bathyal paleodepth for Site 1005. The diversity and abundance of platform-derived benthic foraminifers provide an additional record of platform progradation and flooding events. In much of the Pleistocene section, there is a diverse and abundant benthic fauna, whereas a depauperate assemblage is found in the late Pliocene. In the early Pliocene and the Miocene, a low-diversity, shallow-water assemblage is diluted in the monotonous sediments.
Variations in sedimentation rates reflect the export of bank-derived material to the upper slope. The late Pleistocene sedimentation rate is high (15 cm/k.y.), but it is reduced during the late Pliocene and early Pleistocene (3.5-1.2 Ma) to approximately 2 cm/k.y. This latter rate is characteristic of normal pelagic carbonate sedimentation with little to no addition of platform-derived material. The lower Pliocene sedimentation rate was again high (10 cm/k.y.), indicating that the leeward side of GBB received large amounts of platform-derived material during this period. A hiatus that spans nannofossil zone NN11 (5.6-8.6 Ma) was probably caused by the combination of large-scale slope erosion associated with lower sea level and increased current activity, as there is little to no pelagic sedimentation recorded. An expanded lower upper Miocene section was recovered below the hiatus. The high sedimentation rate of 11 cm/k.y. during nannofossil zones NN9 and NN10 reflects another period of platform shedding, consisting mostly of fine carbonate mud to the upper slope. There is minimal input from the platform during nannofossil zone NN8. Another high input of platform derived material occurred during the late middle Miocene yielding sedimentation rates of 13 cm/k.y. The sedimentation rates are in concert with long-term sea-level changes. The two largest known sea-level falls straddle the middle/late Miocene and early/late Pliocene boundaries and are clearly expressed in the sedimentation rates at Sites 1005 and 1003.
A check shot survey (VSP) was performed that enabled us to correlate the cores to the seismic data. The seismic sequence boundaries correlate well with the biostratigraphic data. Three of the six sequence boundaries coincide with changes in sedimentation rates and the late Miocene hiatus falls exactly on a sequence boundary. In addition, the thick seismic sequences coincide with the periods of high sedimentation rates. There is also a good correlation of the seismic sequence boundaries with the changes in mineralogy. In the Pleistocene section, the sequence boundaries coincide with intervals with lower percentages of high-Mg calcite and aragonite that record reduced platform-derived sediment input. In the Pliocene and Miocene section, the boundaries are either associated with high amounts of insoluble residues or peaks of dolomite. The overall monotonous sediments do not change dramatically at sequence boundaries, but they generally correlate to mudstone intervals that contain a larger amount of nannofossil ooze. The oldest sequence boundary recovered in the middle Miocene is marked by a firmground. The log data correlate well with the seismic sequence boundaries. Not surprisingly, the velocity profile has the strongest correlation.
The log data correlate well with the sedimentary succession and can be used to fill in gaps in low-recovery zones. The strong cyclic nature of the data is the most notable feature recorded in all the logs, specifically from 700 to 385 mbsf (middle Miocene to lower Pliocene) and from 260 to 90 mbsf (upper Pliocene to Pleistocene). The intervening section from 385 to 260 mbsf (lower Pliocene) is marked by an apparent stark diminution of the intensity of the cycles, as reflected in particular by the regularity of the resistivity curve. Between 260 to 90 mbsf, the sharp increases in gamma-ray intensity correlate with increased density, resistivity, and velocity and decreased porosity. This association is indicative of more indurated sediment deposited during periods with a reduced sedimentation rate leading to hardground formation. The logs clearly display that sedimentation was punctuated by periods of decreased sediment input that are marked by the formation of better cemented layers.
In summary, the sedimentary, mineralogical, geophysical, and stratigraphic data record a sedimentary system characterized by variable input of platform-derived sediment that is most likely caused by changing sea level. These fluctuations of input occur on several levels. A high-frequency, cyclic alternation is seen in the sedimentary and log data, whereas the sedimentation rates monitor the long-term changes. The seismic sequences reflect these long term changes and also provide a record on an intermediate scale.
The interstitial water chemistry at Site 1005 yielded interesting geochemical profiles that were subdivided into four zones. The top 45 mbsf, zone 1, displays no change in most measured constituents. This zone probably experiences pervasive flushing of seawater that prevents diffusional gradients from developing between the overlying seawater and the underlying saline fluid. Zone 2 extends to a depth of 190 mbsf and is characterized by a sharp change in the gradient of all major and minor pore-water constituents. Cl- is enriched in this zone, which is interpreted to be an intrusion of saline, sulfate-rich water derived from the shallow platform. High rates of microbial sulfate reduction within this zone reduce the sulfate concentration to zero and give all major ion profiles an anomalous appearance. Most constituents in the underlying zones 3 and 4 display a steady change to the bottom of the hole. A shift in many of the profiles marks the boundary between zone 3 and 4. At this time it is not known to what extent the shifts are influenced by diagenetic alterations within lithological units or fluid flow within these boundaries.
The cores and data collected at Site 1005 provide the necessary information to answer several of the questions addressed before drilling this site. In regard to the sea-level objectives, the data corroborated the results of the more distal Site 1003. Site 1005 also yielded abundant biostratigraphic markers that allow for the precise age dating of sequence boundaries. Furthermore, the ages can be carried along the seismic reflection horizons to Site 1003 where they fall on the same stratigraphic level. This consistency documents that the seismic reflections are indeed also time lines and have a chronostratigraphic significance. This is possible because the impedance that causes seismic reflections is controlled by both original sediment composition and the diagenetic overprint. In these carbonate slope sediments, sediment composition itself is largely controlled by the input of platform-derived sediment, which in turn is related to sea-level fluctuations. Early diagenetic alteration was strongest when the platform was exposed and sedimentation rates were reduced. Thus, both facies and diagenesis are related to sea-level fluctuation, which leaves its expression in physical properties and finally in the seismic sequences.
In regard to the fluid-flow objectives, the geochemical profiles of Site 1005, in conjunction with Sites 1003 and 1004, provided further evidence for fluid flow through the slope of GBB. The consistency of changes in geochemical gradients at distinct stratigraphic horizons indicate that the flow has a horizontal component in the upper sediments that is superimposed on the vertical diffusion in the lower portions of the sedimentary column.