The sedimentary sequence at Hole 1006A is remarkably continuous, expanded, and precisely dated. Therefore, this site is an excellent location to compare isotopic and sedimentary records and will probably become a classic site for late Cenozoic paleoceanography in the low-latitude Atlantic (Eberli, Swart, Malone, et al., 1997). The early Pliocene-middle Miocene bio-cyclostratigraphy shows that the sequence is complete and it provides an accurate estimate of sedimentation rates. The variation in sedimentation rates is crucial for understanding sediment production on the platform over time in relation to sea-level changes. Here, we have an ideal opportunity to study the Great Bahama Bank platform production from early Pliocene to middle Miocene times.
The variation of sedimentation rates in periplatform deposits is mainly controlled by the switching on and off of the carbonate factory during sea-level highstands and lowstands, respectively (Droxler and Schlager, 1985; Reijmer et al., 1988). During flooding, the platform produces increased amounts of neritic material that gets swept onto the upper slope of the basin (highstand shedding). In contrast, during lowstands, the platform production is drastically reduced and in some cases, during large sea-level falls, the platform might have been exposed. If exposed, coarser grained material is deposited onto the very upper slope (Rendle et al., Chap. 6, this volume). The Pleistocene sea-level cycles had a large effect on platform production and shedding of material toward the leeward side of the Great Bahama Bank (Kroon et al., Chap. 2, this volume; Rendle et al., Chap. 6, this volume) according to the above-described scenario. Here, we assume that a similar mechanism was responsible for the cyclic sedimentation in the lower Pliocene-middle Miocene record, although sea-surface temperature variations may also be important. At this stage, we can not prove exactly what caused the cyclic variations because paleoceanographic proxies such as stable oxygen isotope variations have not been produced as yet, but are in progress (McKenzie et el., 1998; McKenzie et al., unpubl. data). Also, the bank had more ramp-like morphology in the Miocene period, so it would have had a slightly different style of carbonate production over the sea-level cycle.
The cyclostratigraphy appears to be extremely useful for building an accurate 'floating' age model based on the assumption that precessional cycling forced the record. This assumption is reasonable because long-term sedimentation rate changes based on cyclostratigraphy are compatible with those based on biostratigraphy. By counting the cycles induced by precessional forcing, accurate sedimentation rate variations can be calculated.
The sedimentation rate at Site 1006 generally varies between 4 and 8 cm/k.y. (Fig. 7), apart from the faster rates of ~5-20 cm/k.y. in Unit 2 (3.6-4.6 Ma). The higher rates are above normal rates of pelagic sedimentation, particularly at the extreme upper end of its range. Carbonate fluxes (there is not much terrigenous material in this section) must have been very high to increase sedimentation rates above normal pelagic and thus the carbonate platform was also shedding material to the slope throughout the cored interval.
The faster sedimentation rate of ~5-20 cm/k.y. (Fig. 7) in Unit 2 (3.6-4.6 Ma) coincides with the interval where there are two resistivity log peaks per precessional cycle, and also where there is an absence of hardgrounds. It is not known why the resistivity log shows a double peak. The precessional cycle is evident in the cores by alternating dark and grey layers. Light grey wackestones/packstones contain high amounts of fine-grained shallow-water bioclasts. Dark grey wackestones/packstones are dominated by pelagic components and contain up to 15% clay (Betzler et al., unpubl. data). The lithology, therefore, suggests a highstand vs. a lowstand deposit during a sea-level cycle at the precessional frequency. However, there is not a clear lithological expression of a double cycle in the cores; thus, the resistivity signal must respond to different parameters that are not yet known such as cyclical cementation horizons.
The elevated sedimentation rates in Unit 2 (Fig. 8) started shortly after the Miocene/Pliocene boundary (5.33 Ma; Hilgen et al., 1995), which occurs at 371 mbsf according to the biostratigraphy. The cyclostratigraphy would put the boundary at similar levels, albeit slightly deeper in the section at 375 mbsf. The timing of the increase in sedimentation rate coincides with the final step in the closure of the Panama Isthmus. We suspect that the closure led to reorganization of surface currents in the Florida Strait and Santaren Channel and the location and strength of the currents may have contributed in shifting depositional centers in the Strait. Another possibility is increased platform shedding into the basin expanding the lower Pliocene section. However, not all the records of Leg 166 sites drilled higher up the slope show an expanded section in this time interval and thus the change in currents and depocenter location is the more likely scenario.
The sedimentation rate changes in Figure 8 show cyclicity at long wavelengths. Spectral analysis of this time-series (Fig. 9) reveals that most of the variability in sedimentation rate is at periodicities of 88, 124, 440, and 1800 k.y., which are very similar to known periodicities in the orbital eccentricity (96, 125, 400, and 2000 k.y.). The pulses of sediment from the bank are paced by the precessional cycle. Thus, the platform produced a carbonate pulse every precessional cycle, through its control on sea level, and perhaps also its control on temperature as a direct result of low-latitude insolation. From our analysis of the sedimentation rate record, the amount of carbonate material shed into the basin during those pulses is controlled by eccentricity, likely because eccentricity modulates the amplitude of the precessional insolation cycle and, therefore, the amplitude of the sea-level variation. It appears that when the platform experiences a sea-level cycle of higher amplitude, it produces and exports more carbonate than when the sea-level cycle is more modest. Short- and long-term eccentricity modulation of the precessional cycles resulted in periods of extreme and minimal platform production.
We conclude that the logs reveal the sedimentation rate variations at Site 1006. The sedimentation rate changes are interpreted to be controlled by both platform production and variability of the currents. The sedimentation rate variations occur in regular cycles controlled by orbital eccentricity. The bundling of precessional cycles into packets of 100 k.y., 400 k.y., and 2 m.y. shows that eccentricity plays an important role in modulating the amplitude of sea-level changes during a period when there was less ice than in the Pleistocene.