DISCUSSION

The ¹⁸O values in the upper portion of the sediments in conjunction with the uniform concentration of Sr²⁺ and Cl^- (Table 6; Fig. 6) in the pore fluids support our ideas of active fluid movement in this portion of the profile (Shipboard Scientific Party, 1997). It is also possible that a period with a very high rate of sedimentation could have produced a pattern in the pore-water geochemistry similar to that observed. In order for this mechanism to be responsible, the entire thickness of the flushed zone must have been deposited over a geologically short period of time. Although this mechanism is unlikely based on the present rates of sedimentation (Shipboard Scientific Party, 1997), it is possible that rates of deposition are outside the range of the chronostratigraphic markers used in this study. Some support for this idea might be provided by the presence of a thinner flushed zone in Sites 1003 and 1007, which shows depositional rates of ~5 cm/k.y. over the upper 50 mbsf compared to 10-15 cm/k.y. at Sites 1003 through 1005.

Within the reproducibility of the replicate ¹⁸O analyses and the standard deviation of our standards, there appears to be some variation in the upper portions of the flushed zone that is not evident in the Cl^- data. It is possible that this variability is real; alternatively, it is possible that it results from nonuniformity in the manner in which the samples were sealed. The increasing ¹⁸O of the pore waters in the flushed zone closer to the margin of the carbonate platform probably indicates that the bottom waters at these sites originated from the surface of the Great Bahama Bank—perhaps during the winter, when cold, dense saline water is known to cascade over the margin.

The increase in the concentration of chloride with increasing depth noted at all the sites during Leg 166 was postulated to be a result of the diffusion of Cl^- from an underlying evaporitic source (Shipboard Scientific Party, 1997). Such a source was drilled at Site 627, north of Little Bahama Bank (Austin, Schlager, et al., 1986), and pore waters showed a similar increase in Cl^- approaching this source (Swart and Guzikowski, 1988). However, analysis of the Na⁺/Cl^- ratios of the interstitial pore-water data from Leg 166 reveals no evidence of a NaCl source. Perhaps a more likely origin is simply evaporated seawater that has become entrapped in rocks during sea-level low stands. However, the evaporated seawater origin is not supported by absence of a correlation between ¹⁸O and D. Therefore, the origin of the increase in salinity is still an enigma: whereas the absence of a correlation between ¹⁸O and D (Fig. 11) suggests the dissolution of NaCl, the absence of a change in the Na/Cl ratio with increasing depth tends to favor the presence of a deep-seated saline fluid.

Typically, pore waters in deep-sea sediments show a decrease in their ¹⁸O values as a result of the interaction between pore water and igneous basement rocks (Lawrence, 1989). In the Bahamas, because the influence of interactions with basement rocks is likely to be minimal at the depths drilled in this study, the observed increase in ¹⁸O at most sites is best explained as a result of the recrystallization of biogenic carbonate components in the presence of a geothermal gradient. The magnitude of the expected changes resulting from recrystallization of carbonates will depend upon the amount of recrystallization, the geothermal gradient, and the ¹⁸O of the initial sediment and pore water. Using the equations presented by previous researchers (Lawrence, 1973, 1989; Lawrence et al., 1976; Killingley, 1983; Stout, 1985), the isotopic composition of the pore water after a specified amount of recrystallization can be calculated from the following equation:

(1)

In this equation:

w_e = ¹⁸O of interstitial water after recrystallization,

C_i = ¹⁸O of calcium carbonate of initial sediment,

w_i = ¹⁸O of interstitial waters before reaction,

M_c = mole fraction of oxygen in carbonate sediment,

M_w = mole fraction of oxygen in pore waters,

R = percentage of recrystallization, and

= fractionation factor between calcite-water.

As an example, Figure 12 shows the evolution of the interstitial water (w_e) in the presence of various different conditions that are likely to occur during burial. Case A represents a geothermal gradient of 50°C/km, an initial sediment ¹⁸O value of +0.0 Peedee belemnite (PDB), an initial water ¹⁸O value of 0.5 SMOW, and a recrystallization rate of 1% per 5 m. Case B represents the same conditions with the exception of a geothermal gradient of 70°C/km. Case C represents the same conditions as Case A except that the initial oxygen isotopic composition of the water is +0.5. In all cases shown in Figure 12, the ¹⁸O increases to values consistent with those observed at the Leg 166 sites.

The only exception to the trend of increasing pore-water ¹⁸O with increasing depth was seen at Site 1006 (Fig. 10). The absence of an increase in ¹⁸O at this site is probably a consequence of reduced amounts of carbonate diagenesis compared to the other more proximal locations (Eberli, Swart, Malone, et al., 1997; Shipboard Scientific Party, 1997). The relative absence of carbonate diagenesis is evident from the excellent preservation of the foraminifers and nannofossils, which made this site a particularly good one for biostratigraphic purposes (Shipboard Scientific Party, 1997).

Water Movement into the Platform

Some support for water being drawn into the platform was provided by unusual chloride profiles, particularly at Sites 1004 and 1005 (Fig. 3). These profiles suggested that less-saline bottom water was being drawn into the platform, perhaps by a mechanism such as Kohout convection (Kohout, 1966; Simms, 1984). An alternative scenario is that more-saline water, originating from the platform top, is sinking downward through the upper sediments. Although the anomaly in the chloride was also present as slightly depressed ¹⁸O values, the data are consistent with either explanation for the origin of the Cl^-anomaly.