-
(5)
Figure 3 shows pH, alkalinity (Davis, Fisher, Firth, et al., 1997), the free energy of calcium carbonate dissolution, the Sr content, and the Mg/Ca and Sr/Ca ratios of the pore waters for Sites 1023-1031. Magnesium and strontium are common substitutes for calcium in calcium carbonates. The calcium data is the shipboard data (Davis, Fisher, Firth, et al., 1997). The Sr concentrations have been measured in Toulouse by inductively coupled plasma-mass spectrometer (ICP-MS) using Indium as an internal standard (Freydier et al., 1995) and are reported elsewhere (Mottl et al., Chap. 9, this volume). We correlate the behavior of Mg and Sr in pore waters to the calcium carbonate saturation state of the pore waters inferred from our calculations.
It can be seen in Figure 3 that most of the pore waters are supersaturated with respect to both calcite and aragonite. At all sites, the reaction affinity increases from close to equilibrium values at the seafloor and reaches an almost constant value at depth of 2-3 kJ/mol. Pore-water calcium (Davis, Fisher, Firth, et al., 1997) and strontium (Mottl et al., Chap. 9, this volume) concentrations decrease in the first 10-20 mbsf. In general, the strontium concentration in interstitial waters of marine sediments increases as the result of calcium carbonate recrystallization (Morse and Mackenzie, 1990, p. 402; Elderfield et al., 1982; Oyun et al., 1995) in accordance with
(5)
in which y > x. The reaction represented by Equation 5 consumes calcium from the pore water and releases strontium to it, hence increasing the Sr/Ca ratio of the aqueous phase. This has been shown (Elderfield et al., 1982; Oyun et al. 1995) for sites where pore-water sampling was not as detailed as that of Leg 168. For example, Elderfield et al. (1982) used sediment and pore-water composition data for DSDP Sites 288 and 289 for which only two samples have been collected in the upper 100 mbsf. During Leg 168, five samples were collected in the first and last cores (first 10 m and last core above basement) and then at least one sample in each core in between. Faure and Powell (1972, p. 79) report that "precipitation of calcite from a solution containing Sr2+ will increase the Sr/Ca ratio of the aqueous phase, while precipitation of aragonite at temperatures below 50°C will decrease this ratio." The Sr/Ca ratio of the aqueous phase in Leg 168 pore waters increases downhole from the seafloor (Fig. 3), consistent with this assertion. Sr substitution for Ca in calcite can explain the decrease in the pore-water Sr concentration in the first tens of meters of the sedimentary column from the seawater value of 89 µmol/kg at the seafloor. This requires that there is no dissolution and subsequent recrystallization of biogenic carbonates.
At Sites 1023, 1024, and 1025, the affinity of the calcite dissolution reaction increases to about 3 kJ/mol and then decreases to reach values close to equilibrium near basement (Fig. 3). It is interesting to note that, at Sites 1024 and 1025, the deepest points are on the equilibrium line. Sr is almost constant at Site 1023 (Fig. 3) and only increases slightly at Site 1024 (Fig. 3B), whereas its value reaches 110 µmol/kg before returning to modern seawater values at Site 1025 (Fig. 3C). It can be also noticed that, at Sites 1023 and 1024, alkalinity reaches its maximum value at a depth where sulfate begins to be totally depleted in the pore waters. It also coincides with the onset of methane production, which occurs at about 50 mbsf at Sites 1023 and 1024 (Davis, Fisher, Firth, et al., 1997). One can see in Figure 3 that the alkalinity maximum is also concomitant with a marked change in the pore-water Mg/Ca and Sr/Ca ratios at Sites 1023 and 1024. These ratios are almost constant at Site 1025 before decreasing below a depth of 20 mbsf. There is no methanogenesis at Site 1025.
In Leg 168 sites, the warmest temperatures have been found at Sites 1026 and 1027, around 63°C, at the sediment/basement interface (Davis, Fisher, Firth, et al., 1997). This temperature homogenization is attributed to the vigor of fluid circulation within the basement. At Site 1026, there is an increase in the affinity of reaction to about 3 kJ/mol, then a decrease to lower values scattered between 0 and 2 kJ below 150 mbsf (Fig. 3D). There is a marked increase in the scatter of the pH data below 130-150 mbsf as well as a trend to higher values (Fig. 3D). This change can also be seen in the Sr data, which show a local maximum at this depth (130 mbsf). The strontium concentration in pore waters reaches 170 µmol/kg before decreasing again when basement is approached (Fig. 3D). Despite this complex variation in Sr concentration with depth, the Mg/Ca and Sr/Ca ratios at Site 1026 resemble those found for the Hydrothermal Transition sites.
At Site 1027 (Fig. 3E) pore waters remain supersaturated down to basement. This is consistent with the massive calcium carbonate occurrence in veins crosscutting basement rocks at this location. Only one example of such calcite-containing veins has been found at Site 1026 (Davis, Fisher, Firth, et al., 1997, p. 128). At Site 1027, the strontium concentration increases to very large values (~360 µmol/kg) before displaying a large decreasing gradient near basement.
Pore-water chemical results from Sites 1028 and 1029 are similar (Fig. 3F, Fig. 3G) with the affinity of reaction being 1-2 kJ/mol throughout the sediment column. At Site 1028 (Fig. 3F), Sr first decreases, then increases to about 135 µmol/kg, and then decreases again close to basement.
At Site 1029, the variation in strontium concentration with depth (Fig. 3G) shows the same complexity as that at Site 1026 (Fig. 3E). There is also a marked reversal in the Sr behavior near basement. The Mg/Ca and Sr/Ca values at this site resemble those found at other sites: an almost linear increase from the seafloor and then a sharp reversal toward low values at depth.
Sediment cores were retrieved at Site 1032 from 195 mbsf to basement at 285 mbsf. The data depicted in Figure 3I show that there is a large scatter in pH at depth. This pH variation leads to reaction affinities between 0.5 and 2 kJ/mol, which may be a little bit lower than was found in deeper sediment sections of other Leg 168 sites.
At Sites 1030 and 1031, which are located above a basaltic outcrop buried under 40-50 m of sediments, upwelling of basement fluid through the sediment cover has been identified by the characteristic shape of the pore-water concentration profiles of elements, especially magnesium and chlorinity (Davis, Fisher, Firth, et al., 1997). Strontium displays such a behavior: its concentration increases from the seawater values (89 µmol/kg) and then reaches a constant value of 110 µmol/kg at Site 1030 and 111 µmol/kg at Site 1031 (Fig. 3H, Fig. 3I). These values are similar to the strontium concentration of the Baby Bare spring fluid (110 µmol/kg; Mottl et al., 1998) and of the basement fluid sampled with the downhole water sampler temperature probe (WSTP; Fisher et al., 1997) at Hole 1026B (111 µmol/kg, Mottl et al., Chap. 9, this volume). Whereas alkalinity generally increases downhole as a result of bacterial sulfate reduction at other Leg 168 sites, it continuously decreases at Sites 1030 and 1031 to very low values that are also comparable to the alkalinity of the Baby Bare spring fluids (Fig. 3H, Fig. 3I). The effects of diagenetic reactions like the decrease in sulfate and the alkalinity production are masked by the upwelling fluid. Because the rate of advection is faster at Site 1031 than at Site 1030, alkalinity at Site 1030 is higher than at Site 1031 (Davis, Fisher, Firth, et al., 1997). Calcite or aragonite are at equilibrium with the pore waters from basement at 45 mbsf up to a depth of 30 mbsf, where the pore waters become supersaturated up to the seafloor (Fig. 3H). Equilibrium between pore waters and calcium carbonate is reached at Site 1031 in almost the entire sediment column, with a trend to slight undersaturation just below the seafloor (Fig. 3I). These two sites are the only ones among the Leg 168 sites where frequent dissolution of foraminifers and coccolith tests have been observed by scanning electronic microscopy (Buatier et al., in press). This is also consistent with micropaleontological observations (X. Su, unpubl. data). Carbonate dissolution is linked to smectite and zeolite formation in altered layers of the sediment (Buatier et al., in press). There is a slight change in the Sr concentration (Fig. 3I) for the last two samples near the sediment/basement interface where the sediment alteration is most intense (Buatier et al., in press).
For an infinite rate of advection, a fluid upwelling through the sediment cover will reach the seafloor unaltered because the rate of advection is faster than the rate of chemical reactions. Inversely, the fluid would be at equilibrium with calcite for an infinite rate of calcite dissolution or precipitation. At Sites 1030 and 1031 the distribution of alkalinity (and of the affinity of the calcite/aragonite dissolution reaction) is the result of the competition between the rate of fluid advection and the rate of chemical reactions. If we assume that a value of the affinity of reaction of about 2 kJ/mol is representative of calcium carbonate precipitation at most Leg 168 sites (Fig. 3), it takes a longer distance (i.e., sediment thickness) to reach this representative value in the case of a rapidly advecting fluid at Site 1031 than in the case of the slower fluid at Site 1030 (Fig. 3H, Fig. 3I).
Several processes can lead to calcium carbonate supersaturation, which can result from the inhibition of precipitation. This can be due to the poisoning of reactive surfaces by organic matter. It is also well known that magnesium and orthophosphate are inhibitors of calcite precipitation. It is interesting to note that organic matter oxidation provides the needed reactant (alkalinity) for calcium carbonate formation at the same time that it provides phosphate to the pore water. A detailed study of the mechanisms of calcium carbonate formation in the sediments of the eastern flank of the Juan de Fuca Ridge is needed to elucidate the problem.
