STABLE ISOTOPIC COMPOSITION
OF CARBONATES

Results

The oxygen and carbon isotopic compositions of calcite in Hole 994C have been analyzed to provide a record of the pelagic carbonate sedimentation and to identify the possible effects of diagenesis on the calcium carbonate fraction. Variations in the vertical ¹⁸O profiles show very sharp negative excursions, with values down to -5.37, within a few levels of the upper 140 m of the sedimentary succession that correspond to the dolomite-rich intercalations; however, there are no specific ¹³C anomalies associated with these ¹⁸O decreases (Table 2; Fig. 2). Except in these specific levels, the values of calcite vary within a narrow range between 1.68 and -0.09 for ¹⁸O, and between 1.00 and -1.55 for ¹³C, as expected for Cenozoic marine deposits. In Holes 995A and 997A, similar low ¹⁸O values of calcite are measured in the dolomite-rich levels (Table 1). The smear-slide observations do not reveal the presence of detrital calcite or of specific features as dedolomitization, that could explain the oxygen isotopic anomalies. It is thus thought that the negative excursions in the ¹⁸O values of calcite from the dolomite-rich intercalations are a result of the crystallization of diagenetic calcite in contact with modified pore waters characterized by low ¹⁸O values.

The oxygen and carbon isotopic compositions of the diagenetic carbonates from the Cape Fear Diapir sediments (Sites 991, 992, and 993) and from the Blake Ridge sediments (Sites 994, 995, and 997) are presented together, although the two areas are not equivalent with respect to the sedimentary succession and to the stability of the gas hydrates (Table 1; Fig. 3). The ¹⁸O values of dolomites generally decrease with depth from high values (maximum 7.53) in the upper dolomite layers to minimal values (minimum -2.68) in the dolomites from levels between 86 and 138 mbsf. In the underlying siderites, an abrupt increase of the ¹⁸O values occurs, and the values remain relatively stable in the range 2.05 to 3.47 within the gas hydrate layer. Below the BSR, the ¹⁸O variations of siderite become greater with values ranging from 1.85 to 5.05. The ¹³C distribution with depth is opposite to the ¹⁸O distribution. Low values are measured in the dolomite of the uppermost 80 m, with the lowest values (-31.26) in the Cape Fear Diapir dolomite, while the minimal values of Blake Ridge dolomite reach -10.21. In the zone between 80 and 138 mbsf, the ¹³C values of the dolomite increase rapidly from -7.42 to 1.66. Beneath this transition zone, siderites exhibit high ¹³C values (1.74 to 7.81) without any systematic change above, within, or below the gas hydrate-rich sediments.

Discussion

The carbonates that are intercalated as nodules and lamina of dolomite and siderite within the sediments of the Blake Ridge are conspicuous diagenetic products related to the presence of gas hydrates. These diagenetic minerals occur also at more discrete concentrations everywhere in the sedimentary succession. Furthermore, it was demonstrated that diagenesis may locally alter the CaCO₃ isotopic composition, either by their recrystallization, or by precipitation as is evidenced by the presence of authigenic calcitic rhombs.

The stable isotopic composition of these carbonates may help to decipher which processes modified the pore water and CO₂ reservoirs and were involved in carbonate precipitation.

The oxygen isotopic composition of a carbonate depends on temperature and on the oxygen isotopic composition of the water where the carbonate precipitates. The carbonate-water ¹⁸O equilibrium may be evaluated for calcite, dolomite, and siderite at any given temperature by the experimental equations:

Calcite 1000 ln = 2.78 10⁶ T^-2 - 2.89 (O'Neil et al., 1969)

Dolomite 1000 ln = 2.62 10⁶ T^-2 + 2.17 (Fritz and Smith, 1970)

Siderite 1000 ln = 3.13 10⁶ T^-2 - 3.50 (Carothers et al., 1988)

In these equations, represents the oxygen isotope fractionation between the carbonate and the water [ carbonate-water = (1000 + carbonate) / (1000 + water)], and T is the absolute temperature.

If it is assumed that the present in-situ temperature corresponds more or less to the equilibrium temperature at which precipitation occurred, the ¹⁸O value of the water in equilibrium with the carbonate may then be calculated. Such calculations were made for the individual diagenetic minerals using the downhole temperature measurements (Paull, Matsumoto, Wallace, et al., 1996; Ruppel, 1997).

During gas hydrate formation, ¹⁸O-rich water molecules enter preferentially into the solid phase so that the remaining liquid water becomes ¹⁸O depleted. The ¹⁸O fractionation factor between hydrate and liquid water has been estimated experimentally to be 1.0026 (Davidson and Leaist, 1983); oxygen isotopic measurements on gas hydrate and interstitial water from Leg 164 give value in the range of 1.0034 to 1.0040 (Matsumoto et al., Chap. 2, this volume). In open systems, the isotopic effects should be negligible except when huge amounts of gas hydrates are formed. In partially closed systems, as is the case for interstitial waters, the fractionation effects may be considerably increased, and the residual solution after gas hydrate crystallization may be depleted by a few per mil in ¹⁸O relative to the original water (Ussler and Paull, 1995). Conversely, the decomposition of gas hydrates liberates ¹⁸O-rich water molecules that can contribute significantly to the ¹⁸O-enrichment of the interstitial solutions.

In the Cape Fear Diapir, the ¹⁸O values calculated for the diagenetic waters range between 0.5 and 1.8. They are thus higher than the average ¹⁸O value of 0.1 of the bottom waters, which correspond to the North Atlantic Deep Water (Craig and Gordon, 1965; Pierre et al., 1991, 1994). The ¹⁸O-enrichment in the interstitial solutions indicates that the diagenetic dolomites may have been precipitated in waters originating partly from gas hydrate decomposition.

In the Blake Ridge area, three zones can be distinguished based on the ¹⁸O values calculated for the diagenetic waters. The uppermost 140 m is characterized by low ¹⁸O values (-1.4 to -7.8) of the calculated diagenetic waters, with rapid and large changes with depth; the ¹⁸O-depletion is interpreted as being caused by the local formation of gas hydrates, under partially closed-system conditions. These anomalously low ¹⁸O values are characteristic of levels where dolomite makes up between 5% to 10% of the total sediment. It is inferred that this type of dolomite was precipitated within ¹⁸O-depleted solutions at times (glacial periods?) when gas hydrate formation was more important than today because of lower bottom-water temperatures. The local inhomogeneity of the diagenesis is also demonstrated by comparing the ¹⁸O values of the water, which are calculated in the levels where calcite and dolomite are both assumed to be diagenetic. These values, which may differ by a few per mil, indicate that these minerals indeed are not cogenetic (Table 1). In the underlying gas hydrate-rich sediments below 140 mbsf, the ¹⁸O values of the calculated diagenetic waters vary within a narrow range (-2 to 0.2). This indicates gas hydrate formation in this layer, where steady-state homogeneous conditions were maintained. Below the BSR, the ¹⁸O values of the calculated diagenetic waters become positive (0.8 to 3.5), showing that ¹⁸O-rich waters are released from gas hydrate decomposition.

The ¹⁸O values calculated for the diagenetic waters differ significantly from the ¹⁸O values measured in the interstitial solutions from Sites 994 and 997 (Matsumoto et al., Chap. 2, this volume), except in the depth range from 300 to 500 m, where the calculated and measured values are very similar. In fact, because of the high vertical flow rate, estimated as 0.2 mm/yr (Egeberg and Dickens, in press), the in situ pore waters are much younger that the sediment. This means that the isotopic compositions of the diagenetic waters deduced from those of the diagenetic carbonates cannot be compared to the isotopic compositions of the interstitial waters from nearby levels.

The carbon isotopic composition of a carbonate is related to the balance of the inorganic and organic sources of carbon, which control the carbon isotopic composition of the CO₂ reservoir. In organic-rich environments, bacterial processes dominate, including (1) bacterial sulfate reduction, in which reduced carbon is oxidized to CO₂, occurs in the uppermost sediments where sulfate ions are available; and (2) downward, in the sulfate-free zone, methane-generating bacteria monitor the fermentation and carbonate reduction processes, both of which produce methane. The simplified equations of these reactions may be written as the following:

2 CH₂O + SO₄^2- 2 CO₂ + S^2- + 2 H₂O (1)

CH₄ + SO₄^2- CO₂ + S^2- + 2 H₂O (2)

2 CH₂O CH₄ + CO₂ (3)

CO₂ + 8e^- + 8 H⁺ CH₄ + 2 H₂O (4)

In bacterial sulfate reduction reactions 1 and 2, the produced CO₂ is depleted in ¹³C because the carbon originates from ¹³C-depleted sources, organic matter (¹³C~-25), and methane (-100 <¹³C <-40).

In the bacterial fermentation reaction 3, the carbon isotopic fractionation between the two by-products may reach 80 (Rosenfeld and Silverman, 1959); the light carbon is preferentially transferred to CH₄, while the CO₂ is highly enriched in ¹³C relative to the initial organic substrate. In the bacterial carbonate reduction reaction 4, the ¹²C-rich CO₂ is used preferentially by bacteria to form CH₄ and the residual CO₂ become enriched in ¹³C as the reaction proceeds. Therefore, the CO₂ produced or remaining during methanogenesis is characterized by high ¹³C values.

The low ¹³C values of the dolomites from the Cape Fear Diapir indicate that CH₄ was the energy source involved in the bacterial sulfate reduction. The range of ¹³C values of the diagenetic carbonates of the Blake Ridge indicates that the isotopic composition of the CO₂ reservoir was most likely controlled primarily by fractionation during the bacterial carbonate reduction process. Similar conclusions were previously proposed by Claypool and Threlkeld (1983) to explain the carbon isotopic behavior of the CO₂ of pore waters from Site 533 of DSDP Leg 76.

Combining the mineralogical information and the stable isotopic data on diagenetic dolomites from Leg 164 provides interesting information. When the ¹⁸O and ¹³C values of dolomites are plotted against the d (104) values, correlations appear between these parameters that are characteristic of the composition of the diagenetic solutions (Fig. 4). The ¹⁸O increase and ¹³C decrease are linearly correlated with the d (104) increase. This suggests that gas hydrate dissociation releases methane, which is used for bacterial sulfate reduction, and enhances Fe incorporation in the dolomite crystal lattice. Conversely, gas hydrate formation reduces the level of incorporation of Fe into the dolomite crystal lattice, but it promotes the precipitation of siderite.