GEOCHEMISTRY

Volatile Hydrocarbons

As part of the shipboard safety and pollution program, volatile hydrocarbons (methane, ethane, and propane) were measured in the sediments of Site 1090 from every core in Hole 1090B using the standard ODP headspace sampling techniques (Table T11; Fig. F20). Headspace methane concentrations were generally low (2-9 parts per million by volume [ppmv]) throughout the sedimentary sequence at Site 1090. Ethane, propane, and other higher molecular weight hydrocarbons were not observed.

Interstitial Water Chemistry

Shipboard chemical analyses of the interstitial water from the sediments at Site 1090 followed the procedures for Sites 1088 and 1089. The results from the shipboard analyses (Table T12; Fig. F21) were obtained from 34 interstitial water samples from Hole 1090B to a depth of 392 mbsf. Generally, interstitial water samples were taken from every core to 259 mbsf (Core 177-1090B-28X). However, Cores 177-1090B-27X, 29X, and 30X were moderately to severely disturbed; the whole rounds taken from these cores were considered unacceptably contaminated and, therefore, were not squeezed. In Core 177-1090B-31X, the core liner was fractured and no interstitial water sample was taken. Fragments of chert were found in Section 177-1090B-31X-CC and the top of Core 177-1090B-32X at ~290 mbsf. Below 326 mbsf, an interstitial water sample was taken from every other core. A second chert layer was recovered in Section 177-1090B-38X-1 at ~340 mbsf (see "Lithostratigraphy" and the "Core Descriptions" contents list for core images).

The interstitial water profiles determined by shipboard analyses for most of the chemical species at Site 1090 show relatively modest gradients downhole to 259 mbsf. Below 290 mbsf, however, there is a significant offset in nearly all of the profiles (Fig. F21). Although we have no interstitial water samples in the interval from 259 to 293 mbsf, the lack of any detectable transitional gradients either above or below this interval suggests that an impermeable layer (presumably the chert layer found in Core 177-1090B-32X) was a barrier to diffusion in interstitial waters. Thus, the interstitial water gradients above the chert layer at 290 mbsf have evolved independently of the interstitial water gradients below this layer since the formation of the impermeable barrier. This unique circumstance may permit insight into the processes controlling the behavior of several dissolved species. There may also be a smaller offset in some profiles below the second chert horizon at 340 mbsf (Fig. F21) but, because of low data density below 290 mbsf, the following discussion will be limited to those profiles that do not show obvious effects of the second chert layer observed at 340 mbsf.

The trends in some of the species would be nearly identical to those observed in Site 1088 if it were not for the chert layer at ~290 mbsf (Fig. F22), whereas other trends are not affected by the layer. In the uppermost part of the profile, chlorinity increases from 555 mM at 2.5 mbsf to 564 mM at 29 mbsf, similar to the increase in chlorinity seen at Site 1088. This trend is clearly not influenced by the deep diffusional barrier, for the reasons discussed in previous chapters (the downward diffusion of Cl^- with a glacial signature). However, the chloride concentration reaches a relatively constant value of ~564 mM at Site 1090, which is lower than the value of ~568 mM at Site 1088 (Fig. F22). This characteristic probably is related to the chert layer, because the water mass bathing the two sites (and diffusing into the interstitial waters) cannot have been much different.

As at Site 1088, Ca⁺² and Mg⁺² show an inverse behavior at Site 1090, and Sr⁺² increases with Ca⁺². Above the chert layer at 290 mbsf in Site 1090, the downhole increase of Ca⁺² and Sr⁺² and the accompanying decrease of Mg⁺² are less pronounced than at Site 1088 (Fig. F22); below the layer, however, the values of these cations conform to the trends extrapolated from Site 1088. In fact, the Mg/Ca values throughout the section at Site 1090 are nearly identical to those observed at Site 1088. These similarities suggest that similar processes govern the sources and sinks of these cations. They also suggest that, after the sources and sinks were restricted by the chert layer at Site 1090, the downward diffusion of seawater acted to smooth out the profiles.

Figure F23 shows a partial comparison of the major cation chemistries at Sites 1088 and 1090. The plot of Mg⁺² vs. Ca⁺² shows that the trends are reasonably linear suggesting conservative behavior over the sampled intervals (i.e., the major processes responsible for the gradients in Ca⁺² and Mg⁺² are occurring at depths deeper in the sediment column). Both Sites 1088 and 1090 also show similar Mg/Ca -1.1 and similar Mg/Ca values with depth, suggesting that the same diagenetic processes are operating at both sites. At least three major processes have been recognized to be important in influencing Ca⁺² and Mg⁺² concentrations in interstitial waters: (1) seawater alteration of basalts that typically produces a Mg/Ca -0.5 (McDuff, 1981), (2) diagenetic alteration of silicic basement or volcanic ash with a Mg/Ca -2 (Baker, 1986), and (3) dissolution of calcite and recrystallization as dolomite, which, in the case of increasing Ca⁺² and decreasing Mg⁺², would produce a Mg/Ca -1 (Shipboard Scientific Party, 1988a); although this occurs only in very high C_org sediments found in continental margin settings. From the shipboard data alone, it is not possible to determine uniquely the combination of these (or other) processes controlling Ca⁺² and Mg⁺² concentrations at Sites 1088 and 1090; the interstitial water Sr⁺² concentrations, however, may offer some insight.

The lower panels in Figure F23 show the Sr⁺² vs. Ca⁺² and Sr⁺² vs. Mg⁺² plots for Sites 1088 and 1090. It is possible, of course, that Sr⁺² is governed by processes independent of those that dictate Mg⁺² and Ca⁺². But the Mg⁺² vs. Ca⁺², Sr⁺² vs. Ca⁺², and Sr⁺² vs. Mg⁺² trends for Site 1090 are all relatively linear and continuous suggesting that, at least at this site, the concentration profiles of all three elements are related. At first glance it appears that Sr⁺² diagenesis is quite different at the two sites, but at least some of the differences between the sites likely were caused by diffusional smoothing that has occurred at Site 1090 as a result of the chert layer at 290 mbsf. (It is important to note that the Sr⁺² gradient is several times greater than either the Ca⁺² or Mg⁺² gradients.) Downhole Sr⁺² increases almost always result from dissolution of biogenic carbonates and recrystallization of authigenic carbonates (Gieskes, 1983; Baker, 1986). The absolute value of the plateau in Sr⁺² probably reflects celestite solubility (Baker, 1986; Baker et al., 1982), but there is no evidence for a decrease in Sr⁺², which might be expected from basalt interactions (Gieskes et al., 1986). Basement at Site 1088 is at ~1300 mbsf, and it is impossible to draw any firm conclusions about basement interaction. At Site 1090, basement is at ~800 mbsf. Thus, the 400-m section should show some evidence of basement influence on interstitial Sr⁺² concentrations if this were an important process.

Therefore, the Mg/Ca slope and the Sr⁺² behavior suggest that carbonate diagenesis at depths greater than 400 mbsf is the primary factor responsible for the observed gradients in Ca⁺², Mg⁺², and Sr⁺² at Site 1090. Still, it will be necessary to model these data with interstitial water ¹⁸O (and possibly also Sr isotopes) to sort out the relative importance of various processes affecting Ca⁺², Mg⁺², and Sr⁺² concentrations. At the very least, it is clear that the gradients in the upper 290 m of the section at Site 1090 were much greater prior to formation of the chert layer at 290 mbsf.

The K⁺ and Li⁺ profiles also suggest a deep sink and source, respectively, below the chert layer. Presumably, interstitial waters are exchanging K⁺ and Li⁺ with basement or basalts deeper in the sediments below the diffusion barrier, whereas diffusional smoothing occurs above the layer. Silicate increases quite gradually to values near opal saturation above the chert layer; it then decreases below the chert layer for reasons not entirely clear, although the decrease could be related to the observation of zeolite at the base of the section (see "Lithostratigraphy").

The redox chemistry of Site 1090 can be characterized generally as suboxic, with sulfate reduction occurring at very low rates, perhaps only in narrow horizons or microenvironments, as evidenced by the relatively small decreases in sulfate with depth. A small, 1- to 2-mM, decrease in sulfate together with small peaks in alkalinity and ammonium at 30 mbsf suggest that sulfate reduction may be somewhat more important near this interval. A small peak in dissolved Fe⁺² of ~8 µM at 40-50 mbsf may be the result of dissolved Fe⁺² diffusing away from organic-rich bands that had previously concentrated iron by localized reducing conditions. It may also be related to a peak in magnetic susceptibility (see "Physical Properties"). More detailed solid phase analyses in this section are required to identify the sources of this dissolved iron. Downhole, Mn⁺² begins increasing at ~58 mbsf, with a broad peak of ~75 µM between 150 and 200 mbsf. The jump in Mn⁺² from 9 to 33 µM between 58 and 66 mbsf may be related to a 10- to 15-cm-thick tephra layer observed just below this level at ~67 mbsf in Section 177-1090B-8H-4 (see "Lithostratigraphy"). The broad peak in Mn⁺² from ~150 to 200 mbsf suggests that relic, solid phase oxidized Mn has survived burial and is actively being reduced. Intact manganese nodules were observed at shallower depths (Cores 177-1090B-8H, 177-1090C-8H, 177-1090D-7H, and 177-1090E-7H; see "Lithostratigraphy" and the "Core Descriptions" contents list for core images), indicating that Mn-oxides do survive burial to significant depths at this site. A second dissolved Fe⁺² peak of ~10 µM is observed at ~235 mbsf. This peak is less well defined, and interpretation is hampered by having only one sample below this depth and the chert layer. However, if the explanation for the broad Mn⁺² peak is correct, the Fe⁺² peak at ~235 mbsf likely represents the standard diagenetic sequence for reduction of potential electron acceptors for organic-matter oxidation in marine sediments. Thus, Mn-oxides are reduced before Fe-oxides by virtue of the higher free energy available for oxidation of organic carbon by Mn-oxides vs. Fe-oxides (Froelich et al., 1979):

O₂ NO₃^- Mn⁺⁴ (as MnO₂) Fe⁺³ (as FeOOH) SO₄^-2 HCO₃^-.

Taken together, these data suggest that the downhole gradient in sulfate may be the remnant of diffusional gradients that were in place before the formation of the chert layer, and may not be the result of sulfate reduction in the upper 250 m of the section. It is likely that some sulfate reduction is occurring or has occurred in these sediments, but it has never become pervasive enough to clear the sedimentary column of reactive Mn- and Fe-oxides. Thus, the gradients in all the major cations and anions, even the biologically active sulfate ion, appear to reflect processes occurring deep in the sedimentary column.

Solid Phase Analysis

The shipboard solid phase analysis at Site 1090 consisted of measurements of inorganic carbon, total carbon (TC), total nitrogen (TN), and total sulfur (TS) (for methods see "Geochemistry" in the "Explanatory Notes" chapter). The results of Hole 1090B are presented in Table T13 and Figure F24. Calcium carbonate (CaCO₃) contents in Hole 1090B range from 0.6 to 90.7 wt%, with an average value of 39.3 wt%. In the Pliocene-Pleistocene sequence above ~65 mbsf (see "Lithostratigraphy") CaCO₃ shows relatively stable and high values averaging ~76 wt%, whereas, below 65 mbsf, CaCO₃ shows high-amplitude short-term fluctuations. These high-amplitude fluctuations in CaCO₃ content correspond to the alternations of light nannofossil ooze and nannofossil-poor terrigenous sediment. Long-term fluctuations in CaCO₃ content, which are shown smoothed with a 7-point running mean in Figure F24, might also be linked to changes in biogenic opal concentrations (see "Lithostratigraphy). These data suggest that the sedimentary sequence in the late Eocene is characterized by low CaCO₃ and high opal content.

Total organic carbon (TOC) contents vary between 0 and 0.61 wt%, with an average value of 0.19 wt%. Most TOC concentrations, measured as the difference between TC and carbonate carbon, are below 0.5 wt%. Many samples have TOC concentrations below the detection limits of the technique, with the difference between TC and carbonate carbon yielding negative values. TN contents are generally low (0.04-0.12 wt%). TS values are also generally low (nearly zero) from 0 to 100 mbsf and from 310 mbsf to the bottom of Hole 1090B. However, slightly higher sulfur contents are observed from 180 to 310 mbsf, corresponding to a late Eocene-Oligocene opal-rich interval (see "Lithostratigraphy"). These data may suggest that the environment of deposition tended toward anoxia because of increased productivity and/or restricted deep-water ventilation in this region during the late Eocene and Oligo-cene. TOC/TN values vary between 0.4 and 10.7, indicating a predominance of marine organic material. Relatively high TOC/TN values are observed in the lower section below 290 mbsf. Pyrolysis analyses were not made because of the organic-carbon-poor nature of the sediments.