Twenty-seven interstitial-water samples were collected at Site 897 from cores ranging from 26 to 636 mbsf. Samples were collected at approximately 10-m intervals in the top 50 m and in the 50 m above basement (when core recovery permitted). Samples were collected at 30-m intervals between 50 and 580 mbsf. Results from the shipboard interstitial-water analyses are presented in Table 14.
Interstitial-water samples spanned lithostratigraphic Unit I through Subunit IIIA. The most significant changes in interstitial-water chemistry occurred at the boundary between Unit I and Subunit IIA. These changes coincide with a major increase in sedimentation rate that occurred at about the Miocene/Pliocene boundary.
The Pleistocene and Pliocene sediments are dominated by turbidite sequences relatively rich in organic carbon. In contrast, the Miocene through Eocene sediments are depleted of reactive carbon (see "Organic Geochemistry" section, this chapter).
The concentration of sulfate decreases from bottom-water concentration (approximately 29 mM) to near zero by 100 mbsf, thereby driving the diffusion of bottom-water sulfate into the sediment (Fig. 58A). Sulfate concentrations remain low until 300 mbsf, where the concentration of sulfate increases to a maximum at 500 mbsf. The depletion of sulfate, in the sequence between 100 and 300 mbsf, is indicative of the presence of reactive organic carbon. The increase in concentration of sulfate below 300 mbsf indicates upward diffusion of sulfate from a deep reservoir into the Pleistocene and Pliocene sequences. This deep reservoir contains sulfate that has been preserved as a result of the relatively slow accumulation rates of the Miocene to late Oligocene sequence (see "Lithostratigraphy" section, this chapter). A gradient is seen between the deep sulfate maximum (~490 mbsf) and the near-basement sediments, which indicates greater demand for oxidant in the Late Cretaceous age sediments (albeit not large enough to deplete the sulfate reservoir) compared to that for the Eocene sediments.
The alkalinity profile reflects variations in the rate of organic carbon degradation through sulfate reduction (Fig. 58B). Alkalinity is normally produced at a rate that is charge equivalent to the rate of sulfate reduction (Gieskes, 1974, 1983). At this site, the alkalinity profile in part mirrors the sulfate profile, but the increase in alkalinity does not reach levels that one would predict from the decrease in sulfate concentration. Alkalinity reaches values greater than 15 mM in the top 150 m of the sediment column and decreases to about 10 mM by 300 m. Alkalinity may have been reduced by means of precipitation of a carbonate phase in the upper 150 mbsf. Below 350 mbsf, alkalinity decreases to a minimum of about 1 mM near basement.
The production of ammonia in sediments results from anoxic carbon degradation; thus, the ammonia profile at this site mirrors the sulfate profile (Fig. 58C). The highest concentrations of ammonia were observed in the Pleistocene and Pliocene sediments of Unit I. Ammonia concentrations decrease from a maximum greater than 4500 M at 250 mbsf to near constant concentrations of about 200 M by 400 mbsf. In the Oligocene to Eocene sediments, the relatively low and constant ammonia concentrations confirm the premise that reactive organic carbon is absent.
Dissolved iron concentrations are elevated with respect to the normal range of seawater values in the organic-rich Pleistocene and Pliocene sediment sequence (0-300 mbsf, Fig. 58D). The concentration of iron decreases below 270 mbsf to values of less than 10 M and remains low through the Miocene to Eocene sequence. The elevated concentrations can be attributed to the release of iron as a result of its reduction by sulfide to the soluble Fe2+ state.
In contrast to those of iron, concentrations of manganese remain relatively low (<6 M) in the Pleistocene and Pliocene sediments (above 300 mbsf, Fig. 59A). However, concentrations increase to values greater than 12 M (maximum 38 M) in the deep Miocene to Eocene sequence (below 300 mbsf), indicating a zone of manganese release. The low concentrations of manganese above 300 mbsf probably are the result of precipitation of an authigenic carbonate phase. Release of manganese from the older sediments might be the result of manganese reduction, but the apparent absence of reactive carbon (see "Organic Geochemistry" section, this chapter) would argue against such a mechanism. The dissolution of a manganese-bearing carbonate phase or phases may be the mechanism of release.
Concentrations of calcium show depletion (30%-40%) with respect to present bottom-water concentrations in the Pleistocene/Pliocene sequence (to about 280 mbsf), but become enriched (> 100%) below 280 mbsf (Fig. 59B). This depletion in calcium is coincident with the zone of maximum alkalinity and probably reflects precipitation of a carbonate phase. The enrichment (26-28 mM) between 280 and 400 mbsf probably reflects a region of silicate mineral alteration and/or carbonate dissolution. The profile below 400 mbsf exhibits only minor changes in the gradient, indicating minimal dissolution or precipitation of carbonate.
Concentrations of magnesium show some depletion with respect to bottom water throughout the sediment column having the greatest depletion (~30%) coincident with the zone of calcium depletion (Fig. 59C). Magnesium concentrations increase slightly from 300 to 545 mbsf, similar to calcium concentrations.
Concentrations of strontium generally increase with depth from a minimum of about 80 M near the sediment surface to a maximum of about 580 M near basement (Fig. 59D). The strontium profile is concave up between the surface and 300 mbsf, indicating some removal in the Pleistocene sediments, with a zone of release in the late Miocene sequence. The concentration maximum at 550 mbsf indicates a second significant zone of strontium release to the interstitial waters. The release of strontium probably is associated with recrystalization of carbonate phases (Gieskes, 1974).
In general, concentrations of potassium decrease through the sediment column, with some narrow zones of release and removal that produce fluctuations in the profile (Fig. 60A). Concentrations of potassium decrease from a maximum of 11 mM near the seafloor to values slightly higher than 2 mM below about 550 mbsf. Unfortunately, the potassium concentration may depend on the temperature at the time of squeezing (Waterman, 1973). Possible artifacts in potassium concentrations as a result of this problem could not be verified because of failure of the in-situ WSTP sampler at this site.
Concentrations of silica show three maxima between 50 and 100 mbsf, between 170 and 200 mbsf, and between 320 and 400 mbsf, all of which indicate alternating zones of silica release and removal (Fig. 60B). Silica concentrations are elevated (up to 800 M) with respect to bottom water through much of the Pleistocene and Pliocene sequences. Silica has been removed from the interstitial water at the boundary between Units I and IIA (~290 mbsf), a sequence dominated by claystones. Silica concentrations show release through the Miocene to the late Oligocene sequence, resulting in a maximum that approaches 1200 M. Below 450 mbsf, silica has been removed from the interstitial water, producing a negative gradient that extends to 580 mbsf.
Concentrations of chloride show anomalous enrichment with respect to seawater from 50 to 100 mbsf, but then return to values near those of bottom water below (Table 14). This chloride peak represents a greater than 20% enrichment, which is greater than one usually associates with changes in paleosalinity. The absence of a comparable increase in concentrations of sodium suggests either contamination or an analytical problem. The chloride values determined by the Dionix chromatograph were found to be less reliable than the titration method. The variation in chloride concentrations below 100 mbsf typically is less than 5% with respect to present-day bottom water and may be associated with mineral hydration. The marked increase in chloride below 600 mbsf probably results from water removal through hydration of basement rock.
The concentration of sodium varies with that of chloride below 100 mbsf, with the exception of a slight negative gradient between 100 and 500 mbsf (Table 14). At 530 mbsf, concentration of sodium is anomalously low, but shows slight enrichment below, similar to that of chloride.