The shipboard geochemistry program focused on inorganic constituents in interstitial water (Table T8). Six samples of the basal sediment from Hole 1201D were also analyzed (Table T9). Routine headspace gas samples were collected, but none contained methane above the detection limit of 3 ppmv.
Fifty-three samples of interstitial (pore) water were recovered by squeezing whole rounds from 34 cores, 1 from Hole 1201A, 9 from Hole 1201B, and 24 from Hole 1201D. The deepest pore water was sampled from just above basement, at 510 mbsf in Hole 1201D. The three holes constitute a single section because they were drilled within ~60 m of one another and their total depths overlap.
Changes in pore water chemistry at Site 1201 reflect lithologic changes. The sedimentary section can be divided into three parts: (1) red to tan silty clay from the seafloor to ~55 mbsf, (2) gray to green volcaniclastic breccia and sandstone and siltstone turbidites from 55 to 507 mbsf, and (3) reddish brown claystone from 507 to 510 mbsf. The volcaniclastic materials are increasingly altered to clays and zeolites with depth (see "Lithostratigraphy").
The composition of the interstitial water at Site 1201 is quite unusual for deep-sea sediments and reflects the extreme alteration of the volcaniclastic materials and sandstones that comprise the lower 250 m of the section. The most unusual feature is the large increase in pH, Ca, and chlorinity. Calcium increases to 270 mmol/kg, 27 times the concentration in seawater, by leaching from the volcaniclastic material. Chlorinity increases to 645 mmol/kg, 20% higher than the seawater value, by loss of water of hydration to the altered basalts. The gain in Ca is balanced by the loss of nearly all of the Mg and K from seawater, but mainly by a 70% decrease in Na, to 140 mmol/kg. As a result, the pore water from Site 1201 at 190 mbsf and below contains far more Ca than Na (Table T9). In seawater, Na comprises about 78% of the cations and chloride about 90% of the anions on a milliequivalent basis (Fig. F61). Subordinate cations are Mg (18%), Ca (3%), and K (1%), whereas subordinate anions are sulfate (9%) and bicarbonate ([alkalinity] 1%). In the average pore water from Site 1201, by contrast, Ca comprises 38% of the cations and, in a Ca-rich sample from 499.7 mbsf, 67% of the cations (Fig. F61). The proportion of chloride increases to 92% in the average pore water from Site 1201 and to 95% in the Ca-rich pore water. The large increase in Ca inevitably saturates the solution with gypsum, causing sulfate to decrease from 28 to 15 mmol/kg as gypsum precipitates. Alkalinity falls from the seawater value of 2.4 to <1 meq/kg as it is consumed by formation of authigenic minerals. These reactions likewise cause the rise in pH to 10.0 from the seawater value of ~8.1, reflecting extreme diagenesis or perhaps zeolite facies metamorphism accompanied by extensive authigenesis.
Between 83 and 147 mbsf in Hole 1201D, the sandstone and siltstone turbidites and breccia were too indurated to yield pore water. Fortunately, this sampling hiatus does not interrupt the smooth depth profiles exhibited by most of the dissolved constituents (Figs. F62, F63, F64, F65). Exceptions are sulfate, Na, and K. Unfortunately, the major inflection point in cross-plots of Ca vs. Na, Cl, Mg, K, sulfate, and alkalinity (not shown) occurs within this interval, connecting generally linear segments above and below. The linear segments are consistent with diffusion as the major process within these depth intervals, whereas the inflection points imply the presence of a chemical reaction zone. Below the hiatus, the decreases in Mg, K, and sulfate with increasing Ca become less steep, whereas the decrease in Na and increase in chlorinity become steeper. (In keeping with these trends, the slope of Na vs. chlorinity is essentially constant with depth.) The slope of alkalinity vs. Ca actually changes from negative (decreasing alkalinity with increasing Ca) within the depth interval of 0-73 mbsf sampled in Hole 1201B to positive (increasing alkalinity) within the interval 147-510 mbsf sampled in Hole 1201D. (The uppermost pore water sample from 83 mbsf in Hole 1201D falls at an intermediate point that is not on either line.) These observations suggest that the most intense uptake of Mg, K, and sulfate is occurring within the upper reaction zone centered at ~110 ± 30 mbsf, whereas the most intense hydration, uptake of Na, and release of Ca is occurring within the basal sediments and possibly within basement. Reactions include alteration of feldspar and glass to smectite and Na-Ca-K zeolites and precipitation of gypsum (see "Lithostratigraphy").
The reality of an upper reaction zone is confirmed by large, broad maxima in dissolved Si, Sr, and Mn and a smaller maximum in fluoride in this depth interval (Fig. F64). These elements are all known to be quite reactive in deep-sea sediments. The Si maximum results from dissolution of silica and/or silicate phases. The Mn maximum implies suboxic to reducing conditions at that depth, and a Sr maximum is typically associated with recrystallization of calcium carbonate. The Li and B profiles resemble those for Na and K, except that the concentrations of Li and B increase slightly with depth between ~150 and 500 mbsf. Aluminum increases steadily with depth between 83 and ~300 mbsf and then levels out.
Comparisons among pore water compositions can also be made on the basis of equilibrium mineral assemblages calculated from thermodynamic principles. The computer code PHREEQC simplifies these calculations (Parkhurst and Appelo, 1999). Pore water precipitates minerals when it is oversaturated with the mineral's chemical constituents and dissolves minerals when it is undersaturated with those constituents. Pore water in equilibrium with minerals neither precipitates nor dissolves those phases. The stability of minerals in aqueous solutions is quantified with the saturation index (SI), which is >0 for minerals that precipitate, <0 for minerals that dissolve, and 0 for equilibrium. Increases in the relative values of negative SIs are also significant because they point to closer approach of mineral precipitation. The uncertainty of SI values is variable because some aqueous systems are well characterized, whereas others are not. For example, calcite SIs are accurate to about ±0.1 SI units, but the uncertainty for zeolites is ±1-2 SI units. The calculations were made for temperature = 3°C and pH = 8.5.
Saturation indices for selected minerals reveal three general trends (Table T10). One is an increase in sepiolite SIs with depth, sepiolite representing a Mg sink. Note, however, that smectite is the true magnesian clay likely to precipitate under alkaline conditions, such as those in the deeper pore water from Site 1201. Saturation indices for this more complex clay are not calculated by PHREEQC. Second, zeolite SIs increase with depth, particularly those of laumontite and leonhardite, which are not present at Site 1201 but were used to model zeolite behavior because they are among the most silica-poor and water-rich zeolites. These zeolites can accommodate Na in exchange for Ca. The sodic zeolite analcime could also accommodate Na. Precipitation of hydrous minerals such as clays and zeolites causes increased chloride concentrations in residual water (Fig. F63A). A third trend is increased SIs with depth for anhydrite and gypsum. Gypsum SIs increase substantially across the indurated zone, implying increased amounts of gypsum precipitation within and below this interval. Gypsum takes up Ca, sulfate, and H2O.
Diagenesis of volcanic sediments has been documented in several western Pacific basins (e.g., Gieskes and Lawrence, 1981; Egeberg et al., 1990; Torres et al., 1995). These findings are based on pore water recovered from DSDP and ODP drill sites. Calcium enrichment accompanied by Mg depletion with increased depth below the seafloor is the characteristic chemical pattern. The same pattern is present at Site 1201.
Fresh volcanic ash, glass, and shards from the Mariana arc contain ~2-13 wt% CaO and 0-10 wt% MgO (Arculus et al., 1995). These materials provide a rich source of Ca in the sediments. An alternative Ca source is carbonate in the sediments. Carbonate dissolution would cause an increase in alkalinity, as is seen in Hole 1201D, between ~200 and 500 mbsf. Smectite was identified in XRD analyses of volcanic materials at Site 1201 (see "Lithostratigraphy"). This clay is the likely sink for Mg. Increases in Ca require Na decreases that are twice as large to maintain charge balance. Sodium can enter clays, zeolites, albitized plagioclase, and other secondary phases. Diagenetic changes such as these can account for the major patterns of pore water chemistry in the altered volcanic sediments at Site 1201.
Six samples of the basal sediment from Hole 1201D were analyzed because of their unusual appearance (Table T9). The basal layer immediately above basement (two samples) is a dark chocolate-brown claystone. This layer grades upward into grayish brown uniform to laminated claystone (two samples) and laminated dark gray claystone (one sample) with pink interlayers (uppermost sample). In general, the uppermost two samples are similar to the lithogenous pelagic clays of Chester and Aston (1976) (Table T9). They have similar amounts of Si, Al, Fe, Mg, Na, and P but only one-half as much Ti and K, and three to five times as much Mn and Ca. The other four samples are progressively more dissimilar. Compared with the lithogenous pelagic clays, the lowermost two have only 80% of the Mg, two-thirds of the Si, Al, and Na, and less than one-half the Ti and K. They are enriched 3-fold in Fe, 6-fold in Ca and P, and up to 12-fold in Mn. Whereas all six samples from Site 1201 are relatively depleted in Ti and K and enriched in Mn and Ca, for the other elements they could be modeled as mixtures of lithogenous pelagic clay, goethite (identified by XRD) (see "Lithostratigraphy"), Mn oxide/hydroxide, and apatite.