Sixty-three samples were collected during the high-resolution interstitial water (IW) program at Site 1115. Three samples were collected from the only core of Hole 1115A, 29 from Hole 1115B, and 31 from Hole 1115C. Whole-round sections were taken down to Core 180-1115C-54R, the deepest core recovered at this site. Only a few whole rounds in the deeper portion of Hole 1115C yielded insufficient IW to preclude determination of the full suite of constituents that had been analyzed in samples from Site 1109. Thus, a nearly complete data set was generated that enables an excellent comparison to be made with Site 1109.
The IW was analyzed for salinity, pH, alkalinity, major cations (Na+, K+, Ca2+, and Mg2+) and anions (Cl- and SO42-), Li+, Sr2+, SiO2, and NH4+. Results of shipboard inorganic chemical analyses are presented in Table T7. The pH values obtained during the alkalinity determination appear to be reliable for the same reasons as described in the Site 1109 chapter (see "Inorganic Geochemistry" in the "Site 1109" chapter and "Inorganic Geochemistry" in the "Explanatory Notes" chapter). Profiles of many inorganic constituents (Figs. F44, F45) display a series of depth intervals in which there are substantially different concentration ranges. Large changes in the concentrations of many of the dissolved constituents are found below 400 mbsf.
In the upper 400 mbsf of the sediments, pH remains in a narrow range (7.5-8.0). A transitional zone (400-600 mbsf) leads to an increase to near pH 9.0. These values are then maintained to the bottom of Hole 1115C. The titration alkalinity profile displays relatively large variations, primarily in the upper 400 mbsf (Fig. F44A). Alkalinity reaches a maximum of 6.7 mM (Table T7), which is only approximately half that observed in sediments from Site 1109 (see "Inorganic Geochemistry" in the "Site 1109" chapter). Dissolved SO42- is strongly depleted downhole and essentially absent from the pore water below a depth of 200 mbsf, other than a very few small excursions at depth. The NH4+ profile (Fig. F44A) displays an initial increase that parallels the alkalinity rise in the upper 66 mbsf. Subsequent downhole variations appear decoupled from alkalinity. The NH4+ maximum at the bottom of Hole 1115C has no alkalinity counterpart.
Salinity varies in the range of 30-38 (Table T7). The minimum is found below the zone of total SO42- depletion, whereas the maximum is at the bottom of Hole 1115C. Salinity does not covary strongly with either Na+ or Cl-. Additionally, these two IW constituents do not covary throughout the entire sediment column (Fig. F44B). Both constituents display a slightly negative concentration gradient down to 250 mbsf, although the dissolved Cl- profile displays more variations in this range than does that of Na+. Concentrations of both constituents increase in the lower 150 m of Hole 1115C. Their profiles are decoupled in the range of 250-600 mbsf.
The K+ profile (Fig. F44B) shows slight depletion relative to seawater between 100 and 300 mbsf of the sedimentary column, followed by a 150- to 200-m-thick transition zone in which dissolved K+ is extensively removed from the pore fluids. Below 600 mbsf, the profile is relatively invariant with concentrations near 1 mM, or ~10% of the seawater value.
Dissolved Li+ (Fig. F44B) exhibits a slightly more variable profile than many other inorganic constituents, although concentrations are generally low (20-55 然) and do not exceed approximately twice the seawater value. The first significant increase in dissolved Li+ below 200 mbsf and its maximum (55 M) near 300 mbsf coincides with a large increase in dissolved SiO2 and the alkalinity maximum, respectively. Other features of the Li+ profile do not appear to correlate strongly with any single IW constituent.
The Ca2+ profile reveals a 50% depletion relative to seawater by the depth where SO42- becomes fully removed, followed by a return to near seawater values by 350 mbsf (Fig. F44C). Below 400 mbsf, Ca2+ displays a sharp increase to a broad maximum of 120-130 mM below 600 mbsf. No significant variation is observed further downhole. The Mg2+ profile is slightly more variable in the upper 400 mbsf, with a number of smooth fluctuations found within the predominantly nannofossil-rich sediments of lithostratigraphic Units I-III (see "Lithostratigraphy"). Below this depth, however, dissolved Ca2+ and Mg2+ exhibit a near antithetical correlation.
The depth ranges in which the Sr2+ profile displays major features (Fig. F44C) appear to correspond more closely to those of Mg2+ than those of Ca+. Variations in the Sr2+ profile, however, are more marked than those of either Ca2+ or Mg2+. Two Sr2+ maxima exist. The first is relatively narrow and is found at 66 mbsf in lithostratigraphic Unit II, whereas the second is broad and extends between 486 and 542 mbsf in sediments of lithostratigraphic Units VI and VII (see "Lithostratigraphy"). This deep-seated broad Sr2+ maximum is situated slightly below where the Ca2+ and Mg2+ profiles cross over, unlike at Site 1109 where these features coincided.
As commonly observed in deep-sea sediments, the dissolved SiO2 profile (Fig. F45) displays greater complexity than those of other pore-water constituents. This is manifested in various excursions in concentration downhole. These, nonetheless, fall mostly within three broadly defined concentration ranges. In the upper 200 mbsf, dissolved SiO2 increases sharply immediately below the mudline then remains between 400 and 500 然. Higher concentrations and larger variations are present between 210 and 390 mbsf, an interval punctuated by two local maxima reaching near 900 然. From 390 to 420 mbsf, dissolved SiO2 concentrations decrease sharply and remain well below 200 然 in the deeper part of the hole (>500 mbsf).
Changes in the chemistry of interstitial fluids at Site 1115 result from many of the same diagenetic reactions that were observed to mediate the pore-water composition at Site 1109. However, significant differences in the IW constituent profiles exist between the two sites and attest to the presence of sedimentary reactions here that do not occur at Site 1109. Coring at Site 1115 also penetrated a forearc sediment sequence (lithostratigraphic Units X-XII; see "Lithostratigraphy") that was not reached at Site 1109. Hence, the lowermost 240 m of the profiles presented here provides insight into processes occurring below the late Miocene unconformity at the top of the forearc sequence (between lithostratigraphic Units IX and X; see "Lithostratigraphy").
The concentrations of IW constituents in the upper 300 mbsf result primarily from the oxidation of organic matter mediated by microbial activity and the concomitant early diagenesis of biogenic carbonates. Within the latter are included the dissolution of aragonite, recrystallization of calcite, dolomite precipitation, and to a very minor extent the formation of ankerite (see "Site 1115 Thin Sections" and "Lithostratigraphy"). Other important processes in the sediment column are the alteration of volcanic matter, formation of authigenic clay and zeolite minerals, and precipitation of pyrite, as well as transformations of pre-existing detrital minerals.
A general reaction was previously given describing the bacterially mediated oxidation of organic matter (e.g., Claypool and Kaplan, 1974; see "Inorganic Geochemistry" in the "Site 1109" chapter). This process leads to strong depletion of SO42- within the first 200 mbsf and increases in both alkalinity and NH4+ concentrations (Fig. F44A). A slight decrease in pH was observed over the first 50 mbsf (Table T7), as expected from the degradation of organic matter. Yet the pH remains higher than typical Pacific seawater at depths of ~1000 m (Millero and Sohn, 1992). The relatively invariant pH in the upper section of the sediments is also considerably higher than values observed in sediments undergoing extensive rates of degradation of organic matter (e.g., Eberli et al., 1997). Therefore, either a lower rate of microbial oxidation of organic matter is present here, or sufficient detrital matter containing Fe exists to buffer the pH to higher values (Ristvet, 1978; Mackenzie et al., 1981). The increased alkalinity in the shallower sediments (~100 mbsf) promotes dissolution and recrystallization of biogenic carbonates (e.g., Morse and Mackenzie, 1990).
Sediments in the upper 100 mbsf contain significant amounts of biogenic aragonite in addition to calcite (see "Site 1115 Thin Sections" and "Lithostratigraphy"). The dissolution of this metastable phase and the recrystallization of calcite are reflected in the dissolved Ca2+, Mg2+, and Sr2+ profiles and in changes in the Ca/Mg ratio (Fig. F44C). The release of dissolved Sr2+ (e.g., Baker, 1986) produces a well-defined maximum of 338 然 at 66 mbsf (Fig. F44C). The disappearance of aragonite below this depth (until well into Hole 1115C) is manifested in a return to lower dissolved Sr2+ further downhole. Recrystallization of calcite is present between ~20 and 150 mbsf as evidenced by close to 50% depletion of Ca2+ relative to seawater at the bottom of this interval (Table T7). The decreasing concentration of dissolved Mg2+ between ~100 and 240 mbsf, a subminimum in alkalinity between 160 and 210 mbsf, and a slightly increasing Ca/Mg ratio, as well as the identification of dolomite in samples subjected to XRD analysis (see "Site 1115 Thin Sections" and "Lithostratigraphy"), are all consistent with dolomite precipitation over this interval. Because dolomite was identified by XRD down to ~300 mbsf (see "Lithostratigraphy"), dolomitization continues, although probably on a more limited basis at least down to this depth. The Ca/Mg ratio variations in this depth interval support this inference. It is plausible that the lower alkalinity maxima observed here relative to Site 1109 reflect consumption of HCO3- during dolomitization, although the greatest extent of dolomite precipitation is found near 160 mbsf, consistent with where the alkalinity subminimum is observed. Ankerite, an Fe-Mg-Mn-enriched calcium carbonate mineral, was tentatively identified by XRD as a minor mineral constituent in samples from 240 mbsf, providing further evidence of calcium carbonate precipitation reactions.
Calcium carbonate mineral reactions are also important in sediments deep within Hole 1115C. These are evidenced by a broad Sr2+ maximum between 480 and 540 mbsf. Here aragonite provides a source of this element in the form of gastropod shells from the inner neritic environment of lithostratigraphic Units VI through the lagoonal setting of lithostratigraphic Unit VIII (see "Lithostratigraphy"). The dissolved Sr2+ maximum is found over a broader interval and has a much lower concentration than in the corresponding units at Site 1109 (see Fig. F70C and "Inorganic Geochemistry," both in the "Site 1109" chapter). Diffusion has likely been more extensive at this site because sediments bounding the dissolved Sr2+ maximum at Site 1109 were much more clay rich, less porous, and underlain by dolerite, which acts as an effective barrier to fluid flow. Other calcium carbonate reactions deep in Hole 1115C likely include calcite cementation of the sediments, as evidenced by a substantial increase in pH below 500 mbsf and the presence of calcareous sandstone and limestone in lithostratigraphic Unit VIII (see "Lithostratigraphic Unit VIII"). The latter coincides with the onset of pH values near 9.0. Maintenance of a high pH downhole is also consistent with the presence of calcite cements.
Increases in the Ca/Mg ratio between 250 and 400 mbsf, where dissolved Mg2+ concentrations actually increase slightly, probably reflect a release of Ca2+ during weathering of volcaniclastic sands and ash layers. This process exerts greater control over the concentrations of these constituents than dolomitization reactions, which are present only to a minimal extent below ~250 mbsf. Indeed, increases in dissolved Ca2+ over this interval are accompanied by substantial increases in dissolved Sr2+ and Li+, as well as very localized SiO2 maxima (Figs. F44B, F44C, F45), all of which are also likely derived from alteration of volcanic matter. Visual core descriptions and results of XRD analyses support this inference. Abundance and thickness of volcaniclastic sand layers are greatest between 250 and 400 mbsf, as is the presence of plagioclase as a major constituent of the sediments (see Fig. F3). The lower dissolved Li+ concentrations between 500 and 600 mbsf may simply reflect a diffusive profile with sources of this constituent located above (e.g., 300-400 mbsf) and below (e.g., 750 mbsf) this zone. Alternatively, a sink for dissolved Li+ may exist near 600 mbsf at the boundary between lithostratigraphic Units X and XI (see "Lithostratigraphic Unit X" and "Lithostratigraphic Unit XI"), possibly clay minerals and zeolites (e.g., De Carlo, 1992). The submaximum in dissolved Li+ deep within forearc sediments of lithostratigraphic Unit XII (see "Lithostratigraphic Unit XII") also indicates a source of this element deep downhole. This requirement is satisfied possibly by the abundant plagioclase observed in this portion of the hole.
Examination of the dissolved K+ profile (Fig. F44B) reveals a steadily decreasing concentration below 100 mbsf. Because the illite observed in this depth range is believed to be detrital, it is unlikely an effective sink for K+. Perhaps the profile simply reflects diffusion with respect to the much steeper removal of this constituent below ~300 mbsf. Froelich et al. (1991) attributed linear negative K+ gradients in sediments to an absence of sedimentary reactions and control by low temperature uptake of this constituent into basement. Although this is clearly not the case here, the nearly linear gradient in the upper 300 mbsf reflects the absence of reactions involving K+ in that section of the sediment column. The absence or very low abundance of K-feldspar within the volcaniclastic sands and ashes of lithostratigraphic Units II and III, which could act as a source, is also consistent with the observed profile. This contrasts with Site 1109, where a significant source of dissolved K+ existed near 400 mbsf (see Fig. F70B and "Inorganic Geochemistry," both in the "Site 1109" chapter). Formation of chlorite or smectite does not occur in the upper half of the sediment column, as confirmed by results of XRD analyses (see "Lithostratigraphy") and cannot contribute to the decrease in dissolved Mg2+ described above. It is only below at least 450 mbsf that Mg2+ begins to be removed substantially from the pore water. The XRD data indicate the appearance of clay minerals that require Mg2+ to form (e.g., chlorite and smectite) below 450 mbsf. However, based only on qualitative XRD results, these minerals are present more often as minor rather than as major constituents. It is also possible that zeolites, which become quite abundant near 600 mbsf, constitute another sink for dissolved Mg2+ (see "Site 1115 Thin Sections" and "Lithostratigraphy").
As observed at Site 1109, large changes are present in the pore-water constituent profiles in sediments that were deposited in neritic environments relative to those deposited at bathyal depths. At Site 1109, these changes occurred primarily below 550 mbsf (see "Lithostratigraphic Unit VII" in "Lithostratigraphy" in the "Site 1109" chapter) and were marked, for example, by a large increase in dissolved Ca2+. A similar trend is found below ~450 mbsf in lithostratigraphic Unit V of Hole 1115C.
At this site, the shapes of the Ca2+, Mg2+, and Ca/Mg ratio profiles between 400 and 600 mbsf are not influenced extensively by calcium carbonate mineral reactions. Rather, they are typical of sediment/pore-water systems in which substantial alteration of basement rock is found. These profiles are consistent with observations previously made at many other ODP sites where alteration of volcanic matter or Layer II basalt was invoked (e.g., Gieskes and Lawrence, 1981; Gieskes, 1983; Torres et al., 1995). Alteration of volcanic matter may also account indirectly for increases in dissolved Cl- (Fig. F44A) observed further downhole. Because Cl- is not normally involved in diagenetic reactions other than halite diagenesis, its concentration increase must occur through hydration of ash as well as the formation of hydrous clay and zeolite minerals, which simply increase the pore-water salinity (Martin et al., 1995). Both types of minerals are observed deep downhole at this site and the covariance of the Na+ and Cl- profiles in lithostratigraphic Unit XII (see "Lithostratigraphic Unit XII") is consistent with the occurrence of such a process. The smaller increases in dissolved Cl- between 300 and 450 mbsf may also derive from this process, although the near invariance or slight depletion of dissolved Na+ require its removal, ostensibly through formation of authigenic feldspars or other phases rich in Na (e.g., Gieskes and Lawrence, 1981).
Profiles of dissolved Li+ and SiO2 (Figs. F44B, F45) display variations that suggest the composition of the pore fluids is mediated by alteration of volcanic matter. Early dissolution of biogenic silica (e.g., diatoms and radiolaria) may also contribute to the observed pore-water SiO2 concentration in the uppermost few meters of Holes 1115A and 1115B. This source of dissolved SiO2, however, becomes progressively less important downhole as the alteration of volcanic glass contained in abundant ash layers in lithostratigraphic Units I and II potentially contributes most of the dissolved SiO2 (see Figs. F29, F45, and "Lithostratigraphy"). The relative invariance of dissolved Li+ in the upper 220 mbsf, an interval in which both volcanic ash and volcaniclastic sands are present, is consistent with either only limited alteration of the most soluble glass in the ash (which contains low abundances of trace elements), or with the volcanic matter below 220 mbsf having a solid phase composition distinct from that above. Because relatively wide fluctuations in dissolved SiO2 concentrations below 220 mbsf correspond roughly to the broad dissolved Li+ maximum here, it is also reasonable to assume that the source of elevated dissolved SiO2 in sediments of lithostratigraphic Units III and IV is a combination of terrigenous and volcaniclastic matter. Magnetic susceptibility exhibits a sharp drop near 220 mbsf (see "Magnetic Susceptibility"). Although further work is needed to elucidate the reason for these changes, compositional differences in sediments above and below this depth appear quite likely.
It is known that rates and stratigraphic occurrences of silica diagenesis in sediments are controlled by a combination of burial time, pressure, temperature, and sediment composition (Torres et al., 1995, and references therein). Reactions leading to the formation of low-permeability horizons during the alteration of volcaniclastic material are also important. Thus, discerning which of these processes controls silica diagenesis is not always simple. For example, a depth relationship was reported by De Carlo (1992) for silica conversions in pelagic to hemipelagic sediments from Sites 762 and 763 on the Exmouth Plateau, Indian Ocean. The relationship existed in spite of the fact that otherwise equivalent lithostratigraphic units were widely separated in depth and volcanic matter was largely absent from these two sites. Thus, burial depth and presumably temperature controlled the observed reaction sequence at these sites. A somewhat analogous yet potentially different situation exists at Sites 1109 and 1115, where equivalent lithostratigraphic units are present at different depths in the sediment column, and where dissolved SiO2 profiles are remarkably similar in terms of depths at which major transitions occur (see "Inorganic Geochemistry" and "Lithostratigraphy" both in the "Site 1109" chapter). The sharp decrease in dissolved SiO2 from >800 然 to <200 然 in the deeper portions of Holes 1109C and 1115C is found near 500 mbsf at both sites. This final transition was observed to be between 500 and 600 mbsf at Sites 762 and 763 (De Carlo, 1992). The Shipboard Scientific Party (1976) reported an equivalent transition near 600 mbsf at Site 317 on the Manihiki Plateau in the Pacific Ocean. Abundant examples exist, however, where dissolved SiO2 profiles and their transitions are found at much shallower depths, such as in areas where basaltic basement is present at sufficiently shallow depths to exert a stronger influence on the overlying sediments (e.g., Gieskes and Lawrence, 1981; Gieskes, 1983; Eberli et al., 1997). At sites containing either volcanic matter disseminated in the sediments or where sediments overlie doleritic sills, a silicification front was identified that is associated with reactions involving increases in Ca2+ and decreases in Mg2+ and K+ in interstitial waters. More elevated temperatures are also known to enhance such reactions (e.g., Sites 474, 482; Gieskes, 1983).
The drop in dissolved SiO2 to concentrations of 100-200 然 is found near the boundary between lithostratigraphic Units VI and VII, the marine sand/brackish silt transition (see "Lithostratigraphic Unit VI" and "Lithostratigraphic Unit VII"), here and in "Lithostratigraphy" in the "Site 1109" chapter. At Site 1109, this transition corresponds to a sharp decrease in porosity (Fig. F71, "Inorganic Geochemistry," "Lithostratigraphy," and "Physical Properties" in the "Site 1109" chapter). Although the covariance with porosity is less distinct here (Fig. F45) than at Site 1109, it remains probable that the transition to low dissolved SiO2 represents a silicification front at both sites, as invoked by Gieskes (1983) for various Deep Sea Drilling Project sites.
Because Sites 1109 and 1115 have similar temperature gradients (31慢搔m-1 and 28慢搔m-1, respectively) and contain abundant volcanic matter in the sediment column, the observed dissolved SiO2 profiles likely reflect a combination of temperature of burial effects and the influence of the alteration of volcanic matter.
The chemical composition of the IW is mediated by a series of sedimentary diagenesis reactions. These include the alteration of volcanic ash layers, volcaniclastic sand layers, and volcanic minerals dispersed throughout the sediments, calcium carbonate dissolution and recrystallization reactions associated with the microbial oxidation of organic matter, as well as other inorganically mediated calcium carbonate mineral reactions and silicification reactions including authigenic clay, zeolite, and quartz formation. Dolomite formation, which was not observed at Site 1109, is important here. The inferences based upon variations in the chemical composition of the IW are supported here by results of XRD analysis, lithostratigraphic observations, and, to a lesser extent, by changes in the physical properties of the sediment in a manner similar to what was observed at Site 1109.