The main objectives for measuring key geochemical parameters of the pore waters at Site 1254 are (1) to determine the chemistry of pore fluid profiles from décollement whole rounds to compare with profiles measured at Site 1040 and to evaluate possible lateral heterogeneity; (2) to determine the composition of fluids and gases along the décollement and evaluate any possible changes through time for hydrologic modeling; (3) to use selected elements, element ratios, and isotopic compositions in the fluids from the décollement and upper fault zone to constrain dehydration reactions at the updip limit of the seismogenic zone; and (4) to determine key depth intervals in which to place the long-term geochemical observatories. These data are important because the escape of fluids to the surface from depth contributes hydrocarbons for gas hydrate formation, affects seawater chemistry for selected elements, supports a deep biosphere, and is intimately linked to deformation, faulting, and the evolution of the décollement. Furthermore, loss of fluid-soluble elements from the shallow slab not only records reactions and processes within the seismogenic zone but also plays a major role in the supply of residual volatiles deeper in the subduction zone and changes the composition of the slab delivered to the depths of magmatism beneath volcanic arcs.
A total of twenty 35- to 45-cm whole rounds were sampled at Site 1254 for pore fluid geochemistry. One to two whole rounds were taken per core from 153.7 to 217.9 mbsf and from 305.7 to 366.4 mbsf; pore water recovery ranged from 3 to 20 mL above 218 mbsf and from 2.5 to 21.0 mL below 305 mbsf. Pore waters were analyzed for Ca, Mg, K, Na, B, Ba, Fe, Mn, Sr, Si, NH4, and SO4 concentrations, as well as for salinity by methods described in "Inorganic Geochemistry" in the "Explanatory Notes" chapter. Pore water geochemistry data are available in Tables T11 and T12. Because of low pore water recovery at Site 1254, samples were not analyzed for alkalinity and pH.
Lithium, Ca, K, and Na were analyzed aboard ship in real time on the ICP-AES between 305 and 366 mbsf to facilitate identification of the zone of maximum fluid flow within the décollement. Lithium and Ca concentrations at Site 1040 (Leg 170) increased abruptly at the top of the décollement between 330 and 354 mbsf, whereas K concentrations decreased. These constituents showed reverse gradients within the transition from the décollement to the underthrust sediments as observed during Leg 170. The objective of real-time chemical monitoring for these constituents was to quickly identify the depth at which these concentration gradients occur to ensure that the underthrust sediments were not penetrated too deeply during coring so that the long-term observatories will sample décollement fluid and not a mixture of décollement and underthrust fluids. The real-time chemical analyses were very successful.
Salinity is below seawater value in all samples analyzed at Site 1254 (Fig. F41) and remains fairly constant above 197 mbsf; however, there is a sharp decrease from 29 at 210 mbsf to 25 (29% lower than seawater salinity) at 218 mbsf. The decrease in salinity is accompanied by a sharp decrease in Cl, K, and Na concentrations (Fig. F41) and by more gradual increases in Ca and Li concentrations (Figs. F42A, F43B). There is also a sharp decrease in salinity from 31 at 330 mbsf to 25 at 345 mbsf, which is accompanied by an abrupt increase in Li concentrations and a marked decrease in K concentrations similar to those observed at 218 mbsf. A sharp increase in salinity from 27 to 32 (only 9% lower than seawater salinity) occurs between 354.5 and 366.4 mbsf and reflects a transition from sediments of Subunit P1B (wedge sediments) (refer to "Lithostratigraphy") to Subunit U1A (underthrust sediments). The salinity in the uppermost underthrust sediments is similar to the uppermost salinity profile at the reference Site 1039 (Leg 170).
The Cl concentration depth profile (Fig. F41B) divides the sediments cored at Site 1254 into two distinct sections. One is a lower than seawater Cl concentration zone (7%-27% lower) in the deformed wedge sediments with superimposed concentration minima in the décollement and at 218 mbsf and less distinct minima from gas hydrate dissociation. The abrupt decrease in Cl concentration between 197.2 and 218.9 mbsf from 516 to 406 mM is accompanied by sharp increases in Ca and Li concentrations and decreases in Mg and K concentrations. It lies ~20 m below the corresponding Cl concentration minimum observed at Site 1040. Within the décollement, the Cl concentration is ~23% more dilute than seawater value and is also accompanied by abrupt increases in Ca and Li concentrations; however, at the 218 mbsf minimum it is 27% more dilute. The higher dilution at 218 mbsf may be an artifact of gas hydrate dissociation, which also dilutes the other components analyzed by the same proportion. Furthermore, if the Cl minima were due to gas hydrate dissociation alone, then the Na/Cl ratio would not be affected; however, the Cl minima at 218 mbsf and within the décollement are accompanied by Na/Cl minima (Table T11). The second distinct section comprises the underthrust sediments (Subunit U1A), which are typified by near-seawater Cl concentrations (~1% lower than seawater concentration).
Sodium concentrations are below seawater concentration (between 7% and 32% lower) in Subunit P1B and approach it (4% lower) in the transition between the décollement and underthrust sediments of Subunit U1A (Fig. F41C). There is a sharp decrease in Na concentrations from 435.91 to 328.54 mM between 197.2 and 217.9 mbsf, which is within the same depth interval that both Li and Ca increase and K and Mg concentrations decrease. Sodium concentrations show a similar decrease between 330.1 and 354.5 mbsf from 427.39 to 389.15 mM, respectively. Both of these depth intervals where Na concentrations drop abruptly are accompanied by a decrease in Cl concentrations and salinity. However, in situ production of such a low-salinity fluid from clay minerals or opal-A dehydration at the very low geothermal gradient at this site is implausible. Simple dilution by gas hydrate dissociation would tend to dilute the Ca concentrations, but rather an increase in Ca concentration is observed (Fig. F42A). There are no known diagenetic reactions that involve sodium uptake at the very low temperatures encountered at this site, except for uptake by clay minerals into the interlayer exchange sites or by zeolite formation. The clay uptake happens almost instantaneously upon exposure to seawater, and no zeolite-rich sediments were observed at this site.
Potassium concentrations are similar to seawater concentration at the top of the cored interval at Site 1254 and steadily decrease with depth from 9.28 mM at 154 mbsf to 4.54 mM (43% of seawater concentration) at 218 mbsf. This decrease in K concentrations was also observed at Site 1040 but occurred at ~200 mbsf at ~20 m above the K minimum at this site. There is another, albeit smaller, decrease in K concentrations between 330 and 350 mbsf from 5.40 to 4.51 mM, respectively. This is the same interval where increases in Ca and Li concentrations and decreases in Mg concentrations are observed. Similar decreases in K concentrations were observed at Site 1040; however, the minimum is shifted ~6 m above that at Site 1040. Potassium concentrations increase sharply from 354.5 to 366.4 mbsf from a value of 4.51 to 12.74 mM (22% higher than seawater concentration), which reflects the transition from the décollement sediments to the underthrust sediments (Subunit U1A). The higher K concentrations at the top of the cored section may be, in part, due to the higher NH4 concentrations at the same depths (Fig. F44B). The NH4 produced may displace K from the clay ion-exchange sites, producing NH4 clays.
Calcium, Mg, Mg/Ca, and Sr concentration depth profiles (Fig. F42) divide the sediment column at Site 1254 into three distinct sections: two are in the deformed wedge Subunit P1B, and the third is below the transition between sediments of the décollement to sediments of the underthrust section (Subunit U1A). Calcium concentrations increase from 5.69 to 18.78 mM (78% higher than seawater concentration) between 161.4 and 218 mbsf, whereas Mg concentrations decrease from 24.02 to 17.68 mM within the same interval. These changes are reflected by the decrease from a value of ~4 (seawater value is 5.4) to a value of ~1 in the Mg/Ca depth profile for Site 1254. This increase in Ca concentrations and decrease in Mg concentrations was observed at Site 1040; however, the peak at Site 1254 occurs ~20 m below the peak at Site 1040. The increase in Ca concentrations observed at ~218 mbsf at Site 1254 coincides with a reverse fault identified in the same depth interval (see "Structural Geology"); however, this geochemical boundary between 161.4 and 218 mbsf appears to be lithologically indistinct (see "Lithostratigraphy"). At 343 mbsf, Ca concentrations increase sharply from 22.64 to 27 mM (more than twice seawater concentration) at 354.5 mbsf. The increase in Ca concentrations is accompanied by a decrease in Mg concentrations from 20.64 to 16.22 mM within the same interval. This region in the sediment column was interpreted as the top of the décollement at Site 1040, and at Site 1254 it may coincide with a brecciated sandy layer that is moderately indurated (Fig. F4). The Mg/Ca ratio remains fairly constant from 306 to 354.5 mbsf and ranges between 0.92 and 0.76. At such low values of Mg/Ca, calcite diagenesis, but not dolomite diagenesis, occurs in the presence of sufficient alkalinity as seen at Site 1040. Also, Ca concentrations are relatively high and Mg concentrations are relatively low within this section. Calcium concentrations decrease sharply at 355 mbsf to 5.39 mM (~50% seawater concentration), whereas Mg concentrations increase to 42.66 mM, representing a transition from Subunit P1B to the underthrust sediments.
Strontium concentrations, unlike Ca concentrations, remain fairly constant above 218 mbsf and range between 79.50 µM at 153.7 mbsf and 91.4 µM at 218 mbsf. Below 306 mbsf, Sr concentrations range between 77.11 µM and 168.18 µM. The highest Sr concentration observed at Site 1254 is at 361 mbsf, which is ~10 m above the Sr peak at Site 1040 and may reflect the transition between the décollement sediments and the underthrust sediments.
The SO4 depletion zone (i.e., SO4 = 0) extends to the depths cored at Site 1254 (Fig. F44A) as well as at Site 1040; therefore, any SO4 analyzed in a sample would indicate drill water contamination. Fortunately, SO4 concentrations were below the detection limit of the ion chromatograph for all but two samples analyzed at Site 1254. The two slightly contaminated samples occur at 161.4 and 360.9 mbsf and represent 3.7% and 6.6% drill water (seawater) contamination, respectively. These two data points have not yet been corrected for drilling-induced contamination; this does not affect the interpretation of the data obtained at Site 1254. The sediments of Subunit P1B are anaerobic and remain within the zero-sulfate zone; therefore, CH4 concentrations are high and provide an environment conducive to gas hydrate formation. The SO4 depletion zone extends below the décollement into the underthrust sediments of Subunit U1A, despite the fact that at Site 1039 the minimum SO4 concentration encountered at ~24 mbsf is 13.2 mM. This suggests that upon underthrusting, the supply of SO4 from seawater by diffusion ceased and the underthrust sulfate-reducing bacteria utilized the remaining SO4 at the top of the section, thereby depleting it to zero concentration, as also observed at Site 1040.
The NH4 concentration depth profile at Site 1254 is similar to the NH4 profile observed at Site 1040 (Fig. F44B). Intense NH4 production occurs at zero-sulfate concentrations; therefore, NH4 concentrations are very high in both Subunit P1B and U1A. Ammonium concentrations generally decrease with depth from 10,796 µM at 153 mbsf to 1,452 µM at 366 mbsf. The very high concentrations in the deformed wedge sediments also suggest that the sediment accumulation rates are similarly high at both Sites 1254 and 1040 and that the pore water system is practically closed. Therefore, the NH4 produced after exchanging with clay minerals remains buried in the section.
Barium concentrations at Sites 1254 and 1040 remain fairly constant between 154 mbsf and 354.5 mbsf (Fig. F43A). Barium concentrations between these depths at Site 1254 range between 8.55 and 14.82 µM. Barium concentrations increase abruptly from 11.17 µM at 354.5 mbsf to 158 µM at 366.4 mbsf and reflect a change in lithology from décollement sediments to underthrust sediments.
Iron and Mn concentration depth profiles are plotted in Figure F44. Iron concentrations above 218 mbsf show little variation with depth except for a small increase in concentration from 2.36 µM at 181.3 mbsf to 5.92 µM at 190.6 mbsf. Below this peak, Fe concentrations range between 2.37 and 2.62 µM. Below 306 mbsf, the Fe profile is more variable, containing two distinct peaks at 306 and 366.4 mbsf of 18.51 and 18.28 µM concentrations, respectively. Manganese concentrations above 218 mbsf are variable and range between 2.68 and 21.73 µM. From 306 to 354.5 mbsf, Mn concentrations remain fairly constant and range between 3.29 and 3.62 µM. Below 354.5 mbsf, there is a sharp increase in Mn concentrations to 10.23 µM at 366.4 mbsf, reflecting the transition from décollement sediments to the underthrust sediments. Because SO4 concentrations are zero within the depths cored at Site 1254, both Fe(OH)3 and MnO2 have been reduced by bacteria. However, above 354.5 mbsf, Fe and Mn concentrations do not show clear trends and may reflect a lithologic variations or a complex combination of redox processes and reprecipitation of mineral phases.
The Li concentration profiles at Site 1254 and Site 1040 are plotted in Figure F43B. Lithium concentrations are approximately twice seawater concentration between 153.7 and 190.6 mbsf and range from 50.78 to 57.36 µM. Lithium concentrations increase abruptly from 57.36 µM at 191 mbsf to 229.14 µM (8.5 times seawater concentration) at 218 mbsf. The increase in Li concentrations is accompanied by an increase in Ca concentrations and by decreases in Mg and K concentrations within the same depth interval. This increase in Li was also observed at Site 1040; however, the peak at Site 1254 is ~20 m below the Li spike at Site 1040. There is a small peak in Li concentrations between 306 and 320 mbsf from 125.1 to 133.82 µM, respectively, which may coincide with a brecciated sandy layer within the sedimentary sequence (Fig. F4). There is another abrupt increase in Li concentrations between 345 and 354.5 mbsf from 171.47 to 239.18 µM, respectively. This sharp increase in Li concentrations is coincident with a maximum in Ca concentrations and a minimum in Mg and K concentrations, as well as a sandy more brecciated sedimentary interval (Fig. F4). From this depth to the bottom of the cored section (366.4 mbsf), Li concentrations rapidly decrease to 25.96 µM, which is ~4% lower than seawater concentration. This decrease in Li clearly reflects the transition from sediments of Subunit P1B to the underthrust sediments.
Silica concentrations generally decrease with depth from 274.3 µM at the top of the cored interval to 71.4 µM at 218 mbsf. They remain fairly constant between 306 and 354.5 mbsf and range from 51.4 to 81.4 µM (Fig. F43C), suggesting more mature silicate diagenesis with all the opal-A transformed to the less soluble Si phases. From 351.5 to 366.4 mbsf, Si concentrations increase sharply from 61.40 µM to concentrations of 549.50 µM, which is due to the dissolution of diatoms and radiolarians that are abundant in the underthrust sediments but rare in the décollement sediments.
Boron concentrations at Site 1254 are similar to those observed at Site 1040 (Fig. F43D) and remain fairly constant between 154 and 354.5 mbsf, with concentrations ranging between 76.68 and 137.76 µM. Between 354.5 and 366.4 mbsf, B concentrations increase sharply from 76.68 to 582.67 µM, respectively. The increase in B concentrations within this interval clearly reflects the transition from deformed wedge sediments to sediments of the underthrust section.
The geochemical excursions in salinity, K, Ca, Mg, Mg/Ca, Li, and CH4 between 210 and 218 mbsf (Subunit P1B) occur within a highly fractured interval interpreted as a fault zone, whereas the excursions at ~351 mbsf may coincide with a brecciated sandy interval that is moderately indurated (see "Structural Geology" and "Lithostratigraphy"). Small peaks in Ca, Li, and CH4 concentrations occur at 330 mbsf and seem to be associated with another sandy brecciated interval that is less indurated than the interval below (Fig. F4). These data suggest that the deep-sourced fluid has migrated along conduits and permeated the lower half of the deformed wedge. Assuming that the geothermal gradient is ~20-30 K/km (see the "Leg 205 Summary" chapter), then the source region must occur at >4 km depth because the minimum temperature required for thermogenic gas formation is 80°-90°C. The minima in K concentrations at 218 and 351 mbsf further suggest that the deformed sediments have been permeated by a fluid from an elevated temperature source of 80°-120°C where the illitization reaction is effective, which consumes K. Also, the K-depletion signature of this fluid provides an approximate upper limit to the temperature at the source of ~<150°C. Above this approximate temperature, fluid-rock reactions leach K from the rocks. Lithium, like K, is partitioned into solids at low to moderate temperatures. At higher temperatures, >100°C but <250°C, Li is released into the fluid phase (Chan and Kastner, 2000). The precise threshold temperatures of partitioning into the solid or fluid phases are as yet unknown. Clay (and other silicate) dissolution or alteration releases B into the fluid phase; however, clay, especially illite, formation consumes B and may be responsible for the low B concentrations within the deformed sediments. The deeply sourced fluid, however, is not enriched in dissolved silica. Gas hydrate dissociation is caused by drilling, and core recovery causes other lesser salinity dilution excursions in the pore water concentrations. These excursions, however, are not accompanied by Ca and Li maxima.
Calcium, Li, and CH4 concentration maxima, as well as K and Mg concentration minima, occur at ~218 mbsf at Site 1254. Similar increases in Ca, Li, and CH4 concentrations, as well as marked decreases in Mg and K concentrations were observed at Site 1040; however, the geochemical discontinuity at Site 1040 occurred between ~180 and 200 mbsf. Therefore, the geochemical boundary at Site 1254 is ~20 m below the same boundary observed during Leg 170. At Site 1040, a fault zone was interpreted to occur at the geochemical discontinuity at ~180 mbsf because of the presence of deformation structures (Kimura, Silver, Blum, et al., 1997). A similar concentration of deformation structures was observed between 180 and 218 mbsf at Site 1254 (see "Structural Geology"). The shift in the geochemical anomalies at Site 1254 to 20 m below those observed at Site 1040 suggests that the movement on the fault is reverse.
The geochemical boundary at ~218 mbsf separates pore fluids typical of clay-rich sediments above from those permeated by a fluid from an elevated temperature source and seems to be lithologically indistinct (see "Lithostratigraphy"). This suggests that the main sediment source has maintained uniform chemical composition throughout the time interval represented by Subunit P1B (see "Lithostratigraphy") and that the changes in pore water chemistry result from fluid infiltration into the lower half of the deformed sediment section. The chemical changes observed in the boundary conduit at ~218 mbsf are similar to those observed near the bottom of the décollement zone that are probably associated with the sandy flow conduit. Except for the biogeochemical components, the pore fluid concentration depth profiles of the underthrust section are similar to those at Site 1039. The concentrations themselves can be either lower or higher than those at Site 1039, reflecting the changes in solubilities and dissolution rates of the major sediment components under the new pressure regime. The higher NH4 concentrations than those at Site 1039 and the zero-sulfate concentrations at the interface between décollement and the underthrust sediments reflect the termination of diffusional communication with bottom seawater. Sulfate reduction, thus, reaches completion in the uppermost few meters of the underthrust hemipelagic section, resulting in somewhat elevated CH4 concentrations within the zero-sulfate depth interval.