Pore fluid barium concentrations at Sites 1039/1253 and Sites 1040/1254 are presented in Table T1. At Site 1039, sulfate concentrations are below seawater value (28.9 mM) at 1.45 meters below seafloor (mbsf) and decrease to a minimum of 13 mM at 24 mbsf (Fig. F3C). Sulfate concentrations increase to values seen in surface sediments by 146 mbsf and increase to near-seawater concentration in the basal carbonate section. Ba concentrations are higher than seawater value (~0.15 µM) and variable from 9 to 146 mbsf, which is approximately the base of lithologic Unit U2 (Fig. F3A). The dissolved Ba2+ concentrations in the upper 146 m of Site 1039 range from 0.378 to 4.257 µM (Table T1). Below 146 mbsf, Ba concentrations are nearly constant and slightly above bottom water concentration. Ba concentrations are more variable at Site 1253, at the base of the pelagic carbonate section, due to fluid-rock reactions with the gabbro sill and metalliferous sediments. The Ba concentrations within lithologic Units U1 and U2 at Site 1039 are as much as 26 times seawater value and may be due to Ba mobilization from organic matter as well as Fe-Mn oxides and oxyhydroxides during organic matter diagenesis, as suggested by McManus et al. (1998). The sharp peaks in the Ba profile within these units, as well as in the methane concentration depth profile (Kimura, Silver, Blum, et al., 1997), suggests active lateral fluid advection along coarse-grained ash layers and other more permeable horizons.
The pore fluid within the prism sediments at Sites 1040 and 1254 is totally depleted in dissolved SO42– to the base of the décollement at 371 mbsf (Fig. F3D). The zero-sulfate zone is also observed in the uppermost underthrust section to a depth of 401 mbsf. The base of lithologic Unit U1 is at 423 mbsf at Site 1040, and within the other two units, sulfate concentrations gradually increase with depth to a value of 28 mM at the base of the pelagic section (Fig. F3D). The depletion of sulfate in the uppermost hemipelagic sediments at Site 1040 suggests that upon underthrusting the supply of sulfate from seawater by diffusion ceased, and the sulfate-reducing bacteria within the underthrust sediments utilized the remaining SO42– at the top of the section. Since the underthrust sediment section is the only source of dissolved SO42–, the depth of sulfate depletion will increase arcward. Ba is above seawater value in the prism sediments, and concentrations range from 6.37 to 13.2 µM, indicating some Ba mobilization from barite dissolution. The maximum Ba concentrations within the prism sediments occur within the upper fault zone and the décollement (Table T1), and Ba concentrations are relatively constant at ~6 µM between these two flow conduits. Drilling during Legs 170 and 205 sampled a deeply sourced fluid within these two intervals originating at temperatures as high as ~150°C (Chan and Kastner, 2000; Hensen et al., 2004). Some distillation of Ba at the depth of the deep-sourced fluid may occur.
Most striking are the extremely high Ba concentrations at the base of the décollement in the underthrust sediment section at both Sites 1040 and 1254. The concentrations in lithologic Unit U1 range from 17.26 to 209.57 µM, with the highest concentration occurring at the base of the décollement at 372 mbsf (Fig. F3B). The concentration at 372 mbsf is ~53 times higher than the Ba concentration of the equivalent sample at Site 1039 and ~1400 times higher than the bottom water value. In contrast, the maximum Ba concentration measured within the décollement and upper fault zone at Site 1254 are 9.95 and 13.2 µM, respectively (Table T1). Thus, the extremely high pore fluid Ba2+ concentration immediately below the décollement is ~20 times higher than the concentrations measured within the two fluid flow conduits in the prism sediments. The sharp discontinuity in pore fluid Ba2+ concentrations between the décollement and uppermost underthrust sediments does not support the suggestion by Saffer and Screaton (2003) that an advective or diffusive flux from the subducted hemipelagic sediments contributes to the Ba signal observed in the décollement and the upper fault zone ~130 m above it, indicating that the two fluid flow systems are effectively decoupled. At 401 mbsf, where SO42– concentrations start to increase, Ba concentrations decrease to values similar to those measured at the reference site (Table T1). The extremely high concentrations in the upper 30 m of the hemipelagic section at Site 1040 indicate intense Ba2+ liberation from the mineral barite due to increased BaSO4 solubility coupled to SO42– depletion. As the subducting sediments move arcward the sulfate depletion zone will become thicker, eventually consuming the pore water SO42– in lithologic Units U2 and U3. These units have higher sedimentary Ba concentrations (see discussion below), and a higher proportion of Ba2+ will be liberated from barite. This process may have a profound impact on the amount of Ba subducted to depths of magma generation. The diagenetic release and transport seaward of Ba from BaSO4 in sulfate-depleted pore water arcward of the trench may reduce previous estimates of bulk sedimentary Ba subducting to depths of magma generation based solely on the reference section seaward of the trench.
Bulk sediment samples were analyzed for Ba to determine if the amount of Ba liberated from barite in the shallow forearc significantly impacts the amount of sedimentary Ba delivered deeper into the subduction zone. The bulk sediment Ba concentration depth profiles for Sites 1039/1253 and 1040/1254 are shown in Figure F4. Ba concentrations are the highest at the base of lithologic Unit U2 and in the pelagic carbonate section (Unit U3) where there is less dilution of biogenic Ba by detrital material. Unit U1 and the top of Unit U2 (hemipelagic sediments) have nearly uniform Ba concentrations. The sharp increase in Ba from 2852 to 7408 ppm between 117.53 and 168.45 mbsf (Fig. F4A) is caused by a change in sedimentation rate from ~46 to ~6 m/m.y. (Kimura, Silver, Blum, et al., 1997). Most Ba concentrations measured at Site 1039 range from 1000 to 3000 ppm, suggesting that barite is a significant fraction of the Ba-containing phases offshore Costa Rica (Eagle et al., 2003). The Ba concentration-depth profile in the underthrust sediments at Site 1040 is almost identical to that at Site 1039, except concentrations in lithologic Unit U1 are lower than those measured in Unit U1 at Site 1039 (Fig. F4B; Table T2). Three samples were analyzed for Ba concentrations in the prism sediments at Site 1254. The concentrations are fairly uniform with depth and are 25%–30% of the concentrations measured in the underthrust section, reflecting dilution of biogenic barite by terrigeneous material deposited by debris and gravity flows (Morris, Villinger, Klaus et al., 2003).
Average bulk Ba compositions of each of the lithologic units (Units 1, 2, and 3) at Sites 1039/1253 and 1040/1254 are presented in Table T3. The total sedimentary Ba lost or gained across the trench as well as the percent difference in concentration of each unit is also presented in Table T3. Unit U1 shows an 18% loss in bulk sedimentary Ba from Site 1039 to Site 1040. Unit U2 displays a 10% loss in Ba, whereas Unit U3 exhibits a 10% gain in Ba between the two sites (Table T3). The sampling resolution within Units U2 and U3 is much lower than in Unit U1. Bulk sediment Ba analyses of additional samples within these units are in progress, and results of these measurements will most likely push the percent difference in Units U2 and U3 closer to zero. It is apparent that the percent difference in Unit U1 is almost twice that in Units U2 and U3, indicating that significant Ba distillation due to barite dissolution had occurred and is occurring, effectively changing the bulk Ba composition of Unit U1 as sulfate becomes depleted. As dissolved sulfate becomes further depleted arcward, greater losses of sedimentary barite must exist not only in Unit U1 but in the deeper Units U2 and U3 as well.
Figure F5 shows cross plots of Ba at Site 1039 vs. Ba at Site 1040 for samples that correlate well across sites. The diagonal line has a slope of 1; therefore, if there were no changes from Site 1039 to Site 1040, then all samples would plot along the 1:1 line. Eighty percent of the samples from lithologic Unit U1 plot well above the line, signifying that the bulk Ba concentrations at Site 1039 are higher than those at Site 1040, whereas samples from Units U2 and U3 plot close to the 1:1 line and are both above and below the line. The pore fluid Ba concentration-depth profile at Site 1040/1254 (Fig. F3B) shows the highest dissolved Ba concentrations to occur immediately below the décollement, thus, assuming a homogeneous sediment section, one would expect the shallowest samples at Sites 1039/1253 to plot the furthest left of the 1:1 line, which is not observed (Fig. F5A). However, the uppermost hemipelagic sediment section at Sites 1039/1253 is heterogeneous, reflecting varying amounts of detrital material input (Kimura, Silver, Blum, et al., 1997) and differing rates of biogenic barite accumulation with depth, and thus with time. Therefore, Figure F5 does not show a clear trend of the deepest samples in Unit U1 plotting closer to the 1:1 line and the shallowest samples plotting furthest from the line. Results of the sequential barite extraction (in progress) will furnish the absolute barite concentrations of the sediment samples at Sites 1039/1253 and will confirm that each sample in the uppermost hemipelagic sediment section has varying initial barite concentrations and differing amounts of detrital Ba.