METHODS

Both pore fluid and solid samples were measured at Sites 1039/1253 and 1040/1254. Most of the sediment samples were from pore water "squeeze cakes," and intervals with ash layers were not sampled. Since dissolved sulfate is only depleted in lithologic Unit U1 pore fluids at Sites 1040/1254, sampling was at a higher frequency in this unit. Pore fluid samples were analyzed for Ba concentrations by inductively coupled plasma–mass spectrometry (ICP-MS), as well as by inductively coupled plasma–optical emission spectrometry (ICP-OES). All of the sediment samples were analyzed for Ba concentrations by ICP-MS.

In order to correlate sediment samples across holes (from Sites 1039/1253 to 1040/1254), percent compaction estimates for the underthrust sediment based on weight percentage CaCO₃ measured shipboard during Leg 170 were employed (Kimura, Silver, Blum, et al., 1997). Assuming that the entire section is underthrust at Costa Rica, the CaCO₃ depth profiles from both sites indicate that there was a 36% reduction in thickness of lithologic Unit U1 and Subunit U2A (hemipelagic clayey section), a 62% reduction of thickness of Subunits U2B and U3A (hemipelagic clayey and transition sections), and a 24% reduction of thickness in Subunits U3B and U3C (pelagic calcareous section) (Kimura, Silver, Blum, et al., 1997). These estimates are similar to those presented in Saffer et al. (2000) based on logged bulk density from Leg 170, where the authors estimate a 33% reduction in thickness of the hemipelagic section by Site 1040 and ~20% reduction in thickness of the lower pelagic calcareous section. To correlate the sample depths at Site 1039/1253 to their appropriate depths at Sites 1040/1254, the following equations were used:

d₁₀₄₀ = (d₁₀₃₉ x t) + 371 and (1)

d_1040(n) = (d₁₀₃₉ x t) + d_{1040(n – 1)}, (2)

where

d₁₀₄₀ = corresponding depth at Site 1040,

d₁₀₃₉ = depth at Site 1039,

t = percent reduction in thickness of each lithologic unit at Sites 1040/1254 determined from the CaCO₃ concentration depth profiles at Sites 1039 and 1040,

371 = depth of the base of the décollement at Site 1040,

d₁₀₃₉ = vertical distance between adjacent samples at Site 1039, and

d_1040(n–1) = cross-correlated depth of the previous sample.

Equation 1 is only used for the shallowest sample at Site 1039, and equation 2 is used for all subsequent samples. Compared with the estimates of Saffer et al. (2000), the cross-hole compaction corrected depths in lithologic Units U1 and U2 at Site 1040 would be nearly identical and depths in Unit U3 would be shifted ~15 m deeper as a result of greater compaction in the transition section based on the CaCO₃ concentration-depth profiles.

A total of 35 pore fluid samples were analyzed for Ba concentrations on a ThermoQuest/Finnigan Element 2 ICP-MS (see Table T1). All sample and standard preparations were made in a clean laboratory. Samples and standards were diluted with double-distilled deionized water containing 0.4-N HNO₃ and spiked with a 1.0-ppb In internal standard. All standard and calibration solutions were prepared from certified stock solutions. Two unknowns and 20 pore fluid samples were analyzed for each batch of analyses. A 1.0-ppb drift standard was analyzed after every four samples. Blanks were interspersed at random during each batch of analyses, and calibration was achieved with five standard solutions ranging from 0.1 to 5.0 ppb. Pore fluid samples were diluted 100 times at Sites 1039/1253, 1000 times in lithologic Unit P1 at Site 1254, 10,000 times in Unit U1 at Sites 1254/1040, and 100 times in Units U2 and U3 at Site 1040 to achieve a final concentration of ~1.0 ppb, thus matching the concentration of the drift standard and internal standard. Prior to analysis, the ICP-MS was tuned using the ¹¹⁵In internal standard to maximize the intensity of the elements to be analyzed, and mass calibrations were performed after every 20 samples. Instrumental drift was corrected online by normalization of the intensity of the analyte with the intensity of the ¹¹⁵In standard. A second drift correction was applied offline using repeated analyses of the 1.0-ppb Ba drift standard made by dilution of the primary certified stock solution. The accuracy and precision of multiple analyses were monitored by repeated analyses of the two unknowns and the 1.0-ppb drift standard. The average accuracy was <1%, and the average precision was <0.65%.

A total of 65 pore fluid samples were analyzed for Ba concentrations by standard addition on a Perkin Elmer Optima 3000 ICP-OES. The average accuracy and precision of the ICP-OES analyses determined by multiple analyses of drift and calibration standards was <4% and <7%, respectively. The results of the ICP-MS and ICP-OES determinations agree fairly well (Fig. F3) and are within the quoted precision of the ICP-OES analyses. Pore water sulfate concentrations at Sites 1039 and 1040 were measured shipboard during Leg 170 by ion chromatography (IC) using a Dionex DX-100. The reproducibility of the analyses, expressed as 1 standard deviations of means of multiple determinations of International Association of Physical Sciences of the Ocean (IAPSO) standard seawater was ~1% (Kimura, Silver, Blum, et al., 1997). Sulfate concentrations at Sites 1253 and 1254 were measured shipboard during Leg 205 by IC using a Dionex DX-120. The reproducibility of the analyses, expressed as percent precision from multiple determinations of IAPSO standard seawater was <2%.

Sediment samples were dried in an oven for 24 hr at 60°C then ground into a fine powder. The samples were weighed before and after drying to determine the weight of water evaporated. Since the sediment samples were taken from pore water squeeze cakes, Ba concentrations of the pore water splits were used to compute the amount of Ba precipitated as salts during the drying process. Ten milligrams of the powdered sample was weighed and placed in a tightly capped, acid-cleaned polytetrafluoroethylene (PTFE) beaker. All sediment digestions were performed in a clean room. Two U.S. Geological Survey (USGS) certified rock standards were digested with each batch of sediments, typically four sediment samples. MAG-1 (marine mud) and SCO-1 (Cody Shale) were chosen as standards because they are marine sediments and have certified values for Ba. The digestion procedure consisted of five steps. Each step included adding an Optima-grade reagent, placing the tightly capped beaker in an ultrasonicator for 60 min, and evaporating to dryness in a PTFE evaporating unit under a heat lamp. The samples were digested by adding 4-N HNO₃ to convert the carbonate to CO₂, adding 30% hydrogen peroxide to oxidize the organic matter, adding a 2:1 mixture of concentrated HF and HNO₃ to digest the sample, and twice treating the samples with concentrated nitric acid. The samples were then diluted 2000-fold by weight with a preprepared 2.5% nitric acid solution in double-distilled deionized water.

All of the sediment samples were analyzed on a ThermoQuest/Finnigan Element 2 ICP-MS. The initial dilution was diluted 500 times by weight to provide a final dilution of 1 x 10⁶ times. All solutions were spiked with a 1.0-ppb In internal standard. The method of analysis by ICP-MS was identical to that performed on the pore fluid samples outlined above. Digestion precision and accuracy were monitored by repeated digestion of reference USGS certified rock standards MAG-1 and SCO-1. Similar final dilutions were made for the sediment standards as the sediment samples. The average percent accuracy and precision of multiple determinations of MAG-1 were <1% and 0.5%, respectively. The average percent accuracy and precision of multiple determinations of SCO-1 were <1.3% and <1%, respectively. A few shipboard X-ray fluorescence (XRF) data analyzed during Leg 170 have been included in Figure F4 and in the summary section of this report. The samples were measured on an ARL 8420 XRF with reported percent accuracy and precision of 2%–3% (Kimura, Silver, Blum, et al., 1997).