IODP Proceedings Volume contents Search | |||||||||||||||||||||
Expedition reports Research results Supplementary material Drilling maps Expedition bibliography | |||||||||||||||||||||
doi:10.2204/iodp.proc.329.102.2011 BiogeochemistryDocumentation of microbial processes in South Pacific Gyre sediment requires a wide range of biogeochemical analyses. Transport-reaction modeling of dissolved metabolite concentrations will allow us to quantify rates of principal net activities in the sediment (e.g., reductions of O2, NO3–, SO42–, Mn[IV], and Fe[III]) (Jørgensen, 2006). Chemical data (e.g., the position of the chloride maximum from the Last Glacial Maximum) and thermal data will be used to help model this transport (D’Hondt et al., 2004). Concentration data will also allow us to calculate mineral stabilities and, in anoxic sediment, to test thermodynamic models of metabolic competition (Hoehler et al., 2000; Wang et al., 2010). The shipboard geochemistry program during Expedition 329 consisted of an ambitious campaign to characterize the geochemical environment of the subseafloor of the South Pacific Gyre. We undertook key biogeochemical analyses on the ship and collected appropriate samples for critical postcruise biogeochemical studies. As described below, components targeted by shipboard analyses included headspace gases, dissolved oxygen, various species of the fixed nitrogen (nitrate, nitrite, and ammonium) and carbonate systems, chloride and sulfate, phosphate, dissolved silica, and a number of dissolved cations. Analyses of the organic and inorganic carbon contents, as well as of the total nitrogen contents, of the sediments were also conducted. Analytical protocols are summarized below. In many cases, these protocols are based on those of Gieskes et al. (1991), Murray et al. (2000), and the IODP user manuals for new shipboard instrumentation, with modifications as indicated. However, as described in the following subsections, in many cases the instrumentation and protocols used during Expedition 329 supersede those cited above. Interstitial water extractionsSedimentary interstitial waters were collected by squeezing whole-core rounds using Manheim squeezers and by sampling using Rhizon soil moisture samplers (Rhizosphere Research Products, Wageningen, The Netherlands; Gribsholt and Kristensen, 2002). Rhizon samples were taken for nitrate, nitrite, ammonium, and nitrate isotope analyses, while squeezed interstitial water samples were used for all other analyses. At Sites U1370 and U1371, interstitial waters gathered by Rhizon samplers were also acquired for shore-based metals analysis (e.g., Fe and Mn). Rhizon sampling extracts interstitial water from sediment by suction filtering into 10 mL plastic syringes through thin tubes of hydrophilic porous polymer that has a mean pore diameter of 0.1 µm. At Sites U1365–U1367, as described further below, and prior to the squeezing of each whole-round core sample, Rhizon samplers (5 cm long filter) were inserted into one end of the whole-round core and a total volume of 2 to 11 mL of interstitial water was extracted (Fig. F9). Core sections were stored cold (9°C) during Rhizon sampling. At Sites U1368–U1371, Rhizon samples were collected from a core interval adjacent to the interstitial water sampling interval. Before deployment, new and previously used Rhizon samplers were soaked in 18.2 MΩ deionized water for several hours, followed by rinsing with 30 mL of 18.2 MΩ water that was suction-filtered through each Rhizon. After washing, Rhizon samplers were left to dry on filter paper. Care was taken to only use completely dry Rhizon samplers. Blanks consisting of 18.2 MΩ water pulled through new and recycled Rhizon samplers analyzed for nitrate onboard, as described below, were all below detection limit. Interstitial water squeezing forces the water out of the sediment by using a titanium squeezer, modified after the stainless steel squeezer of Manheim and Sayles (1974). Gauge pressures of as much as 20,000 lb force were applied to the squeezer using a Carver laboratory hydraulic press to extract the interstitial water. The interstitial water is passed through a prewashed (with 18.2 MΩ water) Whatman number 1 filter fitted above a titanium screen within the squeezer and filtered through a 0.45 µm polysulfone disposable filter (Whatman Puradisc PES) fitted to the syringe that gathers the fluid sample. Sample handling and short term storageDuring this expedition, 415 whole rounds were squeezed for interstitial water analyses and 468 samples were taken by Rhizon sampling for nitrate analyses, as collectively described elsewhere in this volume. Whole rounds for microbiological studies were also often cut from the same core section. Consideration of different interstitial water extraction methods and coordination with microsensor oxygen measurements and microbiological sampling required us to develop an overall strategy to maximize sample throughput and yet preserve the integrity of the samples and analyses. During Expedition 329, the majority of the interstitial water samples were handled in the following way. First, the cores were cut into 1.5 m (nominal) sections on the catwalk. Second, these sections were delivered to the Hold Deck refrigerator (core storage hold) where they were cut into whole rounds for interstitial water and microbiological analyses, curated, and labeled. Third, a Rhizon sampler was immediately inserted into each interstitial water whole round in the Hold Deck refrigerator (Fig. F9) and 2–11 mL of water was extracted into the Rhizon syringe. This Rhizon-extracted water was used for nitrate analysis on board, and aliquots were preserved for shore-based research. Fourth, when the Rhizon sampling was completed, which took as long as 0.1–12 h, the interstitial water whole rounds were delivered to the Biogeochemistry Laboratory. Samples were stored in a 4°C refrigerator in the Biogeochemistry Laboratory until they could be squeezed. The time samples were stored in the Biogeochemistry Laboratory varied from a few minutes to as long as >24 h in some cases, even with the use of six Manheim squeezers and three Carver presses. At Site U1365, all samples were processed as described above. At Site U1366, some whole rounds were cut in the Hold Deck refrigerator but were not sampled using Rhizon samplers. Instead, those whole rounds were immediately delivered to the Biogeochemistry Laboratory for squeezing. The main goal for these samples was to extract interstitial water for shore-based 14C research, but we also analyzed some of these waters on board for dissolved species (e.g., alkalinity, dissolved inorganic carbon [DIC], Ca, Mg, and Sr). These are referred to as the “whole round stored shorter” (WSS) samples. Starting at Site U1367 and continuing for the rest of the expedition, a small subset of the interstitial water samples was cut on the catwalk and immediately delivered to the Biogeochemistry Laboratory for interstitial water extraction by squeezing. These are termed “catwalk” samples throughout this report. Interstitial water was not extracted from these with Rhizon samplers. Slight contrasts between the WSS samples at Site U1366, the catwalk samples at Site U1367 and afterward, and the interstitial waters extracted from whole rounds processed in the Hold Deck refrigerator are described in the individual site chapters in this volume. Because the time that a core arrives on the rig floor is recorded by the drilling crew and because we kept time records of where each whole round was in the sample and handling process, we are able to reconstruct the amount of time it took a whole round to arrive in the Hold Deck refrigerator, the duration of Rhizon sampling, when the whole round was delivered to the Biogeochemistry Laboratory, its storage time in the laboratory refrigerator at 4°C, and the duration of the squeeze. We plotted these various storage and handling times as a function of depth for each site because the cores arrive on the rig floor and thus start in depth order and because the chemical profiles are plotted against depth. These graphs are shown in Figure F10 for all Expedition 329 sites. Graphs are also shown that document how long it took each Rhizon sample to be extracted for Sites U1369, U1370, and U1371 (Fig. F11). Documenting of the time of day in each step of the sample storage and handling process demonstrates that there were no effects of storage on the nitrate profile or that of many other species. Dissolved species of the carbonate system showed effects in certain situations. These graphs also facilitate discussion as to whether occasional anomalous data are due to storage, spurious instrumental effects, or other factors. These effects, particularly for the carbonate system, will be further quantified by postexpedition thermodynamic calculations. Headspace hydrocarbon gasesRoutine analysis of hydrocarbon gas in sediment cores is part of the standard IODP shipboard monitoring of the cores to ensure that the sediments being drilled do not contain unsafe levels of hydrocarbons. The most common method of hydrocarbon monitoring used during IODP expeditions is analysis of gas samples obtained from either core samples (headspace analysis) or from gas expansion pockets visible through the clear plastic core liners (Vacutainer analysis). Because no gas expansion pockets were formed during Expedition 329, Vacutainer analysis samples were not taken during the expedition. Concentrations of hydrocarbon gases were monitored at intervals of 1 sample per core. For the required safety analysis (Kvenvolden and McDonald, 1986), a 3 cm3 bulk sediment sample from a freshly exposed end of a core section was collected immediately after retrieval on deck using an open-ended plastic syringe, extruded into a 20 mL headspace glass vial, immediately capped with a gray butyl rubber septum, and sealed with an aluminum crimp cap. When consolidated or lithified samples were encountered, chips of material were placed in the vial and sealed. The vial was then heated to 70°C for ~30 min prior to analysis. The standard gas analysis program for safety and pollution prevention purposes was complemented by additional headspace analyses (1 per section) following a slightly different approach (Iversen and Jørgensen, 1985; Hoehler et al., 2000; Shipboard Scientific Party, 2003c; Ertefai et al., 2010) with the intent to better constrain the concentrations of adsorbed and dissolved gases. Compared to the rapid safety-oriented protocol, the latter, more time-consuming alternative is more quantitative and may yield more methane. For samples designated for the refined headspace analysis, a 5 cm3 sediment sample was collected from a freshly exposed end of a core section using an open-ended plastic syringe. The sediment sample was collected by penetrating the sediment surface while the plunger was maintained at the sediment surface to prevent contamination from atmospheric gases or trapped air bubbles. After sampling, the syringe was extruded until 3 cm3 of sample remained and the excess was shaved off with a flat spatula flush with the end of the syringe barrel to provide an accurate estimate of the sediment volume within the syringe. The remaining 3 cm3 sediment sample in the syringe was extruded into a 20 mL vial containing 5 mL of 1N NaOH to inhibit biological activity (Iversen and Jørgensen, 1985). Beginning with Site U1368, an additional 5 mL of 1N NaOH was used to decrease headspace volume and assist in mixing the sediment plug. The vial was immediately capped with a gray butyl rubber septum and an aluminum crimp cap. After vigorous manual shaking for ~2 min, the vials were occasionally shaken to completely disperse sediment into the NaOH solution. Samples were subsequently left to stand for at least 24 h at room temperature prior to gas chromatographic analysis. Gas chromatographic analyses of headspace samples for both safety and refined protocols were performed identically. A 5 mL volume of headspace gas was extracted from the sealed sample vial using a standard gas syringe and directly injected into the gas chromatograph. The headspace gas samples were analyzed using an Agilent/HP 6890 Series II gas chromatograph (GC3) equipped with a 2.4 m × 3.2 mm stainless steel column packed with 100/120 mesh HayeSep R and a flame ionization detector set at 250°C. The gas chromatograph oven temperature was programmed to hold for 0.5 min at 80°C, ramp at 30°C/min to 100°C, ramp at 15°C/min to 110°C, and remain at 110°C for 4.5 min, before ramping at 50°C/min to 150°C with a final holding time of 1.8 min. Helium was used as the carrier gas. The GC3 system determines concentrations of methane (C1), ethane (C2), ethene (C2=), propane (C3), and propylene (C3=). Data were collected using the Hewlett Packard 3365 Chemstation data processing program. For both systems, chromatographic response is calibrated to nine different gas standards with variable quantities of low molecular weight hydrocarbons, N2, O2, CO2, Ar, and He and checked on a daily basis. The gas concentrations for the required safety analyses are expressed as component parts per million by volume (ppmv) relative to the analyzed gas. To the extent that sampling procedures are uniform, the differences in the headspace results reflect differences in the amount of gas remaining in the cores. The internal volumes of 15 representative headspace vials were carefully measured before any sites were occupied during Expedition 329 and were determined to average 25.41 ± 0.18 mL. This volume was taken as a constant in calculations of gas concentrations. The volumetric units were converted to concentration units (millimolar) to facilitate comparisons with dissolved interstitial constituents using the relationship
where,
The precision of analysis for the “B” standard was ≤1% for all gases analyzed with the GC3 and the natural gas analyzer. Dissolved hydrogenDissolved hydrogen was quantified in sediment samples using a reduced gas analyzer (RGA; Trace Analytical, Inc., provided by the University of Rhode Island [USA] Geobiology Laboratory). Subcores were taken with 10 cm3 cut-off syringes either on the catwalk or in the Hold Deck core refrigerator. The 10 cm3 subcores were immediately extruded into 40 mL capacity screw-cap vials and filled with water. Care was taken to exclude any air from the vials before closing. Samples for H2 analysis were taken next to whole-round cores used for interstitial water analysis. A headspace was created in the vial by introducing 500 µL of N2 gas that passed through the RGA (bypass gassiest gas was introduced through the septa using a 500 µL gas-tight syringe with a 27 gauge needle. During the injection, an equivalent amount of water was allowed to escape from the vial through a separate needle. The samples were stored upside down for 10–20 h to allow the H2 to partition into the headspace. Just prior sampling, the vials were centrifuged (2 min at 200× g) to move the sediment away from the septa. Afterward, 500 µL of the headspace was extracted using a needle and gas-tight syringe and injected into the RGA. The instrument was fitted with a 250 µL sample loop. The RGA was calibrated using a 102.4 ppmv H2 gas standard (Scott-Marin, Inc.) using a gas mixer to vary the H2 concentration. Dissolved oxygen measurementsCapped and sealed 1.5 m core sections targeted for oxygen analysis were brought into the temperature-controlled room at near in situ temperature (10°C) within 15 min of arriving on deck. A temperature sensor (PT 1000; see optode setup below) was inserted through a hole in the core liner (4 mm) into the core center at the middle and ends (lengthwise) of the sections. Core temperature was between 7° and 12°C when cores arrived in the cold room. Core sections were left until they were at equilibrium with the cold laboratory ambient temperature (9.5°C) throughout the entire 1.5 m core section to ensure that all oxygen measurements were performed under stable temperature conditions. Temperature variation was always <0.5°C within sections and <1°C within cores. Room temperature was continuously monitored and logged on a DICKSON SP175 temperature data logger and varied between 9° and 11°C during all measuring periods. Whole-round sections from subsequent holes were brought to the cold laboratory after geochemical and microbiological sampling in the ship’s Hold Deck core refrigerator, equilibrated to 9.5°C, and measured. Dissolved oxygen measurements were only done on whole rounds longer than 20 cm and within the lengthwise center in order to avoid edge effects. Oxygen measurements were conducted with two independent microsensor methods: amperometric Clark-type oxygen sensors (microelectrodes) and fiber-optic oxygen microsensors (optodes). The oxygen microelectrodes (Revsbech, 1989; Unisense, Aarhus, Denmark) are made of glass with a ~5 cm long tip that is inserted and fixed within a hypodermic needle (1.1 mm diameter; 50 mm length). Electrode signals were amplified and transformed to millivolts (mV) by two-channel picoammeters (PA 2000; Unisense; provided by the University of Southern California Marine Biogeochemistry Laboratory) and directly recorded on a computer using the Profix software (Unisense). The needle-type optodes (Fischer et al., 2009; PreSens, Regensburg, Germany) are miniaturized chemical-optical sensors based on a silica fiber (140 µm diameter) with the tip mounted into a needle-type housing for robustness. Once inserted, the optical fiber is pushed out into the sediment for the measurement. Temperature was measured in each optode measuring point by inserting a temperature sensor (PT 1000) directly adjacent to the optode. The optodes and temperature sensor were connected to a MICROX TX3 (PreSens; provided by the University of Rhode Island Geobiology Laboratory) single-channel fiber-optic oxygen meter, and signals were recorded using the OxiView software (TX3, version 6.02). Both sensor types were calibrated with a two-point calibration (0% and 100% air saturation). Electrodes were calibrated in N2 (0%) and air-purged (100%) filtered seawater taken at the surface of each site, whereas optodes were calibrated in sodium sulfite (Na2SO3)-saturated filtered seawater (0%) and water-saturated air (100%). Both sensor types were calibrated using the PT 1000 temperature sensor. Calibration crosschecks between sensors showed no discrepancies. Dissolved oxygen concentrations were calculated from percent saturation by using the oxygen solubility at in situ salinity and the temperature measured at each sampling point. Electrode and optode measurements were conducted by inserting the sensors into the center of the intact sediment section (3 cm deep) through separate 4 mm holes drilled through the core liner. Electrode readings stabilized within 30–60 s, with a standard deviation <0.4%. Readings were logged for 1 min after stabilizing. Optode signals were left to stabilize for 3 min before being logged per second for 1 min. The standard deviation on optode measurements was <0.3%. Optodes frequently had to be replaced because of wear or breakage. When exchanged, duplicate measurements were conducted to confirm calibrations. Optode and temperature measurements were conducted at 10–20 cm intervals near the sediment surface and close to basement or oxygen depletion zones. Throughout the remaining sediment, optode measurements were taken at 2–4 depths per section (30–75 cm intervals), depending on the sediment cover. Optode measurements within each core were measured in a randomized order, except for at Site U1365. All optode measurements are shown for all sites. Electrode measures were generally performed at 10 cm intervals for the first two sections of cores at 1–3 mbsf and at a similar resolution close to the basalt/sediment interface. At Site U1365, four electrodes that were independently calibrated and connected to separate picoammeters were used simultaneously for the downcore measurements. For the subsequent sites, two independently calibrated electrodes connected to two independent picoammeters were used. Below 3 mbsf, electrode measurements were typically attempted at 20–25 cm intervals, although for some intervals visible disturbance of core sections prevented successful measurements at those intervals. Toward the basaltic basement or near zones of oxygen depletion (when reached), electrode measurements were conducted at 10 cm intervals. For illustrative purposes, the tables and figure for Holes U1365A and U1365B include values from sections that were visibly disturbed before or during the measurement (e.g., by fluid intrusion, sediment mixing, and gas voids) (see “Biogeochemistry” in the “Site U1365” chapter [Expedition 329 Scientists, 2011]). For the remaining sites, values determined using electrodes were not recorded for intervals where attempted measurements were abandoned because of obvious physical disturbance of the core section. Redox potentialRedox (Eh) was measured in sediments from Site U1371 using needle redox electrodes. Redox-potential sensors were connected to a portable or tabletop pH/mV meter (WTW-pH340, WTW GmbH, Weilheim, Germany; pH-m210 Meter Lab, Radiometer Analytical, Lyon, France; provided by the University of Southern California, Marine Biogeochemistry Laboratory). pH and alkalinityAlkalinity and pH in interstitial water were measured immediately after squeezing. The pH was measured with a combination glass electrode, and alkalinity was determined by Gran titration with an autotitrator (Metrohm 809 Titrando; provided by the University of Rhode Island Geobiology Laboratory). Five milliliters of interstitial water was titrated with 0.1 M HCl at 25°C. CRM 104–certified reference material obtained from the laboratory of Professor Andrew Dickson, Marine Physical Laboratory, Scripps Institution of Oceanography (California, USA), was used for calibration of the acid. Standardization was conducted at the beginning and end of a set of samples for each site and after every fifth sample. Dissolved inorganic carbonDIC was measured with the OI Analytical Aurora 1030C total organic carbon (TOC) analyzer, consisting of a syringe module, a sample-stripping manifold, and an infrared CO2 analyzer. Interstitial water samples (1 mL for each injection) were acidified with 0.2 mL of 2 M HCl. The CO2 released was stripped and flowed through the CO2 analyzer. The CO2 Beer-Lambert absorption law was integrated to determine the total CO2 released from the sample. The laboratory standard for DIC was Batch 94 (2015.92 ± 0.85 mM/kg) or Batch 104 (2020.10 ± 0.39 mM/kg) certified reference material obtained from the laboratory of Professor Andrew Dickson, Marine Physical Laboratory, Scripps Institution of Oceanography. The standard was analyzed at least three times at the start of analyses to verify instrument stability and was subsequently analyzed following every fifth sample to constrain instrument drift. No significant drift was observed during each analytical run. The average of all standard measurements was used in the concentration calculations. Concentrations were calculated using the following equation:
where Asample and Astandard are the integrated Beer-Lambert CO2 absorptions of the sample and standard, respectively; Cstandard is the DIC concentration of the standard, the molar concentration (moles/liter) which was taken as the certified molal concentration (moles/kilogram) times a density of 1.025 kg/L. A single determination consisted of four separate injections of sample. The calculated concentrations are based on the average of the latter three absorptions because the first injection tends to show a low value. Chloride and sulfateSulfate and chloride were quantified with a Metrohm 861 Advanced Compact ion chromatograph (provided by the University of Rhode Island Geobiology Laboratory). The ion chromatograph comprises an 853 CO2 suppressor, a thermal conductivity detector, a 150 mm × 4.0 mm Metrosep A SUPP 5 150 column, and a 20 µL sample loop. A Metrohm 837 ion chromatography eluent/sample degasser was coupled to the system. The column oven was set at 32°C. The eluent solution was 3.2 mM Na2CO3 and 1.0 mM NaHCO3. A 1:50 dilution of interstitial water with 18.2 MΩ deionized water was analyzed. Aliquots of a 20 L secondary standard (~1:50 dilution) were used in all analytical runs. This standard was calibrated against International Association for the Physical Sciences of the Oceans (IAPSO) standard seawater (18 alternate analyses of the secondary standard and IAPSO standard). All samples were diluted using the same pipettes as were used in the IAPSO calibration (10 mL dispensette and 200 fixed-volume Eppendorf pipettor). The analysis sequence in an analytical run (36 analyses) was three standards followed by a repetitive sequence of four samples and a standard. In each run, two dilutions (duplicates) of each sample were analyzed. In each run, four IAPSO samples were analyzed to quantify long-term reproducibility and as quality assurance for each run. Reproducibility was also calculated for each run based on the sample and IAPSO duplicates using the method of combined (pooled) standard deviations. If the range between two duplicates was greater than three times the standard deviation, the data were not accepted and another pair of duplicates of the same sample was prepared and analyzed. If the measured concentration of IAPSO standard in a run was not within the expected range (based on the combined standard deviation and 95% confidence interval based on student-t values), the data from the entire run were not accepted and all the samples were reanalyzed. This occurred in only one analytical run. All quantification was based on peak areas. Chloride peak areas of the standard were examined to identify sensitivity drift in each run. If there was no significant drift, the average value within an analytical run of the Cl peak area for the standards was used in concentration calculations. If drift was detected, either a first- or second-order least-squares fit to the response was determined and used for the calculation of concentrations. For sulfate, calculations were based on the measured S/Cl peak area ratio, as this method removes variations caused by temperature-dependent changes in injected volume and sample dilution. Sensitivity drift was corrected in a manner similar to that used for chloride. The reported values are given as a sulfate anomaly, symbolized by ☺, which is the percent difference in the S/Cl ratio of the sample relative to the IAPSO standard,
where
and, for any sample or standard,
The combined standard deviation of Cl– analyses for all duplicates is 0.13%. Because all samples were analyzed in duplicate, the standard error is given as 0.09%. The combined standard deviation of sulfate anomaly (☺) analyses for all duplicates is 0.07% with a standard error of 0.05%. Reported concentrations are based on a Cl– concentration in IAPSO of 559.5 mM and a SO42– concentration of 28.94 mM. Sulfate concentrations were calculated from sulfate anomalies and chloride concentrations with the following formula:
Nitrate measured from interstitial water collected by Rhizon samplerDuring previous subseafloor microbiological research expeditions, it was learned that squeezing to obtain a interstitial water sample generated a high nitrate blank corresponding to between 1 and 5 µM NO3 in the sample. This is an unacceptably high value for a blank. Because of this problem, nitrate concentrations in interstitial water during Expedition 329 were measured using interstitial water from Rhizon samplers. This practice allowed us to effectively quantify even the lowest concentrations of nitrate (<1 µM). Nitrate concentrations were analyzed with a Metrohm 844 UV/VISCompact ion chromatograph (provided by the University of Rhode Island Geobiology Laboratory). A 150 mm × 4.0 mm Metrosep A SUPP 8 150 column was used. The column oven was set at 30°C. The eluent was a 10% NaCl solution filtered through a 0.45 µm filter. Approximately 0.8 mL of interstitial water was injected manually into a 250 µL sample loop. Absorption at the 215 nm channel was used for quantification. A 50 µM sodium nitrate/nitrite standard was run after every second, third, or fourth sample depending on instrument stability. The 50 µM standard was calibrated with CRM 104–certified reference material obtained from the laboratory of Professor Andrew Dickson, Marine Physical Laboratory, Scripps Institution of Oceanography. Spectrophotometric analyses of phosphate and dissolved silicaPhosphate and dissolved silica concentrations in the squeezed interstitial water samples were determined using an OI Analytical discrete analyzer (DA3500) spectrophotometer unit. This is an automated system that controls sample analysis and reagent aspiration, dispensing, heating, and mixing. A 1–2 mL aliquot from the interstitial water sample was pipetted into a DA3500 sample cup. Samples were selected in random order for analysis. Phosphate and dissolved silica analyses were conducted independently, with phosphate being measured first. In the phosphate method, orthophosphate reacts with Mo(VI) and Sb(III) in an acidic solution to form an antimony-phosphomolybdate complex. Ascorbic acid reduces this complex, forming a blue color that is measured at 880 nm. Potassium phosphate monobasic (KH2PO4) dissolved in 18.2 MΩ water was used to produce a calibration curve with 1, 2, 3, 4, and 5 µM concentrations through the autodilution program on the discrete analyzer to check for instrument linearity. A secondary standard (2.5 µM) was prepared from stock solution prepared from potassium phosphate monobasic (KH2PO4) dissolved in 18.2 MΩ water. This standard was used to test accuracy and drift during the analytical runs. An additional certified seawater sample (0.77 µM/kg), from the laboratory of Professor Andrew Dickson was also used to evaluate the accuracy and detection limits of the method. Standard additions of phosphate to pure water and to seawater were tested, and the differences were determined to be negligible. During an analytical run, phosphate was determined in triplicate and, at Site U1368, phosphate was also determined over duplicate runs. Precision was based on pooled standard deviations estimated using the equation (McNaught and Wilkinson, 1997)
where
Silica in solution as silicic acid or silicate is reacted with a molybdate reagent in acid media to form molybdosilicic acid. The complex is reduced by ascorbic acid to form molybdenum blue, measured at 420 nm on the discrete analyzer. A primary standard was prepared by adding a weighed amount (~0.5642 g) of sodium silicofluoride (Na2SiF6) to 18.2 MΩ water. This standard was used to produce a curve of calibration standards according to the instructions given in Gieskes et al. (1991). Standards between 120 and 900 µM were freshly prepared using 5% artificial seawater. We found significant differences in standard curves between those prepared in the artificial seawater as described in Gieskes et al. (1991) and those prepared in pure water. Because of what appears to be a strong dependence of the kinetics of the reaction on time, drift could be substantial. This drift was accounted for by running an independent standard at regular intervals (every five samples). Except for Site U1365 (triplicates), duplicate measurements were performed. Precision was based on pooled standard deviations for duplicate samples that are derived from Equation 7, as discussed above (McNaught and Wilkinson, 1997):
where xi1 and xi2 are duplicate measurements. Analysis of cations in interstitial water by suppressed ion chromatographyConcentrations of dissolved cations (Na+, K+, Ca2+, and Mg2+) in interstitial water were determined by means of suppressed ion chromatography. Aliquots of interstitial water were analyzed on an ICS 3000 ion chromatograph (Dionex). The ICS 3000 combines automated eluent generation and self-regenerating suppression with a conductivity detector. IAPSO primary reference material was used for standardization and quantification. Each aliquot of interstitial water was diluted (1:200) with 18.2 MΩ water and aliquoted to a 10 mL auto-sampler vial. Dilutions were performed manually or by Hamilton autodilutor, depending on time constraints. For each analysis, 25 µL of the sample was injected into the sample loop. Methanesulfonic acid (18 mM) was used as a mobile phase. The pump operated in isocratic mode with a 0.250 mL/min flow rate. The cation exchange IonPac-GS12A (2 mm × 50 mm) guard column and IonPac-CS12A (2 mm × 250 mm) analytical column were operated at ambient temperature (25°C). Suppressor current was set to 14 mA. Detector cell temperature was set at 35°C. Calibration was based on a six-point calibration curve, a blank and five increasingly diluted standards, to bracket the concentrations expected in diluted samples to be analyzed. Diluted IAPSO standard (P145 batch, K15-0.99981) was used for all calibrations. For Site U1365, a 1:200 dilution standard was used, whereas for the remaining sites a 1:180 dilution was used to further increase signal-to-noise ratio. Analytical runs consisted of repetitive sequences of IAPSO standard and four samples. With the exception of Sites U1365 and U1366, in all analytical runs three quality assurance standards per batch were analyzed to assess reproducibility. Two dilutions of each sample were analyzed. Both replicates were rejected when the relative standard deviation between them was >2% of the measured value. Quantification was based on peak area; baselines and peak integration were manually selected by the operator. If drift was observed, sample concentrations were corrected for the drift on a point-to-point basis. Accuracy was determined by triplicate analysis of IAPSO standard. Values of precision and accuracy are reported in each site chapter. Inductively coupled plasma–emission spectrometryMajor and minor elements (Ca, Mg, K, Na, Fe, Ba, B, and Mn) in the interstitial water were determined by inductively coupled plasma–atomic emission spectrometry (ICP-AES) with a Teledyne Prodigy high-dispersion ICP spectrometer. The general method for shipboard ICP-AES analysis of samples is described in Murray et al. (2000) and the user manuals for new shipboard instrumentation with modifications as indicated. Samples and standards were diluted 1:20 using 2% HNO3 for trace element analyses (B, Mn, Fe, Sr, and Ba) and 1:200 for major element analyses (Na, K, Ca, and Mg). Each batch of samples run on the ICP spectrometer contains blanks and solutions of known concentrations (such as IAPSO). Samples were analyzed in batches, commonly one batch per site, to take advantage of achieved calibration and maximize intrasite precision. Each item aspirated into the ICP spectrometer was counted four times from the same dilute solution within a given sample run. Following each run of the instrument, the measured raw-intensity values were corrected for instrument drift and procedural blank. If necessary, a drift correction was applied to each element by linear interpolation between the drift-monitoring solutions. For many of the sites, there was no unidirectional instrumental drift >1% (total through the entire run of multiple hours), and in such cases no drift correction was employed. After drift correction, where appropriate, and blank subtraction, a calibration for each element was calculated using the results from the analyses of known solutions. For the major elements and Sr, because the calibration lines were strongly linear and concentrations deviated only slightly from seawater concentrations, the final concentrations were calculated on the basis of the ratio to the analysis of IAPSO standard. IAPSO standard was measured at least four times through a run (and as many as eight times), each time as an unknown; therefore, it was straightforward to use some IAPSO samples for precise determination of the ratio and other IAPSO samples, not used in this calibration, to provide an independent check on the resultant accuracy and precision. This strategy also enhanced long-term, intersite reproducibility. Typical relative standard deviations of the quadruplicate counts per acquisition were better than 0.5% of the measured value for each element. Precision and accuracy are reported on a site-by-site basis in each site chapter. Sedimentary nitrogen, organic carbon, and inorganic carbonWe routinely sampled 5 cm3 of sediment from the working half of each core on the sampling table after splitting the core. Sample aliquots were freeze-dried for ~24 h and crushed to a fine powder using an agate mortar and pestle. Geochemical analyses included carbonate and elemental analyses (C, H, N, and S). Total inorganic carbon (TIC) concentrations were determined using a UIC 5011 CO2 coulometer. Between 10 and 150 mg of freeze-dried, ground sediment was weighed and reacted with 2N HCl. The liberated CO2 was titrated and the end-point determined using a photodetector. Calcium carbonate content, expressed as weight percent, was calculated from the TIC content, assuming that all evolved CO2 was derived from dissolution of CaCO3, by the following equation:
No correction was made for the presence of other carbonate minerals. Accuracy during individual batches of analyses was determined by running a carbonate standard (100 wt% CaCO3) every five samples. Analytical precision and accuracy for CaCO3 based on an internal calcium carbonate standard were ±0.6% and ±1.2%, respectively. The detection limit for CaCO3, defined here as three times the standard deviation of the blank (2N HCl), was 0.1% for 100 mg of pelagic clay. Total carbon and nitrogen contents of sediment samples were determined with a ThermoElectron Corporation FlashEA 1112 CHNS elemental analyzer equipped with a ThermoElectron packed column CHNS/NCS and a thermal conductivity detector. About 10–40 mg of freeze-dried, ground sediment was weighed into a tin cup and combusted at 900°C in a stream of oxygen. Total carbon content of samples from Site U1365 had to be corrected for carbon contamination associated with the vanadium pentoxide reagent. Correction values are between 0.009% and 0.021%, depending on sample weight during analysis. Replicate analysis of selected samples without vanadium pentoxide shows a good correspondence with the corrected total carbon values. After Site U1365, we discontinued use of vandium pentoxide. All measurements were calibrated to a standard, which was run every five samples. In low-carbonate sediments, cysteine was used as the calibration standard, whereas sulfanilamide was used for carbonate ooze samples. The detection limit was 0.001% for total nitrogen (instrument limit) and 0.002% for total carbon (procedural blank, measured as an empty tin cup). The average standard deviation for total carbon replicates was 0.08% in carbonate-rich samples and 0.01% in clay-rich samples. The average standard deviation for total nitrogen replicates was 0.001% in both carbonate-rich and clay-rich samples. TOC was obtained directly from the CHNS elemental analyzer after treating subsamples (5–20 mg) with 140–200 µL 1N HCl to remove the inorganic carbon fraction. We followed the standard operating procedure for measuring TOC by elemental analysis that was adapted from Verardo et al. (1990) for shipboard use during Expedition 320/321 (Expedition 320/321 Scientists, 2010). However, we increased the amount of acid added from 10 to 20 µL steps. After the final step (i.e., when no reaction was observed after the addition of acid) samples were placed in a 50°C oven to dry overnight. The detection limit for TOC was 0.002% (procedural blank, measured as a silver cup containing HCl, dried and packed into a tin cup). Replicate analysis of selected samples showed small standard deviation (≤0.01%, with a few samples ≤0.02%). |