METHODS

Gas samples were collected during Leg 164 using four different techniques:

1. Free gas samples were extracted by inserting a piercing tool through the core liner and into sediment voids. At shallow depths free core gas was extracted from the core liner by opening a valve to an evacuated container connected to a hypodermic needle. Below about 50 mbsf, the gas within the voids generally escaped under its own pressure into a 60-cm³ syringe.
2. Short sections of core were sealed in gas collection tubes to collect the gases that evolved from core sections as the cores warmed within the core laboratory (Paull, Matsumoto, Wallace, et al., 1996). Selected core sections (cut into the standard 1.5-m-long lengths on the catwalk) were placed in 1.54-m-long tubes that were slightly larger in diameter than the core liners. The tubes were constructed from standard PVC schedule-40 pipe sections (ID ~3.5 in) with end caps that were machined so that they would seal on both ends with O-ring-fitted plates that could be bolted on quickly to make a gas-tight seal. Holes were drilled in the core liners just before the cores were sealed in the tube to let gas escape. Thus, all the gas that evolved within the core after they were sealed in the tubes could be collected and analyzed. Gas from each core flowed out one end of a gas collection tube through tygon tubing into an overturned 1-L graduated cylinder, floating in a bath filled with NaCl-saturated water. Because the cylinders were initially water filled, the volume of water displaced by the gas could be measured directly and sampled through luer-lock fittings by 60-cm³ syringes.
3. The pressure core sampler (PCS) is a downhole tool that seals a small diameter sediment core in an internal chamber at in situ pressure. Upon recovery, the gases within the sealed PCS were vented in a stepwise fashion (Dickens et al., 1997), and aliquots of the gas were collected for compositional and isotopic analysis.
4. Gas samples were collected from five solid gas hydrate samples that were individually placed in a sealed chamber, quickly evacuated, and allowed to dissociate within the closed vessel (Lorenson et al., Chap. 25, this volume).

Relative concentrations of methane, carbon dioxide, and major low molecular-weight hydrocarbon gases (ethane, propane, and butane) were measured on shipboard by gas chromatograph (Paull, Matsumoto, Wallace, et al., 1996). The remaining gas was transferred underwater into inverted 20-mL glass vials, sealed with a rubber stopper, and crimp-sealed with aluminum caps for transport and storage.

¹³C values of CH₄ and CO₂ were measured with a Finnigan MAT 252 gas-chromatograph-isotope-ratio-mass-spectrometer (GC-IR-MS) at the University of North Carolina-Chapel Hill (UNC-CH) using the methods of Merritt and Hayes (1995) and Popp et al. (1995). The results of 251 CH₄ and 137 CO₂ gas analyses are reported in Table 1 and Table 2. Twelve samples contained enough ethane for ¹³C measurements (Table 3).

Forty-one CH₄-bearing samples were analyzed for their hydrogen isotope composition (Table 4). Methane was separated using a cryogenic vacuum line and combusted to CO₂ and water. The water was converted to H₂ gas in a reaction with hot zinc and captured in glass break seals (Coleman et al., 1982; Kendall and Coplen, 1985; Hayes and Johnson, unpubl. data). D measurements were made with a Finnigan MAT 252 mass spectrometer at UNC-CH.

Pore-water samples were stored in flame-sealed glass ampoules for shore-based isotopic measurement of total DIC. DIC within pore-water samples was separated and collected using standard, cryogenic vacuum line techniques (Craig, 1953). Carbon isotope measurements were made on 105 samples (Table 5) using a Delta E mass spectrometer at North Carolina State University.

The bulk organic fraction in 58 sediment samples was analyzed for ¹³C, ¹⁵N, and C:N values (Table 6) using a technique modified from Hedges and Stern (1984). Residual solid phase samples from the shipboard pore-water extraction were broken, and ~50-g pieces from the interior were subsampled and placed in precombusted Pyrex beakers. The samples were freeze dried and then crushed with a mortar and pestle. Dry, powdered samples were placed in a vapor chamber containing 12N hydrochloric acid for 24 hr to remove the inorganic carbon fraction. This technique works well for samples that contain less than 20% CaCO₃; samples with more than 20% CaCO₃ are not reported here because of incomplete CaCO₃ digestion. Subsamples were combusted within a Carlo Erba C/N elemental analyzer and passed directly through the Finnigan MAT 252 GC-IR-MS at UNC-CH providing a ¹³C and ¹⁵N value for each sample. Chromatographic peak areas were used to calculate the atomic C:N value for each sample.

Isotopic values are reported with respect to PDB for ¹³C, SMOW for D, and air for ¹⁵N values. The cumulative (vacuum line, preparation, and mass spectrometer) accuracy of isotopic measurements are +0.2 for ¹³CC0₂, ¹³CCH₄, ¹³CDIC, ¹³COrg. C, ¹³CC₂H₆, and ¹⁵NOrg. C, and +3 for DCH₄ Neal Blair and Howard Mendlovitz, pers. comm., 1997). However, because these samples have experienced obvious degassing, we suspect that the intersample variations inherent in the sampling methods far exceed the laboratory errors for the various gas measurements.