INTRODUCTION

The amount, molecular distribution, and isotopic composition of hydrocarbon gases recovered in marine sediments have frequently been used to infer the origin of the gases and the fate of buried organic carbon (e.g., Claypool and Kaplan, 1974; Paull et al., Chap. 7, this volume). However, sediment cores recovered from deep water and at high pressure by normal coring techniques have lost hydrocarbon gases by the time cores arrive at the surface for sampling. The quantity of gas lost and any potential fractionation of the hydrocarbon distribution or isotopic composition have not been emphasized in the scientific literature because the technology to address these issues--pressure coring with a manifold to collect gas volumes--had been unsuccessful prior to Leg 164 (Kvenvolden et al., 1983; Pettigrew, 1992; Paull, Matsumoto, Wallace, et al., 1996).

Leg 164 sites were drilled into a large reservoir of gas hydrate- and free gas-bearing sediments on the Blake Ridge ~200 km off the coast of South Carolina (Paull, Matsumoto, Wallace, et al., 1996; Dickens et al., 1997). The pressure core sampler (PCS), a tool designed to recover a 1385-cm³ cylindrical sediment core at in situ pressure (Pettigrew, 1992), was used numerous times during Leg 164 with appropriate manifolds to successfully recover incremental gas volumes from pressurized cores of deep-sea sediment (Paull, Matsumoto, Wallace, et al., 1996; Dickens et al., Chap. 43, this volume). Comparisons between the total amount of gas recovered by the PCS (Dickens et al., 1997, and Chap. 43 and Chap. 11, this volume) and by conventional ODP coring techniques at the same depth (especially by XCB; see Paull, Matsumoto, Wallace, et al., 1996) can be striking: often in excess of 95% (and in some cases, at least 99.8%) of the hydrocarbons were lost during the 20-30 min between conventional coring at depth and gas sampling on the surface.

An important issue is whether such extreme and rapid loss of gas from sediment during conventional core retrieval is associated with fractionation of the molecular distribution or isotopic composition of hydrocarbon gases. In this manuscript, we present carbon isotope measurements of CH₄ released from sediment recovered using the PCS during Leg 164. We show that the ¹³C of individual gas increments can have a wide range in ¹³C composition. However, the total ¹³C of volumetrically weighted averages of gas samples from individual PCS cores is similar to the ¹³C of gas sampled from gas voids in conventional sediment cores at comparable depths. We address the cause of carbon isotope variation in incremental gas volumes released from the PCS by comparing the PCS gas data with results of degassing experiments in simple, sediment-free analog systems. Data concerning total amounts of gas (predominantly CH₄) and hydrocarbon composition from PCS cores are reported elsewhere (Paull, Matsumoto, Wallace, et al., 1996; Dickens et al., 1997, and Chap. 43 and Chap. 11,  this volume).