Marine slope failure involving methane-gas-hydrate-bearing sediments is one mechanism for releasing enormous quantities of methane to the ocean and atmosphere. The Storegga Slide, on the Norwegian margin, is the largest known Holocene-aged continental margin slope failure complex and is believed to have occurred in sediments that may have initially contained gas hydrate. Here, we report pore water sulfate gradient measurements that are used as a proxy for the relative amounts of methane that exist in continental margin sediments associated with the colossal Storegga Slide. These measurements suggest that a considerable inventory of methane occurs in sediments adjacent to, and unaffected by, the Storegga Slide events, but indicate that methane is notably absent from sediments on the sole of the slide and distal deposits created by the slide events. Either methane was lost during previous Pleistocene failure events or was never present in significant concentrations within the sediments that failed.
 A lively debate is underway over whether methane release from marine interstitial gas and gas-hydrate reservoirs and Earth's climate are connected [Paull et al., 1991; Dickens et al., 1997; Haq, 1998; Mienert et al., 2001, 2005; Sowers, 2006]. At issue is whether adequate amounts of methane can be released to the ocean from the gas reservoirs at shallow depths within continental margin sediments to increase the dissolved inorganic carbon composition of the ocean or the methane concentration of the atmosphere. Such transfers of methane carbon have been invoked to explain the globally synchronous, geologically abrupt shifts in the marine carbonate carbon isotope record [e.g., Dickens et al., 1997; Hesselbo et al., 2000; Kennett et al., 2000, 2003].
 Methane-bearing gas-hydrate deposits are common within continental margin sediments worldwide and are one of the largest potentially-mobile inventories of carbon on Earth [Kvenvolden and Lorenson, 2001]. Gas hydrates are ice-like solids composed of low molecular weight hydrocarbon gases, principally methane, and water that are stable at the temperatures and pressures that occur within continental margin sediments. Theoretically, gas-hydrate deposits can be easily perturbed by the temperature and pressure changes caused by global climate changes seen in the geologic record [Kennett et al., 2003]. Whether enough methane is rapidly released from these deposits and how this methane is transferred from geologic reservoirs to the ocean and/or atmosphere remain outstanding, unanswered questions [Kvenvolden and Lorenson, 2001].
 Continental margins contain abundant slide scars left by submarine landslides [Hampton et al., 1996]. Seismic reflection profiles from the flanks of these features commonly show bottom simulating reflectors (BSR), which indicate gas hydrate may occur in sediments immediately adjacent to many of these slide scars. This observation implies that sediments involved in many submarine landslide events initially contained gas hydrate [Paull et al., 2000; Maslin et al., 2004].
 Submarine landslides are commonly associated with enough turbulence to destroy the cohesion between sediment grains [Hampton et al., 1996]. As a result, the original fabric of the sediments will be destroyed as the sediment disintegrates into its constituent grains and a dense, turbid water mass is produced. This dense water mass will move energetically downslope as a gravity-driven flow. Such flows can travel great distances before the sediments are re-deposited within deep-sea basins in hydrodynamically sorted beds. During a submarine landslide event methane gas bubbles and pieces of gas hydrate in the disrupted sediment will float upwards into the overlying water column because both phases are less dense than seawater [Paull et al., 2003]. One consequence of a large slope failure event may be the complete loss of methane-gas-hydrate and free gas to the oceans and/or atmosphere.
 The Storegga Slide, located on the Norwegian margin (Figure 1), is the largest known continental margin landslide scar. The last major event occurred 7,250 ± 250 14C years or 8,100 ± 250 calendar years ago [Haflidason et al., 2005] and left a scar and deposits that extend 810 km across the Norwegian continental margin from ∼400 m water depth and fill the adjacent Norwegian Basin in ∼3,900 m of water. Between 2,500 and 3,200 km3 of sediment were involved in the last major event [Haflidason et al., 2004, 2005]. Similar-sized slides have occurred within the area of the last Storegga Slide at ∼100 ky intervals over the last 0.5 Ma [Bryn et al., 2005]. The occurrence of a BSR in seismic reflection profiles on the northern flank of the slide scar indicates the presence of gas hydrates in these sediments today [Bünz et al., 2003]. Presumably, methane hydrates were also contained in the sediments removed by the slides, including the last major event.
 The amount of methane hydrate originally contained within sediments disturbed by the Storegga landslide events is unknown. The facies of the sediments that remain on the northern and southern flanks of the side are different. Glacial sediments occur on the southern flank of the slide while fine-grained hemipelagic material characterizes the northern flank. BSRs have been detected on the northern flank of the Storegga Slide, but not on its southern flank. Nevertheless, estimates can be made using available seismic survey data [Bünz et al., 2003] and methane amounts found in sedimentary sections similar to those that occur on the northern flank of the side which contain a BSR and have been drilled elsewhere [Paull et al., 1996]. If conditions similar to those that currently exist for the northern gas hydrates extended across half the slide scar area prior to the last major slide event, we propose that 50% of the pre-slide sediment volume containing 1% gas hydrate by volume is a reasonable estimate of the pre-slide methane inventory. These estimates indicate that 16 km3 of pure methane-gas-hydrate may have been contained in the sediments that were involved in the last major Storegga landslide event. This amount of gas hydrate contains 1.4 Gt of methane carbon assuming a methane-gas-hydrate stochiometry of CH4•5.9H2O and a density of 0.91 g•cm−3 [Sloan, 1998]. This is approximately 90% of the methane gas inventory of the pre-industrial Holocene atmosphere [Alley et al., 1997; Sundquist and Visser, 2004]. Methane gas dissolved in the sediment pore water could also have been released, further increasing the potential impact of the event. While considerable uncertainty exists about how much gas hydrate originally was contained within these sediment, the rapid release of a quantity of methane similar to that we estimated might have been present would have perturbed the radiative balance of the Earth's atmosphere and left a geochemical record in ice cores and/or deep sea sediments. Thus, the last major Storegga landslide event appears to be the prime test case for evaluating the impact of methane release from large submarine slope failures.
 A globally synchronous signature in the ice core or marine sediment record indicating the addition of methane carbon to the oceans and atmosphere from the last Storegga landslide event has not been identified. The Greenland ice cores at the time of the last Storegga landslide event [Alley et al., 1997; Alley and Agustsdottir, 2005] show a local minimum in the atmospheric methane concentration and indications of cooler climate [Ellison et al., 2006], both of which are inconsistent with the predicted effects associated with a major out-gassing of methane at this time. The carbon isotopic composition of benthic foraminifera in sediment cores from the North Atlantic [Hall et al., 2004] also do not show distinct 13C-depleted spikes in the carbon isotope record at the time of the last major Storegga landslide event.
 Thus, establishing whether gas hydrate was initially present and the fate of the methane gas and gas hydrate contained within the pre-slide sediments is key to evaluating the broader impacts of the Storegga and other mega submarine landslide events. This issue is addressed here using pore water sulfate gradients as a geochemical proxy for subsurface methane distribution.
 Pore water geochemical profiles extending from the seafloor into marine sediments undergo a sequence of progressive changes that reflect the increasing amounts of chemical reduction. The dominant reactions include microbially-mediated consumption of oxygen, followed by sulfate reduction [Claypool and Kaplan, 1974], and methane production below the depth of sulfate depletion. Sulfate needs to be depleted in the pore water before microbial methane production and accumulation will occur [Martens and Berner, 1974]. However, when methane is abundant below the depth of sulfate depletion, sedimentary microbes anaerobically oxidize methane using sulfate as an electron acceptor [Reeburgh, 1976] at this interface between sulfate and methane [Borowski et al., 1996]. As a consequence of anaerobic oxidation of methane at the sulfate-methane interface (SMI), pore water sulfate concentration gradients in the upper sedimentary section are sensitive to variations in the amount of vertical transport of methane to the seafloor. Unlike methane, pore water sulfate concentrations are not altered by pressure changes that occur during sample recovery [Borowski et al., 1999]. Thus, measurements of sulfate gradients provide constraints on the amount of methane accumulation and/or the potential for methane hydrate to occur within near-seafloor sediments. At steady-state conditions, the methane concentration gradient below the SMI is mirrored by the sulfate concentration gradient above the SMI (Figure 2). Sediments overlying shallow gas hydrate accumulations will have larger amounts of methane diffusing upward than sites without gas hydrate. Thus, variations in the depth to the SMI are a relative measure of the upward methane flux near the seafloor. Depth to the SMI has been used to estimate the relative amounts of interstitial methane and/or detectable gas hydrates that occur in the subsurface [Borowski et al., 1999].
 One consequence of the sudden removal of the upper sediment veneer by a major slope failure event like the Storegga Slide is that the entire zone of sulfate-bearing sediments may be removed, exposing sulfate-free sediments on the newly formed seafloor (Figure 2). Over time, sulfate will diffuse back into the sediments. Here we consider whether the post-slide sulfate gradients match the recovery path of sulfate gradients predicted for sediments affected by the last major Storegga landslide event, assuming the amount of gas and gas hydrate within the sediments that failed was similar to that in the sediments on the flanks.
 In August and September 2004 a coring cruise was conducted from the R/V Knorr to investigate the pore water geochemistry of sediments on the flanks, sole, and depositional basin associated with the Storegga Slide scar (Figure 1). Pore waters were extracted from whole-round sediment samples cut from fresh cores at 1-m intervals. Pore water sulfate concentrations were analyzed shipboard using a Dionex DX-120 ion chromatograph. SMI depths were determined directly or by extrapolation (Figure 3).
 Eleven cores of up to 20 m long were collected in 806 to 1,524 m water depth from undisturbed fine-grained hemipelagic sediments on the northern flank of the slide (Figure 1) where BSRs are mapped [Bünz et al., 2003]. Sulfate profiles show that the SMI occurs between 5 and 12 meters below seafloor (mbsf) along the northern flank of the slide (Figure 3). Core lengths, locations, and sulfate concentration data are provided in the auxiliary materials.
 Four cores were taken from the sole of the slide in 1,408 to 1,801 m water depths. These cores contained very firm, apparently over-consolidated clays. Their lithology and position suggest that tens of meters of overburden were removed. Sulfate concentration data in one core extrapolate to an SMI depth of 42 mbsf, while the other cores show nominal decreases in sulfate concentration with depth (Figure 3). These cores are all outside the area where Bünz et al.  indicated the existence of a BSR.
 Five cores were taken in 3,744 to 3,825 m water depths in the Norwegian Basin. These basin cores contained up to 65 cm of hemipelagic sediment that overlies a monotonous sequence of under-consolidated clays that is more than 18 m thick. The homogeneous texture suggests these sediments flowed into the basin as sediment slurry during the last massive Storegga landslide event. Pore water sulfate concentrations are near or slightly higher than seawater-like values throughout the cored interval (Figure 3).
 Sulfate gradients measured in the northern flank of the Storegga Slide indicate that the SMI depths are 5 to 12 mbsf, which are shallow in comparison with other known methane hydrate-bearing sites in similar lithologies [Borowski et al., 1999; Paull et al., 2005]. These SMI depths confirm that conditions suitable for methane accumulation and gas hydrate formation exist in the sediments on the northern flank of the Storegga Slide.
 Sulfate gradients in cores from the northern flank are distinctly steeper than those from either within the slide scar or from sediments that were deposited in the Norwegian Basin as a result of the most recent slide (Figure 3). The sediments that are now on the sole of the Storegga Slide were tens of meters below the seafloor before the slide events removed the overburden [Bryn et al., 2005]. If the sediments exposed on the sole of the slide were initially similar to the sediments on the northern flank today, they would have contained sulfate-free pore waters and an appreciable amount of methane before the slope failure events occurred. Overburden removal would expose methane-bearing and sulfate-free sediments on the seafloor. While re-equilibration via the diffusion of seawater sulfate back into the sediments over the last 8.2 ka will decrease the initially steep gradients, it would not by itself produce a gradient that is less than currently observed on the flank (Figure 3). Apparently the sediments inside the slide scar had distinctly less methane 8.2 ka ago than is currently present outside the slide.
 The sediments that have accumulated in the Norwegian basin were derived from the potentially methane-charged sediments that failed upslope. The sulfate gradients in these basin sediments are inconsistent with significant amounts of methane hydrate being retained within the sediment. The presence of slightly higher-than-seawater sulfate within the basinal deposits suggests that the sediments sampled were completely mixed with oxygenated and sulfate-bearing seawater during transport into the basin. The increase in sulfate with depth suggests that sulfide minerals within the sediment have been oxidized to sulfate. Apparently, the rate of sulfate reduction in these re-deposited sediments has not produced a significant effect on the sulfate profiles in the last 8.2 ka.
 A one-dimensional diffusion model was employed to assess the minimum rate of seawater sulfate diffusion into sulfate-free sediment, which presumably was exposed on the seafloor by the last major Storegga landslide event (Figure 4). The model is based on equation 2.45 of Crank  for a semi-infinite medium when the boundary is kept at constant concentration. In the model the boundary condition at the sediment-water interface was maintained at 28 mM sulfate, no methane was present, and no sulfate was consumed within the sediment. These conditions provide a minimum estimate of the time-scale for equilibration of sulfate-free sediment exhumed by a submarine slide with overlying seawater. The presence of methane in the sediment pore water and its oxidation by anaerobic oxidation of methane will lengthen the re-equilibration time. The free solution diffusion coefficient used for sulfate ion, Do, is 5.8 × 10−6 cm2 sec−1 at 5°C [Li and Gregory, 1974]. An effective sediment diffusion coefficient (D) was computed from D = Do Φ2 [Lerman, 1979] using a sediment porosity (Φ) of 50%, which is conservative for the sediment in the footwall of the Storegga Slide.
 Comparison between the results of this model and the observed sulfate gradients on the sole of the Storegga Slide indicates that it would take considerably longer than 8.2 ky to achieve the observed sulfate gradients (Figure 4). Apparently, significant accumulations of methane-gas-hydrate and methane gas were either never present or lost during the early Storegga landslide events.
 The geochemical data from the Storegga slide indicate that accumulations of sedimentary methane gas and methane-gas-hydrate are not currently within the sediments exposed by the 8.2 ky Storegga landslide event nor contained in the sediments deposited at its distal end. Whether these sediments never contained significant amounts of gas or their original gas content was lost during earlier slide events (∼0.5 Ma) is still unresolved. Either way, these observations explain why the last major Storegga landslide event did not produce a detectible signature in the ice core methane concentrations and carbon isotope record in the North Atlantic region.
 Support for this work was provided by the NSF (OCE-0221366) and the David and Lucile Packard Foundation.