Environ. Sci. Technol. 2008, 42, 7368–7373
Measuring Temporal Variability in Pore-Fluid Chemistry To Assess Gas Hydrate Stability: Development of a Continuous Pore-Fluid Array L A U R A L . L A P H A M , * ,†,‡ JEFFREY P. CHANTON,§ CHRISTOPHER S. MARTENS,† PAUL D. HIGLEY,| HANS W. JANNASCH,⊥ AND J. ROBERT WOOLSEY# Department of Marine Sciences, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599, Department of Oceanography, Florida State University, Tallahassee, Florida 32302, Specialty Devices, Inc., Wylie, Texas 75098, Monterey Bay Aquarium Research Institute, Moss Landing, California 95039, and Center for Marine Resources and Environmental Technology, University of Mississippi, University, Mississippi 38677
Received April 30, 2008. Revised manuscript received July 7, 2008. Accepted July 21, 2008.
A specialized pore-fluid array (PFA) sampler was designed to collect and store pore fluids to monitor temporal changes of ions and gases in gas hydrate bearing sediments. We tested the hypothesis that pore-fluid chemistry records hydrate formation or decomposition events and reflects local seismic activity. The PFA is a seafloor probe that consists of an interchangeable instrument package that houses OsmoSamplers, long-term pore-fluid samplers, a specialized low-dead volume fluid coupler, and eight sample ports along a 10 m sediment probe shaft. The PFA was deployed at Mississippi Canyon 118, a Gulf of Mexico hydrate site. A 170 day record was acquired from the overlying water and 1.3 m below seafloor (mbsf). Fluids were measured for dissolved chloride, sulfate, and methane concentrations and dissolved inorganic carbon and methane stable carbon and deuterium isotope ratios. Chloride and sulfate did not change significantly over time, suggesting the absence of gas hydrate formation or decomposition events. Over the temporal record, methane concentrations averaged 4 mM at 1.3 mbsf, and methane was thermogenic in origin (δ13C-CH4 ) -32.4 ( 3.4‰). The timing of an anomalous 14 mM methane spike coincided with a nearby earthquake (Mw ) 5.8), consistent with the hypothesis that pore-fluid chemistry reflects seismic events.
Introduction Gas hydrates are crystalline compounds of hydrocarbon gases (mainly methane) contained within an ice lattice that form * Corresponding author phone: (850) 645-4639; fax: (850) 6442581; e-mail:
[email protected]. † University of North Carolina at Chapel Hill. ‡ Currently at the Department of Oceanography, Florida State University. § Florida State University. | Specialty Devices, Inc. ⊥ Monterey Bay Aquarium Research Institute. # University of Mississippi. 7368
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
where saturated methane, moderate salinities, high pressure, and low temperature conditions exist (1). They have been found naturally in continental shelf sediments and arctic permafrost, including many sites along the northern Gulf of Mexico (2). Worldwide hydrate deposits are thought to represent one of the largest carbon reservoirs on earth, containing ∼(1-5) × 1015 m3 or 500-2500 Gt of methane carbon (3). In spite of their magnitude, hydrates are considered relatively stable and a small source of atmospheric methane in current global methane budgets (4). If changes occur in any one of the stability conditions (e.g., level of gas saturation, salinity, pressure, or temperature), hydrates become unstable and decompose, possibly releasing enormous amounts of methane into the water column and potentially the atmosphere (5). Ice core records show large methane spikes over geologic time which are thought to be caused from hydrate decomposition events that act as positive feedbacks to global climate change (6). On current time scales, observations of hydrate rafting or decomposition events have also been documented (7, 8), as well as the persistence of hydrate outcrops over several years (9, 10). At hydrate sites in the Gulf of Mexico, the fluid flux from pore fluids to the overlying water has been shown to change over time, including times when the flux reverses (11, 12). Offshore from Peru, indirect evidence suggests that methane is released from hydrocarbon-rich, cold seep sites upon seismotectonic events (9). These previous observations suggest that hydrate deposits are dynamic and that temporal monitoring is needed to fully understand what controls their stability. It is therefore important to understand the factors that control the stability of hydrate deposits to understand their potential role in abrupt climate change (6, 13). Pressure, temperature, and salinity conditions are typically studied, yet in situ gas saturation conditions are rarely known. This study provides direct quantification of in situ dissolved gas and ion concentrations that can be utilized to determine hydrate stability. The purpose of our study was to determine hydrate stability through in situ measurements of temporal variations in methane and salinity pore-fluid concentrations at a hydrate site. We sought to determine whether changes in pore-fluid chemical composition could record hydrate formation or decomposition events. We hypothesized that such events would result in chloride concentrations that differed from background levels. The rationale for this hypothesis is that hydrates exclude salts upon formation (1, 14). Therefore, formation events would be recorded as higher chloride concentrations and decomposition events as lower chloride concentrations. In a tectonically active environment such as the Gulf of Mexico, we also hypothesized that the pore fluids would reflect seismic activity through spikes in methane concentration. These hypotheses were simultaneously tested by deploying a novel pore-fluid array (PFA) sampler that collects and stores pore fluids over time and retains the in situ sample during recovery to avoid degassing methane. Pore fluids collected during deployment were analyzed for chloride, sulfate, and methane concentrations and the stable carbon and hydrogen isotope ratios of methane and dissolved inorganic carbon. The isotope measurements helped decipher the source of the hydrocarbon-rich fluids, either biogenic or thermogenic.
Materials and Methods Study Site. The PFA was deployed within the Mississippi Canyon Lease Block 118 (MC 118), ∼100 miles offshore of 10.1021/es801195m CCC: $40.75
2008 American Chemical Society
Published on Web 08/29/2008
FIGURE 1. MC 118 bathymetric shaded relief map showing the PFA and Sleeping Dragon, a 4 m × 2 m hydrate outcrop (map by Leonardo Macelloni, University of Mississippi). The inset shows the Gulf with MC 118 (black dot) and the locations and timing of two earthquakes (stars). Louisiana in the Gulf of Mexico (28°,51.47′ N, 88°,29.52′ W, 880 m water depth, Figure 1). MC 118 is the site for a longterm gas hydrate monitoring station administered through the Gulf of Mexico Hydrate Research Consortium. Within MC 118, a ∼1 km2 topographic high was found with craterlike depressions and seafloor relief of up to 10 m within a 20 m distance (Leo Macelloni, personal communication). Topographic highs are associated with abundant authigenic carbonate rocks, chemosynthetic communities, and outcropping thermogenic gas hydrates. The mound has distinct areas of higher and lower microbial activity, suggesting significant physical changes over relatively short time intervals (15). Device Design. The PFA was specifically designed for MC 118, but was modeled after similar instruments utilized in convergent margin settings (16-18). It is a long-term seafloorpenetrating sample probe topped by OsmoSamplers (19) that slowly pump pore fluids from different sediment depths up to storage coils housed within an interchangeable instrument package (Figures 2 and 3). The PFA consists of a 10 m long probe shaft with filtered sampling ports, a cement weight, and an instrument package that houses the OsmoSamplers including sample coils, a high-pressure valve, and one side of a fluid coupler. The probe shaft and cement weight allow the PFA to be drop-deployed from a surface ship similar to a gravity corer. With the probe shaft and cement weight in place, the instrument package sits on top of the cement weight via a fluid coupler (the receiver mounted on the weight and the plug on the instrument package). This design allows a submersible to remove and replace instrument packages periodically over time without removing the probe shaft from the sediments, thus minimizing disruption of the sample site during subsequent visits. The continuous sampling provides a temporal record of in situ methane and dissolved ion concentrations. The PFA ports, samplers, coils, and valve are described below. PFA Sample Ports. Pore fluids are initially collected and filtered at eight sample ports that run along a hollow steel
FIGURE 2. A schematic of the PFA. The 10 m probe shaft is deployed, under its own weight, about 7 m deep into the sediment, shown here with eight ports, four of them (closed circles) connected to the four samplers, housed within the removable sampler package, via the fluid coupler. (10 m long × 8 cm square) probe shaft (Figure 2). Ports are evenly spaced 125 cm apart and made up of a stainless steel frit (20 cm diameter, 20 µm stainless steel frits) in contact with small-diameter tubing (0.159 cm outer diameter polyVOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7369
FIGURE 3. Pore-fluid flow direction through the OsmoSampler, fluid coupler, and sample port. Only one sample depth is shown and, thus, one tubing line. At the sample port, pore fluids flow (1) across the frit and are filtered, (2) up the small-diameter tubing along the probe shaft, (3) into the receiver of the fluid coupler, (4) across the plug, (5) across the valve, and (6) into the sample coil. The flow is created by the pumping action of the OsmoSamplers, and the outflow is at 7. (ether ether ketone) (PEEK)). The tubing runs up the inside of the probe shaft and connects each port to a keyed port on a zero dead-volume PEEK coupler. Zero-Dead-Volume Fluid Coupler. A 15 cm tall, singlepin, eight-conduit fluid coupler with a near zero fluid dead zone was designed to allow connection of the removable sampler package to the tubes from the sampler ports (Figures 2-4). This coupler, machined from PEEK, consists of a plug and receiver with 10-15 µm tolerances and radially aligned ports for each fluid conduit. O-ring seals separate adjacent conduits. The receiver is mounted on a plate on top of the steel probe shaft, and the plug is mounted in the instrument sampler package. A system of tapered guides and elastomer spring relieved mating surfaces transforms crude placement of the removable sampler package by a remotely operated vehicle (ROV) into a precise mating of the fluid coupler. Interchangeable Instrument Package. The instrument package is a 45 (H) × 40 (L) × 60 (W) cm fiberglass box that contains the samplers, coils, valve, and coupler socket insert (Figure 2). It can be retrieved and replaced by submersibles. Pore-Fluid Samplers. The OsmoSamplers are illustrated in Figures 3 and 4. These osmotically powered pumps were first designed for subseafloor borehole sampling (17) and later used for sampling submarine vents (20-22) and estuarine waters (19). With slight modifications to the original construction, the samplers were constructed out of 4 in. diameter acrylic tubing with several chambers separated by PVC disks. One disk is mounted with eight osmotic membranes (2ML1, Durect Co.) that separate a saturated salt solution chamber from a deionized (DI) water chamber. This arrangement provides an osmotic pressure differential that creates a slow but continuous flow across the membranes. The instruments require no power, have no moving parts, and are therefore ideal for long-term, deep-sea deployment. Pumping rates were determined in the laboratory prior to 7370
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
deployment at in situ temperatures (5 °C) to be 0.35 ( 0.05 mL/day, which corresponds to ∼90 cm of tubing/day. During the PFA deployment, only four of the eight sample ports were plumbed to OsmoSamplers. Two failed to pump due to leaks across the membranes, probably caused by rupture during deployment or retrieval. The successful samplers collected from the overlying water (OLW) and 1.3 m below seafloor (1.3 mbsf). Sample Coils. Each OsmoSampler pump was connected to a long coil of tubing that stored the pore-water samples (Figure 3). Sample coils were made out of small-diameter (0.159 cm outer diameter) PEEK tubing for the OLW coil and copper tubing for the 1.3 mbsf coil, both rated to pressures of 14 MPa or 200 atm. The high pressure rating is needed to contain the samples at in situ pressures, allowing for in situ methane concentration measurements and exceeding the maximum hydrostatic pressure of 91 atm on the seafloor at MC 118. We chose copper for the pore-fluid measurements to minimize diffusive losses of gases through the tubing walls. Within the narrow-diameter tubing, sample dispersion is minimized due to slow pumping rates and slow diffusion, as discussed in the original sampler application (19). For the OLW sampler, the coil was 85 m long, which, at the measured pumping rate, corresponded to 112 days of sampling time. The 1.3 mbsf coil was 123 m long, corresponding to approximately 170 days of pore-fluid collection. Prior to deployment, the coils were filled with DI water that was first bubbled with nitrogen gas to eliminate methane and then degassed by being boiled for 5 min. Each coil was then gravity filled with DI water before being connected to the sampler. High-Pressure Valve. The sample coils for the OLW and 1.3 mbsf depths were plumbed to a high-pressure valve to avoid losing sample out the tubing ends (Figures 2 and 3). The valve (Valco 12-port, Hastelloy-C) was housed on the side of the instrument package with the valve handle accessible to submersibles or ROVs. Prior to deployment, the valve was in the open position so that the coils were in line with the OsmoSamplers. After deployment and prior to retrieval, the valve was closed on the seafloor by the submersible to isolate the coil and maintain in situ pressures. Sample Flow. The arrangement of the OsmoSamplers, coils, and valve creates a slow but continuous flow across the sampler membranes that draws filtered pore fluid through the sample ports, along the tubing enclosed within the probe shaft, across the fluid coupler and high-pressure valve, and, finally, into the long coils of the gastight tubing (Figure 3). Deployment and Recovery. In May 2005, the fully assembled PFA was deployed from the R/V Pelican and implanted into the sediments of the northwest area of MC 118, about 600 m north (28°,51.4682′ N and 88°,51.4969′ W) of the Sleeping Dragon hydrate outcrop (Figure 1). It was recovered in September 2006 by the submersible Johnson Sea-Link. The PFA was oriented in a fully upright position, protruding out of the sediments approximately 2 m. The instrument package showed minimal biofouling, covered only with hydroids. Before removal of the package from the PFA, the high-pressure valve was closed to seal the sampling tubing. During ascent through the water column, there was no visual indication of sample degassing. On a subsequent dive during the same cruise, a replacement instrument package was successfully mated onto the fluid coupler and PFA probe. This package will be retrieved in the future. It should be noted that due to unavoidable weather problems (Hurricanes Katrina and Rita), the samplers were on the seafloor for 480 days, 3 times longer than coils of the 170 day sampling time allowed by the sample coil length. While this delay could have resulted in a decrease in the pumping rate over time, laboratory tests after deployment showed that the samplers were still pumping at the original rate. Therefore,
FIGURE 4. Photographs of (A) the PFA ready for deployment off the R/V Pelican, (B) the fluid coupler with orange O-rings visible on the plug, and (C) an OsmoSampler with the saturated salt solution, membranes, and deionized water reservoir visible. the time period sampled corresponded to the last 170 days of the deployment time. Subsampling Tubing Coils. The tubing coils were subsampled using a custom-made fraction collector. The collector was connected to the front end of the sample coil and pressurized at the back end. Helium (20 psi, ∼1.25 mL/min) was introduced to displace pore fluids through the coil and into a constant-volume sample loop (3.85 m, 2 mL volume). The loop was attached to a four-port high-pressure switching valve that was also plumbed to a 10 mL gastight syringe and an outlet. Helium flowed until the pore fluids were expelled from the outlet and ∼1.5 mL was collected into a microcentrifuge tube for sulfate and chloride concentrations. The valve was then closed and a 2 mL sample, contained within the loop, extracted into the gastight syringe. Along with the sample, the syringe was also filled with ∼8 mL of helium, and the contents were injected into an evacuated 10 mL gas serum vial and measured for methane concentrations and stable isotope composition. Subsampling then continued for the length of the tubing and for each coil. The subsamples were time-stamped on the basis of the volume collected and the original in situ flow rates. Laboratory experiments were conducted to determine the affect of sample dispersion along the coil during sample extraction. Using high-pressure liquid chromatography and the 3.85 m sample loop, a dye plug was injected and run through the entire 123 m of sample coil. The plug was detectable over 14 m but maintained a sharp, discrete peak, indicating sample integrity. As this deployment was the first of its kind, the raw data set is given and no corrections for dispersion are made. Analytical Methods. Pore-fluid sulfate and chloride concentrations were determined by diluting the samples 1:100 with buffer eluent and injecting them onto a 2010i Dionex ion chromatograph. Methane concentrations were quantified by injecting a 5 mL headspace aliquot into a Shimadzu Mini II gas chromatograph equipped with a PoroPLOT Q column and a flame ionization detector. Certified gas standards served to calibrate the methane. The precision for replicate measurements of single samples was (3% for sulfate, chloride, and methane. Stable carbon and deuterium isotope ratios were obtained with a 5890 HewlettPackard gas chromatograph, coupled to a Finnigan MAT 252 isotope ratio mass spectrometer. The results are reported as δ13C or δD (‰) ) (Rsample/RPDB standard - 1) × 1000
(1)
where R is the ratio of the heavy to light isotope, either 13C/ 12C or D/H.
Results and Discussion PFA Design Assessment. The PFA instrument package was assessed immediately after seafloor retrieval. Visually, the pumps and coils appeared in good condition, with no cracks or leaks, and the salt reservoirs were still supersaturated. However, the high-pressure valve had corroded due to the extended deployment on the seafloor. It is therefore possible that the samples had partially degassed upon ascent and the measured methane concentrations might thus be minimum values. However, the submersible cameras were focused on the sampler box during ascent, and there was no visual indication of degassing. Furthermore, immediately upon retrieval on the ship, the coils were disconnected from the OsmoSampler pumps, outfitted with zero-dead-volume Swagelok valves to maintain the pressure and frozen to minimize microbial activity. Therefore, we assume the copper coils contained the samples and the dissolved methane concentrations measured are still in situ values. As will be shown below, PFA methane concentrations were within the range of other pore-fluid concentrations measured with different in situ pressurized probes (23), which supports the assumption that the samples did not degas, even though the valve failed. An improved PFA design contains the valve within a sealed, oil-filled box to minimize corrosion and ensure pressure retention. Another problem encountered was that high salinities were measured in the two deepest PFA sample depths, possibly because the osmotic membranes ruptured during deployment. These samplers were excluded from the results discussed here. We conclude that the PFA, being deployed similarly to a gravity corer and outfitted with the fluid coupler for sampler package retrieval and replacement, is a promising tool for determining the long-term stability of gas hydrate. Dissolved Ion and Gas Concentrations. Pore fluids were collected at two depths, from OLW and 1.3 mbsf, and represent the last 170 days of the 480 day deployment. There was little temporal change in ion concentrations in the OLW coil (Figure 5A). Chloride (Cl-) concentrations averaged 531 ( 43 mM, and sulfate (SO42-) concentrations averaged 29.6 ( 2.3 mM; the results are shown as Cl- to SO42- ratios which averaged 18.0 ( 0.9, similar to normal overlying water averaging 18.3 (Figure 5A). Chloride concentrations in pore fluids from the 1.3 mbsf sample coil averaged 560 ( 42 mM, and SO42- concentrations VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7371
FIGURE 6. CD diagram for classification of biogenic and thermogenic methane sources, redrawn from ref 26. The PFA, GB (Guaymas Basin (27)), and EPR (East Pacific Rise (27)) are shown. ATM ) atmospheric methane. FIGURE 5. PFA data. (A) Chloride (Cl-) to sulfate (SO42-) concentration ratio (Cl-:SO42-) from OLW (open squares) and 1.3 mbsf (closed squares). The dashed line is the typical seawater Cl-:SO42-. (B) Methane (CH4) concentrations from OLW. (C) Methane concentrations from 1.3 mbsf. (D) CH4 and dissolved inorganic carbon (DIC) stable carbon isotope ratios (δ13C). Each point includes 3% analytical error bars. Black arrows along the x axis indicate the timing of two earthquakes. averaged 19.3 ( 2 mM. The Cl-:SO42- ratios averaged 29.7 ( 2.4 (Figure 5A). The higher Cl-:SO42- values suggest a net sink for sulfate, due to either microbial sulfate reduction or upward advection of SO42--depleted brines. The microbial explanation is supported by previous studies near the PFA location measuring sufficient sulfate reduction activity to produce the observed concentration gradients (15). The lack of significant chloride or sulfate anomalies over time suggests that the PFA did not capture any hydrate formation or decomposition events. In the OLW, methane concentrations were approximately 0.004 mM at the beginning of the time series (t ) 110 days) and increased to 0.1 mM at the end (Figure 5B). These concentrations are high for seawater, which are typically in the nanomolar range (9). Pore-fluid methane concentrations measured at 1.3 mbsf were much higher, ranging between 0.7 and 14 mM but averaging 4.2 ( 2.1 mM throughout the deployment (Figure 5C). The highest concentration measured was a 14 mM spike that occurred at t ) 0 (Figure 5C). The methane spike could have been caused by local tectonic activity. The USGS earthquake search record (http:// neic.usgs.gov/neis/epic/epic_circ.html) reveals that two earthquakes occurred during the PFA deployment (Figures 1 and 5C). Three days prior to retrieval of the PFA, an Mw ) 5.8 earthquake occurred in the Gulf, centered 338 km southwest of MC 118. The timing of this quake coincided with the highest methane concentrations measured in both OLW and 1.3 mbsf coils (Figure 5C). About 120 days before retrieval, there was a small methane increase (Figure 5c) that appears to have occurred several weeks before a smaller earthquake (Mw ) 2.9), centered in Alabama. Previous studies linking tectonic activity and methane release have reported methane concentrations in fluid releases as high as 200 nM, 4 orders of magnitude lower than measured in the overlying water column with the PFA (9). With the seismic data available for the current study, it is impossible to prove a direct causal 7372
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
link between the two quakes and our PFA data. However, the methane spikes we recorded are coincident in time with these quakes. This coincidence is consistent with the hypothesis that temporal changes in pore-fluid chemistry are linked to earthquakes or other local tectonic events (such as salt movement). Future PFA results will be strengthened by coupling the PFA data to records from nearby geophysical arrays, which will provide an extraordinary amount of information about the long-term stability of the MC 118 gas hydrate system. While methane concentrations in the pore fluids were higher than in the overlying water, they were still undersaturated with respect to methane. At the in situ pressure, temperature, and salinity conditions at MC 118, pore fluids are in chemical equilibrium with hydrate when pore fluids contain 70 mM CH4 (24), yet the highest concentration measured in our study was 14 mM. This contrast between theoretical and measured concentrations suggests that hydrates should not have been stable where the PFA was deployed, which is consistent with the lack of outcropping hydrates observed. Additional data from the PFA and the installment of several more PFAs to provide spatial coverage will aid in constraining the geochemical role in hydrate stability and understanding the flux of methane from the sediments. Methane Sources. Sources of the dissolved methane were evaluated using stable carbon and hydrogen isotope ratios (25). The average δ13C-CH4 value in the 1.3 mbsf samples was -32.4 ( 3.4‰ with three very distinct 13C-enriched spikes, with a change in δ13C as high as 10‰ (Figure 5D). This temporal variation in CH4 isotopes could be caused by changes in the relative importance of biogenic and thermogenic sources of gas because thermogenic methane is isotopically more enriched in 13C compared to biogenic methane. The methane stable hydrogen isotope composition was utilized to further differentiate potential sources. The average δD-CH4 value was -136 ( 12‰, and the temporal trend was not highly significant (R2 ) 0.4, data not shown). When plotted on a carbon-hydrogen (C-H) diagram (26), the dissolved methane falls well within the thermogenic region (Figure 6), similar to values from other thermogenic basins (27). The dissolved methane measured at MC 118 appears thermogenic, yet the three large isotopic spikes could result from an input of residual 13C-enriched methane resulting from microbial oxidation. As microbes oxidize and produce 13C-enriched methane, DIC, the product of oxidation,
becomes 13C-depleted (26, 28). Such a coupling between the CH4 and DIC pools has been repeatedly shown in coastal anoxic sediments (29). In the PFA pore fluids, the δ13C-DIC values averaged -13.5 ( 0.9‰ and did not track changes in CH4 isotopes (Figure 5D). Therefore, microbial oxidation does not appear to explain the large isotope spikes. It is possible that the spikes reflect differences in isotopic composition among thermogenic methane sources. Such methane source variability could be controlled by differing the venting rates due to local seismic activity, such as the earthquakes that occurred during the sampling period. If venting rates do change over time, the resulting convection cells could introduce seawater-derived fluids to the sediments and affect the pore-fluid chemistry (30). Future results with the PFA will help understand such processes. Comprehensive spatial and temporal measurements using tools such as the PFA combined with local seismic recorders will be needed to link the seismic activity to variability in pore-fluid composition and thus to hydrate stability in continental margin sediments.
(11)
(12)
(13) (14)
(15)
(16)
Acknowledgments We thank Jean Whelan, the University of Mississippi Machine Shop, Jim Gambony, Don Brewer, Sam Perkins, and Claire Langford for technical assistance, the R/V Pelican captain and crew, and Howard Mendlovitz and Jose´ Mauro Sousa de Moura for providing the hydrogen isotope analysis. We also thank John Crittenden and three anonymous reviewers for their hard work that greatly improved this manuscript. This study was funded by an EPA STAR fellowship (L.L.L.) and the Gulf of Mexico Hydrate Research Consortium with grants from the DOE (DE-FC26-02NT41628), NIUST/STRC (NA16RU1496), and Minerals Management Services (1435-01-02-CA-85273).
(17)
(18) (19)
Supporting Information Available Table giving the pore-water chemistry from OLW and 1.3 mbsf. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Sloan, E. D. Clathrate Hydrates of Natural Gases, 2nd ed.; Marcel Dekker, Inc.: New York, 1998. (2) Sassen, R.; Roberts, H. H.; Carney, R.; Milkov, A. V.; DeFreitas, D. A.; Lanoil, B.; Zhang, C. Free hydrocarbon gas, gas hydrate, and authigenic minerals in chemosynthetic communties of the northern Gulf of Mexico continental slope: Relation to microbial processes. Chem. Geol. 2004, 205, 195–217. (3) Milkov, A. V. Global estimates of hydrate-bound gas in marine sediments: How much is really out there. Earth Sci. Rev. 2004, 66, 183–197. (4) Lelieveld, J.; Crutzen, P. J.; Dentener, F. J. Changing concentration, lifetime and climate forcing of atmospheric methane. Tellus 1998, 50B, 128–150. (5) MacDonald, I. R.; Leifer, I.; Sassen, R.; Stine, P.; Mitchell, R.; Guinasso, N. L., Jr. Transfer of hydrocarbons from natural seeps to the water column and atmosphere. Geofluids 2002, 2. (6) Kennett, J. P.; Cannariato, K. G.; Hendy, I. L.; Behl, R. J. Carbon isotopic evidence for methane hydrate instability during quaternary interstadials. Science 2000, 288, 128–133. (7) MacDonald, I. R., Jr.; Sassen, R.; Brooks, J. M.; Lee, L.; Scott, K. T. Gas hydrate that breaches the sea floor on the continental slope of the Gulf of Mexico. Geology 1994, 22, 699–702. (8) McGee, T.; Woolsey, J. R.; Lapham, L.; Kleinberg, R.; Macelloni, L.; Battista, B.; Knapp, C.; Caruso, S.; Goebel, V.; Chapman, R.; Gerstoft, P. Presented at the 6th International Conference on Gas Hydrates, Vancouver, British Columbia, 2008. (9) Mau, S.; Rehder, G.; Arroyo, I. G.; Gossler, J. Indications of a link between seismotectonics and CH4 release from seeps off Costa Rica. Geochem., Geophys., Geosyst. [Online] 2007, 8, DOI: 10.1029/2006GC001236. (10) Vardaro, M. R.; MacDonald, I. R.; Bender, L. C., Jr. Dynamic processes observed at a gas hydrate outcropping on the
(20) (21)
(22)
(23) (24)
(25) (26) (27) (28) (29) (30)
continental slope of the Gulf of Mexico. Geo-Mar. Lett. [Online] 2005, DOI: 10.1007/s00367-005-0010-2. Tryon, M. D.; Brown, K. M. Fluid and chemical cycling at Bush Hill: Implications for gas- and hydrate-rich environments. Geochem., Geophys., Geosyst. [Online] 2004, 5, Q12004, DOI: 12010.11029/12004GC000778. Solomon, E. A.; Kastner, M.; Jannasch, H.; Robertson, G.; Weinstein, Y. Dynammic fluid flow and chemical fluxes associated with a seafloor gas hydrate deposit on the northern Gulf of Mexico slope. Earth Planet. Sci. Lett. 2008, 270, 95–105. Kvenvolden, K. A. Gas hydratessgeological perspective and global change. Rev. Geophys. 1993, 31, 173–187. Ussler, W.; Paull, C. K. Ion exclusion associated with marine gas hydrate deposits. In Natural Gas Hydrates: Occurrence, Distribution, and Detection; Paull, C. K., Dillion, W. P. Eds.; American Geophysical Union: Washington, DC, 2001; p 315. Lapham, L. L.; Chanton, J. P.; Martens, C. S.; Sleeper, K.; Woolsey, J. R. Microbial activity in surficial sediments overlying acoustic wipe-out zones at a Gulf of Mexico cold seep. Geochem., Geophys., Geosyst. [Online] 2008, 9, Q06001, DOI: 06010.01029/ 02008GC001944. Fisher, A. T.; Wheat, C. G.; Becker, K.; Davis, E. E.; Jannasch, H. W.; Schroeder, D.; Dixon, R.; Pettigrew, T.; Meldrum, R.; MacDonald, R.; Nielsen, M.; Fisk, M.; Cowen, J.; Bach, W.; Edwards, K. Scientific and Technical Design and Deployment of Long-Term, Subseafloor Observatories for Hydrogeologic and Related Experiments, IODP Expedition 301, Eastern Flank of Juan de Fuca Ridge; Integrated Ocean Drilling Program Management International, Inc: College Station, TX, 2005 doi: 10.2204/iodp.proc.301.103.2005. Jannasch, H. W.; Davis, E.; Kastner, M.; Morris, J.; Pettigrew, T.; Plant, J. N.; Solomon, E.; Villinger, H.; Wheat, C. G. CORK II: Long-term monitoring of fluid chemistry and hydrology instrumented boreholes at the Costa Rica subduction zone. In Proceedings of the Ocean Drilling Program, Initial Reports; Morris, Klaus Eds.; Integrated Ocean Drilling Program Management International, Inc: College Station, TX, 2003; Vol. 205. Moran, K.; Farrington, S.; Massion, E.; Paull, C.; Stephen, R.; Trehu, A.; Ussler, W. SCIMPI: A new seafloor observatory system. Oceans 2006, 1–6. Jannasch, H. W.; Wheat, C. G.; Plant, J. N.; Kastner, M.; Stakes, D. S. Continuous chemical monitoring with osmotically pumped water samplers: OsmoSampler design and applications. Limnol. Oceanogr.: Methods 2004, 2, 102–113. Kastner, M.; Jannasch, H. W.; Weinstein, Y.; Martin, J. Presented at the Oceans 2000 MTS/IEEE Conference and Exhibition, Providence, RI, 2001. Wheat, C. G.; Jannasch, H. W.; Plant, J. N.; Moyer, C. L.; Sansone, F. J.; McMurtry, G. M. Continuous sampling of hydrothermal fluids from Loihi Seamount after the 1996 event. J. Geophys. Res. 2000, 105, 19,353–319,367. Solomon, E.; Kastner, M.; Jannasch, H. W.; Weinstein, Y.; Robertson, G.; Aubrey, A. Long-term continuous monitoring of fluid chemistry and flux at the Bush Hill gas hydrate field, Gulf of Mexico using a new flow meter, the MOSQUITO. Eos Trans. AGU, Fall Meet. Suppl. 2004, 85, Abstr. T21C-0539. Lapham, L. L. Ph.D. Thesis, Department of Marine Sciences, University of North Carolina, Chapel Hill, NC, 2007; p 391. Duan, Z.; Mao, S. A thermodynamic model for calculating methane solubility, density, and gas phase composition of methane-bearing aqueous fluids from 273 to 523 K and from 1 to 2000 bar. Geochim. Cosmochim. Acta 2006, 70, 3369–3386. Bernard, B. B.; Brooks, J. M.; Sackett, W. M. Light hydrocarbons in recent Texas contintental shelf and slope sediments. J. Geophys. Res. 1978, 83, 4053–4061. Whiticar, M. J. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chem. Geol. 1999, 161, 291–314. Welhan, J. A. Origins of methane in hydrothermal systems. Chem. Geol. 1988, 71, 183–198. Alperin, M. J.; Reeburgh, W. S.; Whiticar, M. J. Carbon and hydrogen isotope fractionation resulting from anaerobic methane oxidation. Global Biogeochem. Cycles 1988, 2, 279–288. Martens, C. S.; Blair, N. E.; Green, C. D.; Des Marais, D. J. Seasonal variations in the stable carbon isotopic signature of biogenic methane in a coastal sediment. Science 1986, 233, 1300–1303. Aloisi, G.; Wallmann, K.; Haese, R. R.; Salie`ge, J.-F. Chemical, biological and hydrological controls on the 14C content of cold seep carbonate crusts:numberical modeling and implications for convection at cold seeps. Chem. Geol. 2004, 213, 359–383.
ES801195M VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7373
