Methane Flux and Authigenic Carbonate in Shallow Sediments Overlying Methane Hydrate Bearing Strata in Alaminos Canyon, Gulf of Mexico

In June 2007 sediment cores were collected in Alaminos Canyon, Gulf of Mexico across a series of seismic data profiles indicating rapid transitions between the presence of methane hydrates and vertical gas flux. Vertical profiles of dissolved sulfate, chloride, calcium, magnesium, and dissolved inorganic carbon (DIC) concentrations in porewaters, headspace methane, and solid phase carbonate concentrations were measured at each core location to investigate the cycling of methane-derived carbon in shallow sediments overlying the hydrate bearing strata. When integrated with stable carbon isotope ratios of DIC, geochemical results suggest a significant fraction of the methane flux at this site is cycled into the inorganic carbon pool. The incorporation of methane-derived carbon into dissolved and solid inorganic carbon phases represents a significant sink in local carbon cycling and plays a role in regulating the flux of methane to the overlying water column at Alaminos Canyon. Targeted, high-resolution geochemical characterization of the biogeochemical cycling of methane-derived carbon in shallow sediments overlying hydrate bearing strata like those in Alaminos Canyon is critical to quantifying methane flux and estimating methane hydrate distributions in gas hydrate bearing marine sediments.


Introduction
Natural gas hydrates are crystalline clathrate compounds consisting of hydrocarbon guest molecules such as methane (CH4) within a solid water lattice.The largest accumulations of natural gas hydrates occur in offshore and continental (active and passive) margin sediments and, to a lesser extent, in permafrost, where high pressure, low temperatures, and methane concentrations present in excess of solubility promote the formation and stability of solid phase clathrates.Methane hydrates constitute a major organic carbon sink and a vast potential energy source [1].Methane hydrates are also important to global climate and coastal slope stability.
Substantial research has focused on carbon cycling in hydrate bearing sediments.The presence of shallow gas hydrates is normally inferred through geochemical [2,3] and geophysical [4] data interpretation as well as the existence of unique biological communities [5,6].No one line of evidence has proven to be a unique indicator of the presence of shallow gas hydrates [7][8][9].
Authigenic carbonates have been shown to occur in sediments containing CH4 hydrates and in sediments at or near the seafloor over CH4 hydrate deposits [10][11][12][13][14][15][16][17][18][19][20][21].Carbonate structures and concretions can form on the seafloor in areas with a significant CH4 flux [11,14,16,21], carbonate horizons can be formed in the sediments as a result of microbially-mediated oxidation of CH4 [13,19], and authigenic carbonates are often associated with faults that act as conduits for the upward migration of fluids and CH4 [10].Carbonate is an essential base for establishing benthic communities supported by the biogeochemical cycling of CH4 in sediments and bottom waters at cold seeps and CH4 hydrate deposits [6].Carbonate in hydrate bearing sediments can be biogenic and/or authigenic carbonate derived from dissolved inorganic carbon (DIC) in seawater or from DIC generated during the oxidation of organic matter, methane, or non-methane hydrocarbons [17].Variations in the composition of carbonate minerals in the sediments can be due to past changes in the stability limits of gas-hydrate host deposits and can provide a past record of CH4 hydrate destabilization [10][11][12][13]16,20,21].
There exists a close relationship between gas hydrates, CH4 flux, the oxidation of organic matter through sulfate reduction (SR), the anaerobic oxidation of methane (AOM), and authigenic carbonate mineralization in marine sediments containing or overlying CH4 hydrate.High concentrations of authigenic carbonate do potentially indicate elevated rates of AOM, and can provide a record of such processes [12,13,[16][17][18]21].The precipitation of authigenic minerals (aragonite, Mg-calcite, dolomite) in such areas represents a potentially significant carbon sink and plays a role in regulating the flux of methane to the overlying water column and atmosphere [22].
In cold-seep and gas-hydrate influenced sediments, authigenic minerals (aragonite, Mg-calcite, dolomite) can precipitate from the oxidation of methane-rich fluids.Carbonate precipitates as a secondary reaction of the oxidation of organic matter by sulfate reduction (SR) and/or AOM.The net reactions for SR and AOM are: CH4 + SO4 2− → HCO3 − + HS − + H2O Anaerobic Oxidation of Methane (AOM) (2) Whereas aerobic oxidation promotes the dissolution of carbonate in surficial sediments through the lowering of porewater pH, the coupled effect of the microbially-mediated diagenetic reactions shown above, SR and AOM, increase porewater alkalinity (generate bicarbonate).In the presence of seawater-derived cations (Ca 2+ , Mg 2+ ) this can lead to precipitation of authigenic carbonates [23][24][25], shown by the net reaction below: In sediments overlying suspected hydrate deposits, high concentrations of authigenic carbonate can indicate elevated rates of AOM.Precipitation can occur at or around the sulfate-methane interface (SMI) or through a sulfate methane transition (SMT).The SMT is a border between sulfate-bearing sediment above and sulfate-depleted, methane-rich sediments below.The SMT is a zone of intense methane oxidation [2,24].Carbonate precipitation associated with methane oxidation commonly forms discrete mm size concretions in hydrate bearing sediments [20,26].
Temperature, the level of porewater saturation of HCO3 − and cations such as Ca 2+ and Mg 2+ , and sediment redox state are major factors influencing the precipitation of different carbonate minerals.Physical and chemical characteristics of sediments containing hydrates provide distinct diagenetic environments that can promote precipitation and preservation of carbonate minerals.Chemical controls on the formation of specific carbonate species are difficult to evaluate [13,27].Because of this difficulty and the fact that gas hydrates are not generally preserved in conventional core samples, it is necessary to find diagenetic proxies to quantify biogeochemical processes and identify sediments that formerly contained gas hydrate.
Although the interstitial environment controls specific carbonate diagenetic processes, carbonate dissolution or precipitation will produce notable changes in sediment porewater profiles of Ca 2+ , Mg 2+ , and Sr 2+ [13,16,[20][21][22][23][24].Porewaters modified by carbonate diagenesis may be characterized by the direction and gradient of the ratios of these constituents [13,20,21,23,26].Steep vertical and horizontal gradients of Ca 2+ , Mg 2+ , and Sr 2+ can develop on fine scales (centimeters to decimeters) in porewater and sediment of methane bearing strata as a secondary consequence of diffusion, fluid advection, methane supply, and AOM [13,20,21,23,26].Precipitates of authigenic minerals mark areas of fluid flow, and are the result of biogeochemical processes and interaction (mixing) of porewater fluids and ambient seawater [12].Authigenic carbonates can provide and integrated record of such processes [17].
Directly measuring methane fluxes in sediment porewaters is problematic since samples recovered from depth depressurize quickly leading to a change in the partial pressure for methane gas solubility in seawater.Porewater headspace CH4 measurements on samples collected from depressurized sediments, however, do provide at least a relative assessment of spatial variation in methane concentrations.Other biogeochemical indicators exist to aid in quantification and qualification of methane fluxes and methane-derived carbon cycling between organic and inorganic phases in sediments overlying hydrate bearing strata.
Destabilization of gas hydrates leads to a freshening of porewaters that can be seen in Cl − profiles [28].Seawater contains large amounts of sulfate (SO4 2− ), which diffuses downward into pore waters and contributes to diagenesis [2].Under anoxic conditions, SO4 2− depletes as depth increases in sediment due to organoclastic SR of organic matter by microbial activity.Sulfate can also be reduced through AOM.If AOM is the dominant process consuming SO4 2− , usually due to high methane concentrations deeper in the sediments, SO4 2− diffusion into the sediments is inversely related to upward CH4 flux with a 1:1 stoichiometry.Assuming steady-state conditions, the slope of sediment porewater SO4 2− concentration profiles can therefore be used to interpret upward CH4 flux from below.Nonlinearity in the slope of sediment porewater SO4 2− concentration profiles could result from sulfate consumption due to sulfate reduction, sulfate gradient instability, fluid advection, bioturbation, and input of organic matter from sedimentation [3].
The stable carbon isotope ratio, or ratio of stable heavy carbon ( 13 C) to light carbon ( 12 C) of dissolved or gaseous methane (δ 13 C-CH4), in porewaters can be used to distinguish between biogenic and thermogenic sources.Dissolved inorganic carbon (DIC) profiles and δ 13 C-DIC values will also lend information on methane flux and AOM rates [20][21][22][23][24]29].Carbonates formed as a secondary consequence of AOM will have a distinct δ 13 C signature indicative of the parent carbon source.
Dissolved inorganic carbon is the sum of all dissolved inorganic carbon species and is dominated by bicarbonate and carbonate (HCO3 − and CO3 2− ), which both contribute to alkalinity.Comparing alkalinity profiles to the sulfate profiles may help determine if SR or AOM dominates the sulfate depletion process in shallow sediment porewaters.Microbial and thermogenic methane production contribute to the DIC concentration in the form of HCO3 − [20].Calcium (Ca 2+ ) and Mg 2+ both chemically precipitate HCO3 − in the form of marine carbonates.Therefore, the net DIC flux can be calculated by subtracting the Ca 2+ and Mg 2+ flux from the DIC flux estimated from measured porewater concentration values [20,21,29].
If AOM dominates, the net DIC flux should compare to the sulfate flux with a 1:1 ratio above the SMT.If SR dominates, the ratio will approach 2:1.Dissolved inorganic carbon concentrations and δ 13 C may also be used to evaluate area of active methane oxidation release since they alter in response to the presence of methane [30].However, DIC levels are not an absolute confirmation of methane oxidation as increases in bicarbonate can also result from fermentation, or reduction of solid organic matter.A decrease may be attributed to the mineralization of authigenic carbonate [20].Integration of geochemical measurements in sediments and interstitial porewater fluids allows for resolution of the interaction between fluid flow and microbially-mediated diagenic processes in the cycling of methane-derived carbon in shallow sediments overlying methane hydrate bearing strata [20,21,29].
The Gulf of Mexico (GoM) contains both microbial and thermogenic methane hydrates, distinguishable through their carbon isotopic composition (δ 13 C-CH4) [31].Originating from below the bottom of the gas hydrate stability zone (GHSZ), thermogenic methane migrates upward via channels and faults until it can combine with water to become hydrate.The nature of thermogenic methane formation allows for more localized distribution, maturity, and accumulation in massive amounts [32].Because of the presence of fault associated conduits in the GoM thermogenic methane hydrate is common, especially in localized areas [28].Biogenic methane is formed at much shallower depths in sediments through the oxidation of organic matter by bacteria under anoxic conditions both below and within and above the GHSZ.Biogenic methane hydrate deposits are also common in certain areas of the GoM but tend to be less localized and occur at lower concentrations [28,32].Coupling geophysical and geochemical data from shallow sediment to infer methane flux is a method that can be employed to help interpret the location, source, and quantity of methane hydrate reservoirs for potential energy use and climate impact analysis [3].In June 2007, a targeted transect of shallow sediment cores was collected from the seafloor of the Alaminos Canyon region of the GoM in an area where previous seismic surveys indicated rapid transitions between the presence of methane hydrates and vertical gas flux.Geochemical characterization of sediments and porewaters collected were used to show that a significant fraction of the vertical methane flux is cycled into inorganic carbon and that incorporation of methane-derived carbon into dissolved and solid inorganic carbon phases represents a significant carbon sink regulating the flux of methane to the overlying water column and atmosphere at Alaminos Canyon and potentially other site with gas hydrate bearing marine sediments.

Study Area
The Alaminos Canyon (AC) is a deep water canyon (~1000-3000 m) located in the northwestern Gulf of Mexico (Figure 1).The geology of the area is dominated by northeast-southwest trending salt-cored box folds of the Perdido fold belt which lie beneath a thick, mobile salt canopy [33,34].
Research by others has identified Block 818 in AC as a site of significant hydrocarbon flux and oil, gas, and gas hydrate accumulation [33][34][35][36][37][38][39] in a variety of turbidite deposits from sand sheets to amalgamated and leveed channel systems [36].Boswell et al. [34] used well data and 3-D seismic data to show evidence for significant, concentrated gas hydrate accumulation near the AC 818 #1 ("Tigershark") well.Uplift of the Perdido fold belt has raised gas reservoirs in Oligocene Frio sands within the Lower Tertiary deepwater turbidite to shallow depths above the base of the gas hydrate stability (BGHS) zone.The Oligocene Frio sand is very fine-grained, immature volcaniclastic sand with a high porosity [34].
In June 2007, the Gulf of Mexico Gas Hydrate Joint Industry Project (JIP) teamed with the U.S. Naval Research Laboratory (NRL) and the Seep and Methane Hydrate Advanced Research Initiative to conduct fieldwork in AC Block 818 [39].The goal of the NRL-led AC-07 research expedition was to interpret seismic profiles and collect geochemical data to conduct preliminary site characterization of deep sediment hydrate deposits within the region for use in JIP deep drilling efforts.Field sampling and data collection locations for the AC-07 research expedition were based on seismic profiles provide by WesternGeco and reviewed by geophysists and geochemists from WesternGeco, NRL, the U.S. Department of Energy, National Energy Technology Laboratory (NETL) and the U.S. Mineral Management Service (MMS).Seismic reflection maps (WesternGeco) display subtle differences in reflection amplitude created by high levels of sand layering below a 10 m thick pelagic drape and show a focused location of vertical gas migration near an existing well.The seeps are located along a small ridge associated with the up-thrown side of a fault.No gas chimneys are visible in 3-D seismic data.Well data suggests a thick 5 to 50 m of hydrate laden, sandy sediment [39].

Sample Collection
In June 2007, a series of 10 piston cores (PC-04, PC-05, PC-06, PC-07, PC-08, PC-08, PC-14, PC-15, PC-16, PC-21) were collected from R/V Cape Hatteras along WesternGeco inline 986 in AC Block 818 (Figure 2) Seismic data on inline 986 revealed an area of potential gas hydrate accumulation overlying free gas accumulation marked by prominent bottom simulating reflectors (BSRs).The seismic data also suggested possible gas venting through a small seep feature on the seafloor.Piston cores were collected along a roughly 3 km linear transect across the suspected seep feature [39].
Sediment cores were processed using the methods detailed in Coffin et al. [3,39].The sediment cores were collected using a 10 m piston coring system (Milbar Hydro-Test, Inc., Shreveport, LA, USA) that consisted of 2-3 N-80 alloy core barrels lined with 7 cm outer diameter polycarbonate barrels connected by a modified Atlas Bradford connections system.The core weight used was approximately 1400 kg.Trigger weights were set initially at 12-15 m.Typical core penetration depths were between 3 m and 8 m.An NRL Sonardyne Ultra-Short BaseLine (USBL) positioning and tracking system attached 50 m above the core head was used to provide acoustic positioning on all but one piston core deployments.This instrument provided improved accuracy in positioning but deep currents were still responsible for drifts of up to 130 m from the ship's position on the surface.The ship was able to maneuver within a range of 10-40 m from the acoustic positioning.

Sample Processing
Immediately after retrieval, core liners were removed, placed horizontally on deck, and inspected for gas pockets and gas expansion voids.At void spaces, the liner was drilled and gas sampled with a 60 mL polypropylene syringe fitted with a modified 3-way stopcock.Gas samples were then transferred to 30 mL pre-evacuated, glass serum vials fitted with a gastight stopper and aluminum seal for subsequent analysis to determine relative light hydrocarbon concentrations and isotopic ratios.Cores were cut in 10 cm round sections at 25 to 45 cm intervals.The sampling interval was adjusted based upon observations of dark (black) sediment and hydrogen sulfide odor and the appearance of core gas pockets.On average, 20 sediment sections were sampled from each core.
Sediment plugs were immediately collected from each core round after sectioning using a 3 mL polypropylene syringe with the tip cut off.Sediment plugs were transferred to pre-weighed 20 mL serum vials and capped with gastight stoppers and aluminum seals to determine sediment headspace light hydrocarbon concentrations (CH4 through C3H8) [40] as well as δ 13 CCH4(g) ratios.Whole round core sections were then taken to the ship wet laboratory for processing.
Approximately 5 g of wet sediment was collected from each whole round section using a clean Al spatula and transferred into pre-weighed 31-mm snap-tight Petri-dishes.These sub-samples were frozen for use in laboratory measurements of sediment porosity and percent organic carbon.Immediately after the subsample collection, porewater was pressed from the remaining sediment from each round using Reeburgh-style PVC press containers pressurized to ~400 KPa (~60 psi) by low-pressure air applied to a latex sheet placed between the core sections and press gas inflow.Porewater was pre-filtered through Grade 1 Qualitative Filter Paper into gas-tight 60-mL polypropylene syringes and then again through 0.2-μm Acrodisc PES syringe filters (Pall) into ashed (4 h at 450 °C) 20 mL vials.Waters were then further distributed into 1-10 vials for each subsequent analysis and chemically fixed, if necessary to stop microbial activity [3,39].The remaining pressed sediment was wrapped in ashed aluminum foil, sealed in Whirlpack bags, stored frozen at −20 °C, and transported to the land-based laboratory for inorganic and organic carbon concentration and isotope analyses.

Sample and Data Analysis
Headspace methane (CH4) and porewater sulfate (SO4 2− ) and chloride (Cl − ) concentrations were determined onboard ship.Analyses of porewater dissolved inorganic carbon (DIC) concentrations, stable carbon isotope ratios (δ 13 C-DIC), and major cation concentrations (Ca 2+ , Mg 2+ ) were conducted at the NRL laboratory as was solid phase analysis for sediment porosity and carbonate content (CaCO3).
Volumetric CH4 gas concentrations were determined from the 3 mL sediment plugs using headspace techniques and were quantified against certified gas standards (Scott Gas) [40].Analysis was performed using a Shimadzu 14-A gas chromatograph (GC, Kyoto, Japan) equipped with a flame ionization detector and Hayesep-Q 80/100 column.Gases were separated isothermally at 50 °C.Final relative concentrations were calculated using sediment porosity and dry weight data obtained at the U.S. Naval Research Laboratory [40].True CH4 concentrations cannot be reliably measured from porewater samples because pressure reduction during core recovery lowers solubility and transfers dissolved CH4 to the gas phase.Hence, headspace CH4 data presented are used only to provide data comparisons with measured SO4 2− gradients.
Major cation concentrations (Ca 2+ , Mg 2+ ) in sediment porewaters were also measured using a Thermo-Fisher Dionex DX-120 ion chromatograph equipped with an AS-40 autosampler in the laboratory.A 20 mM methanesulfonic acid (CH3SO2OH) eluent was used with a CS-12 column and a Cation Self-Regenerating Suppressor (CSRS Ultra II) at a flow rate of ~0.7 mL/min.Samples again were diluted 1:50 (vol/vol).Calibration standards were prepared in the laboratory and diluted (1:50) IAPSO standard seawater was used as a reference standard.Analytical precision was ±2%-3%.
Porewater DIC concentrations were measured using a UIC CO2 coulometer and standardized to a certified seawater reference material (University of California, San Diego, CA, USA).The conversion of DIC to CO2 and separation from interfering sulfides was conducted according to Boehme et al. [41].
Sediment total carbon and OC (%TC, %SOC) concentrations and δ 13 C values were determined on a Fisons EA 1108 C/H/N analyzer in line with a Thermo Electron Delta Plus XP Isotope Ratio Mass Spectrometer (IRMS) interface via a Conflo II.Pressed sediment was dried at 80 °C, ground with a mortar and pestle, then 15 to 20 mg of sediment was weighed in tin capsules for TC analysis.For SOC analysis, sub-samples were weighed in silver capsules, treated with an excess of 10% HCl and dried in an oven at 70 °C overnight to remove inorganic carbon.A concentration calibration curve for carbon concentration analysis was generated daily by analyzing and acetanilide standard.Sediment carbonate concentrations (%CaCO3) concentrations were estimated by multiplying the difference between %TC and %SOC by the ratio of the molar mass of CaCO3 to carbon, assuming calcite and aragonite as the dominant carbonates: %CaCO3 = (%TC − %SOC) × 8.33 (4) Pore water dissolved inorganic carbon δ 13 C ratios (δ 13 C-DIC) and gas pocket and sediment δ 13 C-CH4 ratios were determined using a Thermo Electron Trace GC (Waltham, MA, USA) equipped with a Varian Porapak-Q column and GC-CIII combustion interface in-line with a Delta Plus XP IRMS [3,42,43].For δ 13 C-DIC analysis, 2 mL porewater samples were treated with 200 μL of 85% H3PO4.The CO2 was extracted from the vial headspace and injected into the GC via a split/splitless inlet in split mode.All δ 13 C-DIC values were normalized through analysis of CO2 and C1-C5 alkanes in NIST RM 8560 (natural gas, petroleum origin).Samples for δ 13 C-CH4 analysis were introduced via an in-line cryogenic focusing system according to the method of Plummer et al. [42].A separate δ 13 C normalization curve was generated for C1-C4 alkanes and used to normalize δ 13 C-CH4 data.Replicate δ 13 C-DIC values varied by less than 0.5‰, and δ 13 C-CH4 by less than 1.0‰ [40].Stable carbon isotope ratios are presented in per mil units compared to a PeeDee Belmenite standard.
Sediment porosity was estimated using the method described by Hoehler et al. [40].Frozen samples (~5-6 g) were thawed and allowed to equilibrate to room temperature.Samples were weighed wet and placed in a drying oven (~50-60 °C) for 24-48 h.Samples were then weighed again after drying.Sediment water content was determined by the difference between wet and dry weight, assuming constant pore water (ρpw) and bottom water (ρsw) density.Porosity (φ) was then determined using the following equation: where: assumed solid matter density (ρsm) = 2.50 g/cm 3 .

Sediment Headspace CH4 and Porewater SO4 2− and Cl − Concentrations
Figure 3 shows measured sediment headspace CH4 and porewater SO4 2− and Cl − concentrations vs. depth for the AC-07 piston cores.With the exception of core PC-06, measured porewater Cl − concentrations were consistently near seawater values (559 mM) in all piston cores.The Cl − profile in PC-06 clearly shows a freshening of porewaters due to the destabilization of solid phase hydrates during core collection.Consequently, measured headspace CH4 values in PC-06 ranged from 1.5 to 18.9 mM throughout the entire length of the core.Measured sediment headspace CH4 values were <0.5 mM in all other cores with the exceptions of PC-07, 14, and 21.In PC-07, 14, and 21, headspace CH4 concentrations were elevated at depth and increased linearly near the base of each core.
Porewater SO4 2− concentrations in all cores generally decreased with depth from near average seawater values (28.9 mM) at the surface (Figure 3).Most piston cores exhibited a linear decrease in porewater SO4 2− concentrations with depth with clear exceptions being PC-06, 08, and 14.The non-linear porewater SO4 2− profile in PC-06 was consistent with the de-stabilization of solid-phase hydrates.In PC-07, 14, and 21 porewater SO4 2− concentrations decreased to near zero limits of detection and intersected increasing headspace CH4 values at depth, indicative of an SMT.

Porewater Ca 2+ and Mg 2+ Concentrations, Ca/Mg Ratios, and Sediment CaCO3 Content
Figure 4 shows measured porewater Ca 2+ and Mg 2+ concentrations, Ca/Mg ratios, and estimated sediment CaCO3 content vs. depth for AC-07 piston cores.Porewater Ca 2+ and Mg 2+ concentrations generally decreased with depth in PC-04, 05, 07, 09, 16, and 21 from near average seawater values of 10.5 mM and 54.1 mM, respectively.The Ca/Mg ratio also decreased with depth in these cores from an average seawater value of ~0.19 porewater SO4 2− concentration profiles (Figure 3), increasing with depth from near average seawater values (2.2 mM) at the surface.In PC-07, 14, 21 porewater DIC concentrations increased with depth to a maximum at the SMT (Figure 3) then decreased below the SMT.As with porewater SO4 2− , Ca 2+ , Mg 2+ , and Ca/Mg ratio profiles, most piston cores exhibited a linear trend of increasing porewater DIC concentrations with depth.Most cores, however, showed some non-linearity in porewater DIC profiles.
Porewater DIC profiles in PC-06, 08, 09, and 14 exhibited the highest degree of non-linearity.Porewater δ 13 C-DIC values in the AC-07 cores were closer to average seawater values at the surface (~0‰) becoming more negative, or isotopically-lighter, with depth.In PC-07, 14, and 21 there was a noticeable inflection point in the porewater δ 13 C-DIC values coincident with the SMT (Figure 3).

Estimated SMT Depths and Sulfate Diffusion Rates
Depths for the sulfate methane transition (SMT) in each core were chosen as the depth where minimum CH4 and SO4 2− concentrations converge plus half the depth to the next whole round core section.For cores where vertical SO4 2− profiles did not reach the limits of detection, a linear extrapolation of the SO4 2− concentration vs. depth profile was used to estimate the depth of the SMT [3,39].The depth of the SMT provides a relative qualitative prediction of vertical CH4 flux.
Since porewater CH4 was not sampled directly, a semi-quantitative, comparative estimate of flux rates between sites was calculated from regression analysis of sediment porewater SO4 2− profiles.A 1:1 ratio of SR to CH4 oxidation is typically present during AOM [44].Assuming steady state conditions where AOM is the dominant process consuming sulfate, SO4 2− diffusion rates can be calculated from the linear fit to the SO4 2− concentration gradient according to Fick's first law [45,46]: where J represents the SO4 2− flux (mmol/m 2 -yr), φ is the sediment porosity, Ds is the diffusion coefficient for sulfate in seawater, c is the range in SO4 2− concentration, and x is the range in depth for the linear section of the SO4 2− porewater profile.Ds in Equation ( 6) is given by: where D0 is assumed to be 2.08 × 10 −2 m 2 /yr, and n = 3 for the clay-silt sediments in this region [47].An average of measured porosity was used in the diffusive flux calculation for each core and, given the shallow cores used in this study, D0 was assumed to remain constant with depth.As per convention, data presented references downward SO4 2− flux as positive (into the sediments) corresponding to an upward flux (negative) for CH4 out of the sediments.In sediment cores with non-linear porewater SO4 2− profiles, where possible, the linear portions of the SO4 2− profile below the apparent mixing depth were selected for calculation of the SO4 2− diffusion rate [44][45][46], otherwise, relative flux between cores was inferred by the depth of the SMT.
In piston cores where SO4 2− profiles were linear, SO4 2− diffusion rates were estimated from the linear fit to the SO4 2− concentration profile, assuming steady state conditions where AOM was the dominant process consuming SO4 2− in a 1:1 ratio to CH4 [46].Likewise the depth of the SMT was estimated for cores where the SMT was not present by extrapolating the SO4 2− concentration gradient to a depth where porewater SO4 2− concentrations equaled zero.A similar approach was employed for piston cores with non-linear profiles (AC-06, 08, 14) by using a linear portion of the SO4 2− profile at depth.These results should be viewed as rough relative estimates only since some of these profiles are likely non-steady state (Table 1).Figure 6 shows estimated SMT depths, sulfate diffusive flux ( ), and average porewater Cl − concentrations (Table 1) over a 3 km linear transect across the suspected seep feature on Inline 986 (Figure 2).The AC-07 cores are referenced to a linear distance from core PC-08.The AC-07 piston cores generally exhibited deep SMTs and low SO4 2− diffusive flux suggesting diffusion-dominated CH4 flux.Sulfate diffusion rates increased and SMT depths decreased closer to the area on the seafloor where the seismic profiles suggested gas venting and/or the presence of gas hydrates, such as those recovered in PC-06 (Figures 2 and 6).However, in general, the estimated sulfate diffusion rates shown Figure 6 suggest that the shallow sediments of Inline 986 overlay a deep-sea, CH4 charged sedimentological environment where most, or all, CH4 diffusing from below is consumed by AOM at the SMT and does not reach the overlying water column [20].
Energies 2014, 7 Figure 6.Estimated SMT depths, sulfate diffusion rates ( ), and average porewater Cl − concentrations over a 3 km transect across the suspected seep feature shown WesternGeco seismic profile for Inline 986 (Figure 2).Distances are referenced to AC-07 PC-08.Core PC-06 contained solid phase hydrates which destabilized during collection.

Diffusive Fluxes of Ca 2+ , Mg 2+ , and DIC and Carbon Mass Balance
In sediments where AOM dominates, 1 mole of SO4 2− should be consumed for each mole of CH4 diffusing upward, producing 1 mole of HCO3 − (Equation ( 2)).This is in contrast to organoclastic SR (Equation (1)), where 2 moles of HCO3 − are produced for each mole of SO4 2− consumed in the breakdown of sediment organic matter.These microbially-mediated processes typically shape the sediment porewater profiles of species like SO4 2− and DIC in organic-rich surface sediments of coastal areas where hydrates are found [21,29,48].For the AC-07 piston cores in this study which were collected from sediments with a low organic matter content and measurable low vertical CH4 flux, it is reasonable to assume that AOM is dominating the consumption of SO4 2− and production of HCO3 − [46].Seismic data and core porewater data support the assumption of a methane-charged environment.The max sediment organic carbon (SOC) measured in all the piston cores collected was 1.2% organic carbon by weight with an average of only 0.6% organic carbon.Model data from Sivan et al. [49] showed that at SOC values <5%, AOM was the dominant microbial process consuming SO4 2− and thus producing HCO3 − [21].
The over-production of HCO3 − in porewaters near the SMT due to AOM can lead to the precipitation of authigenic carbonates in the presence of cations like Ca 2+ and Mg 2+ via Equation (3) [23].In theory, if AOM is dominating CH4 consumption at the SMT in shallow sediments of AC-07 and producing HCO3 − , then downward diffusion of SO4 2− into sediments ( ) should be balanced by change in alkalinity due to DIC flux (dominated by HCO3 − ) at a 1:1 ratio [20,21,23,24,29].However, this simple relation is a gross simplification because of other geochemical process that occur in sediments overlying CH4-charged deep-sea environments.Over-production of HCO3 − in porewaters near the SMT due to AOM can lead to precipitation (or dissolution) of authigenic carbonates in the presence of cations like Ca 2+ and Mg 2+ changing both alkalinity and DIC concentrations near the SMT [23].Methanogenesis or thermogenic production of CH4 occurring below the SMT can also produce an upward DIC flux from below [20,21,23,24,29].Still, using some simple assumptions, estimates can be made in order to constrain relative CH4 fluxes and carbon mass balances at sites like AC-07 [20,21,23,24,29].A net alkalinity flux can be approximated as the net DIC flux [20,21,23,24,29].The net DIC flux can be approximated as DIC diffusion upward through shallow sediments above the SMT (JDIC-shallow), diffusion of deep DIC upward from below the SMT (JDIC-deep), and loss of DIC to authigenic carbonate phases.Loss of DIC to authigenic carbonate phases (JDIC-carbonate) can be estimated by a loss of major porewater cations like Ca 2+ and Mg 2+ as indicated by the downward flux of these cations into sediments from overlying seawater ( + ).
Snyder et al. [20] used this method to constrain the mass balance of carbon across the SMT for sediment core from the Umitaka Spur, Japan.In deep-sea, CH4 charged sedimentological environment like Umitaka Spur and AC-07 where most, or all, CH4 diffusing from below is consumed by AOM at the SMT and does not reach the overlying water column, then the mass balance of carbon at the SMT can be summarized as: The DIC diffusion upward through the shallow sediments above the SMT (JDIC-shallow) and flux of Ca 2+ and Mg 2+ into the sediments from the overlying seawater ( + ) can be approximated using the same methods used for J and Equations ( 6) and ( 7) assuming a D0 for HCO3 − , Ca 2+ , and Mg 2+ of 1.94 × 10 −2 m 2 /yr, 1.36 × 10 −2 m 2 /yr, and 1.26 × 10 −2 m 2 /yr, respectively [50].Given the shallowness of the AC-07 piston cores in relation to the SMT, estimation of the diffusion of DIC upward from below the SMT, or downward to below the SMT, was problematic due to the lack of adequate data points for estimating JDIC-deep.For this study, it is assumed that JDIC-deep << JDIC-shallow.Table 2 shows estimated , JDIC-shallow, and , , and JDIC-net for the AC-07 piston cores that exhibited (near-) linear profiles of relevant porewater species (SO4 2− , Ca 2+ , Mg 2+ , DIC).These cores include PC-04, 05, 09, 16 where the SMT was below the maximum core depth and PC-07 and 21 where the SMT was evident in core porewater profiles.For the AC-07 cores with linear profiles, ≈ −JDIC-shallow + JDIC-carbonate, where JDIC-carbonate = + .Estimated values for plotted against JDIC-net clearly fall along the 1:1 ratio for AOM (Figure 7).
This correlation supports the assumption that AOM is the dominant process consuming SO4 2− in the shallow sediments along AC-07 inline 986.This relationship, however, is a simplification of the actual biogeochemical processes occurring at the SMT in these cores.2).The 1:2 line indicates the molar stoichiometry of SO4 2− consumption to DIC production for organoclastic SR (Equation ( 1)) and the 1:1 line indicates the molar stoichiometry of SO4 2− consumption to dissolved inorganic carbon production for anaerobic oxidation of methane (Equation ( 2)).
The flux values in Table 2 suggest a large contribution of HCO3 − to porewaters near the SMT due to AOM.They also suggest that a large portion of this excess DIC is precipitated as authigenic Ca and Mg carbonates.If this is indeed the case then the DIC pool should strongly reflect the isotopic signature of the CH4 from which is was derived.The contribution of AOM to porewater DIC (HCO3 − ) in each core can be estimated using a mass balance based on the measured δ 13 C-DIC and δ 13 C-CH4 value near the base of cores where the SMT was not present and from the estimated depth of the SMT where it was present [16]: %DIC = ((δ 13 C-DICMIN − δ 13 C-DICSW)/(δ 13 C-CH4MIN − δ 13 C-DICSW)) × 100 (9) where %DICCH4 is the contribution of AOM to porewater DIC (HCO3 − ) in each core at or near the SMT, δ 13 C-DICMIN is the minimum porewater δ 13 C-DIC value at or near the SMT, δ 13 C-DICSW is the background porewater δ 13 C-DIC value (assumed for seawater as 0‰), and δ 13 C-CH4MIN is the minimum measured δ 13 C-CH4 value at or near the SMT.Assuming that AOM is the dominant process consuming SO4 2− at the SMT of the AC-07 cores, the percent of the DIC pool that is incorporated into the authigenic carbonate fraction can be estimated as [20]: %DICcarbonate = (JDIC-carbonate/ ) × 100 (10) The δ 13 C-CH4 values measured at the base of the AC-07 cores or at the SMT ranged from −62‰ to −106‰, suggesting a predominate biogenic source of CH4 (Table 3).The strongly negative, or isotopically lighter, δ 13 C-DIC values (−28‰ to −51‰) shown in Table 3 support the assumption that AOM is significantly contributing HCO3 − to the DIC pool (%DIC = 44%-68%) at or near the SMT of each core.Results suggest that a large fraction (%DICcarbonate = 31% to 59%) of this AOM-derived DIC pool is incorporated into the authigenic carbonate fraction.These estimates are significantly higher than those estimated by others at Umitaka Spur, Japan [20], Hydrate Ridge, and Cascadia Margin [15].3. Estimates of percent contribution of AOM to the DIC pool and the percent contribution of the DIC pool to solid phase carbonates at the SMT for the AC-07 piston cores (Table 2).The percent contribution of AOM to the DIC pool was based on a mass balance using the minimum measured δ 13 C-DIC and δ 13 C-CH4 value near the base of cores where the SMT was not present and from the estimated depth of the SMT where it was present.It is important to note that the flux calculations in Tables 1 and 2, and hence the estimates in Table 3, are based on general assumptions and limited data.The porewater profiles and flux rates used were assumed to be in steady-state.It was assumed that AOM was the dominant process consuming SO4 2− and that no significant amounts of other higher hydrocarbons were present.Presence of higher hydrocarbons, and their subsequent oxidation in porewaters, could result in conditions where > − [30].Note that even when the DIC flux is corrected to JDIC-net = −(JDIC-shallow ) + JDIC-carbonate (Equation ( 8)) near the SMT for the AC-07 cores, in most cases > −JDIC-net (Figure 7).It is also assumed that all CH4-carbon goes to the DIC pool and is not assimilated into microbial biomass or the organic carbon pool.Previous research by others has shown that <2% of methane-derived carbon is assimilated into biomass [21,51] but some studies suggest that cycling of methane-derived carbon into the organic matter pool may be significant [43].Other biogeochemical processes like organoclastic SR, fermentation, thermogenic and biogenic methane production, and incorporation of AOM derived HS − into Fe-sulfide mineral phases can all serve to alter theoretical steady-state porewater profiles for SO4 2− and DIC [20].
Other issues with the flux and budget estimates in this work are that DIC is not a direct measure of HCO3 − concentration nor does a change in HCO3 − directly correspond to a change in CaCO3 as shown in Equation (3).In sediments porewaters like those in this work, approximation of total alkalinity to carbonate alkalinity used in deep-ocean waters does not apply.Total alkalinity is a better measure than DIC for changes in porewater chemistry due to HCO3 − production and carbonate mineralization and precipitation.Changes in the total alkalinity of porewaters can account for changes in major dissolved porewater ions such as Ca 2+ and Mg 2+ .Still, if one assumes that changes in porewater alkalinity in methane-charged sediments like those in this study are largely controlled by excess DIC flux dominated by HCO3 − , then direct measurement of DIC concentrations in porewaters represents an adequate proxy.
Lastly it was assumed that Ca (and Mg) carbonate minerals like calcite and aragonite dominate sediment inorganic carbon in the sediments of AC-07 (Equation ( 4)).Looking at the Ca 2+ and Mg 2+ porewater concentration profiles and the estimated and , rates (Table 2) for the AC-07 cores, it is clear that porewater Mg 2+ concentrations decrease significantly with depth along with Ca 2+ concentrations.This could be indicative of the formation of dolomite at depth and/or the exchange of dissolved Mg 2+ with clays [16,20,52].Dolomite formation at depth, as demonstrated in studies by others on Blake Ridge [52] and the west African Margin [16], could increase , relative to , resulting in an overestimation of JDIC-carbonate and/or a downward flux of DIC below the SMT, thereby impacting JDIC-deep (not measured in this study).The greater decrease in porewater Ca 2+ concentrations relative to Mg 2+ as shown in porewater Ca/Mg ratio profiles in the AC-07 cores (Figure 4), however, is still consistent with the assumption that carbonate formation (calcite and/or aragonite) is the dominant factor controlling JDIC-carbonate.In summary, the JDIC-carbonate values (Table 2) and subsequent %DICcarbonate values (Table 3) in this study likely overestimate the amount of CH4-derived excess DIC that is precipitated as authigenic Ca and Mg carbonates.A common finding of most biogeochemical studies on carbon cycling in methane charged sediments overlying hydrate bearing strata in areas like Alaminos Canyon is that more data and higher resolution data (with depth) is required to develop an adequate carbon budget [20,21,29].Still, this study demonstrates that even in low, diffusive dominated flux areas like Alamos Canyon, the incorporation of CH4-derived excess DIC into authigenic carbonates is significant and simple sediment profiles of SO4 2− , DIC, and other key porewater species provide an adequate first order estimate of CH4carbon fluxes and cycling.

Conclusions
Results of geochemical characterization of sediments and porewaters collected from the seafloor of the Alaminos Canyon region of the GoM, Inline 986, June 2007 suggest a deep sea, CH4 charged, diffusion-dominated sedimentological environment where, except in areas of active fluid advection or exposure of solid phase hydrate near the seafloor (PC-06), most of the CH4 diffusing up through the sediments from the GHSZ below is consumed by AOM at the SMT.Estimated SO4 2− diffusion rates ( ) ranged from 8 to 70 mmol/m 2 -yr.A simple stable isotope mass balance using the minimum measured δ 13 C-DIC and δ 13 C-CH4 values near the base of the AC-07 cores and/or SMT supports the assumption that AOM is significantly contributing HCO3 − to the DIC pool (%DIC = 44%-68%) at or near the SMT.When DIC flux is corrected to JDIC-net to account for the incorporation of CH4-AOM-derived DIC into solid phase carbonates, the resulting stoichiometric balance of 1:1 for :JDIC-net supports the assumption of an AOM dominated system.
Although this study likely overestimates the incorporation of CH4-derived excess DIC into authigenic carbonates, results support the conclusion that incorporation of CH4-derived carbon into dissolved and solid inorganic carbon phases represents a significant carbon sink and plays a role in regulating the flux of methane to the overlying water column and atmosphere at Alaminos Canyon and potentially other site with CH4 hydrate bearing marine sediments.More data and better sampling resolution is required for detailed assessments of carbon cycling in CH4-charged sediments but sediment profiles of SO4 2− , DIC, and other key porewater species used in this study provide an adequate, first-order estimate of CH4carbon fluxes and cycling in Alaminos Canyon.

Figure 1 .
Figure 1.Map showing the location of Alaminos Canyon (AC) in the Gulf of Mexico (GoM).Inlet shows detail of survey block AC 818 containing seismic line 986 (dashed line).Black dots indicate locations of piston cores collected by the U.S. Naval Research Laboratory (NRL) as part of the AC-07 methane hydrate research expedition, June 2007 (Map modified from Google Earth using data from U.S. Geological Survey (USGS), National Archive of Marine Seismic Surveys; Texas A&M University, Gulf of Mexico Coastal Ocean Observing System (GCOOS), Coastal Relief Model for the Gulf of Mexico; and WesternGeco).

Figure 2 .
Figure 2. Locations for piston cores collected during the AC-07 research expedition along the WesternGeco seismic profile for Inline 986.The seismic profile has been colorized to indicate areas of potential hydrate deposits located above interpretation of free gas zones and free gas venting.

Figure 7 .
Figure 7. Scatter plot of vs. JDIC-net for the AC-07 piston cores (Table2).The 1:2 line . Porewater Ca2+and Mg 2+ concentrations and Ca/Mg ratio profiles showed the most variability with depth in PC-06, 08, 14, and 15.The was no clear trend in estimated sediment CaCO3 profiles with depth in any of the piston cores collected along AC-07 Inline 986.Estimated sediment CaCO3 content in all cores varied from 5% to 38% by weight.