Archaea mediating anaerobic methane oxidation ... - Semantic Scholar

Organic Geochemistry 31 (2000) 1175±1187

www.elsevier.nl/locate/orggeochem

Archaea mediating anaerobic methane oxidation in deep-sea sediments at cold seeps of the eastern Aleutian subduction zone Marcus Elvert a,*,1, Erwin Suess a, Jens Greinert a, Michael J. Whiticar b a GEOMAR, Research Center for Marine Geosciences, Wischhofstr. 1-3, D-24148 Kiel, Germany School of Earth and Ocean Sciences, University of Victoria, PO Box 3050, British Columbia, V8W 2Y2, Canada

b

Received 12 October 1999; accepted 1 August 2000 (returned to author for revision 7 January 2000)

Abstract Cold seeps in the Aleutian deep-sea trench support proli®c benthic communities and generate carbonate precipitates which are dependent on carbon dioxide delivered from anaerobic methane oxidation. This process is active in the anaerobic sediments at the sulfate reduction-methane production boundary and is probably performed by archaea working in syntrophic co-operation with sulfate-reducing bacteria. Diagnostic lipid biomarkers of archaeal origin include irregular isoprenoids such as 2,6,11,15-tetramethylhexadecane (crocetane) and 2,6,10,15,19-pentamethylicosane (PMI) as well as the glycerol ether lipid archaeol (2,3-di-O-phytanyl-sn-glycerol). These biomarkers are prominent lipid constituents in the anaerobic sediments as well as in the carbonate precipitates. Carbon isotopic compositions of the biomarkers are strongly depleted in 13C with values of d13C as low as ÿ130.3% PDB. The process of anaerobic methane oxidation is also re¯ected in the carbon isotope composition of organic matter with d13C-values of ÿ39.2 and ÿ41.8% and of the carbonate precipitates with values of ÿ45.4 and ÿ48.7%. This suggests that methane-oxidizing archaea have accumulated within the microbial community, which is active at the cold seep sites. The dominance of crocetane in sediments at one station indicates that, probably due to decreased methane venting, archaea might no longer be growing, whereas high amounts of crocetenes found at other more active stations may indicate recent ¯uid venting and active archaea. Comparison with other biomarker studies suggests that various archaeal assemblages might be involved in the anaerobic consumption of methane. The assemblages are apparently dependent on speci®c conditions found at each cold seep environment. Selective conditions probably include water depth, temperature, degree of anoxia, and supply of free methane. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Aleutian subduction zone; Cold seeps; Authigenic carbonates; Biomarkers; Irregular isoprenoids; Carbon isotopic composition; Crocetane; Crocetenes; PMI; Archaeol

1. Introduction Chemoautotrophic microbial communities inhabiting sediments at cold seeps or living as symbionts in vent macrofauna are important for carbon cycling in deep-sea * Corresponding author. Fax: +1-49-431-600-2928. 1 Present address: Max-Planck-Institute for Marine Microbiology, Celsiussor. 1, 28359 Bremen, Germany. Fax: +1-49421-2028-690; e-mail: melvert@mpi- bremen.de. E-mail address: [email protected] (M. Elvert).

environments, preferentially along convergent continental margins. At the cold seeps, ¯uid venting supports benthic communities and generates authigenic carbonates from the biogeochemical turnover and interaction between ¯uids and ambient bottom water (Suess et al., 1985; Kulm et al., 1986; Wallmann et al., 1997). Growth and metabolism of the associated vent macrofauna are based on a chemoautotrophic food chain which starts with the microbially mediated oxidation of reduced compounds, such as methane or hydrogen sul®de, delivered by active ¯uid venting. For methane, the

0146-6380/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0146-6380(00)00111-X

1176

M. Elvert et al. / Organic Geochemistry 31 (2000) 1175±1187

oxidation to carbon dioxide occurs either by aerobic (Childress et al., 1986) or anaerobic processes (Suess and Whiticar, 1989), with the latter still not completely understood. However, subsequent incorporation of carbon dioxide by organisms in tissues or precipitation of carbonates from oversaturated microenvironments causes a strong carbon isotope shift towards 13C-depleted values often cited as evidence for methane oxidation; e.g. mytilid mussels, vestimentiferan and pogonophoran tube worms, and carbonates are depleted in 13C to values as low as ÿ77% PDB (e.g. Paull et al., 1985; Brooks et al., 1987; Ritger et al., 1987; Suess et al., 1998). Biomarkers found at ancient and recent methane seeps have provided another important piece of evidence supporting methane oxidation under anaerobic conditions (Elvert et al., 1999; Hinrichs et al., 1999; Thiel et al., 1999; Pancost et al., 2000). These authors predominantly identi®ed irregular tail-to-tail isoprenoids and isopranylglycerol diethers such as 2,6,11,15-tetramethylhexadecane (crocetane), 2,6,10,15,19-pentamethylicosane (PMI), 2,3-diO-phytanyl-sn-glycerol (archaeol), 2-O-3-hydroxyphytanyl3-O-phytanyl-sn-glycerol (sn-2-hydroxyarchaeol), and 3O-3-hydroxyphytanyl-3-O-phytanyl-sn-glycerol (sn-3hydroxyarchaeol) with highly depleted carbon isotope values as low as ÿ123.8% PDB from various anaerobic settings. Characteristic settings include methane seeps associated with marine gas hydrates (Elvert et al., 1999; Hinrichs et al., 1999), ancient methane vent systems (Thiel et al., 1999), and methane-rich mud volcanoes (Pancost et al., 2000). The detection of irregular isoprenoids and/or isopranylglycerol diethers, both traditionally believed to be biosynthesized by methanogenic archaea, with such extremely low carbon isotope values prompted these authors to suggest that either certain methanogens themselves are involved in the consumption of methane, operating in reverse in syntrophic cooperation with sulfate reducers (Elvert et al., 1999; Thiel et al., 1999; Pancost, 2000), or that until now unknown methanogens within archaeal lineages evolved to being capable of using methane as their predominant or even exclusive carbon source (Hinrichs et al., 1999). Following these ideas, we analyzed speci®c biomarkers related to anaerobic methane-oxidizing processes from sediments and carbonates at cold seep settings of the eastern Aleutian subduction zone, adjacent to the Aleutian deep-sea trench. These cold seeps are among the deepest observed (4800 m) and therefore, being far removed from the photic zone, are well suited to study chemoautotrophic processes because very little metabolizable particulate organic matter reaches this depth. We especially examined the abundance, carbon isotope values, and signi®cance of biomarkers diagnostic of anaerobic methane oxidation. Moreover, we evaluated the variability of the speci®c biomarkers found in this study compared to those observed at other cold seep environments.

2. Materials and methods 2.1. Study area The study area at the eastern Aleutian subduction zone, referred to as SHUMAGIN sector, was surveyed and sampled during R/V SONNE cruises 97 (SO 97) and 110 (SO 110-lb and SO 110-2), and is shown in Fig. 1a. The tectonic setting, manifestations of venting, and the general sampling strategy have been described earlier by Suess et al. (1998). Widespread methane venting was observed along the entire margin and speci®cally o SHUMAGIN at the intersection of accretionary ridges with tensional faults. These faults occur in canyons landward of the deformation front at water depths around 4800 m and are the result of oblique subduction of the Paci®c plate underneath the Aleutian arc. Colonies of typical seep macrofauna and authigenic carbonate crusts were found. The seep biota consists of bacterial mats, pogonophorans, vestimentiferans, and large colonies of bivalves. The carbon isotope composition of tissues from the seep fauna ranged from ÿ57.1 to ÿ64.3% and thus identi®es methanotrophy as the dominant carbon metabolizing pathway (Suess et al., 1998). Similarly, for authigenic carbonates, d13C values between ÿ42.7 and ÿ50.8% were reported (Greinert, 1998), suggesting that a mixture of biogenic methane, via anaerobic oxidation, and carbon dioxide supplied by sulfate-reducing bacteria was the ultimate carbon source of the authigenic mineralogies. 2.2. Sediment, pore water, and carbonate analysis Contents of Corg were determined from the carbonate free, dried, and homogenized sediment material using a Carlo Erba Nitrogen Analyzer 1500. For carbonate removal, 3 g of wet sediment were treated over night with 15 ml of 10% HCl. After freeze-drying, samples were homogenized by using an agate ball mill. Standard deviations of this method were 0.02%. Sulfate measurements were carried out by ion chromatography and detection by conductivity. Sulfate values are reported in mM and were calibrated with IAPSO-standard seawater. Using duplicate measurements, standard deviations were within 1.5%. The authigenic carbonates were identi®ed by standard X-ray diraction analysis. The speci®c calcite sample selected for extraction of biomarkers was a high Mg-calcite (Greinert, 1998). 2.3. Extraction, chromatographic separation, hydrogenation and derivatization Lipids were extracted ultrasonically from the samples (20±25 g of wet sediment) with 50 ml of methanol/ dichloromethane (2:1, v/v), 50 ml of methanol/dichloromethane (1:2, v/v), and twice with 50 ml of dichloromethane. For the carbonate, 25 g were washed with


1177

Fig. 1. Aleutian subduction zone with (a) area of investigation (SHUMAGIN sector) and (b) detailed map of coring stations TV-G 97 (= TV-guided grab sampler), TV-GKG 40 (= TV-guided box corer), TV-G 43, and TV-G 48 inside a canyon landward of the deformation front.

acetone and dissolved in a 1 l round bottom ¯ask by adding stepwise 500 ml of 1 N HCl and stirring for 6 h. After centrifugation for 5 min at 4000 rpm and decantation of the supernatant, the residue was washed two times with pre-extracted water and the lipids were extracted as described above for wet sediment material. Fractions were separated from the lipid extracts by medium pressure liquid chromatography on 1.3 g silica gel (70±230 mesh, 5% deactivated). Chromatographic separation was by elution with (I) 13 ml of n-hexane (hydrocarbons), (II) 10 ml of dichloromethane/n-hexane (20:80, v/v; esters and ketones), (III) 10 ml of dichloromethane (alcohols), and (IV) 10 ml of methanol/dichloromethane (50:50, v/v; glyco- and phospholipids). Elemental sulfur in the hydrocarbon fractions (I) was removed by passing the fractions over separate short columns ®lled with 1 g of activated copper powder using n-hexane as eluent. Hydrogenation of hydrocarbon fractions was carried out by saturation of 50 ml n-hexane with H2 in fusible glass ampoules pre-®lled with 10 mg of PtO2 and subsequent adding of 50 ml of sample (12 aliquot in n-hexane). After ¯ushing with H2, the ampoules were closed and stored at room temperature for 1 h. Finally, the samples were

directly analyzed by gas chromatography±mass spectrometry (GC±MS). To facilitate gas chromatographic analysis of alcohols, trimethylsilyl (TMS) derivatives were produced. Alcohol fractions were evaporated under a stream of pure nitrogen to near dryness, mixed with 100 ml BSTFA (bis(trimethylsilyl)tri¯uoroacetamide; Supelco), and heated in closed glass ampoules for 2 h at 80 C. Following evaporation to near dryness under nitrogen, the residue was taken up in n-hexane and subsequently analyzed by mass spectrometry. 2.4. Gas chromatography (GC) Gas chromatographic analyses of hydrocarbons were performed using a 30 m apolar DB-5 fused silica capillary column (0.25 mm internal diameter (ID), ®lm thickness 0.25 mm; J&W Scienti®c) in a Carlo Erba 5160 gas chromatograph equipped with an on-column injector and a ¯ame ionization detector. The samples were injected at 60 C. After a 1 min hold time, the oven temperature was raised to 140 C at 10 C/min, then to 310 C at 5 C/min and ®nally kept at 310 C for 25 min. The carrier gas was H2 at a ¯ow rate of 2.5 ml/min. Concentrations for each

1178


compound were determined by adding internal standards (3-methylnonadecane, 2-methylicosane, 5b(H)cholane) with known concentrations prior to GC analysis and are reported in mg/g Corg. Standard deviations for single compounds are below s=2 mg/g Corg except for compounds with more than 50 mg/g Corg (below =10 mg/g Corg). Loss of material during analysis was monitored by adding a recovery standard (n-C40) prior to the overall analytical procedure. In general, typical recoveries were 70±80% relative to n-C40. 2.5. Gas chromatography±mass spectrometry (GC±MS) Hydrocarbons and alcohols (as TMS-derivatives) were identi®ed by GC±MS using a Carlo Erba 8000 gas chromatograph interfaced to a Fisons MD 800 mass spectrometer operated in electron impact (EI-) mode at 70 eV (cycle time 0.9 s, resolution 1000) with a mass range of m/z 40±600 for hydrocarbons and m/z 40±800 for alcohols. The gas chromatograph was equipped with a DB-1 fused silica capillary column (30 m, 0.25 mm ID) coated with cross-linked methyl silicone (®lm thickness 0.25 mm; J&W Scienti®c) using He as carrier gas. The samples were injected in splitless mode (hot needle technique; injector temperature: 285 C) and subjected to the same temperature program given for GC measurements (see Section 2.4.). 2.6. Stable carbon isotope analysis Carbon isotope compositions of hydrocarbons and alcohols were determined using a coupled gas chromatograph±combustion±isotope ratio mass spectrometer (GC±C±IRMS). The mass spectrometer (Finnigan MAT 252) was connected to a Varian 3300 GC equipped with a 50 m CP-Sil 5 CB-MS (0.25 mm ID, 0.4 mm stationary phase; Chrompack). The carrier gas was He at a ¯ow rate of 1.5 ml/min. The samples were on-column injected at 60 C and after 1 min the oven temperature was raised to 140 C at 10 C/min, then to 230 C at 3 C/min, and ®nally to 310 C at 2 C/min at which it was held for 35 min. Carbon isotope ratios are reported in the d notation as per mil (%) deviation from the Pee Dee Belemnite standard (PDB). Internal standards (5b(H)-cholane and n-C36 for hydrocarbons; n-C20 and n-C36 for alcohols) of known isotopic composition were co-injected with each sample for monitoring reproducibility and precision during the project. Analytical reproducibility was on average within 0.2±0.3% for an n-alkane standard (n-C13 to n-C38) with no background, but was much more variable for the complex mixtures analyzed here (up to 1.6%) due to factors such as co-elution or signal to background ratio. Isotopic compositions of alcohols were measured in the form of TMS-derivatives and corrected for the isotopic shift associated with the addition of carbon atoms during derivatization according to Huang et al. (1995).

d13Corg was measured by elemental analysis±isotope ratio mass spectrometry (EA±IRMS) using a Carlo Erba Elemental Analyzer connected via a ConFloTM interface to the Finnigan MAT 252. Analytical reproducibility for duplicate runs was below 0.1%. 3. Results Three sediments and one carbonate sample from four dierent stations at cold seeps were analyzed for biomarkers indicative of anaerobic methane oxidation. The active seep sites along with extensive carbonate crusts and methane anomalies of the bottom water column were observed at a location from the SHUMAGIN sector inside a canyon which crosscuts the third accretionary ridge (Suess et al., 1998) (Fig. 1b). The canyon itself is cut by two faults along N±S and NNW±SSE direction, which probably provide ¯uid pathways and focus diusive ¯uid venting. Pore water analysis showed that sediments analyzed were well within the sulfate reduction zone which starts right below the sediment surface (Fig. 2). Sulfate concentrations reach 10 mM at station TV-G 43 (35 cmbsf) and concentrations