Life at Cold Seeps: a Synthesis of Biogeochemical and Ecological Data from Kazan Mud Volcano, Eastern Mediterranean Sea

Chemical Geology 205 (2004) 367–390 www.elsevier.com/locate/chemgeo

Life at cold seeps: a synthesis of biogeochemical and ecological data from Kazan mud volcano, eastern Mediterranean Sea

Josef P. Wernea,*, Ralf R. Haeseb,1, Tiphaine Zitterc, Giovanni Aloisid,2, Ioanna Bouloubassie, Sander Heijsf, Aline Fiala-Me´dionig, Richard D. Pancosta,3, Jaap S. Sinninghe Damste´ a, Gert de Langed, Larry J. Forneyf,4, Jan C. Gottschalf, Jean-Paul Foucherh, Jean Masclei, John Woodsideb the MEDINAUT and MEDINETH Shipboard Scientific Parties5

a Department of Marine Biogeochemistry and Toxicology, Royal Netherlands Institute for Sea Research (NIOZ), Den Burg, The Netherlands b Department of Geochemistry, Faculty of Earth Sciences, Utrecht University, Utrecht, The Netherlands c Department of Sedimentology and Marine Geology, Faculty of Earth and Life Sciences, Free University Amsterdam, Amsterdam, The Netherlands d Laboratoire d’Oceánographie Dynamique et de Climatologie, Universite´ Pierre et Marie Curie, Paris, France e Laboratoire de Chimie et Biogeochimie Marines, Universite´ Pierre et Marie Curie, Paris, France f Department of Microbiology, Center for Ecological and Evolutionary Studies, University of Groningen, Haren, The Netherlands g Observatoire Oceánologique, Universite´ Pierre et Marie Curie, Banyuls-sur-Mer, France h DRO/Environments Se´dimentaires, IFREMER, Brest, France i Geósciences Azur, Observatoire Oceánologique de Villefranche, Villefranche-sur-Mer, France Received 4 October 2002; received in revised form 30 September 2003; accepted 23 December 2003

Abstract

Recent field observations have identified the widespread occurrence of fluid seepage through the eastern Mediterranean Sea floor in association with mud volcanism or along deep faults. Gas hydrates and methane seeps are frequently found in cold seep areas and were anticipated targets of the MEDINAUT/MEDINETH initiatives. The study presented herein has utilized a multi- disciplinary approach incorporating observations and sampling of visually selected sites by the manned submersible Nautile and by ship-based sediment coring and geophysical surveys. The study focuses on the biogeochemical and ecological processes and conditions related to methane seepage, especially the anaerobic oxidation of methane (AOM), associated with ascending fluids on Kazan mud volcano in the eastern Mediterranean. Sampling of adjacent box cores for studies on the microbiology, biomarkers, pore water and solid phase geochemistry allowed us to integrate different biogeochemical data within a spatially

* Corresponding author. Current address: Large Lakes Observatory and Department of Chemistry, University of Minnesota Duluth, 10 University Dr. 109 RLB, Duluth, MN 55812, USA. Tel.: +1-218-726-7435; fax: +1-218-726-6979. E-mail addresses: [email protected] (J.P. Werne), [email protected] (R.R. Haese). 1 Address as of October 2003: Geoscience Australia, Canberra. 2 Current address: Department of Marine Environmental Geology, GEOMAR, Kiel, Germany. 3 Current address: Organic Geochemistry Unit, School of Chemistry, University of Bristol, Bristol, UK. 4 Current address: Department of Biological Sciences, University of Idaho, Moscow, ID, USA. 5 S. Asjes, K. Bakker, M. Bakker, J.L. Charlou, J.P. Donval, R. Haanstra, P. Henry, C. Huguen, B. Jelsma, S. de Lint, M. van der Maarel, S. Muzet, G. Nobbe, H. Pelle, C. Pierre, W. Polman, L. de Senerpont, M. Sibuet.

0009-2541/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2003.12.031 368 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 highly heterogeneous system. Geophysical results clearly indicate the spatial heterogeneity of mud volcano environments. Results from pore water geochemistry and modeling efforts indicate that the rate of AOM is f6 mol m2 year1, which is lower than at active seep sites associated with conditions of focused flow, but greater than diffusion-dominated sites. Furthermore, under the non-focused flow conditions at Kazan mud volcano advective flow velocities are of the order of a few centimeters per year and gas hydrate formation is predicted to occur at a sediment depth of about 2 m and below. The methane flux through these sediments supports a large and diverse community of micro- and macrobiota, as demonstrated by carbon isotopic measurements on bulk organic matter, authigenic carbonates, specific biomarker compounds, and macrofaunal tissues. Because the AOM community appears to be able to completely oxidize methane at the rate it is seeping through the sediments, ultimate sinks for methane in this environment are authigenic carbonates and biomass. Significant differences in organic geochemical data between this site and those of other cold seep environments, even within the eastern Mediterranean mud volcanoes, indicate that the microbiological communities carrying out AOM varies depending on specific conditions such as methane flux and salinity. D 2004 Elsevier B.V. All rights reserved.

Keywords: Anaerobic oxidation of methane; Cold seep fauna; Gas hydrate; Mud volcano; Biogeochemistry; Microbiology

1. Introduction tidisciplinary investigations combined visual observations from dives aboard the manned submersible R/V Methane cold seeps have become environments of Nautile (launched from the support vessel N/O Nadir) intensive study in recent years for several reasons. and geophysical/acoustic surveys with shore-based Among these are: (1) the potential for methane release sedimentological, (bio)geochemical and (micro)bio- from hydrates and the subsequent impact on climate logical investigations on seafloor sediment samples (Dickens et al., 1995), (2) slope destabilization associ- taken from the Nautile and via box core from the R/V ated with gas hydrate dissociation (Henriet and Mie- Professor Logachev (MEDINAUT/MEDINETH Ship- nert, 1998), (3) energy sources for chemosynthetically board Scientific Parties, 2000). based life (Sassen et al., 1993; Sibuet and Olu, 1998), One of the major objectives of the MEDINAUT/ and (4) as a potential energy resource (Hovland, 2000). MEDINETH expeditions was an in depth biogeo- Furthermore, the anaerobic oxidation of methane chemical and (micro)biological investigation of (AOM) has been identified in methane seep environ- AOM. It is now generally accepted that a microbial ments world-wide, and is believed to scavenge signif- consortium consisting at least of methane-oxidizing icant amounts of methane (Reeburgh et al., 1993) archaea and sulfate-reducing bacteria consumes sulfate which could otherwise be released to the atmosphere, and methane in a 1:1 molar ratio during AOM, potentially contributing to global warming (e.g., Dick- producing sulfide and DIC according to the general- ens et al., 1995). ized net reaction Mud volcanoes, which result from the extrusion of 2 fluid-rich mudflows (mud breccias), are often associ- SO4 þ CH4 ! HS þ HCO3 þ H2O ð1Þ ated with methane seepage, and, given the appropriate pressure and temperature conditions, with gas hydrates as proposed by Reeburgh (1976). While the consor- (Brown, 1990; Milkov, 2000). Mud volcanoes have tium hypothesis for net AOM has been proven through been identified in the eastern Mediterranean (Cita et al., a number of studies (Boetius et al., 2000; Hinrichs et 1981, 1989; Cronin et al., 1997), and active methane al., 1999; Orphan et al., 2001), the specific pathways seepage has been documented on these mud volcanoes of methane-derived carbon incorporation into micro- (Emeis et al., 1996; Limonov et al., 1996; Woodside et bial biomass remain debated (Valentine and Reeburgh, al., 1998). In 1998 and 1999, the joint French–Dutch 2000; Sørensen et al., 2001). Furthermore, the specific MEDINAUT and the Dutch MEDINETH expeditions organisms remain unknown and uncultured, though were made to investigate eastern Mediterranean mud significant advances have been made in narrowing volcanoes and their associated cold seeps. These mul- down the organisms based in large part on phyloge- J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 369 netic studies (Hinrichs et al., 1999; Boetius et al., bers of the MEDINAUT and MEDINETH expeditions. 2000; Orphan et al., 2001; Aloisi et al., 2002). This The focus is on studies of methane cold-seeps on paper is a synthesis of the geophysical, (bio)geochem- Kazan mud volcano, located in the Anaximander ical and (micro)biological studies carried out by mem- Mountains area (Fig. 1).

Fig. 1. (A) Eastern Mediterranean Sea geodynamic map, with location of the main features cited in the text and of the mud diapiric fields within the Mediterranean ridge (triangles)—(1) Prometheus, (2) Pan di Zucchero, (3) Prometheus 2, (4) Olimpi, (5) United Nation. (B) Multibeam bathymetry map of the Anaximander Mountains and location of mud volcanoes (stars). 370 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

2. Geologic setting and regional tectonics: Mediterranean Sea Ryan et al., 1973) beneath the formation of the cold seep environment Anaximander Mountains. The gravity signature (Woodside, 1977; Woodside et al., 1997), lithological Eastern Mediterranean tectonics record the initial facies (Woodside et al., 1997), and surface deformation stages of collision between Africa and Europe, from (Zitter et al., 2003) vary between the eastern and incipient stages south of Crete to more advanced stages western Anaximander Mountains regions, clearly indi- south of Cyprus (and eastwards into the Zagros area cating differences in the geology of the two regions. where continental collision between the Arabian prom- The western mountains belong to the Eastern Hellenic ontory and Europe is now occurring). The complex fore-arc deformation zone (Woodside et al., 2000; Ten deformation of the Eastern Mediterranean results from Veen et al., in press), whereas the eastern mountains the interactions of several plates and micro-plates (the Anaxagoras seamount) form the western limb of (McKenzie, 1972; Reilinger et al., 1997; McClusky the Cyprus Arc and are affected by a large NW–SE et al., 2000) and develops into two cup-shaped arc transpressional wrenching system (Woodside et al., systems, the Hellenic and Cyprus Arcs (Fig. 1). These 2002; Zitter et al., 2003; Ten Veen et al., in press). This two arc systems are characterized by different tectonic area is intensely crosscut by major N150j and N070j regimes, with the Hellenic arc characterized by more faulting, inferred to be active shearing based on both active subduction (Papazachos and Comninakis, 1978; the seismic signature (pop-up flower structures) and the Spakman et al., 1988; Chaumillon and Mascle, 1997) bathymetric expression. and the Cyprus Arc characterized by a quieter tectonic Most of the mud volcanoes discovered in the Anax- and seismic activity (Rotstein and Kafka, 1982; Ben- imander Mountains occur within the Anaxagoras sea- Avraham et al., 1988) where subduction may be halted mount (Fig. 1), associated with both the strike-slip (Robertson, 1998). The submarine Anaximander faults and secondary, probably extensional, faults (Zit- Mountains region (along with the Florence Rise and ter, unpublished results). They are generally identified the Rhodes Basin) is located at the connection between as conical topographic features on the order of 50 to these two arc systems, and as a result has been the 100 m high and 1 km wide (Woodside et al., 1998). subject of several recent studies (Woodside et al., 2000, Nonetheless, the mud volcanoes and associated mud- 2002; Zitter et al., 2003; Ten Veen et al., in press). flows show an important variability in morphology, A series of mud volcano fields exists within the size and acoustic characteristics. The shapes vary from Mediterranean Ridge accretionary prism (Cita et al., conical for Tuzlukush mud volcano, a quite perfectly 1981, 1989, 1996; Cronin et al., 1997; Mascle et al., circular dome for Kula mud volcano, an irregular dome 1999). Two mud volcanoes from the Olimpi field for Kazan mud volcano, to the several km-wide ‘‘mud (Napoli mud volcano and Milano mud volcano) were pie’’ morphology of the exceptionally large Amster- investigated in detail during the Ocean Drilling Pro- dam mud volcano. gram Leg 160 (Robertson et al., 1998). The Anax- imander Mountains mud volcano field occurs in a different context of the eastern Mediterranean colli- 3. Kazan mud volcano sion zone. This mud volcano field was discovered from multibeam mapping during the ANAXIPROBE Kazan mud volcano is a small dome, about 50 m cruise in 1995 based on high backscattering patches high and 1000 m wide, with an elliptical summit at associated with circular relief (Woodside et al., 1997, about 1700 m water depth (Fig. 2). It is located in the 1998), and has received much less vigorous study. southern part of the Anaxagoras seamount (E30j33.5V, The Anaximander Mountains are a complex of three N35j26V) at the interplay between the N150j and seamounts (Fig. 1), rising more than 1000 m above the N070j faults. From the sidescan sonar record, the seafloor, that rifted southwards from Turkey in late volcano shows high to very high backscatter (Fig. 2). Miocene–early Pliocene times (Woodside et al., 1997). The high backscatter signature is believed to be caused The post-Miocene formation of this area is demonstrat- by both the seafloor roughness and the volumetric ed by the absence of evaporites (that were deposited heterogeneity of the mudflows produced by clasts during the Messinian salinity crisis in most parts of the within the mud breccia (Volgin and Woodside, J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 371

Fig. 2. Sidescan sonar image and subbottom profile of Kazan Mud Volcano. Bathymetric map of Kazan mud volcano with the interpretation of sidescan sonar, the position of dive MN10 (in black) and MN12 (in grey), as well as the positions of samples (stars), as follows: (1) MN12BT3, (2) MNLBC18, (3) MNLBC19.

1996). High backscatter is then an indication of the Fig. 2 for dive tracks) from which it was possible to presence of a mudflow, with mud breccia either distinguish two active areas of seepage. The first one is cropping out or near the seafloor (i.e. recent mud- a plateau on the northern part of the summit associated flows). The backscatter level decreases with increasing with a recent mudflow (Fig. 3a); the second is on the thickness of the hemipelagic sediments covering the southeastern flank of the volcano and corresponds to a mudflows (i.e. old mudflows). Several other parame- patch of enhanced backscatter in the sonar imagery ters, such as carbonate crusts on the seafloor or the (Fig. 2). No fresh mud breccia was observed exposed presence of near bottom gas hydrates can influence the on the seafloor, suggesting that the strong backscatter backscatter signature. The tongue-shaped mudflows of came from carbonate crusts and fields of shells ob- Kazan mud volcano extend radially about 1 km away served during the dives (Fig. 3b). In the area surround- from the summit, with recent mudflows located on the ing Kazan mud volcano, faults (Fig. 3c) trend in many northern and eastern sides of the volcano. In contrast to directions (Fig. 2), probably due to the position of the most of the mud volcanoes, part of the summit is volcano at the intersection of major faulting between characterized by a low backscatter area, which is the Anaximenes and the Anaxagoras seamounts (Zit- probably less active (Fig. 2). This observation was ter, unpublished results). Active degassing associated confirmed by dives of the submersible Nautile (see with faulting around Kazan mud volcano is indicated 372 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

4. Methods

4.1. Geophysics

The sidescan sonar and subbottom profiling data were acquired with a MAK-1 system during the 6th Training Through Research (TTR6) expedition. The vehicle (or fish) operating in a Long Range mode at 30- kHz frequency, is towed at 150 m above the seafloor at a speed of 1.5 to 2 knots and is able to insonify a swath 2000 m wide (1000 m on each side). The MAK-1 fish is equipped with a sub-bottom profiler (SBP) operating at a 6-kHz frequency.

4.2. Sampling

Samples were taken from Kazan mud volcano by a variety of methods. Carbonate crust samples were taken either by dives of the submersible Nautile during the MEDINAUT cruise of the R/V Nadir in Novem- ber–December 1998, or by box cores taken during the MEDINETH cruise of the R/V Professor Logachev. Crust sample MNT12BT3 (35j26.007VN, 30j33.53VE, 1706 m water depth (Fig. 2) taken from Kazan is the focus of both organic geochemical and carbonate geochemical/mineralogical studies, and was sampled by submersible dives of the Nautile. Two box cores taken from the summit of Kazan mud volcano during the 1999 MEDINETH cruise of the R/V Professor Logachev (MNLBC18 and MNLBC19, 35j25.936VN, 30j33.704VE, water depth 1673 m and 35j25.950VN, 30j33.679VE, water depth 1673 m, respectively, Fig. 2), were dedicated to geochemical and microbiological analysis. Pore water geochemistry and molecular isotopic geochemistry were studied in core MNLBC19, and microbiological analyses were performed on both MNLBC18 and MNLBC19. It should be noted that the mud breccia sediments investigated in this study were taken by box core Fig. 3. Photographs taken during dives of the submersible Nautile to rather than sampled directly by submersible. There- Kazan mud volcano. (a) Mud breccia showing size variability of fore, the highly precise targeting of active cold seeps clasts in extrusive flows. (b) Example of abundant authigenic was not possible, and the likelihood of actually carbonate crusts and shell fields. (c) Fault scarp demonstrating recent tectonism on Kazan mud volcano. sampling an active seep in this manner is low, given an estimated radius around the coordinates of f80 m as a result of a 5% scope (assumed based on known by typical wipeouts in acoustic facies on the subbot- currents in the Rhodes Gyre area) in the cable lower- tom profiler along a scarp about 1 km west of the ing the box core, and the extremely variable terrain on volcano (Fig. 2). and around Kazan mud volcano (see discussion be- J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 373 low). These samples, therefore, likely represent a mid- 4.5. Geochemistry point in the continuum between highly active seep sites, such as those examined on Hydrate Ridge Pore water analyses are described in detail in Haese (Boetius et al., 2000; Elvert et al., 2001) and those et al. (2003). Carbonate mineralogy and geochemistry in areas of ‘‘normal’’ sediment methane levels, such as analyses are described in Aloisi et al. (2000). Lipid the Kattegat (Bian et al., 2001). concentration and carbon isotopic analyses are described in Werne et al. (2002), Pancost et al. (2000, 4.3. Clay Mineralogy 2001a,b), and Hopmans et al. (2000).

The matrix clay mineral composition of the mud 4.6. Microbiology breccia from several mud volcanoes, and of a hemipelagic sample as reference, was defined on the basis Three depth intervals spanning the zone of AOM of X-ray diffraction (XRD). XRD analyses of the <2 were investigated (0–6, 6–22, >22 cm) because of the Am-size fraction were carried out at different percen- larger sample size needed relative to geochemical tages of relative humidity (0–50–100%) in order to analyses. Total counts of bacteria as well as aggregates identify expanding clay minerals in the complex as described by Boetius et al. (2000) were determined mixture. Carbonate was removed and an internal by epifluorescence microscopy using DAPI (4V,6V-dia- standard was added to prepare samples for analysis. midino-2-phenylindole hydrochloride) and acridin or- Semi-quantitative analyses were performed using ange dyes. Most probable numbers (MPN) were the integrated weighted-peak area and the clay determined according to Heijs et al. (1999) using mineral amount was estimated using the samples appropriate culture media for the functional groups of measured at 50% humidity. Results from the matrix organisms determined according to the German Col- of MNLBC19, sampled at 21 cm deep, are presented lection of Microorganisms and Cell Cultures (www. here. dsmz.de). DNA was isolated from sediment samples following Aloisi et al. (2002), slightly modified with 4.4. Macrobiology respect to extraction times and mechanical lysis. DNA isolates were cleaned by gel-filtration (WIZARD DNA Preparation for transmission electronic microscopy clean up kit, Promega Benelux, Leiden, The Nether- (TEM), included fixing small pieces of bivalve gills lands) and concentrated to a final volume of 120 Al. The (Myrtea sp. and Vesicomya sp.) in 3% glutaraldehyde in total amount of DNA was measured spectrophotomet- 0.4 M NaCl, buffering with 0.1 M cacodylate (pH 7.4) rically. Archaeal 16S ribosomal RNA genes were and post-fixing in 1% osmium tetroxide for 1 h in the amplified from the total chromosomal DNA pool by same buffer. The specimens were dehydrated through a using a group specific archaeal primer set as described graded ethanol series and embedded in Araldite. Semi- by DeLong (1992). For the analysis of the microbial thin sections were stained with toluidine blue. Ultra- communities, terminal-Restriction Fragment Length thin sections were contrasted with uranyl acetate and Polymorphism (t-RFLP) was applied using FAM and lead citrate, and observed under a Hitachi H600 trans- HEX-labeled primers in Polymerase Chain Reaction mission electron microscope. (PCR). After gel purification, 100 ng of PCR product j Stable carbon and nitrogen isotope analyses of was cut for 3 h at 37 C using 2 U of HHA1 digestive the tissues were made on samples frozen and stored enzyme, and analyzed on a ABI 310 automated in liquid nitrogen then freeze dried and acidified in sequencer. the laboratory. The acidified samples were com- busted to CO2 and N2 using standard methods 5. Results and discussion (Bouton, 1991) before being passed through a mass spectrometer (Finnigan Delta E). Carbon isotopes are 5.1. Sediment cores reported relative to PeeDee Belemnite (PDB) and nitrogen isotopes relative to atmospheric molecular Box cores discussed herein span the upper f30 cm nitrogen. of sediment. Sediments consist of grey mud breccia 374 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 with millimetric to centimetric mudstone clasts. The quired for reconstruction of a reasonable methane upper few centimeters of sediment are oxidized. In core concentration profile include near-complete methane MNLBC19, from 22 to 34 cm there is evidence that the oxidation by AOM, an insignificant rate of methano- sediment was gas saturated (this interval corresponds to genesis below the depth of AOM, and an insignificant a temperature of 11.9 jC measured on board). rate of sulfate reduction related to organic matter (OM) decomposition. These assumptions are justified be- 5.2. Mud matrix clay mineralogy cause at this site, AOM is restricted to a narrow depth interval (14–18 cm below sediment surface), as can be The clay mineral composition of the mud matrix of seen from the coinciding discrete depth of the sulfide Kazan mud volcano is largely dominated by smectite maximum, the minimum in the d13C of DIC, and the (85%), with 10% kaolinite, 3% palygorskite and 2% depth of sulfate depletion (Fig. 4). Total organic illite. In comparison with the composition of Pliocene carbon and biomarker concentrations and carbon iso- sediments in the Mediterranean Sea (Melieres et al., topic composition (Fig. 4 and Table 2)showthe 1978) and with the hemipelagic core recovered in the imprint by AOM at approximately the same depth as region of the Florence Rise, the matrix of the mud the pore water indicators, and the lack of AOM-related breccia is considerably enriched in smectite. Such biomarkers in the shallowest sediments justify the smectite-rich formation is typically found in the Med- assumption of (near) complete methane oxidation iterranean Sea in Messinian deposits (Chamley et al., within a discrete depth interval. Sulfate reduction 1978), which suggest a possible deeper, Messinian related to the mineralization of OM and methano- source for these mud breccia matrix sediments. On genesis should be negligible due to the low amount the other hand, the high abundance of hydrated clays (0.2–0.3 wt.%, Fig. 4D) and refractory nature of the seems to indicate the absence of deep burial and OM present (most of the OM in these mud breccia subsequent diagenetic transformation, supporting a sediments is associated with the ascending mud ma- shallow source for the matrix. The smectite in the trix, not with pelagic sedimentation, and is therefore Anaximander area could also be derived from the wea- already largely degraded, Werne et al., 2002). The thering of ophiolitic rocks found on Turkey, Cyprus, most convincing argument for a negligible contribu- and probably partly beneath the Anaximander Moun- tion of sulfate reduction related to the mineralization of tains. As discussed below, the predominance of smec- OM comes from a sulfate/DIC* (=DIC+Ca) molar tite clays has implications for the interpretations of pore ratio of 1 in the depth interval where both variables water geochemistry and the occurrence of methane show a concentration gradient (Haese et al., 2003). hydrates. This clearly reflects the stoichiometery of AOM; sulfate reduction related to OM mineralization would 5.3. Pore water geochemistry result in a ratio of 0.5. If our assumptions are valid, according to the net- Profiles of measured pore water geochemical con- stoichiometery of AOM proposed by Reeburgh stituents are shown in Fig. 4. Since methane is highly (1976) (Eq. (1)), the sulfate flux must be exactly volatile, methane gas bubbles form and may migrate balanced by the methane flux (Borowskietal., laterally and upwards during sample retrieval, leading 1996). Once the methane flux is known, the methane to highly variable measured methane concentrations, depth distribution can be reconstructed based on our which may significantly underestimate the in situ con- assumptions with an analytical solution to the well- centrations (Paull et al., 2000). In order to overcome established one-dimensional, steady-state transport this artifact, the in situ methane concentrations were equation for solutes accounting for diffusion and reconstructed in a recent study on methane dynamics advection (e.g. Berner, 1980). The modeled methane (Haese et al., 2003) by using the unambiguous sulfate depth distribution is shown in Fig. 5. profile and the advective flow velocity of 4 cm year1 According to the model results by Haese et al. estimated for site MNLBC19 using best-fit simulations (2003), a steep increase in the methane concentration of measured depth profiles of conservative solutes (Na, is predicted from the depth of AOM (14–18 cm) down B) (Haese et al., 2003). Additional assumptions re- to a depth of about 1 m before the concentration J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 375

Fig. 4. Depth profiles of pore water and sedimentary geochemical constituents. (A) Sulfate (solid) and sulfide (open) concentration. (B) Dissolved inorganic carbon, DIC, concentration (solid) and carbon isotopic composition (open). (C) Calcium concentration (solid) and methane concentration (open). (D) Bulk organic carbon weight percent (solid) and carbon isotopic composition (open). (E) Bulk carbonate concentration (solid) and isotopic composition (open). In A–C, important depth intervals are marked by different shadings and labeled respectively. Sulfate, 13 13 DIC, d C-DIC, Corg, and d C-Corg profiles were redrawn from Haese et al. (2003), and methane was redrawn from Werne et al. (2002). The zone of anaerobic oxidation of methane, AOM, can be identified by the concurrent depth of sulfate consumption and sulfide and DIC production. Carbonate precipitation is predominant where the calcium gradient is steepest. gradient flattens and an asymptotic maximum methane depth of 2 m and below. Using the approach by concentration is reached at approximately 2 m (Fig. 5). Rempel and Buffet (1997), it was further calculated The predicted asymptotic maximum methane concen- that 3.8% of the initial pore space is filled with gas tration is approximately 130 mmol l1, which matches hydrate in 10,000 years (Haese et al., 2003). the methane saturation concentration with respect to From the simulations of the sulfate and the cor- gas hydrate under the local pressure and temperature responding methane concentration profiles, a depth- conditions (Zatsepina and Buffet, 1997). Thus, at site integrated rate of AOM of f6.0 mol m2 year1 was MNLBC19 gas hydrate formation is expected at a derived (Haese et al., 2003). When compared to cold 376 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

Fig. 5. Model results showing: (A) the best-fit simulation to the measured sulfate concentration profile using the analytical solution of the one- dimensional, steady state transport equation accounting for diffusion and advection, (B) the modeled methane concentration profile, (C) the derived rate of AOM, and (D) the modeled methane concentration profile down to 5-m sediment depth (redrawn from Haese et al., 2003). Measured values are shown for comparison in open diamonds. Note the different depth scales. The grey shaded interval represents the zone of bioirrigation. Once the sulfate flux is calculated, the methane profile can be reconstructed with the assumptions discussed in the text. An asymptotic maximum methane concentration is reached below approximately 2 m and is very close the methane saturation concentration. This suggests gas hydrate formation below a depth of 2 m. The rate of AOM is calculated from the combined sulfate and methane models. Note in B the difference between measured and modeled methane concentrations, suggesting migration and outgassing of methane during core recovery. seep sites at Hydrate Ridge (Boetius et al., 2000) and carbonate concentration nor a discrete minimum in the the Aleutian Trench (Wallmann et al., 1997), the rate of d13C of bulk carbonate can be observed (Fig. 5). This AOM at site MNLBC19 is comparably low. However, finding opposes observations made on the bulk organ- AOM rates of all diffusion-dominated sites, i.e. at Saa- ic carbon record and is related to relatively high nich Inlet, Skan Bay, and Skagerrak, are significantly background (associated with ascending mud matrix) lower than at site MNLBC19 (Haese et al., 2003). carbonate concentrations. This means that the identi- fication of paleo-seep environments based on the bulk 5.4. Bulk carbon geochemistry carbon geochemical record may only be possible in sediments poor in background and pelagic bulk inor- Concentrations and carbon isotopic compositions ganic or organic carbon. The effect of masking the of bulk organic carbon and carbonate are shown in AOM imprint on the carbon records may, however, be Fig. 4. The negative isotopic excursion in bulk organ- compensated by high rates of AOM over long time ic carbon is a result of methane derived biomass from periods (Raiswell, 1988). As discussed below, discrete AOM (Fig. 4). Note however, that the negative nodules of authigenic carbonate and individual bio- excursion in the d13C is presumably only noticeable markers are the most appropriate indicators for tracing because the indigenous bulk organic carbon is very carbon transformations based on the geologic record. low in concentration. Based on a stable isotope mass balance, it was calculated that the biomass-carbon 5.5. Authigenic carbonates derived from AOM only makes up 0.05 wt.% carbon of the dry sediment (Haese et al., 2003). Anaerobic oxidation of methane is often associated AOM leads to the formation of 13C-depleted DIC, with the precipitation of authigenic carbonates (Ritger which diffuses upwards along a concentration gradient et al., 1987; Stakes et al., 1999; Pierre et al., 2000; and is consumed by carbonate precipitation. The Aloisi et al., 2000, 2003), which can be considered a required calcium is replenished by the downward flux terminal sink for methane-derived carbon in marine from the bioirrigated zone (Fig. 4). Despite high rates sediments. AOM in eastern Mediterranean mud vol- of carbonate precipitation as reflected by the steep canoes has resulted in the widespread precipitation of calcium concentration gradient, neither enhanced bulk authigenic carbonates and carbonate crusts (MEDI- J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 377

NAUT/MEDINETH Shipboard Scientific Parties, 2000). The mineralogy and stable carbon and oxygen isotope compositions of these carbonates have been discussed in detail by Aloisi et al. (2000), so only data from carbonates occurring on Kazan mud volcano will be summarized herein. Mud volcano-associated authigenic carbonates in the eastern Mediterranean can be divided into several categories based on mineralogy and morphology, including crusts, nodules, and carbonate-cemented sediments (both lithified pelagites and lithified mud breccia sediments). Crusts range in thickness from millimeter scale up to decimeter scale, with variable morphology ranging from mounds to slabs (Figs. 3b and 6) (Aloisi et al., 2000). The authigenic carbonate crusts are composed of dolomite, aragonite, and calcite in varying proportions, ranging from 70% aragonite (Type ‘‘A’’ crusts of Aloisi et al., 2000)to 90% calcite (Type ‘‘D’’), and carbonate nodules are up to 90% dolomite (Aloisi et al., 2000). Both low- Mg and high-Mg (up to 12 mol% Mg) calcites have been identified (Aloisi et al., 2000). Many of the crusts are associated to some extent with macrofauna such as pogonoforan tube worms and bivalves (Figs. 3b and 6; see discussion of macrofauna below). The carbon isotopic composition of authigenic 13 Fig. 6. (a) Photo taken by submersible Nautile of authigenic carbonates (d Ccarb) on Kazan mud volcano ranges carbonate crust MN12BT3 in situ, prior to sampling. (b) Photo of from f0xto nearly 50x(PDB, Fig. 7; data from sample MN12BT3 in the lab after sampling. Aloisi et al., 2000). Typically, carbonate crusts are much more 13C depleted than poorly lithified mud breccia, which is in turn generally more 13C depleted The oxygen isotopic composition of authigenic 18 than unlithified pelagites, due to the formation of fine- carbonates (d Ocarb) ranges from values of 1xto grained authigenic carbonate minerals (Aloisi et al., +1x(PDB) in the case of nodules and unlithified 2000). Controls on cold-seep carbonate d13Care sediments, and from +2xto approximately +4.5x complex mainly due to uncertainties regarding mixing in lithified sediments and crusts (such as sample 18 ratios of different carbon sources having distinct iso- MN12BT3) (Fig. 7). The enriched d Ocarb signal is topic compositions (bottom water DIC, OM, methane- interpreted by Aloisi et al. (2000) to indicate carbonate derived DIC, DIC in deep fluids) and the extent to precipitation from 18O-enriched diagenetic fluids, rath- which carbon isotopes fractionate during AOM (Aloisi er than as a temperature effect. Possible processes 13 18 et al., 2000).Thed Ccarb at Kazan mud volcano producing O-rich fluids include smectite dehydration suggests that methane is the major source of carbonate and gas hydrate destabilization. Pore water chemistry carbon for crust formation and that this source is and clay mineralogy suggest that the former process somewhat diluted, probably by mixing with relatively might dominate here. In other areas of brine seepage in 13C-rich bottom water or rising fluids. OM decompo- the eastern Mediterranean (such as Nadir Brine Lake sition could play a minor role in contributing to the or Napoli mud volcano in the Olimpi field), extreme DIC pool (Luff and Wallmann, 2003), however, given 18O-enrichment of the carbonate crusts is probably due the modeling results of Haese et al. (2003), such an to precipitation from 18O-enriched fluids sourced from effect is expected to be minimal. relic Messinian brines (Aloisi et al., 2000). However, 378 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

Fig. 7. Cross plot of d13C vs. d18O of carbonates and unlithified sediments from Kazan mud volcano. 13C depletion and 18O enrichment of carbonates relative to unlithified sediments and nodules is believed to indicate authigenic production as a byproduct of AOM (or possibly aerobic oxidation of methane). this hypothesis cannot be applied to the 18O-enriched 5.6.1. Organic geochemistry of authigenic carbonates fluids of Kazan mud volcano, since no brines are found Authigenic crust sample MNT12BT3 has been in the Anaximander Mountains due to the absence of characterized as a lithified pelagite (Aloisi et al., the Messinian evaporite formations. 2000). The carbonate-carbon is 13C depleted (to 46.5xVPDB), clearly indicating a methane-de- 5.6. Organic geochemistry rived source (Aloisi et al., 2000). The suite of archaeal biomarkers identified in this crust includes archaeol (in Molecular isotopic indicators of AOM, such as highest abundance, Table 1), hydroxyarchaeol, croce- 13C depleted compounds typically associated with tane, and pentamethylicosane (PMI, see Appendix A methanogens and sulfate reducers, have been identi- for structures). Of these compounds, those that were fied in samples from various modern cold-seep present in sufficient quantities for carbon-isotopic environments (Hinrichs et al., 1999, 2000; Elvert et analysis are extremely 13C depleted, ranging from al., 2000, 2001; Thiel et al., 2001), including eastern 69x(in the case of PMI) to 112x(archaeol, Mediterranean mud volcanoes (Pancost et al., 2000, Table 1), clearly indicating a methane-oxidation source 2001a; Werne et al., 2002; Aloisi et al., 2002),as for these compounds. Alkyl diethers derived from well as in ancient methane seeps (Thiel et al., 1999) sulfate reducing bacteria (SRB) (series I and II of and non-methane seep environments characterized by Pancost et al., 2001b) were also identified in this crust. AOM such as the Kattegatt Strait (Bian et al., 2001) Hopanoid compounds, such as bishomohopanol, were and the water column of the Black Sea (Schouten et also identified, but the low concentrations precluded al., 2001). isotopic measurement (Bouloubassi et al., unpublished J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 379

Table 1 Hopanoid compounds such as diploptene and dip- Biomarkers and carbon isotopic data from crust MNT12BT3 lopterol were also identified with ‘‘intermediate’’ 13 Compound Relative d C degrees of 13C depletion (f55xto 70x), which abundance (xVPDB) were attributed to aerobic methane-oxidizing bacteria Archaeol ++ 112 (Werne et al., 2002) based in part on decreasing down- Hydroxyarchaeol + n.d. core concentration trends. h,h-bishomohopanol was Crocetane + n.d. PMI:0 + 69 also identified in surficial sediments (Werne and Sin- Series I alkyl diethers + to ++ 94 to 99 ninghe Damste´, unpublished results), as in the crust. Series II alkyl diethers + f70 This compound has been attributed to sulfide-oxidizing ‘‘Background’’ biomarkers + to +++ 23 to 32 bacteria in other studies (Pancost and Sinninghe Dam- Series I and II alkyl diethers refer to Pancost et al. (2001b). ste´, 2003). ‘‘Background’’ biomarkers (e.g., hopanoids, fatty acids, steroids, A series of glycerol dialkyl glycerol tetraethers and n-alkanes derived from bacteria and photoautotrophs) are (GDGTs) was also identified in the sediments of Kazan probably incorporated into authigenic carbonates from the ascending mud matrix that was lithified. See Appendix A for compound mud volcano (Werne and Sinninghe Damste´, unpub- structures. ‘‘n.d.’’ indicates no data. lished results). Based on molecular isotopic studies in other mud volcano cold seep sediments (Pancost et al., 2001a,b) and in the water column of the Black Sea results). In other studies, this compound has been (Schouten et al., 2001), it has been proposed that some attributed to bacteria oxidizing sulfide (Pancost and of these GDGT compounds are derived at least in part Sinninghe Damste´, 2003), which seems reasonable in from anaerobic methane-oxidizing archaea. In Kazan this environment as well given the visual observation mud breccia sediments, however, carbon isotope anal- of white filamentous Beggiatoa-like bacterial mats by ysis of C40 isoprenoids released from these GDGTs by submersible (Fig. 8; and MEDINAUT/MEDINETH chemical degradation show 13C enriched values, in the Shipboard Scientific Parties, 2000) and the identifica- range of 22x, indeed the most 13C enriched com- tion of colorless sulfur bacteria in samples from Kazan pounds in the extractable OM (Werne and Sinninghe mud volcano sediment (Heijs, unpublished data, see Damste´, unpublished results). Such values clearly discussion on microbiology below). Other biomarker indicate a non-methane-related source for these com- compounds identified in comparable or greater abun- pounds, likely derived primarily from Crenarchaeota dance than the AOM consortium biomarkers include ubiquitously occurring in the pelagic zone of the water terrestrially derived n-alkanes, sterols, and fatty acids column (Schouten et al., 2000), and are therefore more which are believed to be associated with OM in the ascending mud breccia based on their typical carbon isotope compositions (f24xto 32x, Table 1).

5.7. Organic geochemistry of mud breccia sediments

The molecular isotopic signature in core MNLBC19 taken from the mud breccia of Kazan mud volcano shows similarities to the crust, but also shows some striking differences. For example, 13C depleted (to 123.8x) archaeal compounds such as archaeol, hydroxyarchaeol, and PMI have been identified in the sediments (Table 2, some data from Werne et al., 2002). We have also identified several non-isoprenoid dialkyl glycerol diethers attributed to SRB, but only series II compounds of Pancost et al. (2001b) were identified in Fig. 8. Photo taken by submersible Nautile showing white patches the sediments (compared to dominantly series I in the that were observed on Kazan mud volcano, believed to be crust). filamentous microbial ‘‘mats’’, perhaps Beggiatoa. 380 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

Table 2 Concentration (mg/g sediment) and carbon isotopic composition (xVPDB) of selected biomarkers Depth (cm) Concentration, mg/g sed d13C, xVPDB 1 2.5 11 17 21 28 1 2.5 11 17 21 28 Pristane 0.23 0.36 0.38 0.48 0.36 0.42 29.1 28.8 27.8 27.9 28.3 31.0 n-C17 0.05 0.06 0.06 0.10 0.08 0.07 – 29.7 – 28.5 27.6 – n-C19 0.12 0.11 0.16 0.22 0.13 0.13 29.5 28.5 28.4 – 29.5 29.0 n-C29 0.31 0.23 0.24 0.29 0.17 0.20 29.7 29.8 30.0 29.4 30.3 30.5 n-C31 0.30 0.27 0.30 0.30 0.16 0.21 31.3 30.7 30.9 30.5 31.0 32.7 Hydroxyarchaeol – 0.19 0.20 0.93 0.61 0.29 – 103 107 111 111 101 Archaeol – 0.07 0.07 0.32 0.16 0.10 – 94.0 94.9 102 101 80.5 PMI:3c – – 0.03 0.04 0.02 – – – 86.7 102 123 53.4 Diether 1 – – 0.01 0.06 0.06 0.03 – – – 90.9 88.5 70.0 Diether 2 – – 0.02 0.08 0.12 0.10 – – – 90.4 96.7 82.0 Diether 3 – – 0.04 0.07 0.10 0.03 – – 85.3 90.8 73.3 61.5 Diether 4 – – 0.01 0.06 0.06 0.05 – – – 89.5 94.5 73.2 Diplopterol 0.20 0.14 – – – – 55.5 62.4 –––– Diploptene 0.08 0.06 0.02 0.01 – – 54.1 59.7 –––– Bishomohopanol 0.22 0.16 – – – – 45.7 52.8 –––– GDGT-1302 0.82 0.45 0.28 0.12 0.20 0.18 Biphytanes released from GDGTs 21.5xto 35.3x. GDGT-1300 0.17 0.09 0.06 0.02 0.05 0.05 GDGT-1298 0.17 0.17 0.13 0.08 0.14 0.10 GDGT-1296 0.06 0.07 0.08 0.04 0.07 0.05 GDGT-1294 0.00 0.00 0.06 0.04 0.06 0.04 Pristane and n-alkanes are matrix-associated organic matter; archaeol, hydroxyarchaeol, and PMI:3c (a tri-unsaturated pentamethylicosane) are derived from anaerobic methane-oxidizing archaea; diethers are derived from syntrophic sulfate reducing bacteria (SRB); hopanoids are from aerobic bacteria, and glycerol diether glycerol tetraethers (GDGTs) are mostly matrix associated, but with some contribution from anaerobic oxidation of methane (AOM). likely to be associated with the ascending mud matrix currently underway to define these differences more than with in situ microbial processes. rigorously by means of the eubacterial and archaeal Numerous indicators of the ‘‘background’’ OM (i.e., 16S rDNA genes. that OM that is associated with the ascending mud Some further information is available, however, matrix) such as n-alkanes (d13C values of f31x) through total bacterial counts as well as counts of the were identified at relatively high abundance in the mud aggregate clusters composed of methane-oxidizing breccia sediments (Werne et al., 2002). The higher Archaea and sulfate-reducing bacteria (cf., Boetius proportion of non-AOM related OM is to be expected et al., 2000). Epifluorescence microscopy utilizing in mud breccia sediments or lithified pelagite carbo- both the DAPI and Acridin Orange methods con- nates compared to more active seep vents or pure firmed the existence of structured consortia, probably carbonate crusts, simply due to dilution. consisting of anaerobic methane-oxidizing archaea and SRB (Boetius, personal communication) (Fig. 5.8. Microbiology 10). Aggregate concentrations in the mud breccia sediments of Kazan mud volcano range from Preliminary results of terminal restriction fragment 1.79107 to 4.2108 cells/cm3, however, the counts length polymorphism (t-RFLP) analyses of 16S rDNA obtained by two methods differed by an order of genes indicate that the microbial communities, though magnitude (Table 3). It is possible that this simply similar, do display definite variability with depth based reflects a difference in the methods (e.g., DAPI on the presence or absence of specific phylotypes in a reflects more ‘‘viable’’ cell counts, and Acridin Or- given community (Fig. 9). Further experiments are ange more ‘‘total’’ counts). It is also possible that

Fig. 9. t-RFLP chromatograms of bacterial and archaeal communities in sediments from Kazan mud volcano, illustrating qualitative differences in microbial community structure at different sediment depths. J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 381 382 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

Fig. 10. Archaea/SRB aggregates identified using fluorescence in situ hybridization (FISH) and stained with DAPI. Archaea are shown here in green and SRB are shown in red.

DAPI numbers may be more reliable, due in part to As these consortia have proven difficult to culture (cf. the very low quantities of DNA that were extractable Boetius et al., 2000; Michaelis et al., 2002), the from these sediment samples. second hypothesis seems more likely. Another interesting feature of the microbiological data is found in the most probable number (MPN) counts. Data from MPN counts indicate the likely Table 4 existence of colorless sulfur bacteria (e.g. sulfide Most probable number (MPN) counts oxidizers) and methanogens in sediments from Kazan Dilution Water 0–6 cm 6–22 cm >22 cm mud volcano above 22 cm depth, but give no indica- WI WII WIII SI SII SIII SI SII SIII SI SII SIII tion of the existence of SRB or aerobic methane Sulfate-reducing bacteria oxidizers (Table 4). These results suggest a number 100 /// /// /// of possibilities. First, it is not surprising to see 1000 evidence of colorless sulfur bacteria in the MPN 10,000 100,000 counts given the visual observation of ‘‘white patches’’ by submersible dives (Fig. 8) which are Colorless sulfur bacteria believed to be sulfide oxidizers similar to Beggiatoa. 100 +++ +++ +++ / / / / / / / / / Second, it is possible that there are no SRB in these 1000 +++ +++ sediments, but this seems very unlikely given the 10,000 +++ 100,000 biomarker and pore water geochemical data discussed. Alternatively, it is possible that the SRB that are living Methanogens syntrophically with the anaerobic methane-oxidizing 100 /// /// /// archaea are not culturable under the conditions uti- 1000 x + lized, and therefore do not appear in the MPN counts. 10,000 + x 100,000

Table 3 Methane-oxidizing bacteria Microbial cell and aggregate counts (DAPI and Acridin Orange) 100 + /// /// /// from core MNLBC18/19 1000 10,000 Sample DAPI counts Acridin Orange Aggregate counts 100,000 interval (cells/cm3) counts (cells/cm3) (aggregates/cm3) (cm) Symbols used are: () indicates not present; (/) indicates not analyzed due to pre-dilution of 100 times; (+), (++), and (+++) 0–6 2.14107 4.2108 few 7 8 indicate the relative level of likely concentration; (x) indicates 6–22 1.7910 2.610 few sample contaminated by oxygen. WI, WII, WIII and SI, SII, SIII >22 2.99107 1.3108 F106 3 indicate water and sediment samples used for culture work, run in aggregates/cm triplicate. J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 383

5.9. Chemosynthetic macrofauna Isorropodon perplexum (Olu-Le Roy et al., 2001, in press; von Cosel and Salas, 2001) and a small mytilid Dense communities composed of small bivalves and attached to carbonate crusts and identified as Idas vestimentiferan tube worms were discovered on east- modiolaeformis (Olu-Le Roy et al., 2001, in press). ern Mediterranean mud volcanoes (Fig. 11, cf. Fig. 2b Some tubes of living worms belonging to Lamellibra- in MEDINAUT/MEDINETH Shipboard Scientific chia sp. (Olu-Le Roy et al., 2001, in press) were also Party, 2000). During the 1998 MEDINAUT cruise, collected at the site. On top of a strongly reduced living individuals of several species were sampled by sediment, white-grey bacterial mats (Fig. 8) were submersible from the top of Kazan mud volcano observed as well as many echinidae sea urchins and a (35j25.983VN, 24j33.594VE, 1707 m water depth; number of unidentified crabs. Fiala-Me´dioni et al., 2001; Olu-Le Roy et al., in press). TEM studies of the gills of Myrtea sp. and Vesico- Among the organisms collected were lucinid bivalves: mya sp., as well as trophosomes of vestimentiferan the most abundant species was identified as Myrtea sp. worms demonstrated that they were composed of (Olu-Le Roy et al., 2001, in press); few samples of bacteriocytes housing sulfide-oxidizing bacteria of another species were found and described as Lucinoma various morphological types as found in other symbi- kasani (Salas and Woodside, 2002).Otherliving otic organisms from vents or cold seeps (Fiala-Me´di- bivalves included a small vesicomyid, identified as oni and Felbeck, 1990; Fisher, 1995). In contrast, the mytilid gill houses both sulfide-oxidizing and meth- anotrophic symbionts (Fiala-Me´dioni et al., 2001) as found in Bathymodiolus puteosrpentis (Cavanaugh et al., 1992) or B. azoricus (Fiala-Me´dioni et al., 2002). Carbon isotopic compositions of the gills of bivalves also suggested chemosynthesis-based metabolism, ranging from values 27.7 xto 30.5 xfor M. sp. and I. perplexum and from 23.6xto 26.6x for the tube worm Lamellibrachia sp. (Fiala-Me´dioni et al., 2001); these values are in the range of those found for other sulfide-oxidizing symbionts in different vent or cold seep sites (Fiala-Me´dioni and Felbeck, 1990; Kennicutt et al., 1992; Fisher, 1995). The values measured on the Kazan mud volcano mytilid I. modiolaeformis are more depleted than those of the lucinid, ranging from 44.2xto 44.6x (Fiala-Me´dioni et al., 2001). Such values reflect the presence of methylotrophic symbionts which may use methane as an energy source as found for B. childressi, a cold seep species collected off Louisiana (Nelson and Fisher, 1995). TEM and isotopic results led to the con- clusion that these species are mainly relying, through their symbionts, on reduced sulfur produced in the sediment, with the exception of the mytilid which is probably able to use both sulfide produced in the sediment and/or methane expelled in the fluids (Olu-Le Roy et al., in press). The chemosynthetic macro- and megafauna are located at sites where fluids are expelled (Olu et al., Fig. 11. Photo taken by submersible Nautile showing types of chemosynthesis-dependent macrofauna observed on Kazan mud 1997; Sibuet and Olu, 1998), but there is currently no volcano. (a) Vestimentiferan tube worms. (b) Lucinid, vesicomyid, statistical evidence for any direct relationship between and mytilid bivalves. methane concentrations and living biomass. With the 384 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 exception of the mytilid, such a relationship appears MPN response for aerobic methane oxidizers in even indirect for Kazan mud volcano species where no the most surficial sediments of Kazan. The lack of mi- methylotrophic-like symbionts were observed. Within crobiological evidence for aerobic methane oxidizers in Mediterranean mud volcano fields, it has been docu- the sediments further suggests that there are no aerobic mented that concentrations of methane increase with organisms utilizing methane escaping the AOM zone depth in the uppermost meter of the mud breccias. In (Werne et al., 2002), despite the presence of aerobic most cases, this appears to be associated with the depth methane-oxidizer biomarkers (diploptene and di- zone of sulfate reduction (de Lange and Brumsack, plopterol) in near surface sediments. An alternative 1998) and results in the generation of inorganic carbon explanation is that these compounds are derived from and hydrogen sulfide which can be used as a substrate an early, rapid phase of aerobic methane oxidation that by the sulfide oxidizing symbionts of the macrofauna. took place immediately after mud breccia emplacement rather than from presently active bacteria. This hypothesis is more attractive given that microbiological data 6. Summary and biogeochemical synthesis indicates the absence ofmethane-oxidizingbacteriacur- rently living in Kazan mud volcano surficial sediments. AOM is a significant biogeochemical process in It is also conceivable, however, that the biomarkers many cold-seep sediments, evident in everything from attributed to aerobic methane oxidizing bacteria by pore water and sedimentary geochemical profiles to Werne et al. (2002) could be derived from other authigenic carbonates to (chemosynthetic) macrofau- bacteria in this environment. Indeed, these biomarkers nal abundance and diversity. In sediments from Kazan are fairly general bacterial biomarkers, and there is mud volcano, measured and modeled pore water strong evidence for the existence of colorless sulfur profiles of the concentrations of sulfate, sulfide, dis- bacteria, both in MPN counts (Table 4) as well as in 13 solved inorganic carbon (DIC) and d CDIC all indi- visual observations of white patches (which may have cate that AOM is presently occurring in the sediments been Beggiatoa-like white filamentous bacteria) made of Kazan mud volcano (Fig. 4, Haese et al., 2003). It by Nautile submersible dives (MEDINAUT/MEDI- is clear from these measured and modeled pore water NETH Shipboard Scientific Parties, 2000). In addition, profiles that sulfate and methane are consumed in the other biomarkers identified in Kazan mud volcano same depth interval, coincident with the production of sediments and other cold seep environments, such as 13 13 sulfide and C-depleted DIC. Indeed, d CDIC values bishomohopanol, have been attributed to sulfide-oxi- of 35x(PDB) require the oxidation of methane to dizing bacteria (this study; Pancost et al., 2002). CO2 (Whiticar and Faber, 1986; Alperin et al., 1988; Under such conditions of complete (or near-com- Whiticar, 1999). 13C depleted authigenic carbonates plete) methane oxidation, only small amounts of meth- and TOC are identified in the zone where AOM is ane are released into the bottom waters. This release is believed to be occurring based on pore water profiles typically associated with major faults where highest (Figs. 4–7). Further evidence of AOM is found in 13C methane fluxes (due to advective flow) can be expec- depleted biomarkers of methane-oxidizing archaea and ted. This aspect of fluid fluxes and methane oxidation SRB (Table 2). Visual observation of microbial con- processes explains why bottom-water methane concen- sortium aggregates through FISH (Fig. 10) and che- trations are so variable over mud volcanoes in the mosynthetic macrofauna such as bivalves and tube eastern Mediterranean and the highest methane con- worms (Fig. 11) confirm the existence of organisms centrations in the deep Mediterranean are identified that are dependent on methane and the byproducts of above Amsterdam mud volcano (Charlou et al., 2003), AOM (such as the sulfide utilized by bacterial sym- which is characterized by intense active tectonism. (Ac- bionts dwelling in the tissues of the macrofauna). tive) faults also serve as preferred fluid pathways The porewater (sulfate, sulfide, DIC) and solid (focused flow), which leads to a high spatial heteroge- phase (bulk OM, and selected biomarkers) geochemical neity of solute fluxes, fluid flow, and microbial and ma- profiles suggest that this process is restricted to a crobenthic abundances, as identified through visual narrow depth interval, where methane is completely observations made during dives of the submersible oxidized. This hypothesis is supported by the lack of Nautile. J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 385

The production of authigenic carbonates as a by- isotopic composition of C40 isoprenoids released from product of AOM represents one of the major terminal GDGTs of Kazan mud volcano are up to 22x, 13 sinks for methane-derived carbon in cold seep envi- significantly C-depleted C40 isoprenoids were iden- ronments. This process also keeps methane out of the tified in authigenic carbonate crusts from Olimpi mud atmosphere thereby reducing the impact of methane volcano (Bouloubassi et al., unpublished results), as on global climate change. The high variability in the well as in seeps from Milano and Amsterdam mud carbon isotopic composition of these carbonates can volcanoes, and crusts, bacterial mats, seeps, and brines be explained by the fact that: (1) they form both as a from Napoli mud volcano (Pancost et al., 2001a). result of chemical supersaturation and microbial cal- These contrasting results in different cold seep loca- cification (Aloisi et al., 2000, 2002); and (2) slight tions, even within eastern Mediterranean mud volca- variations in the depth of precipitation of carbonates noes, strongly suggest that the specific microbial within the sediments will substantially alter their communities vary depending on the geochemistry of isotopic signature due to the steep gradients in d13C the environment, including factors such as methane of the available DIC (Aloisi et al., 2000, 2002). One concentration and flux, and salinity. implication of these observations is that deriving rates The archaeal/SRB aggregates (as determined by the of carbonate formation from pore water gradients may Acridin Orange method) show the greatest abundance underestimate the true calcification rate due to the ad- in the deepest sediment layer analyzed (>22 cm). This ditional contribution from microbes and the very steep contrasts with the pore water geochemistry and solid and likely variable gradients in the concentrations and phase bulk geochemical data, which suggests present- carbon isotope composition of DIC resulting from day maximum AOM rates in the region around 13–15 AOM. cm depth. The lipid profiles, however, display a max- Series I diethers are higher in concentration and imum in AOM indicators (e.g., 13C depleted specific more 13C depleted than series II diethers (ca. 97x biomarkers) at an intermediate sediment depth of 16– compared to 70x; Table 1) in the authigenic 18 cm. The discrepancies in the geochemical and carbonate investigated here. This pattern differs from microbiological data suggest that the historical center that observed in carbonates from Amsterdam mud of AOM in this environment may have migrated volcano (also in the Anaximander Mountains area) vertically in the sediments, likely as a result of varia- where the two series are approximately equivalent in tions in the rate of the methane flux. While such a hy- abundance, as well as to authigenic carbonates from pothesis is not unlikely, it should be noted that the Napoli mud volcano (Olimpi area), where series I microbiological analyses were taken on a separate adja- diethers are lower in concentration than series II cent core, so slight variations in the depth of AOM are diethers (Bouloubassi, unpublished results). Further- expected due to the lateral spatial variability in all more, in mud breccia sediments from Kazan mud aspects of a mud volcano, from fluid venting to surface volcano, only series II diethers were identified. These topography. Furthermore, it is generally acknowledged observations suggest that different SRB could be par- that lipid profiles and bulk OM provide a more ‘‘time- ticipating in the microbial consortium with the archaeal averaged’’ view of the sedimentary processes, while methane-oxidizers in different environments, possibly pore water geochemistry is thought to give more of a as a result of the different salinities of the two environ- ‘‘snapshot’’ view of present-day conditions. DNA has ments. Furthermore, different SRB may be active been observed to degrade on time-scales from 10 min to syntrophic partners with archaea in seep areas charac- 100,000 years, depending on specific circumstances. terized by carbonate precipitation compared to areas These differences in the interpretable time frame of the not precipitating carbonate within the same mud vol- different geochemical and microbiological indicators cano environment. can often make interpretation problematical. Additional support for the hypothesis that very Model calculations estimate a rate of AOM of f6 different microbial consortia are carrying out AOM in mol m2 year1. The advective flow velocity at Kazan different environments is found in the differences mud volcano was found to be 4 cm year1. Compared between GDGT compounds found on Kazan mud to other cold seep sites, the advective flow velocity and volcano and those identified elsewhere. While the the rate of AOM are rather moderate (Haese et al., 386 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

2003). Based on the modeled methane concentration and E. Panoto for technical assistance, and D. Saint profile and the state of methane saturation with respect Hilaire and B. Rivie`re for assistance with TEM sec- to gas hydrate, hydrate formation is expected at a tioning and inclusion. Financial support was provided sediment depth of 2 m and below at site MNLBC19. by the Dutch funding organization NWOALW Whereas gas hydrate stability is thermodynamically (MEDINAUT, MEDINETH, and MEDISED proj- favorable on Kazan mud volcano, the rate of hydrate ects) and the French funding agencies IFREMER, formation is rather slow (Haese et al., 2003). These UPMC (Paris 6) and CNRS (UMR 7621). This findings may best explain the rare occasions of gas manuscript benefited from reviews by two anony- hydrate recoveries during two expeditions to Kazan mous reviewers. [LW] mud volcano.

Appendix A. Biomarker structures Acknowledgements List of abbreviations and acronyms We would like to thank the Captain and crew of the MEDINAUT Joint Dutch–French research expedi- R/V Nadir, the submersible Nautile, and the R/V tion to MEDiterranean mud volcanoes using Professor Logachev for their tireless assistance during the submersible NAUTile in 1998 sampling. We also thank M. Kienhuis, E. Hopmans, MEDINETH Dutch follow-up expedition to MEDI-

Methane oxidizing archaeal biomarkers dialkyl glycerol diethers hopanoids X’ OH O Series I, Y = OH X OY Series II, Y = O OY ' diplopterol Both series,Y’ = OH Archaeol: X = H, X’ = H or sn-3-hydroxyarchaeol: X = H, X’ = OH

n = 14, 15, or 16 diploptene OH SRB diether 1 pentamethylicosane (PMI) O

O OH Crocetane β,β-bishomohopanol

HO Glycerol dialkyl glycerol tetraethers (GDGTs) OO X Eukaryotic biomarker

O Y O OH or X and Y = HO OH O X O HO tetrahymanol

OYO J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 387

NAUT, focused on higher resolution geo- Berner, R.B., 1980. Early Diagenesis: A Theoretical Approach. physical surveys and sediment sampling in Princeton Univ. Press, Princeton, NY, p. 241. Bian, L., Hinrichs, K., Xie, T., Brassell, S.C., Iversen, N., Fossing, 1999 (MEDiterranean NETHerlands) H., Jorgensen, B.B., Hayes, J.M., 2001. Algal and archaeal SBP Sub-Bottom Profiler polyisoprenoids in a recent marine sediment: molecular isotopic DAPI 4V,6V-DiAmidino-2-PhenylIndole evidence for anaerobic oxidation of methane. Geochem. Geo- hydrochloride phys. Geosyst. 2 (1), doi:10.1029/2000GC000112. MPN Most Probable Numbers Boetius, A., Ravenschlag, K., Schubert, C.J., Rickert, D., Widdel, F., Gieseke, A., Amann, R., Jørgensen, B.B., Witte, U., Pfann- PCR Polymerase Chain Reaction kuche, O., 2000. A marine microbial consortium apparently me- HPLC High Performance Liquid Chromatography diating anaerobic oxidation of methane. Nature 407, 623–626. t-RFLP terminal-Restriction Fragment Length Borowski, W.S., Paull, C.K., Ussler III, W. 1996. Marine pore water Polymorphism sulfate profiles indicate in situ methane flux from underlying gas ANAXIPROBE Two phase research program carried hydrate. Geology 24, 655–658. Bouton, T.W., 1991. Stable carbon isotope ratios of natural materi- out in 1995 and 1996 to investigate the als: I. Sample preparation and mass spectrometric analysis. In: origin and evolution of the Anaximander Coleman, D., Fry, B. (Eds.), Carbon Isotope Techniques. Aca- mountains demic Press, San Diego, pp. 155–172. AOM Anaerobic Oxidation of Methane Brown, K.M., 1990. The nature and hydrogeologic significance of DIC Dissolved Inorganic Carbon mud diapers and diatremes for accretionary systems. J. Geo- phys. Res. 95 (B6), 8969–8982. FISH Fluorescence in situ Hybridization 13 Cavanaugh, C.M., Wirsen, C., Jannasch, H.J., 1992. Evidence for d Cx stable carbon isotope composition of sub- methylotrophic symbionts in a hydrothermal vent mussel (Bival- stance x via: Mytilidae) from the Mid-Atlantic Ridge. Appl. Environ. (V)PDB (Vienna) Pee Dee Belemnite, international Microbiol. 58, 3799–3803. standard against which carbon and oxygen Chamley, H., Dunoyer-de-Segonzac, G., Melieres, F., 1978. Clay minerals in Messinian sediments of the Mediterranean Sea. isotope ratios are calibrated DSDP 42, 389–395. PMI Pentamethylicosane Charlou, J.L., Donval, J.P., Zitter, T., Roy, N., Jean-Baptiste, P., SRB Sulfate-Reducing Bacteria Foucher, J.P., Woodside, J., MEDINAUT Scientific Party, GDGT Glycerol Dialkyl Glycerol Tetraether 2003. Evidence of methane venting and geochemistry of brines C Carbon chain with 40 carbons on mud volcanoes of the eastern Mediterranean Sea. Deep-Sea 40 Res., Part 1, Oceanogr. Res. Pap. 50, 941–958. TOC Total Organic Carbon Chaumillon, E., Mascle, J., 1997. From foreland to fore-arc TEM Transmission Electron Microscopy domains: new multichannel seismic reflection survey of the OM Organic Matter Mediterranean Ridge accretionary complex (Eastern Mediterra- nean). Mar. Geol. 138, 237–259. Cita, M.B., Ryan, W.B.F., Paggi, L., 1981. Prometheus mud brec- References cia. An example of shale diapirism in the Western Mediterra- nean Ridge. Ann. Ge´ol. Pays Hell. 30, 543–569. Aloisi, G., Pierre, C., Rouchy, J.-M., Foucher, J.-P., Wood- Cita, M.B., Camerlenghi, A., Erba, E., McCoy, F.W., Castadrori, D., side, J., and the MEDINAUT Shipboard Scientific Party, Cazzani, A., Guasti, G., Gambastiani, M., Lucchi, R., Nolli, V., 2000. Methane-related authigenic carbonates of eastern Med- 1989. Discovery of Mud diapirism on Mediterranean Ridge. A iterranean Sea mud volcanoes and their possible relation to preliminary report. Boll. Soc. Geol. Ital. 108, 537–543. gas hydrate destabilisation. Earth Planet. Sci. Lett. 184, Cita, M.B., Ivanov, M.K., Woodside, J.M., 1996. The Mediterra- 321–338. nean ridge diapiric belt. Mar. Geol. 132, 1–6. Aloisi, G., Bouloubassi, I., Heijs, S.K., Pancost, R.D., Pierre, C., Cronin, B.T., Ivanov, M.K., Limonov, A.F., Egorov, A., Akhmanov, Sinninghe Damste´, J.S., Gottschal, J.C., Forney, L.J., Rouchy, G.G., Akhmetjanov, A.M., Kozlova, E., TTR-5, S.S.P., 1997. J.-M., 2002. CH4-consuming microorganisms and the formation New discoveries of mud volcanoes on the Eastern Mediterra- of carbonate crusts at cold-seeps. Earth Planet. Sci. Lett. 203, nean Ridge. J. Geol. Soc. (Lond.) 154, 173–182. 195–203. de Lange, G.J., Brumsack, H.J., 1998. The occurrence of gas Alperin, M.J., Reeburgh, W.S., Whiticar, M.J., 1988. Carbon hydrates in Eastern Mediterranean mud dome structures as in- and hydrogen isotope fractionation resulting from anaero- dicated by pore-water composition. Gas Hydrates: Relevance to bic methane oxidation. Global Biogeochem. Cycles 2, World Margin Stability and Climate Change. Spec. Publ.-Geol. 279–288. Soc., vol. 137, pp. 167–175. Ben-Avraham, Z., Kempler, D., Ginzburg, A., 1988. Plate tectonics DeLong, E.F., 1992. Archaea in coastal marine environments. Proc. in the Cyprus Arc. Tectonophysics 146, 231–240. Natl. Acad. Sci. U. S. A. 89, 5685–5689. 388 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

Dickens, G.R., O’Neil, J.R., Rea, D.K., Owen, R.M., 1995. Disso- Hopmans, E.C., Schouten, S., Pancost, R.D., van der Meer, M.T.J., ciation of oceanic methane hydrate as a cause of the carbon Sinninghe Damste´, J.S., 2000. Analysis of intact tetraether lipids isotope excursion at the end of the Paleocene. Paleoceanography in archaeal cell material and sediments by high performance 10, 965–971. liquid chromatography/atmospheric pressure chemical ioniza- Elvert, M., Suess, E., Greinert, J., Whiticar, M.J., 2000. Archaea tion mass spectrometry. Rapid Commun. Mass Spectrom. 14, mediating anaerobic methane oxidation in deep-sea sediments at 585–589. cold seeps of the eastern Aleutian subduction zone. Org. Geo- Hovland, M., 2000. Are there commercial deposits of marine chem. 31, 1175–1187. hydrates and ocean sediments?. Energy Explor. Exploit. 18, Elvert, M., Greinert, J., Suess, E., Whiticar, M., 2001. Carbon iso- 339–347. topes of biomarkers derived from methane-oxidizing microbes at Kennicutt, M.C., Burke Jr., R.A., MacDonald, I.R., Brooks, J.M., Hydrate Ridge, Cascadia convergent margin. In: Paull, C., Dil- Denoux, G.J., Macko, S.A., 1992. Stable isotope partitioning in lon, W. (Eds.), Natural Gas Hydrates: Occurrence, Distribution, seep and vent organisms: chemical and ecological significance. and DetectionGeophys. Monogr., vol. 124. Am. Geophys. Un., Chem. Geol. 101, 293–310. pp. 115–129. Limonov, A.F., Woodside, J.M., Cita, M.B., Ivanov, M.K., 1996. Emeis, K.C., Robertson, A.H.F., Richter, C., 1996. Proc. Ocean The Mediterranean Ridge and related mud diapirism: a back- Drill. Program, Initial Rep., vol. 160. Ocean Drilling Program, ground. Mar. Geol. 132, 7–20. College Station, TX. 972 pp. Luff, R., Wallmann, K., 2003. Fluid flow, methane fluxes, carbon- Fiala-Medioni, A., Felbeck, H., 1990. Autotrophic process in inver- ate precipitation and biogeochemical turnover in gas hydrate- tebrate nutrition: bacterial symbiosis in bivalve molluscs. In: bearing sediments at Hydrate Ridge, Cascadia Margin: numer- Kinne, R.K.H., Kinne-Saffran, E., Beyenbach, K.W. (Eds.), An- ical modeling and mass balances. Geochim. Cosmochim. Acta imal Nutrition and Transport Processes: 1. Nutrition in Wild 67, 3403–3421. and Domestics Animals, Comparative Physiology, vol. 5. Mascle, J., Huguen, C., Benkhelil, J., Chamot-Rooke, N., Chaumil- Karger, Basel, pp. 49–59. lon, E., Foucher, J.-P., Griboulard, R., Kopf, A., Lamarche, G., Fiala-Me´dioni, A., Sibuet, M., Gottschal, J.C., Mariotti, A., 2001. Volkonskaia, A., et al., 1999. Images may show start of Euro- Chemosynthesis-based communities associated with fluids in pean–African plate collision. Eos, Transactions, American Geo- deep mediterranean mud-volcanoes. Proceedings, CIESM Con- physical Union, 80, 37, 421, 425, 428. ference. Monte Carlo, Sept. 24–28. McClusky, S., Balassanian, S., Barka, A., Demir, C., Ergintav, S., Fiala-Medioni, A., McKiness, Z.P., Dando, P., Boulegue, J., Mari- Georgiev, I., Gurkan, O., Hamburger, M., Hurst, K., Kahle, H., otti, A., Alayse-Danet, A.M., Robinson, J.J., Cavanaugh, C.M., Kastens, K., Kekelidze, G., King, R., Kotzev, V., Lenk, O., 2002. Ultrastructural, biochemical, and immunological charac- Mahmoud, S., Mishin, A., Nadariya, M., Ouzounis, A., Para- terization of two populations of the mytilid mussel, Bathymo- dissis, D., Peter, Y., Prilepin, M., Reilinger, R., Sanli, I., Seeger, diolus azoricus, from the Mid-Atlantic Ridge. Evidence for a H., Tealeb, A., Toksoz, M.N., Veis, G., 2000. Global Positioning dual symbiosis. Mar. Biol. 141, 1035–1043. System constraints on plate kinematics and dynamics in the Fisher, C.R., 1995. Toward an appreciation of hydrothermal vent eastern Mediterranean and Caucasus. J. Geophys. Res., B: Solid animals: their environment, physiological ecology, and tissue Earth 105/3, 5695–5719. stable isotope values. Seafloor Hydrothermal Systems: Physical, McKenzie, D.P., 1972. Active tectonics of the Mediterranean re- Chemical, Biological, and Geological Interactions. American gion. Geophys. J. R. Astron. Soc. 30, 109–185. Geophysical Union, 297–316. MEDINAUT/MEDINETH Shipboard Scientific Parties, D.P., 2000. Haese, R.R., Meile, C., Van Cappellen, P., De Lange, G.J., 2003. Linking Mediterranean brine pools and mud volcanism. EOS, Carbon geochemistry of cold seeps: methane fluxes and trans- Trans. Am. Geophys. Un. 81, 625–632. formation in sediments from Kazan mud volcano, eastern Med- Melieres, F., Chamley, H., Coumes, F., 1978. X-ray mineralogy iterranean Sea. Earth Planet. Sci. Lett. 212, 361–375. studies, Leg 42A, deep sea drilling project, Mediterranean Heijs, S.K., Jonkers, H.M., van Gemerden, H., Schaub, B.M., Stal, Sea. DSDP 42, 361–383. L.J., 1999. The buffering capacity towards sulfide in sediments Michaelis, W., Seifert, R., Nauhaus, K., Treude, T., Thiel, V., Blu- of a coastal lagoon (Bassin D’Arcachon France): relative impor- menberg, M., Knittel, K., Gieseke, A., Peterknecht, K., Pape, T., tance of chemical and biological processes. Estuar., Coast. Shelf Boetius, A., Amann, R., Jorgensen, B.B., Widdel, F., Peckmann, Sci. 49, 21–35. J., Pimenov, N., Gulin, M., 2002. Microbial reefs in the Black Henriet, J.P., Mienert, J. (Eds.), 1998. Gas Hydrates: Relevance to Sea fueled by anaerobic oxidation of methane. Science 297, World Margin Stability and Climate Change. Special Publica- 1013–1015. tion-Geological Society, vol. 137. 338 pp. Milkov, A.V., 2000. Worldwide distribution of submarine mud vol- Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewer, P.G., Delong, canoes and associated gas hydrates. Mar. Geol. 167, 29–42. E.F., 1999. Methane-consuming archaebacteria in marine sedi- Nelson, D.C., Fisher, C.R., 1995. Chemoautotrophic and Methano- ments. Nature 398, 802–805. trophic Endosymbiotic Bacteria at Deep-Sea Vents and Seeps. Hinrichs, K.-U., Summons, R.E., Orphan, V., Sylva, S.P., Hayes, In: Karl, D. (Ed.), Microbiology of Deep-Sea Hydrothermal J.M., 2000. Molecular and isotopic analysis of anaerobic meth- Vents. CRC Press, Boca, pp. 125–166. ane-oxidizing communities in marine sediments. Org. Geochem. Olu, K., Lance, S., Sibuet, M., Fiala-Me´dioni, A., Dinet, A., 1997. 31, 1685–1701. Spatial distribution of cold seep communities as indicators of J.P. Werne et al. / Chemical Geology 205 (2004) 367–390 389

fluid expulsion patterns through mud volcanoes on the Barbados Reeburgh, W.S., Whalen, S.C., Alperin, M.J., 1993. The role of accretionary prism. Deep-Sea Res., Part 1, Oceanogr. Res. Pap. methylotrophy in the global methane budget. In: Murrell, J.C., 44, 811–841. Kelley, D.P. (Eds.), Microbial Growth on C-1 Compounds. In- Olu-Le Roy, K., Sibuet, M., Levitre, G., Gofas, S., Fiala-Me´dioni, tercept, Andover, UK, pp. 1–14. A., Vacelet, J., and the Medinaut scientific shipboard party (J.P. Reilinger, R.E., McClusky, S.C., Oral, M.B., King, R.W., Toksoz, Foucher, J.W., J. Mascle, G. De Lange, G. Aloisi, J.L. Charlou, M.N., 1997. Global positioning system measurements of pres- J.P. Donval, R. Haese, P. Henry, S. De Lint, M. van der Maarel, ent-day crustal movements in the Arabia–Africa–Eurasia G. Nobbe, C. Pierre, R. Pancost), 2001. Cold seep communities plate collision zone. J. Geophys. Res., B: Solid Earth 102, in the deep Mediterranean Sea (South of Crete and Turkey): 9983–9999. faunal composition and spatial distribution. CIESM Congress Rempel, A.W., Buffett, B.A., 1997. Formation and accumulation Proceedings. Monte Carlo, Sept. 24–28, pp. 36, 40. of gas hydrate in porous media. J. Geophys. Res. 102, Olu-Le Roy, K., Sibuet, M., Levitre, G., Gofas, S., Salas, C., 10151–10164. Fiala-Me´dioni, A., Fouchet J.P., Woodside J., in press. Cold Ritger, S., Carson, B., Suess, E., 1987. Methane-derived authi- seep communities in the deep eastern Mediterranean Sea: genic carbonates formed by subduction-induced pore-water ex- composition and spatial distribution on mud volcanoes. Deep pulsion along the Oregon/Washington margin. GSA Bull. 98, Sea Research. 147–156. Orphan, V.J., House, C.H., Hinrichs, K.-U., McKeegan, K.D., Robertson, A.H.F., 1998. Tectonic significance of the Eratosthenes DeLong, E.F., 2001. Methane-consuming archaea revealed by Seamount: a continental fragment in the process of collision directly coupled isotopic and phylogenetic analysis. Science with a subduction zone in the eastern Mediterranean (Ocean 293, 484–487. Drilling Program Leg 160). Tectonophysics 298, 63–82. Pancost, R.D., Sinninghe Damste´, J.S., 2003. Carbon isotopic com- Robertson, A.H.F., Emeis, K.C., Carmenlenghi, A. (Eds.), 1998. positions of prokaryotic lipids as tracers of carbon cycling in Proceedings ODP, Scientific Results TX (Ocean Drilling Pro- diverse settings. Chem. Geol. 195, 29–58. gram), College station. 817 pp. Pancost, R.D., Sinninghe Damste´, J.S., de Lint, S., van der Maarel, Rotstein, Y., Kafka, A.L., 1982. Seismotectonics of the southern M.J.E.C., Gottschal, J.C., and the MEDINAUT Shipboard Sci- boundary of Anatolia, eastern Mediterranean region: subduction, entific Party, 2000. Biomarker evidence for widespread anaerobic collision and arc jumping. J. Geophys. Res., B 87, 7694–7706. methane oxidation in Mediterranean sediments by a consortium Ryan, W.B.F., Hsu¨, K.J., et al., 1973. Initial Report of the Deep-Sea of methanogenic archaea and bacteria. Appl. Environ. Microbiol. Drilling Project, Leg XIII. U.S. Government Printing Office, 66, 1126–1132. Washington. Pancost, R.D., Hopmans, E.C., Sinninghe Damste´, J.S., and the Salas, C., Woodside, J., 2002. Lucinoma kazani n.sp. (Mollusca: MEDINAUT Shipboard Scientific Party, 2001. Archaeal lip- bivalvia): evidence of a living benthic community associated ids in Mediterranean cold seeps: molecular proxies for an- with a cold seep in the Eastern Mediterranean Sea. Deep-Sea aerobic methane oxidation. Geochim. Cosmochim. Acta 65, Res., Part 1, Oceanogr. Res. Pap. 49, 991–1005. 1611–1627. Sassen, R., Roberts, H.H., Aharon, A., Larkin, J., Chinn, E.W., Pancost, R.D., Bouloubassi, I., Aloisi, G., Sinninghe Damste´, J.S., Carney, R., 1993. Chemosynthetic bacterial mats at cold hydro- and the MEDINAUT Shipboard Scientific Party, 2001. Three carbon seeps, Gulf of Mexico continental slope. Org. Geochem. series of non-isoprenoidal dialkyl glycerol diethers in cold-seep 20, 77–89. carbonate crusts. Org. Geochem. 32, 695–707. Schouten, S., Hopmans, E.C., Pancost, R.D., Sinninghe Damste´, Papazachos, B.C., Comninakis P.E., 1978. Deep structures and J.S., 2000. Widespread occurrence of structurally diverse tet- tectonics of the Eastern Mediterranean. Tectonophysics 45, raether membrane lipids: evidence for the ubiquitous presence 285–296. of low-temperature relatives of hyperthermophiles. Proc. Natl. Paull, C.K., Lorenson, T.D., Dickens, G., Borowski, W.S., Ussler Acad. Sci. 97, 14421–14426. III, W., Kvenvolden, K.A., 2000. Comparison of in situ and Schouten, S., Wakeham, S.G., Sinninghe Damste´, J.S., 2001. An- core gas measurements in ODP Leg 164 bore holes. In: Hol- aerobic methane oxidation by archaea in the euxinic water col- der, G.D., Bishnoi, P.R. (Eds.), Gas Hydrates: Challenges for umn of the Black Sea. Org. Geochem. 32, 1277–1281. the Future. Annals of the New York Academy of Sciences, Sibuet, M., Olu, K., 1998. Biogeography, biodiversity and fluid vol. 912, pp. 23–31. dependence of deeps cold seep communities at active and pas- Pierre, C., Rouchy, J.M., Gaudichet, A., 2000. Diagenesis in the gas sive margins. Deep-Sea Res., Part 2, Top. Stud. Oceanogr. 45, hydrate sediments of the Blake Ridge: mineralogy and stable 517–567. isotope compositions of the carbonate and sulphide minerals. In: Sørensen, K.B., Finster, K., Ramsing, N.B., 2001. Thermodynamic Paull, C.K., Matsumoto, R., Wallace, P.J. (Eds.), Proc. Ocean and kinetic requirements in anaerobic methane oxidizing con- Drill. Program Sci. Results, vol. 164, pp. 139–146. sortia exclude hydrogen, acetate, and methanol as possible elec- Raiswell, R., 1988. Chemical model for the origin of minor lime- tron shuttles. Microb. Ecol. 42, 1–10. stone-shale cycles by anaerobic methane oxidation. Geology 16, Spakman, W., Wortel, M.J.R., Vlaar, N.J., 1988. The Hellenic sub- 41–644. duction zone: a tomographic image and its geodynamic impli- Reeburgh, W.S., 1976. Methane consumption in Cariaco Trench cations. Geophys. Res. Lett. 15, 60–63. waters and sediments. Earth Planet. Sci. Lett. 28, 337–344. Stakes, D.S., Orange, D., Paduan, J.B., Salamy, K.A., Maher, N., 390 J.P. Werne et al. / Chemical Geology 205 (2004) 367–390

1999. Cold-seeps and authigenic carbonate formation in Mon- Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of terey Bay, California. Mar. Geol. 159, 93–109. bacterial formation and oxidation of methane. Chem. Geol. 161, ten Veen, J.H., Woodside, J.M., Zitter, T.A.C., Dumont, J.F., 291–314. Mascle, J., Volkonskaya, A., in press, Neotectonic evolution Werne, J.P., Baas, M., Sinninghe Damste´, J.S., 2002. Molecular of the Anaximander Mountains at the junction of the Hellenic isotopic tracing of carbon flow and trophic relationships in a and Cyprus arcs. Tectonophysics. methane-supported benthic microbial community. Limnol. Oce- Thiel, V., Peckmann, J., Seifert, R., Wehrung, P., Reitner, J., anogr. 47, 1694–1701. Michaelis, W., 1999. Highly isotopically depleted isoprenoids: Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of molecular markers for ancient methane venting. Geochim. Cos- bacterial formation and oxidation of methane. Chem. Geol. 161, mochim. Acta 63, 3959–3966. 291–314. Thiel,V.,Peckmann,J.,Richnow,H.H.,Luth,U.,Rietner,J., Woodside, J.M., 1977. Tectonic element and crust of the eastern Michaelis, W., 2001. Molecular signals for anaerobic methane Mediterranean. Mar. Geophys. Res. 3, 317–354. oxidation in Black Sea seep carbonates and a microbial mat. Woodside, J.M., Ivanov, M.K., Limonov, A.F., 1997. Neotectonics Mar. Chem. 73, 97–112. and fluid flow through seafloor sediments in the Eastern Med- Valentine, D.L., Reeburgh, W.S., 2000. New perspectives on anaer- iterranean and Black Seas: Part I. Eastern Mediterranean obic methane oxidation. Environ. Microbiol. 2, 477–484. SeaIOC Technical Series (Intergovernmental Oceanographic Volgin, L., Woodside, J., 1996. Sidescan sonar images of mud Commission, Unesco) UNESCO 1997, Paris, International, volcanoes from the Mediterranean Ridge: possible causes of vol. 48. 128 pp. variations in backscatter intensity. Mar. Geol. 132, 39–53. Woodside, J.M., Ivanov, M.K., Limonov, A.F.Anaxiprobe Ship- von Cosel, R., Salas, C., 2001. Vesicomyidae (Mollusca: Bivalvia) board Scientifis Parties, 1998. Shallow gas and gas hydrates of the genera Vesicomya, Waisiuconcha, Isorrpodon and Callo- in the Anaximander Mountains region, eastern Mediterranean gonia in the eastern Atlantic and the Mediterranean. Sarsia 86, Sea. In: Henriet, J.-P., Mienert, J. (Eds.), Gas Hydrates: Rele- 333–366. vance to World Margin Stability and Climate ChangeGeological Wallmann, K., Lionke, P., Suess, E., Bohrmann, G., Sahling, H., Society Special, vol. 137, pp. 177–193. Schlu¨ter, M., Da¨hlmann, A., Lammers, S., Greinert, J., von Woodside, J., Mascle, J., Huguen, C., Volkonskaia, A., 2000. The Mirbach, N., 1997. Quantifying fluid flow, solute mixing, Rhodes Basin, a post-Miocene tectonic trough. Mar. Geol. 165, and biogeochemical turnover at cold vents of the eastern 1–12. Aleutian subduction zone. Geochim. Cosmochim. Acta 61, Woodside, J.M., Mascle, J., Zitter, T.A.C., Limonov, A.F., Ergun, 5209–5219. M., Volkonskaia, A., 2002. The Florence Rise, the Western Bend Werne, J.P., Baas, M., Sinninghe Damste´, J.S., 2002. Molecular of the Cyprus Arc. Mar. Geol. 185, 177–194. isotopic tracing of carbon flow and trophic relationships in a Zatsepina, O.Y., Buffett, B.A., 1997. Phase equilibrium of gas hy- methane-supported benthic microbial community. Limnol. Oce- drate: implications for the formation of hydrate in the deep sea anogr. 47, 1694–1701. floor. Geophys. Res. Lett. 24, 1567–1570. Whiticar, M.J., Faber, E., 1986. Methane oxidation in sediment and Zitter, T.A.C., Woodside, J.M., Mascle, J., 2003. The Anaximander water column environments—Isotope evidence. Org. Geochem. Mountains: a clue to the tectonics of Southwest Anatolia. Geol. 10, 759–768. J. 38, 375–394.