Methylotrophic methanogenesis and potential methylated substrates in marine sediment

A Doctoral Dissertation

Guangchao Zhuang

Marum (Zentrum für Marine Umweltwissenschaften) Universität Bremen

Methylotrophic methanogenesis and potential methylated substrates in marine sediment

Dissertation

zur Erlangung des Doktorgrades

der Naturwissenschaften

– Dr. rer. nat. –

Am Fachbereich Geowissenschaften

Der Universität Bremen

vorgelegt von

Guangchao Zhuang

Bremen, October 2014

The PhD thesis was prepared between September 2010 and October 2014 within the Organic Geochemistry Group of the MARUM – Center for Marine Environmental Sciences and Department of Geosciences, University of Bremen, Leobener Str., D-28359 Bremen, Germany.

1st Reviewer: Prof. Dr. Kai-Uwe Hinrichs

2nd Reviewer: Prof. Dr. Andreas P. Teske

Additional examiners:

Prof. Dr. Wolfgang Bach

PD. Dr. Matthias Zabel

Dr. Marshall Bowles

Susanne Alfken

Date of colloquium: 10 November, 2014

Contents

Thesis Abstract ...... I

Zusammenfassung ...... III

Chapter 1 Introduction: Biogeochemistry of and low-molecular-weight

substrates in marine sediments ...... 1

Chapter 2 Gas chromatographic analysis of and ethanol in marine sediment pore waters: Validation and implementation of three

pretreatment techniques ...... 33

Chapter 3 Distribution and isotopic composition of trimethylamine and

dimethylsulfide in anoxic estuarine sediment (Aarhus Bay, Denmark) .. 59

Chapter 4 Multiple evidences for methylotrophic methanogenesis as the dominant

methanogenic pathway in deep-sea hypersaline sediments ...... 89

Chapter 5 Methanogenic substrates, activity and diversity in northwestern Mediterranean Sea: Quantitative assessment of methanogenic pathways

in marine sediments ...... 131

Chapter 6 Conclusions and perspectives ...... 169

Acknowledgements ...... 175

Abstract

Abstract

Methane is the simplest and a potent that plays important roles in atmospheric chemistry, the global , and the formation of gas hydrates in marine sediment. Microbial production of methane is the terminal step during the degradation of organic matter. It is generally thought that methane is predominantly produced from hydrogenotrophic and acetoclastic methanogenesis, while methylotrophic methanogenesis and its relative importance for methane production in marine sediments remain largely unconstrained. The main objective of this study is to constrain potential methylated substrates and methylotrophic methanogenic activities, and further evaluate the importance of methylotrophic methanogenesis in marine sediment. As the lack of knowledge on in situ concentrations of methylated compounds impedes our understanding on their quantitative contribution to methane production, the first step was to determine the concentrations and carbon isotopic composition of methylated compounds using newly-developed methods. Quantitative or isotopic analysis of methanol, trimethylamine (TMA) and dimethylsulfide (DMS) in marine sediment and pore waters were achieved using gas chromatographic approaches in combination with a range of pretreatment techniques. Using these protocols, the concentrations and distributions of methylated compounds were determined in a variety of marine sediments from Aarhus Bay in Denmark, Orca Basin in the Gulf of Mexico and Gulf of Lions in the northwestern Mediterranean Sea. To further constrain the importance of methylotrophic methanogenesis, two case studies combining the newly-developed methods as well as various biogeochemical analyses were performed in hypersaline sediment and estuarine sediment. In hypersaline sediment of Orca Basin, multiple lines of evidences from abundances of methanogenic substrates, carbon isotope systematics between methane and substrates, thermodynamic calculations, stable isotope tracer and radiotracer experiments as well

I Abstract

as gene and lipid biomarkers collectively confirmed that methylotrophic methanogenesis was the dominant methanogenic pathway in Orca Basin sediments. Furthermore, the distribution of methanogenic substrates, activity and diversity were characterized to quantitatively estimate the relative importance of different methanogenic pathways in estuarine sediment of the northwestern Mediterranean Sea. The results showed that both methylotrophic and hydrogenotrophic methanogenesis contributed to the formation of methane in the reduction zone, with methylotrophic methanogenesis accounting for 13%-74% of the total methane production. In the sulfate-depleted sediments, hydrogenotrophic methanogenesis dominated methanogenic pathway (67%-97%), whereas acetoclastic methanogenesis contributed up to 31% of methane production in organic-rich sediment. In contrast, the contribution of methylotrophic methanogenesis to the total methanogenic activity was negligible in the methanogenic zone (< 1%). Collectively, new constraints from methylated compounds and the metabolic activities improve our quantitative understanding on methylotrophic methanogenesis in different marine sediment settings. The findings in this thesis provide more comprehensive insights into the relative importance of methylotrophic methanogenesis in marine sediment.

II Zusammenfassung

Zusammenfassung

Methan ist der einfachste Kohlenwasserstoff und ein effizientes Treibhausgas, das eine wichtige Rolle sowohl in der Chemie der Atmosphäre, dem globalen Kohlenstoffkreislauf als auch bei der Bildung von Gashydraten in marinen Sedimenten spielt. Mikrobielle Produktion von Methan ist der finale Schritt im Abbau organischen Materials. Es wird allgemein angenommen, dass Methan vorwiegend hydrogenotroph oder acetoklastisch gebildet wird. Das Auftreten und die Bedeutung methylotropher Methanogenese in marinen Sedimenten sind jedoch weitgehend unbekannt. Ziel der vorliegenden Arbeit ist es, potentielle methylierte Substrate zu identifizieren und methylotrophe methanogene Aktivität zu quantifizieren um die Bedeutung methylotropher Methanogenese in marinen Sedimenten zu evaluieren. Da die Konzentrationen methylierter Verbindungen im Sediment und deren Beitrag zur Methanproduktion weitestgehend unbekannt sind, bestand der erste Schritt darin, Konzentrationen und isotopische Zusammensetzung methylierter Verbindungen mit Hilfe neu entwickelter Methoden zu messen. Die quantitative und isotopische Analyse von Methanol, Trimethylamin (TMA) und Dimethylsulfid (DMS) in marinem Sediment und Porenwasser wurde durch die Kombination von Gas-Chromatographie und einer Reihe von Probenvorbereitungsschritten erreicht. Mit Hilfe dieser Techniken wurden die Konzentrationen und Verteilungen methylierter Verbindungen in verschiedenen Sedimenten, aus der Aarhusbucht (Dänemark), dem Orca-Becken (Golf von Mexiko) und dem nordwestlichen Mittelmeer, bestimmt. Um die Bedeutung methylotropher Methanogenese zu bestimmen, wurden die neu entwickelten Methoden mit weiteren biogeochemischen Analysen kombiniert und in zwei Fallstudien an hypersalines und ästuarines Sediment angewandt. Im hypersalinen Sediment des Orca-Beckens zeigte die Kombination aus den Konzentrationen methanogener Substrate, Kohlenstoffisotopien von Methan und Substraten, Inkubationsexperimenten mit stabilen und radioaktiven Kohlenstoffisotopen sowie thermodynamischer Berechnungen und Lipid- und

III Zusammenfassung

Genbiomarkern, dass methylotrophe Methanogenese der dominierende methanogene Reaktionsweg im Orca-Becken ist. Weiterhin wurden die Verteilung methanogener Substrate und methanogene Aktivität in ästuarinen Sedimenten aus dem nordwestlichem Mittelmeer charakterisiert um die relative Bedeutung der verschiedenen methanogenen Reaktionswege zu quantifizieren. Die Ergebnisse zeigen, dass sowohl methylotrophe als auch hydrogenotrophe Methanogenese zur Bildung von Methan in der Sulfatreduktionszone beitragen. Methylotrophe Methanogenese trug hierbei 13 bis 74% zur gesamten Methanproduktion bei. In Sulfat-freien Sedimenten war hydrogenotrophe Methanogenese der dominierende Reaktionsweg (67-97%) während acetoklastische Methanogenese bis zu 31% zur Methanproduktion in Organik-reichen Sedimenten beitrug. Im Vergleich dazu war der Anteil methylotropher Methanogenese an der gesamten methanogenen Aktivität innerhalb der methanogenen Zone vernachlässigbar (< 1%). Zusammenfassend verbessern die neuen Erkenntnisse zur Verteilung methylierter Substrate und methylotropher methanogener Aktivität das quantitative Verständnis methylotropher Methanogenese in verschiedenen marinen Sedimenten. Die Ergebnisse dieser Arbeit ermöglichen somit ein tieferes Verständnis der relativen Bedeutung methylotropher Methanogenese in marinen Sedimenten.

IV Chapter 1

Chapter 1

Introduction

Biogeochemistry of methane and low-molecular-weight substrates in

marine sediments

1 Chapter 1

1.1 Methane production during organic matter degradation in marine sediment

Methane is the smallest but the most abundant hydrocarbon on Earth. As one of the most potent greenhouse gases, methane plays important roles in Earth’s greenhouse effect, atmospheric chemistry and the global carbon cycle. The major sources of include wetlands, enteric in animals (e.g., termites, ruminants), rice cultivation, burning and fossil fuel (Table 1.1). Though the contribution of the ocean to the global methane budget is limited, marine sediment is the largest reservoir of methane (500-10,000 gigatons) on Earth (Kvenvolden, 2002; Milkov, 2004).

Table 1.1 Global methane gross production and net emission. Data from Reeburgh, 2007.

Source Gross production Net emission (Tg yr-1) (Tg yr-1) Nature sources Wetlands 142 115 Termites 44 20 Oceans and freshwaters 85.3 10 Hydrates 10 5 Anthropogenic sources Rice cultivation 577 100 Ruminants 80 80 62 40 Biomass burning 55 55 Coal mining 35 35 Gas production 58 40 Others 54 25 Total 1202.3 525

In marine sediment, methane is primarily formed from the breakdown of organic matter, which can occur microbially or through thermo-chemical processes. The organic matter, derived from the euphotic zone of the ocean by photosynthesis, is largely mineralized in the oxic water column during sedimentation, and only a small fraction is transported to the sediment surface (Ducklow et al., 2001). With the involvement of various , the degradation of this small part of organic

2 Chapter 1

matter would be the driving force for early diagenesis (Rullkötter, 2006). As macromolecular compounds such as structural carbohydrates, proteins, nucleic acids and lipid complexes (Fig. 1.1), can not be directly metabolized by prokaryotic (Jørgensen, 2000), these polymers are first hydrolyzed to oligomers or monomers, e.g., amino acids, simple sugars, and long-chain fatty acids, with extracellular enzymes produced by the (Arnosti, 2011). Subsequently, fermenting bacteria degrade the monomeric compounds into a limited number of fermentation products including volatile fatty acids and alcohols (e.g., formate,

acetate, propionate, ethanol), as well as H2 and CO2. Through a further fermentation step, the products may be concentrated to the key metabolites such as acetate, H2 and

CO2.

Fig. 1.1 The stepwise degradation of organic matter in marine sediments (Modified from Konhauser, 2009).

3 Chapter 1

During early diagenesis, the degradation of organic matter could be coupled to various terminal electron-accepting processes. Depending on the energy yield, these processes follow the sequence of aerobic respiration, nitrate reduction, metal reduction (e.g., manganese and ), sulfate reduction and methane production (Froelich et al., 1979). Ideally, this sequence could be reflected from the vertical profile with biogeochemical zonations in marine sediments (Fig. 1.2). With increasing depth into the sediment, the decreasing energy yield from methane production restricts methanogenic microorganisms to utilizing a limited number of substrates, such as H2/CO2, acetate and methylated compounds (Fig. 1.1). The formation of methane is the terminal step of organic carbon degradation, which could account for 5-10% of organic matter mineralization (Canfield et al., 2005; Hinrichs and Boetius, 2003). The residual refractory organic matter that escapes mineralization will be deposited to the deeper sediment, and constitute one of the major carbon reservoirs on Earth (Hedges and Keil, 1995). Increasingly heated with depth during burial, the recalcitrant organic matter could be further thermally altered to methane and higher .

Fig. 1.2 Pathways of organic carbon degradation and related biogeochemical zonations in marine

sediments (modified from Konhauser, 2009 and White, 2013). [CH2O] symbolizes organic matter of unspecified composition and standard free energy yields (ΔG0) refer to per mol of organic carbon.

4 Chapter 1

1.2 Methane source in marine environment

In general, methane could be produced biogenically by microbial or thermogenic degradation of organic matter, and abiogenically by chemical reactions independent of organic matter.

1.2.1 Abiogenic methane

At minor amounts on a global scale, abiogenic methane could be formed by either high-temperature magmatic processes in volcanic and geothermal areas, or via low-temperature (<100 °C) gas-water-rock reactions in continental settings (Etiope and Sherwood Lollar, 2013). The most widely invoked mechanism for the generation of abiotic methane could be the catalytic hydrogenation of carbon oxides (CO and

CO2) by Fischer-Tropsch-type reactions (Eqs. 1.1 and 1.2), especially in serpentinized ultramafic rocks (Charlou et al., 2002; Horita and Berndt, 1999). During diagenesis of oceanic rocks, could be released from a variety of processes such as serpentinization and radiolysis (Smith et al., 2005). With the aid of hydrogen, the transformation of the carbon gas molecules could occur on the surface of metal catalysts (e.g., nickel, iron, chromium). This synthetic pathway has been experimentally investigated under hydrothermal conditions, at temperatures above

200 °C and high pressures (McCollom and Seewald, 2006), and the synthesis is also theoretically possible at low temperature.

nCO + 2nH2 = —(CH2)n— + nH2O (1.1)

CO2 + 4H2 = CH4 + 2H2O (1.2) Instead of the involvement of a heterogeneous catalyst or gas phase, abiotic

production of methane may occur in aqueous solution in presence of CO2, CO and H2

at relatively high temperatures (> 150 °C) (Seewald et al., 2006). Aqueous CO2 is reduced to methane through a series of redox reactions with , formaldehyde or methanol as intermediates (Eqs. 1.3-1.6). Such abiotic reactions involving aqueous carbon compounds in hydrothermal conditions may influence metabolic pathways utilized by organisms that inhabit submarine vent environments.

5 Chapter 1

CO2 + H2 → HCOOH (1.3)

HCOOH + H2 → CH2O + H2O (1.4)

CH2O + H2 → CH3OH (1.5)

CH3OH + H2 → CH4 + H2O (1.6)

1.2.2 Biogenic methane 1.2.2.1 Microbial production of methane

Basically, the overwhelming majority of methane is produced by methanogenic microorganisms during the degradation of organic matter. With respect to the utilization of low-molecular-weight substrates, microbial methanogenic pathways could be classified into three types: hydrogenotrophic methanogenesis from carbonate reduction with hydrogen, acetoclastic methanogenesis from acetate disproportionation, and methylotrophic methanogenesis from methylated compounds including methanol, and methylsulfides (Fig. 1.3; Table 1.2). Furthermore, as sulfate-reducing bacteria could thermodynamically outcompete with for

H2/CO2 and acetate, these two substrates are acknowledged as competitive substrates. The methylated compounds, which are generally not used by sulfate reducers, are termed as non-competitive substrates for methanogens (Oremland and Polcin, 1982).

Table 1.2 Methanogenic pathways and standard Gibbs free energies for methanogenic reactions.

Reaction ΔGo (kJ mol-1 reaction)a I Hydrogenotrophic methanogenesis - + 4H2, aq + HCO3 + H → CH4, aq + 3H2O -229 II Acetoclastic methanogenesis - - CH3COO + H2O → CH4, aq + HCO3 -15 III Methylotrophic methanogenesis - + 4CH3OH → 3CH4, aq + HCO3 + H + H2O -223 + - + + 4CH3NH3 + 3H2O → 3CH4, aq + HCO3 + 4NH4 + H -136 + - + + 4(CH3)2NH2 + 6H2O → 6CH4, aq + 2HCO3 + 4NH4 + 2H -261 + - + + 4(CH3)3NH + 9H2O → 9CH4, aq + 3HCO3 + 4NH4 + 3H -402 - + 2(CH3)2S + 3H2O → 3CH4, aq + HCO3 + 2H2S + H -46 - + 4CH3SH + 3H2O → 3CH4, aq + HCO3 + 4H2S + H -59 aThe standard free energies were calculated from the free energy of formation of the most abundant ionic at neutral pH (Rossini et al., 1952).

6 Chapter 1

In addition to these key substrates, quaternary amines such as choline and glycine betaine as direct methanogenic substrates have been recently demonstrated in pure cultures (L'Haridon et al., 2014; Watkins et al., 2014; Watkins et al., 2012). These compounds extend the substrates range for methanogenesis, while their role as methanogenic substrates in natural environments requires further validation.

Fig. 1.3 The generation of low-molecular-weight substrates and methanogenic pathways in marine sediments (Modified from Konhauser, 2009).

Despite that methane is the main metabolic product of methanogens, methane could also be produced anaerobically or aerobically from other microorganisms at minor quantities as a byproduct of (Whitman et al., 2006). For example, trace amount methane formation has been observed in proteolytic clostridia cultures growing on L-methionine (Rimbault et al., 1988). The archaeal species from Archaeoglobus could also emit methane in the cultures (Mori et al., 2008). More strikingly, phosphorus-starved marine microbes could decompose methylphosphonic

7 Chapter 1

acid with concomitant release of methane, and this finding provides a plausible explanation for the methane paradox in the aerobic ocean (Karl et al., 2008; Metcalf et al., 2012).

1.2.2.2 Thermogenic production of methane

Escaping early diagenetic mineralization, refractory organic matter (kerogen) would be exported to the deep subsurface due to subduction or sedimentation driven subsidence. Methane and other hydrocarbons will be generated from the thermo-altered organic matter in the increasingly heated sediments with depth (Fig. 1.4).

Fig. 1.4 General scheme of hydrocarbon generation from sedimentary rock (Tissot and Welte, 1984).

Specifically, kerogen slowly releases gaseous and liquid hydrocarbons at depths where temperatures exceed 50 °C (Schoell, 1988; Tissot and Welte, 1984). Economical oils are formed at temperatures between 60 and 120 ºC during early catagenesis. Wet gas, a condensate consisting of methane, other lighter hydrocarbons and liquid hydrocarbons, is essentially produced during the stage of catagenesis. Towards the end of catagenesis the proportion of methane rises rapidly in the gaseous products, due to the presumed thermal cracking of previously evolved hydrocarbons.

8 Chapter 1

With increasing temperature and kerogen maturity, metagenesis occurs at the last stage of thermal alteration of organic matter. During metagenesis, methane is the only hydrocarbon released, and the carbon-rich solid residues remain as inert kerogen.

1.3 Low-molecular-weight methanogenic substrates in marine sediment 1.3.1 Competitive substrates for methanogenesis

1.3.1.1 H2/CO2

Molecular hydrogen is a central metabolite in marine sediments. Hydrogen can be supplied through fermentation and thermal alteration of sedimentary organic matter (Hoehler et al., 1998; Seewald, 2003), abiotic production via serpentinization (Kelley et al., 2005), or radiolysis of water (D'Hondt et al., 2009). The metabolism of hydrogen could be coupled to various terminal processes such as sulfate reduction and

methane production (Cord-Ruwisch et al., 1988). Generally, H2 concentration is thought to be mediated by microorganisms involving both production and

consumption, and a threshold is maintained that allows H2 production and consumption thermodynamically feasible (Lin et al., 2012). As a consequence, the

threshold concentrations of H2 increase with the decreasing redox potentials of hydrogenotrophic processes, e.g., nitrate reduction < iron and manganese reduction < sulfate reduction < methanogenesis < homoacetogenesis. Thus, hydrogenotrophic methanogens are not able to compete with sulfate-reducing bacteria for hydrogen in the presence of sulfate. Once the available dissolved sulfate is exhausted, hydrogenotrophic methanogenesis become more important process, usually at greater sediment depth. Due to the accumulated sizable bicarbonate pool, in most cases, hydrogenotrophic methanogenesis is the dominant methanogenic pathway in marine sediment (Whiticar, 1999).

1.3.1.2 Acetate

9 Chapter 1

In addition to hydrogen, volatile fatty acids are another class of important intermediates during organic matter degradation. The water-soluble C2-compound

acetate could be produced from organic matter fermentation or CO2 reduction via the acetyl-CoA pathway (acetogenesis) (Drake, 1994). Likewise, acetate serves as an important substrate for a variety of redox processes (e.g., Hedderich and Whitman, 2006). Due to the rapid turnover, the concentrations of acetate are typically maintained at low levels (<10 μM) in surface sediments (Parkes et al., 2007; Wellsbury et al., 1997), while elevated abundance of 10 mM or higher can be reached in the deeply-buried sediments (Wellsbury et al., 1997). With the recently developed technique that allows online carbon isotopic determination of acetate (Heuer et al., 2006), it is possible to infer the metabolic pathways that produce or consume acetate, and further link the inferred metabolic variation to geochemical zonation (e.g., Heuer et al., 2009). Similar to hydrogen, acetate is preferentially utilized by sulfate-reducing bacteria, and acetoclastic methanogenesis only occur in low sulfate environments. For example, acetate could contribute to 70% of the methane production in freshwater sediment (Conrad, 1999; Winfrey and Zeikus, 1977). In marine sediment, acetoclastic methanogenesis occur until sulfate is depleted, and the importance might depend on the organic matter lability and substrate availability (Crill and Martens, 1986).

1.3.2 Non-competitive substrates for methanogenesis 1.3.2.1 Methanol

Volatile alcohols (e.g., methanol and ethanol) are also important metabolites in marine sediments. Under aerobic or anaerobic conditions, methanol could be produced from the decomposition of lignin or pectin (Donnelly and Dagley, 1980; Schink and Zeikus, 1980). Lignin is phenolic polymer and constitutes the natural composite material in vascular plants. The methoxylated monomers of lignin could release the methoxy group with the formation of methanol during degradation by bacteria or fungi. Pectin is a common constituent of plant and algal cells and is a polymer of α-(l,4)-galacturonic acid that is partially methoxylated at the carboxy

10 Chapter 1

groups. Likewise, the methoxy group is released as methanol during anaerobic decomposition of pectin. Diverse pectinolytic strains of Clostridium, Erwinia, and Pseudomonas species produce methanol as a major end product during growth on pectin (Schink and Zeikus, 1982).

1.3.2.2 Methylated amines

Methylated amines, e.g., monomethylamine, dimethylamine and trimethylamine, are small organic nitrogen compounds and ubiquitous in the marine environment. The major source for the methylated amines, in particular trimethylamine (TMA), is from the degradation of a veriety of precursors including glycine betaine (N,N,N-trimethylglycine), choline (N,N,N-trimethylethanolamine) and trimethylamine N-oxide (TMAO). In marine organisms, quaternary amines such as glycine betaine and TMAO are accumulated by cells in response to salinity or water stresses. These compatible solutes maintain a higher intracellular osmotic potential than that of the extracellular environment thereby keeping the cellular turgor pressure constant by preventing water loss through osmosis under marine conditions (Yancey et al., 1982). Anaerobically, TMA could be produced from betaine with the involvement of fermentative such as Sporomusa, or sulfate reducer Desulfuromonas acetoxidans (Heijthuijsen and Hansen, 1989; Möller et al., 1984). In parallel, choline could be degraded to TMA by sulfate-reducing bacteria (Hayward and Stadtman, 1959). In a coculture of Desulfovibrio strain G1 and barkeri strain Fusaro (DSM 804), it was found that choline was degraded to TMA by sulfate reducer, and the produced TMA was subsequently used by methanogen to produce methane (Fiebig and Gottschalk, 1983). Recent studies have shown that betaine and choline could be used as direct methanogenic substrates in pure cultures of strains, allowing methanogens to gain energy from these substrates without the need for syntrophic partners (Watkins et al., 2014; Watkins et al., 2012). Furthermore,

11 Chapter 1

TMAO could also serve as the terminal electron acceptor in of bacteria with the reduction to TMA (Strøm et al., 1979).

1.3.2.3 Methylated sulfides

Dimethylsulfide (DMS) is the dominant volatile compound produced biogenically in the ocean (e.g., Charlson et al., 1987). In the seawater, DMS is mainly derived from dimethylsulfoniopropionate (DMSP), a tertiary sulfonium compound that acts as an osmolyte in a wide variety of marine phytoplankton (e.g., Keller et al., 1989). The conversion of DMSP to DMS involves an elimination reaction which could be catalyzed either by OH- or through enzymatic cleavage (Eq. 1.7). Under natural conditions, DMS is dominantly produced through enzymatic cleavage of DMSP, and the alkaline catalysis reaction is very slow at the seawater pH of ~8. In the presence of strong base, DMSP could be decomposed to DMS following a 1:1 ratio, and the concentration of DMSP could be determined from the quantitative production of DMS after the treatment with alkali (Dacey and Blough, 1987).

+ - - + (CH3)2S CH2CH2COO → (CH3)2S + CH2=CHCOO + H (1.7) DMSP could occur in the sediments because of the settling of algal detritus, production by benthic algae, and release by rooted plants. The degradation of DMSP in marine sediment involves a cleavage to acrylate and DMS, successive demethylations to 3-mercaptopropionate or demethiolation to (Kiene and Taylor, 1988; van der Maarel and Hansen, 1997) (Fig. 1.5). Furthermore, the degradation of sulfur-containing amino acids, the methylation of methanethiol (Kiene and Capone, 1988; Kiene and Taylor, 1988), and the reduction of dimethylsulfoxide by sulfate-reducing bacteria (Jonkers et al., 1996) could contribute to the production of DMS in marine sediment (Fig. 1.5). In addition to transformation from organic sulfur compounds, methylated sulfides could be microbially derived through linking organic or inorganic carbon to hydrogen sulfide. For example, O-methyl groups (methyl groups bonding to an atom, e.g., methanol) can be transferred microbially to hydrogen sulfide to

12 Chapter 1

form methanethiol, with additional methylation under certain circumstances yielding DMS (Finster et al., 1990; Lomans et al., 2002). Microbial conversion of inorganic carbon to DMS has also been observed in anoxic lake sediment (Lin et al., 2010). In a methanogen pure culture, DMS and methanethiol could be produced by Methanosarcina acetivorans when carbon monoxide served as the only electron donor (Moran et al., 2008).

Fig.1.5 Microbial transformation of methylated sulfur compounds in marine sediments.

In contrast to hydrogen and acetate, quite limited information is available on those methylated methanogenic substrates in marine sediment. A main reason for this gap is the lack of suitable analytical protocols to quantify these labile and relatively volatile trace compounds. The conventional procedures for volatile compounds, such as headspace analysis for hydrocarbon gases, could not suffice for successful determination of these methylated compounds. For example, the expected low concentrations, and complete water miscibility of methanol necessitates an efficient preconcentration step for detection with small volume pore waters. For methylamines, the basic nature and high polarity pose further difficulty for trace analysis due to strong adsorption on surfaces of common labware materials or column. Because of these analytical challenges, the concentrations of methanol, methylated amines or sulfides are barely reported and their biogeochemistry are largely unconstrained in marine sediment (Fitzsimons et al., 1997; Lee and Olson, 1984; Wang and Lee, 1990).

13 Chapter 1

Despite the lack of knowledge on in situ concentrations of methylated substrates, there is a persistent interest in methylotrophic methanogenesis in marine sediment. Oremland et al. (1982b) firstly reported that methanol and TMA were important substrates for methane production in anoxic salt marsh sediments, and the production rate and the produced amount of methane were not affeced by the presence of sulfate (e.g., the addition of 20 mM sulfate). The hypothesis that methanogenesis and sulfate reduction could operate concurrently was further tesified in the estuarine sediment (Oremland and Polcin, 1982). In the presence of sulfate, methanogenesis was not inhibited in sediments amended with methanol, TMA, or methionine, but was retarded with hydrogen and acetate as substrates, and sulfate reduction was stimulated by acetate and hydrogen but not methanol and TMA. In the inertidal sediment of Lowes Cove, Maine, TMA could account for 35.1 to 61.1% of total methane production (King et al., 1983), although a large fraction of trimethylamine and methanol could be also catabolized via non-methanogenic process. At high concentrations (mM range), DMS and methanethiol are mainly converted to methane, and a significant fraction might be consumed by sulfate-reducing bacteria at low concentrations (μM range) (Kiene et al., 1986; Kiene and Visscher, 1987). In the deep carobonate sediment of the southern Australian continental margin, high concentrations of methane were observed in the sulfate redution zone, and the authors assumed that methanogenesis from the non-competitive substrates contributed to the coexistence of methane and sulfate (Mitterer, 2010; Mitterer et al., 2001). More intriguingly, methylotrophic methanogenesis has been proposed as the main methanogenic pathway in a variety of hypersaline sediments, e.g., Big , Nevada (Oremland et al., 1982a), the Napoli mud volcano, Eastern Mediterranean Sea (Lazar et al., 2011), hypersaline ponds in Baja California Sur, Mexico, and northern California, USA (Kelley et al., 2012). Accordingly, methylotrophic methanogens also have been detected or isolated with culture independent or dependent methods from various sedimentary environments (e.g., Lazar et al., 2012; Lazar et al., 2011; Oremland et al., 1989; Singh et al., 2005; Sowers and Ferry, 1983).

14 Chapter 1

Though multiple lines of indications or evidences suggest the potential methane production from methylated compounds, the significance of methylotrophic methanogenesis is not fully understood in marine sediment. To better understand this poorly constrained methanogenic pathway, the main focus of this PhD thesis will be methylated compounds and their role as methanogenic substrates in marine sediment.

1.4 Stable carbon isotope analysis of methane and precursors

Most elements including carbon are mixtures of atoms with different masses as a result of a variable number of neutrons. The two stable isotopes of carbon, 12C and 13C, occur naturally at a proportion of approximately 99:l. The stable isotopic composition of carbon is expressed as δ-notation [‰], defined with the following equation (1.8):

13 δ C = (Rsample/Rstandard – 1) × 1000 (‰) (1.8) where R is the ratio of 13C and 12C, and the sample is measured relative to reference standard (Vienna PeeDee Belemnite, VPDB). The information encoded in the stable carbon isotopic composition has enormously advanced our understanding of related biogeochemical processes in the past several decades (e.g., Hayes et al., 1990; Heuer et al., 2009). In particular for methane, the isotopic signatures have been used extensively to identify and quantify the methanogenic pathways (e.g., Conrad, 2005; Whiticar, 1999). Abiotic methane has a wide range of δ13C, either 13C enriched (e.g., > -20‰) or depleted (e.g., between -30‰ and -47‰) (Etiope and Sherwood Lollar, 2013). Thermogenic methane, roughly but not exclusively, ranges from -50‰ to -20‰. In contrast, the microbial formation of methane results in more negative δ13C values from -110‰ to -50‰, which is clearly distinguished from those with abiotic or thermogenic source. The of methane is controlled by the isotopes of precursors and the kinetic isotope effects during the metabolism of different compounds by methanogens. In principle, the substrate molecules with lighter isotopes (i.e., 12C), would diffuse and react more rapidly, thus these molecules are utilized more

15 Chapter 1

frequently by microbes. This discrimination could lead to the strong depletion in 13C of microbial CH4 relative to the precursor substrates. As the fractionation varies with respect to substrates, the produced methane with different extents of depletion could be used as a proxy to distinguish the biological methanogenic pathways or estimate their relative contributions to total methane production (e.g., Kotsyurbenko et al., 2004; Krüger et al., 2002). For example, methane derived from acetate might have δ13C values of -70‰ to -50‰, while hydrogenotrophic methanogenesis produces methane with δ13C values between -110‰ to -60‰ due to the larger fractionation factor (-49‰ to -95‰) (Whiticar et al., 1986). In contrast, due to the lack of information on the isotopes of methylated compounds, little is known about the magnitude of isotope fractionation and the stable isotopic composition of methane produced from methylotrophic methanogenesis. Few studies investigated the isotopic discrimination between methylated substrates and methane with methylotrophic methanogen enrichment or pure cultures (Table 1.3). From these studies, it could be inferred that methylotrophic methanogenesis was associated with significant isotopic fractionation, and the largest fractionation was from methanol induced methanogenesis (63‰ to 83‰), followed by TMA (37‰ to 71‰) and DMS (44‰ to 54‰). Thus it could be expected that methane derived from methylated substrates should be relatively depleted in 13C. It should be noted that isotopic proxies could be obscured and influenced by other factors, such as methane oxidation, substrate limitation, or multiple sources of substrates. For example, Holler et al. (2009) demonstrated that anaerobic oxidation of methane could drive an enrichment of 12‰-39‰ for residual methane. With limited substrates, the kinetic isotope effect will significantly decrease, resulting in greater 13C-enrichment of the metabolic product (e.g., methane) (Kelley et al., 2012; Yoshinaga et al., 2014). Furthermore, it is found that the stable isotope proxy model is not directly applicable for interpreting methanogenic pathways in a dynamic and complex environment with multiple methanogenic substrates, including both competitive and non-competitive substrates (Parkes et al., 2012).

16 Chapter 1

Table 1.3 Fractionation factor associated with methylotrophic methanogenesis. Substrate Fractionation factor (‰) Culture Referencesa 13 13 (δ Csubstrate-δ Cmethane) Methanol 72-77 Methanol-grown 1 methanogen enrichment 73-75 Methanosarcina barkeri 2 63 Methylotrophic coccoid 3 methanogen (strain GS-16) 83 Methanosarcina barkeri 4 68-77 -- 5 TMA 37 Methylotrophic coccoid 3 methanogen (strain GS-16) 50.2 Methanosarcina barkeri 6 71 Methanococcoides burtonii 6 39 -- 5 67 Methanosarcina barkeri 4 DMS 44 Methylotrophic coccoid 3 methanogen (strain GS-16) 44-54 -- 5 aReferences: 1. Oremland et al., 1982a; 2. Krzycki et al., 1987; 3. Oremland et al., 1989; 4. Londry et al., 2008; 5. Whiticar, 1999; 6. Summons et al., 1998.

1.5 Methane oxidation in marine environment

In the atmosphere, methane could be naturally oxidized through photochemical radical reactions. The hydroxyl radicals, derived from the UV degradation of ozone, could attack methane to form methyl radicals (Levy, 1971), and the methyl radicals would be ultimately oxidized to formaldehyde (Grosjean, 1995). Before escaping to the atmosphere, most of the biogenic methane are consumed aerobically or anaerobically. For example, aerobic methanotrophy (Eq. 1.9) is the major sink of methane in terrestrial habitats (Reeburgh, 2007).

CH4 + 2O2 → CO2 + 2H2O (1.9) In contrast, anaerobic oxidation of methane (AOM) could account for 80-85% of marine methane oxidation (Hinrichs and Boetius, 2003). In marine sediment, the process of AOM is usually coupled with sulfate reduction (Eq. 1.10), and mediated by a syntrophic consortium of methanogen and sulfate-reducing bacteria.

17 Chapter 1

2- - - CH4 + SO4 → HCO3 + HS + H2O (1.10) Thus, AOM largely occurs where methane and sulfate overlap, a typical zonation defined as sulfate methane transition zone (SMTZ). At SMTZ, methane is completed consumed and the upwards methane flux are significantly regulated. Depending on the burial rate of reactive organic matter, the depth of the methanogenic zone, and the transport velocity of methane and sulfate and their consumption rates, SMTZs are found at decimeters to tens of meters below the seafloor (Knittel and Boetius, 2009). The highest AOM rates are generally observed in the SMTZ, which covers a wide range from a few pmol cm-3 day-1 in subsurface of deep margins, to several tens nmol cm-3 day-1 in surface coastal sediments (Knittel and Boetius, 2009).

1.6 Methanogenic and methanotrophic 1.6.1 Methanogens

Life on Earth is grouped into three Domains, Archaea, Bacteria and Eukarya (Woese et al., 1990). Methanogens are strictly anaerobic Archaea belonging to the phylum of and produce methane as the major end product of metabolism. Despite the limited substrates range, methanogens are phylogenetically diverse. Based on 16S rRNA gene analysis, methanogens are classified into seven orders: Methanopyrales, Methanococcales, Methanobacteriales, , , Methanocellales (Sakai et al., 2008), and Methanoplasmatales (Paul et al., 2012). Species from orders Methanopyrales, Methanococcales, Methanobacteriales, Methanomicrobiales, Methanocellales are generally

hydrogenotrophic methanogens that utilize H2/CO2 to produce methane, some of which could also use formate and secondary alcohols as methanogenic substrates (Garcia et al., 2000). Members of the order Methanosarcinales are the most metabolically versatile methanogens, as they can metabolize both competitive and non-competitive substrates (e.g., H2/CO2, acetate, methanol, methylamines). All the obligatory methylotrophic methanogens that utilize methylated compounds without hydrogen (e.g., Methanococcoides, ) are found in this order.

18 Chapter 1

Members of the Methanosaeta in this order are the only methanogens that exclusively utilize acetate for methanogenesis. As the recently proposed seventh order of methanogen, Methanoplasmatales are mostly retrieved from intestinal samples (e.g., termite and cockroach guts) (Paul et al., 2012), and both the pure culture Methanomassiliicoccus luminyensis and the enrichment cultures from this order produce methane through obligate hydrogen dependence of methanol reduction (Dridi et al., 2012). Through the metabolism of substrates, methanogens perform methanogenic reactions with the involvement of a complex series of enzymes. However, despite the different carbon substrates used for methanogenesis, all methanogens share the common final step in which the methyl- reductase (MCR) catalyzes the reduction of the into methane through the reaction between the methyl coenzyme M and the (Ferry, 1992). As a result, the mcrA gene encoding the MCR is specific and useful functional gene marker targeting methanogens in the environment.

1.6.2 Anaerobic methanotrophic archaea (ANME)

In marine sediment, three distinct clusters of Euryarchaeota, namely ANME-1, ANME-2, and ANME-3 (anaerobic oxidation of methane) could mediate the process of AOM (Knittel and Boetius, 2009). ANME-1 are distantly related to the orders Methanosarcinales and Methanomicrobiales, and could occur as single cells or monospecific aggregates in association with sulfate-reducing bacteria of the Desulfosarcina/Desulfococcus group form the Deltaproteobacteria (Knittel et al., 2005; Michaelis et al., 2002). It has been found that ANME-1 dominated the diffusion-driven SMTZs and microbial mats of the Black Sea (Knittel et al., 2005; Michaelis et al., 2002; Niemann et al., 2005). Both ANME-2 and ANME-3 belong to the order Methanosarcinales, while ANME-3 are more closely related to genera of Methanococcoides and Methanolobus. The consortium of ANME-2 and associated Desulfosarcina/Desulfococcus are largely detected in various cold seep ecosystems

19 Chapter 1

(Knittel et al., 2005; Orcutt et al., 2005; Orphan et al., 2002). ANME-3 are typically found in aggregates with Desulfobulbus-related bacteria as a sulfate-reducing partner (Lösekann et al., 2007). The close phylogenetic relationship between methanotrophic and methanogenic archaea and the biogeochemical link between both processes in the carbon cycle might suggest a coevolution of their biochemistry (Knittel and Boetius, 2009). In the current reverse methanogenesis hypothesis, the initial step in methane oxidation is essentially a reversal of methyl-coenzyme M reduction catalyzed by MCR in methanogenesis. First evidence for the presence of MCR in sediments from AOM zones or enrichment cultures comes from the identification of novel mcrA genes that could be assigned to ANME-1 and ANME-2 (Hallam et al., 2003). Further surveys of metagenomic libraries revealed the presence of nearly all genes typically associated with methanogenesis in ANME-1, and to a lesser extent in ANME-2 cells (Hallam et al., 2004). Consequently, the highly conserved mcrA genes could also be used as the key gene markers for ANMEs. Furthermore, recent studies also demonstrated that ANME could function as normal methanogen for net production of methane (Bertram et al., 2013; Lloyd et al., 2011).

1.6.3 Lipid biomarkers

Specific lipid biomarkers and the stable carbon isotope signatures have been widely used for the identification of methanogenic and methanotrophic communities in natural environments (Hinrichs et al., 1999; Pancost et al., 2000). Due to the close phylogenetic and biochemical relationships between the two groups, methanogen and ANMEs have similar lipids biomarkers. Specifically, these biomarkers include the isoprenoidal glycerol diethers isoprenoidal glycerol ethers (2,3-di-O-phytanyl-sn-glycerol), sn2- (2-O-(3'-hydroxy-3',7', 11',15'-tetramethyl)hexadecyl-3-O-phytanyl-sn-glycerol), the branched hydrocarbon 2,6,10,15,19-pentamethylicosane (PMI) and crocetane (Koga and Morii, 2005; Rossel et al., 2008; Schouten et al., 1997). Despite the similarity of lipids biomarker structure,

20 Chapter 1

the stable isotopes of lipid could be used as a diagnostic tool to distinguish methanogens and . For example, lipid biomarkers of autotrophic methanogens have intermediately depleted carbon compositions (-30 to -60‰), while ANME lipids are characterized with highly depleted carbon isotope compositions, e.g., -60‰ and -130‰ (Hinrichs et al., 1999; Hinrichs et al., 2000; Pancost et al., 2000). However, Londry et al. (2008) reported a large fractionation by -50‰ between methylated substrate (e.g., TMA, methanol) and lipids, suggesting methanogens could produce more 13C-depleted lipids than commonly realized.

1.7 Objectives and structures of the thesis

To summarize, microbial production of methane is an important terminal process during organic matter degradation, as well as an essential link in marine carbon cycle. Methylotrophic methanogenesis is the least understood methanogenic pathway, and the role of methylated compounds as methanogenic substrates remains largely unconstrained. Therefore, the objective of this PhD thesis is to constrain potential methylated substrates and methylotrophic methanogenic activities, and estimate the relative importance of methylotrophic methanogenesis for methane production in marine sediments. The major questions need to be addressed are: 1) What are the concentrations of methylated compounds in marine sediments? Due to the analytical challenge, the in situ concentrations of methylated compounds are poorly constrained in marine sediment. New sensitive analytical methods will be helpful to determine the concentrations of volatile methylated compounds (e.g., volatile alcohols, TMA and DMS) in marine pore water and sediments. 2) What is the biogeochemical importance of methylated compounds in marine sediments?

21 Chapter 1

So far little is known about the biogeochemistry of methylated compounds in marine sediment. The knowledge on the abundance and distribution would improve our understanding on their biogeochemical importance in marine sediments. 3) What is the dominant methanogenic pathway in deep-sea hypersaline sediment? Based on in vitro studies, methane is believed to be produced from non-competitive compounds in hypersaline environment, while the evidence is not straightforward for this hypothesis. Further comprehensive investigations could provide more convincing evidence to decipher the source of methane in the deep-sea hypersaline sediments. 4) What is the relative importance of different methanogenic pathways in marine sediments? Currently, our understanding of methane production is largely limited to hydrogenotrophic and acetoclastic methanogenesis, and scarce in vitro studies on methylotrophic methanogenesis are mostly from the salt marsh or hypersaline sediments. The comprehensive investigation on methanogenic activities from both competitive and non-competitive substrates in typical marine sediments would provide new insights into the quantitative importance of different methanogenic pathways. 5) What are the potential factors that could influence the methanogenic pathways in marine sediments? The relative importance of different methanogenic pathways might differ from condition to condition, and methanogenic activities measured at geochemically distinct sediments or zonations (e.g., SMTZ), could help to better understand the influence of environmental factors on methane production from different pathways. In this thesis, Chapter 2 reports three different pretreatment methods for trace analysis of volatile alcohols in sedimentary pore waters. With these methods, methanol and ethanol are measured in the pore waters for the first time. Chapter 3 describes the distribution and isotopic composition of TMA and DMS in marine sediment with the new developed analytical protocols, and quantitatively estimates

22 Chapter 1 the contribution of DMS for methane production with isotope mass balance calculations. Chapter 4 presents an integrated study combining biogeochemical analyses, experimental incubation with stable isotope tracers and radiotracers, and gene and lipid biomarkers, to constrain the dominant methanogenic pathway in deep-sea hypersaline sediment. Chapter 5 reports the distribution of methanogenic substrates, methanogenic activity and diversity in the Mediterranean Sea sediments, and further explores the relative importance of different methanogenic pathways and potential influence factors on methane production. Finally, all the major observations are briefly reiterated in Chapter 6, together with future perspectives.

1.8 References

Arnosti, C., 2011. Microbial extracellular enzymes and the marine carbon cycle. Annual Review of Marine Science 3, 401-425. Bertram, S., Blumenberg, M., Michaelis, W., Siegert, M., Krüger, M., Seifert, R., 2013. Methanogenic capabilities of ANME‐archaea deduced from 13C‐labelling approaches. Environmental Microbiology 15, 2384-2393. Canfield, D.E., Kristensen, E., Thamdrup, B., 2005. Aquatic Geomicrobiology. Elsevier. Charlou, J.L., Donval, J.P., Fouquet, Y., Jean-Baptiste, P., Holm, N., 2002. Geochemistry of high

H2 and CH4 vent fluids issuing from ultramafic rocks at the Rainbow hydrothermal field (36°14′N, MAR). Chemical Geology 191, 345-359. Charlson, R., Lovelock, J., Andreae, M., Warren, S., 1987. Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326, 655-661. Conrad, R., 1999. Contribution of hydrogen to methane production and control of hydrogen concentrations in methanogenic soils and sediments. FEMS Microbiology Ecology 28, 193-202. Conrad, R., 2005. Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Organic Geochemistry 36, 739-752. Cord-Ruwisch, R., Seitz, H.-J., Conrad, R., 1988. The capacity of hydrogenotrophic anaerobic bacteria to compete for traces of hydrogen depends on the redox potential of the terminal electron acceptor. Archives of Microbiology 149, 350-357. Crill, P.M., Martens, C.S., 1986. Methane production from bicarbonate and acetate in an anoxic marine sediment. Geochimica et Cosmochimica Acta 50, 2089-2097.

23 Chapter 1

D'Hondt, S., Spivack, A.J., Pockalny, R., Ferdelman, T.G., Fischer, J.P., Kallmeyer, J., Abrams, L.J., Smith, D.C., Graham, D., Hasiuk, F., 2009. Subseafloor sedimentary life in the South Pacific Gyre. Proceedings of the National Academy of Sciences 106, 11651-11656. Dacey, J.W., Blough, N.V., 1987. Hydroxide decomposition of dimethylsulfoniopropionate to form dimethylsulfide. Geophysical Research Letters 14, 1246-1249. Donnelly, M., Dagley, S., 1980. Production of methanol from aromatic acids by Pseudomonas putida. Journal of Bacteriology 142, 916-924. Drake, H.L., 1994. Acetogenesis, acetogenic bacteria, and the acetyl-CoA “Wood/Ljungdahl” pathway: past and current perspectives, Acetogenesis. Springer, pp. 3-60. Dridi, B., Fardeau, M.-L., Ollivier, B., Raoult, D., Drancourt, M., 2012. Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon isolated from human faeces. International Journal of Systematic and Evolutionary Microbiology 62, 1902-1907. Ducklow, H.W., Steinberg, D.K., Buesseler, K.O., 2001. Upper ocean carbon export and the . Oceanography 14, 50-58. Etiope, G., Sherwood Lollar, B., 2013. Abiotic methane on Earth. Reviews of Geophysics 51, 276-299. Ferry, J.G., 1992. Biochemistry of methanogenesis. Critical Reviews in Biochemistry and Molecular biology 27, 473-503. Fiebig, K., Gottschalk, G., 1983. Methanogenesis from choline by a coculture of Desulfovibrio sp. and Methanosarcina barkeri. Applied and Environmental Microbiology 45, 161-168. Finster, K., King, G.M., Bak, F., 1990. Formation of methylmercaptan and dimethylsulfide from methoxylated aromatic compounds in anoxic marine and fresh water sediments. FEMS Microbiology Letters 74, 295-301. Fitzsimons, M.F., Jemmett, A.W., Wolff, G.A., 1997. A preliminary study of the geochemistry of methylamines in a salt marsh. Organic Geochemistry 27, 15-24. Froelich, P.N., Klinkhammer, G., Bender, M.a.a., Luedtke, N., Heath, G.R., Cullen, D., Dauphin, P., Hammond, D., Hartman, B., Maynard, V., 1979. Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Geochimica et Cosmochimica Acta 43, 1075-1090. Garcia, J.-L., Patel, B.K., Ollivier, B., 2000. Taxonomic, phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe 6, 205-226. Grosjean, D., 1995. Atmospheric chemistry of biogenic hydrocarbons: Relevance to the Amazon. Quimica Nova 18, 184-201. Hallam, S.J., Girguis, P.R., Preston, C.M., Richardson, P.M., DeLong, E.F., 2003. Identification of methyl coenzyme M reductase A (mcrA) genes associated with methane-oxidizing archaea.

24 Chapter 1

Applied and Environmental Microbiology 69, 5483-5491. Hallam, S.J., Putnam, N., Preston, C.M., Detter, J.C., Rokhsar, D., Richardson, P.M., DeLong, E.F., 2004. Reverse methanogenesis: testing the hypothesis with environmental genomics. Science 305, 1457-1462. Hayes, J., Freeman, K.H., Popp, B.N., Hoham, C.H., 1990. Compound-specific isotopic analyses: a novel tool for reconstruction of ancient biogeochemical processes. Organic Geochemistry 16, 1115-1128. Hayward, H.R., Stadtman, T.C., 1959. ANAEROBIC DEGRADATION OF CHOLINE I., Vibrio cholinicusn. sp: Fermentation of choline by an anaerobic, cytochrome-producing bacterium1. Journal of Bacteriology 78, 557-561. Hedderich, R., Whitman, W.B., 2006. Physiology and biochemistry of the methane-producing Archaea, The . Springer, pp. 1050-1079. Hedges, J.I., Keil, R.G., 1995. Sedimentary organic matter preservation: an assessment and speculative synthesis. Marine Chemistry 49, 81-115. Heijthuijsen, J.H.F.G., Hansen, T.A., 1989. Anaerobic degradation of betaine by marine Desulfobacterium strains. Archives of Microbiology 152, 393-396. Heuer, V., Elvert, M., Tille, S., Krummen, M., Mollar, X.P., Hmelo, L.R., Hinrichs, K.-U., 2006. Online δ13C analysis of volatile fatty acids in sediment/porewater systems by liquid chromatography-isotope ratio-mass spectrometry. Limnology and Oceanography: Methods 4, 346-357. Heuer, V.B., Pohlman, J.W., Torres, M.E., Elvert, M., Hinrichs, K.-U., 2009. The stable carbon isotope biogeochemistry of acetate and other dissolved carbon species in deep subseafloor sediments at the northern Cascadia Margin. Geochimica et Cosmochimica Acta 73, 3323-3336. Hinrichs, K.-U., Boetius, A., 2003. The anaerobic oxidation of methane: new insights in and biogeochemistry, Ocean Margin Systems. Springer, pp. 457-477. Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewer, P.G., DeLong, E.F., 1999. Methane-consuming archaebacteria in marine sediments. Nature 398, 802-805. Hinrichs, K.-U., Summons, R.E., Orphan, V., Sylva, S.P., Hayes, J.M., 2000. Molecular and isotopic analysis of anaerobic methane-oxidizing communities in marine sediments. Organic Geochemistry 31, 1685-1701. Hoehler, T.M., Alperin, M.J., Albert, D.B., Martens, C.S., 1998. Thermodynamic control on hydrogen concentrations in anoxic sediments. Geochimica et Cosmochimica Acta 62, 1745-1756. Holler, T., Wegener, G., Knittel, K., Boetius, A., Brunner, B., Kuypers, M.M., Widdel, F., 2009.

25 Chapter 1

Substantial 13C/12C and D/H fractionation during anaerobic oxidation of methane by marine consortia enriched in vitro. Environmental Microbiology Reports 1, 370-376. Horita, J., Berndt, M.E., 1999. Abiogenic methane formation and isotopic fractionation under hydrothermal conditions. Science 285, 1055-1057. Jonkers, H.M., Der Maarel, M.J., Gemerden, H., Hansen, T.A., 1996. Dimethylsulfoxide reduction by marine sulfate‐reducing bacteria. FEMS Microbiology Letters 136, 283-287. Jørgensen, B.B., 2000. Bacteria and marine biogeochemistry, Marine Geochemistry. Springer, pp. 173-207. Karl, D.M., Beversdorf, L., Björkman, K.M., Church, M.J., Martinez, A., Delong, E.F., 2008. Aerobic production of methane in the sea. Nature Geoscience 1, 473-478. Keller, M.D., Bellows, W.K., Guillard, R.R., 1989. production in marine phytoplankton. Biogenic sulfur in the environment 393, 167-182. Kelley, C.A., Poole, J.A., Tazaz, A.M., Chanton, J.P., Bebout, B.M., 2012. Substrate limitation for methanogenesis in hypersaline environments. Astrobiology 12, 89-97. Kelley, D.S., Karson, J.A., Früh-Green, G.L., Yoerger, D.R., Shank, T.M., Butterfield, D.A., Hayes, J.M., Schrenk, M.O., Olson, E.J., Proskurowski, G., 2005. A serpentinite-hosted ecosystem: the Lost City hydrothermal field. Science 307, 1428-1434. Kiene, R., Capone, D., 1988. Microbial transformations of methylated sulfur compounds in anoxic salt marsh sediments. Microbial Ecology 15, 275-291. Kiene, R., Taylor, B., 1988. Demethylation of dimethylsulfoniopropionate and production of thiols in anoxic marine sediments. Applied and Environmental Microbiology 54, 2208-2212. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. Kiene, R.P., Visscher, P.T., 1987. Production and fate of methylated sulfur compounds from methionine and dimethylsulfoniopropionate in anoxic salt marsh sediments. Applied and Environmental Microbiology 53, 2426-2434. King, G., Klug, M., Lovley, D., 1983. Metabolism of acetate, methanol, and methylated amines in intertidal sediments of Lowes Cove, Maine. Applied and Environmental Microbiology 45, 1848-1853. Knittel, K., Boetius, A., 2009. Anaerobic oxidation of methane: progress with an unknown process. Annual Review of Microbiology 63, 311-334. Knittel, K., Lösekann, T., Boetius, A., Kort, R., Amann, R., 2005. Diversity and distribution of methanotrophic archaea at cold seeps. Applied and Environmental Microbiology 71, 467-479.

26 Chapter 1

Koga, Y., Morii, H., 2005. Recent advances in structural research on ether lipids from archaea including comparative and physiological aspects. Bioscience, Biotechnology, and Biochemistry 69, 2019-2034. Konhauser, K.O., 2009. Introduction to geomicrobiology. John Wiley & Sons. Kotsyurbenko, O.R., Chin, K.J., Glagolev, M.V., Stubner, S., Simankova, M.V., Nozhevnikova, A.N., Conrad, R., 2004. Acetoclastic and hydrogenotrophic methane production and methanogenic populations in an acidic West‐Siberian peat bog. Environmental Microbiology 6, 1159-1173.

Krüger, M., Eller, G., Conrad, R., Frenzel, P., 2002. Seasonal variation in pathways of CH4

production and in CH4 oxidation in rice fields determined by stable carbon isotopes and specific inhibitors. Global Change Biology 8, 265-280. Krzycki, J., Kenealy, W., DeNiro, M., Zeikus, J., 1987. Stable carbon isotope fractionation by Methanosarcina barkeri during methanogenesis from acetate, methanol, or -hydrogen. Applied and Environmental Microbiology 53, 2597-2599. Kvenvolden, K.A., 2002. Methane hydrate in the global organic carbon cycle. Terra Nova 14, 302-306. L'Haridon, S., Chalopin, M., Colombo, D., Toffin, L., 2014. Methanococcoides vulcani sp. nov., a novel marine methylotrophic methanogen; using betaine, choline and N, N-dimethylethanolamine for methanogenesis, isolated from the Napoli Mud Volcano in the Eastern Mediterranean Sea; and emendation of the genus Methanococcoides. International Journal of Systematic and Evolutionary Microbiology, ijs. 0.058289-058280. Lazar, C.S., John Parkes, R., Cragg, B.A., L'Haridon, S., Toffin, L., 2012. Methanogenic activity and diversity in the centre of the Amsterdam Mud Volcano, Eastern Mediterranean Sea. FEMS Microbiology Ecology 81, 243-254. Lazar, C.S., Parkes, R.J., Cragg, B.A., L'Haridon, S., Toffin, L., 2011. Methanogenic diversity and activity in hypersaline sediments of the centre of the Napoli mud volcano, Eastern Mediterranean Sea. Environmental Microbiology 13, 2078-2091. Lee, C., Olson, B.L., 1984. Dissolved, exchangeable and bound aliphatic amines in marine sediments: initial results. Organic Geochemistry 6, 259-263. Levy, H., 1971. Normal atmosphere: Large radical and formaldehyde concentrations predicted. Science 173, 141-143. Lin, Y.-S., Heuer, V.B., Goldhammer, T., Kellermann, M.Y., Zabel, M., Hinrichs, K.-U., 2012.

Towards constraining H2 concentration in subseafloor sediment: A proposal for combined analysis by two distinct approaches. Geochimica et Cosmochimica Acta 77, 186-201. Lin, Y., Heuer, V., Ferdelman, T., Hinrichs, K.-U., 2010. Microbial conversion of inorganic carbon

27 Chapter 1

to dimethyl sulfide in anoxic lake sediment (Plußsee, Germany). Biogeosciences 7, 2433-2444. Lloyd, K.G., Alperin, M.J., Teske, A., 2011. Environmental evidence for net methane production and oxidation in putative ANaerobic MEthanotrophic (ANME) archaea. Environmental Microbiology 13, 2548-2564. Lomans, B., Van Der Drift, C., Pol, A., Op den Camp, H., 2002. Microbial cycling of volatile organic sulfur compounds. Cellular and Molecular Life Sciences 59, 575-588. Londry, K.L., Dawson, K.G., Grover, H.D., Summons, R.E., Bradley, A.S., 2008. Stable carbon isotope fractionation between substrates and products of Methanosarcina barkeri. Organic Geochemistry 39, 608-621. Lösekann, T., Knittel, K., Nadalig, T., Fuchs, B., Niemann, H., Boetius, A., Amann, R., 2007. Diversity and abundance of aerobic and anaerobic methane oxidizers at the Haakon Mosby Mud Volcano, Barents Sea. Applied and Environmental Microbiology 73, 3348-3362. McCollom, T.M., Seewald, J.S., 2006. Carbon isotope composition of organic compounds produced by abiotic synthesis under hydrothermal conditions. Earth and Planetary Science Letters 243, 74-84. Metcalf, W.W., Griffin, B.M., Cicchillo, R.M., Gao, J., Janga, S.C., Cooke, H.A., Circello, B.T., Evans, B.S., Martens-Habbena, W., Stahl, D.A., 2012. Synthesis of methylphosphonic acid by marine microbes: a source for methane in the aerobic ocean. Science 337, 1104-1107. Michaelis, W., Seifert, R., Nauhaus, K., Treude, T., Thiel, V., Blumenberg, M., Knittel, K., Gieseke, A., Peterknecht, K., Pape, T., 2002. Microbial reefs in the Black Sea fueled by anaerobic oxidation of methane. Science 297, 1013-1015. Milkov, A.V., 2004. Global estimates of hydrate-bound gas in marine sediments: how much is really out there? Earth-Science Reviews 66, 183-197. Mitterer, R.M., 2010. Methanogenesis and sulfate reduction in marine sediments: A new model. Earth and Planetary Science Letters 295, 358-366. Mitterer, R.M., Malone, M.J., Goodfriend, G.A., Swart, P.K., Wortmann, U.G., Logan, G.A., Feary, D.A., Hine, A.C., 2001. Co‐generation of hydrogen sulfide and methane in marine carbonate sediments. Geophysical Research Letters 28, 3931-3934. Möller, B., Oßmer, R., Howard, B., Gottschalk, G., Hippe, H., 1984. Sporomusa, a new genus of gram-negative anaerobic bacteria including Sporomusa sphaeroides spec. nov. and Sporomusa ovata spec. nov. Archives of Microbiology 139, 388-396. Moran, J.J., House, C.H., Vrentas, J.M., Freeman, K.H., 2008. Methyl sulfide production by a novel carbon monoxide metabolism in Methanosarcina acetivorans. Applied and Environmental Microbiology 74, 540-542.

28 Chapter 1

Mori, K., Maruyama, A., Urabe, T., Suzuki, K.-I., Hanada, S., 2008. Archaeoglobus infectus sp. nov., a novel thermophilic, chemolithoheterotrophic archaeon isolated from a deep-sea rock collected at Suiyo Seamount, Izu-Bonin Arc, western Pacific Ocean. International Journal of Systematic and Evolutionary Microbiology 58, 810-816. Niemann, H., Elvert, M., Hovland, M., Orcutt, B., Judd, A., Suck, I., Gutt, J., Joye, S., Damm, E., Finster, K., 2005. Methane emission and consumption at a North Sea gas seep (Tommeliten area). Biogeosciences 2, 335-351. Orcutt, B., Boetius, A., Elvert, M., Samarkin, V., Joye, S.B., 2005. Molecular biogeochemistry of sulfate reduction, methanogenesis and the anaerobic oxidation of methane at Gulf of Mexico cold seeps. Geochimica et Cosmochimica Acta 69, 4267-4281. Oremland, R.S., Kiene, R.P., Mathrani, I., Whiticar, M.J., Boone, D.R., 1989. Description of an estuarine methylotrophic methanogen which grows on dimethyl sulfide. Applied and Environmental Microbiology 55, 994-1002. Oremland, R.S., Marsh, L., DesMarais, D.J., 1982a. Methanogenesis in Big Soda Lake, Nevada: an alkaline, moderately hypersaline desert lake. Applied and Environmental Microbiology 43, 462-468. Oremland, R.S., Marsh, L.M., Polcin, S., 1982b. Methane production and simultaneous sulphate reduction in anoxic, salt marsh sediments. Nature 296, 143-145. Oremland, R.S., Polcin, S., 1982. Methanogenesis and sulfate reduction: competitive and noncompetitive substrates in estuarine sediments. Applied and Environmental Microbiology 44, 1270-1276. Orphan, V.J., House, C.H., Hinrichs, K.-U., McKeegan, K.D., DeLong, E.F., 2002. Multiple archaeal groups mediate methane oxidation in anoxic cold seep sediments. Proceedings of the National Academy of Sciences 99, 7663-7668. Pancost, R.D., Damsté, J.S.S., de Lint, S., van der Maarel, M.J., Gottschal, J.C., 2000. Biomarker evidence for widespread anaerobic methane oxidation in Mediterranean sediments by a consortium of methanogenic archaea and bacteria. Applied and Environmental Microbiology 66, 1126-1132. Parkes, R.J., Brock, F., Banning, N., Hornibrook, E.R.C., Roussel, E.G., Weightman, A.J., Fry, J.C., 2012. Changes in methanogenic substrate utilization and communities with depth in a salt-marsh, creek sediment in southern England. Estuarine, Coastal and Shelf Science 96, 170-178. Parkes, R.J., Cragg, B.A., Banning, N., Brock, F., Webster, G., Fry, J.C., Hornibrook, E., Pancost, R.D., Kelly, S., Knab, N., 2007. Biogeochemistry and biodiversity of methane cycling in subsurface marine sediments (Skagerrak, Denmark). Environmental Microbiology 9,

29 Chapter 1

1146-1161. Paul, K., Nonoh, J.O., Mikulski, L., Brune, A., 2012. “Methanoplasmatales,” Thermoplasmatales-related archaea in termite guts and other environments, are the seventh order of methanogens. Applied and Environmental Microbiology 78, 8245-8253. Reeburgh, W.S., 2007. Oceanic Methane Biogeochemistry. Chemical Reviews 107, 486-513. Rimbault, A., Niel, P., Virelizier, H., Darbord, J.C., Leluan, G., 1988. L-Methionine, a precursor of trace methane in some proteolytic clostridia. Applied and Environmental Microbiology 54, 1581-1586. Rossel, P.E., Lipp, J.S., Fredricks, H.F., Arnds, J., Boetius, A., Elvert, M., Hinrichs, K.-U., 2008. Intact polar lipids of anaerobic methanotrophic archaea and associated bacteria. Organic Geochemistry 39, 992-999. Rossini, F.D., Wagman, D.D., Evans, W.H., 1952. Selected values of chemical thermodynamic properties. US Government Printing Office Washington, DC. Rullkötter, J., 2006. Organic matter: the driving force for early diagenesis, Marine Geochemistry. Springer, pp. 125-168. Sakai, S., Imachi, H., Hanada, S., Ohashi, A., Harada, H., Kamagata, Y., 2008. Methanocella paludicola gen. nov., sp. nov., a methane-producing archaeon, the first isolate of the lineage ‘Rice Cluster I’, and proposal of the new archaeal order Methanocellales ord. nov. International Journal of Systematic and Evolutionary Microbiology 58, 929-936. Schink, B., Zeikus, J., 1980. Microbial methanol formation: a major end product of pectin metabolism. Current Microbiology 4, 387-389. Schink, B., Zeikus, J., 1982. Microbial ecology of pectin decomposition in anoxic lake sediments. Journal of General Microbiology 128, 393-404. Schoell, M., 1988. Multiple origins of methane in the Earth. Chemical Geology 71, 1-10. Schouten, S., Van Der Maarel, M.J.E.C., Huber, R., Damste, J.S.S., 1997. 2,6,10,15,19-Pentamethylicosenes in Methanolobus bombayensis, a marine methanogenic archaeon, and in Methanosarcina mazei. Organic Geochemistry 26, 409-414. Seewald, J.S., 2003. Organic-inorganic interactions in petroleum-producing sedimentary basins. Nature 426, 327-333. Seewald, J.S., Zolotov, M.Y., McCollom, T., 2006. Experimental investigation of single carbon compounds under hydrothermal conditions. Geochimica et Cosmochimica Acta 70, 446-460. Singh, N., Kendall, M.M., Liu, Y., Boone, D.R., 2005. Isolation and characterization of methylotrophic methanogens from anoxic marine sediments in Skan Bay, Alaska: description of Methanococcoides alaskense sp. nov., and emended description of Methanosarcina baltica. International Journal of Systematic and Evolutionary Microbiology 55, 2531-2538.

30 Chapter 1

Smith, N., Shepherd, T., Styles, M., Williams, G., 2005. Hydrogen exploration: a review of global hydrogen accumulations and implications for prospective areas in NW Europe, Geological Society, London, Petroleum Geology Conference series. Geological Society of London, pp. 349-358. Sowers, K.R., Ferry, J.G., 1983. Isolation and characterization of a methylotrophic marine methanogen, Methanococcoides methylutens gen. nov., sp. nov. Applied and Environmental Microbiology 45, 684-690. Strøm, A., Olafsen, J., Larsen, H., 1979. Trimethylamine oxide: a terminal electron acceptor in anaerobic respiration of bacteria. Journal of General Microbiology 112, 315-320. Summons, R.E., Franzmann, P.D., Nichols, P.D., 1998. Carbon isotopic fractionation associated with methylotrophic methanogenesis. Organic Geochemistry 28, 465-475. Tissot, B., Welte, D., 1984. Diagenesis, catagenesis and metagenesis of organic matter. Petroleum Formation and Occurrence. Springer Berlin Heidelberg, pp. 69-73. van der Maarel, M.J., Hansen, T.A., 1997. Dimethylsulfoniopropionate in anoxic intertidal sediments: a precursor of methanogenesis via dimethyl sulfide, methanethiol, and methiolpropionate. Marine Geology 137, 5-12. Wang, X.-C., Lee, C., 1990. The distribution and adsorption behavior of aliphatic amines in marine and lacustrine sediments. Geochimica et Cosmochimica Acta 54, 2759-2774. Watkins, A.J., Roussel, E.G., Parkes, R.J., Sass, H., 2014. Glycine Betaine as a Direct Substrate for Methanogens (Methanococcoides spp.). Applied and Environmental Microbiology 80, 289-293. Watkins, A.J., Roussel, E.G., Webster, G., Parkes, R.J., Sass, H., 2012. Choline and N, N-Dimethylethanolamine as direct substrates for methanogens. Applied and Environmental Microbiology 78, 8298-8303. Wellsbury, P., Goodman, K., Barth, T., Cragg, B.A., Barnes, S.P., Parkes, R.J., 1997. Deep marine biosphere fuelled by increasing organic matter availability during burial and heating. Nature 388, 573-576. White, W.M., 2013. Geochemistry. John Wiley & Sons. Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161, 291-314. Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and

freshwater environments: CO2 reduction vs. acetate fermentation-Isotope evidence. Geochimica et Cosmochimica Acta 50, 693-709. Whitman, W.B., Bowen, T.L., Boone, D.R., 2006. The methanogenic bacteria, The Prokaryotes. Springer, pp. 165-207.

31 Chapter 1

Winfrey, M., Zeikus, J., 1977. Effect of sulfate on carbon and electron flow during microbial methanogenesis in freshwater sediments. Applied and Environmental Microbiology 33, 275-281. Woese, C.R., Kandler, O., Wheelis, M.L., 1990. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proceedings of the National Academy of Sciences 87, 4576-4579. Yancey, P.H., Clark, M.E., Hand, S.C., Bowlus, R.D., Somero, G.N., 1982. Living with water stress: evolution of osmolyte systems. Science 217, 1214-1222. Yoshinaga, M.Y., Holler, T., Goldhammer, T., Wegener, G., Pohlman, J.W., Brunner, B., Kuypers, M.M., Hinrichs, K.-U., Elvert, M., 2014. Carbon isotope equilibration during sulphate-limited anaerobic oxidation of methane. Nature Geoscience 7, 190-194.

32 Chapter 2

Chapter 2

Gas chromatographic analysis of methanol and ethanol in marine

sediment pore waters: Validation and implementation of three

pretreatment techniques

Guang-Chao Zhuang1,2,*, Yu-Shih Lin1,3, Marcus Elvert1, Verena B. Heuer1, Kai-Uwe Hinrichs1

Published in Marine Chemistry, 160, 82-90 (2014).

1. Organic Geochemistry Group, MARUM Center for Marine Environmental Sciences, University of Bremen, 28359 Bremen, Germany 2. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, College of Chemistry and Chemical Engineering, Ocean University of China, 266100 Qingdao, China 3. Department of Oceanography, National Sun Yat-Sen University, No. 70 Lienhai Rd, Kaohsiung 804, Taiwan *Corresponding author: Guang-Chao Zhuang ([email protected])

33 Chapter 2

2.1 Abstract

Low-molecular-weight (LMW) alcohols are produced during the microbial degradation of organic matter from precursors such as lignin, pectin, and carbohydrates. The biogeochemical behavior of these alcohols in marine sediment is poorly constrained but potentially central to carbon cycling. Little is known about LMW alcohols in sediment pore waters because of their low concentrations and high water miscibility, both of which pose substantial analytical challenges. In this study, three alternative methods were adapted for the analysis of trace amounts of methanol and ethanol in small volumes of saline pore waters: direct aqueous injection (DAI), solid-phase microextraction (SPME), and purge and trap (P&T) in combination with gas chromatography (GC) coupled to either a flame ionization detector (FID) or a mass spectrometer (MS). Key modifications included the desalination of samples prior to DAI, and the use of a threaded midget bubbler to purge small-volume samples under heated conditions and the addition of salt during P&T. All three methods were validated for LMW alcohol analysis, and the lowest detection limit (60 nM and 40 nM for methanol and ethanol, respectively) was achieved with the P&T technique. With these methods, ambient concentrations of volatile alcohols were determined for the first time in marine sediment pore waters of the Black Sea and the Gulf of Mexico. A strong correlation between the two compounds was observed and tentatively interpreted as being controlled by similar sources and sinks at the examined stations.

2.2 Introduction

Volatile alcohols such as methanol and ethanol are metabolic intermediates generated during the microbial degradation of organic matter under both oxic and anoxic conditions. Methanol production from pectin and lignin by aerobic and anaerobic microorganisms has been experimentally demonstrated (Donnelly and Dagley, 1980; Schink and Zeikus, 1980) or proposed (Bache and Pfennig, 1981). Ethanol is a well-known fermentation product and is mainly generated from

34 Chapter 2

carbohydrates under anoxic conditions (Eichler and Schink, 1984). Degradation of methanol and ethanol can be coupled to various terminal electron-accepting processes including aerobic respiration (Kolb, 2009), nitrate reduction (Payne, 1973), iron reduction (Daniel et al., 1999), sulfate reduction (Rabus et al., 2006), and methanogenesis (King et al., 1983; Summons et al., 1998; Yoshioka et al., 2009). Among all the microbial processes that involve volatile alcohols, methylotrophic methanogenesis has been the focus of most biogeochemical studies. Methylotrophic methanogenesis is a microbial reaction in which low-molecular-weight (LMW) methylated substrates, including methanol, are converted to methane. The methanogens who use these “non-competitive” substrates can coexist with sulfate-reducing bacteria (Kiene et al., 1986; Oremland et al., 1982), whereas those utilizing hydrogen and acetate, the other two major substrates for methane production, are usually out-competed by sulfate reducers. In other words, methylotrophic substrates allow methanogens to produce methane in anoxic, sulfate-bearing estuaries (King et al., 1983; Oremland and Polcin, 1982) and deep-sea sediment (Mitterer, 2010; Yoshioka et al., 2009). Hence, the biogeochemistry of C-1 and C-2 alcohols is important for our understanding of carbon cycling in marine sediment and other environments rich in lignocellulosic materials, such as soil and lignite coal. To date, a quantitative assessment of the role of volatile alcohols in the biodegradation of organic matter and methylotrophic methanogenesis has been difficult because their in-situ concentrations in soils and sediments have rarely been reported. This gap reflects the lack of a routine sampling protocol and the analytical difficulty associated with the low concentration of volatile alcohols. Several recent studies that employed proton transfer reaction mass spectrometry coupled to different pretreatment techniques successfully quantified methanol and ethanol in seawater (Beale et al., 2011; Kameyama et al., 2010; Williams et al., 2004). Nevertheless, the requirements of large sample volumes (several liters) has hampered the application of these techniques to the interstitial waters of soils and sediments, for which typically only a few milliliters of sample are available. The goal of this study is to validate and optimize three different pretreatment

35 Chapter 2

techniques for the quantification of methanol and ethanol in marine sediment pore waters. Direct aqueous injection (DAI) and solid-phase microextraction (SPME) coupled with gas chromatography (GC) have previously been used for methanol and ethanol analysis in petroleum-contaminated water or body fluids (Gurka et al., 1992; Hong et al., 1999; Lee et al., 1999; Lee et al., 1998; Potter, 1996), but their applications to pore waters have not yet been evaluated. Furthermore, we constructed a purge and trap (P&T) system for the quantification of both methanol and ethanol in small-volume fluid samples. We compared the strengths and weaknesses of the different methods, and implemented these methods to the analysis of sediment pore-waters samples recovered in the Black Sea and Gulf of Mexico.

2.3 Materials and Methods 2.3.1 Chemicals

Methanol (hypergrade for liquid chromatography-mass spectrometry) and ethanol (high-performance liquid chromatography gradient grade) standards were purchased from Merck (Darmstadt, Germany) and Carl Roth GmbH (Karlsruhe,

Germany), respectively. Artificial seawater was prepared from NaCl, KCl, NaHCO3,

MgCl2·6H2O (Sigma-Aldrich Chemie GmbH, Steinheim, Germany), KBr and

CaCl2·2H2O (Merck, Darmstadt, Germany) according to the recipe of (Kester et al., 1967).

2.3.2 Sampling and geochemical analysis

Sediment and pore-water samples were collected by a gravity corer from Station 22 (36° 29.5'E, 42° 13.5'N; ~840 m water depth) in the southeastern Black Sea (RV Meteor cruise M72/5, May-June 2007) and by a multicorer from Station 13 (90° 17.3'W, 27° 7.4'N, 2068 m water depth) in the Gulf of Mexico (RV Atlantis cruise AT18-2, November 2010). The cores were stored at 4 to 8 °C and processed within 24 hours after retrieval. Decompression during core retrieval might lead to modification

36 Chapter 2 of the pore water chemistry, but an evaluation of this effect on the LMW alcohol concentration is beyond the scope of the present study. Sampling protocols differed slightly due to equipment limitations of the two cruises. At Station 22, pore-water samples were extracted from the intact sediment core (ø = 9 cm) via Rhizon suction samplers (0.1 µm porous polymer, Rhizosphere Research, Wageningen, the Netherlands; Seeberg-Elverfeldt et al., 2005), which had been rinsed with at least 30 mL of dilute hydrochloric acid (pH = 1-2) followed by a 30 mL rinse with Milli-Q water. At Station 13, pore-water samples were extracted from 2.5 - 5 cm thick slices of sediment by centrifugation at 5000 rpm for 20 min. Centrifugation may cause cell lysis due to decompression or other factors, whereas the use of Rhizons minimizes the artifacts and disturbance during sampling (Seeberg-Elverfeldt et al., 2005). Because shipboard analysis of LMW alcohols was not possible during the two expeditions, the pore-water samples were stored at -20 °C. As the effect of preservation by freezing samples for volatile alcohols analysis could not be quantitatively assessed, we cannot entirely rule out the possibility of sample alteration during storage. For methane analysis, 3-5 mL of sediment were collected with cut-off plastic syringes, and mixed with either 10 mL of 2.5% NaOH in 20 mL glass serum vials (Station 22) or with 2 mL of 1 M NaOH in a 30 mL glass serum vials (Station 13) that were closed with a butyl septum and crimp cap. For Station 22, details of sediment sampling and gas analysis have been described elsewhere (Riedinger et al., 2010). At Station 13, methane was analyzed by GC-FID (SRI 8610C GC, SRI Instruments, USA) and pore water sulfate was measured using zinc acetate-fixed samples with a Metrohm 883 IC

Basic Plus ion chromatograph equipped with an external CO2 suppressor.

2.3.3 DAI-GC-MS operating conditions

Direct aqueous injection was performed on a Trace GC 2000 (Thermo Finnigan, Milan, Italy) coupled to a DSQ II MS (Thermo Finnigan, Texas, USA) after the standards and pore-water samples had been desalted to reduce column maintenance (Potter, 1996). Desalting was achieved with Dionex OnGuard II Ag and H cartridges

37 Chapter 2

(Thermo Fisher GmbH, Idstein, Germany). The cartridges were flushed with 10 mL of deionized water to remove trace contaminants and to condition the resin, before 3 mL of sample solution, carried in a glass syringe, were pushed through the cartridges at a flow rate of approximately 2 mL/min. The initial 2 mL effluent was discarded, and the remainder was collected in a glass vial for later analysis. To assess the desalting procedure, salinity of the effluent was measured with a refractometer (ATAGO CO., LTD, Tokyo, Japan), the accuracy and reproducibility of which were 1‰ and 0.5‰, respectively. The optimized GC conditions were: programmable temperature vaporization (PTV) injection mode; a split ratio of 1:10; injection temperature of 250 °C; injection volume of 1 µL; helium carrier gas at 3 mL/min; oven temperature at 60 °C for 3 min, ramped to 240 °C at 40 °C/min, and held at 240 °C for 5 min. We used an OPTIMA® WAXplus column (30 m × 0.25 mm I.D., 0.25 µm film thickness; Macherey-Nagel GmbH, Düren, Germany), which has a polar stationary phase of polyethylene glycol and is suitable for aqueous injection. No guard column was used. The MS was operated in selected ion monitoring (SIM) mode with the ion fragment of 31 (m/z) selected for both alcohols. Other MS parameters were: ionization voltage of 70 eV; ion source at 200 °C; transfer line temperature of 250 °C.

2.3.4 Headspace SPME-GC-FID operating conditions

Carboxen-polydimethylsiloxane (CAR/PDMS) fibers (75 μm, Supelco, Bellefonte, USA) recommended for methanol analysis (Lee et al., 1999) were used to extract the alcohols. A blank injection was performed to clean and condition the fiber before sample analysis. One milliliter of standard or sample was transferred to a 4-mL vial that contained ~0.5 g Na2SO4 (Carl Roth GmbH, Karlsruhe, Germany). The vial was immediately sealed with a polytetrafluoroethylene (PTFE) septum and crimp-capped. To extract the analytes, the fiber was exposed for 20 min to the headspace of the sample container while the solution was magnetically stirred (500 rpm) at 30 °C. After extraction, the analytes were thermally desorbed for 2 min in a

38 Chapter 2

split/splitless injector (250 °C) equipped with a deactivated SPME liner (Supelco). Analysis was performed using a Hewlett-Packard GC 5890 series II (Hewlett-Packard GmbH, Waldbronn, Germany) equipped with an FID and an Alltech Heliflex® AT-Q capillary column (30 m × 0.32 mm I.D.; Alltech Associates, Inc., Deerfield, US) that held a porous divinylbenzene polymer stationary phase. The GC-FID conditions were: splitless injection and the split/splitless purge valve opened for 2 min after injection; helium carrier gas at 3 mL/min; oven temperature of 120 °C upon injection, raised to 240 °C at 20 °C/min, and held at that temperature for 2 min. All optimization tests were carried out using a mixed standard containing 580 µM of methanol and 430 µM of ethanol in Milli-Q water.

2.3.5 P&T-GC-FID operating conditions

A cryogenic P&T system, modified from (Andreae and Barnard, 1983), was used for the alcohol preconcentration (Fig. 2.1). Our modifications include the following: (1) We employed a 20-mL glass vial fitted with a threaded midget bubbler (Supelco) instead of a bubbling chamber as our sparging chamber. (2) Rather than trapping and separating on the same chromatography column, the alcohols were first trapped cryogenically in a PTFE tube and then released upon heating into the GC capillary column for separation and detection. (3) All tubing and connectors in the system were made of PTFE or perfluoroalkoxy because stainless steel and brass were found to significantly reduce the methanol signal. A helium carrier stream, which was controlled and monitored with a toggle valve (Swagelok, Hamburg, Germany) and gas flow controller (Valco Instruments Co. Inc., Houston, US), passed through a pressure gauge (Supelco) and the threaded midget bubbler to purge the solution in the 20-mL vial that was heated in a water bath to 85 °C. To introduce the sample, the sample vial was disconnected and 6 mL of standard or sample were injected into the

vial to fully immerse the glass frit of the bubbler. After addition of Na2SO4 to the saturation level (~2.5 g), the vial was immediately connected to the system for purging. A Pyrex glass tube (12 cm long and 10 mm O.D.) filled with K2CO3

39 Chapter 2

(Sigma-Aldrich Chemie GmbH, Steinheim, Germany) to two-thirds of the total volume was used to remove moisture from the carrier gas stream which then entered the cold trap via a six-port Sulfinert™ valve (Restek GmbH, Bad Homburg, Germany). A PTFE tube (50 cm long, 1/8 inch O.D.) was used for the cold trap and looped twice to form 5-cm diameter coils. The coils were placed 10 cm below the connections to the six-port valve and submerged in liquid nitrogen during purging. After 30 min of purging and trapping, the trap was transferred into a heating sleeve and heated for 3 min at 200 °C to vaporize the analytes. The heating sleeve was made of silicon heating tape (MiL Heating Systems GmbH, Ubstadt-Weiher, Germany) that was wrapped to form a bag covered with aluminum foil, and connected to an electronic temperature regulator through a sensor (MiL Heating Systems GmbH). The analytes were then introduced into the GC by switching the six-port valve.

Fig. 2.1 Purge and trap GC-FID system used for pore water alcohol analysis.

Analysis was performed using a Hewlett-Packard GC 5890 series II equipped with an FID and Alltech Heliflex® AT-Q capillary column. The oven temperature was programmed to rise from 40 °C to 240 °C at 20 °C/min and then maintained constant for 6 min to fully eliminate water in the column. Other GC parameters were the same as those described for the SPME method. Concentrations of methanol and ethanol standards used for optimization were 11.6 μM and 8.5 μM, respectively.

40 Chapter 2

2.3.6 Standards, contamination tests, and procedural blank

DAI standards were prepared in artificial seawater to evaluate the efficiency of desalting cartridges and SPME standards were prepared with Milli-Q water. Tap water was used for the most sensitive P&T method since tap water consistently yielded the lowest level of LMW alcohols compared to our laboratory supply of deionized or Milli-Q water. The use of tap water to prepare standard solutions has also been reported in previous studies of volatile compounds (e.g., Zwank et al., 2003). Milli-Q or tap water rather than artificial seawater was used for SPME and P&T standards, as final standard and sample analyses of these two methods required extra salt addition and the preparation of artificial seawater could increase the risk of contamination. Furthermore, storage of pore-water samples was avoided in the vicinity of possible contamination sources, and analyses were conducted in an alcohol-free lab, as confirmed by the absence of alcohols in blank samples that were exposed to ambient lab air for up to two weeks. Procedural blank tests confirmed that new Rhizons did not contribute detectable contaminants when they had been carefully rinsed with diluted acid and Milli-Q water. Nevertheless, repeated use of the Rhizons increased the risk of contamination and we found evidence for contamination (e.g., ~2 nmol of both alcohols were detected in 6-ml rinsates, P < 0.01) in test Rhizons. Hence, we recommend that Rhizons be rinsed thoroughly with diluted acid and Milli-Q water before use, and the final fraction of the eluting Milli-Q be collected as procedural control from a representative set of Rhizons, analyzed and treated as a blank in the evaluation of methanol and ethanol concentrations in pore-water samples.

2.3.7 Statistical analyses

The significance of differences in blanks in SPME and P&T optimization conditions was tested using one-way analysis of variance (one-way ANOVA) with a significance level (α) set at 0.05.

41 Chapter 2

2.4 Results and discussion 2.4.1 Optimization of direct aqueous injection, solid phase microextraction, and purge and trap analysis 2.4.1.1Direct aqueous injection (DAI)

Whereas DAI does not involve a pre-concentration step and is therefore the simplest of the three investigated techniques, the typically small injection volumes permitted by GC limits the sensitivity of DAI. Multiple tests carried out in this study showed that the performance and applicability of DAI for LMW alcohols analysis in pore waters depend on optimization of (1) salt removal, (2) injection conditions, and (3) choice of column and detector. Since salt can cause deterioration of the column, pore-water samples must be desalinated before GC analysis. In this study, we employed Dionex OnGuard II Ag and H cartridges to desalt the standard and sample solution. These cartridges have been used to remove chloride ions for the ion chromatographic analysis of acetate (Lang et al., 2010), but have not been used in combination with DAI-GC. Refractometric analysis of the cartridge eluates showed complete removal of salts from artificial seawater. DAI-GC analysis confirmed the absence of methanol or ethanol bleeding from the cartridge and a high recovery (> 90%; Fig. 2.2) of both alcohols from desalted standards. To determine optimal conditions, the PTV and split/splitless injector were systematically compared. Use of the PTV injector with split ratio of 1:10 resulted in superior peak shape and reproducibility compared to the split/splitless injector. The injection volume was limited to 1 μL, as larger volumes caused backflush and affected the reproducibility and sensitivity of the analysis.

42 Chapter 2

(a) 7 Spiked MilliQ water Cartridge-treated artificial seawater standard 6 ) 5 2

10 5 R = 0.995

4 R2 = 0.996 3

2

Response(peak area × 1

0 0 100 200 300 400 500 Methanol (µM)

(b) Spiked MilliQ water 10 Cartridge-treated artificial seawater standard ) 5 8 10 R2 = 0.998 6 R2 = 0.997 4

2 Response(peak area ×

0 0 50 100 150 200 250 300 350 Ethanol (µM)

Fig. 2.2 Five-point calibration curves for (a) methanol and (b) ethanol performed with spiked Milli-Q water and cartridge-treated artificial seawater with DAI-GC-MS. Difference in slopes can be explained by slight losses during cartridge treatment.

The OPTIMA® WAXplus column has a polar stationary phase of polyethylene glycol that is suitable for aqueous injection and gave sharp sequential peaks of methanol and ethanol, followed by water. Both DAI-GC-FID and DAI-GC-MS methods were tested for detection, but the former was later only used a few times to measure standards in order to determine extraction efficiencies of other methods (cf. Table 2.1). DAI-GC-FID was less favored for routine analysis because of frequent

43 Chapter 2

flame extinction by the injected water, a problem acknowledged earlier for DAI-GC-FID (Potter, 1996). With the SIM mode of GC-MS, both alcohols were detected as well-resolved peaks atop a flat baseline, facilitating peak integration and quantification. After elution of ethanol, the MS data acquisition was terminated to minimize the amount of water entering the ionization chamber.

2.4.1.2 Solid phase microextraction (SPME)

SPME can be conducted by either immersing the fibers into the solution or exposing the fibers to the headspace of the sample vial. The latter approach, the headspace SPME, has the additional advantage of being less sensitive to matrix interferences (Cassada et al., 2000). After initial tests provided evidence that headspace SPME with salting-out agents yielded a higher recovery of methanol and ethanol standards than direct immersion (Fig. 2.3a), we further investigated several parameters that could affect the extraction efficiency of headspace SPME, including (1) type and dosage of salting-out agent (Fig. 2.3a), (2) extraction temperature (Fig. 2.3b), and (3) fiber exposure duration (Fig. 2.3c). Standards treated with saturated

Na2SO4 (Fig. 2.3a, P < 0.001) at an extraction temperature of 30 °C (Fig. 2.3b, P < 0.01) provided the largest signals for both alcohols. The methanol response did not improve after 10 min of extraction (Fig. 2.3c, P = 0.11), whereas acceptable equilibrium for ethanol was achieved after 20 min (Fig. 2.3c, P < 0.01). Thus, the final optimized headspace SPME conditions were set for a 30 °C extraction for 20 min of Na2SO4-saturated sample. Our optimized operating conditions are very close to those reported for alcohol analysis in other non-salty sample matrices (Sales and de Lourdes Cardeal, 2003) and attest to the relative insensitivity of headspace SPME to matrix interferences. Further improvements of the extraction efficiency of methanol and ethanol by headspace SPME may require fundamental analytical developments, such as the design of new fiber materials that are highly specific to LMW alcohols.

44 Chapter 2

(a) 2000 Direct immersion without salt addition Headspace without salt addition* 10% NaCl (w/w) 20% NaCl (w/w) 1600 30% NaCl (w/w) Saturated NaCl

10% Na2SO4 (w/w) 30% Na2SO4 (w/w) Saturated Na SO 1200 2 4

800

Relativeintensity (%) 400

0 Methanol Ethanol Salt addition

(b) 200 180 Methanol

Ethanol 150 (%) intensity relative Ethanol 160

120 120 90 80 60

40 30 Methanolrelative intensity(%)

0 0 20* 30 40 50 60 70 80 Extraction temperature (oC)

(c) 140 320

120 280 Ethanol relative intensity (%) intensity relative Ethanol 240 100 200 80 160 60 120 40 Methanol 80 Ethanol 20

Methanolrelative intensity(%) 40

0 0 0 10* 20 30 40 Exposure time (min)

Fig. 2.3 Effect of (a) salt addition, (b) extraction temperature, and (c) exposure time on the relative signal intensities of methanol and ethanol measured by headspace SPME. The relative signal intensity (%) was defined as the peak area at optimization conditions relative to peak area at reference conditions (reference condition was considered as 100%, and denoted with * in each panel). Error bars represent the standard deviations from the replicate measurements.

2.4.1.3 Purge and trap

The main challenge of LMW alcohol analysis by P&T is the high water miscibility and polar nature of the analytes, both of which result in poor extraction efficiency. In previous studies of seawater (Beale et al., 2011), this problem was mainly circumvented by purging large sample volumes. However, in our case, the

45 Chapter 2

sample volume was limited and the P&T system required fine tuning of operating conditions to maximize the extraction efficiency. We employed a 20-mL glass vial fitted with a threaded midget bubbler, which allows for mild heating of the solution in a water bath and the addition of salt during purging. Furthermore, we systematically assessed a series of purging conditions (purge time, purge flow, purge temperature and ionic strength) their impact on the extraction efficiency and evaluated dehydration protocols before the analytes entered the cold trap. Various purge parameters including purge time (10 to 35 min), purge flow (20 to 150 mL/min), temperature of the water bath (ambient to 85 °C), and the use of salting-out agents (NaCl or Na2SO4) were assessed to improve the sensitivity. The extraction efficiency generally increased with longer purge time, higher purge flow, and higher water-bath temperature (Fig. 2.4a, P < 0.001 for methanol and P < 0.01 for ethanol; Fig. 2.4b, P < 0.001 for both alcohols; Fig. 2.4c, P < 0.001 for both alcohols). The ultimate purge conditions were set at 30 min with a flow rate of 120 mL/min at 85 °C; purge conditions harsher than this were found to cause trap clogging. Addition of salting-out agents increased the extraction efficiency without contribution to trap clogging, but the use of saturated NaCl (36% w/w) blocked the bubbler and led to a

reduced response (Fig. 2.4d, P < 0.01 for both alcohols). Therefore, Na2SO4 at saturation (40% w/w) was chosen as our salting-out condition. Since relatively harsh purge conditions vaporize a significant amount of water, the risk of trap clogging increases. An efficient water elimination unit with a low alcohol affinity is sorely needed for drying the purge gas before it enters the liquid nitrogen trap. We evaluated two dehydration options, one with a Pyrex glass tube filled with dehydration reagents and the other with an empty U-shaped glass tube immersed in a slush bath (-25 °C) prepared with a mixture of o-xylene and liquid

nitrogen. Among all tested dehydration reagents, K2CO3 was favored over CaCl2,

MgSO4, CaSO4 and molecular sieve (3 Å) because of the relatively high response and

minimum risk of trap clogging under our purge condition. The K2CO3 drier also allowed for higher recoveries of 181% and 226% for methanol and ethanol,

respectively, relative to the U-shaped cold-bath drier. Thus, the K2CO3 drier was

46 Chapter 2

employed for subsequent analyses. Before each run, the whole system was flushed for 3 min with pure purge gas to clean the transfer lines and avoid carryover. No

carryover effect was observed when K2CO3 was replaced in the drier after the run of a concentrated sample.

Fig. 2.4 Effect of (a) purge time, (b) purge flow, (c) purge temperature, and (d) salt addition on the response of methanol and ethanol pre-concentrated with P&T. The relative signal intensity (%) was defined as the peak area at optimization conditions relative to peak area at reference conditions (reference condition was considered as 100%, and denoted with * in each panel). Error bars represent the standard deviations from the replicate measurements.

2.4.2 Validation and comparison of DAI-GC-MS, SPME-GC-FID and P&T-GC-FID

Table 2.1 summarizes sample requirements, sample processing time, and the performance of three pretreatment techniques under the optimized conditions.

47 Chapter 2

Table 2.1 Validation and comparison of different methods for methanol and ethanol analysis

Method Sample volume Detection limit Linearity (R2)b Reproducibility (%)c Recovery & Extraction Run time (mL) (µM)a efficiency (%)d (min)

Methanol Ethanol Methanol Ethanol Methanol Ethanol Methanol Ethanol DAI 3 for 7.4 3.1 0.996 0.997 8 7 97 ± 8 91 ± 9 12 GC-MS preparation (23-450 μM) (15-302 μM) (112 μM) (76 μM) (1 μL for injection) SPME 1 5.2 0.4 0.995 0.999 7&5 5&7 0.13 ± 0.01 4.0 ± 0.5 30 GC-FID (11.6-116 μM) (2.5-49 μM) (15 μM& (12 μM& 580 μM) 430 μM) P&T 6 0.06 0.04 0.999 0.996 7 6 7.9 ± 1.1 7.6 ± 1.2 50 GC-FID (0.2-5.8 μM) (0.2-4.3 μM) (11.6 μM) (8.5 μM) a Method detection limits (MDLs) were determined as 3.14 times the standard deviation of replicate measurements (n = 7) of standards spiked with low concentrations of analytes according to EPA. b Linearity of each method was performed within different concentration ranges (indicated in the bracket, n = 5). c Reproducibility was calculated from the relative standard deviation of repeated measurements of the same standard solution (n = 6) at different concentrations (indicated in the bracket). d The recovery for DAI was calculated from the ratios of GC response of cartridge-treated artificial seawater standards to that of spiked Milli-Q water. The extraction efficiency for SPME and P&T was calculated as the ratio of the peak area obtained by SPME or P&T to that of peak area obtained by direct injection of an equivalent mass of standards on the same detector (FID). Errors indicated the confidence intervals with a confidence level of 95%, n = 5.

48 Chapter 2

The method detection limits (MDLs) for methanol and ethanol during implementation of the DAI-GC-MS method were 7.4 and 3.1 µM, respectively. SPME has a similar MDL for methanol (5.2 µM), but the MDL for ethanol was one order of magnitude lower (0.4 µM). The lowest MDLs for methanol (60 nM) and ethanol (40 nM) were achieved by the P&T procedure with a sample volume of 6 mL. Excellent linearity of the external calibration curves was achieved for all three methods (Fig. 2.2 and Table 2.1). The analytical reproducibility of these methods, assessed by repeatedly measuring the same standard solution (N = 6), were 5-8% for both alcohols. The recovery of DAI-GC-MS, calculated from the ratios of the GC response of cartridge-treated artificial seawater standards to that of spiked Milli-Q water, was 97 ± 8% and 91 ± 9% (P = 95%, n = 5) for methanol (23-450 μM) and ethanol (15-302 μM) standards when combined with the desalting procedure (Fig. 2.2), or nearly quantitative recovery of both alcohols. Due to the high water-miscibility of alcohols, the absolute extraction efficiencies of SPME and P&T for methanol (SPME: 0.13 ± 0.01%, P&T: 7.9 ± 1.1%; P = 95%, n = 5) and ethanol (SPME: 4.0 ± 0.5%, P&T: 7.6 ± 1.2%; P = 95%, n = 5) were much lower than the recovery of DAI. In spite of the low extraction efficiencies, the good reproducibility and linear response ensured accurate quantification of the SPME and P&T methods. DAI-GC-MS is a versatile method for alcohol analysis with short processing time, requiring simple equipment and straightforward sample pretreatment. Its nearly quantitative extraction efficiency makes it a potentially suitable method to be coupled with on-line stable carbon isotope analysis (δ13C). In recent decades, compound-specific δ13C analysis has advanced the understanding of critical biogeochemical processes, e.g., δ13C of methane has been applied extensively for the identification and quantification of methane sources and sinks (Whiticar et al., 1999). Likewise, the δ13C composition of volatile alcohols could provide further insights into their biogeochemical role in the environment. Nevertheless, the MDL of the DAI-GC-MS method is limited by the minute injection volume of 1 μL. Furthermore, frequent system maintenance for the DAI method is required because of troublesome effects (e.g., peak tailing, sensitivity, and reproducibility) due to water and buildup of

49 Chapter 2

non-volatile constituents in the liner, analytical column, and MS ion source. Hence, DAI is a good option for fast monitoring of alcohols in systems with concentrations in the higher micromolar range, e.g., possibly in laboratory incubations of sediments and soils. SPME provides a solvent-free means to extract volatile compounds from aqueous solutions, which is particularly advantageous for samples with a complex matrix. Compared with the DAI-GC-MS approach, the SPME-GC-FID method fails to significantly improve the methanol MDL but does lower the MDL of ethanol by one order of magnitude. The absolute extraction efficiencies of SPME were the poorest among the three methods. With slightly larger sample volume requirements and longer processing time, the P&T procedure has the lowest MDLs, one to two orders of magnitude lower than those of the other two methods. As demonstrated below, the P&T procedure provides sufficient sensitivity for sediment pore-water samples. Therefore, P&T is recommended for accurate analysis of natural samples with trace levels of alcohols. Nevertheless, the low absolute extraction efficiencies of volatile alcohols for the P&T and SPME methods will disallow stable isotope analysis, as isotope fractionation may occur during extraction. All three techniques were validated for volatile alcohol analysis, and the method of choice depends on the expected concentration range in the samples and the available sample volume.

2.4.3 Analysis of marine sediment pore-water samples

The application of the optimized analytical methods for methanol and ethanol was assessed using sediment pore-water samples retrieved from the Black Sea and Gulf of Mexico. This is the first time that the concentrations of methanol and ethanol in deep-sea sediment have been determined quantitatively. Because sample volume limitations did not permit a parallel comparison of all three methods, we applied two of the validated methods in each sample set to evaluate their reliability. Black Sea samples were analyzed by DAI-GC-MS and P&T-GC-FID and samples from the Gulf

50 Chapter 2

of Mexico by SPME-GC-FID and P&T-GC-FID (Fig. 2.5). We note that the extraction efficiency of alcohols and other potential analytes from sediment of the two stations might not be identical. This is not only because of the different pore-water extraction protocols applied, but also because of the different mineralogy, which, as has been observed for other LMW compounds (Ertefai et al., 2010; Ijiri et al., 2009; Lee and Olson, 1984), may affect the adsorption behaviors of both compounds on particles. In the Black Sea, P&T-GC-FID revealed relatively high alcohol concentrations in the shallowest sample at 51 cm below the seafloor (cmbsf) (methanol: 6 µM, ethanol: 3 µM) and at 507 cmbsf (methanol: 5 µM, ethanol: 5 µM), and low alcohol concentrations of <1 μM at all the other depths (Fig. 2.5). Analyses by DAI-GC-MS confirmed the presence of elevated concentrations of methanol (14-15 µM) and ethanol (6-7 µM) at 51 and 547 cmbsf with the measured values were slightly higher than those obtained by P&T-GC-FID. At all the other depths, the concentrations were below the MDL of the DAI method. Although the measured concentrations varied, the vertical distributions of both alcohols obtained by DAI and P&T were generally consistent, and the discrepancy might be attributed to the different pretreatments of the sample matrix and the bias of quantification with different detectors. In the Gulf of Mexico, methanol was detectable by SPME-GC-FID with concentrations ranging from 11-78 µM except for a few depths (2.5, 7.5, and 22.5 cmbsf) where the concentrations were below the MDL (Fig. 2.5). Ethanol was detected at all depths, varying from 3 to 62 μM. P&T-GC-FID analysis revealed a similar pattern for the downcore distribution. The lowest methanol concentration of 2 μM was found at 5 cmbsf, and the maximum of 69 µM was at 32.5 cmbsf. Ethanol concentrations varied from 3 to 43 μM with a maximum also recorded at 32.5 cmbsf. The concentration values obtained with the P&T method were in excellent agreement with the results obtained by SPME (Fig. 2.5).

51 Chapter 2

Fig. 2.5 Depth profiles of methanol (a and d), ethanol (b and e), CH4 (c and f) and sulfate concentrations (f) in pore waters from Black Sea and Gulf of Mexico sediments. Sulfate concentrations were not available for the Black Sea samples. Error bars in (d) and (e) (some are smaller than the symbol in the figure) represent the standard deviations from duplicated measurements. MDL: method detection limit.

One intriguing observation is that methanol and ethanol concentrations are linearly correlated at both stations (Fig. 2.6a). This behavior is strikingly different from that of the corresponding acids, i.e., formic acid and , which show distinct concentration profiles in marine sediments (D’Hondt et al., 2003). In order to verify that our observed correlation was not caused by a procedural artifact or by contamination, we conducted additional experiments with the P&T method. Natural seawater was spiked with methanol and ethanol at various concentration ratios (1:10, 1:2, 2:1, 10:1) and the solutions were analyzed by P&T-GC-FID. The results showed

52 Chapter 2

that the measured methanol and ethanol concentrations were independent of each other and reflected the actual spiked concentrations (Fig. 2.6b).

Fig. 2.6 (a) The relationship between methanol and ethanol in pore waters from Black Sea (+) and Gulf of Mexico (×) sediments. Note that the extremely high concentration at 32.5 cmbsf from Gulf of Mexico was not included in the regression analysis, as the correlation would be significantly driven by this single data point. (b) Signal of different concentration ratios of methanol (blue) and ethanol (red) spiked into seawater detected by P&T-GC-FID. Methanol and ethanol ratios are 1:10 (diamond), 1:2 (rectangle), 2:1 (triangle), and 10:1 (circle).

Contamination during processing in our laboratory was also ruled out by various tests (see section 2.3.6) and, furthermore, it is highly unlikely that the sediments were contaminated by equal amounts of both alcohols during two distinct expeditions during which different sampling protocols were applied by different science crews and using different equipment, etc. We therefore conclude that the methanol and ethanol concentration profiles reflect their biogeochemical behavior in these marine

53 Chapter 2

sediment cores. The correlation implies that both alcohols may originate from the same organic matter pool, e.g., lignocellulosic materials, and are consumed by similarly active sinks. Further screening of a more diverse set of samples will be required to provide a mechanistic understanding of the biogeochemical processes governing the distribution of these compounds in marine sediments. Based on the sparse data available to date, it would appear that, unlike volatile fatty acids (D’Hondt et al., 2003; Ijiri et al., 2012), the pore-water distribution of the two alcohols in the sediments of the Gulf of Mexico is unrelated to the concentrations of bulk and nitrogen. The accumulation of alcohols at 51 cmbsf in the Black Sea, presumably, might be associated with active and net production from relatively fresh and abundant macromolecular precursors in shallower sediment (e.g., lignin or pectin for methanol: Donnelly and Dagley, 1980; Schink and Zeikus, 1980; lignin distribution: Staniszewski et al., 2001; cellulose for ethanol: Eichler and Schink, 1984). The subsurface alcohol maximum in the Black Sea, on the other hand, appears at the transition to the methanogenic zone (Fig. 2.5), where methane concentrations increase sharply and the highest rates of anaerobic oxidation of methane can be found (Knab et al., 2009; Riedinger et al., 2010). DAI-GC analysis of pore-water samples taken from another Black Sea station (cf. Riedinger et al., 2010; data not shown) confirmed the relationship between the two alcohols and the transition to the methanogenic zone because alcohols were only detectable (methanol = 16.9 µM, ethanol = 12.2 µM) in this horizon. Hoehler et al. (1999) reported the accumulation of other fermentation

products such as acetate and H2 in the sulfate-methane transition zone. They postulated that it reflects a temporary decoupling of production by fermentative bacteria and consumption by terminal microorganisms (Alperin et al., 1994). Hence, the elevated alcohol levels at the transition to the methanogenic zone may reflect a similar production-consumption imbalance accompanying the shift in the dominant terminal electron-acceptor process. Two additional features stand out in the Gulf of Mexico profile. First, the methanol concentration was particularly low (2 μM) in the uppermost sampling

54 Chapter 2

interval (0-5 cmbsf). This is attributed to the presence of oxygen in the bottom water, which promotes aerobic methanol oxidation (Kolb, 2009). Second, both alcohols reached tens of micromolar concentrations at the deepest sampling depth (32.5 cmbsf). This peak does not likely result from a local production-consumption imbalance, as inferred for the Black Sea sample, because geochemical data do not show a transition in the redox regime. Additional downcore surveys will be necessary to explain the distribution pattern of these alcohols in near-surface sediment.

2.5 Conclusion

We adapted and validated three methods to quantify methanol and ethanol in marine sediment pore waters and successfully analyzed deep-sea samples for the first time using all three methods. Results obtained by different methods were in good agreement, but the P&T-GC-FID method was superior to the other two in terms of MDL. The main findings for the natural sediment samples were: 1) in both the Black Sea and Gulf of Mexico sediments, methanol and ethanol concentrations were correlated, implying common sources and sinks; 2) slightly higher concentrations of alcohols were detected at the transition to the methanogenic zone in the Black Sea sediment; 3) average concentrations of both alcohols were higher in the Gulf of Mexico than in the Black Sea. Given the limited sampling carried out in this study, the extent to which these features represent the general distribution of alcohols in marine sediment should be further evaluated.

2.6 Acknowledgements

We thank the captains, crew as well as the scientists of RV Meteor cruise M72/5 and RV Atlantis cruise AT18-2 for their kind assistance during sampling. E. P. Reeves, X. Prieto-Mollar and J. Wendt are collectively acknowledged for their suggestions and support on method development. F. Elling provided additional geochemical data for sediments collected from the Gulf of Mexico. We would like to cordially thank the

55 Chapter 2 associate editor and the anonymous reviewers for their constructive comments that improved the quality of the manuscript. The study was supported by the DFG through the Research Center/Excellence Cluster MARUM-Center for Marine Environmental Sciences, Project GB2. G.-C. Zhuang was sponsored by the China Scholarship Council (CSC) and the Bremen International Graduate School for Marine Sciences (GLOMAR).

2.7 References

Alperin, M., Albert, D., Martens, C., 1994. Seasonal variations in production and consumption rates of in an organic-rich coastal sediment. Geochimica et Cosmochimica Acta 58, 4909-4930. Andreae, M., Barnard, W., 1983. Determination of trace quantities of dimethyl sulfide in aqueous solutions. Analytical Chemistry 55, 608-612. Bache, R., Pfennig, N., 1981. Selective isolation of Acetobacterium woodii on methoxylated aromatic acids and determination of growth yields. Archives of Microbiology 130, 255-261. Beale, R., Liss, P.S., Dixon, J.L., Nightingale, P.D., 2011. Quantification of oxygenated volatile organic compounds in seawater by membrane inlet-proton transfer reaction/mass spectrometry. Analytica Chimica Acta 706, 128-134. Daniel, R., Warnecke, F., Potekhina, J.S., Gottschalk, G., 1999. Identification of the syntrophic partners in a coculture coupling anaerobic methanol oxidation to Fe(III) reduction. FEMS Microbiology Letters 180, 197-203. Donnelly, M., Dagley, S., 1980. Production of methanol from aromatic acids by Pseudomonas putida. Journal of Bacteriology 142, 916-924. Eichler, B., Schink, B., 1984. Oxidation of primary aliphatic alcohols by Acetobacterium carbinolicum sp. nov., a homoacetogenic anaerobe. Archives of Microbiology 140, 147-152. Gurka, D.F., Pyle, S.M., Titus, R., 1992. Environmental analysis by direct aqueous injection. Analytical Chemistry 64, 824-831.

Hoehler, T.M., Albert, D.B., Alperin, M.J., Martens, C.S., 1999. Acetogenesis from CO2 in an anoxic marine sediment. Limnology and Oceanography, 662-667. Hong, S., Duttweiler, C.M., Lemley, A.T., 1999. Analysis of methyl tert.-butyl ether and its degradation products by direct aqueous injection onto gas chromatography with mass spectrometry or flame ionization detection systems. Journal of Chromatography A 857, 205-216.

56 Chapter 2

Kameyama, S., Tanimoto, H., Inomata, S., Tsunogai, U., Ooki, A., Takeda, S., Obata, H., Tsuda, A., Uematsu, M., 2010. High-resolution measurement of multiple volatile organic compounds dissolved in seawater using equilibrator inlet–proton transfer reaction-mass spectrometry (EI-PTR-MS). Marine Chemistry 122, 59-73. Kester, D.R., Duedall, I.W., Connors, D.N., Pytkowicz, R.M., 1967. Preparation of Artificial Seawater. Limnology and Oceanography 12, 176-179. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. King, G.M., Klug, M., Lovley, D., 1983. Metabolism of acetate, methanol, and methylated amines in intertidal sediments of Lowes Cove, Maine. Applied and Environmental Microbiology 45, 1848-1853. Kolb, S., 2009. Aerobic methanol-oxidizing Bacteria in soil. FEMS Microbiology Letters 300, 1-10. Lang, S.Q., Butterfield, D.A., Schulte, M., Kelley, D.S., Lilley, M.D., 2010. Elevated concentrations of formate, acetate and dissolved organic carbon found at the Lost City hydrothermal field. Geochimica et Cosmochimica Acta 74, 941-952. Lee, X.P., Kumazawa, T., Kondo, K., Sato, K., Suzuki, O., 1999. Analysis of methanol or formic acid in body fluids by headspace solid-phase microextraction and capillary gas chromatography. Journal of Chromatography B: Biomedical Sciences and Applications 734, 155-162. Lee, X.P., Kumazawa, T., Sato, K., Seno, H., Ishii, A., Suzuki, O., 1998. Improved extraction of ethanol from human body fluids by headspace solid-phase microextraction with a carboxen-polydimethylsiloxane-coated fiber. Chromatographia 47, 593-595. Mitterer, R.M., 2010. Methanogenesis and sulfate reduction in marine sediments: A new model. Earth and Planetary Science Letters 295, 358-366. Oremland, R.S., Marsh, L.M., Polcin, S., 1982. Methane production and simultaneous sulphate reduction in anoxic, salt marsh sediments. Nature 296, 143-145. Oremland, R.S., Polcin, S., 1982. Methanogenesis and sulfate reduction: competitive and noncompetitive substrates in estuarine sediments. Applied and Environmental Microbiology 44, 1270-1276. Payne, W., 1973. Reduction of nitrogenous oxides by microorganisms. Bacteriological Reviews 37, 409-452. Potter, T.L., 1996. Analysis of petroleum-contaminated water by GC/FID with direct aqueous iunjection. Ground Water Monitoring & Remediation 16, 157-162.

57 Chapter 2

Rabus, R., Hansen, T., Widdel, F., 2006. Dissimilatory sulfate-and sulfur-reducing prokaryotes. The Prokaryotes 2, 659-768. Schink, B., Zeikus, J., 1980. Microbial methanol formation: a major end product of pectin metabolism. Current Microbiology 4, 387-389. Summons, R.E., Franzmann, P.D., Nichols, P.D., 1998. Carbon isotopic fractionation associated with methylotrophic methanogenesis. Organic Geochemistry 28, 465-475. Williams, J., Holzinger, R., Gros, V., Xu, X., Atlas, E., Wallace, D., 2004. Measurements of organic species in air and seawater from the tropical Atlantic. Geophysical Research Letters 31, L23S06. Yoshioka, H., Sakata, S., Cragg, B.A., Parkes, R.J., Fujii, T., 2009. Microbial methane production rates in gas hydrate-bearing sediments from the eastern Nankai Trough, off central Japan. Geochemical Journal 43, 315-321.

58 Chapter 3

Chapter 3

Distribution and isotopic composition of trimethylamine and dimethylsulfide in anoxic estuarine sediment (Aarhus Bay, Denmark)

Guang-Chao Zhuang1, 2,*, Yu-Shih Lin1,3, Marshall W. Bowles1, Verena B. Heuer1, Mark A. Lever4,#, Marcus Elvert1, Kai-Uwe Hinrichs1

In preparation for Biogeoscience.

1. MARUM Center for Marine Environmental Sciences, University of Bremen, 28359 Bremen, Germany 2. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, College of Chemistry and Chemical Engineering, Ocean University of China, 266100 Qingdao, China 3. Department of Oceanography, National Sun Yat-Sen University, No. 70 Lienhai Rd, Kaohsiung 804, Taiwan 4. Center for Geomicrobiology, Department of Bioscience, Aarhus University, Ny Munkegade 116, DK-8000 Aarhus C, Denmark #Current address: Institute of Biogeochemistry and Pollution Dynamics, Department of Environmental Sciences, Eidgenössische Technische Hochschule Zürich *Corresponding author: Guang-Chao Zhuang ([email protected])

59 Chapter 3

3.1 Abstract

Trimethylamine (TMA) and dimethylsulfide (DMS) are low-molecular-weight metabolites and methanogenic substrates in marine sediments. In this study, simultaneous quantification of TMA and DMS or dimethylsulfoniopropionate (DMSP) was accomplished in small volumes of marine pore waters and sediments through an optimized purge and trap technique in combination with gas chromatography coupled to mass spectrometry. Natural abundances and vertical distributions of TMA and DMS(P) (i.e., DMS or DMSP) were characterized in pore waters and sediments from Aarhus Bay, Denmark. Exchangeable TMA (0.3-6.6 μmol kg-1 wet sediment) and total base-extractable TMA (2-18 μmol kg-1) in the solid phase were much more abundant than the dissolved pool in pore waters (< 20 nM), indicating strong adsorption of TMA to sediments. Likewise, total base-hydrolyzable DMSP (DMS(P)t) in sediment was at least three orders of magnitude higher (11-65 μmol kg-1) than the combined pool of DMS and dissolved DMSP (DMS(P)d) in pore waters (1-12 nM). TMA and DMS(P)t accumulated in the surface sediment, and slightly decreased with depth, consistent with their origin from the decomposition of sedimentary phytoplankton debris and subsequent removal due to adsorption or microbial consumption. The stable carbon isotopic composition of TMA and DMS(P)t in sediments was investigated for the first time with headspace analysis. The δ13C values of TMA were around -38‰ relative to VPDB and varied little with depth, suggesting a large fraction of TMA could avoid biodegradation due to the adsorption to sediment. Base-hydrolyzable DMS(P)t was progressively enriched from -23.0‰ to -18.9‰ with increasing depth, indicating the microbial degradation of DMS(P) in marine sediment. Isotope mass balance calculation revealed that a maximum of 9% of DMS(P)t could be utilized by methanogen, and the generated methane could account for 1% to 29% of methane in the sulfate reduction zone. The occurrence of TMA and DMS(P) provided important substrates for methylotrophic methanogenesis, which might contribute to the methane production in the sulfate-reducing sediments.

60 Chapter 3

3.2 Introduction

Methylated amines (e.g., monomethylamine, dimethylamine, trimethylamine) and sulfides (e.g., dimethylsulfide, methanethiol) are organic nitrogen and sulfur compounds that are commonly found in the marine environment (e.g., Charlson et al., 1987; King, 1988). In particular for trimethylamine (TMA) and dimethylsulfide (DMS), they could be formed from a number of potential precursors (e.g., glycine betaine, dimethylsulfoniopropionate (DMSP)). The precursors are largely, but not solely, formed within phytoplankton (i.e., functioning as osmolytes) and can be converted to TMA and DMS while in the water column or upon deposition at the seafloor (Belviso et al., 2006; Keller et al., 1999). After formation or deposition within marine sediments, TMA and DMS constitute highly energetic substrates for methylotrophic methane producing organisms (Kiene et al., 1986; King, 1984b; Oremland and Polcin, 1982; van der Maarel and Hansen, 1997). The connection between phytoplankton productivity, TMA and DMS, and methane biogeochemistry adds impetus to further developing our understanding of these critical intermediary species. In addition to TMA supplied directly to marine sediments by marine algae, animals and bacteria (Budd and Spencer, 1968), TMA may be produced as intermediates during the anaerobic degradation of quaternary amines such as choline and glycine betaine (King, 1988). Similarly, marine sediments receive inputs of phytoplankton derived organic matter containing the natural osmolyte and cryoprotector, DMSP, which is enzymatically degraded to DMS (e.g., Keller et al., 1989; Zhuang et al., 2011). Under anoxic conditions, DMS can also be generated from bacterial reduction of dimethylsulfoxide (Kiene and Capone, 1988), methylation of methanethiol (Lomans et al., 2002), or microbial incorporation of inorganic substrates

(e.g., CO, CO2) and organic methoxylated compounds (Finster et al., 1990; Lin et al., 2010; Moran et al., 2008). TMA and DMS can be important methanogenic substrates in marine sediments (Kiene, 1988; King, 1984a; Mitterer, 2010; Watkins et al., 2012). Unlike the major

61 Chapter 3

methanogenic substrates, hydrogen and acetate, the utilization of methylotrophic substrates by methanogens is not subject to competition with sulfate-reducing bacteria (Kiene et al., 1986; Oremland et al., 1982). Consequently, these non-competitive substrates could potentially drive methanogenesis in sulfate-bearing sediment. Culture independent molecular methods have been used to detect putative methylotrophic methanogens in a number of environments: tidal-flat sediments (Wilms et al., 2006), brine pools (Joye et al., 2009), and mud volcanoes (Lazar et al., 2012). Furthermore, methylotrophic methanogens utilizing TMA have been successfully enriched or isolated from a variety of sedimentary environments (e.g., Lazar et al., 2012; Oremland et al., 1989; Singh et al., 2005; Sowers and Ferry, 1983). Despite of the potential significane for methane production, TMA or DMS are rarely quantitatively measured in marine sediments or pore water due to analytical challenges (Andreae, 1985; Fitzsimons et al., 2001; Lee and Olson, 1984). For example, the requirement of a large sample volume (> several tens mL) and the time-consuming pretreatment with manipulative steps in the published methods such as distillation (Lee and Olson, 1984) or microdiffusion (Kamil Abdul-Rashid et al., 1991; Yang et al., 1993) make the routine application of these methods for pore water TMA analysis difficult. In this study, we quantified TMA and DMS in marine sediment and pore waters in one analytical step with a small volume (~5 mL) using a simple setup of purge and trap coupled to gas chromatography-mass spectrometry (P&T-GC-MS). Our goal was to measure concentrations of ΤΜΑ and DMS(P) (i.e., DMS or DMSP), determine the δ13C isotope composition in the solid phase, and further address their biogeochemistry and potential for methane production. We performed this research at a well studied site within Aarhus Bay, Denmark, for which the intricasies of methane cycling and the general sedimentary biogeochemical cycles are well known (Holmkvist et al., 2011; Jensen et al., 1995; Thamdrup et al., 1994).

3.3 Materials and experimental 3.3.1 Chemicals and standard solutions

62 Chapter 3

Trimethylamine hydrochloride (98%) and DMS (≥ 99.0%) were purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). The stock solution of trimethylamine hydrochloride (20 mM) was prepared in Milli-Q water and stored in a 50 mL vial sealed without a headspace using polytetrafluoroethene (PTFE)/silicone septum (Supelco, Bellefonte, USA). DMS stock solution (5 mM) was prepared in methanol (hypergrade for LC-MS, Merck, Darmstadt, Germany). Both stock solutions were stored at 4 °C in the dark. Standard solutions for the calibration were obtained by diluting the stocks with Milli-Q water. Ethyl methyl sulfide (EMS, 96%, Sigma-Aldrich Chemie GmbH) was used as the internal standard.

3.3.2 Pore water and sediment sampling

Pore water and sediment samples were collected from Aarhus Bay, Denmark, in April 2012. The Aarhus Bay (Fig. 3.1) is a semi-enclosed bay on the Baltic Sea-North Sea transition and has been previously characterized for geochemistry and microbial activities (Jørgensen, 1996; Holmkvist et al., 2011). Two sites, M1 (56°07.0762' N, 10°20.8078' E) and M5 (56°06.1977' N, 10°27.4822' E) were investigated with water depths of 15 m and 27.5 m, respectively (Fig. 3.1). Rumohr Lot cores of up to 70 cm below seafloor (cmbsf) were taken with minimal disturbance to the surface sediment horizons, stored at 4 °C, and processed within 24 hours after retrieval. Pore water samples were extracted from sediment via Rhizon soil moisture samplers (0.1 µm porous polymer, Rhizosphere Research, Wageningen, the Netherlands) (Seeberg-Elverfeldt et al., 2005). Rhizons were rinsed with at least 30 mL of dilute hydrochloric acid (pH 1-2) followed by 30 mL of Milli-Q water before use. The collected pore water was split into aliquots for analyses of sulfate, TMA, and DMS. Sediment samples for TMA, DMS(P) or total organic carbon (TOC) were taken with 10 mL cut-off syringes and introduced into 40 mL vials sealed with screw caps. For methane analysis, 5 mL sediment samples were collected with cut-off plastic syringes and transferred into 25 mL glass serum vials closed with a butyl septum and crimp cap. The serum vials contained 8 mL of 2.5% NaOH to terminate biological activity

63 Chapter 3

in the sediment. All the pore-water and sediment samples were stored at either 4 °C (methane analyses) or –20 °C before analysis.

Fig. 3.1 Sampling sites M1 and M5 in Aarhus Bay, Denmark.

3.3.3 Geochemical analysis

Methane was quantified on a gas chromatograph equipped with a flame ionization detector using the headspace method (SRI 310C GC, SRI Instruments, USA) (Jørgensen and Parkes, 2010). The carbon isotope composition of methane was obtained using a ThermoFinnigan Trace GC Ultra equipped with a 30 m × 0.32 mm ID Carboxen-1006 PLOT column coupled to a ThermoFinnigan Delta Plus XP isotope ratio mass spectrometer (IRMS) via a Finnigan combustion interface-III (Ertefai et al., 2010). Pore water sulfate was measured by an ICS-3000 ion chromatography system (Dionex Corporation, USA) (Glombitza et al., 2014). TOC were determined in a Carlo Erba NA-1500 CNS analyzer after decarbonating the sediment samples with H2SO3 (5-6 w/w%) (Langerhuus et al., 2012).

64 Chapter 3

3.3.4 Purge and trap apparatus and method optimization for quantitative analysis of TMA and DMS(P)

TMA and DMS were preconcentrated by means of a cryogenic purge and trap (P&T) system (Fig. 3.2), in which volatile analytes were stripped from the aqueous phase with a stream of inert gas and subsequently trapped cryogenically prior to analysis. Specifically, a helium carrier stream (40 mL/min), controlled and monitored with a toggle valve (Swagelok, Hamburg, Germany) and a gas flow controller (Valco Instruments Co. Inc., Houston, US), entered the sparging chamber via a pressure gauge (Supelco) and a 1 mL plastic syringe (B. Braun, Melsungen, Germany). The syringe flange was cut, the plunger removed, and the tip fitted with a long hypodermic needle (Gauge 21, 120 mm; B. Braun, Melsungen, Germany). A 50 mL fluorinated ethylene propylene vial (Thermo Scientific Inc., USA) was used as the sparging chamber (Fig. 3.2), as it offered improvements over the salinized glass sparging chamber described by Glob and Sørensen (1987). The sparging chamber was sealed with a screw cap and a PTFE septum, and heated in a water bath (60 °C) during purging. The carrier gas stream exited from the sparging chamber through a second syringe fitted with a short needle (Gauge 21, 40 mm; B. Braun, Melsungen, Germany) and passed through a moisture condenser. The condenser was made of PTFE tubing (6.4 mm O.D.) submerged in a cryo-bath (–25 °C) filled with a mixture of o-xylene and liquid nitrogen, without compromising TMA and DMS, as tests showed that glass tube filled with chemical dryer tended to cause poor reproducibility probably due to adsorption. The dry gas stream then entered the cold trap immersed in liquid nitrogen. The cold trap was made of 1.6 mm O.D. PTFE tubing, which was looped to form coils of 5 cm in diameter and connected to a six-port Sulfinert™ valve (Restek GmbH, Bad Homburg, Germany). After 10-min of purging and trapping, the trap was transferred into a heating sleeve (200 °C), gas flow was directed to the GC by switching the six-port valve, and the vaporized analytes were collected for 1 min prior to GC analysis. The heating sleeve was made of silicon heating tape (MiL Heating Systems

65 Chapter 3

GmbH, Ubstadt-Weiher, Germany), which was wrapped to form a bag, covered with aluminum foil, and connected to an electronic temperature regulator through a sensor (MiL Heating Systems GmbH). All tubing, connectors, chamber, and traps were made of PTFE or perfluoroalkoxy (PFA) (Swagelok, Hamburg, Germany) to minimize adsorption during analysis. The whole system was purged for 4 min prior to the addition of samples to avoid carryover. After this step, the cold trap was immerged in liquid nitrogen for 3 min to fully cool down before purging of the sample started.

Fig. 3.2 Purge and trap GC-MS system used for analyzing TMA and DMS(P) in the pore waters and sediment.

The quantitative analysis of TMA and DMS was performed on a Trace GC (Thermo Finnigan, Milan, Italy) coupled to a DSQ II MS (Thermo Finnigan, Texas, USA). An advanced base-deactivated Rtx®-Volatile Amine Column (30 m × 0.32 mm I.D., Restek GmbH, Homburg, Germany) was used for separation of the analytes. The optimized GC conditions were: split/splitless injector with a base deactivated liner (4.0 mm I.D., Restek GmbH, Homburg, Germany); split ratio, 1:10; injection temperature, 250 °C; carrier gas, helium at 2 mL/min; oven temperature, starting at 40 °C for 5 min, ramped to 250 °C at 40 °C/min, and held at 250 °C for 3 min. The MS with an electron impact source was operated in selected ion monitoring (SIM) mode. The major ions were: m/z 58 and 59 for TMA, m/z 47 and 62 for DMS, and m/z 61 for EMS. We selected m/z 58 and 62 for quantification. Other MS parameters

66 Chapter 3

were: ionization voltage, 70 eV; ion source, 200 °C; transfer line temperature, 250 °C. During optimization, a 5 mL aliquot of mixed standards (200 nM TMA and 20 nM DMS), together with 1 mL of 5 M NaOH, 25 μL of 1 μM EMS, and 0.2 mL of 20

mM NH4Cl, was injected through the septum to the sparging chamber before analysis. NaOH is necessary to release TMA from the hydrochloride form, as TMA (with the pKb values of 4.19) exists primarily in the non-volatile protonated form in the marine

environment (pH ~8). NH4Cl helps to improve the recovery of TMA by competing with TMA for adsorption sites on the surface of labware and column (Dalene et al., 1981). Measurements of standards showed that compared to non-spiked solutions,

NH4Cl-amended standards had higher response and better reproducibility of TMA without compromising the response of DMS. Furthermore, we systematically evaluated the influence of purge conditions (purge flow, purge time, purge temperature and ionic strength) with respect to purging efficiency. The significance of differences between different P&T conditions was tested using one-way analysis of variance (one-way ANOVA) with the significance level (α) set at 0.05. Multiple runs with different flows of 40, 60, 80 mL min-1 demonstrated that the impact of purge flow on TMA was insignificant (Fig. 3.3a, ANOVA, P = 0.417), whereas DMS response decreased markedly with the enhanced flow (Fig. 3.3a, ANOVA, P < 0.001). Prolonged purging did not improve the TMA signal after 10 min and caused a lower response of DMS (Fig. 3.3b). Rising temperature from ambient to 80 °C improved the detection of TMA (Fig. 3.3c, ANOVA, P < 0.001) but resulted in a considerable loss

of DMS (Fig. 3.3c, ANOVA, P < 0.001). Addition of Na2SO4 did not affect the purging efficiency of TMA (Fig. 3.3d, ANOVA, P = 0.163) but compromised the analysis of DMS (Fig. 3.3d, ANOVA, P < 0.001). Collectively, for concurrent quantification of TMA and DMS, the ultimate purge condition was set to be 10 min with a flow rate of 40 mL/min at 60 °C without salt addition. Under the optimized purge and trap condition, linear calibration and good repeatability were achieved for both compounds (Table 3.1). The method detection limit (MDL) was 20 nM for TMA and 0.1 nM for DMS, respectively, with a 5 mL sample volume. Taking the purge and GC running time into account, the total process

67 Chapter 3 time for each sample is 25 min. The short sample processing time, the relatively low requirement of sample volume, and the capability to analyze two distinct methylated substrates simultaneously make our method a suitable assay for reconnaissance studies of TMA and DMS in sedimentary environments.

Fig. 3.3 The effects of (a) purge flow, (b) purge time, (c) purge temperature, and (d) salt addition on the response of TMA and DMS measured with P&T-GC-MS. In each panel, 100% recovery corresponds to the highest signal observed. Error bars represent the standard deviation of triplicate measurements, and optimal conditions are denoted with * in each panel.

Table 3.1 Linearity, repeatability and detection limit for TMA and DMS analysis with P&T-GC-MS. Compound Selected Detection Repeatability Linearity ion masses limit 2 (m/z) (nM) Concentration R range (nM) TMA 58, 59 20 6% 7% 30-600 nM 0.990 (50 nM) (200 nM) DMS 47, 62 0.1 5% 7% 0.5-20 nM 0.998 (2 nM) (20 nM)

68 Chapter 3

3.3.5 Definition and measurements of different pools

For pore water analysis, a 5 mL aliquot of aqueous sample, with 1 mL of 5 M

NaOH, 25 μL of 1 μM EMS, and 0.2 mL of 20 mM NH4Cl, was introduced to the sparging chamber and treated with the same procedure as the standards. However, due to the presence of DMSP in the pore water, NaOH could act as a reagent that cleaves DMSP and generates DMS with 1:1 stoichiometry (Dacey and Blough, 1987). Thus, the term of DMS(P)d was used to refer to the combined pool of dissolved DMS and DMSP determined by our method in the pore water. The mixture was stored at 4 °C in the dark overnight to allow for complete deprotonation of TMA and conversion of DMSP to DMS. For analysis of sediment, we mixed 0.5 g sediment with 4.5 mL of artificial

seawater, 1 mL of 5 M NaOH, 20 μL of 50 μM EMS, and 0.2 mL of 20 mM NH4Cl. Tests performed with standards suggested that precursors of TMA were unlikely to interfere with the measurements of TMA, as the addition of NaOH only resulted in the conversion of 0.1% of glycine betaine to TMA and no TMA was formed from trimethylamine oxide or choline. Thus the measured TMA was defined as the total base-extractable pool of TMA in the sediment, which might include dissolved, exchangeable or bound TMA in the sediment. Due to alkaline cleavage, total base-hydrolyzable DMSP in the sediment, including adsorbed DMS, DMS(P)d, and particulate DMSP, were collectively quantified as one pool, designated with the term DMS(P)t. Likewise, the slurries were stored at 4 °C in the dark overnight prior to analysis. Furthermore, a fraction of amine compounds could reversibly adsorb to particle exchange sites (Lee and Olson, 1984; Wang and Lee, 1990). To quantify the exchangeable pool of TMA in the sediments, ~4 g wet sediments were extracted with 20 mL 1 M LiCl for 2 hours with continuous agitation. After centrifugation, the supernatant was treated with the same procedure for TMA analysis in the pore water.

3.3.6 Headspace method for carbon isotope analysis of TMA and DMSP in sediment

69 Chapter 3

We used a headspace method to determine the stable carbon isotopic composition of base-extracted TMA and DMS in the sediment. Around 6 g sediments were placed in a 10-mL glass vial, amended with 2 mL of a 5 M NaOH, and sealed immediately with butyl stoppers. The samples were homogenized by vigorous shaking, and stored upside down at 4 °C in the dark overnight. Prior to analysis, sediment samples were heated at 60 °C for 10 min, and 800 μL of gaseous samples from the headspace were taken with gas-tight syringe and injected manually into the GC-IRMS system. Stable carbon isotope of analysis was performed with a ThermoFinnigan Trace GC Ultra coupled to a ThermoFinnigan Delta Plus XP isotope ratio mass spectrometer via a Finnigan combustion interface-III. The δ13C values are expressed relative to that for Vienna-Pee Dee Belemnite (VPDB), which are defined

13 13 12 by the equation δ C (‰) = (Rsample/Rstandard – 1) × 1000, with R = C/ C and Rstandard = 0.0112372 ± 2.9 × 10−6. The column and temperature gradient for separation were adopted from the purge and trap method for quantitative analysis. To improve the detection limit, the split ratio was set to be 1:2 with a column flow of 5 mL min-1. To assess the precision and accuracy of the carbon isotopic analysis, we analyzed different concentrations of TMA standards and compared the δ13C values obtained by GC-IRMS to trimethylamine hydrochloride determined with a Flash 2000 Organic Elemental Analyzer coupled via a Conflo IV interface (Thermo Finnigan MAT GmbH) to a Delta V Plus isotope ratio mass spectrometer (EA-IRMS). The reference value for DMS was obtained by injecting 200 µL of pure standard into water-free headspace vials and after heating to 60 °C (23 °C above the boiling point of DMS) for 10 min, the δ13C value of the gaseous DMS standard in the headspace in the absence of an aqueous phase was compared with aqueous standards to evaluate the isotopic accuracy of DMS at different concentrations (Lin et al., 2010). The average δ13C values of TMA at concentrations from 10 µM to 1000 µM was -43.5‰ ± 0.5‰ (n = 21), in agreement with the results obtained from EA-IRMS measurements (-43.4‰ ± 0.1‰, n = 10) (Fig. 3.4). By comparison, δ13C values of DMS standards between 10 µM and 200 µM were generally consistent with the gaseous standards (-43.4‰ ± 0.2‰, n = 4), while a positive shift of > 1.6‰ was observed at 5 µM, 500 µM and

70 Chapter 3

1000 µM. As all TMA and DMS(P) concentrations in our environmental samples were within the optimal range defined by the standard measurements, the headspace method is adequate for precise carbon isotopic analysis of TMA and DMS(P) in marine sediment.

Fig. 3.4 Stable carbon isotopic analysis of TMA (a) and DMS (b) at different concentrations with 3 mL standards. Filled square represent the δ13C values of TMA and DMS, open diamond designate the corresponding peak area, and dashed line are the reference values from pure standards. Gray areas indicate the range of peak area from environmental samples. Error bars represent the standard deviation of triplicate measurements.

3.4 Results 3.4.1 Geochemistry

At Site M1, low micromolar levels of methane were detected in the presence of sulfate (Fig. 3.5-A). Methane concentrations progressively increased to 14 μM at 46 cmbsf, while the concentrations of sulfate decreased from 23 mM at 1 cmbsf to 16 mM at 54 cmbsf. The carbon isotopic values of methane became increasingly more negative with depth and reached -60‰ at 50 cmbsf (Fig. 3.5-B). Furthermore, the δ13C values of TOC increased from -21.3‰ to -18.1‰ at 18 cmbsf, and varied little at deeper depths (Fig. 3.5-B).

71 Chapter 3

In contrast to M1, sulfate were reduced already in the upper 50 cmbsf (Fig. 3.6-A) at Site M5, and sulfate concentration decreased to 0.2 mM at 54 cmbsf. Accordingly, a clear sulfate methane transition zone (SMTZ) occurred at ~45 cmbsf, below which methane accumulated. The δ13C values of methane were most positive (up to -57‰) in the upper 10 cmbsf, decreased and stayed constant at ~-63‰ between 10 and 38 cmbsf at M5 (Fig. 3.6-B). The trend of 13C depletion in methane continued in the SMZT, where the δ13C values shifted from -65‰ to -83‰, as commonly observed in this zone (cf. Yoshinaga et al., 2014). In the underlying methanogenic zone, the δ13C values were highly negative (~-81‰), in accordance with the typical range for hydrogenotrophic methanogenesis (Whiticar, 1999). Like the M1 profile, the δ13C values of TOC slightly increased in the upper 20 cmbsf (-22.9‰ to -18.7‰) and leveled off until 60 cmbsf.

3.4.2 Distribution of TMA and DMS(P) in the Aarhus Bay sediment

With the optimized method, the abundance and distribution of TMA, DMS, and DMSP were investigated in the pore waters and sediments. At Site M1, dissolved TMA in the pore waters was below the MDL of 20 nM at most depths (data not shown), whereas the concentration of TMA in solid phase sediment was relatively high. Exchangeable TMA accumulated in the upper 5 cmbsf (> 5 μmol kg-1 wet sediment), decreased gradually to 0.5 μmol kg-1 at 17 cmbsf and became constant with respect to depth until 50 cmbsf (Fig. 3.5-C). The total base-extractable TMA in the sediment was more abundant (2-15 μmol kg-1) and had a downcore distribution pattern similar to the exchangeable pool (Fig. 3.5-D). The δ13C values of base-extractable TMA varied slightly between 15 and 26 cmbsf, with an average value of -37.9 ± 0.9‰. The pool sizes of DMS(P)d in the pore waters were small (~3 nM), and the concentrations declined from 12 nM at 1 cmbsf to ~3 nM at 7 cmbsf and varied slightly in deeper depth (Fig. 3.5-E). The total pool of hydrolyzable DMS(P)t (17-65 μmol kg-1) was at least three orders of magnitude higher than DMS(P)d (Fig. 3.5-F). DMS(P) concentrations in the bottom water (17 nM, Fig. 3.5-F) were lower

72 Chapter 3 than in the surface sediment (65 μmol kg-1), and DMS(P)t in the sediment generally decreased with depth. Furthermore, the δ13C values of evolved DMS from base hydrolyzed sediment ranged from -21.9‰ to -18.6‰, and became slightly positive with depth.

Fig. 3.5 Depth profiles of methane and sulfate (A), δ13C of methane and TOC (B), exchangeable TMA (C), base-extractable TMA and δ13C of TMA (D), dissolved DMS(P)d (E), base-hydrolyzable DMS(P)t and δ13C of DMS(P)t (F) in pore waters and sediments at Site M1, Aarhus Bay. Note that exchangeable TMA, base-extractable TMA and base-hydrolyzable DMS(P)t was expressed as μmol per kg wet sediment. The concentrations of base-hydrolyzable DMS(P)t in the bottom water was also shown at 0 cmbsf (F), and the unit was nM. Error bars for δ13C of TMA and DMS(P) represent the standard deviations from the duplicate measurements.

At Site M5, TMA was also not detectable in the pore waters (data not shown). Exchangeable TMA was highest at 1 cmbsf (6.5 μmol kg-1) and decreased from 1.5 μmol kg-1 to 0.3 μmol kg-1 below 5 cmbsf (Fig. 3.6-C). The base-extractable pool of TMA in the sediment was particularly large in the upper 5 cmbsf (>10 μmol kg-1) and

73 Chapter 3 decreased with depth to a value of ~3 μmol kg-1 below 10 cmbsf (Fig. 3.6-D). Similar to M1, the δ13C values of TMA were between -36.4‰ and -39.2‰, and varied slightly with depth. For DMS(P)d, higher concentrations were observed between 12 and 20 cmbsf (Fig. 3.6-E). The maximum DMS(P)t (44 μmol kg-1) was detected at a shallow subsurface depth of 3 cmbsf (Fig. 3.6-F), below which DMS(P)t decreased with increasing depth. Furthermore, the δ13C values of evolved DMS increased with depth, from -23.4‰ at 1 cmbsf to -19.1‰ at 62 cmbsf.

Fig. 3.6 Depth profiles of methane and sulfate (A), δ13C of methane and TOC (B), exchangeable TMA (C), base-extractable TMA and δ13C of TMA (D), dissolved DMS(P)d (E), base-hydrolyzable DMS(P)t and δ13C of DMS(P)t (F) in pore waters and sediments at Site M5, Aarhus Bay. Note that exchangeable TMA, base-extractable TMA and base-hydrolyzable DMS(P)t was expressed as μmol per kg wet sediment. The concentrations of base-hydrolyzable DMS(P)t in the bottom water was also shown at 0 cmbsf (F), and the unit was nM. Error bars for δ13C of TMA and DMS(P) represent the standard deviations from the duplicate measurements.

74 Chapter 3

3.5 Discussion 3.5.1 Source and sink of TMA and DMS in marine sediment

The concentrations of TMA in Aarhus Bay were comparable to previous observations from offshore sediments (Lee and Olson, 1984), but were significantly lower than those in salt marsh or intertidal sediments (Table 3.2). TMA could be derived from either direct release of marine biota (Budd and Spencer, 1968), or through the degradation of osmolyte precursors (King, 1984a). The presence of marsh grass could affect levels through direct excretion or input of organic detritus (Wang and Lee, 1990), and lead to elevated concentrations of dissolved TMA (up to a few micromolar) in salt marsh sediments (Fitzsimons et al., 2001; Wang and Lee, 1994). Benthic fauna was also proposed to influence the spatiotemporal variation of TMA in intertidal sediment (Sørensen and Glob, 1987). In our study sites, the permanently submerged estuarine sediment distinguished itself from salt marsh and intertidal sediment in that pore waters were deficient in TMA (<20 nM). With water depths of 15-30 m, the sources of TMA are already limited in Aarhus Bay sediment. At Sites M1 and M5, both exchangeable TMA and total base-extractable TMA in the solid phase were much larger than the dissolved TMA pool. This distribution, also described in previous studies (Fitzsimons et al., 2001; Fitzsimons et al., 1997; Wang and Lee, 1993), is likely caused by the strong adsorption of TMA on particles. With the reported average adsorption coefficients between exchangeable TMA and dissolved TMA (K = 250) (Wang and Lee, 1994), we can calculate the theoretical concentrations of dissolved TMA in Aarhus Bay. The calculated TMA concentrations in the pore waters at both sites ranged from 1 to 24 nM, which is consistent with TMA being below our detection limit (<20 nM). Laboratory adsorption studies demonstrated that the adsorption behavior of TMA might be influenced by pore water salinity, types of clay minerals, and organic matter content of the sediment (Wang and Lee, 1990). Specifically, adsorption coefficients slightly decreased with reduced organic carbon content and salinity. The lower salinity in the estuarine sediment (< 30‰) relative to the salt marsh, could elevate the adsorption of TMA due to the

75 Chapter 3

higher availability of ion-exchange sites, and further diminish the pool size of TMA in the pore water.

Table 3.2 TMA and DMS(P) concentrations determined with different methods in marine sediments. Dissolved Exchangeable Base-hydrolyzable Sediment settings aRef. and (μM) (μmol kg-1) (μmol kg-1) methods TMA 2.2-2.4 - - Intertidal Sediments, Lowes 1 Cove, Maine, USA 0.4-29 22 - Buzzards Bay, Massachusetts, 2 US 0-0.14 0.2-2 - Eastern Tropical North Pacific 2 Ocean, continental margin of Mexico 0-10 0.1-15 - Norsminde Fjord, Denmark 3 0-~5.1 - 5-238 Westend Saltpond, St. Croix, US 4 0.01-0.5 1-70 - Flax Pond salt marsh, 5 Long Island, New York, US 0-50 - 10-980 Salt marsh, Oglet Bay, UK 6 <4.7 - <4600 Salt marsh, East Anglian 7 Estuary, UK 0.03-2.23 - 4370-10000 mudflats, the Thames Estuary, 8 UK <0.02 0.3-6.6 2.0-15.7 Aarhus Bay, Denmark 9 DMS(P) ~100 intertidal sediment, 10 Westerschelde Estuary, Netherlands 55.4±12.2 Northern Norwegian fjords 11 8.7±4.1 Inshore, Southern North Sea 11 1-2 nM 15-65 Aarhus Bay, Denmark 12 a References and methods: 1. King, 1983; headspace GC-FID. 2. Lee and Olson, 1984; distillation GC with chemiluminescent nitrogen detector. 3. Sørensen and Glob, 1987; purge and trap GC-FID. 4. King, 1988; headspace GC-FID. 5. Wang and Lee, 1994; static diffusion GC-NPD. 6. Fitzsimons et al., 1997; micro-diffusion GC-NPD. 7. Fitzsimons et al., 2001; micro-diffusion GC-NPD. 8. Fitzsimons et al., 2006; micro-diffusion GC-NPD. 9, 12. this study; purge and trap GC-MS. 10. van Bergeijk, et al. 2002; headspace GC-FID. 11. Belviso, et al., 2006; purge and trap GC-FPD.

The total base-extractable TMA pool was 2-10 times greater than the exchangeable TMA pool, and significant correlations were observed between the exchangeable TMA and the total base-extractable TMA at M1 (R2 = 0.85) and M5 (R2

76 Chapter 3

= 0.81). These data collectively suggested that the addition of base liberated both the exchangeable and bound TMA, similar to those approaches of extracting fixed amines in the sediments with HCl or HF (Lee and Olson, 1984; Wang and Lee, 1994). Furthermore, TMA in the solid phase had the highest concentration in near surface sediment and generally decreased with depth, implying that it could be produced from the decomposition of sedimentary phytoplankton debris but subsequently removed due to physical adsorption or microbial consumption. The adsorbed fraction was partly reversible under natural conditions (Fitzsimons et al., 2006), as noted by the decrease in concentration with depth, and the desorbed exchangeable TMA might have fueled microbial processes. Previous experiments examining the decomposition of TMA also demonstrated that amended 14C-labeled TMA disappeared rapidly due to adsorption or consumption, while most of the adsorbed amines would eventually be remineralized (Wang and Lee, 1995). In general, enzymatic cleavage of DMSP has been considered as the major source of DMS in marine environments (Keller et al., 1989). In diatom-dominated intertidal sediments, instantaneous increases of pore water DMS(P) up to 1 μM have been reported as a response of DMSP-containing microorganisms to osmotic shocks (Van Bergeijk et al., 2002). Analogous to TMA, the concentrations of DMS(P)d (1-12 nM) and DMS(P)t (15-65 μmol kg-1) in the estuarine sediment of Aarhus Bay were lower than those in the intertidal sediment (Table 3.2), but were comparable to those reported for the fjord sediment (Belviso et al., 2006), suggesting a reduced flux of DMSP into the sediment with increasing water depth. However, the two sites showed different downcore distribution for DMS(P)d. At M1, DMS(P)d concentration peaked at the sediment-water interface, decreased sharply from 0 to 5 cmbsf, and leveled at 2-5 nM at greater depths except for the sample at 50 cmbsf. Such a pattern was in agreement with phytodetritus being the main source of DMS(P) in the pore waters. In contrast, at M5, the level of DMS(P)d concentrations reached a maximum between 12 and 20 cmbsf, while the distribution of DMS(P)t showed a general decreasing trend with depth, similar to that of M1. The decoupling between the dissolved and total pools implies that other sources than sedimentary DMSP were contributing to the

77 Chapter 3

dissolved DMS pool. Site M5 has a shallow SMTZ, suggesting intensive sulfate reduction and presumably accumulation of hydrogen sulfide in the upper 50 cmbsf of the sediment. Involvement of hydrogen sulfide in the production of DMS and methanethiol from organic and inorganic carbon has been demonstrated before (Lin et al., 2010; Lomans et al., 2002) and may provide an explanation for the subsurface DMS(P)d maximum, although the exact mechanism requires further studies. Furthermore, DMS(P)t concentrations in the bottom water (<20 nM) were much lower than that in the surface sediment (Fig. 3.5F, 3.6F), indicating the dominance of particulate DMSP in the total base-hydrolyzable pool. DMSP accumulated in the surface sediment and decreased slightly with depth at both sites, suggesting the microbial degradation of DMSP under anoxic conditions. In marine sediments, DMSP could be degraded either through enzymatic cleavage to acrylate and DMS, or through successive demethylations of the sulfur atom to 3-methiolpropionate and 3-mercaptopropionate (Kiene and Taylor, 1988; Taylor and Gilchrist, 1991). These labile degradation products of DMSP, including DMS, could be further mineralized by microorganisms, e.g., denitrifying bacteria, sulfate reducers, and methanogens (Kiene and Capone, 1988; Visscher and Taylor, 1993).

3.5.2 Stable carbon isotopic signatures of TMA and DMS(P) in sediment

Although the low concentrations of TMA and DMS in the pore water did not allow for isotopic analysis, it was possible to examine the isotopic signature of the TMA and DMS from base extracted sediment. Interestingly, the δ13C values of TMA were strongly offset from the isotopic composition of TOC with TMA being depleted in 13C relative to TOC by 20‰. The precursors of TMA such as glycine betaine and choline are derived from marine algae, thus their isotopic values are primarily dependent on the production of biomass and the isotopic fractionation associated with photosynthetic processes. The C1 units for quaternary methylammonium groups of glycine betaine and choline mainly originate from the non-carboxyl carbon of serine and glycine. A previous study has showed that non-carboxyl carbon of serine is 6‰ to

78 Chapter 3

9‰ 13C enriched relative to TOC in a range of photosynthetic organisms (Abelson and Hoering, 1961). Likewise, non-carboxyl carbon of glycine varied from 6‰ depletion to 3‰ enrichment dependent of the organisms (Abelson and Hoering, 1961). Thus, the carbon in glycine betaine or choline should be generally 13C-enriched or similar with respect to the isotopic composition of the photosynthetic TOC. As a consequence, the relatively large discrimination between TMA and TOC might be caused by the isotopic fractionation during the degradation reaction of TMA from glycine betaine or choline. Once produced, TMA can be consumed by microorganisms, and these processes are also associated with isotopic effects. For example, the utilization of TMA by methanogens could cause a large fractionation between TMA and methane (37‰-71‰) (Londry et al., 2008; Summons et al., 1998). Accordingly, the residual TMA pool should be 13C-enriched due to the preferential utilization of lighter 12C isotopes. Although the δ13C values of TMA did not show significant variations with depth at M1 and M5, this was not contradictive to the assumption on microbial consumption of TMA. The measured δ13C values were from the base extracted pool of TMA, which included both bioavailable and bound TMA in sediment. Instead, the homogeneous isotopes of TMA could indicate that a large fraction of TMA was adsorbed to sediment and less available for biodegradation, consistent with our observation of the difference of TMA pool sizes in the pore water and sediments. Therefore, the δ13C values of TMA in the base extracted pool might reflect an isotopic balance between the adsorption, desorption and biological consumption of TMA in the sediments, as these processes collectively controlled the TMA pools in different phases. In contrast to TMA, the isotopic values of DMS(P) at both sites were comparable to those of TOC. Similar to glycine betaine and choline, DMSP is also produced in marine algae, and it is largely synthesized from methionine (Gage et al., 1997). Thus, it could be also inferred that the isotopic signature of DMSP depends on photosynthetic processes and the carbon in DMSP should be similarly 13C-depleted relative to TOC. Unlike TMA, the evolved DMS for isotopic analysis is from DMSP

79 Chapter 3

through base hydrolysis, thus the measured isotope values should be indicative of the carbon signature of DMSP. The isotopic difference between measured TMA and DMS(P), also suggested the potential fractionation occurred during the microbial production of TMA from precursors, which might lead to the relatively 13C depleted isotope of TMA. At both sites, the δ13C values of DMS(P) became increasingly more positive with depth, indicating the microbial degradation of DMS(P), during which 13C-depleted DMS(P) was preferentially utilized by microbes. A strong negative correlation between the concentrations and isotopes of DMS(P)t was observed at M5 (R2 = 0.49), further attesting to the enrichment of 13C in the residual pool with the degradation of DMSP. Among those microbial processes that involves DMS or DMSP degradation, the methane production from DMS are associated with large fractionation (Whiticar, 1999), which might result in the enrichment of 13C in DMS(P).

3.5.3 Potential implication for methanogenesis from TMA and DMS(P)

At M1 and M5, small amount of methane were detected in the sulfate-reducing sediments. As hydrogenotrophic methanogenesis and acetoclastic methanogenesis are thermodynamically inhibited by sulfate reduction (Oremland et al., 1982), the occurrence of methane in the presence of sulfate has been proposed to result from the in situ production from methylated compounds, although this process is often obscured by both the diffusive flux from underlying horizons and the concurrent oxidation of methane (Valentine, 2011, references therein). At M5, the relatively constant δ13C values of methane between 10 and 38 cmbsf (~-63‰) might suggest a balance between methane production and oxidation, as anaerobic oxidation of methane (AOM) could drive the progressive 13C-enrichment of methane in the sulfate reduction zone (Holler et al., 2009; Whiticar, 1999). As non-competitive substrates, both TMA and DMS(P) were relatively abundant in the sulfate-reducing sediments at M1 and M5. The fact that their concentrations decreased with depth, also implied that they could be potentially degraded and

80 Chapter 3

metabolized by methylotrophic methanogens to produce methane. In particular for DMS(P), the progressive enrichment of 13C strongly suggested the microbial degradation of DMS(P), which also provided an avenue to estimate the contribution of DMS(P) for methane production. According to the carbon isotope mass balance, the relationship between isotope ratios and the fraction of the remaining substrates could be addressed by the following equation (Whiticar, 1999):

13 13 δ Csubstrate, t = δ Csubstrate, i + ɛln f (3.1)

13 13 where δ Csubstrate, t is the isotopic value of substrate (i.e., DMS) at time t, δ Csubstrate, i is the initial isotope value of substrate, f is the fraction of initial substrate remaining at time t, and ɛ is the isotope fractionation factor. To estimate the maximum potential of DMS(P) for methane production, we assumed that the produced DMS during DMSP degradation were exclusively utilized by methanogens, and the isotopic values of DMS are identical to those of base-hydrolyzable DMS(P). As previous studies have reported the fractionation factor of 44-54‰ between DMS and methane during methanogenesis with cultures, we adapted an average value of 49‰ for calculations (Oremland et al., 1989; Whiticar, 1999). Taking the isotopic composition of DMS(P) at 1 cmbsf as the initial value, we could calculate that a maximum of 9% of DMS(P) could be utilized by methanogens to produce methane in the sulfate reduction zone. Accordingly, the generated methane from DMS(P) could be calculated with measured DMS(P) concentrations based on the following disproportionation reaction (Eq. 3.2, Garcia et al., 2000).

2 (CH3)2S + 2 H2O → 3 CH4 + CO2 + 2H2S (3.2) By comparison, the generated methane from the decomposition of DMS(P) could account for 1% to 29% of methane in the sulfate reduction zone, although the present methane was a result of balance between production and consumption. Although DMS(P) did not entirely contribute to methane production from our calculations, the importance of methylotrophic methanogenesis should not be neglected in the sulfate reduction zone, if all methylated substrates, e.g., methanol, monomethylamine, dimethylamine were taken into account. Furthermore, molecular analysis demonstrated that the archaeal genus of Methanococcoides, which could

81 Chapter 3

metabolize methylated compounds as sole substrates, dominated the methanogen populations at 20 cmbsf at M1 (Chen, unpublished data). Together these indications suggest methylated compounds could be important substrates for methylotrophic methanogens, and methylotrophic methanogenesis potentially occurred in the sulfate reduction zone. To fully understand the relative contribution of methylated compounds for total methane production, further constraints on methanogenic activity should be pursued, e.g., rate measurements combining radiotracers with the in situ concentrations.

3.6 Conclusion

The simultaneous measurements of TMA and DMS(P) from marine sediments were accomplished with gas chromatographic approaches in combination with a purge and trap system for quantification or a headspace method for carbon isotope analysis. In sediment samples from the Aarhus Bay, TMA and DMS(P) were several orders of magnitude more abundant in the sediment than in the pore water, and the solid-phase content decreased with depth, indicating the microbial degradation of TMA and DMS(P) under anoxic conditions. The isotopes of TMA were strongly 13C-depleted (-38‰), indicating potential isotopic fractionation during the production of TMA from the precursors. The relatively constant TMA δ13C values with depth might suggest a large fraction of TMA was not available for mineralization due to the adsorption to sediment. In contrast, the progressive enrichment of 13C in DMS(P) with depth indicated the microbial degradation of DMS(P). Based on the isotope mass balance calculations, about 9% of DMS(P) could be utilized by methanogen to produce methane. Given the abundance of methylated compounds in marine sediment, their contributions for methane production and the importance of methylotrophic methanogenesis deserve further quantitative evaluations.

3.7 Acknowledgements

82 Chapter 3

The study was supported by the DFG through the Research Center/Excellence Cluster MARUM-Center for Marine Environmental Sciences. G. C. Zhuang was jointly sponsored by the Chinese Scholarship Council (CSC) and the Bremen International Graduate School for Marine Sciences (GLOMAR). We are grateful to B. B. Jørgensen for valueable comments to improve the manuscript. Special thanks go to H. Røy, as well as crews from Aarhus University for their kind assistance in sampling. We thank E. P. Reeves, X. Prieto-Mollar and J. Wendt for their support on method development, and M. Yoshinaga for discussion on isotope mass balance calculations.

3.8 References

Abelson, P.H., Hoering, T., 1961. Carbon isotope fractionation in formation of amino acids by photosynthetic organisms. Proceedings of the National Academy of Sciences of the United States of America 47, 623-632. Andreae, M.O., 1985. Dimethylsulfide in the water column and the sediment porewaters of the Peru upwelling area. Limnology and Oceanography 30, 1208-1218. Belviso, S., Thouzeau, G., Schmidt, S., Reigstad, M., Wassmann, P., Arashkevich, E., Stefels, J., 2006. Significance of vertical flux as a sink for surface water DMSP and as a source for the sediment surface in coastal zones of northern Europe. Estuarine, Coastal and Shelf Science 68, 473-488. Budd, J., Spencer, C., 1968. The utilisation of alkylated amines by marine bacteria. Marine Biology 2, 92-101. Charlson, R., Lovelock, J., Andreae, M., Warren, S., 1987. Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326, 655-661. Dacey, J.W., Blough, N.V., 1987. Hydroxide decomposition of dimethylsulfoniopropionate to form dimethylsulfide. Geophysical Research Letters 14, 1246-1249. Dalene, M., Mathiasson, L., Jönsson, J.Å., 1981. Trace analysisi of free amines by gas-liquid chromatography. Journal of Chromatography A 207, 37-46. Ertefai, T.F., Heuer, V.B., Prieto-Mollar, X., Vogt, C., Sylva, S.P., Seewald, J., Hinrichs, K.-U., 2010. The biogeochemistry of sorbed methane in marine sediments. Geochimica et Cosmochimica Acta 74, 6033-6048. Finster, K., King, G.M., Bak, F., 1990. Formation of methylmercaptan and dimethylsulfide from methoxylated aromatic compounds in anoxic marine and fresh water sediments. FEMS

83 Chapter 3

Microbiology Letters 74, 295-301. Fitzsimons, M., Kahni-Danon, B., Dawitt, M., 2001. Distributions and adsorption of the methylamines in the inter-tidal sediments of an East Anglian estuary. Environmental and Experimental Botany 46, 225-236. Fitzsimons, M.F., Jemmett, A.W., Wolff, G.A., 1997. A preliminary study of the geochemistry of methylamines in a salt marsh. Organic Geochemistry 27, 15-24. Fitzsimons, M.F., Millward, G.E., Revitt, D.M., Dawit, M.D., 2006. Desorption kinetics of ammonium and methylamines from estuarine sediments: consequences for the cycling of nitrogen. Marine Chemistry 101, 12-26. Gage, D.A., Rhodes, D., Nolte, K.D., Hicks, W.A., Leustek, T., Cooper, A., Hanson, A.D., 1997. A new route for synthesis of dimethylsulphoniopropionate in marine algae. Nature 387, 891-894. Garcia, J.-L., Patel, B.K., Ollivier, B., 2000. Taxonomic, phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe 6, 205-226. Glob, E., Sørensen, J., 1987. Determination of dissolved and exchangeable trimethylamine pools in sediments. Journal of Microbiological Methods 6, 347-355. Glombitza, C., Pedersen, J., Røy, H., Jørgensen, B.B., 2014. Direct analysis of volatile fatty acids in marine sediment porewater by two-dimensional ion chromatography-mass spectrometry. Limnology and Oceanography: Methods 12, 455-468. Holler, T., Wegener, G., Knittel, K., Boetius, A., Brunner, B., Kuypers, M.M., Widdel, F., 2009. Substantial 13C/12C and D/H fractionation during anaerobic oxidation of methane by marine consortia enriched in vitro. Environmental Microbiology Reports 1, 370-376. Holmkvist, L., Ferdelman, T.G., Jørgensen, B.B., 2011. A cryptic sulfur cycle driven by iron in the methane zone of marine sediment (Aarhus Bay, Denmark). Geochimica et Cosmochimica Acta 75, 3581-3599. Jensen, H.S., Mortensen, P., Andersen, F., Rasmussen, E., Jensen, A., 1995. Phosphorus cycling in a coastal marine sediment, Aarhus Bay, Denmark. Limnology and Oceanography 40, 908-917. Jørgensen, B.B., 1996. Case study-Aarhus bay. Eutrophication in Coastal Marine Ecosystems, 137-154. Jørgensen, B.B., Parkes, R.J., 2010. Role of sulfate reduction and methane production by organic carbon degradation in eutrophic fjord sediments (Limfjorden, Denmark). Limnology and Oceanography 55, 1338-1352. Joye, S.B., Samarkin, V.A., MacDonald, I.R., Hinrichs, K.-U., Elvert, M., Teske, A.P., Lloyd, K.G., Lever, M.A., Montoya, J.P., Meile, C.D., 2009. Metabolic variability in seafloor brines

84 Chapter 3

revealed by carbon and sulphur dynamics. Nature Geoscience 2, 349-354. Kamil Abdul-Rashid, M., Riley, J.P., Fitzsimons, M.F., Wolff, G.A., 1991. Determination of volatile amines in sediment and water samples. Analytica Chimica Acta 252, 223-226. Keller, M., Kiene, R., Matrai, P., Bellows, W., 1999. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. I. Batch cultures. Marine Biology 135, 237-248. Keller, M.D., Bellows, W.K., Guillard, R.R., 1989. Dimethyl sulfide production in marine phytoplankton. Biogenic Sulfur in the Environment 393, 167-182. Kiene, R., 1988. Dimethyl sulfide metabolism in salt marsh sediments. FEMS Microbiology Letters 53, 71-78. Kiene, R., Taylor, B., 1988. Demethylation of dimethylsulfoniopropionate and production of thiols in anoxic marine sediments. Applied and Environmental Microbiology 54, 2208-2212. Kiene, R.P., Capone, D.G., 1988. Microbial transformations of methylated sulfur compounds in anoxic salt marsh sediments. Microbial Ecology 15, 275-291. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. King, G.M., 1984a. Metabolism of trimethylamine, choline, and glycine betaine by sulfate-reducing and methanogenic bacteria in marine sediments. Applied and Environmental Microbiology 48, 719-725. King, G.M., 1984b. Utilization of hydrogen, acetate, and “noncompetitive”; substrates by methanogenic bacteria in marine sediments. Geomicrobiology Journal 3, 275-306. King, G.M., 1988. Distribution and metabolism of quaternary amines in marine sediments. John Wiley and Sons, Chichester, United Kingdom. Langerhuus, A.T., Røy, H., Lever, M.A., Morono, Y., Inagaki, F., Jørgensen, B.B., Lomstein, B.A., 2012. Endospore abundance and D: L-amino acid modeling of bacterial turnover in holocene marine sediment (Aarhus Bay). Geochimica et Cosmochimica Acta 99, 87-99. Lazar, C.S., John Parkes, R., Cragg, B.A., L'Haridon, S., Toffin, L., 2012. Methanogenic activity and diversity in the centre of the Amsterdam Mud Volcano, Eastern Mediterranean Sea. FEMS Microbiology Ecology 81, 243-254. Lee, C., Olson, B.L., 1984. Dissolved, exchangeable and bound aliphatic amines in marine sediments: initial results. Organic Geochemistry 6, 259-263. Lin, Y., Heuer, V., Ferdelman, T., Hinrichs, K.-U., 2010. Microbial conversion of inorganic carbon to dimethyl sulfide in anoxic lake sediment (Plußsee, Germany). Biogeosciences 7, 2433-2444.

85 Chapter 3

Lomans, B., Van der Drift, C., Pol, A., den Camp, H.O., 2002. Microbial cycling of volatile organic sulfur compounds. Cellular and Molecular Life Sciences CMLS 59, 575-588. Londry, K.L., Dawson, K.G., Grover, H.D., Summons, R.E., Bradley, A.S., 2008. Stable carbon isotope fractionation between substrates and products of Methanosarcina barkeri. Organic Geochemistry 39, 608-621. Mitterer, R.M., 2010. Methanogenesis and sulfate reduction in marine sediments: A new model. Earth and Planetary Science Letters 295, 358-366. Moran, J.J., House, C.H., Vrentas, J.M., Freeman, K.H., 2008. Methyl sulfide production by a novel carbon monoxide metabolism in Methanosarcina acetivorans. Applied and Environmental Microbiology 74, 540-542. Oremland, R.S., Kiene, R.P., Mathrani, I., Whiticar, M.J., Boone, D.R., 1989. Description of an estuarine methylotrophic methanogen which grows on dimethyl sulfide. Applied and Environmental Microbiology 55, 994-1002. Oremland, R.S., Marsh, L.M., Polcin, S., 1982. Methane production and simultaneous sulphate reduction in anoxic, salt marsh sediments. Nature 296, 143-145. Oremland, R.S., Polcin, S., 1982. Methanogenesis and sulfate reduction: competitive and noncompetitive substrates in estuarine sediments. Applied and Environmental Microbiology 44, 1270. Seeberg-Elverfeldt, J., Schlüter, M., Feseker, T., Kölling, M., 2005. Rhizon sampling of pore waters near the sediment/water interface of aquatic systems. Limnology and Oceanography: Methods 3, 361-371. Singh, N., Kendall, M.M., Liu, Y., Boone, D.R., 2005. Isolation and characterization of methylotrophic methanogens from anoxic marine sediments in Skan Bay, Alaska: description of Methanococcoides alaskense sp. nov., and emended description of Methanosarcina baltica. International Journal of Systematic and Evolutionary Microbiology 55, 2531-2538. Sørensen, J., Glob, E., 1987. Influence of benthic fauna on trimethylamine concentrations in coastal marine sediments. Marine Ecology Progress Series. Oldendorf 39, 15-21. Sowers, K.R., Ferry, J.G., 1983. Isolation and characterization of a methylotrophic marine methanogen, Methanococcoides methylutens gen. nov., sp. nov. Applied and Environmental Microbiology 45, 684-690. Summons, R.E., Franzmann, P.D., Nichols, P.D., 1998. Carbon isotopic fractionation associated with methylotrophic methanogenesis. Organic Geochemistry 28, 465-475. Taylor, B.F., Gilchrist, D.C., 1991. New routes for aerobic biodegradation of dimethylsulfoniopropionate. Applied and Environmental Microbiology 57, 3581-3584. Thamdrup, B., Fossing, H., Jørgensen, B.B., 1994. Manganese, iron and sulfur cycling in a coastal

86 Chapter 3

marine sediment, Aarhus Bay, Denmark. Geochimica et Cosmochimica Acta 58, 5115-5129. Valentine, D.L., 2011. Emerging topics in marine methane biogeochemistry. Annual Review of Marine Science 3, 147-171. Van Bergeijk, S., Sch nefeldt, K., Stal, L., Huisman, J., 2002. Production and consumption of dimethylsulfide (DMS) and dimethylsulfoniopropionate (DMSP) in a diatom-dominated intertidal sediment. Marine Ecology Progress Series 231, 37-46. van der Maarel, M.J.E.C., Hansen, T.A., 1997. Dimethylsulfoniopropionate in anoxic intertidal sediments: a precursor of methanogenesis via dimethyl sulfide, methanethiol, and methiolpropionate. Marine Geology 137, 5-12. Visscher, P.T., Taylor, B.F., 1993. Aerobic and anaerobic degradation of a range of alkyl sulfides by a denitrifying marine bacterium. Applied and Environmental Microbiology 59, 4083-4089. Wang, X.-C., Lee, C., 1990. The distribution and adsorption behavior of aliphatic amines in marine and lacustrine sediments. Geochimica et Cosmochimica Acta 54, 2759-2774. Wang, X.-C., Lee, C., 1993. Adsorption and desorption of aliphatic amines, amino acids and acetate by clay minerals and marine sediments. Marine Chemistry 44, 1-23. Wang, X.-C., Lee, C., 1994. Sources and distribution of aliphatic amines in salt marsh sediment. Organic Geochemistry 22, 1005-1021. Wang, X.-C., Lee, C., 1995. Decomposition of aliphatic amines and amino acids in anoxic salt marsh sediment. Geochimica et Cosmochimica Acta 59, 1787-1797. Watkins, A.J., Roussel, E.G., Webster, G., Parkes, R.J., Sass, H., 2012. Choline and N, N-Dimethylethanolamine as direct substrates for methanogens. Applied and Environmental Microbiology 78, 8298-8303. Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161, 291-314. Wilms, R., Sass, H., Köpke, B., Köster, J., Cypionka, H., Engelen, B., 2006. Specific bacterial, archaeal, and eukaryotic communities in tidal-flat sediments along a vertical profile of several meters. Applied and Environmental Microbiology 72, 2756-2764. Yang, X.H., Lee, C., Scranton, M.I., 1993. Determination of nanomolar concentrations of individual dissolved low molecular weight amines and organic acids in seawater. Analytical Chemistry 65, 572-576. Yoshinaga, M.Y., Holler, T., Goldhammer, T., Wegener, G., Pohlman, J.W., Brunner, B., Kuypers, M.M., Hinrichs, K.-U., Elvert, M., 2014. Carbon isotope equilibration during sulphate-limited anaerobic oxidation of methane. Nature Geoscience 7, 190-194. Zhuang, G., Yang, G., Yu, J., Gao, Y., 2011. Production of DMS and DMSP in different

87 Chapter 3 physiological stages and salinity conditions in two marine algae. Chinese Journal of Oceanology and Limnology 29, 369-377.

88 Chapter 4

Chapter 4

Multiple evidences for methylotrophic methanogenesis as the dominant methanogenic pathway in deep-sea hypersaline sediments

Guang-Chao Zhuang1, 2,*,#, Felix J. Elling1,*, Lisa M. Nigro3, Vladimir Samarkin4, Samantha B. Joye4, Andreas Teske3, Kai-Uwe Hinrichs1

In preparation for Environmental Microbiology.

1. Organic Geochemistry Group, MARUM Center for Marine Environmental Sciences, University of Bremen, 28359 Bremen, Germany 2. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, College of Chemistry and Chemical Engineering, Ocean University of China, 266100 Qingdao, China 3. Department of Marine Sciences, University of North Carolina at Chapel Hill, Chapel Hill, 27599 NC, USA 4. Department of Marine Sciences, University of Georgia, 30602 Athens, GA, USA *These authors contributed equally to this work. #Corresponding author: Guang-Chao Zhuang ([email protected])

89 Chapter 4

4.1 Abstract

Among Earth’s most extreme habitats, dark, deep and anoxic brines host unique microbial ecosystems that remain largely unexplored. As the terminal step of organic matter degradation, methanogenesis is a potentially significant but poorly constrained process supporting carbon cycling in deep-sea hypersaline environments. Here we integrated biogeochemical and phylogenetic analyses as well as incubation experiments to unravel the origin of methane in hypersaline sediments of Orca Basin in the northern Gulf of Mexico. Significant concentrations of methane (up to 1.4 mM) coexisted with high concentrations of sulfate (16-43 mM) in two sediment cores retrieved from the southern and northern sections of Orca Basin. Strong 13C-depletion

13 (δ CH4: -76.7 to -89.1‰) as well as high methane to ethane ratios (600-3000) pointed to a biological source of methane. While low concentrations of competitive substrates limited the significance of hydrogenotrophic and acetoclastic methanogenesis, the high abundance of non-competitive methylated substrates and their precursors (e.g., methanol, trimethylamine, dimethylsulfide, dimethylsulfoniopropionate) suggested that methane could be generated through methylotrophic methanogenesis in the hypersaline sediment of Orca Basin. The carbon isotope systematics of methylated substrates and methane supported methylotrophic methanogenesis as the dominant source of methane in Orca Basin sediments. Thermodynamic calculations demonstrated that hydrogenotrophic and acetoclastic methanogenesis were unlikely to occur under in situ conditions, while methylotrophic methanogenesis from methylated substrates was highly favorable. Stable isotope tracer and radiotracer experiments with 13C bicarbonate, acetate and methanol as well as 14C labeled methylamine indicated that methylotrophic methanogenesis was the predominant methanogenic pathway in Orca Basin sediment. Based on 16S rRNA gene sequences, halophilic methylotrophic methanogens of the genus Methanohalophilus dominated the benthic archaeal communities in the northern basin but also occurred in the southern basin. High abundances of methanogen lipid biomarkers such as intact polar and polyunsaturated and

90 Chapter 4 hydroxyarchaeols were detected in sediments from the northern basin but were of lower abundance in sediments from the southern basin. Collectively, the availability of methylated substrates, thermodynamic calculations, experimentally determined methanogenic activity as well as gene and lipid biomarkers strongly suggested the occurrence and predominance of methylotrophic methanogenesis in Orca Basin sediment. Our findings elucidate the significance of methylotrophic methanogenesis for methane production in the presence of sulfate in deep marine hypersaline environments.

4.2 Introduction

Hypersaline environments, commonly defined as those with salinities exceeding 50‰, are widespread on earth including coastal salt marshes and sabkhas, inland salt lakes as well as deep-sea brine pools (Boetius and Joye, 2009; McGenity, 2010). Among the most extreme habitats, dark and anoxic seafloor brine lakes have been discovered in the deep sea of the Gulf of Mexico, the Mediterranean Sea and the Red Sea, which are characterized by extremely high salinity, strong stratification, and steep geochemical gradients (Antunes et al., 2011; Joye et al., 2005). As halophiles are among the candidates for the earliest life forms on earth, studying the microbial ecology of brine ecosystems may yield insights into the origins, and limits of (early) microbial life on earth and potentially other planets (Mancinelli et al., 2004). Despite harsh conditions, a diversity of microorganisms involved in fundamental biogeochemical processes, e.g., cycling of methane and sulfur, potentially support microbial ecosystems in these chemically distinct habitats (Joye et al., 2009). In particular, biological production of methane has been proposed to play a key role in supporting carbon cycling in deep-sea brines and underlying sediments (Charlou et al., 2003; Lazar et al., 2011; van der Wielen et al., 2005). Microbial production of methane by methanogenic Archaea is the terminal step in the degradation of organic matter in anoxic environments. Methanogenic pathways are commonly classified with respect to the type of carbon source, i.e., CO2 reduction

91 Chapter 4

by available hydrogen (hydrogenotrophic methanogenesis), acetate disproportionation (acetoclastic methanogenesis) and methanogenesis from methylated substrates, such as methanol, methylamines, and methyl sulfides (methylotrophic methanogenesis). Thermodynamically, methanogens are out-competed for hydrogen and acetate by sulfate reducers, and methane formation via these ‘competitive’ substrates is generally inhibited until sulfate is depleted. In the presence of sulfate, methanogens may circumvent competition by utilizing ‘non-competitive’ methylated substrates, as these are consumed exclusively or preferentially by methylotrophic methanogens (Oremland et al., 1982b). The stable carbon isotopic signature of methane has been extensively used in the past for identifying formation pathways of methane in the environment (Conrad, 2005; Whiticar, 1999; Whiticar et al., 1986). Presumably, thermogenic generation of methane at high temperatures yields methane with δ13C values >-50‰. In contrast, strong fractionation during biogenic methanogenesis leads to isotopically lighter methane, e.g., acetoclastic methanogenesis produces methane with δ13C values in the range of -70‰ to -50‰ while hydrogenotrophic methanogenesis yields δ13C values of -110‰ to -60‰ (Whiticar, 1999). Although little is known about the isotopic composition of methane derived from methylotrophic methanogenesis in the environment, large carbon isotope fractionation between methanol or trimethylamine and methane in pure cultures suggests that methane produced from these substrates is similarly depleted as that produced from acetoclastic or hydrogenotrophic methanogenesis (Londry et al., 2008; Penger et al., 2012; Rosenfeld and Silverman, 1959; Summons et al., 1998). In hypersaline environments, methylotrophic methanogenesis is thought to be of greater importance because of the high concentrations of sulfate found in these environments and the abundance of potential methylated substrates and their precursors (McGenity, 2010). The methylated substrates could be derived from the decomposition of organic matter, such as methanol from lignin and pectin (Donnelly and Dagley, 1980; Schink and Zeikus, 1980). Methylamines and dimethyl sulfide (DMS) could be derived from the breakdown of osmolytes, e.g., glycine betaine and dimethylsulfoniopropionate (DMSP), which are ubiquitously synthesized by

92 Chapter 4 phytoplankton or other microbes inhabiting hypersaline environments (Oren, 1990; Zhuang et al., 2011). As methanogens inhabiting hypersaline environments need to divert more energy for the biosynthesis or uptake of compatible solutes, methylotrophic methanogenesis might be more favorable than acetoclastic and hydrogenotrophic methanogenesis due to its higher energy yield (McGenity, 2010; Oren, 1999; Roberts, 2005). Although this hypothesis was proposed decades ago and supported by sparse in vitro methanogenic activity measurements and molecular analyses in hypersaline environments (Lazar et al., 2011; Orphan et al., 2008; Smith et al., 2008), to our knowledge, very limited information was available on the in situ abundance of methylated substrates, impeding our understanding of the relative importance of methylotrophic methanogenesis. In addition, geochemical and microbiological investigations of deep-sea brines mostly focused on the brine-seawater interface while benthic microbial communities and associated processes remain largely unconstrained. In this study, we sought to constrain methanogenic pathways by combining geochemical, physiological, and microbiological analyses of hypersaline sediments from Orca Basin, which contains the largest seafloor brine pool in the Gulf of Mexico covering approximately 200 km² (Pilcher and Blumstein, 2007). Natural abundances of major competitive and non-competitive substrates, stable carbon isotope compositions of methane and precursors, methanogenic activity with stable isotope tracers and radiotracers as well as methanogen lipid biomarkers and 16S rRNA genes were comprehensively characterized to decipher the origin of methane in Orca Basin. The relationships between available substrates, methanogenic activity and methanogen populations provided new insights into the relative importance of different methanogenic pathways in deep-marine hypersaline sediments, in particular from non-competitive substrates.

4.3 Materials and methods 4.3.1 Site description and sampling protocol

93 Chapter 4

Orca Basin is a 25 km long and 15 km wide depression on the continental slope of the northern Gulf of Mexico, hosting an anoxic hypersaline brine pool in the lowermost 200 m (Fig. 4.1; Shokes et al., 1977). Sediment samples were collected using a multicorer during expedition AT18-2 of the research vessel Atlantis, in November 2010. Three multicores (60 cm) were recovered from the southern sub-basin (MUC-6, N 26.9080°, W 91.3348°, water depth: 2432 m), the northern sub-basin (MUC-7, N 27.0000°, W 91.2832°, water depth: 2339 m), and a non-brine ‘reference site’ (N 27.1235°, W 90.2878°, water depth: 2068 m) on the Louisiana-Texas continental slope, respectively (Fig. 4.1). Due to overpenetration of the multicore, depth below seafloor is unknown for the Orca Basin cores. Sediment cores were transferred to a 4 °C cold room after retrieval and sectioned at room temperature within 12 hours using a core extruder. Samples for C1-C6 hydrocarbon concentration and isotope analysis were taken by transferring 5 mL sediment with a cut-off syringe into a 30 mL serum vial containing 2 mL of 1 M NaOH, crimp-capped immediately, and stored at room temperature before analysis. For hydrogen analysis, 2 mL of sediment was sampled using a cut-off syringe, transferred to helium purged 10 mL headspace vials and crimp-capped. The samples were stored and equilibrated for 96 h at in situ temperature (5 °C) before analysis. Sediment samples for intact polar lipids, trimethylamine (TMA) and DMSP in the solid phase were stored at -20 °C. Sediment samples for DNA extraction were stored at -80 °C. Pore water was extracted from sediments by centrifugation at 5000 rpm for 20 min. Pore water samples for dissolved inorganic carbon (DIC), volatile fatty acids, alcohols, DMS, TMA were stored at -20 °C. Samples for sulfate analysis were prepared by adding 50 µL of 50%

HCl to 1 mL of pore water, bubbled with N2 for three minutes and stored at 4 °C.

Samples for chloride analysis were preserved by adding 100 µL of HNO3 to 1.5 mL of pore water. The residual sediments from pore water extraction were collected for determination of the content and stable isotopic composition of total organic carbon (TOC). Live sediments for incubation experiments were taken from a parallel multicore sectioned into three 20 cm intervals, transferred to glass bottles, flushed

with N2 and stored at 4 °C.

94 Chapter 4

Fig. 4.1 Location of Orca Basin and the reference site in the northern Gulf of Mexico (a) and bathymetric map of Orca Basin (b, 20 m horizontal resolution, based on multibeam data acquired during RV Atlantis expedition 18-2) showing multicorer sampling locations in the southern and northern sub-basins.

4.3.2 Biogeochemical analyses

Salinity of pore water was analyzed with a handheld refractometer (Atago, Co. LTD., Japan) after dilution. Samples for determination of dissolved hydrocarbon concentrations were prepared by adding 5 mL sediment and 2 mL of 1 M NaOH into a 30 mL serum vial. Vials were sealed with butyl rubber stoppers, crimp-capped, shaken to dissolve sediment and stored upside down at room temperature for 24 hours until measurement. Concentrations of C1-C6 hydrocarbons were determined by injecting l mL headspace gas into a SRI 8610C gas chromatograph (SRI Instruments, USA) equipped with a flame ionization detector (FID). Methane and other hydrocarbons were separated on a Hayes-Sep-D packed column (2m) using a temperature program from 91 °C (hold for 1.8 min) to 200 °C (hold for 3 min) with 25 °C min-1 (Joye et al., 2011). Concentrations were calculated by comparison to a certified mixed standard (Scott Specialty Gases). The analytical precision determined from replicate measurements was better than 1.2%. Analysis of δ13C values of methane was

95 Chapter 4

performed on a gas chromatography combustion isotope ratio mass spectrometry (GC-C-IRMS) system combining a ThermoFinnigan Trace GC Ultra with a DELTA Plus XP mass spectrometer via a ThermoFinnigan GC Combustion III interface. Methane was separated isothermally at 40 °C for 8 min on a Carboxen-1006 PLOT fused-silica capillary column (0.32 mm × 30 m; Supelco, Inc., USA) with a constant helium flow of 3 mL min-1 (Ertefai et al., 2010). Sulfide was quantified using a spectrophotometer (Cline, 1969). Chloride concentrations of diluted pore water samples were quantified using a Dionex DX5000 ion chromatograph (Joye et al., 2004; Joye et al., 2010). Sulfate concentrations of diluted pore water samples were measured on the same system after filtration through an OnGuard II Ag/H matrix ion resin cartridge. Headspace hydrogen concentrations were measured after 4 days of incubation at 5 °C on a Peak Performer 1 Reduced Gas Analyzer equipped with a mercury oxide detector (Peak Laboratories, USA) (Hoehler et al., 1998; Lin et al., 2012). Pore water hydrogen concentrations were calculated from headspace

concentrations and the solubility constant for H2 corrected for temperature and salinity (Crozier and Yamamoto, 1974) as described by Lin et al. (2012). DIC concentration and carbon isotopic composition was measured by injecting 1 mL headspace gas after acidification of pore water with 0.2 mL of 0.1 mol L-1 phosphoric

acid (H3PO4) into a gas chromatography isotope ratio mass spectrometry consisting of a Hewlett-Packard 5890 GC equipped with a 6 m Poroplot Q column held at 35 °C and a Finnigan Mat DELTA S IRMS. Concentrations and carbon isotopic compositions of volatile fatty acids in the pore water were analyzed by using a liquid chromatography-isotope ratio mass spectrometry system, which consisted of a ThermoFinnigan Surveyor high performance liquid chromatography (HPLC) equipped with a VA Nucleogel Sugar 810-H column (300 × 7.8 mm; Macherey-Nagel, Germany) and a guard column of the same material linked to a DELTA Plus XP IRMS via a ThermoFinnigan LC Isolink interface (Heuer et al., 2006). Methanol in pore water was determined by a laboratory-built purge and trap pretreatment system connected to a Hewlett-Packard GC 5890 series II with an FID and Alltech Heliflex® AT-Q capillary column (30 m × 0.32 mm I.D.; Alltech Associates, Inc., Deerfield, US)

96 Chapter 4

(Zhuang et al., 2014). Dissolved pool of DMS and DMSP (DMS(P)d thereafter), and dissolved TMA in the pore water, and total base-hydrolyzable DMS and DMSP (DMS(P)t thereafter) and base-extractable TMA in the sediment were released by the addition of NaOH, preconcentrated with a purge and trap procedure, and analyzed on a Trace GC 2000 (Thermo Finnigan, Milan, Italy) coupled to a DSQ II MS (Thermo Finnigan, Texas, USA) with a Rtx®-Volatile Amine Column (30 m × 0.32 mm I.D., Restek GmbH, Homburg, Germany). The detection limit for DMS and TMA was 0.1 nM and 20 nM respectively. Stable carbon isotopic ratios of sediment-extracted TMA and DMSP were analyzed using a GC-C-IRMS system combining a ThermoFinnigan Trace GC Ultra with a DELTA Plus XP mass spectrometer via a ThermoFinnigan GC Combustion III interface. In brief, ~5 g sediment was placed in a 10-mL glass vial to which 1 mL of a 5 M NaOH was added and sealed immediately with a butyl stopper. After heating at 60 °C for 10 min, 800 μL gas samples were taken from the headspace

13 for GC-C-IRMS analysis. Determination of TOC and δ CTOC were conducted on pre-weighed, freeze-dried, decarbonatized sediment (corrected for weight of precipitated salts) using a ThermoScientific Flash 2000 elemental analyzer coupled to a ThermoScientific DELTA V Plus IRMS via a ThermoScientific ConFlo IV interface.

4.3.3 Stable isotope tracer experiment and radiotracer methanogenic activity measurements

Methanogenic activity in Orca Basin sediments was independently assessed at the University of Bremen using stable isotope tracers and at the University of Georgia, Athens, using radioactive tracers. The potential of different substrates to support methanogenesis in Orca Basin sediments was assessed in slurry incubations with 20% 13C-labeled substrates (bicarbonate, acetate, methanol), representing key substrates for three typical methanogenic pathways. Slurries were prepared in a glove box under a reducing

atmosphere (N2:H2, 97:3) from 1:1 sediment and anoxic artificial brine medium (major ion composition based on Schijf, 2007). Five different combinations of three

97 Chapter 4

labeled substrates and headspace gases were amended to sediment from the southern basin (MUC-6, 0-20 cm), the northern basin (MUC-7, 0-20 cm), a non-brine reference site as a positive control, and sterilized samples (autoclaved prior to substrate amendment) as a negative control. Substrate/headspace gas combinations included 1)

30 mM bicarbonate with a headspace of 80% H2 and 20% CO2, 2) 30 mM bicarbonate

with a headspace of 97% N2 and 3% H2, 3) 800 µM acetate with a headspace of N2, 4)

500 μM methanol with a headspace of N2, 5) no substrate with a headspace of N2. All samples were incubated for 181 days in a water bath in the dark at 30 °C and were constantly, gently agitated. Production of methane from the 13C-labeled substrates was determined by monitoring the stable carbon isotope composition of methane in the headspace at nine time points. The samples were stored by injecting 1.5 mL gas to replace a pH 1 to 2 saturated NaCl solution (pH adjusted by 1 M HCl) in a fully filled 10 mL headspace vial. The samples were placed upside down and stored at 4 °C until processing. Methane production rates were calculated from methane concentrations

13 13C (Cmethane), the increases of the fractions of C in methane (ΔF methane) over time (t),

13 13C and the fractions of C in the amended substrates (e.g., F methanol= 0.2) (Eq. 4.1; cf. Wegener et. al, 2012):

13C ΔF methane Methane production rate = Cmethane × 13C (4.1) F substrate × t The fractions of F13C in methane and substrates were calculated from the isotope ratios F13C = R13C/12C/(R13C/12C + 1), where R was the ratio of 13C and 12C. Methanogenesis rate measurements were conducted through radiotracer experiments with 14C-labeled substrates on sediment from the northern basin, using a method described previously (Bowles et al., 2011). Briefly, intact sediment samples were introduced into a modified Hungate tube, and injected with different radiotracers (14C-bicarbonate, 14C-acetate, 14C-methylamine). After 120 h of incubation at 15 °C, microbial activities were terminated by injecting 2.5 mL 2 M NaOH into the samples through a syringe and needle, and headspace was created by dragging the plunger to the base of the Hungate tube. 14C-methane from the headspace was stripped out with

14 14 air, purified from volatile C-labeled non-methane components such as CO2,

98 Chapter 4

14C-acetate, 14C-methylamine and 14C-methanol in series of traps and combusted in

14 Ni-Ti alloy tubing filled with CuO needles at 800 °C. C-CO2 formed from 14C-methane was trapped in ScintiSafe Gel cocktail mixed with Carbosorb E (5:1 ratio) and counted on a LS Beckman 6500 liquid scintillation counter. Methanogenesis rates from methylamine were expressed as turnover rate as the pore water concentrations were unknown.

4.3.4 Intact polar and core lipid analysis

Wet sediment samples were amended with an internal standard (C21-PC, Avanti Polar Lipids Inc., Alabaster, AL, USA) prior to solvent extraction using a modified Bligh & Dyer protocol (Bligh and Dyer, 1959) after Sturt et al. (2004). The total lipid

extract (TLE) was dried under a stream of N2 and stored at -20 °C until measurement. Intact polar and core lipids were quantified by injecting an aliquot of the TLE dissolved in methanol:dichloromethane (9:1, v:v) on a Dionex Ultimate 3000 HPLC system connected to a Bruker maXis Ultra-High Resolution quadrupole time-of-flight tandem mass spectrometer (qTOF-MS) equipped with an ESI ion source operating in positive mode (Bruker Daltonik, Bremen, Germany). The mass spectrometer was set to a resolving power of 27000 at m/z 1222 and every analysis was mass-calibrated by loop injections of a calibration standard and correction by lock mass, leading to a mass accuracy of better than 1-3 ppm. Ion source and other MS parameters were optimized by infusion of standards into the eluent flow from the LC system using a T-piece. Analyte separation was achieved using reversed phase HPLC on an Acquity

UPLC BEH C18 column (1.7 µm, 2.1 x 150 mm, Waters, Eschborn, Germany) maintained at 65 °C as described by Wörmer et al. (2013). In brief, analytes were eluted at a flow rate of 0.4 ml min-1 using linear gradients of methanol:water (85:15, eluent A) to methanol:isopropanol (50:50, eluent B) both with 0.04% formic acid and

0.1% NH3. The initial condition was 100% A held for 2 min, followed by a gradient to 15% B in 0.1 min and a gradient to 85% B in 18 min. The column was then flushed

99 Chapter 4

with 100% B for 8 min. Lipids were identified by retention time as well as by accurate molecular mass and isotope pattern match of proposed sum formulas in full scan mode as well as MS2 fragment spectra. Integration of peaks was performed on extracted ion chromatograms

+ + + of ±10 mDa width and included the [M+H] , [M+NH4] , and [M+Na] ions. Where applicable, double charged ions were included in the integration. Lipid abundances were corrected by the relative response of representative archaeal lipid standards versus the internal standard determined via an external calibration curve (cf. Elling et al., 2014). The abundances of glycerol dibiphytanyl glycerol tetraethers (GDGTs), hydroxy GDGTs (OH-GDGTs) and glycerol dialkanol diethers (GDDs) were corrected for the response factor of purified acyclic GDGT. The abundances of archaeol (AR) and hydroxyarchaeol (OH-AR) and their unsaturated derivatives were corrected for the response of purified AR. Similarly, the abundances of intact polar lipids were corrected for the response factors of purified monoglycosidic archaeol (1G-AR), diglycosidic archaeol (for 2G-AR, phosphoinositol-AR (PI-AR), 2G-OH-AR, PI-OH-AR, and glycosyl-phosphatidylglycerol-AR (G-PG-OH-AR)), acyclic 1G-GDGT (for 1G- and 1G-OH-GDGTs), 2G-GDGT (for 2G- and 2G-OH-GDGTs) and 1G-PG-GDGT (for HPH-GDGTs).

4.3.5 Molecular analysis

DNA was extracted with the MOBIO (Carlsbad, CA) PowerSoil kit following the manufacturer’s protocol. Prior to extraction, samples were centrifuged at maximum speed (16000 x g) for 5 minutes and overlying liquid decanted. The archaeal 16S rRNA gene was amplified with Speedstar DNA polymerase (Takara) using universal primers B8f (Huber et al., 2002) and 1492r (Teske et al., 2002) and the manufacturers recommended concentration for the buffer, dNTPs and tap polymerase. Amplification was performed in a thermal cycler (iCycler, Biorad) under the following conditions: initial denaturation at 95 °C for 4 minutes, 25 cycles of 95 °C for 10 seconds, 52 °C for 15 seconds and 72 °C for 20 seconds, a final

100 Chapter 4

extension of 72 °C for 10 minutes. PCR products were electrophoresed on a 1% agarose gel and visualized with ultraviolet light after staining with GelRed (Biotium). PCR products were purified with the Qiagen PCR purification kit.Clone libraries were constructed from PCR products with the TOPO TA Cloning Kit and TOP10 chemically competent cells (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Individual white colonies were arbitrarily picked and sent to Genewiz (South Plainfield, NJ) for Sanger sequencing using vector primers M13 F and M13 R. DNA alignments and sequence-associated taxonomic assignments were determined using the SILVA 111 small subunit rRNA database (Pruesse et al., 2012; Quast et al., 2012) and the ARB phylogenetic and sequence analysis program (Ludwig et al., 2004).

4.3.6 Thermodynamic calculations

The Gibbs free energy per mole of reaction under in situ conditions was calculated for different metabolic pathways (Table 4.1). Standard-state free energy yields (ΔGo) were calculated and corrected for temperature with Gibbs functions, ΔGo = ΔHo – TΔSo (Stumm and Morgan, 1981), using thermodynamic data of products and

2- - - + reactants from Wagman et al. (1982). The activity of SO4 , HCO3 , HS , and NH4 was computed through the input of major ions (Ca2+, Mg2+, Na+, K+, Cl-, dissolved

2- + - inorganic carbon, SO4 , NH4 and HS ) (Schijf, 2007) at in situ pH and temperature with Visual MINTEQ version 3.0, using the specific ion interaction theory activity coefficient model; activity coefficients of dissolved gases were set to 1.

101 Chapter 4

Table 4.1 Reactions and standard Gibbs free energies at 5 °C. Reaction ΔGo (kJ mol-1 reaction) Hydrogen-based methanogenesis -229.9 - + 4H2, aq + HCO3 + H → CH4, aq + 3H2O Hydrogen-based sulfate reduction -298.6 2- + 4H2, aq + SO4 + 2H → H2S + 4H2O Acetate-based methanogenesis -13.9 - - CH3COO + H2O → CH4, aq + HCO3 Acetate-based sulfate reduction -82.7 - 2- + - CH3COO + SO4 + H → H2S + 2HCO3 Methanol-based methanogenesis -224.5 - + 4CH3OH → 3CH4, aq + HCO3 + H + H2O Dimethylsulfide-based methanogenesis -45.9 - + 2(CH3)2S + 3H2O → 3CH4, aq + HCO3 + 2H2S + H Trimethylamine-based methanogenesis -401.7 + - + + 4(CH3)3NH + 9H2O → 9CH4, aq + 3HCO3 + 4NH4 + 3H Anaerobic oxidation of methane -229.9 - + CH4, aq + 3H2O + 4H2, aq → HCO3 + H Homoacetogenesis -215.9 - + - 4H2, aq + 2HCO3 + H → CH3COO +4H2O

4.4 Results and discussion 4.4.1 Geochemistry of Orca Basin sediments

Pore water salinities varied slightly between 261 and 273‰ in the southern basin (Fig. 4.2a-A) and decreased with depth from 255‰ to 222‰ in the northern basin (Fig. 4.2b-A). Such high salinities limit the active microbial community to extremely halophilic bacteria and archaea that are defined to thrive at salinities over 180‰ (Oren, 2008). The average concentration of chloride in the pore waters from the southern basin was 5134 mM (Fig. 4.2a-A), about eight times higher than at the reference site (550 mM, Fig. 4.4-A). In the northern basin, chloride concentrations were slightly lower, with an average value of 4453 mM between 32.5 and 52.5 cm (Fig. 4.2a-A). The difference in salinity and chloride concentrations might be related to spatial variations in the composition of the Orca Basin brine. Given the greater depth in the southern basin (2480 m), a higher salinity layer could be present to fill the deeper trough compared to the northern basin (ca. 2350 m).

102 Chapter 4

Fig. 4.2 Depth profiles of geochemical parameters A) salinity and chloride, B) sulfate and sulfide, 13 13 C) TOC and δ C-TOC, D) methane and ethane, E) δ C-CH4 in pore water and sediments from (a) southern and (b) northern Orca Basin.

Pore water sulfate concentrations varied between 38 and 43 mM in the southern basin (Fig. 4.2a-B). In contrast, sulfate declined with sediment depth from 30 mM to 16 mM in the northern basin concurrent with the appearance of low amounts of sulfide (Fig. 4.2b-B), potentially indicating the occurrence of sulfate reduction at these depths. In Orca Basin, sulfide production from sulfate reduction might be masked by precipitation of sulfide with abundant reduced iron and metastable iron sulfides (Wiesenburg et al., 1985). The TOC content of the sediment in the southern (average of 3.7%, Fig. 4.2a-C) and northern basin (average of 5.3%; Fig. 4.2b-C) was much higher than at the reference site (average of 0.6%, Fig. 4.4-C). The average δ13C values of TOC (-21.4‰ and -22.0‰) suggested a mixed terrestrial and marine planktonic source of organic matter (Fig. 4.2a, b-C). The high TOC content in Orca Basin sediments likely reflects a combination of high terrigenous and surface

103 Chapter 4

productivity-derived input and enhanced preservation (Kennicutt et al., 1992; Northam et al., 1981). Despite the presence of sulfate, considerable methane concentrations were observed in Orca Basin sediment. Methane accumulated in the upper 25 cm of the southern basin sediments (~800 μM), dropped markedly to 3 μM at 35 cm, and varied slightly between 0.5 μM and 3 μM below 35 cm (Fig. 4.2a-D). Methane was depleted in 13C within a narrow range of -75.5 to -76.7‰ throughout the core (Fig. 4.2a-E). In the northern basin, methane was more abundant and fluctuated between 300 and 1300 µM (Fig. 4.2b-D). The δ13C values of methane was -89.6‰ at 2.5 cm, became progressively less depleted with increasing depth, and reached a minimum of -83.1‰ near the base of the core (57.5 cm; Fig. 4.2b-E). In other deep-sea brine pools, methane has been assumed to originate from biogenic, thermogenic, or mixed biogenic and thermogenic sources (Charlou et al., 2003; Joye et al., 2005). Charlou

and colleagues (2003) reported that CH4 was microbially produced in brine overlying mud volcanoes of the eastern Mediterranean Sea, with indications from δ13C values of methane (-65.6‰) and methane to ethane ratios higher than 1000. In contrast, high δ13C values of methane (-30 to 40‰) and low methane to ethane ratios (~57) in the Red Sea brine pools might indicate a thermogenic origin of methane (Faber et al., 1998). Joye et al. (2005) suggested a mixed thermogenic and biogenic source of hydrocarbon gases in Gulf of Mexico seafloor brines based on the predominance of methane (> 98%) and intermediate δ13C values of methane (-66‰ to -60‰). Although biogenic methane is generally characterized by strongly negative δ13C values (<< -50‰), it has been demonstrated that methane with relatively enriched δ13C values (~ -40‰) was biologically produced in hypersaline ponds, as substrate limitation was decreasing isotopic fractionation during methanogenesis (Kelley et al., 2012). By comparison, the strongly depleted δ13C values of methane in Orca Basin (-77‰ to -89‰) are strongly indicative of microbial methanogenesis. A biogenic origin of methane is further supported by low ethane concentrations (mostly <1 μM; Fig. 4.2a, b-D) and methane to ethane ratios in the range of 600-3000. In contrast to the brine pool, methane concentrations were very low in the sulfate-bearing sediment

104 Chapter 4

at the reference site (mostly <1 μM; Fig. 4.4-B). Methanogenesis from hydrogen and acetate is likely to be inhibited due to competition for these substrates with sulfate reducers. In contrast, the coexistence of substantial amounts of biological methane with sulfate at both brine sites might result from methanogenesis from non-competitive substrates. Furthermore, as high salinity favors methanogenesis from non-competitive substrates such as methanol, DMS, and TMA, it could be expected that methylotrophic methanogenesis drives methane production in the hypersaline sediment of Orca Basin. Under the assumption that substrate limitation leads to 13C-enriched methane (Kelley et al., 2012), the depletion of methane isotope in the Orca Basin may reflect the availability of substrates for methanogenesis.

4.4.2 Potential methanogenic substrates in Orca Basin

To further constrain the source of methane in Orca Basin, potential methanogenic precursors including competitive substrates, i.e., hydrogen and acetate as well as non-competitive substrates, e.g., methanol, DMS, TMA, and DMSP, were quantified in pore waters and sediments. Hydrogen concentrations were uniformly low in the sediment of the southern and northern basins (~0.25 nM; Fig. 4.3a, b-A), which was lower by three orders of magnitude than values reported for shallow seafloor brines in the Gulf of Mexico

(Joye et al., 2009). In marine sediments, low H2 concentrations may be sustained by close coupling of microbial H2 production and consumption via interspecies hydrogen transfer in syntrophic relationships (Lin et al., 2012). The thermodynamic threshold concentrations of H2 depend on the redox potential of the respective anaerobic respiration pathway, following the sequence of nitrate reduction < iron and manganese reduction < sulfate reduction < methanogenesis < homoacetogenesis (Lin et al., 2012). Presumably, H2 concentrations in Orca Basin sediment might be maintained below the respective thresholds by sulfate reducers, thereby thermodynamically inhibiting hydrogenotrophic methanogenesis and homoacetogenesis (cf. section 4.4.3).

105 Chapter 4

Fig. 4.3 Depth profiles of potential methanogenic substrates A) hydrogen, B) DIC and δ13C-DIC, C) methanol, D) base-extractable TMA and δ13C-TMA, E) dissolved DMS(P)d, F) base-hydrolyzable DMS(P)t and δ13C-DMS(P) in pore water and sediments from the (a) southern and (b) northern Orca Basin.

DIC concentrations were generally lower than 10 mM with δ13C values of ~ -20‰ (Fig. 4.3a, b-B), indicating the source from organic matter remineralization (δ13C-TOC ~ -22‰). Acetate and other volatile fatty acids were below the detection limit (~ 3-5 µM; Heuer et al., 2006) both in the southern and northern basin. Contrary to previous studies that demonstrated high concentrations of acetate and consequent methanogenesis in brines and underlying sediments in the Gulf of Mexico and the Mediterranean Sea (Joye et al., 2009; Lazar et al., 2011), the absence of acetate marginalized the role of acetoclastic methanogenesis in Orca Basin. Furthermore, the highly depleted δ13C value of methane (-77 to -89‰) might rule out acetoclastic methanogenesis as the major source of methane in Orca Basin. Carbon isotopic fractionation during acetoclastic methanogenesis has been shown to be in the range from -35‰ to -7‰ in pure cultures (Krzycki et al., 1987; Penning et al., 2006; Valentine et al., 2004). Fermentation of organic matter would supposedly yield acetate with a δ13C signature close to that of TOC (~-22‰) and could therefore not

106 Chapter 4

explain the strong 13C-depletion of methane. Alternatively, acetate may be derived

13 from homoacetogenesis utilizing H2 and CO2 (δ C-DIC ~-20‰), which results in strong carbon isotope fractionation in the range of -50 to -60‰ (Gelwicks et al., 1989). Therefore, methanogenesis from homoacetogenic acetate could theoretically yield δ13C signatures similar to those observed in Orca Basin sediments. At the low hydrogen concentrations in Orca Basin however, both sulfate reducers and methanogens would likely out-compete homoacetogens (Lin et al., 2012), rendering a scenario of methanogenesis from 13C-depleted acetate unlikely. As hydrogen and acetate are unlikely substrates for methanogenesis in Orca Basin, we investigated the availability of several methylated substrates. Methanol was relatively abundant in the southern basin, with an average concentration of 6.5 μM (Fig. 4.3a-C). In the northern basin, methanol concentrations were lower (~1 μM), with an exception of 3.7 μM at the sediment surface (2.5 cm) (Fig. 4.3b-C). The presence of methanol in the southern and northern basins implied an in situ imbalance of methanol production and consumption. Known sources of methanol include fermentation of lignin and pectin, e.g., fermentative bacteria such as Haloanaerobium praevalens were found to be involved in pectinolytic activity in hypersaline environments (Ollivier et al., 1994). The utilization of methanol by methanogens has been documented in Big Soda Lake, Nevada, a moderately hypersaline lake (Oremland et al., 1982a), as well as the extremely hypersaline Dead Sea (Marvin-DiPasquale et al., 1999). Thus, methanogenesis from methanol seems feasible for Orca Basin, which is further supported by incubation experiments (section 4.4.4). Dissolved TMA was only detectable in the pore water at 2.5 cm (southern basin: 48 nM; northern basin: 41 nM; detection limit: 20 nM), whereas the total base-extractable TMA in the sediment was much more abundant in both the southern (3.4-7.0 μmol kg-1, Fig. 4.3a-D) and northern basin (2.2-5.5 μmol kg-1, Fig. 4.3b-D). By comparison, TMA concentrations in the sediment at the reference site were generally lower by a factor of three (0.8-2.3 μmol kg-1, Fig. 4.4-D). In hypersaline environments, abundant TMA might be derived from the breakdown of compounds

107 Chapter 4

such as glycine betaine that are biosynthesized as compatible solutes by halophilic microorganisms (Oren, 1990). Zhilina and Zavarzin (1990) isolated a halophilic homoacetogen from cyanobacterial mats of Sivash (Crimea) that produced methylamines and acetate from glycine betaine. Although TMA has been proposed as a central intermediate for methanogenesis in hypersaline environments owing to the ubiquity of precursor compounds (McGenity, 2010), this assumption lacks direct evidence from in situ abundance measurements, especially in deep-sea hypersaline sediments. The observation of elevated TMA concentrations in Orca Basin sediments highlights the importance of this low molecular weight metabolite in hypersaline environments. The pore water concentrations of the dissolved pool of DMS and DMSP (DMS(P)d) were only in the lower nanomolar range (~1.9 nM in the southern basin and ~1.4 nM in the northern basin, Fig. 4.3a, b-E) which may be attributed to the rapid turnover of this labile compound (Kiene and Capone, 1988). The total base-hydrolyzable DMSP (DMS(P)t) including DMS, dissolved and particulate DMSP were three orders of magnitude higher than those of DMS(P)d in the pore water of the southern and northern basins, with an average of 8.5 and 5.4 μmol kg-1, respectively (Fig. 4.3a, b-F). Analogous to TMA precursors, DMSP acts as an osmolyte in a wide variety of marine phytoplankton (Keller et al., 1989; Zhuang et al., 2011). Sedimentary DMSP might be derived from phytoplankton biomass exported from the overlying water column, which was reflected by an overall decrease in abundance with depth. The concentrations of DMS(P)t in the hypersaline sediments were about twice those in the reference site (3.3 μmol kg-1, Fig. 4.4-E), indicating increased preservation of organic matter from marine surface production in the brine pool. It has been demonstrated that DMSP and its decomposition products such as DMS and methanethiol could significantly stimulate methane production in a variety of anoxic environments including hypersaline sediments (Kiene et al., 1986; van der Maarel and Hansen, 1997). In addition to being a precursor of non-competitive substrates, DMSP could be degraded to other methanogenic substrates such as low-molecular-weight fatty acids, e.g., acrylate, propionate, and acetate (Kiene and Taylor, 1988). Collectively, the high abundance of precursor molecules in the

108 Chapter 4

sediments would facilitate methanogenesis from methylated substrates in Orca Basin.

Fig. 4.4 Depth profiles of geochemical parameters and methylated substrates A) salinity and chloride, B) methane and sulfate, C) TOC and δ13C-TOC, D) methanol and base-extractable TMA, E) dissolved DMS(P)d and base-hydrolyzable DMS(P)t in the pore water and sediments from the reference site.

Further evidence for methylotrophic methanogenesis in Orca Basin sediment could be inferred from δ13C systematics of methane and potential precursors. The δ13C value of base-extractable TMA in the sediment was on average -36.5‰ in the southern and northern basin (Fig. 4.3a, b-D). Based on pure culture studies, utilization of TMA should yield methane δ13C values in the range of 39-71‰ more depleted than the substrates (Londry et al., 2008; Summons et al., 1998; Whiticar, 1999). Thus δ13C values of TMA-derived methane in Orca Basin sediments should range between -75.5 and -106.5‰. The δ13C signature of DMS evolved from base-hydrolysis of DMSP was between -15.6 and -14.1‰ (Fig. 4.3a, b-F) and became progressively depleted with depth, indicating preferential utilization of 12C during microbial degradation of DMSP. Assuming a fractionation factor of 44-54‰, δ13C values of DMS-derived methane should range between -59 and -69‰ (Oremland et al., 1989; Whiticar, 1999). Methanogenesis from methanol would yield methane with δ13C values of -84 to -104‰ assuming δ13C of TOC (-21‰) as a proxy for δ13C of methanol (Krzycki et al., 1987; Londry et al., 2008; Oremland et al., 1989; Whiticar, 1999). In sum, the δ13C values of methane of -77 ‰ to -89 ‰ in Orca Basin sediments are in the expected range for methylotrophic methanogenesis utilizing predominantly TMA and methanol and to a

109 Chapter 4

lesser extent DMS and DMSP. Therefore, methanogenesis in Orca Basin is likely to a substantial degree fueled by non-competitive methylated substrates.

4.4.3 Thermodynamic calculations

Owing to the low hydrogen concentrations, free energy yields of hydrogenotrophic methanogenesis could not meet the biological energy quantum for life sustainment (-15 kJ mol-1, Hoehler, 2004; Schink and Stams, 2006) in sediments from southern and northern Orca Basin under in situ conditions (southern: ΔG: -14 to 14 kJ mol-1; northern: ΔG: 8 to 13 kJ mol-1; Fig 4.5a, b). The threshold for acetoclastic methanogenesis was slightly exceeded below 12.5 cm in the northern basin (ΔG: -14 to -19 kJ mol-1, Fig. 4.5b), and methanogenesis from acetate was also feasible in the southern basin (ΔG: -18 to -35 kJ mol-1, Fig. 4.5a), assuming acetate concentrations of 5 μM (limit of detection). Nevertheless, sulfate reduction with both hydrogen and acetate was energetically more favorable in the northern (ΔG: -48 to -56 kJ mol-1) and southern basin (ΔG: -57 to -65 kJ mol-1), indicating thermodynamic inhibition of hydrogenotrophic and acetoclastic methanogenesis in Orca Basin. In contrast, methanogenesis from methanol could have high energy yields in the southern (ΔG: -210 to -272 kJ mol-1, Fig. 4.5a) and northern Orca Basin (ΔG: -188 to -209 kJ mol-1, Fig. 4.5b). Likewise, methane production from DMS under in situ conditions was a highly exergonic process at both the northern and southern basin (Fig. 4.5a, b). High in situ energy yields of TMA-driven methanogenesis (< -600 kJ mol-1) in the southern and northern basin (Fig. 4.5a, b) reflected the importance of TMA as a key metabolic intermediate in hypersaline environments. Theoretically, sulfate reduction coupled with the metabolism of methylated substrates could provide more energy than methylotrophic methanogenesis, but it has been demonstrated that sulfate-reducing bacteria generally do not use these non-competitive substrates (Oremland and Polcin, 1982). Indeed, only very few known sulfate-reducing bacteria are capable of using methanol as their sole carbon substrate (Braun and Stolp, 1985), and no sulfate-reducing bacteria have been isolated yet which can grow on

110 Chapter 4 methylamines. In addition to methanogenesis, anaerobic oxidation of methane (AOM) might be expected to occur in the sediments due to the presence of abundant methane and sulfate. Methane oxidation to bicarbonate and hydrogen sulfide was thermodynamically unfavorable in the northern basin (ΔG: > -15 kJ mol-1), but could potentially occur in the southern basin (ΔG: -43 to -64 kJ mol-1). Homoacetogensis from H2/CO2 was energetically unfeasible in both southern and northern Orca Basin.

Fig. 4.5 Depth profiles of Gibbs free energy (ΔG) yield in the (a) southern and (b) northern Orca Basin. H-MOG: hydrogen-based methanogenesis; Ac-MOG: acetate-based methanogenesis, we used 5 μM where acetate concentrations below the detection limit; TMA-MOG: TMA-based methanogenesis, we used 20 nM where TMA concentrations were below the detection limit; DMS-MOG: DMS-based methanogenesis; Methanol-MOG: methanol-based methanogenesis; H-SR: hydrogen-based sulfate reduction, sulfide was assumed as 10 μM; Ac-SR: acetate-based sulfate reduction. The dashed line indicates the minimum biological energy quantum for life sustainment (-15 kJ mol-1).

111 Chapter 4

As halophiles need to divert energy for the maintenance of their osmotic equilibrium via the biosynthesis or uptake of organic compatible solutes and active ion pumping, high energy gains from methylotrophic methanogenesis could facilitate the survival of extremely halophilic methanogens in Orca Basin. A substrate-dependence of the upper salinity limit of methanogens was also observed in pure cultures, where methylamines facilitated growth to higher salinities (salinity: 270) than hydrogen and carbon dioxide (salinity: 120) as well as acetate (salinity: 40) (Oren, 1999). Environmental studies have demonstrated that these salinities should not be considered as the upper limit of activity in situ, e.g., acetoclastic methanogenesis has been detected in seafloor brines in the Gulf of Mexico at salinities exceeding 100 (Joye et al., 2009), but may rather reflect the relative importance of methanogenic pathways at different salinities. The relative energy yield from different methanogenic pathways indicated that methylotrophic methanogenesis with either methanol, DMS or TMA as substrate might be more significant than hydrogenotrophic and acetoclastic methanogenesis in hypersaline sediments of Orca Basin.

4.4.4 Methanogenic activity

13 13 Stable isotope tracer experiments with C-bicarbonate and H2, C-acetate, and 13C-methanol were conducted to trace the utilization of different substrates for methanogenesis in Orca Basin sediments. In sediment from the southern basin and the

13 reference site, δ CH4 did not change significantly over time for all types of substrates added. Likewise, samples from the northern basin amended with 13C-bicarbonate and

13 13 H2 as well as C-acetate did not show significant increase of δ CH4 after 181 days of

13 incubation (Fig. 4.6). However, a linear increase of δ CH4 over time was observed in

13 13 samples amended with C-methanol. δ CH4 in the duplicate incubations increased progressively from -87‰ at day 3 to -56‰ and -87‰ to -67‰ at day 112, respectively and did not increase further (Fig. 4.6). Based on mass balance

112 Chapter 4

calculations, the methane production rates were 2.5 ×10-4 and 4.1 ×10-4 nmol L-1 d-1 at day 112 for duplicate incubations. The killed controls did not show significant

13 changes in δ CH4 over time.

13 13 Fig. 4.6 Time series of δ C-CH4 in duplicate sediment samples amended with either C-labeled bicarbonate, acetate or methanol as well as killed control samples from northern Orca Basin.

Methanogenic activities in the southern basin could be too low to be captured by the stable isotope tracer experiment or could be confined to deeper sediment layers. Indeed, methane concentrations in sediment from the southern basin were lower than those in the northern basin, while methanol was more abundant in the southern basin. Based on the incubation experiments, enhanced utilization of methanol by relatively higher methanogenic activity might explain low methanol concentrations in the northern basin. The incorporation of 13C exclusively from methanol into methane suggested that methylotrophic methanogenesis was the only methanogenic pathway operating in the hypersaline sediment of Orca Basin. Similarly, 13C-enriched methane was only produced from 13C-labeled methylamines and methanol but not from labeled acetate and bicarbonate in hypersaline ponds in Baja California Sur, Mexico, and northern California, USA (Kelley et al., 2012). Radiotracer experiments using 14C-labeled bicarbonate, acetate and methylamine

113 Chapter 4

support the significance of methylotrophic methanogenesis in Orca Basin sediments. Methanogenesis from 14C-labeled methylamine was detected in the shallow sediment of the northern basin. The turnover rates of methylamine were 1.2×10-6 and 2.1× 10-6 per day at 2.5 cm and 7.5 cm, respectively. In contrast, no 14C uptake into methane from labeled bicarbonate and acetate was observed, suggesting that hydrogenotrophic and acetoclastic methanogenesis were not occurring in Orca Basin sediments. Lazar et al. (2011) demonstrated that methylotrophic methanogenesis was the only significant methanogenic pathway in shallow hypersaline sediments (0-40 cm) in the centre of the Napoli mud volcano in the Eastern Mediterranean Sea. However, methylamine turnover rates in Orca Basin were by two orders of magnitude lower in Orca Basin sediments than in the Napoli mud volcano (Lazar et al., 2011), which could be attributed to the lower salinity in the Napoli mud volcano (< 3 M Cl-) compared to Orca Basin (~5 M Cl-). Despite the relatively low methanogenic activity, the combination of stable isotope and radioactive tracer experiments attested a dominance of methylotrophic methanogenesis in the hypersaline sediments of Orca Basin.

4.4.5 Diversity of archaeal 16S ribosomal RNA genes

Sanger sequencing targeting the archaeal 16S rRNA gene revealed diverse archaeal populations in Orca Basin sediments. Sequences related to the Marine Group I (MG-I) dominated the assemblages in the surface sediment (0-5 cm, 28 archaeal sequences, 61% MG-I) and at 55-60 cm (37 archaeal sequences, 73% MG-I) of the southern basin (Fig. 4.7). The large number of MG-I sequences may be indicative of the preservation of water column-derived DNA, which has been demonstrated to occur over geologic timescales in anoxic sediments (Corinaldesi et al., 2011). Other sequences were related to the Marine Group II (MG-II) Euryarchaeota (9%) and the euryarchaeal clades Halobacteriales (4%) and Methanohalophilus (13%). In particular, archaea of the genus Methanohalophilus are obligate halophilic methanogens that have been previously detected in a wide range of hypersaline

114 Chapter 4

environments (Oremland et al., 1982b; Orphan et al., 2008). Cultured members of this genus are obligate that could grow at salinities of up to 250 and

metabolize a variety of C1 compounds including methylamines, methanol and

methylsulfides, rather than H2, formate, or acetate (Liu et al., 1990; Paterek and Smith, 1988). In the deep sediment sample (55-60 cm) from the southern basin, sequences related to the Marine Group III, the Deep-Sea Hydrothermal Vent Euryarchaeotal Group 5 (DHVE-5) and putative anaerobic methane oxidizers of the ANME-1a, ANME-1b and ANME-2c clades were detected (Fig. 4.7). The occurrence of ANME sequences and the consequent AOM might explain the strong decrease of methane concentrations below 30 cm sediment depth in the southern basin (Fig. 4.2a-D). While thermodynamic calculations indicate that AOM is feasible in sediment from the

-1 13 southern basin (ΔG: -43 to -64 kJ mol ) this is not reflected in the δ CH4 profile, which may be related to the balance of methane flux and methane oxidation rates.

Fig. 4.7 Relative abundances of archaeal 16S rRNA gene sequences at 0-5 cm and 55-60 cm sediment depth in southern and northern Orca Basin.

Archaeal sequence diversity in the surface sediment (0-5 cm, 56 archaeal sequences) of the northern basin was similar to the southern basin with the exception of a higher relative abundance of Methanohalophilus sequences (32%; Fig. 4.7) and occurrence of sequences related to the ANME-2a and the Miscellaneous Crenarchaeotic Group (MCG). In the deep sediment sample (50-55 cm, 33 archaeal sequences), Methanohalophilus accounted for 85% of the sequences, the rest being related to MG-I. The elevated abundance of Methanohalophilus-related sequences in

115 Chapter 4 the northern basin clone library reflected the higher potential for methylotrophic methanogenesis compared to the southern basin inferred from substrate concentrations and labeling experiments. Collectively, 16S rRNA gene analysis provided phylogenetic evidence for the existence of a methylotrophic methanogen community in the hypersaline sediment of Orca Basin and supported the observation of higher methanogenic activity in the northern basin.

4.4.6 Lipid biomarkers

To constrain the archaeal community composition and identify biomarkers indicative for methane cycling, we analyzed the relative abundances of archaeal core and intact polar membrane in the sediment of Orca Basin. Overall, the relative abundances of major core and intact polar lipid groups were highly similar across the sampled sediment intervals as well as between the southern and northern basin sites (Fig. 4.8). Archaeal tetraether lipids (GDGTs) and derivatives such as hydroxylated GDGTs and GDDs with zero to five pentacyclic rings were the major components of the core lipid pool in Orca Basin sediments. These compounds are sourced mainly by aerobic planktonic Thaumarchaeota and ubiquitously found in the marine water column and sediments (Schouten et al., 2000; Wakeham et al., 2007) and are therefore reflective of the high input of marine organic matter in Orca Basin (Tribovillard et al., 2008; Tribovillard et al., 2009). The diether lipids archaeol and hydroxyarchaeol as well as their unsaturated homologues containing up to seven unsaturations comprised 22-43% of the core lipids. In contrast to GDGTs, high abundances of archaeol and hydroxyarchaeol are typically not observed in marine water column and sediment samples (cf. Turich and Freeman, 2011; Turich et al., 2007). While archaeol can be found in cultured representatives of all archaeal clades (Koga and Morii, 2005; Schouten et al., 2008), high abundances of archaeol and specifically hydroxyarchaeol in marine sediments are often related to the occurrence of methanogenic and methanotrophic archaea (Hinrichs et al., 1999;

116 Chapter 4

Rossel et al., 2008; Sprott et al., 1990). Similarly, cellular membranes of extremely halophilic archaea consist primarily of archaeol but are devoid of hydroxyarchaeol (Kates, 1993; Koga and Morii, 2005). As the relative abundances of archaeol and hydroxyarchaeol as well as unsaturated hydroxyarchaeols are strongly correlated in Orca Basin sediments (Fig. 4.8), a common methanogenic or methanotrophic archaeal source of these lipids is very likely.

Fig. 4.8 Relative abundances of archaeal core and intact polar lipids in sediment from the southern (a) and northern (b) Orca Basin. Core lipid nomenclature: AR: archaeol; uns ARs: summed unsaturated archaeols; OH-AR: hydroxyarchaeol; uns OH-ARs: summed unsaturated hydroxyarchaeols; GDGTs: glycerol dibiphytanyl glycerol tetraethers; OH-GDGTs: hydroxy GDGTs; GDDs: glycerol dialkanol diethers. Intact polar lipid nomenclature indicates headgroup and core lipids combinations: 1G: monoglyosyl; 2G: diglycosyl; PI: posphoinositol; G-PG: glycosyl-phosphatidylclycerol; HPH: hexose-phosphohexose.

Polyunsaturated archaeols have so far not been reported from environmental samples, but have been found in several cultivated extremely halophilic Halobacteriales strains (Gibson et al., 2005; Nichols and Franzmann, 1992; Qiu et al.,

117 Chapter 4

1998) and appear to be involved in adaptation of the lipid membrane to high salinities (Dawson et al., 2012). Similarly, polyunsaturated archaeols and hydroxyarchaeols are also synthesized by the methylotrophic methanogen Methanococcoides burtonii in response to cold temperatures (Franzmann et al., 1992; Nichols et al., 2004). In Orca Basin sediments, the relative abundance of unsaturated archaeols was not correlated with either archaeol or saturated and unsaturated hydroxyarchaeols, suggesting disparate sources of these lipids (Fig. 4.9). Therefore, unsaturated archaeols in Orca Basin likely originate from Halobacteriales-like archaea while unsaturated hydroxarchaeols along with saturated archaeol and hydroxyarchael could be synthesized by methanogenic or methanotrophic archaea.

Fig. 4.9 Linear correlation coefficients (p <0.05) calculated from relative abundances of core and intact polar lipids in Orca Basin sediments.

Intact polar lipids accounted for 3 to 7% of the total archaeal lipids in Orca Basin sediments, which was within the range reported for marine sediments, with an exception of 14% at 50-55 cm sediment depth in the northern basin (Lipp and Hinrichs, 2009; Liu et al., 2011). Among the intact polar lipids, intact polar glycosidic and phospho-glycosidic GDGTs and hydroxy GDGTs with zero to five cyclopentyl moieties dominated the intact polar lipid pool in most samples (Fig. 4.8). These lipids

118 Chapter 4

are typical for planktonic Thaumarchaeota (Elling et al., 2014; Schouten et al., 2008) but may also be produced in significant amounts by benthic archaea, e.g., methanogens and ANME-1 and ANME-2 methanotrophs (Koga et al., 1993; Lipp and Hinrichs, 2009; Rossel et al., 2008; Tornabene and Langworthy, 1979). Additionally, GDGTs are not known to be produced by extremely halophilic archaea (Kates, 1993). The relative abundances of intact polar GDGTs were also not significantly correlated to the abundances of core and intact polar archaeols, indicating different sources of GDGTs and archaeols in Orca Basin. Given the high abundance of (potentially preserved) Marine Group I Thaumarchaeota 16S rRNA gene sequences in Orca Basin sediments, preservation of water column-derived thaumarchaeal intact polar lipids (cf. Schouten et al., 2010; Xie et al., 2013) in the anoxic brine may contribute significant amounts of intact polar GDGTs to the sedimentary lipid pool (Harvey and Kennicutt II, 1992; Kennicutt et al., 1992). Intact polar diether lipids consisted of saturated archaeols and hydroxyarchaeols with glycosidic and phosphatidic headgroups that are found both in cultivated methanogens (Koga et al., 1993) as well as uncultivated putative methanotrophic archaea of the ANME-1, ANME-2 and ANME-3 clades (Rossel et al., 2008). Polyunsaturated intact polar archaeols as well as headgroups that are specific for Halobacteriales, such as PGP-Me (Kates, 1993), were not detected, suggesting that Orca Basin sediments likely do not harbor significant populations of this archaeal clade. The relative abundances of intact polar and core archaeols and hydroxyarchaeols were significantly correlated, indicating that these intact polar lipids are likely produced within Orca Basin (Fig. 4.9). Intact polar lipid diversity as well as relative abundance of archaeal diether lipids was highest at 50-55 cm sediment depth in the northern basin (72%, Fig. 4.8), potentially indicating high levels of methanogenic biomass and/or activity in this sample. This is consistent with high abundances of 16S rRNA gene sequences related to methylotrophic methanogens (section 4.4.5). In conclusion, unusual polyunsaturated core archaeols found in Orca Basin are likely derived from extremely halophilic Halobacteriales-like archaea inhabiting the

119 Chapter 4

brine of Orca Basin. In contrast, highly specific polyunsaturated core hydroxyarchaeols as well as intact polar archaeols and hydroxyarchaeols indicate that the archaeal community in the sediments of Orca Basin is primarily composed of methanogenic and/or methanotrophic archaea.

4.5 Conclusions

Significant amounts of methane were observed in hypersaline sediments from southern and northern Orca Basin. Although previous biogeochemical investigations suggested a potential of methylated compounds as methanogenic substrates in hypersaline environments, the formation of methane in Orca Basin and other large deep-sea hypersaline basins remained largely unconstrained. In this study, multiple lines of evidence suggested that methylotrophic methanogenesis is the major source of methane in Orca Basin sediments: 1) Strongly depleted δ13C values of methane in the range of -76.7‰ to -89.1‰ as well as methane to ethane ratios of 600-3000 pointed to a biogenic origin of methane in Orca Basin. 2) Hydrogenotrophic and acetoclastic methanogenesis were unlikely to occur in Orca Basin due to competitive inhibition by sulfate reducers in the presence of abundant sulfate, low concentrations of H2 and no detectable acetate. In contrast, methylated substrates and their precursors were highly abundant relative to the non-hypersaline reference site, suggesting that these substrates could potentially fuel methane production in the hypersaline sediments of Orca Basin. 3) Thermodynamic calculations demonstrated that methylotrophic methanogenesis was much more favorable than hydrogenotrophic and acetoclastic methanogenesis under in situ conditions. High energy gains from methylotrophic methanogenesis could facilitate the survival of halophilic methanogens, as halophiles require more energy to maintain an osmotically balanced and functional cytoplasm. 4) Stable isotope tracer experiments with different 13C-labeled substrates indicated 13C incorporation into methane only from methanol and exclusively in

120 Chapter 4

sediments from northern Orca Basin. In addition, methanogenesis from 14C-labeled methylamine but not from bicarbonate or acetate was detected in shallow sediments of the northern basin, attesting that methylotrophic methanogenesis was the dominant methanogenic pathwayin Orca Basin sediment. 5) 16S rRNA gene sequences related to obligately halophilic methylotrophic methanogens of the genus Methanohalophilus were abundant in the clone library sequences constructed from DNA extracted from the sediment of the southern Orca Basin and dominated archaeal 16S rRNA gene sequence libraries in the subsurface of the northern basin. High abundances of methanogen lipid biomarkers, e.g., intact polar and polyunsaturated archaeols and hydroxyarchaeols, were detected in Orca Basin sediments. In conclusion, biogeochemical, experimental, and phylogenetic evidence confirmed the presence of an active methanogenic community utilizing abundant methylated substrates and sustaining the high concentrations of methane that are observed in the hypersaline sediments of Orca Basin. Methylotrophic methanogenesis and other processes involving methylated compounds as central intermediates might also be significant contributors to carbon cycling in other deep sea hypersaline basins such as in the Mediterranean and Red Seas.

4.6 Acknowledgements

We gratefully acknowledge the captain, crew and participating scientists of RV Atlantis cruise AT18-2. The cruise and the sampling program in the deep Gulf of Mexico and Orca Basin were supported by NSF Microbial Observatories and Microbial Interactions and Processes (Collaborative Research: A Microbial Observatory examining Microbial Abundance, Diversity, Associations, and Activity at Seafloor Brine Seeps. PIs: S. B. Joye, I. MacDonald and A. Teske). Verena Heuer, Yu-Shih Lin and Marcos Yoshinaga are thanked for supporting incubation experiments as well as biogeochemical and lipid analyses. We thank Jenny Wendt for supporting TOC analyses. The study was supported by the DFG through the Research

121 Chapter 4

Center/Excellence Cluster MARUM-Center for Marine Environmental Sciences. G.-C. Zhuang was sponsored by the Chinese Scholarship Council (CSC) and the Bremen International Graduate School for Marine Sciences (GLOMAR).

4.7 References

Antunes, A., Ngugi, D.K., Stingl, U., 2011. Microbiology of the Red Sea (and other) deep‐sea anoxic brine lakes. Environmental Microbiology Reports 3, 416-433. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911-917. Boetius, A., Joye, S., 2009. Thriving in salt. Science 324, 1523-1525. Bowles, M.W., Samarkin, V.A., Joye, S.B., 2011. Improved measurement of microbial activity in deep-sea sediments at in situ pressure and methane concentration. Limnology and Oceanography: Methods 9, 499-506. Braun, M., Stolp, H., 1985. Degradation of methanol by a sulfate reducing bacterium. Archives of Microbiology 142, 77-80. Charlou, J., Donval, J., Zitter, T., Roy, N., Jean-Baptiste, P., Foucher, J., Woodside, J., 2003. Evidence of methane venting and geochemistry of brines on mud volcanoes of the eastern Mediterranean Sea. Deep Sea Research Part I: Oceanographic Research Papers 50, 941-958. Cline, J.D., 1969. Spectrophotometric determination of hydrogen sulfide in natural waters. Limnology & Oceanography 14, 454-458. Conrad, R., 2005. Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Organic Geochemistry 36, 739-752. Corinaldesi, C., Barucca, M., Luna, G.M., Dell’Anno, A., 2011. Preservation, origin and genetic imprint of extracellular DNA in permanently anoxic deep-sea sediments. Molecular Ecology 20, 642-654. Crozier, T.E., Yamamoto, S., 1974. Solubility of hydrogen in water, sea water, and sodium chloride solutions. Journal of Chemical and Engineering Data 19, 242-244. Dawson, K.S., Freeman, K.H., Macalady, J.L., 2012. Molecular characterization of core lipids from halophilic archaea grown under different salinity conditions. Organic Geochemistry 48, 1-8. Donnelly, M., Dagley, S., 1980. Production of methanol from aromatic acids by Pseudomonas putida. Journal of Bacteriology 142, 916-924. Elling, F.J., Könneke, M., Lipp, J.S., Becker, K.W., Gagen, E.J., Hinrichs, K.-U., 2014. Effects of

122 Chapter 4

growth phase on the membrane lipid composition of the thaumarchaeon Nitrosopumilus maritimus and their implications for archaeal lipid distributions in the marine environment. Geochimica et Cosmochimica Acta 141, 579-597. Ertefai, T.F., Heuer, V.B., Prieto-Mollar, X., Vogt, C., Sylva, S.P., Seewald, J., Hinrichs, K.-U., 2010. The biogeochemistry of sorbed methane in marine sediments. Geochimica et Cosmochimica Acta 75, 6033-6048. Faber, E., Botz, R., Poggenburg, J., Schmidt, M., Stoffers, P., Hartmann, M., 1998. Methane in Red Sea brines. Organic Geochemistry 29, 363-379. Franzmann, P., Springer, N., Ludwig, W., Conway de Macario, E., Rohde, M., 1992. A Methanogenic Archaeon from Ace Lake, Antarctica: Methanococcoides burtonii sp. nov. Systematic and Applied Microbiology 15, 573-581. Gelwicks, J.T., Risatti, J.B., Hayes, J., 1989. Carbon isotope effects associated with autotrophic acetogenesis. Organic Geochemistry 14, 441-446. Gibson, J.A., Miller, M.R., Davies, N.W., Neill, G.P., Nichols, D.S., Volkman, J.K., 2005. Unsaturated diether lipids in the psychrotrophic archaeon Halorubrum lacusprofundi. Systematic and Applied Microbiology 28, 19-26. Harvey, H.R., Kennicutt II, M.C., 1992. Selective alteration of Sargassum lipids in anoxic sediments of the Orca Basin. Organic Geochemistry 18, 181-187. Heuer, V., Elvert, M., Tille, S., Krummen, M., Mollar, X.P., Hmelo, L.R., Hinrichs, K.-U., 2006. Online δ13C analysis of volatile fatty acids in sediment/porewater systems by liquid chromatography-isotope ratio-mass spectrometry. Limnology and Oceanography: Methods 4, 346-357. Hinrichs, K.-U., Boetius, A., 2003. The anaerobic oxidation of methane: new insights in microbial ecology and biogeochemistry, Ocean Margin Systems. Springer, pp. 457-477. Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewer, P.G., DeLong, E.F., 1999. Methane-consuming archaebacteria in marine sediments. Nature 398, 802-805. Hoehler, T., 2004. Biological energy requirements as quantitative boundary conditions for life in the subsurface. Geobiology 2, 205-215. Hoehler, T.M., Alperin, M.J., Albert, D.B., Martens, C.S., 1998. Thermodynamic control on hydrogen concentrations in anoxic sediments. Geochimica et Cosmochimica Acta 62, 1745-1756. Huber, H., Hohn, M.J., Rachel, R., Fuchs, T., Wimmer, V.C., Stetter, K.O., 2002. A new phylum of Archaea represented by a nanosized hyperthermophilic symbiont. Nature 417, 63-67. Joye, S.B., Boetius, A., Orcutt, B.N., Montoya, J.P., Schulz, H.N., Erickson, M.J., Lugo, S.K., 2004. The anaerobic oxidation of methane and sulfate reduction in sediments from Gulf of

123 Chapter 4

Mexico cold seeps. Chemical Geology 205, 219-238 Joye, S.B., Bowles, M.W., Samarkin, V.A., Hunter, K.S., Niemann, H., 2010. Biogeochemical signatures and microbial activity of different cold-seep habitats along the Gulf of Mexico deep slope. Deep-Sea Research Part II: Topical Studies in Oceanography 57, 1990-2001. Joye, S.B., MacDonald, I.R., Montoya, J.P., Peccini, M., 2005. Geophysical and geochemical signatures of Gulf of Mexico seafloor brines. Biogeosciences 2, 295-309. Joye, S.B., Samarkin, V.A., MacDonald, I.R., Hinrichs, K.-U., Elvert, M., Teske, A.P., Lloyd, K.G., Lever, M.A., Montoya, J.P., Meile, C.D., 2009. Metabolic variability in seafloor brines revealed by carbon and sulphur dynamics. Nature Geoscience 2, 349-354. Kates, M., 1993. Biology of halophilic bacteria, Part II. Experientia 49, 1027-1036. Keller, M.D., Bellows, W.K., Guillard, R.R., 1989. Dimethyl sulfide production in marine phytoplankton. In Biogenic sulfur in the environment. Saltzman, E.S., Cooper, W.J. (eds.). Washington, D.C.: American Chemical Society, pp. 167-182. Kelley, C.A., Poole, J.A., Tazaz, A.M., Chanton, J.P., Bebout, B.M., 2012. Substrate limitation for methanogenesis in hypersaline environments. Astrobiology 12, 89-97. Kennicutt, M., Macko, S.A., Harvey, H.R., Bidigare, R.R., 1992. Preservation of Sargassum under anoxic conditions: molecular and isotopic evidence. Organic matter: Productivity, Accumulation, and Preservation in Recent and Ancient Sediments, 123-141. Kiene, R., Capone, D., 1988. Microbial transformations of methylated sulfur compounds in anoxic salt marsh sediments. Microbial Ecology 15, 275-291. Kiene, R., Taylor, B., 1988. Demethylation of dimethylsulfoniopropionate and production of thiols in anoxic marine sediments. Applied and Environmental Microbiology 54, 2208-2212. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. Koga, Y., Morii, H., 2005. Recent advances in structural research on ether lipids from archaea including comparative and physiological aspects. Bioscience, Biotechnology, and Biochemistry 69, 2019-2034. Koga, Y., Nishihara, M., Morii, H., Akagawa-Matsushita, M., 1993. Ether polar lipids of methanogenic bacteria: structures, comparative aspects, and biosyntheses. Microbiological Reviews 57, 164-182. Krzycki, J., Kenealy, W., DeNiro, M., Zeikus, J., 1987. Stable carbon isotope fractionation by Methanosarcina barkeri during methanogenesis from acetate, methanol, or carbon dioxide-hydrogen. Applied and Environmental Microbiology 53, 2597-2599. Lazar, C.S., Parkes, R.J., Cragg, B.A., L'Haridon, S., Toffin, L., 2011. Methanogenic diversity and

124 Chapter 4

activity in hypersaline sediments of the centre of the Napoli mud volcano, Eastern Mediterranean Sea. Environmental Microbiology 13, 2078-2091. Lin, Y.-S., Heuer, V.B., Goldhammer, T., Kellermann, M.Y., Zabel, M., Hinrichs, K.-U., 2012. Towards constraining H2 concentration in subseafloor sediment: A proposal for combined analysis by two distinct approaches. Geochimica et Cosmochimica Acta 77, 186-201. Lipp, J.S., Hinrichs, K.-U., 2009. Structural diversity and fate of intact polar lipids in marine sediments. Geochimica et Cosmochimica Acta 73, 6816-6833. Liu, X., Lipp, J.S., Hinrichs, K.-U., 2011. Distribution of intact and core GDGTs in marine sediments. Organic Geochemistry 42, 368-375. Liu, Y., Boone, D.R., Choy, C., 1990. Methanohalophilus oregonense sp. nov., a methylotrophic methanogen from an alkaline, saline aquifer. International Journal of Systematic Bacteriology 40, 111-116. Londry, K.L., Dawson, K.G., Grover, H.D., Summons, R.E., Bradley, A.S., 2008. Stable carbon isotope fractionation between substrates and products of Methanosarcina barkeri. Organic Geochemistry 39, 608-621. Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Buchner, A., Lai, T., Steppi, S., Jobb, G., Förster, W., 2004. ARB: a software environment for sequence data. Nucleic Acids Research 32, 1363-1371. Mancinelli, R., Fahlen, T., Landheim, R., Klovstad, M., 2004. Brines and evaporites: analogs for Martian life. Advances in Space Research 33, 1244-1246. Marvin-DiPasquale, M., Oren, A., Cohen, Y., Oremland, R.S., 1999. Radiotracer studies of bacterial methanogenesis in sediments from the Dead Sea and Solar Lake (Sinai). In Microbiology and Biogeochemistry of Hypersaline Environments. Oren, A. (ed.). Boca Raton: CRC Press, pp 149-160. McGenity, T., 2010. Methanogens and methanogenesis in hypersaline environments, Handbook of Hydrocarbon and Lipid Microbiology. Springer, pp. 665-680. Nichols, D.S., Miller, M.R., Davies, N.W., Goodchild, A., Raftery, M., Cavicchioli, R., 2004. Cold adaptation in the Antarctic archaeon Methanococcoides burtonii involves membrane lipid unsaturation. Journal of Bacteriology 186, 8508-8515. Nichols, P.D., Franzmann, P.D., 1992. Unsaturated diether phospholipids in the antartic methanogen Methanococcoides burtonii. FEMS Microbiology Letters 98, 205-208. Northam, M.A., Curry, D.J., Scalan, R.S., Parker, P.L., 1981. Stable carbon isotope ratio variations of organic matter in Orca Basin sediments. Geochimica et Cosmochimica Acta 45, 257-260. Ollivier, B., Caumette, P., Garcia, J.-L., Mah, R., 1994. Anaerobic bacteria from hypersaline environments. Microbiological Reviews 58, 27-38.

125 Chapter 4

Oremland, R.S., Kiene, R.P., Mathrani, I., Whiticar, M.J., Boone, D.R., 1989. Description of an estuarine methylotrophic methanogen which grows on dimethyl sulfide. Applied and Environmental Microbiology 55, 994-1002. Oremland, R.S., Marsh, L., DesMarais, D.J., 1982a. Methanogenesis in Big Soda Lake, Nevada: an alkaline, moderately hypersaline desert lake. Applied and Environmental Microbiology 43, 462-468. Oremland, R.S., Marsh, L.M., Polcin, S., 1982b. Methane production and simultaneous sulphate reduction in anoxic, salt marsh sediments. Nature 296, 143-145. Oremland, R.S., Polcin, S., 1982. Methanogenesis and sulfate reduction: competitive and noncompetitive substrates in estuarine sediments. Applied and Environmental Microbiology 44, 1270-1276. Oren, A., 1990. Formation and breakdown of glycine betaine and trimethylamine in hypersaline environments. Antonie van Leeuwenhoek 58, 291-298. Oren, A., 1999. Bioenergetic aspects of halophilism. Microbiology and Molecular Biology Reviews 63, 334-348. Oren, A., 2008. Microbial life at high salt concentrations: phylogenetic and metabolic diversity. Saline Systems 4, 13. Orphan, V., Jahnke, L., Embaye, T., Turk, K., Pernthaler, A., Summons, R., Des Marais, D., 2008. Characterization and spatial distribution of methanogens and methanogenic in hypersaline microbial mats of Baja California. Geobiology 6, 376-393. Paterek, J.R., Smith, P.H., 1988. Methanohalophilus mahii gen. nov., sp. nov., a Methylotrophic Halophilic Methanogen. International Journal of Systematic Bacteriology 38, 122-123. Penger, J., Conrad, R., Blaser, M., 2012. Stable carbon isotope fractionation by methylotrophic methanogenic archaea. Applied and Environmental Microbiology 78, 7596-7602. Penning, H., Claus, P., Casper, P., Conrad, R., 2006. Carbon isotope fractionation during acetoclastic methanogenesis by Methanosaeta concilii in culture and a lake sediment. Applied and Environmental Microbiology 72, 5648-5652. Pilcher, R.S., Blumstein, R.D., 2007. Brine volume and salt dissolution rates in Orca Basin, northeast Gulf of Mexico. AAPG bulletin 91, 823-833. Pruesse, E., Peplies, J., Glöckner, F.O., 2012. SINA: accurate high-throughput multiple sequence alignment of ribosomal RNA genes. Bioinformatics 28, 1823-1829. Qiu, D.F., Games, M.P., Xiao, X.Y., Games, D.E., Walton, T.J., 1998. Application of high‐performance liquid chromatography/electrospray mass spectrometry for the characterization of membrane lipids in the haloalkaliphilic archaebacterium Natronobacterium magadii. Rapid Communications in Mass Spectrometry 12, 939-946.

126 Chapter 4

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., Peplies, J., Glöckner, F.O., 2012. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research, gks1219. Roberts, M.F., 2005. Organic compatible solutes of halotolerant and halophilic microorganisms. Saline Systems 1, 1-30. Rosenfeld, W.D., Silverman, S.R., 1959. Carbon isotope fractionation in bacterial production of methane. Science 130, 1658-1659. Rossel, P.E., Lipp, J.S., Fredricks, H.F., Arnds, J., Boetius, A., Elvert, M., Hinrichs, K.-U., 2008. Intact polar lipids of anaerobic methanotrophic archaea and associated bacteria. Organic Geochemistry 39, 992-999. Schijf, J., 2007. Alkali elements (Na, K, Rb) and alkaline earth elements (Mg, Ca, Sr, Ba) in the anoxic brine of Orca Basin, northern Gulf of Mexico. Chemical Geology 243, 255-274. Schink, B., Stams, A.J., 2006. Syntrophism among prokaryotes. Springer. Schink, B., Zeikus, J., 1980. Microbial methanol formation: a major end product of pectin metabolism. Current Microbiology 4, 387-389. Schouten, S., Hopmans, E.C., Baas, M., Boumann, H., Standfest, S., Könneke, M., Stahl, D.A., Damsté, J.S.S., 2008. Intact membrane lipids of “Candidatus Nitrosopumilus maritimus,” a cultivated representative of the cosmopolitan mesophilic group I . Applied and Environmental Microbiology 74, 2433-2440. Schouten, S., Hopmans, E.C., Pancost, R.D., Damsté, J.S.S., 2000. Widespread occurrence of structurally diverse tetraether membrane lipids: evidence for the ubiquitous presence of low-temperature relatives of hyperthermophiles. Proceedings of the National Academy of Sciences 97, 14421-14426. Schouten, S., Middelburg, J.J., Hopmans, E.C., Sinninghe Damsté, J.S., 2010. Fossilization and degradation of intact polar lipids in deep subsurface sediments: a theoretical approach. Geochimica et Cosmochimica Acta 74, 3806-3814. Shokes, R.F., Trabant, P.K., Presley, B.J., Reid, D.F., 1977. Anoxic, hypersaline basin in the northern Gulf of Mexico. Science 196, 1443-1446. Smith, J.M., Green, S.J., Kelley, C.A., Prufert‐Bebout, L., Bebout, B.M., 2008. Shifts in methanogen community structure and function associated with long‐term manipulation of sulfate and salinity in a hypersaline microbial mat. Environmental Microbiology 10, 386-394. Sprott, G., Ekiel, I., Dicaire, C., 1990. Novel, acid-labile, hydroxydiether lipid cores in methanogenic bacteria. Journal of Biological Chemistry 265, 13735-13740. Stumm, W., Morgan, J.J., 1981. Aquatic chemistry: an introduction emphasizing chemical equilibria in natural waters. John Wiley.

127 Chapter 4

Sturt, H.F., Summons, R.E., Smith, K., Elvert, M., Hinrichs, K.U., 2004. Intact polar membrane lipids in prokaryotes and sediments deciphered by high‐performance liquid chromatography/electrospray ionization multistage mass spectrometry—new biomarkers for biogeochemistry and microbial ecology. Rapid Communications in Mass Spectrometry 18, 617-628. Summons, R.E., Franzmann, P.D., Nichols, P.D., 1998. Carbon isotopic fractionation associated with methylotrophic methanogenesis. Organic Geochemistry 28, 465-475. Teske, A., Hinrichs, K.-U., Edgcomb, V., de Vera Gomez, A., Kysela, D., Sylva, S.P., Sogin, M.L., Jannasch, H.W., 2002. Microbial diversity of hydrothermal sediments in the Guaymas Basin: evidence for anaerobic methanotrophic communities. Applied and Environmental Microbiology 68, 1994-2007. Tornabene, T.G., Langworthy, T.A., 1979. Diphytanyl and dibiphytanyl glycerol ether lipids of methanogenic archaebacteria. Science 203, 51-53. Tribovillard, N., Bout-Roumazeilles, V., Algeo, T., Lyons, T.W., Sionneau, T., Montero-Serrano, J.C., Riboulleau, A., Baudin, F., 2008. Paleodepositional conditions in the Orca Basin as inferred from organic matter and trace metal contents. Marine Geology 254, 62-72. Tribovillard, N., Bout-Roumazeilles, V., Sionneau, T., Serrano, J.C.M., Riboulleau, A., Baudin, F., 2009. Does a strong pycnocline impact organic-matter preservation and accumulation in an anoxic setting? The case of the Orca Basin, Gulf of Mexico. Comptes Rendus Geoscience 341, 1-9. Turich, C., Freeman, K.H., 2011. Archaeal lipids record paleosalinity in hypersaline systems. Organic Geochemistry 42, 1147-1157. Turich, C., Freeman, K.H., Bruns, M.A., Conte, M., Jones, A.D., Wakeham, S.G., 2007. Lipids of marine Archaea: Patterns and provenance in the water-column and sediments. Geochimica et Cosmochimica Acta 71, 3272-3291. Valentine, D.L., Chidthaisong, A., Rice, A., Reeburgh, W.S., Tyler, S.C., 2004. Carbon and hydrogen isotope fractionation by moderately thermophilic methanogens. Geochimica et Cosmochimica Acta 68, 1571-1590. van der Maarel, M.J.E.C., Hansen, T.A., 1997. Dimethylsulfoniopropionate in anoxic intertidal sediments: a precursor of methanogenesis via dimethyl sulfide, methanethiol, and methiolpropionate. Marine Geology 137, 5-12. van der Wielen, P.W., Bolhuis, H., Borin, S., Daffonchio, D., Corselli, C., Giuliano, L., D'Auria, G., de Lange, G.J., Huebner, A., Varnavas, S.P., 2005. The enigma of prokaryotic life in deep hypersaline anoxic basins. Science 307, 121-123. Wagman, D.D., Evans, W.H., Parker, V.B., Schumm, R.H., Halow, I., 1982. The NBS tables of

128 Chapter 4

chemical thermodynamic properties. Selected values for inorganic and C1 and C2 organic substances in SI units. DTIC Document. Wakeham, S.G., Amann, R., Freeman, K.H., Hopmans, E.C., Jørgensen, B.B., Putnam, I.F., Schouten, S., Sinninghe Damsté, J.S., Talbot, H.M., Woebken, D., 2007. Microbial ecology of the stratified water column of the Black Sea as revealed by a comprehensive biomarker study. Organic Geochemistry 38, 2070-2097. Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161, 291-314. Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and

freshwater environments: CO2 reduction vs. acetate fermentation—Isotope evidence. Geochimica et Cosmochimica Acta 50, 693-709. Wiesenburg, D.A., Brooks, J.M., Bernard, B.B., 1985. Biogenic hydrocarbon gases and sulfate reduction in the Orca Basin brine. Geochimica et Cosmochimica Acta 49, 2069-2080. Xie, S., Lipp, J.S., Wegener, G., Ferdelman, T.G., Hinrichs, K.-U., 2013. Turnover of microbial lipids in the and growth of benthic archaeal populations. Proceedings of the National Academy of Sciences 110, 6010-6014. Zhilina, T.N., Zavarzin, G.A., 1990. Extremely halophilic, methylotrophic, anaerobic bacteria. FEMS Microbiology Letters 87, 315-322. Zhuang, G.-C., Lin, Y.-S., Elvert, M., Heuer, V.B., Hinrichs, K.-U., 2014. Gas chromatographic analysis of methanol and ethanol in marine sediment pore waters: Validation and implementation of three pretreatment techniques. Marine Chemistry 160, 82-90. Zhuang, G., Yang, G., Yu, J., Gao, Y., 2011. Production of DMS and DMSP in different physiological stages and salinity conditions in two marine algae. Chinese Journal of Oceanology and Limnology 29, 369-377.

129

130 Chapter 5

Chapter 5

Methanogenic substrates, activity and diversity in northwestern

Mediterranean Sea: Quantitative assessment of methanogenic

pathways in marine sediments

Guang-Chao Zhuang1,2,*, Verena B. Heuer1, Cassandre S. Lazar1,3, Tobias Goldhammer1, Vladimir Samarkin4, Jenny Wendt1, Samantha B. Joye4, Andreas P. Teske3, Kai-Uwe Hinrichs1

In preparation for Geochimica et Cosmochimica Acta.

1. MARUM Center for Marine Environmental Sciences, University of Bremen, 28359 Bremen, Germany 2. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, College of Chemistry and Chemical Engineering, Ocean University of China, 266100 Qingdao, China 3. Department of Marine Sciences, University of North Carolina at Chapel Hill, Chapel Hill, 27599 NC, USA 4. Department of Marine Sciences, University of Georgia, 30602 Athens, GA, USA *Corresponding author: Guang-Chao Zhuang ([email protected])

131 Chapter 5

5.1 Abstract

Microbial production of methane is an important terminal process during the organic matter degradation in marine sediments. It is generally thought that methane is predominantly produced from hydrogenotrophic and acetoclastic methanogenesis, while the importance of methylotrophic methanogenesis from methylated compounds remains largely unconstrained. To estimate the relative contribution of different substrates and pathways for methane production, we performed various biogeochemical, experimental and molecular genetic analyses to characterize substrates abundance, methanogenesis rates and methanogen communities in the sediments of Gulf of Lions, northwestern Mediterranean Sea. Methylated substrates such as methanol, trimethylamine (TMA), dimethylsulfide (DMS) were detected in the pore water and sediments, and showed vertical and spatial variability. In the sulfate-reducing sediments, the turnover time to methane for methanol (~44 days) and methylamine (~25 days) were much faster than acetate (> 105 days) and bicarbonate (> 106 days). Methanogenesis rates measurement revealed that methylotrophic methanogenesis and hydrogenotrophic methanogenesis contributed to methane production in the presence of sulfate, and methanogenesis from methanol could account for 13%-74% of the total methane production in the sulfate reduction zone. In the sulfate-depleted sediment, hydrogenotrophic methanogenesis dominated the methanogenic pathway (67%-97%), acetate could also make a significant contribution to methane production in organic-rich sediment (up to 31%), whereas the contribution of methylotrophic methanogenesis could be neglected in the methanogenic zone (< 1%). These results indicated that the activity of sulfate-reducing bacteria and organic matter could significantly influence the methanogenic pathways in marine sediments. Methanogenic communities based on methyl coenzyme M reductase gene (mcrA) analysis were generally in accordance with the distribution of methanogenic activity from different substrates. Our study quantitatively addressed the importance of methylotrophic methanogenesis in marine sediment, and provided new insights into relative

132 Chapter 5

contribution of methanogenic pathways for methane production.

5.2 Introduction

Marine sediment is the largest global reservoir of the potent greenhouse gas, methane, and most methane is generated from microbial methanogenesis (Milkov, 2004). During organic matter degradation, methane is produced by complex communities consisting of hydrolytic, fermentative, acetogenic and methanogenic microorganisms. As the ultimate process, methanogens produce methane from a number of substrates (e.g., bicarbonate, acetate, methylated compounds) that are generated by these complex communities (Conrad et al., 1989). Methanogens are strictly anaerobic archaea, which are sufficiently divergent to span many orders (e.g., Methanomicrobiales, Methanosarcinales and Methanocellales) (Garcia, et al, 2000; Sakai et al., 2008). In spite of the collective knowledge of specific processes and organisms that lead to methane production, and the significant quantities of methane found in marine sediments, relatively little is known of the pathways for methane production in sediments.

In marine sediment, methanogenesis from H2/CO2 and acetate is generally acknowledged as the major source of methane (Conrad, 2005), while the importance of methylotrophic methanogenesis from methylated compounds remains obscure. Early biogeochemical investigations (Oremland and Polcin, 1982) and more recent model (Mitterer, 2010) suggested that methylotrophic methanogenesis might account for the methane production in the presence of sulfate, as hydrogenotrophic and acetoclastic methanogens are unable to effectively compete with sulfate reducers for these substrates. Nevertheless, most observations of methylotrophic methanogenesis are based on in vitro studies from salt marsh, intertidal or hypersaline sediments (Parkes et al., 2012; King et al., 1983; Lazar et al., 2011b), where sulfate concentrations are high and methylated substrates such as methylamines are potentially abundant (King, 1988; Wang and Lee, 1994). In typical marine sediments, which have geochemical stratifications with a clear transition from sulfate to methane

133 Chapter 5

rich sediments (termed the sulfate methane transition zone, hereafter SMTZ), the relative importance of methylotrophic methanogenesis is unclear, in both the sulfate rich and sulfate depleted zones. Furthermore, limited information on the pool size and spatial distribution of methylated substrates in marine sediment makes it difficult to assess the methanogenic activities (e.g., methanogenesis rate measurement with radiotracers), though works reporting methane production quantification in marine sediments are sparse (Parkes et al., 2012; King et al., 1983). Modern cultivation independent genetic approaches have facilitated the detection of methanogens communities in various environments (Banning et al., 2005; Cadillo-Quiroz et al., 2006; Chan et al., 2005; Orphan et al., 2008; Parkes et al., 2010), providing important additional information about the metabolic potential for methanogenesis. The alpha subunit of methyl coenzyme M reductase (mcrA) is a highly conserved gene that catalyzes the final step of methanogenic pathway, and is a specific and useful functional gene marker for targeting methanogens and methanotrophs (Rudolf, 1998). With such molecular tools, depth related changes in methanogen community composition have been determined in marine sediment (Dhillon et al., 2005; Lazar et al., 2011a; Parkes et al., 2005). Nevertheless, few studies have addressed how methanogen community structure relates to the utilization of different substrates (Parkes et al., 2007; Webster et al., 2006; Yoshioka et al., 2010), in particular, those for methylated substrates in marine sediment (Parkes et al., 2012; Lazar et al., 2012). The objective of this study is to investigate the production of methane from different substrates and pathways, with the focus on methylotrophic methanogenesis, and further quantify the relative importance of different methanogenic pathways in marine sediments. Therefore, we conducted a suite of biogeochemical and molecular genetic analyses in marine sediments of northwestern Mediterranean Sea. The concentrations and distribution of potential methanogenic substrates, depth related methanogenic activities from different 14C-labeled substrates and the methanogen communities were collectively characterized in the sediment of Gulf of Lions. Additionally, these analyses were performed at geochemically distinct sites with

134 Chapter 5

geochemical zonations, which could help to better understand the influence of the environmental factors on methanogenic substrates and pathways.

5.3 Materials and methods 5.3.1 Site description and sampling protocol

Sediment cores were retrieved during expedition POS450 of the research vessel Poseidon, from Gulf of Lions, Mediterranean Sea in April, 2013 (Fig. 5.1).

Fig. 5.1 Sampling sites in the Gulf of Lions, northwestern Mediterranean Sea.

In the Gulf of Lions, a number of sites were sampled to track the terrestrial input from the Rhône River to the shelf (Fig. 5.1). Sites GeoB17306 and GeoB17307 were located in the pro-delta of the Rhône River and Sites GeoB17308 and GeoB17305 were off the river mouth (Heuer et al., 2014). The Rhône River is the major freshwater source in the Mediterranean Sea (Cauwet et al., 1990; Sempéré et al., 2000). As considerable riverine input of nutrients and wind-driven upwelling stimulate marine , the Rhône delta and the adjacent area is one of the most productive regions in the phosphorous-limited, oligotrophic Mediterranean Sea (Ludwig et al., 2009). High marine productivity and significant terrestrial input also influence the quantity and quality of the organic matter in marine sediment, thus

135 Chapter 5

the Gulf of Lions is ideal to investigate the generation and metabolism of low-molecular-weight compounds, and the influence of environmental factors on benthic communities and associated biogeochemical processes such as methanogenesis. During the cruise, pore water and sediments samples were collected from both multicorer (MUC) cores and gravity cores (Table 5.1).

Table 5.1 Locations and water depth of the sampled cores in Gulf of Lions. Station Longitude Latitude Water depth (Equipment) (m) Multicorer GeoB17305-3 43° 13.80' N 4° 30.61' E 61 GeoB17306-1 43° 18.95' N 4° 52.18' E 30 GeoB17307-5 43° 18.24' N 4° 51.53' E 52 GeoB17308-1 43° 16.20' N 4° 43.79' E 62

Gravity corer GeoB17306-2 43° 18.97' N 4° 52.16' E 30 GeoB17308-4 43° 16.21' N 4° 43.80' E 61

After retrieval, MUC cores were sectioned immediately at room temperature using a core extruder. Sediment samples for methane analysis were taken by transferring 2-3 mL sediment with a 5 mL cut-off syringe into a 22 mL serum vial, crimp-capped immediately, and stored at room temperature before onboard analysis. Another set of samples with addition of 5 mL 1M NaOH were stored in 22 mL serum vial at 4 °C for shore-based isotopic analysis of methane. Subsamples of sediments for total organic carbon (TOC), trimethylamine (TMA), dimethylsulfide (DMS) or dimethylsulfoniopropionate (DMSP) in the solid phase were collected using cut-off plastic syringes, transferred to headspace vials, and stored at -20 °C (Heuer et al., 2014). Pore water samples were extracted from sediment via Rhizon soil moisture samplers (0.1 µm porous polymer, Rhizosphere Research, Wageningen, the Netherlands) (Seeberg-Elverfeldt et al., 2005). Rhizons were rinsed with at least 30 mL of dilute hydrochloric acid (pH 1-2) followed by 30 mL of Milli-Q water before use. The collected pore waters were split into aliquots for analyses of sulfate, acetate,

136 Chapter 5

dissolved inorganic carbon (DIC), methanol, TMA, DMS and stored at -20 °C before analysis. Gravity cores were cut into 1 m segments on deck and pore water samples were extracted from holes drilled into the core liner in the same manner as from the MUC cores. Syringe samples for solid phase analysis (e.g., methane, hydrogen, TMA, DMS) were taken from freshly exposed sediments through small sampling ports (ca. 2 cm×3 cm), which were cut into the closed core liners. For hydrogen analysis, a sediment sample of 2-3 mL was extruded into a 12-mL headspace vial, immediately sealed with a thick black butyl stopper, crimp capped, and flushed with N2 (purity = 99.999%) for at least 1 min. Samples were stored at 4 °C for shore-based incubation and analysis (Heuer et al., 2014). For methanogenesis rate measurements, ~2 mL sediments were introduced into the modified Hungate tubes through the cut end, and a plunger modified from the butyl rubber stopper was added from the cut end to move the sediment to the threaded end of the tube (Bowles et al., 2011; Orcutt et al., 2005). Then the septum was placed onto the sediment without headspace, and the screw cap was tightened. The gas-tight sample tubes filled with sediments were stored at 4 °C before incubation. Quadruplicate samples (one control plus triplicate samples) for

- each substrate (HCO3 , acetate, methanol, methylamine) at various depths were taken from gravity cores (GeoB17306 and GeoB17308). Sediments for mcrA gene analysis were sampled with cut-off syringes, stored in sterile Falcon tubes, immediately frozen at -32 °C (i.e., the coldest temperature available), shipped on dry ice and stored at -80 °C in the laboratories on shore (Heuer et al., 2014).

5.3.2 Biogeochemical analyses

Concentrations of dissolved methane were measured on board by injecting 100-500 µL headspace gas into a gas chromatograph (Trace GC Ultra, ThermoFinnigan) equipped with a CP-PoraBOND Q (Varian Inc.) column and a flame ionization detector (FID) (Ertefai et al., 2010). Analysis of δ13C values of methane was performed on a gas chromatography combustion isotope ratio mass spectrometry (GC-C-IRMS) system combining a ThermoFinnigan Trace GC Ultra with a DELTA

137 Chapter 5

Plus XP mass spectrometer via a ThermoFinnigan GC Combustion III interface (Ertefai et al., 2010). Sulfate concentrations were determined by ion chromatography (Metrohm 861 Advanced Compact IC, column A Supp 5, conductivity detection after chemical suppression). Quantitative and isotopic analysis of volatile fatty acids (VFAs) in the pore water were performed with a liquid chromatography-isotope ratio mass spectrometry system, which consisted of a ThermoFinnigan Surveyor high performance liquid chromatography (HPLC) equipped with a VA Nucleogel Sugar 810-H column (300 × 7.8 mm; Macherey-Nagel, Germany) and a guard column of the same material linked to a DELTA Plus XP IRMS via a ThermoFinnigan LC Isolink interface (Heuer et al., 2006). Methanol in pore water was determined by a laboratory-built purge and trap pretreatment system connected to a Hewlett-Packard GC 5890 series II with a FID and Alltech Heliflex® AT-Q capillary column (30 m × 0.32 mm I.D.; Alltech Associates, Inc., Deerfield, US) (Zhuang et al., 2014). Dissolved pool of DMS and DMSP (DMS(P)d thereafter), and dissolved TMA in the pore water, and total base-hydrolyzable DMS and DMSP (DMS(P)t thereafter) and base-extractable TMA in the sediment were released by the addition of NaOH, preconcentrated with a purge and trap procedure, and analyzed on a Trace GC 2000 (Thermo Finnigan, Milan, Italy) coupled to a DSQ II MS (Thermo Finnigan, Texas, USA) with a Rtx®-Volatile Amine Column (30 m × 0.32 mm I.D., Restek GmbH, Homburg, Germany). Due to alkaline cleavage, DMSP (the precursor of DMS) could quantitatively convert to DMS (1:1) in our method. Hence, DMS(P)d refers to the combined pool of dissolved DMS and DMSP in the pore water, and DMS(P)t refers to the total base-hydrolyzable pool that includes DMS, DMS(P)d, and particulate DMSP in the sediment. Furthermore, to quantify exchangeable TMA that reversibly adsorb to particle exchange sites in the sediments (Lee and Olson, 1984; Wang and Lee, 1990), ~4 g wet sediments were extracted with 20 mL 1 M LiCl for 2 hours with continuous agitation. After centrifugation, the supernatant were treated as the same procedure for

13 TMA analysis in the pore water. Determination of TOC and δ CTOC were conducted on pre-weighed, freeze-dried, decarbonatized sediment using a ThermoScientific Flash 2000 elemental analyzer coupled to a ThermoScientific DELTA V Plus IRMS

138 Chapter 5

via a ThermoScientific ConFlo IV interface. DIC concentration and carbon isotopic composition was measured by injecting 1 mL headspace gas after acidification of pore

water with 0.2 mL of 0.1 M phosphoric acid (H3PO4) into a gas chromatography isotope ratio mass spectrometry consisting of a Hewlett-Packard 5890 GC equipped with a 6 m Poroplot Q column held at 35 °C and a Finnigan Mat DELTA S IRMS. Headspace hydrogen concentrations were measured with a Peak Performer 1 Reduced Gas Analyzer equipped with a mercury oxide detector (Peak Laboratories, USA) (Lin et al., 2012).

5.3.3 Methanogenic activity

Methanogenesis rates were measured using 14C-labeled bicarbonate, acetate

14 - (1,2- C-CH3COO ), and methanol. Relative turnover time to CH4 was determined using 14C-labeled methylamine, as in situ concentration data were not available. Sediment samples were injected with 100 μL of 14C-labeled bicarbonate (600 kBq), acetate (210 kBq), methanol (298 kBq) or methylamine (143 kBq) in batches of four (one control plus measurement in triplicate) and incubated at in situ temperature for 7 days. For controls, 2.5 mL of 2 M NaOH were injected into the sample and homogenized before isotope addition. Incubations were terminated by injecting 2.5 mL 2 M NaOH into the samples using a syringe and needle, and this was possible by the vacuum created by pulling the plunger to the base of the Hungate tube. The

14 methane production rate was determined by quantitatively converting CH4 in the

14 14 headspace to CO2 and trapping the evolved CO2. Briefly, vessels were shaken

vigorously and purged with a gentle flow of CO2 free compressed air. The gas stream

14 was passed through 2 M NaOH and pelletized NaOH to remove CO2 and then flowed through an 800 °C titanium-nickel alloy column packed with copper oxide

14 14 14 used to catalyze the oxidation of CH4 to CO2. The resulting CO2 was trapped in a mixture of Carbosorb E (Perkin Elmer) and the scintillation cocktail Permafluor E (Perkin Elmer). The efficiency of the trapping solution was >98%, tested by adding a

14 known activity of CO2. Immediately following purging and trapping, the cocktail

139 Chapter 5

was counted on a liquid scintillation counter. Rates were calculated using substrate

14 concentrations measured in separate samples, and activities recovered in CH4 pool (Eq. 5.1):

14 14 MOG rate = Csubstrate ×α/t (DPM- CH4/DPM- Csubstrate) (5.1) where MOG is the rate of methanogenesis from bicarbonate, acetate, and methanol

–3 –1 (nmol substrate reduced cm d ), Csubstrate is the concentration of different substrates (nmol cm–3), α is the isotopic fractionation factor (Assumed: bicarbonate: 1.04; acetate: 1.02; methanol: 1.07; methylamine: 1.06) (Krzycki et al., 1987; Summons et

14 al., 1998; Whiticar, 1999), t is the period of incubation (days), DPM- CH4 is the

14 activity recovered in the products pool (DPM), and DPM- Csubstrate is the activity of the substrate injected into sample (DPM). Relative turnover times for 14C-labeled substrates were calculated based on the time in days required to metabolize to

14 methane the total amount of added Csubstrate (Eq. 5.2):

14 14 Turnover time = DPM- Csubstrate/ (DPM- CH4/t) (5.2)

5.3.4 Molecular analysis

DNA was extracted from 10 g of sediment using the UltraClean® Mega Soil DNA Isolation Kit (MO BIO, CA). The mcrA genes were amplified by PCR using the ME1 (GCMATGCARATHGGWATGTC) and ME2 (TCATKGCRTAGTTDGGRTAGT) primers previously developed to specifically target methanogens (Hales et al., 1996). The PCR conditions were as follows: denaturation at 94 °C for 40 s, annealing at 50 °C for 1 min 30 s, and extension at 72 °C for 3 min for 30 cycles. PCR products were cloned using the TOPO XL cloning kit (Invitrogen, CA), and chemically transformed into Escherichia coli TOPO10 One Shot cells according to the manufacturer’s instructions. Sequences were obtained by Genewiz (South Plainfield, NJ) on an ABI Prism 3730xl sequencer using the M13R primer. Sequences were analyzed using the NCBI BLASTN search program within GeneBank (http://blast.ncbi.nlm.nih.gov/Blast) (Altschul et al., 1990).The mcrA

140 Chapter 5

sequences were translated into amino acid sequences using BioEdit and amino acid sequences were aligned using ClustalX (Larkin et al., 2007). Sequence data were analyzed with the MEGA4.0.2 program (Tamura et al., 2007). The phylogenetic trees were calculated by the neighbor-joining analysis. The robustness of inferred topology was tested by bootstrap resampling (1,000).

5.4 Results and discussion 5.4.1 Geochemistry of pore water and sediment from MUC cores

At Sites GeoB17306 and GeoB17307 in the pro-delta, sulfate was reduced with respect to depth, as evidenced by sulfate concentrations decreasing from 31 mM at the sediment-seawater interface to 25 mM and 10 mM, respectively (Fig. 5.2-A, B). Methane concentrations increased with depth, and reached 29 μM and 46 μM at 27.5 cm below seafloor (cmbsf), respectively (Fig. 5.2-A, B). In contrast, sulfate concentrations were generally constant with depth at Sites GeoB17308 and GeoB17305, and the concentrations of methane were consistently lower than 3 μM (Fig. 5.2-C, D). TOC content declined along the transect proceeding away from the Rhône River. The average TOC content was 1.13% at GeoB17306, and further from the river, decreased to 0.90% and 0.61% at GeoB17308 and GeoB17305, respectively (Fig. 5.2, E-H). The clear decrease in TOC reflected the reduced terrestrial organic matter input with increasing distance from the river mouth. The stable carbon isotopic value (δ13C) of TOC was -27.4‰ at GeoB17306, and became increasingly positive at GeoB17308 (-25.4‰) and GeoB17305 (-24.5‰) (Fig. 5.2, I-L), suggesting a gradual increase of the relative proportion of organic matter from marine productivity.

141 Chapter 5

Fig. 5.2 Depth profiles of methane and sulfate (A-D), TOC (E-H), and δ13C of TOC (I-L) in pore waters and sediments from retrieved MUC cores in the Gulf of Lions, Mediterranean Sea.

5.4.2 Distribution of potential methanogenic substrates in the pore water and sediments from MUC cores

Generally, methanol concentrations were below 2 μM in the sedimentary pore waters (Fig. 5.3, A-D). The vertical profiles displayed two different trends, 1) methanol was more concentrated in the uppermost sediment horizons and decreased slightly with depth, such as GeoB17308 (Fig. 5.3-C); 2) methanol concentrations were

142 Chapter 5 low in the surface sediment, increased with depth, and reached the maximum at depth between 8 cmbsf and 22 cmbsf (GeoB17305, GeoB17306, GeoB17307) (Fig. 5.3-A, B, D). The change of methanol concentrations with depth might reflect an in situ production and consumption imbalance, controlled by different microbial processes. For example, methanol could be derived from lignin and pectin in marine sediment (Donnelly and Dagley, 1980; Schink and Zeikus, 1980), and the decomposition of these relatively abundant and fresh organic matter could contribute to the accumulation of methanol in the surface sediment (e.g., GeoB17308) (Zhuang et al., 2014). On the other hand, due to the presence of oxygen in the bottom water, methanol could be oxidized aerobically by methylotrophic bacteria (Kolb, 2009), which could lead to the low concentrations of methanol in the upper sediments (e.g., GeoB17306, GeoB17307). As the degradation of organic matter persisted, methanol was released at higher levels in deeper sediment. Methanol peaked at shallower depth (7 cmbsf) at Site GeoB17307, while the maximum occurred at 22 cmbsf at Site GeoB17305, which might be related to the extent of organic matter mineralization. Once produced, methanol could be metabolized through various terminal processes such as nitrate reduction (Payne, 1973), iron reduction (Daniel et al., 1999), sulfate reduction (Cord-Ruwisch and Ollivier, 1986; Rabus et al., 2006), and methanogenesis (King et al., 1983), and these reactive processes in marine sediment could counteract methanol production and maintain methanol at lower concentrations in deeper sediments. Furthermore, methanol concentrations generally declined along the transect off the Rhône River, e.g., the average concentration of methanol at GeoB17306 in the pro-delta was three times that of Site GeoB17305 off the river mouth. Such a spatial variation indicated the influence of terrestrial organic matter input on the methanol distribution, as the main precursors of methanol such as lignin and pectin are largely terrigenous. Lignin is a major component of the woody tissues of vascular land plants and absent from marine organisms (Hedges et al., 1997; Hedges and Mann, 1979), and pectin is a component of photosynthetic biomass found in the plant and algal cell walls (Schink and Zeikus, 1982). The reduced terrestrial organic matter

143 Chapter 5 supply along the transect from pro-delta to the shelf could result in a decreasing abundance of lignin and pectin in sediment, thus influence the spatial distribution of methanol.

Fig. 5.3 Depth profiles of methanol and ethanol (A-D), exchangeable TMA (E-H), base-extractable TMA (I-J) and base-hydrolyzable DMS(P)t (M-P) in pore waters and sediments from retrieved MUC cores in Gulf of Lions, Mediterranean Sea. Note that exchangeable TMA, base-extractable TMA and base-hydrolyzable DMS(P) was expressed as μmol per kg wet sediment.

144 Chapter 5

In comparison to methanol, ethanol concentrations were lower and generally less than 0.5 μM in the pore waters (Fig. 5.3, A-D). As an important fermentation product, ethanol is mainly generated from carbohydrates under anoxic conditions (Eichler and Schink, 1984). Nevertheless, the formation of reduced fermentation products such as ethanol is energetically less favorable than the production of acetate with concomitant transfer of hydrogen to hydrogenotrophic bacteria (Schink et al., 1985). Consequently, the fermentative ethanol production is presumed to be less preferred under substrate-limiting conditions in natural habitats, which lead to the low abundance of ethanol in the pore waters. In addition, the anaerobic degradation of ethanol could further reduce the ethanol pool size. In the presence of sulfate, ethanol could be oxidized to acetate by sulfate-reducing bacteria with concomitant reduction of sulfate to sulfide (Schink et al., 1985). At low sulfate conditions, ethanol oxidation could be coupled with interspecies hydrogen transfer to methanogens by some sulfate reducers (Bryant et al., 1977). Some homoacetogenic bacteria could combine ethanol oxidation to acetate with reductive acetate formation from carbon dioxide (Eichler and Schink, 1984). The thermophilic marine methanogen isolate organophilum can also directly utilize ethanol (Frimmer and Widdel, 1989). Thus, limited ethanol production during fermentation and multiple putative degradations could explain the low concentrations of ethanol observed in the pore waters. TMA was barely detectable at most depths in the pore waters (< 20 nM), whereas the pool sizes in sediment were relatively large (Fig. 5.3, E-L). The concentration of exchangeable TMA was between 0.4 and 7.0 μmol kg-1 (Fig. 5.3, E-H), and the base-extractable TMA was more abundant (0.7-11 μmol kg-1) in the sediments (Fig. 5.3, I-L). The difference between the dissolved pool of TMA and solid phase could be attributed to the strong adsorption of TMA on the sediments. Amine compounds can be reversibly adsorbed to particle exchange sites as cations, and this adsorption could significantly influence their distribution in sediments (Fitzsimons et al., 2006; Wang and Lee, 1990, 1993). Although TMA was absent in the pore water, the abundant exchangeable TMA (extracted with LiCl to compete for adsorption), indicated the high potential of bioavailable TMA that could be utilized

145 Chapter 5

for microbial processes in the sediments. Furthermore, exchangeable TMA generally comprised 40%-70% of the total base-extractable pool that included dissolved, exchangeable and bound TMA in the sediments, suggesting a substantial fraction of TMA is sorbed reversibly to the sediment. Significant correlations were also observed between the exchangeable TMA and the base-extractable TMA at all study sites (R2 = 0.68-0.99). Likewise, the combined pool of dissolved DMS and DMSP in the pore waters was small (<4 nM), and accumulated in the upper 10 cmbsf. The concentrations of base-hydrolyzable DMSP in the sediments were about three orders of magnitude higher, ranging from 1.5 and 13.4 μmol kg-1 (Fig. 5.3, M-P). Generally, TMA and DMS accumulated in the surface sediment and decreased with depth. In marine environment, TMA and DMS could be derived from a number of potential precursors (e.g., choline, glycine betaine, DMSP), which are largely formed within phytoplankton (i.e., functioning as osmolyte). Upon deposition at the seafloor (Belviso et al., 2006; Keller et al., 1999), these precursors can be degraded and lead to the accumulation of TMA and DMS in the uppermost sediment horizons. After formation within sediments, TMA and DMS constitute highly energetic substrates for methanogenesis or non-methanogenic processes (Kiene et al., 1986; King, 1984; Oremland and Polcin, 1982; van der Maarel and Hansen, 1997). Due to the microbial consumption and rapid turnover, the concentrations of TMA and DMS in the pore waters were maintained at a low level. Furthermore, no clear variations for TMA and DMS(P) (i.e., DMS and DMSP) were observed among different sites, indicating the weak influence of terrestrial input on their spatial distribution. In contrast to methanol, the precursors of TMA and DMS have a marine origin, and the production of TMA and DMS could be related to marine productivity and associated biogeochemical processes. The concentrations of chlorophyll were comparable among sites GeoB17305, GeoB17308, GeoB17307 (~1.1 mg m-3) (Heuer et al., 2014), suggesting the riverine nutrients input did not significantly influence the phytoplankton abundance thus the distribution of TMA and DMS(P).

146 Chapter 5

5.4.3 Methanogenic activities

Methanogenesis from major substrates representing different pathways were assessed with sediment collected from gravity cores at Geo17306 and GeoB17308. At GeoB17306, intensive sulfate reduction was concentrated in the top 100 cmbsf, and sulfate concentrations decreased from 30 mM to < 1mM at this interval (Fig. 5.4-A). Methane concentrations were generally low (< 0.1 mM) in the sulfate rich sediment, increased rapidly to ~5.3 mM at 105 cmbsf and remained high in the deeper layer of the 500 cm core (Fig. 5.4-A). Between 80-100 cmbsf, a discrete SMTZ significantly regulated the methane flux from the underlying methanogenic zone. The δ13C values of methane varied slightly between -73.1‰ and -80.8‰ below 100 cmbsf, and shifted to -62.3‰ at 81.5 cmbsf, due to the rigorous oxidation of methane at the SMTZ (Fig. 5.4-B).

Fig. 5.4 Depth profiles of methane and sulfate (A), δ13C of methane (B), DIC and δ13C of DIC (C), hydrogen (D), acetate and δ13C of acetate (E), methanol and ethanol (F), exchangeable TMA and base-extractable TMA (G) and base-hydrolyzable DMS(P)t (H) in pore waters and sediment from retrieved GC core at Site GeoB17306.

147 Chapter 5

Hydrogen concentrations determined with headspace equilibration method varied between 0.18 and 0.57 nM (Fig. 5.4-D). The δ13C values of DIC positively increased from -23.7‰ at 50 cmbsf to 3.8‰ at 500 cmbsf (Fig. 5.4-C), suggesting the increasing methane production from carbonate reduction. Acetate concentrations were below 8 μM, and the δ13C values of acetate were between -29.8‰ to -32.4‰ at several depths with concentrations higher than 5 μM that allowed for isotopic analysis (Fig. 5.4-E). For methylated substrates, methanol concentrations were generally lower than 2 μM (Fig. 5.4-F), and TMA and DMS(P)t were relatively abundant in the solid phase of sediments (Fig. 5.4-G, H). In the surface sediment at GeoB17306, methanogenic activity from bicarbonate and acetate were low, whereas the turnover of methanol and methylamine were much faster (Fig. 5.5, A-D). The average turnover time of methanol and methylamine in the upper 70 cmbsf was 44 and 25 days, respectively (Fig. 5.5-C, D), about ~3-5 orders of magnitude faster than bicarbonate (1.2×106 days) and acetate (3.1×104 days) (Fig. 5.5-A, B). In contrast to previous studies, the turnover of methanol or methylamine was much faster than those from gas hydrate-bearing, hypersaline, or salt marsh creek sediments (Parkes et al., 2012; Lazar et al., 2011b; Yoshioka et al., 2009). Methane production rates from methanol and bicarbonate were comparable (~0.02 nmol cm-3 d-1), both of which were about 100 times of acetoclastic methanogenesis (Fig. 5.5, A-C). Below 90 cmbsf, methanol and methylamine turnover stabilized and the turnover time were similar to those in the surface sediment. In contrast, the turnover of bicarbonate and acetate became much more rapid, in particular for acetate, the turnover time reduced considerably to 43 days. Changes in turnover time to methane for different substrates reflected the methane production rates. Methanogenesis rates

-3 -1 from H2/CO2 increased to a maximum of 1.83 nmol cm d at 120 cmbsf, remained at the elevated rates to the base of the core. The average of acetate methanogenesis rate was 0.14 nmol cm-3 d-1 in the methanogenic zone, which was thousand times higher than those in surface sediments. Both hydrogenotrophic and acetoclastic rates were in agreement with the wide ranges of rates measured in various marine sediment settings

-3 including seep or non-seep sediments (H2/CO2 methanogenesis: < 0.01-30 nmol cm

148 Chapter 5

d-1, acetate methanogenesis: < 0.001-60 nmol cm-3 d-1) (O'Sullivan et al., 2013; Orcutt et al., 2005; Parkes et al., 2007).

Fig. 5.5 Depth distributions of turnover times and rates of methanogenic substrates at Site GeoB17306. SMTZ was marked with gray bars.

In comparison to Site GeoB17306, sulfate depleted from 31 mM to 1 mM at 275 cmbsf at GeoB17308, and methane accumulated to 1 mM below the SMTZ, with negative isotopic values between -89.0‰ and -97.0‰ (Fig. 5.6-A, B). DIC concentrations were much lower than GeoB17306 (Fig. 5.6-C), indicating the lower mineralization rate at GeoB17308. Likewise, the abundance of methanol were much less than GeoB17306 (Fig. 5.6-F). Acetate concentrations increased slightly to 5 μM at 345 cmbsf, with the δ13C value of -27.2‰ (Fig. 5.6-E). The turnover of bicarbonate was relatively fast in the upper 75 cmbsf (3.1×105 days), increased to 1.4×107 days in the deeper sulfate reduction zone at 200 cmbsf, and decreased progressively to 8.6×105 days at 350 cmbsf in the methanogenic zone (Fig. 5.7-A). The turnover of acetate was fastest at 75 cmbsf (6724 days), and reduced by a factor of 10 in the deeper sediment (Fig. 5.7-B). Methanol and methylamine displayed a parallel trend with increasing turnover time along depth (Fig. 5.7-C, D). High methanogenic activities were detected for methanol (31 days) and methylamine (30 days) in the upper sediment. The turnover time of these methylated substrates increased markedly with depth, and reached the maximum of 7.4×104 and 1.1×105 days at 350 cmbsf. Consistent with the turnover rate, methanogenesis rates from

149 Chapter 5 bicarbonate, acetate and methanol were relatively high in the surface sediment, and generally decreased with depth (Fig. 5.7, A-C).

Fig. 5.6 Depth profiles of methane and sulfate (A), δ13C of methane (B), DIC and δ13C of DIC (C), hydrogen (D), acetate and δ13C of acetate (E), methanol and ethanol (F), exchangeable TMA and base-extractable TMA (G) and base-hydrolyzable DMS(P)t (H) in pore waters and sediment from retrieved GC core at Site GeoB17308.

Fig. 5.7 Depth distributions of turnover times and rates of methanogenic substrates at Site GeoB17308. SMTZ was marked with gray bars.

150 Chapter 5

Geochemical analysis on methane and precursors, and 14C activity measurements indicated methanogenesis was an important process at Site GeoB17306. In the upper sediment, methane occurred at low concentrations, and methylotrophic methanogenesis was important methanogenic pathway in the presence of high levels of sulfate. Although it was generally thought that sulfate reducer thermodynamically outcompete with methanogen for hydrogen, and the concentrations of hydrogen (< 1 nM) was also not sufficient to support methane production (Lovley and Goodwin, 1988), hydrogenotrophic methanogenesis occurred in the sulfate-reducing sediments. The incomplete inhibition of hydrogenotrophic methanogenesis could be explained that methane was formed in organic rich microenvironments such as the interior portions of decaying organisms or particulate organic matter, while at the same time sulfate reduction occurred in the surrounding sediment matrix (Martens and Berner, 1974). Furthermore, the isotopic values of methane also suggested the in situ production of methane in the sulfate reduction zone. Escaping rigorous AOM at SMTZ, the upwards diffusive methane should be progressively 13C-enriched in the sulfate reduction zone (Whiticar, 1999). In our case, the δ13C values of methane shifted from -80.8‰ below SMTZ to -62.3‰ at 81.5 cmbsf, but then negatively increased to -69.4‰ at 13.5 cmbsf. This change strongly suggested that highly 13C-depleted methane from other source counteracted the enrichment of 13C in

methane caused by AOM. Accordingly, methanogenesis from H2/CO2 or methanol could produce the highly 13C-depleted methane due to the large fractionation between substrates and methane (Whiticar, 1999), thus the occurrence of methane in the sulfate reduction zone could be a collective result of the upwards diffusion, AOM and in situ

production from methylated substrates or H2/CO2. Consistent with the rate measurement, acetoclastic methanogenesis was unlikely to occur as acetate could produce methane with δ13C values between -65‰ to -50‰ (Whiticar et al., 1986), which was not depleted enough to balance the enriched isotopes due to AOM. Until sulfate was exhausted, acetate and hydrogen became available for methanogens, and methane were largely produced from H2/CO2 or to a less extent acetate. For example,

-3 -1 highest methanogenic rate was from H2/CO2 (1.83 nmol cm d ) at 120 cmbsf,

151 Chapter 5 followed by acetate methanogenesis with a rate of 0.06 nmol cm-3 d-1, and the least was from methanol (0.008 nmol cm-3 d-1). In agreement with rate measurements, both the negative δ13C values of methane (~-75.4‰) and the increasingly enriched 13C of DIC suggested the dominance of hydrogenotrophic methanogenesis below SMTZ. In contrast to GeoB17306, SMTZ occurred at a greater depth between 270 cmbsf and 290 cmbsf at GeoB17308, suggesting the lower mineralization rate and microbial activity. Accordingly, much lower methanogenic activities were detected at this site except in the upper 75 cmbsf. The methane production rate from methanol was 0.007

-3 -1 nmol cm d at 25 cmbsf, which was slightly higher than H2/CO2 methanogenesis (0.004 nmol cm-3 d-1). At deeper depths within sulfate reduction zone, methanogenesis rate from H2/CO2 exceeded that from methanol, and hydrogenotrophic methanogenesis became the dominant methanogenic pathway. Below SMTZ, hydrogenotrophic methanogenesis was predominant, and highly 13C depleted methane (-93‰) were produced in the methanogenic zone. Although the acetate was also present at GeoB17306, acetoclastic methanogenesis was extremely low along the entirety core (< 4×10-4 nmol cm-3 d-1). By comparison, methanogenesis rates in the methanogenic zone at GeoB17308 were lower by 2-3 orders of magnitude than Geo17306, suggesting that methanogenesis might be less important at GeoB17308. Surprisingly, both hydrogenotrophic and acetoclastic methanogenesis rates were much higher in the surface sulfate-reducing sediment than those in the methanogenic zone, indicating that the activity of sulfate-reducing bacteria might be not the limited factor for methane production from competitive substrates at this site.

5.4.4 Methanogenic diversity

Molecular analysis with mcrA genes revealed diverse methanogen populations and marked changes within different geochemical zonations. At 20-45 cmbsf of GeoB17306, sequences were affiliated with Methanomicrobiales (49%), Methanosarcinales (17%) as well as ANMEs (34%, i.e., ANME-3, ANME-2a) (Fig. 5.8). Generally, species from Methanomicrobiales are strictly hydrogenotrophic

152 Chapter 5

methanogens. Among the order of Methanosarcinales, a large number of sequences were related to Methanosarcina, which are capable of metabolizing the major substrates from three methanogenic pathways (Table 5.2). In contrast, the genera of Methanococcoides and Methanolobus are obligately methylotrophic methanogens, and they exclusively grow on methylated compounds (Table 5.2). Hence, the detection of hydrogenotrophic and methylotrophic methanogens correlated well with

the occurrence of methane production from H2/CO2 and methylated compounds in the sulfate reduction zone. Furthermore, the abundance of ANMEs at this depth interval indicated AOM potentially occurred in the sulfate reduction zone. At 75-90 cmbsf, sequences closely related to ANME-2a dominated the retrieved sequences (55%; Fig. 5.8), which could be related to the intensive methane oxidation at the SMTZ (~90 cmbsf). The majority of methanogens belonged to Methanomicrobiales (38%, Fig. 5.8), and a few sequences were affiliated with the genus of Methanosarcina (Table 5.2). Interestingly, a new group of Methanoplasmatales was also detected at this depth. This group was recently discovered in termite guts and has been proposed as the seventh order of methanogen (Paul et al., 2012). Cultured or enriched members from this order could only metabolize methanol with hydrogen to produce methane (Dridi et al., 2012). The presence of Methanoplasmatales in marine sediments, suggested that the new order might be more widespread in natural environment than previously thought. In the deeper sediment between 379-395 cmbsf, Methanomicrobiales still dominated the methanogen communities (48%), while Methanosarcinales (mainly Methanosarcina) were slightly more abundant in relative proportion than the sulfate reduction zone (23%, Fig. 5.8). Since higher activity from acetoclastic methanogenesis has been observed in the methanogenic zone, the increasing abundance of Methanosarcina suggested acetate might be the major substrates for this versatile genus of methanogen. Additionally, the occurrence of ANME at this depth was not puzzling, as the scatted methane profile might be an indication of AOM in the methanogenic zone. Treude et al. (2014) also reported active AOM in the methanogenic zone of Alaskan-Beaufort continental margin sediments.

153 Chapter 5

Table 5.2 Methanogen communities based on mcrA genes analysis and potential substrates utilization from Mediterranean Sea sediments. Station Methaneogen mcrA gene Sequence Potential substrate utilization (Depth) orders clone libraries number (References) GeoB17306 1 20-45 cm Methanomicrobiales Methanogenium marinum 1 H2/CO2, formate 2,3 1 H2/CO2, formate 4 Methanoplanus limnicola 2 H2/CO2, formate 2,3 Methanomicrobiales 36 H2/CO2, formate

Methanosarcinales Methanosarcina mazei 11 H2/CO2, acetate, methylamines, methanol3 Methanolobus bombayensis 2 Methylamines, methanol, dimethylsulfide5 Methanococcoides 2 Methylamines, methanol, N,N-Dimethylethanolamine, betaine, choline3,6,7,8, ANME-3 8 ANME-2a 20 9 75-90 cm Methanoplasmatales Methanoplasmatales 2 Methanol, methanol/H2 Methanomicrobiales Methanomicrobiales 33 Refer to above. Methanosarcinales Methanosarcina mazei 2 Refer to above.

Methanosarcina 2 H2/CO2, acetate, methanol, methylamines, methanethiol, dimethylsulfide3, 10 ANME-2a 47 379 -395 cm Methanomicrobiales Methanomicrobiales 44 Refer to above. Methanosarcinales Methanosarcina mazei 13 Refer to above. Methanosarcina 8 Refer to above. ANME-2a 27 GeoB17308 3 225-240 cm Methanobacteriales Methanobrevibacter 2 H2/CO2, formate Methanomicrobiales Methanomicrobiales 14 Refer to above. 11 Methanocellales Methanocella paludicola 1 H2/CO2, formate 12,13 Methanocella 15 H2/CO2, formate Methanosarcinales Methanosarcina mazei 2 Refer to above. Methanosarcina 9 Refer to above. References: 1. (Chong et al., 2002); 2. (Liu and Whitman, 2008); 3. (Garcia et al., 2000); 4. (Wildgruber et al., 1982); 5. (Kadam et al., 1994); 6. (Watkins et al., 2014); 7. (Watkins et al., 2012); 8. (L'Haridon et al., 2014); 9. (Paul et al., 2012); 10. (Galagan et al., 2002); 11. (Sakai et al., 2008). 12. (Sakai et al., 2010); 13. (Lü and Lu, 2012).

Unfortunately, the amplification of mcrA gene was only successful between 225-240 cmbsf at GeoB17308, probably due to the low abundance of methanogens at

154 Chapter 5

this site. Compared to GeoB17306, the methanogen communities were more diverse, with additional genus recovered from orders of Methanocellales and Methanobacteriales. Sequences from Methanomicrobiales, Methanocellales (e.g., Methanocella) and Methanosarcinales (e.g., Methanosarcina) accounted for 33%, 37%, and 26% of the assemblages, respectively, and most of the species from these orders could reduce bicarbonate with the aid of hydrogen to produce methane. Accordingly, hydrogenotrophic methanogenesis was the main methanogenic pathway at this depth interval, and the activity was 100 times higher than methanogenesis from acetate and methanol. Thus, the indications from mcrA gene analysis were closely related to the geochemical profiles and depth distribution of methanogenic activities at both sites, providing additional information for the metabolism of different substrates by methanogens.

Fig. 5.8 Depth distributions of relative abundance of methanogen populations at Site GeoB17306 and GeoB17308 based on mcrA genes clone libraries.

5.4.5 Relative importance of different methanogenic pathways and potential influence factors in marine sediment

Based on the integrated observations from geochemistry, methanogenic activity,

155 Chapter 5

and communities in the sediment of northwestern Mediterranean Sea, we could quantitatively estimate the relative importance of methanogenic pathways. The distinct biogeochemistry and methanogenic capabilities observed at Sites GeoB17306 and GeoB17308, also made it possible to compare the importance of methanogenic pathways and the factors influencing methanogenic activity in different biogeosystems. Site GeoB17306 could be a representative of a biogeosystem with relatively high TOC, high microbial activity and abundant populations. In this system, methanogenesis is an important terminal process, and methane is produced with high rates. In the sulfate reduction zone, the availability and rapid turnover of methylated compounds lead to the significant methane production via methylotrophic methanogenesis. For example, methanol could account for 43%-74% of total measured methane production (Fig. 5.9). Due to the competition from sulfate reducers, the contribution of acetate for methane production could be neglected in the presence of sulfate (e.g., < 1%), although acetoclastic methanogenesis was detected at a low rate. Hydrogenotrophic methanogenesis was not completely inhibited, and H2/CO2 could make a significant contribution to methane production (e.g., 26%-57%). Below the sulfate reduction zone, methane was clearly produced, and hydrogenotrophic methanogenesis became the dominant methanogenic pathway with the depletion of sulfate (e.g., 67%-97%). Acetoclastic methanogenesis also could be active, and its contribution for methane production could increase, e.g., a maximum of 31% of at 480 cmbsf. In contrast, the role of methylotrophic methanogenesis was largely limited, and much less methane could be produced from methylated substrates in the methanogenic zone (e.g., < 1%). In contrast, Site GeoB17308 was a biogeosystem characterized by lower TOC content, lower microbial activity and limited microbial communities. In this regarding, methanogenesis is less robust and methane was produced at much lower rates. In comparison to Site GeoB17306, methylated compounds were still important methanogenic substrates in the sulfate reduction zone, while their contribution for methane production decreases due to the lower availability of substrates (methanol

156 Chapter 5

methanogenesis: 13%-60%). At the SMTZ and within the methanogenic zone,

methane was predominantly produced from H2/CO2 (e.g., 88%-97%). A minor fraction of methane could be produced from acetate in both sulfate reduction and methanogenic zones (generally < 3%), indicating the insignificance of acetoclastic methanogenesis in lower TOC sediment. From the geochemical and methanogenic variability in these different systems, it could be inferred that the most important control for methane production might be the activity of sulfate-reducing bacteria and the supply of organic matter, given the similarity of other environmental factors such as pH, salinity, temperature. In both biogeosystems, the quantitative importance of methylotrophic methanogenesis was particularly highlighted in the presence of sulfate reducers, as its contribution could be up to 74% in the surface sulfate-reducing sediment but less than 1% in the deeper methanogenic zone. Similar to our observations, King et al. (1983) reported that 35.1-61.1% of methane production in slurries of sulfate rich intertidal sediments was from TMA, while methylamines and methanol only accounted for 1% and 5% of total methane production in the low sulfate lake sediments (Lovley and Klug, 1983). Furthermore, the role of acetate as methanogenic substrate also hinged on the thermodynamic control from sulfate-reducing bacteria in system with high TOC input, reflected from the different contributions of acetoclastic methanogenesis in the sulfate reduction and methanogenic zones at GeoB17306. Nevertheless, the activity of sulfate reducers may be not the major factor at GeoB17308, as methanogenic rates from acetate were much higher in the sulfate reduction zone than methanogenic zone. Thus, the importance of acetoclastic methanogenesis could also be a function of organic matter lability or supply rates. For example, acetate could contribute to a significant amount of methane production in the methanogenic zone at GeoB17306 (up to 31%), whereas the contribution was negligible at GeoB17308 (generally < 3%). Previous study also showed that acetate could account for 20-29% of the total methane production in organic-rich sediment from Cape Lookout Bight (Crill and Martens,

1986). Under both circumstances, although methane production from H2/CO2 was also affected by the activity of sulfate reducers, hydrogenotrophic methanogenesis not

157 Chapter 5 only significantly contributed to methane production in the sulfate-reducing sediment, but was the predominant pathway in methanogenic zone, indicating the importance of hydrogenotrophic methanogenesis in marine sediment.

Fig. 5.9 Schematic illustrating the relative importance of different methanogenic pathways in marine sediments at Site GeoB17306 and Site GeoB17308, Mediterranean Sea. The relative importance of methanogenic pathways were estimated with methanogenesis rates from bicarbonate, acetate and methanol.

Furthermore, as methane production was detected in the presence of sulfate in both these biogeosystems, our results also have global implications for methane budget. In the surface sediment, the endogenous methane from hydrogenotrophic and methylotrophic methanogenesis could lead to methane escaping to the water column

158 Chapter 5

before being oxidized, which could be further emitted to the atmosphere through air-sea exchange. The methylated compounds, relatively abundant in marine environment, could also stimulate the methane production in anoxic microniche in the ocean. Hence, the surface sediment and overlying water column could be a likely source of atmospheric methane due to the occurrence of hydrogenotrophic or methylotrophic methanogenesis.

5.5 Conclusion

In the present study, integrated investigations on methanogenic substrates, activity and diversity were performed to quantitatively estimate the relative contribution of different methanogenic pathways in marine sediment, northwestern Mediterranean Sea. As potential non-competitive substrates, methylated substrates such as methanol, TMA and DMS were present in the pore water and sediments. The spatial and vertical distributions of these substrates were either influenced by terrestrial organic matter input or regulated by microbial production or consumption

processes. Methane production from H2/CO2, acetate, methanol and methylamine was all detected along the sediment cores with radiotracer experiment, and relative activity could be influenced by environmental factor such as organic matter reactivity and the activity of sulfate-reducing bacteria. In the sulfate-reducing sediment, methylotrophic methanogenesis contributed to 13%-74% of the total methane production. Hydrogenotrophic methanogenesis dominated the methane production with the depletion of sulfate (67%-97%), and acetate could contribute to a significant amount of methane production in organic-rich sediment. Methanogenic community diversity correlated well with the geochemical profiles and the distribution of methanogenic rates from different pathways. Our study highlights the importance of methylotrophic methanogenesis in the sulfate-reducing sediments, and provides the quantitative assessment on the relative contribution of three different methanogenic pathways for methane production.

159 Chapter 5

5.6 Acknowledgements

We thank the captains, crew as well as the scientists of RV Poseidon cruise POS450 for their kind assistance during sampling. We are grateful for M. Bowles for the valuable comments on the manuscript. We also thank J. Arndt for the DIC measurements. The study was supported by European Research Council (ERC grant 247153, DARCLIFE; PI: K.-U. Hinrichs) and the Deutsche Forschungsgemeinschaft (DFG) through the Research Center/Excellence Cluster MARUM-Center for Marine Environmental Sciences. G. C. Zhuang was jointly sponsored by the Chinese Scholarship Council (CSC) and the Bremen International Graduate School for Marine Sciences (GLOMAR).

5.7 References

Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool. Journal of Molecular Biology 215, 403-410. Banning, N., Brock, F., Fry, J.C., Parkes, R.J., Hornibrook, E.R.C., Weightman, A.J., 2005. Investigation of the methanogen population structure and activity in a brackish lake sediment. Environmental Microbiology 7, 947-960. Belviso, S., Thouzeau, G., Schmidt, S., Reigstad, M., Wassmann, P., Arashkevich, E., Stefels, J., 2006. Significance of vertical flux as a sink for surface water DMSP and as a source for the sediment surface in coastal zones of northern Europe. Estuarine, Coastal and Shelf Science 68, 473-488. Bowles, M.W., Samarkin, V.A., Joye, S.B., 2011. Improved measurement of microbial activity in deep-sea sediments at in situ pressure and methane concentration. Limnology and Oceanography: Methods 9, 499-506. Bryant, M., Campbell, L.L., Reddy, C., Crabill, M., 1977. Growth of Desulfovibrio in lactate or

ethanol media low in sulfate in association with H2-utilizing methanogenic bacteria. Applied and Environmental Microbiology 33, 1162-1169. Cadillo‐Quiroz, H., Bräuer, S., Yashiro, E., Sun, C., Yavitt, J., Zinder, S., 2006. Vertical profiles of methanogenesis and methanogens in two contrasting acidic peatlands in central New York State, USA. Environmental Microbiology 8, 1428-1440.

160 Chapter 5

Cauwet, G., Gadel, F., de Souza Sierra, M., Donard, O., Ewald, M., 1990. Contribution of the Rhône River to organic carbon inputs to the northwestern Mediterranean Sea. Continental Shelf Research 10, 1025-1037. Chan, O.C., Claus, P., Casper, P., Ulrich, A., Lueders, T., Conrad, R., 2005. Vertical distribution of structure and function of the methanogenic archaeal community in Lake Dagow sediment. Environmental Microbiology 7, 1139-1149. Chong, S.C., Liu, Y., Cummins, M., Valentine, D.L., Boone, D.R., 2002. Methanogenium

marinum sp. nov., a H2-using methanogen from Skan Bay, Alaska, and kinetics of H2 utilization. Antonie van Leeuwenhoek 81, 263-270. Conrad, R., 2005. Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Organic Geochemistry 36, 739-752. Conrad, R., Andreae, M., Schimel, D., 1989. Control of methane production in terrestrial ecosystems. Exchange of Trace Gases between Terrestrial Ecosystems and the Atmosphere., 39-58. Cord-Ruwisch, R., Ollivier, B., 1986. Interspecific hydrogen transfer during methanol degradation by Sporomusa acidovorans and hydrogenophilic anaerobes. Archives of Microbiology 144, 163-165. Crill, P.M., Martens, C.S., 1986. Methane production from bicarbonate and acetate in an anoxic marine sediment. Geochimica et Cosmochimica Acta 50, 2089-2097. Daniel, R., Warnecke, F., Potekhina, J.S., Gottschalk, G., 1999. Identification of the syntrophic partners in a coculture coupling anaerobic methanol oxidation to Fe(III) reduction. FEMS Microbiology Letters 180, 197-203. Dhillon, A., Lever, M., Lloyd, K.G., Albert, D.B., Sogin, M.L., Teske, A., 2005. Methanogen diversity evidenced by molecular characterization of methyl coenzyme M reductase A (mcrA) genes in hydrothermal sediments of the Guaymas Basin. Applied and Environmental Microbiology 71, 4592-4601. Donnelly, M., Dagley, S., 1980. Production of methanol from aromatic acids by Pseudomonas putida. Journal of Bacteriology 142, 916-924. Dridi, B., Fardeau, M.-L., Ollivier, B., Raoult, D., Drancourt, M., 2012. Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon isolated from human faeces. International Journal of Systematic and Evolutionary Microbiology 62, 1902-1907. Eichler, B., Schink, B., 1984. Oxidation of primary aliphatic alcohols by Acetobacterium carbinolicum sp. nov., a homoacetogenic anaerobe. Archives of Microbiology 140, 147-152. Ertefai, T.F., Heuer, V.B., Prieto-Mollar, X., Vogt, C., Sylva, S.P., Seewald, J., Hinrichs, K.-U., 2010. The biogeochemistry of sorbed methane in marine sediments. Geochimica et

161 Chapter 5

Cosmochimica Acta 74, 6033-6048. Fitzsimons, M.F., Millward, G.E., Revitt, D.M., Dawit, M.D., 2006. Desorption kinetics of ammonium and methylamines from estuarine sediments: consequences for the cycling of nitrogen. Marine Chemistry 101, 12-26. Frimmer, U., Widdel, F., 1989. Oxidation of ethanol by methanogenic bacteria. Archives of Microbiology 152, 479-483. Galagan, J.E., Nusbaum, C., Roy, A., Endrizzi, M.G., Macdonald, P., FitzHugh, W., Calvo, S., Engels, R., Smirnov, S., Atnoor, D., 2002. The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Research 12, 532-542. Garcia, J.-L., Patel, B.K., Ollivier, B., 2000. Taxonomic, phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe 6, 205-226. Hales, B.A., Edwards, C., Ritchie, D.A., Hall, G., Pickup, R.W., Saunders, J.R., 1996. Isolation and identification of methanogen-specific DNA from blanket bog peat by PCR amplification and sequence analysis. Applied and Environmental Microbiology 62, 668-675. Hedges, J., Keil, R., Benner, R., 1997. What happens to terrestrial organic matter in the ocean? Organic Geochemistry 27, 195-212. Hedges, J.I., Mann, D.C., 1979. The lignin geochemistry of marine sediments from the southern Washington coast. Geochimica et Cosmochimica Acta 43, 1809-1818. Heuer, V.B.; Aiello, I.W.; Elvert, M.; Goldenstein, N.I.; Goldhammer, T.; Könneke, M.; Liu, X.; Pape, S.S.; Schmidt, F.A.; Wendt, J.K.; Zhuang, G.: Report and preliminary results of R/V POSEIDON cruise POS450, DARCSEAS II – Deep subseafloor Archaea in the Western Mediterranean Sea: Carbon Cycle, Life Strategies, and Role in Sedimentary Ecosystems, Barcelona (Spain) – Malaga (Spain), April 2 – 13, 2013. Berichte, MARUM – Zentrum für Marine Umweltwissenschaften, Fachbereich Geowissenschaften, Universität Bremen, No. 305, 42 pages. Bremen, 2014. ISSN 2195-9633. Heuer, V., Elvert, M., Tille, S., Krummen, M., Mollar, X.P., Hmelo, L.R., Hinrichs, K.-U., 2006. Online δ13C analysis of volatile fatty acids in sediment/porewater systems by liquid chromatography-isotope ratio-mass spectrometry. Limnology and Oceanography: Methods 4, 346-357. Kadam, P.C., Ranade, D., Mandelco, L., Boone, D.R., 1994. Isolation and characterization of Methanolobus bombayensis sp. nov., a methylotrophic methanogen that requires high concentrations of divalent cations. International Journal of Systematic Bacteriology 44, 603-607. Keller, M., Kiene, R., Matrai, P., Bellows, W., 1999. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. I. Batch cultures. Marine Biology 135,

162 Chapter 5

237-248. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. King, G., Klug, M., Lovley, D., 1983. Metabolism of acetate, methanol, and methylated amines in intertidal sediments of Lowes Cove, Maine. Applied and Environmental Microbiology 45, 1848-1853. King, G.M., 1984. Utilization of hydrogen, acetate, and “noncompetitive”; substrates by methanogenic bacteria in marine sediments. Geomicrobiology Journal 3, 275-306. King, G.M., 1988. Distribution and metabolism of quaternary amines in marine sediments. John Wiley and Sons, Chichester, United Kingdom. Kolb, S., 2009. Aerobic methanol-oxidizing Bacteria in soil. FEMS Microbiology Letters 300, 1-10. Krzycki, J., Kenealy, W., DeNiro, M., Zeikus, J., 1987. Stable carbon isotope fractionation by Methanosarcina barkeri during methanogenesis from acetate, methanol, or carbon dioxide-hydrogen. Applied and Environmental Microbiology 53, 2597-2599. L'Haridon, S., Chalopin, M., Colombo, D., Toffin, L., 2014. Methanococcoides vulcani sp. nov., a novel marine methylotrophic methanogen; using betaine, choline and N, N-dimethylethanolamine for methanogenesis, isolated from the Napoli Mud Volcano in the Eastern Mediterranean Sea; and emendation of the genus Methanococcoides. International Journal of Systematic and Evolutionary Microbiology, ijs. 0.058289-058280. Larkin, M.A., Blackshields, G., Brown, N., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947-2948. Lazar, C.S., John Parkes, R., Cragg, B.A., L'Haridon, S., Toffin, L., 2012. Methanogenic activity and diversity in the centre of the Amsterdam Mud Volcano, Eastern Mediterranean Sea. FEMS Microbiology Ecology 81, 243-254. Lazar, C.S., L'Haridon, S., Pignet, P., Toffin, L., 2011a. Archaeal populations in hypersaline sediments underlying orange microbial mats in the Napoli Mud Volcano. Applied and Environmental Microbiology 77, 3120-3131. Lazar, C.S., Parkes, R.J., Cragg, B.A., L'Haridon, S., Toffin, L., 2011b. Methanogenic diversity and activity in hypersaline sediments of the centre of the Napoli mud volcano, Eastern Mediterranean Sea. Environmental Microbiology 13, 2078-2091. Lee, C., Olson, B.L., 1984. Dissolved, exchangeable and bound aliphatic amines in marine sediments: initial results. Organic Geochemistry 6, 259-263.

163 Chapter 5

Lin, Y.-S., Heuer, V.B., Goldhammer, T., Kellermann, M.Y., Zabel, M., Hinrichs, K.-U., 2012.

Towards constraining H2 concentration in subseafloor sediment: A proposal for combined analysis by two distinct approaches. Geochimica et Cosmochimica Acta 77, 186-201. Liu, Y., Whitman, W.B., 2008. Metabolic, phylogenetic, and ecological diversity of the methanogenic archaea. Annals of the New York Academy of Sciences 1125, 171-189. Lovley, D.R., Goodwin, S., 1988. Hydrogen concentrations as an indicator of the predominant terminal electron-accepting reactions in aquatic sediments. Geochimica et Cosmochimica Acta 52, 2993-3003. Lovley, D.R., Klug, M.J., 1983. Methanogenesis from methanol and methylamines and acetogenesis from hydrogen and carbon dioxide in the sediments of a eutrophic lake. Applied and Environmental Microbiology 45, 1310-1315. Lü, Z., Lu, Y., 2012. Methanocella conradii sp. nov., a thermophilic, obligate hydrogenotrophic methanogen, isolated from Chinese rice field soil. PloS One 7, e35279. Ludwig, W., Dumont, E., Meybeck, M., Heussner, S., 2009. River discharges of water and nutrients to the Mediterranean and Black Sea: major drivers for ecosystem changes during past and future decades? Progress In Oceanography 80, 199-217. Martens, C.S., Berner, R.A., 1974. Methane production in the interstitial waters of sulfate-depleted marine sediments. Science 185, 1167-1169. Milkov, A.V., 2004. Global estimates of hydrate-bound gas in marine sediments: how much is really out there? Earth-Science Reviews 66, 183-197. Mitterer, R.M., 2010. Methanogenesis and sulfate reduction in marine sediments: A new model. Earth and Planetary Science Letters 295, 358-366. O'Sullivan, L.A., Sass, A.M., Webster, G., Fry, J.C., Parkes, R.J., Weightman, A.J., 2013. Contrasting relationships between biogeochemistry and prokaryotic diversity depth profiles along an estuarine sediment gradient. FEMS Microbiology Ecology 85, 143-157. Orcutt, B., Boetius, A., Elvert, M., Samarkin, V., Joye, S.B., 2005. Molecular biogeochemistry of sulfate reduction, methanogenesis and the anaerobic oxidation of methane at Gulf of Mexico cold seeps. Geochimica et Cosmochimica Acta 69, 4267-4281. Oremland, R., Polcin, S., 1982. Methanogenesis and sulfate reduction: competitive and noncompetitive substrates in estuarine sediments. Applied and Environmental Microbiology 44, 1270-1276. Orphan, V., Jahnke, L., Embaye, T., Turk, K., Pernthaler, A., Summons, R., Des Marais, D., 2008. Characterization and spatial distribution of methanogens and methanogenic biosignatures in hypersaline microbial mats of Baja California. Geobiology 6, 376-393. Parkes, R.J., Brock, F., Banning, N., Hornibrook, E.R.C., Roussel, E.G., Weightman, A.J., Fry,

164 Chapter 5

J.C., 2012. Changes in methanogenic substrate utilization and communities with depth in a salt-marsh, creek sediment in southern England. Estuarine, Coastal and Shelf Science 96, 170-178. Parkes, R.J., Cragg, B.A., Banning, N., Brock, F., Webster, G., Fry, J.C., Hornibrook, E., Pancost, R.D., Kelly, S., Knab, N., 2007. Biogeochemistry and biodiversity of methane cycling in subsurface marine sediments (Skagerrak, Denmark). Environmental Microbiology 9, 1146-1161. Parkes, R.J., Sass, H., Webster, G., Watkins, A.J., Weightman, A.J., O'Sullivan, L.A., Cragg, B.A., 2010. Methods for studying methanogens and methanogenesis in marine sediments, Handbook of Hydrocarbon and Lipid Microbiology. Springer, pp. 3799-3826. Parkes, R.J., Webster, G., Cragg, B.A., Weightman, A.J., Newberry, C.J., Ferdelman, T.G., Kallmeyer, J., Jørgensen, B.B., Aiello, I.W., Fry, J.C., 2005. Deep sub-seafloor prokaryotes stimulated at interfaces over geological time. Nature 436, 390-394. Paul, K., Nonoh, J.O., Mikulski, L., Brune, A., 2012. “Methanoplasmatales,” Thermoplasmatales-related archaea in termite guts and other environments, are the seventh order of methanogens. Applied and Environmental Microbiology 78, 8245-8253. Payne, W., 1973. Reduction of nitrogenous oxides by microorganisms. Bacteriological Reviews 37, 409-452. Rabus, R., Hansen, T., Widdel, F., 2006. Dissimilatory sulfate-and sulfur-reducing prokaryotes. The Prokaryotes 2, 659-768. Rudolf, T., 1998. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology 144, 2377-2406. Sakai, S., Conrad, R., Liesack, W., Imachi, H., 2010. Methanocella arvoryzae sp. nov., a hydrogenotrophic methanogen isolated from rice field soil. International Journal of Systematic and Evolutionary Microbiology 60, 2918-2923. Sakai, S., Imachi, H., Hanada, S., Ohashi, A., Harada, H., Kamagata, Y., 2008. Methanocella paludicola gen. nov., sp. nov., a methane-producing archaeon, the first isolate of the lineage ‘Rice Cluster I’, and proposal of the new archaeal order Methanocellales ord. nov. International Journal of Systematic and Evolutionary Microbiology 58, 929-936. Schink, B., Phelps, T.O.M.J., Eichler, B., Zeikus, J., 1985. Comparison of ethanol degradation pathways in anoxic freshwater environments. Journal of General Microbiology 131, 651-660. Schink, B., Zeikus, J., 1980. Microbial methanol formation: a major end product of pectin metabolism. Current Microbiology 4, 387-389. Schink, B., Zeikus, J., 1982. Microbial ecology of pectin decomposition in anoxic lake sediments. Journal of General Microbiology 128, 393-404.

165 Chapter 5

Seeberg-Elverfeldt, J., Schlüter, M., Feseker, T., Kölling, M., 2005. Rhizon sampling of pore waters near the sediment/water interface of aquatic systems. Limnology and Oceanography: Methods 3, 361-371. Sempéré, R., Charrière, B., Van Wambeke, F., Cauwet, G., 2000. Carbon inputs of the Rhone River to the Mediterranean Sea: biogeochemical implications. Global Biogeochemical Cycles 14, 669-681. Summons, R.E., Franzmann, P.D., Nichols, P.D., 1998. Carbon isotopic fractionation associated with methylotrophic methanogenesis. Organic Geochemistry 28, 465-475. Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24, 1596-1599. Treude, T., Krause, S., Schweers, J., Dale, A.W., Coffin, R., Hamdan, L.J., Sulfate reduction and methane oxidation activity below the sulfate-methane transition zone in Alaskan-Beaufort continental margin sediments: Implications for deep sulfur cycling. Geochimica et Cosmochimica Acta. van der Maarel, M.J.E.C., Hansen, T.A., 1997. Dimethylsulfoniopropionate in anoxic intertidal sediments: a precursor of methanogenesis via dimethyl sulfide, methanethiol, and methiolpropionate. Marine Geology 137, 5-12. Wang, X.-C., Lee, C., 1990. The distribution and adsorption behavior of aliphatic amines in marine and lacustrine sediments. Geochimica et Cosmochimica Acta 54, 2759-2774. Wang, X.-C., Lee, C., 1993. Adsorption and desorption of aliphatic amines, amino acids and acetate by clay minerals and marine sediments. Marine Chemistry 44, 1-23. Wang, X.-C., Lee, C., 1994. Sources and distribution of aliphatic amines in salt marsh sediment. Organic Geochemistry 22, 1005-1021. Watkins, A.J., Roussel, E.G., Parkes, R.J., Sass, H., 2014. Glycine betaine as a direct substrate for methanogens (Methanococcoides spp.). Applied and Environmental Microbiology 80, 289-293. Watkins, A.J., Roussel, E.G., Webster, G., Parkes, R.J., Sass, H., 2012. Choline and N, N-Dimethylethanolamine as direct substrates for methanogens. Applied and Environmental Microbiology 78, 8298-8303. Webster, G., John Parkes, R., Cragg, B.A., Newberry, C.J., Weightman, A.J., Fry, J.C., 2006. Prokaryotic community composition and biogeochemical processes in deep subseafloor sediments from the Peru Margin. FEMS Microbiology Ecology 58, 65-85. Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161, 291-314. Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and

166 Chapter 5

freshwater environments: CO2 reduction vs. acetate fermentation-Isotope evidence. Geochimica et Cosmochimica Acta 50, 693-709. Wildgruber, G., Thomm, M., König, H., Ober, K., Richiuto, T., Stetter, K.O., 1982. Methanoplanus limicola, a plate-shaped methanogen representing a novel family, the Methanoplanaceae. Archives of Microbiology 132, 31-36. Yoshioka, H., Maruyama, A., Nakamura, T., Higashi, Y., Fuse, H., Sakata, S., Bartlett, D., 2010. Activities and distribution of methanogenic and methane‐oxidizing microbes in marine sediments from the Cascadia Margin. Geobiology 8, 223-233. Yoshioka, H., Sakata, S., Cragg, B.A., Parkes, R.J., Fujii, T., 2009. Microbial methane production rates in gas hydrate-bearing sediments from the eastern Nankai Trough, off central Japan. Geochemical Journal 43, 315-321. Zhuang, G.-C., Lin, Y.-S., Elvert, M., Heuer, V.B., Hinrichs, K.-U., 2014. Gas chromatographic analysis of methanol and ethanol in marine sediment pore waters: Validation and implementation of three pretreatment techniques. Marine Chemistry 160, 82-90.

167

168 Chapter 6

Chapter 6

Conclusions and perspectives

169 Chapter 6

6.1 Conclusions

This PhD project aims to constrain the least acknowledged methanogenic pathway, methylotrophic methanogenesis in marine sediments. The performed studies improved our current understanding on methylated substrates and methylotrophic methanogenesis, and provided new insights into its quantitative importance for methane production in marine sediment. The major conclusions are: 1) Novel analytical methods allow for the quantitative or isotopic analysis of methylated substrates in marine sediments. In this thesis, new sensitive methods with different pretreatment procedures were established for the trace analysis of methylated compounds in marine sediment or pore waters. With these protocols, the concentrations or isotopes of methanol, ethanol, TMA, DMS or DMSP were successfully determined in small volumes of marine pore waters or sediments. 2) Low-molecular-weight methylated compounds are present in marine sediment or pore waters. The abundance and distribution of methylated compounds were investigated in a variety of marine sediments. Methanol and ethanol concentrations were generally lower than 10 μM in the pore water. The concentrations of TMA and DMS were at the nanomolar range in the pore water, while the base-extractable TMA and base-hydrolyzable DMS were at least three orders of magnitude higher in the sediment. The distribution of methylated compounds showed spatial and vertical variability, which could be related to the production and consumption processes. The occurrence of these methylated compounds suggested that they could be potential substrates for methylotrophic methanogenesis. 3) Methylotrophic methanogenesis was the dominant methanogenic pathway in hypersaline sediments. Large amounts of methane occurred in the presence of high concentrations of sulfate in Orca Basin, and strong evidences from substrate abundance, carbon isotopes systematics, thermodynamic calculations, experimental incubations and lipid and gene biomarkers confirmed that methane could be produced from methylotrophic methanogenesis with the utilization of methylated substrates. The findings highlight the particular importance of methylotrophic methanogenesis in

170 Chapter 6 hypersaline environments. 4) Methylotrophic methanogenesis could significantly contribute to methane production in the sulfate-reducing sediment. In estuarine sediments from northwestern Mediterranean Sea, the relative importance of different methanogenic pathways were quantitatively assessed with depth related methanogenesis rate measurements. The turnover time for methylated substrates such as methanol (~44 days) and methylamine (~25 days) were much faster than acetate (> 105 days) and bicarbonate (> 106 days) in the surface sulfate-reducing sediments. Consequently, methylotrophic methanogenesis could account for 13-74% of the total methane production in the sulfate reduction zone. In the sulfate-depleted sediments, hydrogenotrophic methanogenesis dominated the methane production (67%-97%), acetoclastic methanogenesis could also supply a significant amount of methane in organic-rich sediment (31%), while the contribution was neglected from methylotrophic methanogenesis (< 1%). These results also demonstrated that the activity of sulfate-reducing bacteria and the supply of organic matter could significantly influence methanogenic pathways in marine sediment.

6.2 Perspectives

Although new constraints from this study have advanced our qualitative and quantitative understanding on methylated compounds and methylotrophic methanogenesis in marine sediments, there is still space to be improved or open questions could be further addressed: 1) The precursors and the generation of methylated compounds in marine sediment. Whereas the low-molecular-weight methylated compounds are present in the marine sediment, their precursors and the processes how they are produced are not fully constrained. For example, the degradation of lignin or pectin might contribute to the formation of methanol (Donnelly and Dagley, 1980; Schink and Zeikus, 1980). In addition to known precursors from osmolytes (e.g., glycine betaine, DMSP), amino acids could also be decomposed to TMA and DMS (Kiene and Capone, 1988; King, 1984). In the carbonate sediment, it has been assumed that high abundance of amino

171 Chapter 6

acid (e.g., glycine, methionine) provided a source of methylated substrates for methanogens that allow them to produce significant amounts of methane in the sulfate

reduction zone (Mitterer, 2010; Mitterer et al., 2001). For these assumptions, it would be helpful to look for more direct evidences, e.g., with in vitro incubation experiments. 2) The metabolism of methylated compounds by non-methanogenic process. Despite that methylated compounds are non-competitive substrates for methanogens, previous studies also demonstrated that the consumption of these compounds could be related to non-methanogenic processes (Daniel et al., 1999; Kiene et al., 1986; King et al., 1983). Thus questions might arise which other microorganisms could utilize these substrates and what is the relative importance of methanogenic or non-methanogenic pathways for their consumption. The answer for these questions may help us better understand the biogeochemistry of low-molecular-weight compounds in marine sediment. 3) Isotopic analysis of volatile alcohols. Given the abundance and the rapid turnover of methanol, the role of methanol as substrates for methanogen or other microbes might be more important than previously recognized. To better understand the metabolic pathways involving methanol production or consumption, the carbon isotopes of methanol could provide valuable information, as inferred from acetate metabolism in marine sediment (Heuer et al., 2009). However, the HPLC method for acetate (Heuer et al., 2006) can not suffice the isotopic analysis of methanol due to the low concentration in the pore water. A possible way for it is to couple the developed purge and trap system to the GC-IRMS, while the fractionation effect during separation between liquid and gas phase needs to be considered and requires further examination. 4) The assimilation of methylated compounds into biomass. In addition to serving as energy source, methylated compounds are important source of carbon for marine methylotrophs. The uptake of methanol by denitrifying microorganisms has been demonstrated through DNA-SIP experiment (Ginige et al., 2004; Kalyuhznaya et al., 2009). Both turnover to methane and assimilation into lipid of methanol by

172 Chapter 6

ANME was observed with SIP in ANME-dominated microbial mat of Black Sea (Bertram et al., 2013). Thus it would be attractive to link the biomarker or in-situ community with the assimilation activity of methylated compounds with SIP or radiotracer assays. 5) Factors influencing methanogenic pathways and potential implications for methanogenesis in the marine subseafloor. As demonstrated in this thesis, the relative importance of different methanogenic pathways might vary from habitat to habitat, from site to site, and factors such as salinity, the activity of sulfate-reducing bacteria and organic matter could significantly influence the production of methane. Likewise, other environmental factors, such as pH or temperature could also affect methane production. In deep marine subseafloor, it has been shown that acetate could be abundantly released with increasing temperature during burial, and acetoclastic methanogenesis was considerably stimulated due to the high availability of acetate (Wellsbury et al., 1997). As degradation products from organic matter, methylated compounds, e.g., methanol, could also be expected to be generated in the increasingly heated sediment. If so, methylated compounds could be high energetic substrates for methylotrophic methanogenesis or other processes in deep biosphere.

6.3 References

Bertram, S., Blumenberg, M., Michaelis, W., Siegert, M., Krüger, M., Seifert, R., 2013. Methanogenic capabilities of ANME‐archaea deduced from 13C‐labelling approaches. Environmental Microbiology 15, 2384-2393. Daniel, R., Warnecke, F., Potekhina, J.S., Gottschalk, G., 1999. Identification of the syntrophic partners in a coculture coupling anaerobic methanol oxidation to Fe(III) reduction. FEMS Microbiology Letters 180, 197-203. Donnelly, M., Dagley, S., 1980. Production of methanol from aromatic acids by Pseudomonas putida. Journal of Bacteriology 142, 916-924. Ginige, M.P., Hugenholtz, P., Daims, H., Wagner, M., Keller, J., Blackall, L.L., 2004. Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization-microautoradiography to study a methanol-fed denitrifying microbial

173 Chapter 6

community. Applied and Environmental Microbiology 70, 588-596. Heuer, V., Elvert, M., Tille, S., Krummen, M., Mollar, X.P., Hmelo, L.R., Hinrichs, K.-U., 2006. Online δ13C analysis of volatile fatty acids in sediment/porewater systems by liquid chromatography-isotope ratio-mass spectrometry. Limnology and oceanography: Methods 4, 346-357. Heuer, V.B., Pohlman, J.W., Torres, M.E., Elvert, M., Hinrichs, K.-U., 2009. The stable carbon isotope biogeochemistry of acetate and other dissolved carbon species in deep subseafloor sediments at the northern Cascadia Margin. Geochimica et Cosmochimica Acta 73, 3323-3336. Kalyuhznaya, M.G., Martens‐Habbena, W., Wang, T., Hackett, M., Stolyar, S.M., Stahl, D.A., Lidstrom, M.E., Chistoserdova, L., 2009. Methylophilaceae link methanol oxidation to denitrification in freshwater lake sediment as suggested by stable isotope probing and pure culture analysis. Environmental Microbiology Reports 1, 385-392. Kiene, R., Capone, D., 1988. Microbial transformations of methylated sulfur compounds in anoxic salt marsh sediments. Microbial Ecology 15, 275-291. Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G., 1986. Metabolism of reduced methylated sulfur compounds in anaerobic sediments and by a pure culture of an estuarine methanogen. Applied and Environmental Microbiology 52, 1037-1045. King, G., Klug, M., Lovley, D., 1983. Metabolism of acetate, methanol, and methylated amines in intertidal sediments of Lowes Cove, Maine. Applied and Environmental Microbiology 45, 1848-1853. King, G.M., 1984. Metabolism of trimethylamine, choline, and glycine betaine by sulfate-reducing and methanogenic bacteria in marine sediments. Applied and Environmental Microbiology 48, 719-725. Mitterer, R.M., 2010. Methanogenesis and sulfate reduction in marine sediments: A new model. Earth and Planetary Science Letters 295, 358-366. Mitterer, R.M., Malone, M.J., Goodfriend, G.A., Swart, P.K., Wortmann, U.G., Logan, G.A., Feary, D.A., Hine, A.C., 2001. Co‐generation of hydrogen sulfide and methane in marine carbonate sediments. Geophysical Research Letters 28, 3931-3934. Schink, B., Zeikus, J., 1980. Microbial methanol formation: a major end product of pectin metabolism. Current Microbiology 4, 387-389.

Wellsbury, P., Goodman, K., Barth, T., Cragg, B.A., Barnes, S.P., Parkes, R.J., 1997. Deep marine

biosphere fuelled by increasing organic matter availability during burial and heating. Nature

388, 573-576.

174 Acknowledgements

Acknowledgements

For me, the four years in Bremen is definitely one of the most important time of my life, where not only I pursue my dreams in science, but also have new experiences outside of academia. Looking back at these years, I have to say, there are so many people I would like to thank for helping me have an easy life and wonderful scientific experience. First of all, I would like to express my extreme gratitude to my supervisor Kai-Uwe Hinrichs. It was him who gave me the opportunity to study in the world-leading group in the field of organic geochemistry. During these years, he always guided me with endless support, great patience, and fruitful discussion, and inspired me whenever I was depressed. I feel so lucky and proud as his PhD student. Furthermore, I owe a huge debt of gratitude to my co-supervisors Mandy Joye (University of Georgia, Athens, USA) and Andreas Teske (University of North Carolina, Chapel Hill, USA). Mandy endowed me the external research opportunities and I enjoyed very much during the two-month research stay in her lab. Andreas contributed a lot to my projects and kindly agreed to be my second supervisor. My PhD project could not be accomplished without their supports and suggestions. I would like to deeply acknowledge my other committee members. In the first two years, Yu-Shih Lin was always the first person who came into my mind when I met problems in science. She trained me from the small details of method development to the big scientific pictures with her brilliant ideas. Marcus Elvert was always supportive from the technical assistance to the scientific discussion. Verena Heuer was also an extremely helpful and patient mentor, and her comments greatly improve the quality of this thesis. I benefited a lot from the collaboration with my group members and scientists from other institutes. Marshall Bowles always gave me feedbacks at the first time no matter whatever I requested, and inspired me when I was frustrated and helpless. Felix Elling was an excellent and hard-working lab mate, who generously shared his

175 Acknowledgements

data and ideas with me, and helped me translate the abstract. Weichao Wu was a nice office mate, Chinese friend, and brilliant colleague that I could discuss with. The technicians, Xavier Prieto, Jenny Wendt, Evert Kramer, Heidi Taubner, Jessica Arndt and our secretary, Birgit Schmincke provided excellent technical and administrative support in the labs and offices. Vladimir A. Samarkin from UGA introduced me the radiotracers technique not only hand by hand, but with funny jokes. Mark Lever and Bo Jørgensen from Aarhus University organized a wonderful short cruise tailed for my project, and provided me valuable samples, post-cruise data and discussions. I would also like to thank my friends, colleagues, and lab mates, either sailed with me, helped me in the labs or shared experiences and thoughts with me. They are: Cassandre Lazar, Lisa Nigro, Marcos Yoshinaga, Kevin Becker, Gonzalo V. Gómez

Sáez, Shuchai Gan, Jan Schröder, Martin Könneke, Frauke Schmidt, Tobias

Goldhammer, Travis Meador, Lars Wörmer, Julius Lipp, Florence Schubotz, Thomas Evans, Rishi Adhikari, Nadine Goldenstein, Xiaolei Liu, Sitan Xie, Rong Zhu. I appreciate my Chinese brothers and friends in Bremen, Xiting, Fei, Songtao, Wei, Yiting, Haiyan, Yanlong, Jianguo, Jia, Min. When I was lost in my life, they always accompanied with me, and inspired me to move on. The time we spent together was enjoyable and memorable. None of this work would have been possible without financial support. I thank the funding from the DFG (MARUM) and NSF (UGA). GLOMAR is acknowledged for providing the travel grants and organizing interesting courses. I am thankful to the support from Chinese Scholarship Council (CSC) and grant of the Gottfried-Wilhelm Leibniz Price for Kai-Uwe Hinrichs, which jointly ensure the accomplishment of my PhD study. Finally, I am and will be forever indebted to my family, my parents, my wife, my parents-in-law, my uncle, my aunt, my sister and my brother-in-law, my brothers. You are the constant source of love, joy, inspiration, advices and any unconditional support. I am so lucky and grateful that I could share my life with all of you.

176

Name: Guangchao Zhuang Datum: 06.10.2014

Anschrift:

E r k l ä r u n g

Hiermit versichere ich, dass ich

1. die Arbeit ohne unerlaubte fremde Hilfe angefertigt habe,

2. keine anderen als die von mir angegebenen Quellen und Hilfsmittel benutzt habe und

3. die den benutzten Werken wörtlich oder inhaltlich entnommenen Stellen als solche kenntlich gemacht habe.

______, den ......

------(Unterschrift)