Activity, diversity and community structure of aerobic methane-oxidizing and carbon dioxide-producing bacteria in soils from the Canadian high Arctic

Christine Martineau Department of Natural Resource Sciences Microbiology Unit McGill University, Montreal May 2011

A thesis submitted to McGill University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

© Christine Martineau, 2011 Abstract The fate of soil organic carbon stocked in permafrost environments is a major concern in the context of global warming. In this thesis, the bacterial populations implicated in two important aerobic microbially-driven processes of the carbon cycle, aerobic methane oxidation and carbon dioxide production, were studied in different soils from the Canadian high Arctic. A protocol for the safe and sensitive detection of DNA in cesium chloride density gradients for stable isotope probing of DNA, a recent and widely used technique in microbial ecology allowing for the identification of microorganisms implicated in the degradation of a specific substrate, was developed. Using this protocol, active methanotrophic bacteria from the genera Methylobacter and Methylomonas were identified in active layer soils from Eureka, in the Canadian high Arctic. These soils had the capacity to oxidize methane at 4°C and at room temperature (RT), but the oxidation rates were greater at RT and were significantly enhanced by nutrient amendment. Bacterial populations implicated in aerobic methane oxidation and carbon dioxide production were studied in three different soils with highly distinctive physico-chemical characteristics from Axel Heiberg Island, in the Canadian high Arctic. Using microarray and clone library analyses of the particulate methane monooxygenase gene (pmoA), putative atmospheric methane oxidizers from the uncultured genotypes “upland soil cluster gamma” and “upland soil cluster alpha” were detected for the first time in Arctic soils and were associated with near neutral and acidic pH conditions, respectively. The overall methanotrophic bacterial diversity in these soils was higher than previously described for other Arctic soils and the community composition differed depending on the soil type.

Potential methane oxidation rates of the soils at low and high methane concentrations were positively correlated to the relative abundance of genotype “upland soil cluster gamma”. Differences in the bacterial community structure in the three soils from Axel Heiberg Island were detected at the genera/species levels using microarrays of the 16S rRNA gene and were related to soil pH and seasonal changes. Shifts in

i community structure were also detected at the phyla/classes levels by real-time PCR (qPCR) of the 16S rRNA gene, with the soil carbon dioxide production rate being positively correlated to the relative abundance of bacterial groups previously described as copiotrophs (Alphaproteobacteria, Bacteroidetes, and Betaproteobacteria). The results from this study indicated that bacterial communities in high Arctic soils play an important role in two aerobic processes of the carbon cycle, methane oxidation and carbon dioxide production. Methanotrophic bacteria and methane oxidation were detected in these soils and might be implicated in the reduction of methane emissions from the melting permafrost in the context of global warming. Beside, the relatively higher abundance of copiotrophic bacterial taxa in high Arctic soils with high organic matter content might lead, upon warming, to a rapid increase in soil carbon dioxide production. Further research is needed to assess the relevance of these findings under in situ conditions in a warming climate.

ii Résumé Le sort du carbone contenu dans le pergélisol est une source de préoccupation majeure dans le contexte des changements climatiques. Dans cette thèse, les populations bactériennes impliquées dans deux processus du cycle du carbone, l’oxydation du méthane et la production du dioxyde de carbone, ont été étudiées dans différents sols du haut Arctique canadien. Un protocole pour la détection sensible et sécuritaire de l’ADN dans les gradients de chlorure de césium, étape cruciale d’une technique appelée « stable isotope probing of DNA » qui permet d’identifier les bactéries impliquées dans la dégradation d’un substrat, a d’abord été développé. Ce protocole a par la suite permis d’identifier des bactéries méthanotrophes actives appartenant aux genres Methylobacter et Methylosarcina dans des sols provenant d’Eureka, dans le haut Arctique canadien. La capacité de ces sols à dégrader le méthane a été observée à 4°C et à température ambiante, mais les taux de dégradation étaient nettement plus élevés à température ambiante tout en étant grandement stimulés par l’ajout de nutriments. Les bactéries impliquées dans l’oxydation du méthane et la production de dioxyde de carbone ont été étudiées dans trois sols présentant des caractéristiques physico-chimiques distinctes provenant de l’île d’Axel Heiberg, dans le haut Arctique canadien. Grâce à l’utilisation de biopuces à ADN et de librairies de clones ciblant le gène codant pour l’enzyme méthane monooxygénase particulaire (pmoA), des bactéries potentiellement impliquées dans l’oxydation du méthane atmosphérique appartenant aux génotypes « upland soil cluster gamma » et « upland soil cluster alpha » ont été identifiées pour la première fois dans des sols de l’Arctique. La présence de ces deux génotypes était respectivement associée à des sols de pH neutres et acides. De manière générale, la diversité des populations méthanotrophes détectées dans les sols de l’île d’Axel Heiberg était supérieure à ce qui est généralement observé dans d’autres sols de l’Arctique et variait selon le type de sol. La capacité des sols à dégrader le méthane à des concentrations faibles et élevées était positivement corrélée à la présence dans les sols du génotype « upland soil cluster gamma ».

iii L’utilisation d’une biopuce à ADN ciblant le gène codant pour l’ARN ribosomal 16S a permis la détection de différences dans la structure des communautés bactériennes au niveau genre/espèce dans les trois sols de l’île d’Axel Heiberg. Ces différences étaient associées au pH du sol et aux changements de saisons. Des différences dans la structure des communautés bactériennes ont également été détectées au niveau phylum/classe par PCR en temps réel ciblant le gène codant pour l’ARN ribosomal 16S de différents taxons bactériens. La capacité de production du dioxyde de carbone des sols était positivement corrélée à l’abondance relative de taxons bactériens présentant des attributs copiotrophes (Alphaproteobacteria, Bacteroidetes, and Betaproteobacteria). Les résultats de cette étude indiquent que les communautés bactériennes jouent un rôle important dans deux processus du cycle du carbone, l’oxydation du méthane et la production de dioxyde de carbone, dans les sols du haut Arctique canadien. L’oxydation du méthane, de même que les bactéries méthanotrophes associées à ce processus, ont été détectées dans ces sols, et pourraient donc jouer un rôle important dans la réduction des émissions de méthane liées au dégel du pergélisol suite au réchauffement climatique. D’autre part, l’abondance relative plus élevée de bactéries copiotrophes dans des sols du haut Arctique canadien contenant de grandes quantités de carbone pourrait conduire, suite à un réchauffement du climat, à une augmentation de la production de dioxyde de carbone. Des recherches plus approfondies seront nécessaires afin de déterminer si les résultats observés dans cette étude auront une incidence en conditions in situ dans le contexte des changements climatiques.

iv Contributions to knowledge

The work presented in this thesis contributed to the advancement of knowledge in several ways: ‐ A new protocol for the safe and sensitive detection of DNA in cesium chloride density gradients for stable isotope probing (SIP) assays was developed. This protocol was successfully used to detect active methanotrophs in high Arctic soils. The protocol was also successfully applied to other DNA-SIP assays in our laboratory, and other research groups indicated their interest in using this protocol to replace the traditional ethidium bromide protocol.

‐ Active methanotrophic bacteria were identified for the first time in high Arctic soils using DNA-SIP, confirming the importance of Type I methanotrophs as active members of the methanotrophic bacterial communities in this environment.

‐ This study is, to the best of my knowledge, the first study on methanotrophic bacteria in Arctic soils that includes a wide diversity of soil types with distinctive physico-chemical characteristics.

‐ Methane (CH4) oxidation at a low CH4 concentration, a high-affinity process implicated in atmospheric CH4 uptake that has rarely been reported in the literature for Arctic soils, was detected.

‐ Putative atmospheric methane-oxidizing bacteria from the “upland soil cluster gamma” and “upland soil cluster alpha” genotypes were detected for the first time in Arctic soils, and an overall wider diversity of methanotrophic bacteria than what was previously reported for cold environments was observed.

‐ A positive correlation between the relative abundance of copiotrophic bacterial taxa and carbon dioxide (CO2) production was demonstrated in the high Arctic soil environments examined.

‐ A simple multiple regression model in which the CO2 production rate in high Arctic soil is linked to the relative abundance of different groups of bacteria was proposed.

v Acknowledgements

I would like to thank the following persons:

Dr. Charles Greer, my supervisor, for giving me the opportunity to conduct this research project; for his patience and support at all times; and for enabling me to work in his laboratory with such a wonderful team.

Dr. Lyle Whyte, my co-supervisor, for his input in the project and financial support and for giving me the invaluable opportunity to travel with his team to the Arctic multiple times. These are unforgettable moments.

Dr Étienne Yergeau, for his friendship and precious collaboration. Without his statistical and writing help, much of this work would not have the same value.

All the research officers and technicians from the environmental microbiology group at BRI. Each has been great collaborators, providing me with help and support throughout the years. They were also great company to me and made a routine day in the lab a lot of fun. Special thanks to Claude Masson and Diane Labbé for their technical input in my project.

My family. My parents, for their eternal support. There are no words to say how important you have been throughout this journey. My three sisters, for their friendship and presence at all times. They are a great source of inspiration for their baby sister. My nephews and nieces, for the joy they bring in my life. My life partner Kerlin, for love, care and inspiration. And Rose-Alexia, our little treasure...

The great friends I have made throughout my time as a Ph.D student. Élisabeth, Silvia, Claudie, Josée, Roli, Thibault, Valentina, Lia. I feel fortunate to have met so many great persons all in the same place...

I also would like to thank the following agencies for their financial support: the Natural Sciences and Engineering Research Council of Canada (NSERC), le Fond Québécois de la Recherche sur la Nature et les Technologies (FQRNT), the International Polar Year (IPY‐MERGE), the Northern Science Training Program (NSTP), the Polar Continental Shelf Program (PSCP) and McGill University.

vi

This work is dedicated to Rose-Alexia, who gave me the strength, the time and the inspiration to achieve this...

vii Contribution of authors

For the manuscripts presented in Chapters 3 and 4 (published), the candidate was responsible for conducting the research, analyzing the data and preparing the manuscripts, with guiding and critical editing from the co-authors, Dr Whyte and Dr Greer. For the manuscripts presented in Chapter 5 and 6 (in preparation for publication), the candidate was responsible for conducting the research, with input from Dr Levente Bodrossy (pmoA microarrays, Chapter 5) and Dr Étienne Yergeau (real-time PCR of the 16S rRNA genes, Chapter 6). The candidate and Dr Yergeau performed data analyses for these manuscripts. The candidate was responsible for the preparation of manuscripts from Chapters 5 and 6, with guiding and critical editing from the co-authors, Dr Bodrossy (Chapter 5), Dr Yergeau, Dr Whyte, and Dr Greer (Chapters 5 and 6).

viii Table of contents Abstract i Résumé iii Contributions to knowledge v Acknowledgements vi Dedication vii Contribution of authors viii Table of contents ix List of tables xiv List of figures xv List of abbreviations xvii

Chapter 1 - General introduction 1.1 Problematic 1 1.2 Research objectives 3 1.3 Thesis outline 4

Chapter 2 – Literature review 2.1 The carbon cycle in the Arctic environment in the context of global warming 5 2.1.1 The permafrost carbon pool 5 2.1.2 The fate of the permafrost carbon pool in the context of global warming 5 2.1.3 Carbon dioxide in the atmosphere and the global CO2 cycle 7 2.1.4 Methane in the atmosphere and the global CH4 cycle 8 Methane in permafrost environments 10 The soil CH4 sink 10

2.2 Microbes and microbial processes associated with the CH4 cycle in cold environments 11 2.2.1 Methanogenic archaea 11 2.2.2 Anaerobic oxidation of methane 12 2.2.3 Aerobic methanotrophic bacteria 13 The CH4 oxidation pathway 14 High affinity methane-oxidizing bacteria 16 Facultative methanotrophic bacteria 18 2.2.4 Ecology of aerobic methanotrophic bacteria 18 2.2.5 Effect of soil parameters on methanotrophic bacterial community structure and CH4 oxidation 19 Temperature 19 Soil pH 20 Nitrogen 20 Methane concentration 21 Climatic changes 22 2.2.6 Methanotrophic bacteria in cold environments 22

ix Methanotrophic bacterial isolates from cold environments 22 Methanotrophic bacterial activity, abundance and diversity in cold environments 23

2.3 Microbes and microbial processes associated with soil respiration in cold environments 25 2.3.1 Microbial abundance and diversity in Arctic soils 26 Comparison between active layer and permafrost 27 Comparison between Arctic soils and soils from lower latitudes 29 2.3.2 Effect of soil parameters on soil bacteria and soil respiration 29 Temperature 29 Soil pH 31 Nutrients 31 Vegetation and organic matter content 32 Connecting text 32

Chapter 3 - Development of a SYBR safeTM technique for the sensitive detection of DNA in cesium chloride density gradients for stable isotope probing assays 3.1 Abstract 35 3.2 Main text 36 3.3 Acknowledgements 41 Connecting text 42

Chapter 4 - Stable isotope probing analysis of the diversity and activity of methanotrophic bacteria in soils from the Canadian high Arctic 4.1 Abstract 43 4.2 Introduction 44 4.3 Material and Methods 46 4.3.1 Site description, soil sampling and soil characterization 46 4.3.2 Soil incubation and CH4 degradation 47 4.3.3 DNA extraction, fractionation and quantification 48 4.3.4 Real-time PCR 48 4.3.5 PCR amplification and DGGE analyses 51 4.3.6 Phylogenetic, cluster and statistical analyses 52 4.3.7 Nucleotide sequence accession numbers 53 4.4 Results 53 4.4.1 Soil characteristics 53 4.4.2 Real-time PCR 54 4.4.3 Methane oxidation rates 55

x 4.4.4 13C-labelling of the DNA and DGGE analysis of the 16S rRNA gene 56 4.4.5 DGGE analysis of the pmoA gene 58 4.4.6 Sequence analysis of the 16S rRNA and pmoA genes 60 4.5 Discussion 64 4.6 Conclusions 69 4.7 Acknowledgements 70 Connecting text 71

Chapter 5 - Detection of putative atmospheric methane-oxidizing bacteria in soils from the Canadian high Arctic 5.1 Abstract 72 5.2 Introduction 73 5.3 Material and methods 75 5.3.1 Site description, soil sampling and soil characterization 75 5.3.2 DNA extraction, purification and quantification 76 5.3.3 pmoA microarray 77 5.3.4 Clone libraries 78 5.3.5 Statistical analyses 78 5.4 Results 80 5.4.1 Methane concentration and CH4 degradation 80 5.4.2 Microarray analyses of the pmoA/amoA genes 81 Microarray with the pmoA PCR 82 Microarrays with the pmoA/amoA PCR 82 Methanotrophic bacterial community structure 82 5.4.3 Clone libraries 85 pmoA clone libraries 85 pmoA/amoA clone libraries 86 Phylogenetic diversity of pmoA/amoA genes from high Arctic soils 86 5.4.4 Correlation between methanotrophic bacterial groups and soil CH4 oxidation rates 89

5.5 Discussion 90 5.5.1 Detection of putative methane-oxidizing bacteria in high Arctic soils 90 5.5.2 Methane oxidation at low and high CH4 concentrations and links with the methanotrophic bacterial communities 91 5.5.3 Diversity of methanotrophic bacteria in high Arctic soils 93 5.6 Conclusions 95 5.7 Acknowledgements 96 Connecting text 97

xi Chapter 6 - Differences in major bacterial taxa in three high Arctic soils lead to different carbon dioxide production rates 6.1 Abstract 98 6.2 Introduction 99 6.3 Material and methods 101 6.3.1 Site description, soil characterization and CO2 production 101 6.3.2 DNA extraction, purification and quantification 102 6.3.3 Real-time PCR 102 6.3.4 16S rRNA gene PCR amplification and microarray analysis 102 6.3.5 Statistical analyses 103 6.4 Results 104 6.4.1 Soil characteristics 104 6.4.2 16S rRNA gene microarray analysis 106 6.4.3 Real-time PCR 108 6.4.4 Carbon dioxide production rate and in situ CO2 flux 108 6.5 Discussion 110 6.5.1 Links between CO2 production and bacterial community structure 110 6.5.2 Links between soil characteristics and bacterial community structure 111 6.6 Conclusions 114 6.7 Acknowledgements 114 Chapter 7 – General discussion and conclusions

7.1 Discussion and conclusions related to aerobic CH4 oxidation 115 7.1.1 Development of a protocol for the sensitive detection of DNA in CsCl density gradients 115 7.1.2 Methane-oxidizing capacity of soils from the Canadian high Arctic 115 7.1.3 Diversity and community composition of the methanotrophs in different high Arctic soils 116 7.1.4 Links between the methanotrophic bacterial community composition, soil characteristics and methane oxidation capacity 117

7.2 Discussion and conclusions related to CO2 production 117 7.2.1 Carbon dioxide production capacity of different soils from the Canadian high Arctic 117 7.2.2 Links between the bacterial community composition in different high Arctic soils and the soil characteristics 118 7.2.3 Links between the bacterial community composition of high Arctic soils and CO2 production 119

7.3 Summary 120

xii List of references 121

Appendices - Microarray hybridization patterns 138

xiii List of tables

Table 2.1 - Comparison of all described families of aerobic methanotrophic bacteria. 15

Table 4.1 - Primers used in this study. 50

Table 4.2 - Physicochemical parameters of soil samples 32D, 33D and 34D from Eureka collected in 2006 and 2007. 54

Table 5.1 - Methane concentration and CH4 degradation rates at CH4 concentrations of 15 and 1000 ppm for three different Arctic soil types collected in 2008 (winter conditions) and 2009 (summer conditions). 81

Table 5.2 - Number of sequences, number of OTUs and Shannon diversity index of 3 pmoA clone libraries and 4 pmoA/amoA clone libraries from three different high Arctic soils. 86

Table 5.3 - Pearson correlation coefficients between the relative abundance of different groups of methanotrophs based on the microarray analyses of samples amplified with the pmoA or pmoA/amoA primers and the CH4 oxidation rates at 15 or 1000 ppm. 89

Table 6.1 - Soil description. 105

Table 6.2 - Soil characteristics. 105

Table 6.3 - Multiple regression coefficients for CO2 production rates. 109

xiv List of figures

Figure 2.1 - Global CH4 sources. 9

Figure 2.2 - Pathways for the oxidation of CH4. 16

Figure 3.1 - Comparison of the CsCl density gradient ultracentrifugation with SYBR safe™ or EtBr. 38

Figure 3.2 - Cesium chloride density gradient ultracentrifugation with SYBR safe™ of different concentrations of 13C-labelled DNA. 39

Figure 3.3 - Cesium chloride density gradient ultracentrifugation with SYBR safe™ of environmental DNA. 40

Figure 4.1 - Quantitative PCR of pmoA gene and 16S rRNA gene of type I methanotrophs. 55

Figure 4.2 - Methane oxidation rates of soil samples 32D, 33D, and 34D from Eureka. 56

Figure 4.3 - DGGE of PCR-amplified 16S rRNA gene fragments from DNA fractions retrieved from CsCl density gradients. 58

Figure 4.4 - Cluster analysis of 16S rRNA and pmoA gene DGGE banding patterns. 59

Figure 4.5 - Phylogenetic relationships of bacterial 16S rRNA gene sequences from heavy DNA fractions for soil samples 32D, 33D, and 34D from Eureka. 62

Figure 4.6 - Phylogenetic relationships of bacterial pmoA deduced amino acid sequences from heavy DNA fractions for soil samples 32D, 33D, and 34D from Eureka. 63

Figure 5.1 - Methanotrophic bacterial community composition of the acidic soils, upland tundra soils and wet meadow soils from Axel Heiberg Island. 84

Figure 5.2 - Methanotrophic bacterial community composition based on pmoA/amoA clone libraries from an acidic soil, an upland tundra soil, and a wet meadow soil from Axel Heiberg Island. 87

Figure 5.3 - Phylogenetic relationships of deduced amino acid sequences based on pmoA/amoA clone libraries from an acidic soil, an upland tundra soil and a wet meadow soil from Axel Heiberg Island. 88

Figure 6.1 - Phylogenetic community composition of the acidic soils, upland tundra soils and wet meadow soils from Axel Heiberg Island. 107

xv Figure A.1 - DNA microarray hybridization patterns of pmoA PCR products amplified from upland tundra soils and acidic soils from Axel Heiberg Island. 139

Figure A.2 - DNA microarray hybridization patterns of pmoA/amoA PCR products amplified from upland tundra soils and wet meadow soils from Axel Heiberg Island. 140

xvi List of abbreviations amoA Gene coding for the α-subunit of the ammonia monooxygenase ANME Anaerobic methanotrophic archaea ANOVA Analysis of variance AOB Ammonia-oxidizing bacteria AOM Anaerobic oxidation of methane Blast Basic local alignment search tool C Carbon CH4 Methane CO2 Carbon dioxide CsCl Cesium chloride DGGE Denaturing Gradient Gel Electrophoresis DNA Deoxyribonucleic acid EtBr Ethidium bromide IPTG Isopropyl-β-D-thiogalactopyranoside MMO Methane monooxygenase NCBI National Center for Biotechnology Information NMS Nitrate mineral salts medium OTU Operational Taxonomic Unit PCR Polymerase Chain Reaction Pg Petagrams PLFA Phospholipid fatty acid pMMO Particulate methane monooxygenase pmoA Gene coding for the α-subunit of the particulate methane monooxygenase ppb parts per billion ppm parts per million PVPP Polyvinylpolypyrrolidone qPCR Real-time PCR or quantitative PCR RNA Ribonucleic acid rRNA Ribosomal ribonucleic acid RT Room temperature SIP Stable isotope probing sMMO Soluble methane monooxygenase Tg Teragrams USCα Upland soil cluster alpha USCγ Upland soil cluster gamma UV Ultra-violet X-gal 5-bromo-4-chloro-3-indolylbeta-D-galactopyranoside

xvii Chapter 1 - General introduction 1.1 Problematic Permafrost regions occupy approximately 23.9% of the exposed land area of the Northern Hemisphere (Zhang et al., 2008). In the past 100 years, the average temperatures in the Arctic regions have increased at almost twice the global average rate (McBean et al., 2005) and the melting of permafrost is one of the most important impacts of global warming in these high latitude environments. Theoretical modelling suggests that as much as 90% of the near- surface permafrost could thaw by the end of the 21st century (Lawrence and Slater, 2005). While it has been generally reported that 15% of the total soil organic carbon for the 0-100 cm depth is stocked in permafrost (Post et al., 1982), a recent estimate indicates that the northern permafrost region contains as much as 50% of the global organic carbon pool (Tarnocai et al., 2009). The presence of these large amounts of carbon in permafrost is raising serious concerns whether melting permafrost, and the resulting increase in microbial activity, might be a source of extensive emissions of the greenhouse gases carbon dioxide (CO2) and methane (CH4) to the atmosphere. However, the fate of the carbon trapped in permafrost will both depend on environmental parameters (temperature, precipitation) and soil characteristics (quality and quantity of organic matter, nutrient availability, water content), but also on the microbial community implicated in carbon cycling. It is, therefore, crucial to understand the links between soil microorganisms, soil characteristics and carbon turnover in Arctic soils. In this thesis, the bacterial populations implicated in two aerobic microbially-driven processes of the carbon cycle, aerobic CH4 oxidation and CO2 production, were studied in different soils from the Canadian high Arctic.

Methane, which is 23 times more potent than CO2 as a greenhouse gas (IPCC, 2007), is produced by methanogenic archaea under anaerobic conditions. These microorganisms are known to inhabit permafrost environments (Rivkina et al., 2007; Simankova et al., 2003) and their capacity to produce CH4 at cold temperatures has been reported (Ganzert et al., 2007; Metje and Frenzel, 2007; Rivkina et al., 2007; Wagner et al., 2005). Their methanogenic activity is

1 expected to increase as permafrost soil temperature increases (Ganzert et al., 2007). Methane produced in anaerobic zones by methanogenic archaea can be oxidized in aerobic zones by aerobic methanotrophic bacteria. These bacteria utilize CH4 as a sole carbon and energy source and might contribute, therefore, to the reduction of CH4 emissions from melting permafrost. Some high-affinity methanotrophic bacteria also have the capacity to oxidize CH4 at atmospheric concentrations (~1.7 ppm) and are responsible for the atmospheric CH4 uptake by upland soils (Kolb, 2009). Methanotrophic bacterial diversity and activity have been previously assessed in a limited number of soils from cold environments (Dedysh et al., 1998; Liebner et al., 2009; Liebner and Wagner, 2007; Omelchenko et al., 1993; Pacheco-Oliver et al., 2002; Vecherskaya et al., 1993; Wagner et al., 2005; Wartiainen et al., 2003) but the pool of knowledge concerning the activity and diversity of methanotrophic bacterial populations in high Arctic soils is limited. Carbon dioxide is the most important greenhouse gas found in the atmosphere (IPCC, 2007). It plays an important role in the natural cycle of carbon, involving exchanges of large amounts of carbon between the ocean, the terrestrial biosphere and the atmosphere. In the Arctic environment, one important source of CO2 comes from soil respiration, which is principally driven by soil microorganisms, including soil bacteria. Some bacterial phyla and classes were shown to react coherently to organic matter inputs and were, therefore, classified as copiotrophs or oligotrophs (Fierer et al., 2007). Oligotrophs grow slowly, are adapted to low carbon inputs and have a diversified metabolic capacity, while copiotrophs are fast growing and are adapted to large inputs of carbon. Thus, copiotrophs are typically found in soils with high carbon turnover rates and oligotrophs in soils with low carbon turnover rates. The relative abundance of copiotrophic and oligotrophic groups of bacteria in high Arctic soils might influence the rate of CO2 production in these soils and it is thus crucial to characterize these bacterial groups in Arctic soils.

2 1.2 Research objectives The aim of this project was to get a better understanding of the links between soil microorganisms, soil characteristics and carbon turnover in different soils from the Canadian high Arctic. The general objective of this thesis was, therefore, to characterize the activity, diversity and community structure of the bacterial populations implicated in two aspects of the carbon cycle, CH4 oxidation and CO2 production. Specific objectives of the thesis were defined independently for each process:

Objectives related to CH4 oxidation: 1. Develop a protocol for the sensitive detection of DNA in cesium chloride density gradients for stable isotope probing (SIP) assays to allow for the use of this technique in the study of active methanotrophs in high Arctic soils; 2. Evaluate the capacity of soils from the Canadian high Arctic to oxidize

CH4 under various incubation conditions, including different temperatures,

nutrient amendments and CH4 concentrations; 3. Identify and characterize the diversity and community composition of the methanotrophs in different high Arctic soils using DNA-SIP, microarrays and clone libraries; 4. Link the methanotrophic bacterial community composition of high Arctic

soils to the soil characteristics and CH4 oxidation capacity.

Objectives related to CO2 production:

1. Evaluate the CO2 production capacity of different soils from the Canadian high Arctic; 2. Characterize the bacterial community composition in soils from the Canadian high Arctic using microarrays and real-time PCR (qPCR) of the 16S rRNA genes and link the microbiological data to the soil physico- chemical characteristics; 3. Link the bacterial community composition of high Arctic soils to the soil

CO2 production capacity, with a specific emphasis on copiotrophic and oligotrophic groups of bacteria.

3 1.3 Thesis outline This thesis includes four manuscripts, two of which have been published (Chapters 3 and 4), and two in preparation for publication (Chapter 5 and 6). An overview of the relevant literature is presented in Chapter 2, including background information on the carbon cycle in the Arctic environment and a presentation of the bacteria and bacterial processes associated with the CH4 and

CO2 cycles. Chapter 3 presents the technical development of a protocol for the safe and sensitive detection of DNA in cesium chloride density gradients for DNA-SIP, a recent technique that allows for the identification of microorganisms implicated in the degradation of a specific substrate. In Chapter 4, an experiment in which DNA-SIP was used to study the active methanotrophic bacterial populations in soils from Eureka, in the Canadian high Arctic, is presented. Chapter 5 and 6 present two experiments conducted on three different soils with highly distinctive physico-chemical characteristics collected in Axel Heiberg Island, in the Canadian high Arctic. In Chapter 5, methanotrophic bacterial communities in these soils were characterized using microarray and clone library analyses of the particulate methane monooxygenase gene (pmoA), a key gene involved in the initial oxidation reaction of CH4. The microbiological data were linked to soil physico-chemical characteristics and CH4 degradation rates at low and high CH4 concentrations. In Chapter 6, microarrays and real-time PCR (qPCR) targeting 16S rRNA genes were used to characterize the bacterial communities in the Axel Heiberg Island soils and the microbiological data were linked to soil physico-chemical characteristics and CO2 production rates. Chapter 7 presents a general discussion and conclusions in relation with the research objectives and a summary of the thesis.

4 Chapter 2 – Literature review

2.1 The carbon cycle in the Arctic environment in the context of global warming

2.1.1 The permafrost carbon pool Permafrost is defined as subsurface earth material remaining below 0°C for a minimum of two consecutive years (Permafrost subcommittee, 1988). It is widespread in the Arctic and boreal regions of the Northern Hemisphere, where permafrost regions occupy approximately 23.9% of the exposed land area (Zhang et al., 2008), which represents close to 16% of the global soil area (Tarnocai et al., 2009). The active layer in permafrost regions refers to the surface ground that thaws during the summer. The thickness of the active layer varies from a few tens of centimetres to more than 2 m in zones of continuous permafrost, but can reach several meters in zones of discontinuous permafrost. Organic carbon pools in permafrost regions are estimated to 191.29 petagrams (Pg) for the 0–30 cm depth, 495.80 Pg for the 0–100 cm depth, and 1024.00 Pg for the 0–300 cm depth (Tarnocai et al., 2009). The northern permafrost regions contain approximately 1672 Pg of carbon in total, of which 1466 Pg would occur in permafrost. This 1672 Pg of carbon represents approximately 50% of the global soil organic carbon pool (Tarnocai et al., 2009).

2.1.2 The fate of the permafrost carbon pool in the context of global warming The Arctic regions are known to be highly affected by global warming. The temperature has been increasing by 0.09°C per decade over the last century in the Arctic (north of the 60°), while the increase was of 0.06°C per decade in the Northern Hemisphere in general (McBean et al., 2005). Since 1966, the increase per decade has reached 0.4°C (McBean et al., 2005), fuelling concerns about the effect of global warming on polar regions. Theoretical modelling suggests that as much as 90% of the near-surface permafrost could thaw by the end of the 21st century as a result of these increasing temperatures (Lawrence and Slater, 2005). Trying to evaluate the fate of the carbon stocked in the permafrost and in soils in

5 general upon warming is extremely difficult because of the high number of factors potentially implicated, as described by Davidson and Janssens (2006). Briefly, if warming induces an increase in soil carbon decomposition through increased microbial activity and carbon is transferred to the atmosphere, a positive feedback to climate change could occur. However, if warming induces increasing plant growth and the accumulation of carbon in plant material exceeds the increase in decomposition, a negative feedback to climate change could occur. Evaluating the effect of climate warming on the soil carbon is made particularly difficult by the fact that decomposition of organic compounds includes microbial processes with a wide range of kinetic properties and temperature sensitivities (Davidson and Janssens, 2006). Moreover, several environmental parameters have the potential to affect soil carbon microbial decomposition, including the quantity and quality of organic carbon, the temperature, the soil water content and nutrient availability (Schuur et al., 2008). Additionally, the occurrence in the soil of oxic or anoxic conditions could be particularly important in determining the end product of microbial decomposition; if the carbon emerging from thawing permafrost is degraded under oxic condition, carbon dioxide (CO2) will be the main greenhouse gas being produced, while anoxic conditions will lead primarily to the production of CH4 (Schuur et al., 2008). In a recent study, Schuur et al. (2009) found that permafrost areas that thawed over the past 15 years were loosing annually 40% more carbon than areas subjected to minimal thaw, but had overall net ecosystem carbon uptake because of an increase in plant growth that compensated for these losses. However, areas that thawed decades earlier lost even more old carbon, which resulted in a net ecosystem carbon release despite increased plant growth. These results indicate that higher losses of soil carbon than plant carbon uptake with permafrost thaw over long timescales could make permafrost a large carbon source following climate warming (Schuur et al., 2009).

6 2.1.3 Carbon dioxide in the atmosphere and the global CO2 cycle Carbon dioxide is the most important greenhouse gas found in the atmosphere and CO2 emissions from human activities are considered the largest factor contributing to climate change (IPCC, 2007). The concentration of CO2 in the atmosphere has been increasing from pre-industrial values of about 280 parts per million (ppm) to nearly 380 ppm in 2005 (IPCC, 2007). The increase in atmospheric CO2 concentration is mainly associated with human activities, primarily the burning of fossil fuels and deforestation, but also changes in land use and management (IPCC, 2007). Carbon dioxide plays an important role in the carbon cycle, which involves exchanges of large amounts of carbon between the ocean, the terrestrial biosphere, and the atmosphere. Carbon dioxide is converted to biomass through photosynthesis by chlorophyll-containing organisms (plants, but also phytoplankton and Cyanobacteria), while processes like respiration and burning are responsible for the release of CO2 to the atmosphere. Carbon dioxide is also exchanged between the atmosphere and the ocean, reacting with water to form bicarbonate and carbonate ions.

The emission of CO2 from soils is recognized as one of the major sources in the global carbon cycle, and small variations to the soil respiration process could have a strong impact on atmospheric CO2 concentrations (Schlesinger and

Andrews, 2000). The CO2 emissions from soil respiration are estimated to 75 Pg of carbon annually (Schlesinger and Andrews, 2000). Soil CO2 emissions are mainly the result of autotrophic respiration by plant and heterotrophic respiration by soil microorganisms through the oxidation of plant-derived organic matter

(Falkowski et al., 2000). Although CO2 emissions from soils are an important source of carbon to the atmosphere, terrestrial ecosystems are globally responsible for the uptake of around 2.8 Pg of carbon annually (Arneth et al., 2010). This net terrestrial carbon uptake mainly depends on the balance between net primary productivity through photosynthesis and carbon losses from soil through soil respiration.

In northern latitudes environments (above 50°N), the CO2 balance of terrestrial ecosystems was found to be strongly affected by the temperature

7 increases occurring in autumn and spring, which have reached 1.1°C and 0.8°C over the past two decades, respectively (Piao et al., 2008). Both photosynthesis and respiration were found to increase in response to autumn warming, but the increase in respiration was higher than the increase in photosynthesis (Piao et al., 2008). In contrast, photosynthesis increased more than respiration in response to spring warming conditions. These results indicate that northern terrestrial ecosystems might be net sources of CO2 during the autumn and that, if future warming in autumn occurs faster than in spring, the capacity of northern ecosystems to act as a carbon sink may be reduced more and earlier, than previously suggested (Piao et al., 2008).

2.1.4 Methane in the atmosphere and the global CH4 cycle Methane is the second most important anthropogenic greenhouse gas after

CO2, yet it is 23 times more potent as greenhouse gas (IPCC, 2007). The concentration of CH4 in the atmosphere has been increasing from pre-industrial values of about 715 parts per billion (ppb) to currently about 1770 ppb (Conrad, −1 2009). The growth rate of atmospheric CH4 was 12 ppb year in the 1980s but has decreased since the early 1990s to reach 4 ppb year−1 since 1999 (Bousquet et al., 2006). In a study looking at the processes controlling the variations in CH4 emissions between 1984 and 2003, Bousquet et al. (2006) found that wetland emissions were mainly responsible for the inter-annual variability of CH4 sources while fire emissions generally played a smaller role. Their results also showed that the decrease in atmospheric CH4 growth during the 1990s was related to a reduction in anthropogenic emissions. Since 1999, anthropogenic emissions of

CH4 have started to increase again, but the effect of this increase on the growth rate of atmospheric CH4 has been reduced by the decrease in wetland emissions.

However, the authors indicated that atmospheric CH4 concentrations may increase again if emissions from wetlands return to their 1990s levels (Bousquet et al., 2006).

The global CH4 cycle has been reviewed recently by Conrad (2009). The −1 total CH4 sources are in the order of 500–600 teragrams (Tg) CH4 year . The

8 individual sources of CH4 are shown in Figure 2.1. About 69% of all CH4 sources are the result of CH4 production by methanogenic archaea (biogenic sources), while 25% of these sources are associated with mining, combustion of fossil fuels, and burning of biomass (non-biogenic sources). Wetlands are the largest individual source of biogenic CH4. The largest CH4 sink is the photochemical oxidation of CH4 with OH radicals, which represents more than 80% of the total

CH4 sinks. The two other sinks are the diffusion of CH4 into the stratosphere and microbial CH4 oxidation in soils (Conrad, 2009). The total CH4 sources are slightly higher than the total CH4 sinks, causing the growth in atmospheric CH4 concentrations observed since pre-industrial times.

Figure 2.1 - Global CH4 sources in per cent of the total budget of about 500–600 Tg CH4 per year. Source: (Conrad, 2009).

Net CH4 flux between the earth’s surface and the atmosphere is the result of the balance between CH4 production and CH4 oxidation (Conrad, 2009).

Indeed, for most CH4 sources, rates of CH4 production are much larger than rates of CH4 emission because a large proportion of the CH4 produced is consumed by methanotrophic bacteria before reaching the atmosphere. In rice fields, about 20% of the CH4 produced is oxidized by methanotrophic bacteria, thus reducing the net

CH4 flux from these environments (Conrad, 2009). Another example is the emission of CH4 from gas hydrates, which is strongly reduced through the activity of anaerobic methane oxidizers (ANME) (Reeburgh, 2007).

9 Methane in permafrost environments

Significant amounts of CH4 are preserved as CH4 hydrates at an average depth of several hundred meters in the permafrost (Rivkina et al., 2007). Methane is also found in near-surface permafrost layers and it could be liberated to the atmosphere as permafrost melts (Rivkina et al., 2007). Actual CH4 emissions from soils from high latitudes would represent up to 7% of the global sources of

CH4 and largely depend on the soil humidity, temperature and vegetation type (Rivkina et al., 2007). Emissions from tundra soils were found to decrease to a low level after the growing season but then increased significantly during the freeze-in period, most probably because the CH4 that had accumulated in the soil was squeezed out through frost action (Mastepanov et al., 2008).

The soil CH4 sink

The soil CH4 sink represents approximately 5% of the global CH4 budget, but high variations are observed when looking at different estimates of the amount −1 of CH4 consumed annually by soils. While a value of 10 Tg CH4 y had been previously reported (Fung et al., 1991), a study including 318 annual estimates from a wide diversity of ecosystems estimated the total soil annual CH4 uptake at −1 −1 36 ± 23 Tg yr , this value being reduced to 22±12 Tg CH4 y if the measurements were stratified by climatic zone, ecosystem, and soil texture (Dutaur and Verchot, 2007). A more recent estimate, calculated using a model for the consumption of atmospheric CH4 by soil, estimated that 24.8-28 Tg of CH4 are consumed annually by soil (Curry, 2009). The model projects that the global annual mean CH4 soil sink should increase by 8% by 2100. The largest absolute increases are expected to occur in cool temperate and subtropical forest ecosystems and would be associated with increasing temperature and reduced water stress. However, the largest relative increases in consumption (>40%) would occur in the boreal forest, tundra and polar desert environments of the high northern latitudes (Curry, 2009). These increases are expected to be linked to enhanced soil diffusivity of CH4 resulting from lower soil moisture.

10 2.2 Microbes and microbial processes associated with the CH4 cycle in cold environments

2.2.1 Methanogenic archaea

Methanogenic archaea, the microorganisms responsible for CH4 production under anaerobic conditions, are important members of the microbial communities in cold environments (Cavicchioli, 2006). They are known to inhabit permafrost environments (Rivkina et al., 2007; Simankova et al., 2003) and their capacity to produce CH4 at cold temperatures has been reported (Ganzert et al., 2007; Metje and Frenzel, 2007; Rivkina et al., 2007; Wagner et al., 2005).

Methanogenic archaea, which belong to the phylum Euryarchaeota, produce CH4 through two main pathways: (1) hydrogenotrophic methanogenesis, in which CO2 is reduced to CH4 using dihydrogen as a reductant, and (2) acetoclastic methanogenesis, which involves the fermentation of acetate to CH4 and CO2. Both pathways, as well as gene sequences related to microorganisms associated with hydrogenotrophic and acetoclastic methanogenesis, have been detected in permafrost environments (Ganzert et al., 2007; Metje and Frenzel, 2007). So far, only a few psychrophilic and psychrotolerant strains of methanogens have been isolated from cold environments and growth temperatures as low as –2.3°C have been reported for these microorganisms (Wagner and Liebner, 2010). The level of methanogenic activity and the archaeal biomass were found to be well correlated to the soil CH4 concentration in permafrost environments (Wagner et al., 2007), indicating that methanogenic archaea are producing CH4 under in situ conditions. The abundance and composition of the methanogenic population in permafrost environments were shown to be similar to that of methanogenic communities of temperate soil ecosystems (Wagner et al., 2005). The highest abundance of methanogenic archaea in permafrost environments was detected in the active layer, with populations of up to 3 x 108 cells g-1 soil (Kobabe et al., 2004). Yergeau et al. (2010) used metagenomic analysis to characterize the microbial population in active layer and 2m permafrost samples from the

Canadian high Arctic, with a focus on microorganisms implicated in the CH4 cycle. Methanogens were detected in both samples, with a dominance of the

11 genus Methanomicrobia at both sampling depths (30.1% and 38.2% of the archaeal sequences retrieved from the active layer and the permafrost, respectively). Sequences related to the genera Methanobacteria, Methanococci and Methanopyri were also detected in these samples (Yergeau et al., 2010). Methanogens have been detected in other types of cold environments besides permafrost (Wadham et al., 2008). Active methanogens related to the order Methanosarcinales were found in subglacial sediments from Robertson

Glacier, in the Canadian Rockies (Boyd et al.). In a study looking at the CH4 concentration at different depths in a 3,053 m deep ice core from the Greenland ice sheet, higher CH4 concentrations were detected in the lowest 90 m of the core, and these high concentrations were associated with the presence of methanogenic archaea (Tung et al., 2005). Sequences related to methanogenic archaea, as well as CH4 of biological origin, were found in deep marine sediments, also characterized by their cold temperatures (Inagaki et al., 2006). Finally, methanogens have been shown to represent a great proportion of the microbial population in cold marine waters and can also be found in sea ice (Cavicchioli, 2006).

2.2.2 Anaerobic oxidation of methane Anaerobic oxidation of methane (AOM) is an important process in oceans, where over 90% of the 85–300 Tg of CH4 produced annually by methanogenesis would be consumed by anaerobic methanotrophic archaea (ANME) (Knittel and Boetius, 2009). Anaerobic methanotrophic archaea are a group of Euryarchaeota that acquire energy exclusively from the AOM with sulphate or other electron acceptors as the final electron acceptor. Anaerobic oxidation of methane has been described for different oceanic ecosystems, including hydrothermal vents and cold seep ecosystems. Anaerobic oxidation of methane was also observed in terrestrial habitats like mud volcanoes, landfills and anoxic freshwater lakes (Knittel and Boetius, 2009). Recently, sequences related to the ANME-1 cluster have also been detected in a subzero hypersaline methane seep in the Canadian high Arctic

(Niederberger et al., 2010).

12 The biochemical process of AOM remains unclear and, to date, no ANME have been isolated. One hypothesis on the AOM process is generally accepted:

ANME would utilize the enzymes of methanogenesis in reverse for CH4 oxidation and provide electrons to an autotrophic sulfate-reducing partner (Knittel and Boetius, 2009). Although AOM has been seen exclusively as a sulfate-dependent process for several years, other electron acceptors are more energetically favourable. In a study looking at AOM in marine methane seep sediments from the Eel River Basin in California, it was shown that microorganisms from this environment can utilize manganese and iron to oxidize CH4, indicating that marine AOM can be associated with a larger variety of oxidants than previously thought (Beal et al., 2009). Because manganese and iron are largely discharged into oceans by rivers, AOM using these electron acceptors might be important in the global CH4 cycle.

A novel pathway for CH4 oxidation, nitrite-driven AOM by oxygenic bacteria, was recently described (Ettwig et al., 2010). This process is conducted by the newly described bacterium Methylomirabilis oxyfera, an apparently anaerobic denitrifying bacterium which encodes, transcribes and expresses the aerobic pathway for CH4 oxidation. It was shown that, to achieve CH4 oxidation, M. oxyfera bypasses the denitrification intermediate nitrous oxide by the conversion of two nitric oxide molecules to dinitrogen and oxygen, which is then used to oxidize CH4 (Ettwig et al., 2010).

2.2.3 Aerobic methanotrophic bacteria Methane oxidation in aerobic ecosystems is due to the activity of methanotrophic bacteria, which utilize CH4 as a sole carbon and energy source. The main phyla of methanotrophic bacteria and their characteristics are presented in Table 2.1. Most known aerobic methanotrophs are divided into two major groups (Type I and Type II) based on phylogeny and carbon assimilation pathways (Bowman, 2006). Type I methanotrophs (or Gammaproteobacteria methanotrophs) belong to the family Methylococcaceae in the Gammaproteobacteria subdivision, while Type II methanotrophs (or

13 Alphaproteobacteria methanotrophs) belong to the families Methylocystaceae and Beijerinckiaceae in the Alphaproteobacteria subdivision (Bowman, 2006; Op den Camp et al., 2009). In addition to these two groups, methanotrophic bacteria belonging to the phylum Verrucomicrobia have been recently described (Dunfield et al., 2007; Op den Camp et al., 2009). Methanotrophs possess the enzyme methane monooxygenase (MMO), responsible for the oxidation of CH4 to methanol. The MMO exists in two forms: the membrane-bound particulate methane monooxygenase (pMMO) and the cytoplasmic soluble methane monooxygenase (sMMO). The pMMO has been detected in all methanotrophs with the exception of Methylocella spp., while the sMMO is expressed under low copper concentrations and is found in some methanotrophs, including Methylocella spp. Because of its occurrence in most known methanotrophs, the gene encoding the α-subunit of the pMMO (pmoA) is widely used as a biological marker to study methanotrophic bacteria.

The CH4 oxidation pathway

The CH4 oxidation pathway as found in aerobic methanotrophs is presented in Figure 2.2 (Hanson and Hanson, 1996). The first step of this pathway is the oxidation of CH4 to methanol through the activity of the sMMO or the pMMO (see previous section for more details about these enzymes). The methanol is then oxidized to formaldehyde by the enzyme methanol dehydrogenase. The formaldehyde can be assimilated into organic carbon by the ribulose monophosphate pathway, found in Type I methanotrophs, or the serine pathway, found in Type II methanotrophs. Methanol can also be further oxidized to formate and finally to CO2 through the activity of formaldehyde dehydrogenase and formate dehydrogenase, respectively. These two final steps generate the reducing power required for the metabolism of CH4.

14 Table 2.1 - Comparison of all described families of aerobic methanotrophic bacteria (adapted from Op den Camp et al., 2009).

Group Type I Type II n.a. Phylum and class Proteobacteria (Gammaproteobacteria) Proteobacteria (Alphaproteobacteria) Verrucomicrobia Family Methylococcaceae Methylocystaceae Beijerinckiaceae Methylacidiphilaceae Genera Methylococcus, Methylocaldum, Methylocystis, Methylosinus Methylocella, Methylocapsa Methylacidiphilum Methylohalobius, Methylothermus, Methylobacter, Methylomicrobium, Methylomonas, Methylosarcina, Methylosoma, Crenothrix, Clonothrix, Methylosphaera Internal membranes or Membrane bundles perpendicular to the Membrane stacks along the cell Methylocapsa: single membrane Carboxysome-like structures compartments cell envelope periphery, parallel to the cell stack parallel to the long axis or vesicular membranes envelope Methylocella: cytoplasmic membrane invaginations or vesicles Lowest reported growth pH 5.0 (many species) 4.4 (Methylocystis heyeri) 4.2 (Methylocapsa acidiphila) 0.8 (M. fumariolicum SolV) Highest reported growth pH 11 (Methylomicrobium buryatense) 7.5 (Methylocystis heyeri) 7.5 (Methylocella spp.) 6.0 (M. infernorum V4) Lowest growth temp. (°C) 3.5 (Methylobacter psychrophilus) 5.0 (Methylocystis heyeri) 4.0 (Methylocella silvestris) 37 (M. kamchatkense Kam1) Highest growth temp. (°C) 67 (Methylothermus thermalis) 40 (many species) 30 (Methylocella spp.) 65 (M. fumariolicum SolV) Major PLFAs (more than 14:0 (1–24%) 16:1ω8c (0–41%) 16:1ω8c (0–29%) 18:1ω7c (59–83%) i14:0 (7–22%) 15% of total in any species) 16:0 (4–63%) 18:1ω7c (0–60%) 18:1ω7c (10–37%) a15:0 (13–31%) 16:1ω5t (0–30%) 18:1ω9c (0–35%) 18:1ω8c (32–74%) 18:0 (14–42%) 16:1ω7c (8–57%) Carbon fixation pathway Ribulose monophosphate pathway, Serine cycle Serine cycle, Calvin-Benson- Serine cycle, Calvin– Calvin– Benson–Bassham cycle (rarely) Bassham cycle Benson–Bassham cycle G+C mol% 43–65 60–67 60–63 40.8–45.5

N2 fixation +/− +/− + +/− sMMO +/− +/− +/− − pMMO + + +/− + Obligately methylotrophic + + +/− + n.a., not applicable; G+C mol%, proportion of guanine and cytosine in DNA molecules; PLFAs, phospholipid fatty acids; + all species positive, - all species negative, +/- some species positive.

15

Figure 2.2 - Pathways for the oxidation of CH4 and assimilation of formaldehyde. sMMO, soluble methane monooxygenase; pMMO, particulate methane monooxygenase; MDH, methanol dehydrogenase; CytC, cytochrome c; FADH, formaldehyde dehydrogenase; FDH, formate dehydrogenase. Source: (Hanson and Hanson, 1996).

High affinity methane-oxidizing bacteria

While the process of atmospheric CH4 uptake by soils is well described (Dunfield, 2007) (see section 2.1.4) and has been known for several years, the bacteria implicated in the oxidation of CH4 at atmospheric concentrations (~1.7 ppm) are not well characterized. Early studies suggested that methanotrophs can be divided into low affinity methanotrophs, related to cultured bacteria having the capacity to oxidize CH4 at high concentrations, and high-affinity methanotrophs, related to uncultured bacteria having the capacity to oxidize CH4 at atmospheric concentration (Bender and Conrad, 1992). These atmospheric methane-oxidizing bacteria would be specialized oligotrophs adapted to low CH4 concentrations and possessing a MMO with a higher substrate affinity compared to most cultivated methanotrophic bacteria (Bender and Conrad, 1992). Recent studies have shown that bacteria from the genus Methylocystis produce two isozymes of the pMMO with different oxidation kinetics associated with low and high CH4 affinity, respectively (Baani and Liesack, 2008). These bacteria have the capacity to

16 sustain oxidation at low CH4 concentrations, but they might require additional energy sources for long-term survival and growth (Baani and Liesack, 2008; Knief and Dunfield, 2005). Particulate methane monooxygenase gene sequences from uncultivated bacteria are frequently detected in soils having the capacity to degrade atmospheric CH4 (Angel and Conrad, 2009; Holmes et al., 1999; Knief et al., 2003; Kolb, 2009). These uncultured putative atmospheric methane-oxidizing bacteria can mainly be separated into two groups: one being related to the Alphaproteobacteria and referred to as the “upland soil cluster alpha” (USCα) and the other being related to the Gammaproteobacteria methanotrophs and referred to as the “upland soil cluster gamma” (USCγ) (Kolb, 2009). The USCα genotype is widely detected in forest soils, where it often dominates the methanotrophic bacterial populations (Kolb, 2009). Sequences related to this group were also found in other soil types, including grassland (Horz et al., 2005) and acidic peat (Chen et al., 2008a). The USCγ genotype, which has been less frequently detected in soils, was first described in a study looking at the methanotrophic bacterial diversity in 35 different upland soils (Knief et al., 2003). More recently, pmoA transcripts related to USCγ were found to be dominant in a desert soil (Angel and Conrad, 2009). A third group of pmoA gene sequences, closely related to the ammonia monooxygenase gene (amoA) of ammonia-oxidizing bacteria, has been detected in some forest soils with atmospheric methane-oxidizing capacities (Knief et al., 2006; Kolb et al., 2005). These amoA-related sequences belong to a methanotrophic bacterial group referred to as “Cluster 1” (Kolb, 2009), which also includes bacterial isolates from Arctic soils with highly divergent pmoA genes (Pacheco-Oliver et al., 2002). However, the presence of these bacterial isolates in Arctic soils has not been associated with high-affinity methane- oxidizing activity (Pacheco-Oliver et al., 2002).

17 Facultative methanotrophic bacteria Although most known methanotrophic bacteria are restricted to one- carbon compounds as carbon and energy sources, a few examples of cultivated methanotrophic bacteria having the potential to use multicarbon compounds were recently described. Members of the genus Methylocella, which are the only methanotrophs that do not possess the pMMO, were found to have the capacity to use as their sole energy source the one-carbon compounds CH4 and methanol, as well as the multicarbon compounds acetate, pyruvate, succinate, malate, and ethanol (Dedysh et al., 2005). More recently, the bacterium Methylocapsa aurea was also described as a facultative methanotroph having the capacity to utilize acetate as a growth substrate (Dunfield et al., 2010). The isolate belongs to the family Beijerinckiaceae of the class Alphaproteobacteria and is most closely related to the obligate methanotroph Methylocapsa acidiphila. Crenothrix polyspora is another facultative methanotroph due to its capacity to utilize acetate or glucose in the absence of CH4 (Stoecker et al., 2006).

2.2.4 Ecology of aerobic methanotrophic bacteria Aerobic methanotrophic bacteria are widely distributed in nature and are abundant at the interface of anoxic and oxic environments, where they reduce CH4 emissions to the atmosphere. Both cultivation and cultivation independent methods have been used to study their distribution in various environments, such as rice fields, landfills, freshwater sediments and upland soils. Most methanotrophic bacterial isolates are neutrophilic, although acidophilic and alkaliphilic species have been described (Table 1.1). Members of the Verrucomicrobia methanotrophs can grow at pH values below 1, while the Type I methanotroph Methylomicrobium buryatense can tolerate pH’s as high as 11 (Op den Camp et al., 2009). Methanotrophs are also generally mesophilic, but some psychrophilic methanotrophs (growth optimum below 15°C), like Methylobacter psychrophilus (Omelchenko et al., 1996) and Methylosphaera hansonii (Bowman et al., 1997), have been isolated from cold environments. Although true thermophilic methanotrophs (growth optimum above 60°C) have not been isolated

18 to date, several isolates from the genera Methylothermus, Methylococcus and Methylocaldum have growth optima above 42°C (Trotsenko and Khmelenina, 2002; Tsubota et al., 2005). While a wide diversity of both Type I and Type II methanotrophs, mainly members of the genera Methylomonas, Methylobacter, Methylomicrobium, Methylococcus, Methylocaldum, Methylocystis and Methylosinus, have been detected in rice fields and landfills, freshwater environments are generally characterized by the dominance of Type I over Type II methanotrophs (Chen and Murrell, 2010). Type I methanotrophs were 1 to 2 orders of magnitude more abundant than Type II methanotrophs in Lake Washington sediment (Costello et al., 2002) while 90% of the clones in pmoA clone libraries from Lake Constance were Type I methanotrophs (Rahalkar and Schink, 2007). The genera Methylomonas, Methylobacter, Methylosarcina, Methylococcus, and Methylosoma are typically found in these environments. A similar trend is generally reported for methanotrophic bacteria in cold environments (see section 2.2.6 for more details). Methanotrophic bacteria have also been isolated from several extreme environments, including hot springs, volcanic mud and soda lakes (Chen and Murrell, 2010).

2.2.5 Effect of soil parameters on methanotrophic bacterial community structure and CH4 oxidation

Temperature Methanotrophic bacterial community structure and activity can be affected by different soil characteristics as well as by the environmental conditions.

Because it is an enzymatic process, CH4 oxidation is strongly affected by temperature, with generally higher rates of oxidation being observed at higher soil temperatures (Einola et al., 2007). A seasonal effect, which mainly reflects modifications to the soil temperature, was also found to have an impact on the methanotrophic bacterial community structure in Alpine soils (Abell et al., 2009). The lowest abundance of both Type Ia and Type II methanotrophs as assessed by real-time PCR (qPCR) was observed in the winter, while the highest abundance of

19 Type Ia methanotrophs occurred in autumn and the highest abundance of Type II methanotrophs occurred in summer (Abell et al., 2009).

Soil pH As is the case for bacterial communities in general, soil pH plays an important role in structuring methanotrophic bacterial communities. In a study looking at the methanotrophic bacterial diversity from 35 different upland soils, including forests, meadows and farmlands, a correlation between the occurrence of the putative atmospheric methane-oxidizing genotypes USCγ and USCα and the soil pH was observed, with USCγ sequences being detected more frequently in soils with pH values above 6.0 than in more acidic soils and the opposite trend being observed for USCα sequences (Knief et al., 2003). Members of Cluster I were also reported as important members of methanotrophic bacterial populations in neutral soils with atmospheric methane-oxidizing activity (Kolb, 2009; Kolb et al., 2005).

Nitrogen

Several studies have shown the importance of nitrogen in regulating CH4 oxidation (Bodelier and Laanbroek, 2004). The inhibitory effect of the addition of nitrogen-based fertilizer on soil CH4 oxidation was first reported for forest soils, where a reduction of CH4 uptake of 33% was observed when ammonium nitrate was added (Steudler et al., 1989). Further studies have shown that the inhibition was mainly associated with ammonia-based fertilizers (King and Schnell, 1994), while nitrate has been found to be inhibitory at very high concentrations only (Bodelier and Laanbroek, 2004). Several mechanisms have been proposed to explain the inhibitory effect, including the osmotic stress due to salt addition and the competitive inhibition of MMO by ammonia. A positive effect of nitrogen addition to soil, even in the form of an ammonia-based fertilizer, was also reported in several studies (Bodelier and Laanbroek, 2004). An example of that is the strong increase in CH4 oxidation in a rice field soil following ammonia amendment, which was also associated with the specific enrichment of Type I

20 methanotrophs from the genera Methylomicrobium and Methylocaldum (Noll et al., 2008). Nitrogen is an essential element to bacterial growth and its addition to nitrogen-limited environments can have, therefore, a positive effect on methanotrophic bacterial activity.

Methane concentration

The CH4 oxidation process is strongly affected by the CH4 concentration, with generally increasing CH4 oxidation rates being observed with increasing CH4 concentrations until saturation is reached (Pawlowska and Stepniewski, 2006). Methane concentration can also affect the methanotrophic bacterial taxa found in a specific environment. It has been previously suggested in the literature that Type

I methanotrophs are dominant in environments with moderate CH4 concentrations while Type II methanotrophs are dominant at higher CH4 concentrations (Hanson and Hanson, 1996). Recent studies have shown that this distribution pattern is modified at CH4 concentrations below 1,000 ppm. Incubation of hydromorphic 13 soils at low (5-30 ppm) or high (430-570 ppm) CH4 concentrations resulted in significant modifications to the pattern of 13C-labelled phospholipid fatty acids (PLFAs), with PLFAs associated with Type II methanotrophs mainly being labelled at low CH4 concentrations and PLFAs associated with both Type I and

Type II methanotrophs being labelled at high CH4 concentrations (Knief et al.,

2006). Uncultivated and cultivated Type II methanotrophs were found to be responsible for most of the CH4 oxidation at low CH4 concentrations (< 30 ppm) while the activity of Type I methanotrophs was stimulated by elevated CH4 concentrations (500 ppm) (Knief et al., 2006). In a study performed on bacteria from the genera Methylocystis, Methylosinus, Methylocaldum and Methylobacter isolated from upland soils, two Methylocystis strains required the lower CH4 concentrations for growth (10–100 ppmv), while Type I methanotrophs generally needed higher CH4 concentrations for growth (Knief and Dunfield, 2005). Type II methanotrophs, therefore, appear to be more oligotrophic than Type I methanotrophs.

21 Global changes Global changes are expected to have a strong impact on several environmental parameters that could potentially affect methanotrophic bacterial populations. The effect of four global change-associated modifications (CO2 concentration, temperature, precipitation and nitrogen deposition) applied individually or in combination on methanotrophic bacterial populations was tested in California upland grassland soils (Horz et al., 2005). None of the treatments significantly affected the overall diversity of the methanotrophic bacterial population. However, the simulated global changes did modify the community composition, with the relative abundance of Type II methanotrophs significantly decreasing under elevated precipitation and elevated temperature. In contrast, the relative abundance of a novel methanotroph clade related to the genotype USCγ and referred to as “JR2” increased with elevated precipitation and temperature (Horz et al., 2005).

2.2.6 Methanotrophic bacteria in cold environments Methanotrophic bacterial isolates from cold environments Methanotrophic bacteria were isolated from various cold environments such as northern tundra soils (Dedysh et al., 2004; Omelchenko et al., 1996; Pacheco-Oliver et al., 2002), polar lakes (Bowman et al., 1997) and permafrost (Trotsenko and Khmelenina, 2005). The first psychrophilic methanotroph was recovered from a tundra soil of the Polar Urals and was classified as Methylobacter psychrophilus (Omelchenko et al., 1996). Two other methanotrophic bacteria, Methylosphaera hansonii and Methylomonas scandinavica, were described as true psychrophiles and were isolated from a saline Antarctic meromictic lake and a Swedish deep igneous rock aquifer, respectively (Trotsenko and Khmelenina, 2002). All other methanotrophic bacteria isolated from cold environments are psychrotolerant rather than psyschrophilic, with optimal growth temperatures above 20°C (Trotsenko and Khmelenina, 2005). A species closely related to M. psychrophilus was isolated from an Arctic wetland soil of the Svalbard Islands, Norway (Wartiainen et al.,

22 2006). Even though this new species, designated as Methylobacter tundripaludum, shared high similarity of the 16S rRNA gene with M. psychrophilus, results of DNA hybridization and biochemical tests genotypically and phenotypically differentiated this bacterium from other known Methylobacter species. Acidic environments are abundant in cold regions of the Northern Hemisphere and methanotrophic bacteria adapted to these environments have also been isolated (Dedysh et al., 2004; Dedysh et al., 2002; Dedysh et al., 2000; Dunfield et al., 2003). These bacteria were classified as four new species of two new genera, Methylocella (Ml. palustris, Ml. silvestris, Ml. Tundrae) and Methylocapsa (Ma. acidiphila), representing a novel lineage of methanotrophic bacteria in the family Beijerinckiaceae of the Alphaproteobacteria. Methanotrophic bacteria distantly related to all previously know methanotrophs were also isolated from Canadian Arctic tundra soils (Pacheco-Oliver et al., 2002). Phylogenetic analysis of the pmoA genes amplified from the isolates revealed novel pmoA gene sequences with low similarities to the pmoA gene from known bacteria. These novel pmoA genes were more closely related to the amoA gene of ammonia-oxidizing bacteria than to the pmoA gene of methanotrophs. Although they were not fully characterized, analysis of the 16S rRNA gene of these isolates showed that they were distantly related to Type II methanotrophs from the genera Methylosinus and Methylocystis (Pacheco-Oliver et al., 2002).

Methanotrophic bacterial activity, abundance and diversity in cold environments Methanotrophs inhabit different types of aquatic and terrestrial cold ecosystems, as reviewed by Trotsenko and Khmelenina (2005). Methanotrophs were detected in the sediments and water column of different meromictic lakes in Antarctica, but their number, as a proportion of total bacteria, was lower than in temperate lakes (Bowman et al., 1997). Permanently cold deep ground waters from Northern countries are also commonly found to harbour methanotrophic bacterial populations, with a clear dominance of type I methanotrophs (Trotsenko and Khmelenina, 2005).

23 A few studies were conducted on methanotrophs in soils from cold environments located in Northern Europe and Siberia (Dedysh et al., 1998; Liebner et al., 2009; Liebner and Wagner, 2007; Omelchenko et al., 1993; Vecherskaya et al., 1993; Wagner et al., 2005; Wartiainen et al., 2003). Among the most studied sites are the low-centered ice-wedge polygons from the Lena

Delta, Siberia (N 72°22′, E 126°28′), where several studies were conducted to assess the activity, distribution, abundance, and diversity of methanotrophic bacteria (Liebner et al., 2009; Liebner and Wagner, 2007; Wagner et al., 2003). Methane oxidation rates in soils from the Lena Delta were affected differently by the temperature, with maximum CH4 oxidation rates detected at 21°C in the upper active layer and at 4°C in the deep active layer zones, indicating a dominance of psychrophilic methanotrophs close to the permafrost (Liebner and Wagner, 2007). A maximum abundance of 2 x 108 methanotrophic bacteria g-1 soil was reported for the active layer soil from this site (Liebner and Wagner, 2007). Type I methanotrophs were dominant throughout the active layer but their abundance was variable according to depth, while Type II methanotrophs were more abundant close to the permafrost (Liebner and Wagner, 2007; Wagner et al., 2005). Sequence analysis of the 16S rRNA and pmoA genes showed that the methanotrophic bacterial diversity in these soils is similar at both sampling depths and is restricted to the genera Methylobacter and Methylosarcina, both members of the Type I methanotrophic bacteria (Liebner et al., 2009). The dominance of Type I methanotrophs observed in soils from the Lena Delta was also confirmed in several other soils from cold environments. Methanotrophic bacteria from the genus Methylobacter, followed by Methylomonas and Methylocystis, were found to be the most abundant bacteria by immunofluorescent microscopy in Russian tundra bog soils in which CH4 oxidation was detected throughout the active layer (Vecherskaya et al., 1993). Using primers specific to the 16S rRNA gene of Type I and Type II methanotrophs, most Type I methanotrophs detected in three different wetland soils from the islands of Svalbard, Norway were related to Methylobacter while detected Type II methanotrophs were related to Methylosinus (Wartiainen et al.,

24 2003). Type I methanotrophs were detected in metagenomic libraries from an active layer sample and a 2 m permafrost sample from the Canadian high Arctic, while no genes related to Type II methanotrophs were detected (Yergeau et al., 2010). Quantification of the 16S rRNA gene of types I and II methanotrophs in these soil samples also indicated that Type I methanotrophs were dominant (Yergeau et al., 2010). Although very little information on the potential of methanotrophic bacteria to mitigate CH4 emissions under in situ conditions in cold terrestrial environments is available, the results from one study conducted in northern

Canada (54.6°N) indicated that biological CH4 oxidation can reduce CH4 emissions from 58 to 92% according to in vitro experiments and approximately 20% based on in situ experiments (Popp et al., 2000). Moreover, in situ consumption of CH4 at concentrations ranging from atmospheric (1.7 ppm) to 500 ppm by low latitude (53° N) tundra soils at a temperature of 7°C has been reported (Whalen and Reeburgh, 1990). These results indicate that methanotrophic bacteria could play a significant role in mitigating CH4 emissions from melting permafrost and could also be an important sink for atmospheric

CH4.

2.3 Microbes and microbial processes associated with soil respiration in cold environments

While microorganisms implicated in the CH4 cycle belong to defined groups of bacteria and archaea, microorganisms implicated in the CO2 cycle, and more precisely in soil respiration, are extremely diverse and belong to all groups of bacteria, archaea and fungi. Therefore, this section will focus exclusively on the microbial populations in Arctic soils and on the effect of the soil parameters on microbial community structure and soil respiration. Although some information is provided on archaea and fungi in this section, soil bacteria are mainly covered.

25 2.3.1 Microbial abundance and diversity in Arctic soils Microbial cell counts as high as 109 cells g-1 of soil have been reported in the Arctic (Hansen et al., 2007), but significant variations are observed from site to site. Moreover, a large proportion of these cells are non-viable due to the harsh conditions of the Arctic environment (Steven et al., 2009). Approximately 70 bacterial isolates with a wide range of metabolic capacities, including aerobic and anaerobic heterotrophy, sulfate-reduction, methanotrophy, and methanogenesis, have been recovered from permafrost environments (Steven et al., 2009). Members of the Gram-positive phyla Firmicutes and Actinobacteria are among the most frequently isolated bacteria from cold environments (Steven et al., 2008). Bacterial isolates from cold environments have some distinctive characteristics compared to isolates from other environments. They are generally well adapted to cold temperatures and growth at subzero temperatures has been reported (Steven et al., 2007; Steven et al., 2008), although most isolates are psychrotolerant rather than psychrophilic. Bacterial isolates from permafrost environments are also often halotolerant, a characteristic that likely allows them to circumvent the low water availability characteristic of frozen environments (Steven et al., 2009). Cold temperatures have a strong impact on bacterial structure and activity, and bacterial cells present several adaptative features in response to cold (Rodrigues and Tiedje, 2008; Steven et al., 2009). A widely used strategy to sustain cell activity at low temperatures is to produce cold-adapted enzymes with enhanced catalytic efficiency. Psychrophilic enzymes are characterized by their high flexibility, which is the result of different features including higher numbers of hydrophobic side chains and weaker intramolecular bonds. Increased membrane fluidity, as conferred by an increase in unsaturated fatty acids and a higher proportion of short chain fatty acids, is another common characteristic observed in microbes exposed to cold temperatures. Exposure to cold is also known to induce modifications to the proteins being synthesized by the cell, and the proteins expressed following exposure to cold are referred to as cold adapted proteins (Hebraud and Potier, 1999). Chaperones and pyruvate

26 dehydrogenases are proteins that are commonly associated with cold adaptation (Rodrigues and Tiedje, 2008). Culture-independent studies performed on soils from permafrost environments have shown that a wide diversity of bacteria and archaea inhabit these environments. The Actinobacteria and Proteobacteria are often the dominant members of the bacterial communities (Steven et al., 2007; Vishnivetskaya et al., 2006; Yergeau et al., 2010), while 16S rRNA gene sequences related to both major archaeal phyla, the Euryarchaeota and the Crenarchaeota, have been observed (Steven et al., 2007; Steven et al., 2008). Some members of the Actinobacteria are recognized for their capacity to degrade cellulose, chitin and other complex C compounds (Yergeau et al., 2010). In addition, several classes of Proteobacteria are known for their ability to rapidly turnover soil carbon (see section 2.3.2). Consequently, these bacterial phyla might play an important role in organic matter degradation in cold environments.

Comparison between active layer and permafrost Soils from Arctic environments are characterized by two distinct zones: the active layer and the permafrost (see section 2.1.1 for definitions). Several differences in the environmental conditions prevailing in these two soil zones can be observed. The active layer is subject to freeze and thaw cycles, with temperatures fluctuating between -35 and +15°C, while temperatures in the permafrost are constantly below 0°C (-12°C to -20°C in the Arctic), with <1°C of seasonal fluctuation in the top 10 m of permafrost (Gilichinsky et al., 2008; Vishnivetskaya et al., 2006). Winter freezing of thawed soil in the active layer creates fissures in the soil, enabling gas diffusion, while the permafrost layer acts as a physical barrier limiting gas diffusion (Steven et al., 2007). The water availability can be variable in the active layer of soil and can reach saturation during the spring thaw, while the availability of liquid water in permafrost is low (Vishnivetskaya et al., 2006). Because of these important differences in environmental conditions between the active layer and the permafrost, differences in the microbial

27 communities inhabiting these environments are expected. Indeed, the viable cell counts and the diversity of the isolates from 2 m permafrost from the Canadian high Arctic were lower than what was found in the active layer (Steven et al., 2008). Based on data from different studies, Gilichinsky et al. (2008) also reported an average 100-fold decrease in cells in the permafrost layer. Another example of the differences observed between the active layer samples and samples from deeper soil zones is the higher abundance of some bacterial groups, like Actinobacteria and Firmicutes, near the permafrost table (Wagner et al., 2009) or in the permafrost (Yergeau et al., 2010). It was suggested that the higher abundance of these bacterial groups in the permafrost might be related to their better capacity to endure the harsher conditions found in deeper soil zones (Yergeau et al., 2010). Although important differences in microbial abundance between the active layer and the permafrost have been reported, some studies have suggested that minimal distinctions in microbial diversity and community structure occur between these two environments. Steven et al. (2008) found that the microbial diversity, as assessed by clone libraries of the 16S rRNA gene, was similar in the active layer soil, permafrost table and permafrost horizons of a core from the Canadian high Arctic. In a study looking at the phylogenetic and functional community composition in an active layer sample and a 2 m deep permafrost sample from the same site, Yergeau et al. (2010) found similar community compositions at both sampling depths. The Actinobacteria were the dominant group in both samples, with a codominance of the Betaproteobacteria being observed in the 2 m permafrost (Yergeau et al., 2010). The functional community composition was also similar at both sampling depths, and the authors indicated that the microorganisms present in the permafrost at this site can be seen as a representative subset of the microorganisms present in the active layer (Yergeau et al., 2010).

28 Comparison between Arctic soils and soils from lower latitudes Microbial communities in Arctic soils are exposed to particularly severe environmental stresses and, thus, these soils are expected to harbor unique bacterial communities. However, information from an increasing number of sites indicate that the bacterial community structure in Arctic soils might not be so different from the bacterial community structure of soils from lower latitudes (Chu et al., 2010; Neufeld and Mohn, 2005). The structure of the bacterial community was found to be similar in Arctic tundra soils and boreal forest soils, with a dominance of Proteobacteria and high abundances of Actinobacteria, Acidobacteria, Firmicutes, Bacteroidetes, Verrucomicrobia, and Cyanobacteria (Neufeld and Mohn, 2005). Moreover, a greater bacterial diversity was found in the Arctic tundra soils than for boreal forest soils, with the highest diversity being detected in the soil from the most northern location (82°N) (Neufeld and Mohn, 2005). Recently, Chu et al. (2010) performed an extended study of the bacterial diversity in soils collected from 29 heath tundra sites located close to the top of exposed ridges in the Canadian, Alaskan and European Arctic using a high resolution bar-coded pyrosequencing technique. The bacterial community composition and diversity of the Arctic soils was compared to the bacterial community composition in 85 soils from a wide range of lower latitude sites. Bacterial communities were found to be as variable within Arctic soils as across the lower latitude soils, while the phylotype richness and phylogenetic diversity in Arctic soils were not significantly lower than in the lower latitude soils. The authors concluded that, although environmental stresses are particularly severe in Arctic tundra, the bacterial community structure of Arctic soils does not appear as fundamentally distinct from that of lower latitude soils (Chu et al., 2010).

2.3.2 Effect of soil parameters on soil bacteria and soil respiration Temperature Temperature is known to have a strong impact on microbial respiration. Although increasing respiration rates with increasing temperatures have been reported in many soil warming experiments (Hartley et al., 2008; Reth et al.,

29 2005; Rustad et al., 2001), the initial positive response often declines with time (Eliasson et al., 2005; Luo et al., 2001; Melillo et al., 2002; Rustad et al., 2001). Because alterations to the microbial community structure following soil warming or in response to seasonal changes were observed in several studies (Lipson and Schmidt, 2004; Pietikäinen et al., 2005; Wallenstein et al., 2007; Zhang et al., 2005), it has been suggested that this phenomenon of acclimatation was related to changes in microbial respiration due to modifications to the microbial community composition (Hartley et al., 2008). Other hypotheses that could explain the transitory effect of warming on soil respiration include a decay in the labile soil cabon pool (Melillo et al., 2002) and a physiological response of microbes to warming leading to a reduced carbon-use efficiency (Allison et al., 2010). In a warming experiment on subarctic soils, as opposed to what is generally observed, the acclimatation process did not occur, and the medium term effect of temperature increases on soil respiration was to further increase rates instead of leading to a decline (Hartley et al., 2008). These results indicate that, at least in the medium term, the warming-induced changes in the microbial community in

Arctic soils might lead to increasing rates of CO2 production and, thus, to important losses of the carbon stocked in soils from cold environments. Besides its direct impact on soil respiration, increasing temperatures in the

Arctic will affect several factors that could influence the net CO2 flux from terrestrial ecosystems in this region. Climate warming can stimulate plant growth by increasing the photosynthetic rates and extending the length of the growing season (Myneni et al., 1997). Climate warming can also induce modifications to the plant community composition and distribution, with increased growth of shrubs and grasses (Dormann and Woodin, 2002; Walker et al., 2006), a decrease of cover by mosses and lichens (Walker et al., 2006), and decreased species diversity and evenness (Walker et al., 2006). These modifications to the plant community composition could in turn affect the CO2 fluxes from Arctic soils. Higher temperatures and decomposition rates can also increase nutrient availability (Chapin et al., 1995), which is also known to have an effect on plant growth and soil respiration. Since nitrogen is the main limiting factor in Arctic

30 ecosystems (Jonasson et al., 1999; Tarnocai and Campbell, 2006), increased nitrogen inputs following warming could stimulate plant growth and carbon storage. However, microbes are an important sink for nutrients and microbial immobilization was shown to limit plant uptake when nutrients are in low quantities (Jonasson et al., 1999). Previous studies in Arctic soils reported a moderate impact of warming on the nitrogen cycle, influenced by time, soil moisture, and plant cover (Deslippe et al., 2005; Jonasson et al., 1999; Shaw and Harte, 2001).

Soil pH Soil pH was previously described as the main factor influencing the bacterial community composition in different environments including arable soils (Baker et al., 2009; Rousk et al., 2010) and wetland soils (Hartman et al., 2008), but also across a wide range of ecosystems at the continental scale (Fierer and Jackson, 2006; Lauber et al., 2009). As observed in other soil types, the variation in the bacterial community in Arctic soils was found to be strongly related to pH but not to other soil characteristics (Chu et al., 2010). The relative abundance of the main bacterial groups varied according to soil pH, with positive correlations being observed for the Alphaproteobacteria, Actinobacteria, Betaproteobacteria and Bacteroidetes but negative correlation being observed for the Acidobacteria. As previously suggested by Rousk et al. (2010), one hypothesis that could explain the importance of pH in structuring soil bacterial communities might be related to the narrow pH range at which most bacterial taxa can grow. Although the relationship between soil pH and respiration is not as clear as it is for parameters like temperature and organic matter content, a positive correlation between soil pH and soil respiration was reported in some studies (Andersson and Nilsson, 2001; Reth et al., 2005) and is most probably related to the strong impact of soil pH on microbial community structure.

31 Nutrients Nutrient amendment modifies the soil microbial community composition, as shown in a recent study assessing the effect of short and long-term fertilization on bacterial communities in moist acidic tundra (Campbell et al., 2010). Alphaproteobacteria and Gammaproteobacteria were more abundant in long-term fertilized samples compared with control soils, while a decline in the total number of taxa was observed (Campbell et al., 2010). Fertilization was also found to have a strong impact on soil respiration. In a long-term fertilization experiment in an -1 Alaskan tundra soil, nutrient amendment caused a net loss of almost 2 kg C m2 over 20 years (Mack et al., 2004). Although the plant production doubled during the experiment, losses of carbon and nitrogen from deep soil layers were higher than the increased carbon and nitrogen storage in plant biomass, indicating that the projected release of soil nutrients associated with warming at high latitudes may amplify carbon release from soils, causing a net loss of carbon (Mack et al., 2004).

Vegetation and organic matter content The presence of a vegetative cover and the resulting input in organic matter is often described as an important parameter affecting soil bacterial communities. Microbial abundance is higher in vegetated sites than in non- vegetated sites (Yergeau et al., 2007). Moreover, vegetated soils harbour distinct bacterial communities as compared to non-vegetated soils (Thomson et al., 2010; Yergeau et al., 2007) and within vegetated soils, vegetation type can also influence the bacterial community structure (Wallenstein et al., 2007). Interestingly, several bacterial phyla and classes were shown to react coherently to organic matter inputs and were, therefore, classified as copiotrophs or oligotrophs (Fierer et al., 2007). Oligotrophs grow slowly, are adapted to low carbon inputs and have a diversified metabolic capacity, while copiotrophs are fast growing and are adapted to large inputs of carbon. Thus, copiotrophs are typically found in soils with high carbon turnover rates and oligotrophs in soils with low carbon turnover rates. Fierer et al. (2007) looked at the relationships

32 between the relative abundance of the major bacterial taxa in 71 soils collected throughout the United States representing a wide range of soil characteristics. Among all the soil characteristics measured, the carbon mineralization rate was the best predictor of the relative abundance of the different bacterial taxa. A positive correlation between the abundance of the Bacteroidetes and the Betaproteobacteria and the carbon mineralization rates was detected, while there was a negative correlation between the abundance of the Acidobacteria and the carbon mineralization rates (Fierer et al., 2007). The authors suggested that, because of their higher occurrence in soils with high carbon availability and their positive response to carbon input, the Bacteroidetes and Betaproteobacteria possess attributes that can be associated with copiotrophic organisms, while the Acidobacteria possess characteristics related to oligotrophic organisms. This classification of bacterial taxa into copiotrophs and oligotrophs and their respective abundance in soils with high and low carbon turnover rates is supported by data from several other studies (Lipson, 2007; Thomson et al., 2010; Wallenstein et al., 2007). A meta-analysis of results from 16 clone libraries from bulk soils and 16 clone libraries from 16 rhizosphere soils showed that Acidobacteria are less abundant and β-, α- and γ-Proteobacteria are more abundant in rhizosphere soils, characterized by their high nutrients and carbon availability, than in bulk soils (Fierer et al., 2007). Moreover, higher rates of respiration were detected in vegetated soils than in non-vegetated soils, corresponding to higher abundances in these soils of Alphaproteobacteria and Acidobacteria, respectively (Thomson et al., 2010). Because of its importance in the soil carbon turnover and associated CO2 production, the relative abundance of copiotrophic and oligotrophic bacterial taxa has to be assessed in Arctic soils in the context of permafrost thaw.

33 Connecting text As presented in this literature review, the ongoing process of climate change could have a major impact on the fate of the soil organic carbon, and more precisely on the carbon stocked in permafrost environments. The decomposition of these large amounts of carbon could lead to important emissions of greenhouse gases (CO2, CH4) and, because of their implication in carbon cycling, microbial processes will play an important role in controlling the net gas fluxes from Arctic soils. The aim of this study was, therefore, to study the bacterial communities implicated in two processes related to the carbon cycle in high Arctic soils, CH4 oxidation and CO2 production. The activity, diversity, and community structure of these bacterial populations was assessed using techniques like microcosms assays, stable isotope probing of DNA, microarrays, and clone libraries, in different experiments conducted on high Arctic soils from Eureka (Ellesmere Island) and Axel Heiberg Island, Nunavut, Canada. The results from these experiments are presented in the following chapters.

34 Chapter 3 - Development of a SYBR safeTM technique for the sensitive detection of DNA in cesium chloride density gradients for stable isotope probing assays

Christine Martineau1,2, Lyle G. Whyte2 and Charles W. Greer1

1Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Ave., Montreal, QC, Canada, H4P 2R2. 2Department of Natural Resource Sciences, McGill University, Macdonald Campus, 21,111 Lakeshore, Ste-Anne-de-Bellevue, QC, Canada, H9X 3V9

Published in: Journal of Microbiological Methods, May 2008. 73: 199-202.

3.1 Abstract SYBR safeTM, a fluorescent nucleic acid stain that is less hazardous than ethidium bromide (EtBr), was used as a replacement for EtBr in cesium chloride (CsCl) density gradients for DNA stable isotope probing (DNA-SIP) assays. The separation of 12C- and 13C-labelled DNA using SYBR safeTM gave similar results to those obtained using EtBr, while the detection limit of DNA was enhanced by the use of SYBR safeTM. Our results also demonstrated that SYBR safeTM can be applied successfully to the ultracentrifugation of DNA from environmental samples exposed to a 13C-labelled substrate. SYBR safeTM is a safe, sensitive and effective alternative to the use of EtBr in CsCl density gradients for DNA-SIP assays.

35 3.2 Main text Stable isotope probing (SIP) of DNA is a technique that had become an important tool in microbial ecology (Radajewski et al., 2000). In this technique, compounds labelled with stable isotopes like 13C or 15N are provided as substrates for the growth of microorganisms in environmental samples. These microorganisms incorporate the stable isotopes into their cellular compounds, including phospholipid fatty acids (PLFA), RNA and DNA, which can be used as biomarkers to identify the bacterial population implicated in the metabolism of the substrate of interest. 13C-labelled DNA is the biomarker that has been the most extensively used in SIP studies. The 13C-labelled DNA can be separated from non-labelled 12C-DNA by cesium chloride (CsCl) density gradient ultracentrifugation and screened through various phylogenetic and functional analyses. Using DNA-SIP, the active microbial populations implicated in the degradation of various compounds have been identified (Friedrich, 2006). One of the critical steps of the DNA-SIP technique is the CsCl density gradient ultracentrifugation. During this step, the 13C-DNA is separated from the 12C-DNA based on their respective buoyant densities. High concentrations of ethidium bromide (EtBr) are commonly added in the CsCl gradients to allow for the visualization of the DNA under UV light following the ultracentrifugation. This technique has numerous disadvantages. Ethidium bromide, because of its tendency to intercalate between DNA bases, is known as a carcinogenic and mutagenic agent (Singer et al., 1999). Moreover, several studies have shown that UV exposure can damage DNA through photochemical reactions (Cariello et al., 1988; Emanuele et al., 2005) and subsequently compromise its use in downstream molecular biology analyses (Gründemann and Schomig, 1996; Hartman, 1991). Finally, EtBr wastes require safe disposal involving considerable environmental and monetary costs. In the present study, we tested the use of SYBR safeTM (Invitrogen, Carlsbad, CA) as a replacement for EtBr in the CsCl density gradient ultracentrifugation step of DNA-SIP assays. SYBR safeTM is a fluorescent nucleic acid stain that has a lower mutagenic potential than EtBr and is classified as non-

36 hazardous (http://probes.invitrogen.com/media/publications/494.pdf). Moreover, it was developed to be visualized optimally using a blue light transilluminator that does not damage the DNA. The results obtained in this study also demonstrated that the use of SYBR safeTM improves the detection limit of DNA in CsCl density gradients. The first step of this study was to verify that a similar separation of the 12C- and 13C-DNA could be obtained using SYBR safeTM instead of EtBr in the CsCl density gradients. 12C- and 13C-DNA was obtained through the cultivation of Methylosinus trichosporium on 12C- or 13C-methane. DNA was extracted following the protocol described by Pospiech and Neumann (1995) and was quantified using the PicoGreen® dsDNA quantitation assay (Invitrogen, Carlsbad, CA). Five μg of each 12C- and 13C-DNA from M. trichosporium was loaded in the ultracentrifugation tubes and duplicates were performed for each DNA stain. For the ultracentrifugation with EtBr, the tubes were prepared as described by Radajewski et al. (2004). For the ultracentrifugation with SYBR safeTM, we determined that the tubes had to be prepared according to the following protocol. The volume of the DNA extract was adjusted to 700 μL with TE (10:1) buffer. One gram of CsCl was dissolved in this solution and 25 μL of 10 000X SYBR safeTM were added. The resulting mixture was loaded into a 13x51 mm polyallomer Quick-Seal centrifuge tube (Beckman, Fullerton, CA, USA). The tubes were filled with a 1 g·mL-1 CsCl solution and were heat-sealed. The density of the resulting solution was 1.72 g·mL-1, as determined using a refractometer (Reichert Abbe Mark II Refractometer, Reichert, Depew, NY, USA). Tubes were ultracentrifuged at 265 000xg for 16 h in a Vti80 rotor (Beckman, Fullerton, CA, USA) using an ultracentrifuge (L8-70M ultracentrifuge, Beckman, Fullerton, CA, USA). Tubes containing SYBR safeTM were visualized using the Safe ImagerTM blue light transilluminator (Invitrogen, Carlsbad, CA) while tubes containing EtBr were visualized using a long wave UV hand light (Model UVL 56, UVP, Upland, CA, USA). Typical results for the ultracentrifugation of 5 μg of each 12C- and 13C- DNA from M. trichosporium with SYBRTM safe or EtBr are presented in Figure

37 3.1. Results were consistent between duplicates (data not shown). Ultracentrifugation with SYBR safeTM resulted in the efficient separation of the two DNA bands (Figure 3.1A). Band separation was similar to what was observed when the 12C- and 13C-DNA were ultracentrifuged with EtBr (Figure 3.1).

Figure 3.1 - Comparison of the CsCl density gradient ultracentrifugation of 5 μg of 12C- and 13C-DNA from Methylosinus trichosporium with SYBR safe™ (A) or EtBr (B).

The detection limit of DNA in CsCl gradients prepared with SYBR safeTM as a dye was determined. Five ultracentrifugation tubes were set up, ultracentrifuged and visualized as described previously for SYBR safeTM, each one containing 2000 ng of 12C-DNA and either 0, 100, 200, 500, 1000 or 2000 ng of 13C-DNA from M. trichosporium (Figure 3.2). Using this approach, we found that a band containing 100 ng of DNA was faint but detectable in the CsCl density gradient, while a 200 ng band was easily visualized. The ultracentrifugation in CsCl density gradients of 100 ng and 200 ng of 12C-DNA from M. trichosporium was repeated twice to confirm the value of the detection limit (data not shown).

38

Figure 3.2 – Cesium chloride density gradient ultracentrifugation with SYBR safe™ of 0, 0.1, 0.2, 0.5, 1.0 and 2.0 μg of 13C-labelled DNA with a constant amount of 2.0 μg of 12C-DNA, both from Methylosinus trichosporium. Arrows indicate the position of the 13C-DNA band.

We tested the potential of using SYBR safeTM for the ultracentrifugation in CsCl density gradients of DNA from environmental samples exposed to a 13C- labelled substrate. Two microcosms containing 20 g of an Arctic surface soil sample collected in Eureka (Ellesmere Island, Nunavut) were incubated at 4 ºC 13 with 20 mL of NMS medium until 5 or 40 mL of CH4 were consumed. DNA was extracted from 10 g of soil from each of the two microcosms following the protocol described by Fortin et al. (2004), but without performing the PVPP purification step. Ultracentrifugation tubes containing the DNA extracted from each microcosm were set up as described above, ultracentrifuged at 177 000xg for 40 h and visualized with the Safe ImagerTM blue light transilluminator. CsCl density gradient ultracentrifugation of DNA from an environmental 13 TM sample incubated under CH4 using SYBR safe resulted in two bands corresponding to the 12C-DNA and 13C-DNA (Figure 3.3). After the consumption 13 of 5 mL of CH4, the two DNA bands were clearly visible (Figure 3.3A), while a 13 13 longer incubation with up to 40 mL of CH4 led to a stronger C-DNA band and a fainter 12C-DNA band (Figure 3.3B). SYBR safeTM has routinely been used for the ultracentrifugation in CsCl density gradients of environmental DNA for more than a year in our laboratory with similar results.

39

Figure 3.3 – Cesium chloride density gradient ultracentrifugation with SYBR safe™ of DNA extracted from an environmental sample that had metabolized either 5 mL (A) or 40 13 mL (B) of CH4.

The results obtained in this study showed that SYBR safeTM is an effective replacement for EtBr in CsCl density gradients. The separation of the 12C- and 13C-DNA bands, which is crucial for SIP analysis, was achieved using SYBR safeTM as a dye in the CsCl density gradient, both for pure culture DNA and for DNA from an environmental previously exposed to a 13C-labelled compound. Our results also demonstrated that the use of SYBR safeTM with the Safe ImagerTM blue light transilluminator provided a sensitive detection of DNA in CsCl density gradients. While our detection limit was of 2 μg of DNA with EtBr (data not shown) and detection limits between 0.5 and 2 μg have been reported by others (Cadisch et al., 2005; Neufeld et al., 2007b), we were able to visualize a band containing amounts as low as 100 ng of DNA using SYBR safeTM. This 5-20 fold increase in sensitivity can play an important role in SIP analysis, where the labelling of DNA with 13C can be challenging. The success of a SIP analysis resides not only in the labelling of enough DNA to be able to detect it, but also on short incubation times that limit the extent of cross-feeding. The use of a more sensitive dye provides the opportunity to detect lower amounts of DNA and, therefore, to reduce incubation times.

40 An interesting aspect of the use of SYBR safeTM in CsCl density gradients is that it is optimally visualized using the Safe ImagerTM, a blue light transilluminator that does not damage the DNA. The use of a blue light transilluminator can be of critical importance when the DNA retrieved from the gradient is to be used in molecular techniques that require intact, undamaged DNA. As an example, the 13C-DNA retrieved from a CsCl density gradient ultracentrifugation tube can be used to create a metagenomic library (Dumont et al., 2006). Such an approach, that has the potential of increasing the amount of information obtained through SIP studies, includes a cloning step that can be negatively affected by previous exposure of the DNA to UV (Gründemann and Schomig, 1996; Hartman, 1991). Stable isotope probing of DNA is a powerful tool in microbial ecology and it has the potential to provide a great deal of information on microbial activity in environmental samples. However, the application of this technique can be limited by the ability to detect the labelled DNA. Here, we developed and tested a modification of the CsCl density gradient step of the SIP assay that permits a more sensitive detection of DNA. This modification, which consisted of replacing EtBr by SYBR safeTM in the CsCl solution, is simple and, when combined with the use of the Safe ImagerTM blue light transilluminator, provides several other advantages including increased safety for the user, much less potential damage to the resulting DNA, and does not generate hazardous waste material. Therefore, we suggest that SYBR safeTM is applicable to DNA-SIP and a significant improvement over the currently employed EtBr–based technique.

3.3 Acknowledgements The authors would like to thank Diane Labbé for her technical support and David F. Juck for collecting the Arctic soil sample and taking the pictures. Logistical support from the Canadian Polar Continental Shelf Project (PCSP) is gratefully acknowledged. CM was supported by a NSERC postgraduate scholarship for the duration of this project.

41 Connecting text

In Chapter 3, a protocol for the safe and sensitive detection of DNA in CsCl density gradients for DNA-SIP, a technique that allows for the identification of microorganisms implicated in the degradation of a specific substrate, was developed. Although DNA-SIP had been used previously to characterize active methane-oxidizing bacteria from different environments including landfills, peat and forest soils, active methanotrophs in Arctic soils had never been explored. In Chapter 4, the new developed protocol was used to study the activity and diversity of active methanotrophic bacteria in three soils from Eureka, in the Canadian high Arctic.

42 Chapter 4 - Stable isotope probing analysis of the diversity and activity of methanotrophic bacteria in soils from the Canadian high Arctic

Christine Martineau1,2, Lyle G. Whyte2, and Charles W. Greer1

1National Research Council Canada, Biotechnology Research Institute, Montréal, QC, Canada 2Department of Natural Resource Sciences, McGill University, Ste. Anne de Bellevue, QC, Canada

Published in: Applied and Environmental Microbiology, September 2010. 76: 5773-5784

4.1 Abstract

The melting of permafrost and its potential impact on methane (CH4) emissions is a major concern in the context of global warming. Methanotrophic bacteria have the capacity to mitigate CH4 emissions from melting permafrost. Here, we used real-time PCR (qPCR), stable isotope probing (SIP) of DNA, denaturing gradient gel electrophoresis (DGGE) fingerprinting and sequencing of the 16S rRNA and pmoA genes to study the activity and diversity of methanotrophic bacteria in active layer soils from Ellesmere Island in the Canadian high Arctic. Results showed that most of the soils had the capacity to oxidize CH4 at 4°C and at room temperature (RT), but the oxidation rates were greater at RT than at 4°C and were significantly enhanced by nutrient amendment. The DGGE banding patterns associated with active methanotrophic bacterial populations were also different depending of the temperature of incubation and the addition of nutrients. Sequencing of the 16S rRNA and pmoA genes indicated a low diversity of the active methanotrophic bacteria, with all methanotroph 16S rRNA and pmoA gene sequences being related to Type I methanotrophs from Methylobacter and Methylosarcina. The dominance of Type I over Type II methanotrophs in the native samples was confirmed by quantitative PCR of the 16S rRNA gene with primers specific for these two groups of bacteria. The 16S

43 rRNA and pmoA gene sequences related to Methylobacter tundripaludum were found in all soils, regardless of the incubation conditions, and they might, therefore, play a role in CH4 degradation in situ. This work is providing new information supporting the potential importance of Methylobacter spp. in Arctic soils found in previous studies and contributes to the limited body of knowledge on methanotrophic activity and diversity in this extreme environment.

4.2 Introduction Permafrost regions occupy approximately 23.9% of the exposed land area of the Northern Hemisphere (Zhang et al., 2008). In the past 100 years, the average temperatures in the Arctic regions have increased at almost twice the global average rate (IPCC, 2007). The melting of permafrost is one of the most important impacts of global warming on these high latitude environments, and theoretical modelling suggests that as much as 90% of the near-surface permafrost could thaw by the end of the 21st century (Lawrence and Slater, 2005). While it has been generally reported that 15% of the total soil organic carbon for the 0-100 cm depth is stocked in permafrost (Post et al., 1982), a recent estimate indicates that the northern permafrost region contains as much as 50% of the global organic carbon pool (Tarnocai et al., 2009). Carbon stocked in permafrost is now regarded as one of the most important carbon-climate feedbacks because of the size of the carbon pool and the intensity of climate forcing at high latitudes (Schuur et al., 2008; Schuur et al., 2009). The presence of these large amounts of carbon in permafrost is raising serious concerns whether melting permafrost, and the resulting increase in microbial activity, might be a source of extensive emissions of the greenhouse gases carbon dioxide (CO2) and methane (CH4) to the atmosphere. The actual emissions of CH4, from soils of high latitudes have been estimated to represent about 25% of the emissions from natural sources (Fung et al., 1991). Methane, which is 25 times more potent than CO2 as a greenhouse gas (IPCC, 2007), is produced by methanogenic archaea under anaerobic conditions. These microorganisms are known to inhabit permafrost environments (Rivkina et al., 2007; Simankova et al., 2003) and their capacity to produce CH4 at cold

44 temperatures has been reported (Ganzert et al., 2007; Metje and Frenzel, 2007; Rivkina et al., 2007; Wagner et al., 2005). Their methanogenic activity is expected to increase as permafrost soil temperature increases (Ganzert et al.,

2007). Moreover, large amounts of CH4 are stocked as CH4 hydrates in permafrost at an average depth of several hundred metres (MacDonald, 1990). Methane is also found in near surface permafrost layers and could potentially be liberated to the atmosphere as permafrost melts (Rivkina et al., 2007). Methane can be oxidized in aerobic zones by aerobic methanotrophic bacteria or in anaerobic zones by anaerobic methanotrophic archaea (see Knittel and Boetius 2009 for a review). Anaerobic methane oxidizers were not covered in the context of this study, which focussed exclusively on aerobic methanotrophs.

These bacteria utilize CH4 as a sole carbon and energy source through the activity of the enzyme methane monooxygenase (MMO). Most known aerobic methanotrophs are divided into two major groups (Type I and Type II) based on phylogeny and carbon assimilation pathways (Bowman, 2006). Type I methanotrophs, also known as γ-Proteobacteria methanotrophs (Op den Camp et al., 2009) belong to the family Methylococcaceae within the γ-Proteobacteria subdivision, while Type II methanotrophs (α-Proteobacteria methanotrophs) belong to the family Methylocystaceae in the α-Proteobacteria subdivision

(Bowman, 2006). Because of their capacity to oxidize CH4, aerobic methanotrophs can significantly reduce CH4 emissions to the atmosphere and play an important role in the global CH4 cycle (Conrad, 1996; Hanson and Hanson, 1996). Methanotrophic activity has been observed in cold environments and methanotrophs might contribute to the reduction of CH4 emissions from melting permafrost. Aerobic methanotrophic bacteria from cold environments have been reviewed in detail elsewhere (Trotsenko and Khmelenina, 2005). Most studies addressing methanotrophs from cold environments were conducted on soils from very few sites located in Northern Europe and Siberia (Dedysh et al., 1998; Liebner et al., 2009; Liebner and Wagner, 2007; Omelchenko et al., 1993; Wagner et al., 2005; Wartiainen et al., 2006; Wartiainen et al., 2003), while methanotrophic bacterial populations in soils from the

45 Canadian high Arctic remain mostly unexplored (Pacheco-Oliver et al., 2002). In addition, most of these studies were conducted at low latitudes and the pool of knowledge concerning the activity and diversity of methanotrophic bacterial populations in high Arctic soils is limited. The question being addressed in this study is whether there are active methanotrophs in the active layer soil in the high

Arctic. Therefore, the present work had two objectives: (i) to evaluate the CH4 oxidation capacity of three active layer soils from the Canadian high Arctic under various incubation conditions and (ii) to identify and characterize the diversity of the active methanotrophs in these soils using stable isotope probing (SIP) of DNA and sequencing of the 16S rRNA and pmoA genes. With this work, we identify for the first time active methanotrophs in high Arctic soils through the use of DNA- SIP.

4.3 Material and Methods

4.3.1 Site description, soil sampling and soil characterization The three soils (32D, 33D and 34D) used in this study were collected around a water retention pond located in Eureka, Ellesmere Island, Nunavut (80°0.029 N, 85°50.367 W). A fuel spill had occurred close to this site in 1990 and these soils had been collected in 2006 to monitor for hydrocarbon contamination. Two of these samples (34D in 2006 and 33D in 2007) had low, but detectable concentrations of petroleum hydrocarbons (see the Soil characteristics section for more details). Composite samples from the active layer of soil just below the surface (top 10 cm) were aseptically collected in July 2006 and 2007 and were kept frozen at -20°C until use. Soil analyses (pH, Mehlich-III, total nitrogen, nitrate, cationic exchange capacity (CEC)) were performed by Agridirect Inc. (Longueuil, Québec) according to their standard operating methods (http://www.agridirect.ca/systeme/nosmethodes.asp#sol). Organic matter content was determined by loss on ignition following combustion for 16 h at 375°C (Beaudoin, 2003) and the water content was determined at 105°C using a moisture analyzer (MA 30, Sartorius, Mississauga, ON). Total petroleum hydrocarbons

46 concentrations were measured using the technique described by the “Centre d’expertise en analyse environnementale du Québec (1997)”.

4.3.2 Soil incubation and methane degradation Soil incubations were performed under conditions similar to the conditions described in previous DNA-SIP studies (Cebron et al., 2007b; Chen et al., 2008b; Morris et al., 2002). Briefly, twenty grams of soil (wet weight) were mixed into a slurry with 20 mL of nitrate mineral salts (NMS) medium (Whittenbury et al., 1970) (NMS treatment), or added directly into 120 mL glass serum vials (No NMS treatment). NMS medium is a growth medium used for the enrichment and isolation of methanotrophic bacteria and was added to stimulate CH4 degradation by the soils, which were expected to have low CH4 oxidation rates. The vials were 13 crimp-sealed with a butyl rubber stopper and 1 mL of CH4 (Sigma-Aldrich, Oakville, ON) was injected into the headspace of each microcosm. The samples were incubated at 4°C or at room temperature (RT), corresponding respectively to optimal growth temperatures for psychrophilic and psychrotolerant bacteria (Gounot and Russel, 1999), on a rotary shaker at 160 rpm in the case of the NMS treatment. Triplicate microcosms were prepared for each treatment and autoclaved samples were used as negative controls for CH4 degradation. For each soil 13 12 treatment, a fourth series of microcosms, in which CH4 was replaced by CH4, was prepared and used as a negative control for DNA-SIP. Headspace CH4 concentration was monitored by gas chromatography as described by Roy and Greer (2000). Methane concentration was determined using a standard curve of serial dilutions of 99% pure CH4 in nitrogen. Soil CH4 oxidation rates were calculated from the initial linear decrease in CH4 concentration over time. When more than 90% of the CH4 had been consumed, the microcosms were flushed with sterile air to avoid the accumulation of CO2 and either 1 mL (NoNMS treatment) 13 12 or 5 mL (NMS treatment) of CH4 or CH4 were added to the headspace. When the degradation of more than 90% of the CH4 was completed, the samples were frozen at –20°C.

47

4.3.3 DNA extraction, fractionation and quantification DNA was extracted from 10 g of the native samples and from one of the triplicates for each treatment of incubation following the protocol described by Fortin et al. (2004). DNA crude extracts from the native samples were purified using PVPP/Sephacryl spin columns (Jugnia et al., 2009) and used for PCR- 13 DGGE and quantitative PCR assays. DNA crude extracts from the CH4 incubated samples were loaded onto cesium chloride (CsCl) density gradients together with SYBR safeTM (Invitrogen, Carlsbad, CA), as described by Martineau et al. (2008). Control tubes containing 12C and 13C-DNA from Methylosinus trichosporium were also prepared and used to determine the expected position of the DNA when it was not visible in the environmental samples (Radajewski et al., 2000). The DNA was centrifuged at 265 000xg for 16 h in a Vti80 rotor (L8-70M ultracentrifuge, Beckman, Fullerton, CA, USA). The DNA was visualized with the Safe ImagerTM blue light transilluminator and the total DNA was collected with a syringe. A second ultracentrifugation was performed at 177 000xg for 40 h to remove humics and other potential PCR inhibitors. After this second ultracentrifugation, 250 µL DNA fractions were collected and DNA was recovered from the CsCl by ethanol precipitation. DNA concentration in the extracts from the non-incubated samples and from each fraction was determined using the PicoGreen® dsDNA quantitation assay (Invitrogen, Carlsbad, CA).

4.3.4 Real-time PCR A list of all primers used in this study is presented in Table 4.1. Real-time PCR (qPCR) of the 16S rRNA genes of Type I and Type II methanotrophs was performed as described by Yergeau et al. (2010), with some modifications. The primer U785F was used in combination with either the primer MethT1bR or the primer MethT2R for the quantification of 16S rRNA genes of Type I and Type II methanotrophs, respectively. The primer MethT1bR was designed to target the

48 16S rRNA gene of Type I methanotrophs from the genera Methylomonas, Methylobacter, Methylomicrobium and Methylococcus while the primer MethT2R was designed to target Type II methanotrophs from the genera Methylosinus and Methylocystis (Wise et al., 1999). qPCR of the pmoA gene was performed using the general assay for methanotrophic bacteria described by Kolb et al. (2003) using the primers A189F and Mb661R, with some modifications. These primers can amplify the pmoA gene of most known methanotrophs, with some exceptions (Dunfield et al., 2003; Stralis-Pavese et al., 2004). For all assays, qPCR reactions were performed in 20 µl volumes using the iQ SYBR green supermix (Bio-Rad Laboratories, Hercules, CA) on a Rotor-Gene 3000 apparatus (Corbett Life Science, Sydney, Australia). Reactions were set up as per the manufacturer’s instructions, with 0.5 ng of purified DNA extract from native samples. After 3 min of initial denaturation at 95°C, PCR cycling was performed as follows: 40 cycles of 30 s at 95°C, 30 s at 55°C, and 30 s at 72°C. For the pmoA gene, an additional 15 s reading step at 82°C was added at the end of each cycle and the fluorescence signal was acquired at the end of this step to avoid the fluorescence from primer dimers. Standards were made from 10-fold dilutions of linearized plasmids containing the gene fragment of interest that was cloned from amplified pure culture DNA. The limit of detection (LOD) was determined for each qPCR assay used in this study. Briefly, qPCR of a 2-fold dilution series of linearized plasmids containing the gene fragment of interest were performed in 6 replicates, starting with a concentration of approximately 1000 gene copies/reaction to a concentration of less than 1 gene copy/reaction. The results were used to determine the number of copies that is necessary for reliable detection of the gene using the LOD tool from the GenEx Pro software (version 4.4.2, MultiD Analyses AB, Göteborg, Sweden). When a cut-off value of 34 cycles and a level of confidence of 95% were applied, LODs of 5.43 copies reaction-1 (pmoA gene), 12.56 copies reaction-1 (16S rRNA gene of Type I methanotrophs) and 17.56 copies reaction-1 (16S rRNA gene of Type II methanotrophs) were determined. The specificity of each reaction was verified by melting-curve analysis. Lambda

49 DNA was used to correct for potential PCR inhibitors present in soil extracts (Beller et al., 2002). Briefly, equal volumes of diluted soil DNA extracts and a cloned 500-bp fragment of bacteriophage lambda (105 copies per µl) were mixed. When the recovery of lambda was below 100%, quantification values for all other genes were corrected accordingly. PCR inhibition ranged from 59% to 89% and qPCR values for all other assays were corrected accordingly.

Table 4.1 - Primers used in this study.

Target gene E. coli Primer name Sequencec Reference positiona Bacterial 16S rRNA 785-803 U785F 5'-GGATTAGATACCCTGGTAG-3' (Baker et al., 2003) 341-357 U341Fb 5'-CCTACGGGAGGCAGCAG-3' (Baker et al., 2003), (Muyzer et al., 1993) 785-803 U803R 5'-CTACCAGGGTATCTAATCC-3' (Baker et al., 2003) Type I methanotroph 988-1006 MethT1bR 5'-GATTCYMTGSATGTCAAGG-3' (Wise et al., 1999) 16S rRNA Type II methanotroph 997-1017 MethT2R 5′-CATCTCTGRCSAYCATACCGG-3′ (Wise et al., 1999) 16S rRNA pmoA A189Fbcd 5′- GGNGACTGGGACTTCTGG-3′ (Holmes et al., 1995) Mb661Rcd 5′- CCGGMGCAACGTCYTTACC-3′ (Costello and Lidstrom, 1999) a Refers to 16S rRNA gene primers b When used for DGGE, a GC clamp is attached to the 5'-end (see text) c N=A, C, T or G; Y = C or T; R = A or G; M = A or C; S = C or G d When used for DGGE, degenerate bases (N, M, Y) were replaced by inosine (I) (Jugnia et al., 2009).

50 4.3.5 PCR amplification and DGGE analyses Microbial patterns associated with the DNA from the native soil samples and from the fractions collected from the CsCl density gradients were investigated using Denaturing Gradient Gel Electrophoresis (DGGE) (Muyzer et al., 1993) of PCR amplified 16S rRNA and pmoA gene fragments. PCR amplifications targeting the bacterial 16S rRNA gene were performed on a minimum of 2-3 DNA fractions from the 13C-DNA region (“heavy” DNA fractions) and 2-3 DNA fractions from the 12C-DNA region (“light” DNA fractions). For both gene amplifications, a GC clamp with the sequence 5'- CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGG-3' (Sheffield et al., 1989) was attached to the 5’-end of the forward primer. The primers U341F and U803R were used for the PCR amplification of the bacterial 16S rRNA gene, while the primers A189F and Mb661R, in which the degenerate bases were replaced by inosine (Candrian et al., 1991), were used for PCR amplification of the pmoA gene (see Table 4.1 for more details on the primers). Each 50 μL PCR mixture contained 25 ng of template DNA, 0.5 μM of each oligonucleotide primer, 200 μM of each dNTP, 1 mM MgCl2, 5 μg of BSA DNase-free (Amersham Biosciences, Baie d'Urfé, Québec, Canada), and 2.5 U of Taq polymerase in 1 × buffer (Amersham Biosciences). To increase the specificity of the amplification, both genes were amplified by a touchdown PCR method (Don et al., 1991). For the 16S rRNA gene, the annealing temperature was set to 65°C and decreased by 1°C at every cycle for 10 cycles, and then 15 additional cycles were performed at 55°C. For the pmoA gene amplification, the annealing temperature was set to 60°C and decreased by 0.5°C at every cycle for 20 cycles, and then 10 additional cycles were performed at 50°C. After 5 min of initial denaturation at 94°C, each PCR cycle consisted of 1 min at 94°C, 1 min at annealing temperature, and 3 min at 72°C. For each sample, 300 ng of amplicon were loaded onto an 8% (wt vol-1) acrylamide gel containing a 30 to 65% (16S rRNA gene) or 25 to 55% (pmoA gene) denaturing gradient, where 100% denaturant consists of 7 M urea and 40% formamide. Gels were run at 60°C for 16 h at 80 V using a Bio-Rad Dcode

51 universal mutation detection system (Bio-Rad Laboratories, Mississauga, ON, Canada). Gels were stained with SYBR gold (Invitrogen, Carlsbad, CA) and visualized with a FluorImager System model 595 (Molecular Dynamics, Sunnyvale, CA). Selected DGGE bands were excised from the gels and eluted in 60 µl of water at 4°C overnight. One microliter of DNA elution was reamplified with the appropriate corresponding primers without the GC clamp as follows: an initial denaturation of 5 min at 94°C, followed by 25 cycles of 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min, and a final elongation at 72°C for 30 min. The PCR reactions were purified using the Illustra GFX PCR DNA purification kit (GE Healthcare, Baie d’Urfé, Québec). Sequencing was performed by the “Laboratoire de synthèse et d’analyse d’acides nucléiques” at Université Laval (Ste-Foy, Québec) with a capillary ABI Prism 3100 sequencer.

4.3.6 Phylogenetic, cluster and statistical analyses Sequences were analyzed and manually corrected using BioEdit v7.0.5 (Ibis Bioscience, Carlsbad, CA) and were checked for chimeras with Bellerophon (Huber et al., 2004), using Huber-Hugenholtz correction. The gene sequences (16S rRNA gene) and the deduced amino acid sequences (pmoA gene) were submitted for comparison to the GenBank databases using the BlastN and BlastP algorithms, respectively. The sequences were aligned with their relatives using ClustalW, and phylogenetic trees were constructed with the MacVector 7.2 software package (Accelrys, San Diego, CA). The neighbor-joining algorithm was used with Jukes-Cantor correction (16S rRNA gene) or Poisson correction (pmoA deduced amino acid sequences). The robustness of inferred topologies was tested by 1,000 bootstrap resamplings of the neighbor-joining data. DGGE gel image analysis was performed using GelComparII (Applied Maths, Austin, TX). Cluster analyses of 16S rRNA and pmoA DGGE patterns associated with one selected heavy DNA fraction from each sample were performed using the Dice similarity coefficient, and UPGMA dendrograms were generated.

52 All statistical analyses for the CH4 oxidation rate data were carried out in R (v 2.7.1; The R Foundation for Statistical Computing). Normality was tested using the “shapiro.test” function. Because transformation failed to normalize data and, therefore, meet the assumptions of parametric analysis of variance (ANOVA), data were analyzed using Kruskal-Wallis non-parametric ANOVA. Analyses were performed using the “kruskal.test” function of the “pgirmess” library.

4.3.7 Nucleotide sequence accession numbers The 16S rRNA and pmoA gene sequences obtained in this study have been deposited in the GenBank database under accession numbers HM564340 to HM564353 and HM564354 to HM564378, respectively.

4.4 Results

4.4.1 Soil characteristics The mean annual temperatures at Eureka were -16.7°C in 2006 and - 17.8°C in 2007, with a mean monthly maximum temperature of 11.8°C in July 2007 and a mean monthly minimum temperature of -42.8°C in March 2007 (http://www.climate.weatheroffice.ec.gc.ca/index.html). Results of soil analyses are presented in Table 4.2. The three soils had near neutral pH values and were characterized by low levels of total nitrogen and phosphorus. The total nitrogen concentration was below the detection limit (<0.2%) for each soil and for both sampling years, while phosphorus concentrations slightly above the detection limit (9 kg ha-1) were found in samples 32D in 2006, 33D in 2007 and 34D in 2007. Organic matter content ranged between 3.8 and 5.06 %, while total petroleum hydrocarbons (TPH) were only detected in samples 34D in 2006 and 33D in 2007 at low concentrations.

53 Table 4.2 - Physicochemical parameters of soil samples 32D, 33D and 34D from Eureka collected in 2006 and 2007.

Value for physicochemical parameter of soil samplesb 32D 33D 34D Physicochemical parametera 2006 2007 2006 2007 2006 2007 pH 7.0 7.1 7.0 7.1 7.2 6.7 buffer-pH 7.4 7.5 7.4 7.5 >7.5 7.1 Mehlich-III K (kg/ha) 381 375 414 676 443 667 Mg (kg/ha) 1820 1580 1150 1580 1360 1780 Ca (kg/ha) 2560 3010 4120 1950 3540 3250 P (kg/ha) 10 <9 <9 19 <9 12

Total N (%) <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 Nitrate (ppm) 0.55 0.60 0.81 <0.50 <0.50 0.50 Saturation K (%) 3.1 3.3 3.2 6.8 3.8 4.2 Mg (%) 47.7 44.8 28.4 51.3 37.5 36.9 Ca (%) 40.3 51.2 61.3 38.0 58.7 40.3 K+Mg+Ca (%) 91.1 99.3 92.8 96.1 100 81.5

Estimated CEC (meq/100g) 14.2 13.1 15.0 11.4 13.5 18.0

Organic matter (%) 4.22 ± 0.10 4.36 ± 0.10 4.21 ± 0.13 3.80 ± 0.15 4.60 ± 0.12 5.06 ± 0.10 Water content (%) 20.06 ± 1.13 20.08 ± 0.90 16,11 ± 0.34 18.06 ± 1.09 19.73 ± 0.55 22.64 ± 0.72 TPH (mg/kg) ND ND ND 167 65.27 ND aMehlich-III is a technique to determine water-soluble and exchangeable K, Mg, Ca and P; Saturation, percent base saturation; CEC, cationic exchange capacity (milliequivalents/100g); TPH, total petroleum hydrocarbons. bFor the organic matter content and water content, values are means of triplicates (standard errors of the means are given). ND, TPH not detectable (<50 mg/kg).

4.4.2 Real-time PCR quantification The pmoA gene was detected in all the native samples with the exception of sample 32D in 2006 (Figure 4.1A) and copy numbers ranged between 1.9 x 103 and 1.14 x 104 copies ng-1 of total DNA. The 16S rRNA gene of Type I methanotrophs was detected in all samples (Figure 4.1B), with copy numbers ranging between 1.47 x 104 and 4.53 x 104 copies ng-1 of total DNA. For all samples, the abundance of the 16S rRNA gene of Type II methanotrophs was below the limit of detection of 17.56 copies/reaction (data not shown).

54

Figure 4.1 – Real-time PCR (qPCR) of pmoA gene (A) and 16S rRNA gene of type I methanotrophs (B) in native soil samples 32D, 33D, and 34D from Eureka, collected in 2006 and 2007. The abundance of the 16S rRNA gene of type II methanotrophs was below the limit of detection of 17.56 copies reaction–1 for all samples (data not shown). N.D., not detected.

4.4.3 Methane oxidation rates Methane oxidation was detected for all samples, with or without NMS medium and at both incubation temperatures except for sample 33D in 2007, which did not oxidize CH4 under any of the tested conditions (Figure 4.2). No

CH4 oxidation was detected for the autoclaved samples. The CH4 oxidation rates were significantly (P < 0.0001) affected by the addition of NMS medium, all rates being much higher when samples were incubated with NMS medium in soil slurries (Figure 4.2B) than without NMS medium (Figure 4.2A). Rates were also significantly affected by the temperature of incubation (P = 0.0043), with all rates being significantly higher at RT than at 4°C except for samples 32D and 34D in 2006 when incubated without NMS medium (Figure 4.2B). No significant effect on the CH4 oxidation rates were detected due to the sampling site (P = 0.524) or the sampling year (P = 0.904). Methane oxidation rates increased over time for the soil-NMS medium slurries, while incubation without NMS medium was generally characterized by a decrease in the CH4 oxidation rate over time (data not shown). Because of this

55 reduction over time for most of the samples incubated without NMS medium, sample 32D in 2007 was the only sample that degraded enough CH4 to be further analysed using DNA-SIP. Total incubation times for the samples supplemented with NMS medium varied between 8 and 12 days at RT and between 31 and 45 days at 4°C. The incubation of sample 32D in 2007 without NMS medium was stopped after 200 days for both the RT and 4°C treatments, although CH4 degradation was not completed in the 4°C incubation.

Figure 4.2 - Methane oxidation rates of soil samples 32D, 33D, and 34D from Eureka, collected in 2006 and 2007 incubated in microcosms without (A) or with (B) the addition of NMS medium at 4°C ( ) or at room temperature (RT) ( ). Values are means of triplicates, and the error bars represent the standard errors of the means. Asterisks indicate significantly different means for the temperature treatment within each sample at p ≤ 0.05 according to Kruskal-Wallis non-parametric ANOVA. N.D., not detected; dw, dry weight.

4.4.4 13C-labelling of the DNA and DGGE analysis of the 16S rRNA gene The heavy DNA fraction (13C-labeled) was clearly visible in the CsCl gradients below the light DNA fraction (12C-labeled) for all extracts from soils 13 incubated with CH4 in the presence of NMS medium. The only exception was sample 32D in 2007 incubated without NMS medium, so heavy DNA was

56 collected from a position in the gradient that was based on the position of 13C- DNA from a pure culture. 13 13 To identify microorganisms that utilized CH4 and incorporated C into their DNA, PCR-DGGE of the 16S rRNA gene was performed on 2-3 heavy and light DNA fractions collected from the CsCl gradients. No profile difference was observed between the heavy and light DNA fractions of samples incubated with 12 CH4 (data not shown) while a clear enrichment of some DGGE bands in the 13 heavy DNA fractions was observed when the samples were incubated with CH4 (Figure 4.3). These enriched bands correspond to the bacteria that directly or 13 indirectly assimilated the CH4. For the samples incubated with NMS medium, clearly distinct banding profiles characterized by low numbers of intense DGGE bands were found in the heavy DNA fractions (Figure 4.3A). For sample 32D in 2007 incubated without NMS medium, less clearly unique banding profiles were found in the heavy DNA fractions and several bands were common to all heavy and light DNA fractions (Figure 4.3B). This could reflect that only a small amount of DNA was 13C-labelled and that a 12C-DNA background is still visible in the heavy DNA fractions. PCR-DGGE analysis of the 16S rRNA gene from the native soil samples did not successfully detect any known methanotrophic bacteria, suggesting that they were initially below the detection limit of the technique (data not shown). For each sample the DGGE banding pattern of the 16S rRNA gene amplified from the heavy DNA was selected for cluster analyses, and resulted in two main clusters based on their incubation with or without NMS medium (Figure 4.4A). Within the samples incubated with NMS medium, the DGGE banding patterns of samples incubated at 4°C were more closely related to each other than to the DGGE banding patterns of samples incubated at RT. No clear effect of the sampling site or the sampling year on the clustering of the 16S rRNA DGGE banding patterns was observed.

57

Figure 4.3 – Typical DGGE of PCR-amplified 16S rRNA gene fragments from DNA fractions retrieved from CsCl density gradients. For each sample, 2 or 3 heavy (H) DNA fractions and 2 or 3 light (L) DNA fractions were selected for DGGE analysis. Bands that are clearly enriched in the heavy DNA fractions compared to the light DNA fractions are indicated by black arrowheads. (A) DGGE pattern obtained for samples incubated with the addition of NMS medium (sample 33D collected in 2006 incubated at room temperature (RT) is presented). (B) DGGE pattern of DNA fractions from sample 32D 2007 incubated at 4°C or at room temperature (RT) without NMS medium. The M lanes contain markers.

4.4.5 DGGE analysis of the pmoA gene PCR amplification and DGGE of the pmoA gene was performed on the DNA fractions previously selected as being characteristic of the heavy DNA for the 16S rRNA gene and the corresponding pmoA DGGE banding patterns were analysed through clustering. Incubation temperature had the strongest effect on the pmoA DGGE banding patterns, with samples incubated at 4°C clustering separately from the samples incubated at RT (Figure 4.4B). The effect of the incubation with or without NMS medium on the DGGE patterns of the pmoA gene was not as clear as it was for the 16S rRNA gene, and sample 32D in 2007 incubated without NMS at 4°C and RT did not form a separate cluster. No clear effect of the sampling site or the sampling year on the clustering of the pmoA gene DGGE patterns was detected. The pmoA gene could not be amplified from the

58 native samples when using the A189F and Mb661R primers with the GC clamp, so DGGE could not be performed.

Figure 4.4 - Cluster analysis of 16S rRNA (A) and pmoA (B) gene DGGE banding patterns of selected heavy DNA fractions collected from CsCl density gradients for Eureka soil samples 32D, 33D, and 34D collected in 2006 and 2007 incubated with NMS medium and for sample 32D collected in 2007 (32D2007) incubated without NMS medium (NoNMS). The bands that were extracted for nucleotide sequencing are numbered.

59 4.4.6 Sequence analysis of the 16S rRNA and pmoA genes Specific phylogenetic information for the 16S rRNA and pmoA genes was determined by sequencing of the individual DGGE bands. Unless otherwise indicated in Figure 4.4, all bands that migrated to the same position in a gel gave identical sequencing results. Sequences with a length between 420 and 440 base pairs (16S rRNA gene) or between 400 and 480 base pairs (130-160 deduced amino acids, pmoA gene) were obtained and were submitted for comparison to the GenBank database. Bands with 98-99% sequence identity to the 16S rRNA gene of Methylobacter tundripaludum (Bands 4, 8, 13 and 20) were found in all samples, with or without NMS medium and at both incubation temperatures, with the exception of sample 32D in 2007 incubated with NMS medium at RT (Figure 4.4A). All samples incubated at RT were characterized by the presence of bands with 96-97% identity to the 16S rRNA gene of Methylobacter luteus (Bands 10 and 14), while these bands did not appear in the samples incubated at 4°C (Figure 4.4A). When the samples were incubated without NMS medium, some of the 16S rRNA gene DGGE bands enriched in the heavy DNA fractions were found to be related to bacteria that are not currently known as methanotrophs. Bands related to Myxobacterium sp. SMH-27-4 (Band 13b, 97% sequence identity) and to an uncultured Bacteroidetes (Band 12, 98% sequence identity) were enriched in the heavy DNA DGGE banding patterns of sample 32D incubated at RT without NMS medium (Figure 4.4A). Moreover, a band related to Methylophilus sp. ECd5 (Band 18, 97% sequence identity), was found at both incubation temperatures (Figure 4.4A). Some sequences related to methylotrophs from the genera Methylophilus and Methylotenera (Bands 11 and 15, 93% and 97% identity, respectively) were also detected in samples incubated with NMS medium, especially in the samples incubated at RT (Figure 4.4A). A phylogenetic tree of the 16S rRNA gene sequences from bands enriched in the heavy DNA fractions of all samples tested and their closest relatives is presented in Figure 4.5. All 16S rRNA gene sequences that were related to known methanotrophs shared more than 96% identities with cultivable Type I

60 methanotrophs of the Gamma-Proteobacteria. These methanotroph sequences mainly formed two groups among the genus Methylobacter, one group being related to Methylobacter luteus (AF304195) and the other to Methylobacter tundripaludum (AJ414655). Sequences related to Methylosarcina quisquiliarum (NR025040) were found exclusively in sample 32D in 2007 incubated without NMS medium at both incubation temperatures. Several of the 16S rRNA gene sequences found in the Eureka samples were closely related to 16S rRNA gene clones from cold environments, including an uncultured Methylotenera related clone from a Himalayan glacier (EU809719), an uncultured Beta-proteobacterium related clone from Siberia ground water (AJ583175), an uncultured Methylobacter related clone from the Lena Delta (EU124849) and an uncultured bacteria related clone from a borehole in Finland (FJ823170). The pmoA DGGE patterns of samples incubated at 4°C were characterized by the presence of several bands migrating at different positions in the gel but sharing as much as 100% homology (ex.: bands 1, 2 and 3, Figures 4.4B). With the exception of bands 16, 20, 26, 27 and 31, which shared 93-95% identities with Methylosarcina lacus (AY007286), all the pmoA amino acid sequences retrieved from the heavy DNA fractions of Eureka samples shared more than 94% identities to members of the genus Methylobacter (Figure 4.6). As it was the case for the 16S rRNA gene, several sequences from the 4°C incubations were related to Methylobacter tundripaludum (AJ414658).

61

Figure 4.5 - Phylogenetic relationships of bacterial 16S rRNA gene sequences obtained from DGGE bands enriched in the heavy DNA fractions collected from CsCl density gradients for soil samples 32D, 33D, and 34D from Eureka, Ellesmere Island, Nunavut, Canada, collected in 2006 13 and 2007 incubated in the presence of CH4 with or without NMS medium and at 4°C or at room temperature. The tree was inferred by neighbor-joining analysis of 434 homologous positions of sequence from each band. Escherichia coli (GenBank accession no. AJ567606) was used as the outgroup. Numbers on the nodes are the bootstrap values (percentages) based on 1,000 replicates (values above 70 are presented). The scale bar indicates the estimated number of base changes per nucleotide sequence position. The numbered DGGE bands presented in Figure 4.4 are shown in normal type. Known genotypes are shown in boldface type. GenBank accession numbers are shown after the species in smaller type. Lake Constance is located in Germany, Switzerland, and Austria at the northern foot of the Alps.

62

Figure 4.6 - Phylogenetic relationships of pmoA deduced amino acid sequences obtained from DGGE bands enriched in the heavy DNA fractions collected from CsCl density gradients for soil samples 32D, 33D, and 13 34D from Eureka, collected in 2006 and 2007, incubated in the presence of CH4 with or without NMS medium at 4°C or at room temperature. The tree was inferred by neighbor-joining analysis of 114 homologous positions of deduced amino acid sequence from each band. The ammonia monooxygenase subunit A amino acid sequence from Nitrosococcus oceani (GenBank accession no. AAB57809) was used as the outgroup. Numbers on the nodes are the bootstrap values (percentages) based on 1,000 replicates (values above 70 are presented). The scale bar indicates the estimated number of base changes per nucleotide sequence position. The numbered DGGE bands presented in Figure 4.4 are shown in normal type. Known genotypes are shown in boldface type. GenBank accession numbers are shown after the species in smaller type. Lake Washington is located in King County, Washington state. Transbaikal soda lake is located in the Transbaikal region of Russia.

63 4.5 Discussion

All the Arctic soils examined in this study had the capacity to oxidize CH4 at RT and at 4°C, with and without NMS medium, except sample 33D in 2007.

All our soil samples had higher CH4 oxidation rates at RT than at 4°C, which indicates that the active methanotrophic bacterial populations in these Arctic soils are psychrotolerant rather than psychrophilic: quite active at low temperatures (4– 7°C) but with higher optimal growth temperatures (15–30°C). Liebner and

Wagner (2007) also found maximum CH4 oxidation potentials at 21°C for upper active layer soils from the Lena Delta, Siberia, while deeper active layer soils were more active at 4°C. Several studies have also reported that microbial communities in Arctic soils, including permafrost soils, are dominated by psychrotolerant rather than by psychrophilic bacteria (Aislabie et al., 2006; Gilichinsky, 2002; Mohn and Stewart, 2000; Steven et al., 2007). All the Eureka samples were characterized by their low levels in nutrients essential to bacterial growth, like nitrogen and phosphorus, which is typical of high latitude Arctic soils (Tarnocai and Campbell, 2006). According to our results for CH4 oxidation rates, this lack of essential nutrients may be an important factor limiting the activity of methanotrophic bacteria in Arctic soils. Indeed, when the samples were incubated with the addition NMS medium, a growth medium that was designed for the enrichment and isolation of methanotrophic bacteria

(Whittenbury et al., 1970), CH4 oxidation rates were 15 to 100 times greater than for the corresponding samples incubated without NMS medium. In trials performed on sample 32D in 2007, we found that the addition of only the nitrogen component of the NMS medium resulted in CH4 oxidation rates that were similar to the NMS treatment, while the addition of tap water or the phosphorus component only had intermediary effects (data not shown). Therefore, nitrogen appears as the main limiting nutrient to CH4 oxidation in our samples. Nitrogen is known to be a strong regulator of CH4 oxidation in soils, as reviewed by Bodelier and Laanbroek (2004). While several studies have shown that the application of ammonium-based fertilizers can strongly reduce CH4 consumption by different soil types, a strong stimulation of soil CH4 consumption following application of

64 ammonium and nitrate-based fertilizers is also frequently observed (Bodelier and Laanbroek, 2004).

Surprisingly, we did not detect CH4 degradation for sample 33D in 2007 under any of the incubation conditions tested, even though this sample was oxidizing CH4 in 2006. This corresponded with the detection of low amounts of TPH in the sample in 2007 (Table 4.2) and to increased respiratory activity in the microcosms (data not shown). Our hypothesis is that the input of an exogenous carbon source at this site led to the development of an active heterotrophic bacterial population that outcompeted methanotrophic bacteria and limited their activity in the microcosms. However, this hypothesis still needs to be confirmed. While the addition of NMS medium and the incubation temperature were found to greatly increase CH4 oxidation rates in the Eureka soils, a strong impact of these two parameters on the bacterial diversity of the heavy DNA fractions from DNA-SIP assays was also observed. Cluster analysis showed that 16S rRNA DGGE banding patterns of heavy DNA fractions from samples incubated without NMS medium were clearly distinct from those of samples incubated with NMS medium. DGGE banding patterns of heavy DNA fractions from samples incubated with NMS at 4°C were more closely related to each other than to the DGGE banding patterns of samples incubated at RT for both the 16S rRNA and pmoA genes, while no clear effect of the sampling site or the sampling year was detected. A previous DNA-SIP study performed on an agricultural field soil had shown that the incubation with NMS medium can modify the community structure and reduce the diversity of the methanotrophic bacterial populations of the heavy DNA fraction (Cebron et al., 2007b). In addition, it was shown by Mohanty et al. (2007) that the relative abundance of different methanotrophic bacterial populations in a rice field soil and a forest soil is affected by the temperature of incubation. The absence of a clear effect of the sampling site on the active methanotrophic bacterial populations is not surprising as the three sites were located close to each other and the soil samples had similar physico- chemical characteristics (Table 4.2). However, it is also possible that this absence of effect of the sampling site or year is due to the incubation conditions to which

65 the samples were exposed. These conditions might have selected for specific bacterial populations in all the samples.

Concomitant with the CH4 oxidation results, sequencing of 16S rRNA and pmoA genes from the heavy DNA fractions of Eureka soil samples incubated 13 under CH4 indicated that the active methanotrophic bacterial populations were related to psychrotolerant rather than to psychrophilic bacteria. Several sequences of the 16S rRNA and pmoA genes found were related to Methylobacter tundripaludum, a psychrotolerant methanotrophic bacterium with an optimal growth temperature of 23 °C, isolated from an Arctic wetland soil from Svalbard (Wartiainen et al., 2006). When samples were incubated at 4°C with NMS medium, almost 100% of the sequences found in the heavy DNA fractions were related to M. tundripaludum. With one exception, bacteria related to this bacterium were active in all our samples, at both incubation temperatures, with and without NMS medium. Sequences related to M. tundripaludum were, therefore, detected even under the conditions that were the most closely related to the natural environmental conditions (without NMS medium, at 4°C). These results indicate that psychrotolerant methanotrophs related to M. tundripaludum may play a more important role in CH4 degradation in situ than other bacterial taxa that were detected under more selective conditions (with NMS medium, at RT), like M. luteus. In addition, 16S rRNA gene sequences related to Methylosarcina quisquiliarum were found exclusively in sample 32D in 2007 incubated without NMS medium at both incubation temperatures and might, therefore, also be ecologically relevant. Despite some minor differences, the phylogenetic trees generated with the pmoA amino acid and 16S rRNA gene sequences retrieved from SIP showed a high degree of similarity when compared to known methanotrophic bacteria. Some differences between the 16S rRNA and pmoA genes were detected in the closest relatives at the species level, but these differences appeared to be related to a lack of information for the pmoA gene in the databases. For example, no pmoA sequence is available in GenBank for M. luteus, which was found to be closely related to some of our sequences for the 16S rRNA gene. Overall, all sequences

66 detected in this study shared high identities (>96% for the 16S rRNA gene and >93% for the pmoA amino-acid sequences) to cultured bacteria. Although methanotrophs are present in the native Eureka samples, as indicated by their capacity to degrade CH4 and by the qPCR detection of genes specific to methanotrophic bacteria (pmoA gene, 16S rRNA gene of Type I and Type II methanotrophs), these bacteria do not appear to be dominant members of the bacterial population. PCR-DGGE of the 16S rRNA gene of the native soil samples did not detect sequences related to known methanotrophic bacteria, while the pmoA gene could not be amplified using the A189F and Mb661R primers with the GC clamp, so DGGE could not be performed. Our results also showed that the overall diversity of active methanotrophic bacteria as detected by DNA-SIP in the Eureka soil samples was low. All methanotroph 16S rRNA and pmoA gene sequences found in the heavy DNA fractions were related to Type I methanotrophs from the genera Methylosarcina and Methylobacter, while no sequence related to Type II methanotrophs were detected. Interestingly, these results are consistent with those of a recent study by Liebner et al. (2009), who used DGGE and cloning of 16S rRNA and pmoA gene fragments to study the methanotrophic bacterial diversity at two depths in an active layer soil from the Lena Delta, Siberia. For both genes and using either DGGE or cloning, they also found sequences exclusively related to Type I methanotrophs from the genera Methylobacter and Methylosarcina, with a strong dominance of sequences related to Methylobacter (Liebner et al., 2009). Several characteristics, including high latitudes, pH and organic matter content, shared by the Eureka soils analysed in this study and the soils from the Lena Delta studied by Liebner et al. (2009) could explain the similarities in their methanotrophic bacterial populations. These characteristics might be imposing selective pressure toward a restricted group of methanotrophs that have the capacity to colonize these extreme environments. However, it is also possible that the diversity of the active methanotrophic bacteria in the Eureka samples was negatively affected by the enrichment process 13 that would have occurred during incubation with CH4.

67 The dominance of Type I over Type II methanotrophs in the native samples was confirmed through the quantitative detection of the 16S rRNA gene using primers specific to these two groups of bacteria. Earlier studies on soils from the Lena Delta describing the activity and abundance of methanotrophic bacteria showed that Type I methanotrophs were the most abundant methanotrophs (Liebner and Wagner, 2007) and that the abundance of a phospholipid fatty acids related to Type II methanotrophs was low, if not undetectable, throughout the active layer of soil (Wagner et al., 2005). Immunofluorescent microscopy using 14 antibodies against the majority of known methanotrophic species in Russian tundra bog soils also indicated the dominance of members from the genus Methylobacter (Vecherskaya et al., 1993). Yergeau et al. (2010) also found a dominance of Type I over Type II methanotrophs in other active layer and permafrost soils from Eureka using qPCR of the 16S rRNA gene. Using both DNA-SIP and qPCR, we now provide new evidence indicating the importance of Type I methanotrophs in high Arctic soils. Stable isotope probing of DNA is a technique that has some limitations, as discussed by many authors (Dumont and Murrell, 2005; Neufeld et al., 2007a). Low substrate incorporation and short incubation times can result in poor labelling of the 13C-DNA, which becomes indistinguishable from the 12C-DNA background. High concentrations of the substrate and long incubation times can lead to significant enrichment and cross-feeding, a mechanism by which non- targeted microorganisms incorporate 13C into their DNA through the metabolism of by-products derived from the targeted organisms. In this study, and in most

DNA-SIP studies, CH4 concentrations that are higher than those expected to be found in situ and long incubation times, depending of the treatment, were used in order to obtain good 13C-labelling of the DNA. Because of the potential enrichment of low-affinity methanotrophs under these conditions, our results have to be cautiously interpreted in relation to the methanotrophic bacterial populations active under in situ conditions. Several of the 16S rRNA gene sequences enriched in the heavy DNA fractions from our samples were related to bacteria that were not previously

68 reported as methanotrophs. Most of these sequences were related to bacteria from the genera Methylophilus and Methylotenera, both members of the ß- Proteobacteria. This observation has often been associated with SIP experiments and sequences related to bacteria known for their capacity to degrade methanol, the first compound produced by methanotrophs during CH4 oxidation, which are widely detected in SIP studies targeting methanotrophic bacteria (Cebron et al., 2007a; Cebron et al., 2007b; Lin et al., 2004). While it is generally hypothesized that the presence of these sequences in the heavy DNA fractions is due to cross- feeding, it is extremely challenging to determine if the bacteria were 13C-enriched through direct or indirect incorporation of the 13C. Recent studies have shown that the diversity of methanotrophic bacteria is broader than what was generally reported (see Conrad 2009 for a review). Therefore, the hypothesis that the enrichment in the heavy DNA fractions of bacteria that were not previously 13 known as methanotrophs through direct utilization of CH4, and not through cross-feeding, has to be considered.

4.6 Conclusions In this study, we showed that active methanotrophic bacterial populations are present in soils from the Canadian high Arctic, but that their CH4 oxidation capacity is limited by nutrient availability and low ambient temperature. These limitations raise questions on the potential of methanotrophic bacteria to oxidize

CH4 in situ and their ability to limit emissions of CH4 from melting permafrost.

Future studies employing in situ soil CH4 flux measurements and CH4 isotope signature analyses (Whiticar, 1999), should provide a better insight into the role played by methanotrophic bacteria in the carbon cycle in Arctic environments. Typically, SIP experiments require incubation conditions and substrate concentrations that differ considerably from the natural environment and the data may have to be interpreted with caution (Neufeld et al., 2007a). However, most of the results presented in this study using DNA-SIP are consistent with previous studies demonstrating the important role of Type I methanotrophs of the genus Methylobacter in the Arctic environment. The consistent detection of sequences

69 related to Methylobacter tundripaludum in all our samples and under various incubation conditions suggests that this bacterium might play an important role in

Arctic permafrost soils and might contribute to the reduction of CH4 emissions in melting permafrost environments as a result of global warming.

4.7 Acknowledgements The authors would like to thank David F. Juck for collecting the soil samples. Logistical support from the Canadian Polar Continental Shelf Project (PCSP) is gratefully acknowledged. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Program, Northern Supplements Program, Special Research Opportunities IPY Program. Christine Martineau was supported by NSERC and FQRNT postgraduate scholarships for the duration of this project. Additional funding for Martineau was provided by the Department of Indian and Northern Affairs - Northern Scientific Training Program (NSTP).

70 Connecting text Results from Chapter 4 indicated that high Arctic soils had the capacity to degrade CH4 and that this activity was related to the presence in these soils of Type I methanotrophs, mainly bacteria related to Methylobacter tundripaldudum. Some questions were raised by this first study on methanotrophic bacteria in soils from the Canadian high Arctic. Can high Arctic soils oxidize CH4 at concentrations closer to the atmospheric concentration, and can methanotrophic bacteria associated with high-affinity CH4 oxidation be detected in these soils? Moreover, since methanotrophic bacterial populations are known to differ depending on the environmental conditions, would it be possible to detect a wider diversity of methanotrophic bacteria in high Arctic soils by looking at soils with more distinctive characteristics? In Chapter 5, these questions were assessed in a study looking at methanotrophic bacterial populations implicated in CH4 degradation at low and high CH4 concentrations in three different soils with highly distinctive physico-chemical characteristics from Axel Heiberg Island, in the Canadian high Arctic.

71 Chapter 5 - Detection of putative atmospheric methane- oxidizing bacteria in soils from the Canadian high Arctic

Christine Martineau1,2, Levente Bodrossy3, Etienne Yergeau1,2, Lyle G. Whyte2 and Charles W. Greer1

1National Research Council Canada, Biotechnology Research Institute, Montréal, QC, Canada 2Department of Natural Resource Sciences, McGill University, Ste. Anne de Bellevue, QC, Canada 3Environmental Genomics Team, CSIRO Marine & Atmospheric Research, Hobart, Tasmania, Australia

In preparation for publication

5.1 Abstract

The melting of permafrost and the potential for methane (CH4) emissions are a major concern in the context of global warming. Arctic soils are known for their capacity to take up atmospheric CH4, and this capacity could increase in response to climate change. Methane oxidation in Arctic soils might, therefore, not only reduce CH4 emissions from the melting permafrost but could also consume CH4 from the atmosphere. In this study, we used microarray and clone library analyses of the particulate methane monooxygenase gene (pmoA) to characterize the methanotrophic bacterial communities at two sampling depths in three different soil types representative of different Arctic settings. The microbiological data were linked to soil physico-chemical characteristics and CH4 degradation rates at low and high CH4 concentrations. The pmoA gene sequences related to two main groups of putative atmospheric methane oxidizers, the “upland soil cluster gamma” (USCγ) and the “upland soil cluster alpha” (USCα), were detected for the first time in Arctic soils and were associated with near neutral and acidic pH, respectively. Methane oxidation rates in soils at both low and high CH4 concentrations were higher in soils where the USCγ genotype was detected. These results indicated that methanotrophic bacteria in high Arctic soils

72 might have the capacity to mitigate CH4 emissions from the melting permafrost, but also to act as a sink for atmospheric CH4 due to the activity of these high- affinity methane-oxidizing bacteria.

5.2 Introduction Increasing temperatures in the Arctic are expected to result in the melting of up to 90% of the near-surface permafrost by 2100 (Lawrence and Slater, 2005). The presence of large amounts of carbon in permafrost is raising serious concerns whether melting permafrost, and the resulting increase in microbial activity, might be a source of methane (CH4), a greenhouse gas 23 times more potent than carbon dioxide (CO2) (IPCC, 2007), to the atmosphere (Davidson and Janssens, 2006). However, through the activity of methanotrophs, soils can also represent a significant sink for CH4 (Curry, 2009). For instance, atmospheric CH4 uptake has been reported for various well-aerated soils (Knief et al., 2003) including Arctic soils (Roslev and Iversen, 1999; Whalen and Reeburgh, 1990) and could be responsible for the degradation of 24.8-28 Tg of CH4 annually (Curry, 2009). The highest increase (>40%) in atmospheric CH4 soil consumption in simulated warmer climates are expected to occur in boreal forest, tundra and polar desert environments (Curry, 2009). Thus, in the context of global warming, CH4 oxidation by Arctic soils might not only reduce CH4 emissions from the melting permafrost, but also contribute to atmospheric CH4 uptake. Methane oxidation in soils is due to the activity of methanotrophic bacteria, which utilize CH4 as a sole carbon and energy source. Most known aerobic methanotrophs are divided into two major groups (Type I and Type II) based on phylogeny and carbon assimilation pathways (Bowman, 2006). Type I methanotrophs belong to the family Methylococcaceae in the Gammaproteobacteria subdivision, while Type II methanotrophs belong to the family Methylocystaceae in the Alphaproteobacteria subdivision (Bowman, 2006). In addition to these two groups, new methanotrophic bacteria belonging to the phylum Verrucomicrobia were recently described (Dunfield et al., 2007; Op den Camp et al., 2009). The enzyme, methane monooxygenase (MMO), is a key

73 enzyme in the CH4 oxidation process and exists in two forms: the particulate methane monooxygenase (pMMO) and the soluble methane monooxygenase (sMMO). Because of its occurrence in most known methanotrophs, with the exception of Methylocella spp., the gene encoding the α-subunit of the pMMO (pmoA) is widely used as a biological marker to study methanotrophic bacteria.

While the process of CH4 uptake by soils is well described (Dunfield, 2007) and has been known for several years, the bacteria implicated in the oxidation of CH4 at atmospheric concentrations (~1.7 ppm) are not well characterized. These atmospheric methane-oxidizing bacteria would be specialized oligotrophs adapted to low CH4 concentrations and possessing a MMO with a higher substrate affinity compared to most cultivated methanotrophic bacteria (Bender and Conrad, 1992). Recent studies have shown that bacteria from the genus Methylocystis produce two isozymes of the pMMO with different oxidation kinetics associated with low and high CH4 affinity, respectively (Baani and Liesack, 2008). These bacteria have the capacity to sustain oxidation at low CH4 concentrations, but they might require other energy sources for growth (Baani and Liesack, 2008; Knief and Dunfield, 2005). Particulate methane monooxygenase gene sequences from uncultivated bacteria are frequently detected in soils having the capacity to degrade atmospheric CH4 (Angel and Conrad, 2009; Holmes et al., 1999; Knief et al., 2003; Kolb, 2009). These uncultured putative atmospheric methane-oxidizing bacteria can mainly be separated into two groups: one being related to the Alphaproteobacteria and referred to as the “upland soil cluster alpha” (USCα) and the other being related to the Gammaproteobacteria and referred to as the “upland soil cluster gamma” (USCγ) (Kolb, 2009). A third group of pmoA gene sequences, closely related to the ammonia monooxygenase gene (amoA) of ammonia-oxidizing bacteria, has been detected in some forest soils with atmospheric methane-oxidizing capacity (Knief et al., 2006; Kolb et al., 2005). Interestingly, these amoA-related sequences belong to a methanotrophic bacterial group referred to as “Cluster 1” (Kolb, 2009), which also includes bacterial isolates from Arctic soils with highly divergent pmoA genes (Pacheco-Oliver et al., 2002). However, the presence of

74 these bacterial isolates in Arctic soils has not been associated with high-affinity

CH4 oxidizing activity (Pacheco-Oliver et al., 2002). Several studies have provided a better understanding of atmospheric methane-oxidizing bacteria in forest soils, as reviewed by Kolb (2009), and a limited number of studies have focussed on these bacterial communities in non- forested upland soils, but they remain largely unexplored in Arctic soils. In this paper, we looked at the diversity of methanotrophic bacteria in three different soil environments encountered in the Canadian high Arctic, including well aerated upland tundra soils. Microarray and clone library analyses of the pmoA gene were used to characterize the methanotrophic bacterial communities at two sampling depths. The microbiological data were linked to soil physico-chemical characteristics and CH4 degradation rates at low and high CH4 concentrations.

Different CH4 oxidation rates and highly distinctive methanotrophic bacterial populations were detected depending on the soil type. The pmoA gene sequences related to two main groups of putative atmospheric methane oxidizers, USCγ and USCα, were detected for the first time in Arctic soils. Methane oxidation rates at both low and high CH4 concentrations were linked to the presence in the soils of the genotype USCγ. These results indicated that methanotrophic bacteria in high

Arctic soils might have the capacity to mitigate CH4 emissions from the melting permafrost, but also to act as a sink for atmospheric CH4 due to the activity of high-affinity methane-oxidizing bacteria.

5.3 Material and methods

5.3.1 Site description, soil sampling and soil characterization Soil samples were collected within a 2 km radius from the McGill Arctic Research Station (MARS), located 8 km inland at Expedition Fjord, Axel Heiberg Island, Nunavut, in the Canadian high Arctic (79.433N, 90.766W). Three different soil types presenting distinct characteristics were selected for this study and were grouped into acidic soils (AC; 2 sites in 2008, 1 site in 2009), upland tundra soils (UT; 2 sites in 2008, 1 site in 2009) and wet meadow soils (WM; 2 sites in 2008, 2 sites in 2009). Soil samples were aseptically collected in April

75 2008 and July 2009 from the surface (top 15 cm) and at the bottom of the active layer of soil (approximate depth of 40 cm), for a total of 20 soil samples, and were kept frozen at -20ºC until used. Soil descriptions and characteristics are presented on page 105. Methane concentrations in the soil samples were determined by extracting the gas from the soil pore water solution using the protocol described by Wagner et al. (2003), with some modifications. Briefly, 3 g of soil were added to a 20 mL vial containing 6 mL of saturated sodium chloride solution. The vial was crimp-sealed, vortexed for 30 s and incubated for one hour at 80°C to allow for the transfer of the gases from the soil solution to the headspace of the vial.

CH4 degradation rates at low (15 ppm) and high (1000 ppm) CH4 concentrations were determined by measuring the headspace CH4 concentration over time in 20 mL crimp-sealed vials containing 5 g of soil samples incubated at 10°C. For incubations at a CH4 concentration of 1000 ppm, the CH4 concentration in the headspace was determined by gas chromatography as described by Roy and Greer

(2000). For incubations at a CH4 concentration of 15 ppm, the CH4 concentration in the headspace was determined using a Varian 3800 gas chromatograph coupled to a FID detector. The gas sample (250 uL) was injected onto a 7 m x 2 mm I.D. Chromosorb 102 packed column (Supelco, Bellefonte, PA). The column was o heated at 40 C for 3 minutes while the injector and detector were maintained at

125oC and 150oC, respectively. Argon was used as carrier gas. The CH4 calibration curve was made from dilutions of 99% pure CH4 in nitrogen (from 1.5 ppm to 35 ppm).

5.3.2 DNA extraction, purification and quantification DNA was co-extracted with RNA from 3 g of soil samples using the MoBio RNA PowerSoil Total RNA Isolation Kit and eluted from the purification column using the RNA PowerSoil DNA Elution Accessory Kit (MoBio Laboratories, Carlsbad, CA) following the manufacturer’s instructions. DNA extracts were purified using the OneStep PCR Inhibitor Removal Kit (Zymo Research, Orange, CA). DNA concentration in the purified extracts was

76 determined using the PicoGreen® dsDNA quantitation assay (Invitrogen, Carlsbad, CA).

5.3.3 pmoA microarray A microarray targeting the pmoA/amoA genes of methanotrophs that covers the whole known diversity of these bacteria (~200 probes) was used as a methanotrophic bacterial community fingerprinting method (Bodrossy et al., 2003; Stralis-Pavese et al., 2004). pmoA and/or amoA genes were amplified from all DNA extracts using two different approaches, as described by Stralis-Pavese et al. (2004). The first approach was used to amplify the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries (referred to as “pmoA/amoA” in the text). The second approach was a two-step PCR specific to the pmoA and related genes from methanotrophs but not for amoA and related genes (referred to as “pmoA” in the text). Thus, with this second approach, no signal on the microarray was expected for probes related to ammonia oxidizers, the Methylocapsa acidiphila group, the Arctic methanotrophs with highly divergent pmoA genes (Pacheco-Oliver et al., 2002) and from RA21 (Holmes et al., 1999) and related sequences. A minimum of three 50 µL PCR reactions were pooled for each sample. PCR amplification, in vitro transcription and hybridization protocols were as described by Bodrossy et al. (2003) and modified by Stralis-Pavese et al. (2004). PCR amplification with both PCR approaches was not successful for all the samples. All the acidic soil DNA extracts were successfully amplified using the two-step PCR targeting the pmoA and closely related genes, but not using the primers targeting the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries. The reverse situation was observed for the wet meadow samples, while DNA extracts from the upland tundra soil could generally be amplified using both PCR approaches. Since the taxa targeted by the two PCR approaches were quite different, the microarray and clone library analyses were performed separately for each approach. A total

77 of 10 samples were amplified using the two-step pmoA PCR and 14 samples were amplified using the pmoA/amoA PCR.

5.3.4 Clone libraries pmoA/amoA clone libraries were constructed for one representative site from each of the three soil types (two sampling depths per site) using the same PCR products that were used for the microarray analyses. The PCR products were purified with the PureLink™ PCR Purification Kit (Invitrogen Canada, Burlington, ON, Canada) and ligated using the QIAGEN PCR cloning kit (QIAGEN, Mississauga, ON, Canada) at an insert: vector ratio of 3:1 following the manufacturer’s instructions. The ligations were transformed in SURE Competent Cells (Stratagene, La Jolla, CA) following the manufacturer’s instructions.

Transformants were selected on Luria-Bertani medium supplemented with ampicillin, 5-bromo-4-chloro-3-indolylbeta-D-galactopyranoside (X-Gal), and isopropyl-beta-D-thiogalactopyranoside (IPTG). Ninety-five randomly selected white colonies from each library were screened for the correct insert size. Of these, 48 clones were randomly selected from each library and sent for sequencing. With this number of clones, the rarefaction curves of each library tended to approach the saturation plateau and no further sequencing was performed.

5.3.5 Statistical analyses

Statistical analyses for the soil parameters, CH4 degradation rates and microarray analyses were performed in R (version 2.9.0, The R Foundation for Statistical Computing). For analysis of variance (ANOVA) of the soil parameters and CH4 degradation rates, normality was tested using the “shapiro.test” function. When necessary, data were log transformed to meet parametric ANOVA assumptions. ANOVA and post hoc Tukey HSD tests were then carried out using the “aov” and “TukeyHSD” functions, respectively. When transformations failed to normalize data, Kruskal-Wallis and associated multiple comparison tests were performed using the “kruskal.test” and “kruskalmc” functions of the “pgirmess”

78 library, respectively. The effect of sampling depth was tested using the paired Wilcoxon signed rank test (“wilcox.test” function). For the microarray analysis, probe signal intensities were normalized to positive controls on the same array and expressed in relative intensities of the individual hybridization potential of each probe as assessed during validation of the array with pure cultures and clones relative intensities (Stralis-Pavese et al., 2004). When the relative intensity of a signal was below 0.025, it was considered as not detected and the value was set to 0. Probes that were not detected in any of the samples and probes with broad specificity were removed from the analyses. Bray-Curtis distance matrices were calculated using the “vegdist” function of the “vegan” library and principal coordinate analyses were carried out using the “cmdscale” function based on the square root of the distance. The sums of the relative intensities of probes corresponding to the main groups of methanotrophic bacteria were added to the ordination as supplementary variables that were not involved in the calculation of the ordination. Pearson linear correlation of the sums of relative intensities of probes corresponding to the main groups of methanotrophic bacteria with CH4 oxidation rates at CH4 concentrations of 15 ppm and 1000 ppm were calculated using the “cor.test” function. Deduced amino acid sequences from all clone libraries (269 sequences) were aligned using ClustalW and were used for subsequent analyses in the program MOTHUR (Schloss et al., 2009). Uncorrected pairwise distance matrixes were generated using the “dist.seqs“ function and were used to classify the sequences into operational taxonomic units (OTU) (“cluster” function) with a cut off value of 7%, which was suggested as an appropriate cut off value to determine methanotrophic species-level OTUs using pmoA amino acid sequences (Degelmann et al., 2010). Shannon diversity index calculations were generated for each library using the “summary.single” function. Libraries were compared using the LIBSHUFF program (Singleton et al., 2001). One representative amino acid sequence from each OTU was selected using the “get.oturep” function and was submitted for comparison to the GenBank database using the BlastP algorithm. Sequences from each OTU, sequences from their closest relatives and from the

79 known main groups of methanotrophs were aligned and phylogenetic trees were constructed with the MacVector 7.2 software package (Accelrys, San Diego, CA). The neighbour-joining algorithm was used with Poisson correction.

5.4 Results

5.4.1 Methane concentration and CH4 degradation

Sampling date had a significant effect on the CH4 concentration in the soils (p=0.0221), with higher values being found in the samples from the winter of 2008 than from the summer of 2009. Sampling depth also had a significant effect on this parameter (p=0.0020), with CH4 concentrations being generally higher in the 40 cm deep soils than in the top soils (Table 5.1).

Methane degradation rates at low (15 ppm) and high (1000 ppm) CH4 concentrations were significantly (p=0.0432) and nearly significantly (p=0.0538) affected by the sampling date, with higher values being detected for the summer 2009 soils than for the winter 2008 soils. The sampling depth had a significant effect on the CH4 degradation rate at 15 ppm (p=0.0143), with higher values in the top soils than in the 40 cm deep soils, but not on the CH4 degradation rate at 1000 ppm (p=0.4316) (Table 5.1).

Because of the strong effect of the sampling date on CH4 concentrations and CH4 degradation rates, it was impossible to combine data from both sampling years in statistical analyses and to test the effect of the soil type due to the insufficient number of replicates within each sampling year. However, trends in the effect of the soil type on these parameters can still be observed. Among the samples from 2008, CH4 degradation at a concentration of 15 ppm was detected exclusively for the acidic and upland tundra top soils (Table 5.1). Methane degradation at 15 ppm was detected for all soils collected in 2009, with generally higher values being found for the upland tundra soil, intermediate values for acidic soils, and lower values for the wet meadow soils. Methane degradation at 1000 ppm was detected for all samples. Within each sampling year, the samples showing the highest CH4 degradation rates at 15 ppm and 1000 ppm were upland tundra top soils.

80

Table 5.1 - Methane concentration and CH4 degradation rates at CH4 concentrations of 15 and 1000 ppm for three different Arctic soil types collected in 2008 (winter conditions) and 2009 (summer conditions).

a Sampling Soil type CH4 concentration CH4 degradation (15 ppm) CH4 degradation (1000 ppm) year (nmoL CH4/g soil) (nmoL/g soil/day) (nmoL/g soil/day) top 40 top 40 top 40 2008 ACb 0.47 1.04 0.020 n.d. 0.79 0.78 UTb 0.78 1.03 0.031 n.d. 1.42 0.63 WMb 0.46 1.10 n.d. n.d. 0.61 0.59 2009 AC 0.46 0.55 0.218 0.016 1.80 0.52 UT 0.28 0.50 3.904 0.193 7.12 1.23 WMb 0.39 0.60 0.014 0.006 0.75 1.96 aAC: Acidic soil; UT : Upland tundra soil; WM : Wet meadow soil bValues are averages from two different samples n.d.: not detected; top: value for the top soils. 40: value for the 40 cm deep soils.

5.4.2 Microarray analyses of the pmoA/amoA genes A microarray targeting the pmoA/amoA genes of methanotrophs that covers the whole known diversity of these bacteria (~200 probes) was used as a methanotrophic bacterial community fingerprinting method. Two different PCR approaches were tested to amplify the pmoA and/or amoA genes in DNA extracts from the 20 soil samples (see section 5.3.3) but none of the two approaches successfully amplified all the samples. Thus, a total of 10 samples were amplified using the two-step pmoA PCR and 14 samples were amplified using the pmoA/amoA PCR. Because of the important differences in the specificity of the primers used in these two PCR approaches, results obtained with each PCR approach were analyzed separately. Microarray hybridization patterns for the 10 samples from acidic and upland tundra soils amplified using the pmoA PCR approach and the 14 samples from upland tundra and wet meadow soils amplified using the pmoA/amoA PCR approach are presented in Figures A.1 and A.2 of the Appendices, respectively.

81 Microarray with the pmoA PCR The acidic soils were characterized by high relative signal intensities for several probes related to Type Ia methanotrophs (Methylobacter-Methylomonas- Methylomicrobium-Methylosarcina group), while very few of these taxa were detected in the upland tundra soils. In contrast, high relative signal intensities were detected for two probes targeting uncultured bacteria from the “upland soil cluster gamma” (USCγ) and a probe targeting the Methylohalobius- Methylothermus group in the upland tundra soils, while these organisms were not dominant in the acidic soils. High relative signal intensities for probes targeting uncultured bacteria from the “upland soil cluster alpha” (USCα) were mostly associated with the acidic top soils, although some probes were also detected in upland tundra soils.

Microarrays with the pmoA/amoA PCR For both the upland tundra soils and the wet meadow soils, high relative signal intensities were detected for several of the probes targeting methanotrophs with a pmoA gene closely related to the amoA gene of ammonia-oxidizing bacteria, including members of the Svalbard clade and the Arctic methanotrophs with highly divergent pmoA genes. Several probes targeting putative atmospheric methane oxidizers from the USCα, USCγ and “tropical upland soil cluster 2” (TUSC2) groups were detected in the upland tundra soils, while no signals for these probes were detected in the wet meadow soils. No probe could be specifically associated with the wet meadow soils, where bacteria related to ammonia-oxidizing bacteria were dominant.

Methanotrophic bacterial community structure

Principal coordinate analysis (PCoA) was used to produce ordination graphs in which the samples were positioned according to their similarity as assessed by the pmoA/amoA gene microarray. The PCoA results for the samples amplified with the pmoA and the pmoA/amoA PCR approaches are presented in Figures 5.1A and 5.1B, respectively. In the ordination of the samples amplified

82 with the pmoA PCR approach (Figure 5.1A), the acidic soils were clearly separated from the upland tundra soils on the first axis of the ordination. Acidic soils formed two different groups according to their sampling depth, top soils being separated from the 40 cm deep soils on the second axis of the ordination (Figure 5.1A). No grouping of the samples according to their sampling date was observed. In the ordination of the samples amplified with the pmoA/amoA PCR approach (Figure 5.1B), samples were also generally separated according to the soil type but the grouping was not as clear as for the pmoA primers and implicated both axes. No grouping of the samples according to their sampling date or depth was observed.

83 1.0 A AC1-top

USCα

UT2-top AC3-top

AC2-top Type Ib UT1-top

Axis 1 = 21.56% USCγ Type II UT3-40 TUSC-2 AC3-40 Type Ia UT3-top AC2-40

AC1-40

-1.0 -1.0 Axis 1 = 40.40% 1.0

WM1-top

1.0 B Svalbard clade

WM1-40 WM4-top WM4-40 Axis 2 = 15.90% Type 1a

UT2-top AMO-related WM3-40 WM2-top Type II Arctic soils UT1-40 UT1-top Baltic sea clade WM3-top UT2-40 USCα TUSC-2 Nitrosospira-Nitrosovibrio Type Ib USCγ WM2-40 UT3-top -0.4 UT3-40 -0.8 1.0 Axis 1 = 28.95% Figure 5.1 - Methanotrophic bacterial community composition of the acidic soils (AC), upland tundra soils (UT) and wet meadow soils (WM) from Axel Heiberg Island, in the Canadian high Arctic. (A) Principal coordinate analysis based Bray-Curtis similarity calculated from pmoA/amoA microarray hybridization patterns of DNA extracted from 6 AC soil samples and 4 UT soil samples amplified using a two-step PCR approach targeting the pmoA and closely related genes. (B) Principal coordinate analysis based on Bray-Curtis similarity calculated from pmoA/amoA microarray hybridization patterns of DNA extracted from 6 UT soil samples and 8 WM soil samples amplified using a PCR approach targeting the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries. “Top” and “40” indications refer to the sampling depth (within the top 15 cm or at a depth of 40 cm, respectively). Numbers 1 and 2 refer to samples from 2008 (winter); numbers 3 and 4 refer to samples from 2009 (summer). The arrows represent the relative abundance of major groups of methanotrophic bacteria and were added as supplementary variables after the analysis. USCα, upland soil cluster alpha; USCγ, upland soil cluster gamma; TUSC2, tropical upland soil cluster 2; AOB, ammonia-oxidizing bacteria.

84 5.4.3 Clone libraries A total of 7 clone libraries were constructed, 3 with PCR products from samples amplified using the pmoA PCR approach (pmoA clone libraries) and 4 with PCR products from samples amplified using the pmoA/amoA PCR approach (pmoA/amoA clone libraries). Forty-eight clones with an insert of the correct size were randomly selected and sequenced from each library. In the pmoA clone library of the 40 cm deep acidic soil, 15 sequences contained more than one open reading frame, while in the pmoA/amoA clone library of the wet meadow top soil clone library, 21 sequences did not produce any significant match to any known GenBank sequence. These sequences, as well as a few bad quality sequences found in each library, were removed from subsequent analyses. The final number of sequences analyzed, the number of OTUs detected and the Shannon diversity index for each library are presented in Table 5.2. Using a species-level cut off value of 7% (Degelmann et al., 2010), a total of 15 OTUs were found among the 269 putative amino acid sequences from the 7 libraries. The highest numbers of OTUs as well as the highest Shannon diversity indexes were detected in the pmoA and the pmoA/amoA clone libraries from the upland tundra top soil (Table 5.2). Paired comparisons using LIBSHUFF showed that all libraries were significantly different from each other, with P-values below 0.0001. pmoA clone libraries The three pmoA clone libraries presented completely distinct methanotrophic bacterial populations (Figure 5.2). The clone library from the acidic top soil contained only two OTUs, both being related to the USCα genotype. The clone library of the 40 cm deep acidic soil contained three OTUs, two of which were related to Type Ia methanotrophs. The third OTU, which was also found in the upland tundra top soil clone library, was related to uncultured putative atmospheric methane oxidizers from the USCγ genotype. The upland tundra top soil clone library had the highest number of OTUs, 5 of them being related to the USCγ genotype and one being related to the USCα genotype.

85 Table 5.2 - Number of sequences, number of OTUs and Shannon diversity index of 3 pmoA and 4 pmoA/amoA clone libraries from three different high Arctic soils. Number of Number of Shannon diversity Clone librarya sequences OTUsb indexb pmoA clone libraries AC-top 43 2 0.1105 AC-40 cm 30 3 0.4677 UT-top 44 6 1.1728 pmoA/amoA clone libraries UT-top 44 5 1.0559 UT-40 cm 45 4 0.8726 WM-top 24 1 0.0000 WM-40 39 3 0.7269 aAC: Acidic soil; UT: Upland tundra soil; WM: Wet meadow soil; top: value for the top soils. 40: value for the 40 cm deep soils. bDetermined using a cutoff value of 7%.

pmoA/amoA clone libraries Several of the OTUs found in the four pmoA/amoA clone libraries from the upland tundra and wet meadow soils at both sampling depths were related to Arctic methanotrophs with highly divergent pmoA genes (Figure 5.2). Four different OTUs related to these bacteria were detected among the four clone libraries, most of them being detected in 2-3 different samples. Besides these amoA-related pmoA genes, the upland tundra top soil clone library also had 2 OTUs related to the USCγ genotype and one related to the USCα genotype. One OTU related to Nitrosospira sp. NpAV was exclusively detected in the wet meadow top soil clone library.

Phylogenetic diversity of pmoA/amoA genes from high Arctic soils A phylogenetic tree presenting pmoA/amoA deduced amino acid sequences from each OTU found in the 7 clone libraries and from the major known groups of uncultured and cultured methanotrophic bacteria is presented in Figure 5.3. The 15 OTUs were mainly divided into five distinct groups among methane- and ammonia-oxidizing bacteria (Figure 5.3). Four OTUs clustered with the Arctic methanotrophs with highly divergent pmoA genes and three OTUs clustered with pmoA sequences from the USCα genotype, both groups belonging to the Alphaproteobacteria. Two OTUs clustered with Type Ia methanotrophs and five

86 OTUs clustered with sequences from the USCγ genotype, both groups belonging to the Gammaproteobacteria. One OTU clustered with an ammonia-oxidizing bacterium from the genus Nitrosospira (Betaproteobacteria). Three of the OTUs that were related to USCγ putative atmospheric methane oxidizers were distantly related to their uncultured closest relatives, forming new clusters within this group.

100% OTU1 Type Ia OTU2 methanotrophs OTU3 80% OTU4 Uncultured putative atmospheric methane OTU5 oxidizers Gamma- OTU6 60% OTU7 OTU8 Uncultured putative atmospheric methane OTU9 oxidizers Alpha- 40% OTU10 OTU11 Arctic methanotrophs OTU12 with higly divergent 20% OTU13 pmoA gene OTU14 OTU15 Nitrosospira sp. NpAV 0% AC-top AC-40 UT-top UT-top UT-40 WM-top WM-40

pmoA clone libraries pmoA/amoA clone libraries

Figure 5.2 - Methanotrophic bacterial community composition based on pmoA/amoA clone libraries from an acidic soil (AC), an upland tundra soil (UT), and a wet meadow soil (WM) from Axel Heiberg Island, in the Canadian high Arctic. OTUs were defined according to a methanotrophic species-level cutoff value of 7% (Degelmann et al., 2010). Gene libraries were retrieved from pmoA PCR products amplified using a two-step PCR approach targeting the pmoA and closely related genes (pmoA clone libraries) or a PCR approach targeting the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries (pmoA/amoA clone libraries). Analysis was performed with 142 homologous positions from 269 pmoA/amoA deduced amino acid sequences. “Top” and “40” indications refer to the sampling depth (within the top 15 cm or at a depth of 40 cm, respectively).

87 Crenothrix polyspora ABC59822 Uncultured bacterium AAZ06152 (River plain aquifer) 98 UT1-top-AH5 (OTU12) 91 UT-40-B4 UT-40-G8 88 Methanotroph K3-16 AAN62857 72 WM-top-36 93 WM-40-2E10 (OTU14) Arctic methanotrophs with 90 Uncultured bacterium AAN62859 (Arctic soil) highly divergent pmoA genes 100 UT-top-AD9 (Alphaproteobacteria) Uncultured bacterium CAI30576 (Hydromorphic soil) UT-40-E12 (OTU13) WM-40-B11 Methanotroph K2-14 AAN62856 UT-40-D2 (OTU11) 97 Uncultured bacterium CAD56343 (Water distribution system) Uncultured bacterium ACS72314 (Desert soil) 100 Nitrosospira sp. NpAV AAB53437 Ammonia‐oxidizing bacteria 89 80 WM-40-H4 (OTU15) (Betaproteobacteria) Methylacidiphilum infernorum V4 (pmoA 2) ABX56604 100 Methylacidiphilum infernorum V4 (pmoA 1) ABX56601 Meth ylocapsa aurea CBA11953 Uncultured bacterium RA14 AAD47927 99 98 UT-top-AC8 (OTU8) Uncu ltured bacterium ACN22486 (Pine forest soil) USCα 99 UT-top-PA3 100 100 (Alphaproteobacteria) Uncultured bacterium CBJ05749 (Glacier forefield) 92 75 AC-top-A4 (OTU9) 89 Uncultured bacterium ABV25575 (Acidic peatland) 94 AC-top-C4 (OTU10) Methylosinus trichosporium AAA87220 99 Methylocystis parvus AAQ10310 99 Uncultured bacterium ABQ10698 (Landfill) AC-40-B5 (OTU2) Methylosarcina lacus AAG13081 100 93 [Methylomicrobium album AAA87217 Type Ia methanotrophs 98 AC-40-D11 (OTU1) (Gammaproteobacteria) 100 Uncultured bacterium BAI22710 (Rice field) Methylosoma difficile ABD13901 Methylomonas methanica AAA87218 Methylobacter sp. PP39D ABD62307 Methylothermus thermalis AAX37294 95 Methylohalobius crimensis CAE46474 Methylococcus capsulatus AAQ10311 Methylocaldum sp. T-025 BAF49660 UT-top-PG8 (OTU3) 99 Uncultured bacterium CAI30586 (Hydromorphic soil) 76 AC-40-C10 UT-top-PC10 84 UT-top-AD7 (OTU7) 85 Uncultured bacterium CAE22498 (Upland soil) USCγ Uncultured bacterium ACS72344 (Desert soil) 71 (Gammaproteobacteria) Uncultured bacterium ACS72360 (Desert soil) UT-top-PB8 (OTU6) 99 79 99 UT-top-PB11 UT-top-PF12 (OTU5) 100 UT-top-AE4

0.1 UT-top-PG6 (OTU4) 100 UT-top-PH9

88 Figure 5.3 - Phylogenetic relationships of deduced amino acid sequences based on pmoA/amoA clone libraries from an acidic soil (AC), an upland tundra soil (UT) and a wet meadow soil (WM) from Axel Heiberg Island, in the Canadian high Arctic. The tree was inferred by neighbor-joining analysis of 142 homologous positions of deduced amino acid sequences from each OTU, their closest relatives and from the known main groups of methanotrophs. OTUs were defined according to a methanotrophic species-level cutoff value of 7%. The PmoA amino acid sequence from Crenothrix polyspora ABC59822 was used as the outgroup. Numbers on the nodes are the bootstrap values (percentages) based on 1,000 replicates (values above 70 are presented). The scale bar indicates the estimated number of base changes per nucleotide sequence position. The OTU number corresponding to representative sequences from each OTU is indicated in brackets. The main groups of methanotrophic bacteria identified in this study are indicated in red type. “Top” and “40” indications refer to the sampling depth (within the top 15 cm or at a depth of 40 cm, respectively)

5.4.4 Correlation between methanotrophic bacterial groups and soil CH4 oxidation rates

Based on the microarray analyses of samples amplified with the pmoA and the pmoA/amoA primers, the sum of relative intensities of probes related to the USCγ and to the tropical upland soil cluster 2 (TUSC2) genotypes were significantly positively correlated to the soil CH4 oxidation rates at CH4 concentrations of 15 ppm and 1000 ppm (Table 5.3). No other significant correlations between the sum of the relative intensities of probes corresponding to specific groups of methanotrophic bacteria and CH4 oxidation rates were detected.

Table 5.3 - Pearson correlation coefficients between the relative abundance of different groups of methanotrophs based on the microarray analyses of samples amplified with the pmoA or pmoA/amoA primers and the CH4 oxidation rates at 15 or 1000 ppm. Methanotrophic pmoA primers pmoA/amoA primers a bacterial group CH4 oxidation rate at CH4 oxidation rate at 15 ppm 1000 ppm 15 ppm 1000 ppm Type Ia n.s. n.s. n.s. n.s. 0.787 USCγ 0.666 (0.0355) 0.741 (0.0141) (<0.001) 0.829 (<0.001) Type Ib n.s. n.s. n.s. n.s. Type II n.s. n.s. n.s. n.s. USCα n.s. n.s. n.s. n.s. TUSC2 0.756 (0.011) 0.858 (0.002) 0.697 (0.006) 0.762 (0.002) AOB-related n.d. n.d. n.s. n.s. aUSCα, upland soil cluster alpha; USCγ, upland soil cluster gamma; TUSC2, tropical upland soil cluster 2; AOB, ammonia-oxidizing bacteria. n.s.: correlation is not significant (p-value>0.05); n.d.: bacterial group not detected using this PCR approach. The p-value is given in brackets.

89 5.5 Discussion

5.5.1 Detection of putative methane-oxidizing bacteria in high Arctic soils In the context of global warming, the presence of large amounts of carbon in permafrost is raising serious concerns about whether melting permafrost, and the expected increase in microbial activity, might be a source of extensive emissions of CH4 to the atmosphere. Methanotrophic bacteria in Arctic soils have the potential to reduce these CH4 emissions and act as a sink for atmospheric CH4 through high affinity enzymatic activity of the pMMO. In this study, we have detected for the first time pmoA genes sequences related to the two main groups of putative atmospheric methane oxidizers in high Arctic soils using both microarray analyses and clone libraries of the pmoA and/or amoA genes. Putative atmospheric methane-oxidizing bacteria related to the upland soil cluster gamma (USCγ) were found to be important members of the methanotrophic bacterial communities in upland tundra top soils, while putative atmospheric methane- oxidizing bacteria related to the upland soil cluster alpha (USCα) were mainly associated with the acidic top soils. Even though there are no bacterial isolates for these two groups of putative atmospheric methane-oxidizing bacteria, their pmoA genes have been previously detected in upland soils having the capacity to oxidize atmospheric CH4 (Kolb, 2009). The USCα genotype is widely detected in forest soils, where it often dominates the methanotrophic bacterial populations (Kolb, 2009). Sequences related to this group were also found in other soil types, including grassland (Horz et al., 2005) and acidic peat (Chen et al., 2008a). The USCγ genotype, which has been less frequently detected in soils, was first described in a study looking at the methanotrophic bacterial diversity in 35 different upland soils (Knief et al., 2003). Its presence was associated with soils having a pH above 6, while USCα-related sequences were detected in more acidic soils (Knief et al., 2003). More recently, pmoA transcripts related to USCγ were found to be dominant in a desert soil with a pH of 8.6 showing in situ atmospheric

CH4 uptake (Angel and Conrad, 2009). Interestingly, the results from this study also indicated that the soil pH is an important factor influencing the distribution of the two main groups of putative

90 atmospheric methane oxidizers in high Arctis soils, with the USCγ genotype being associated with upland tundra soils (pH 6.4) and the USCα genotype being associated with acidic soils (pH 4.8). Different methanotrophic bacteria have different pH optima (Bowman, 2006), which could explain this distribution. Moreover, atmospheric methane-oxidizing bacteria from the USCα genotype are phylogenetically related to Methylocapsa acidiphila (Dedysh et al., 2002), a Type II methanotroph isolated from acidic peat, and they might, therefore, represent a clade of acidophilic bacteria. Previous work using microarray analysis of the 16S rRNA gene showed that the bacterial community structure in the three soil types analysed in this study was also strongly affected by the soil pH, with acidic soils harbouring highly distinctive bacterial populations from the soil types with a more neutral pH (see Chapter 6). The effect of the soil pH on the bacterial community structure has been widely described in the literature (Chu et al., 2010; Fierer and Jackson, 2006; Rousk et al., 2010), and it appears that methanotrophic bacterial communities implicated in the oxidation of atmospheric CH4 are also strongly affected by this parameter.

5.5.2 Methane oxidation at low and high methane concentrations The CH4 degradation capacity of the three soil types was tested in microcosms at low (15 ppm) and high (1000 ppm) CH4 concentrations. All soils had the capacity to oxidize CH4 at 1000 ppm and most soils had the capacity to oxidize CH4 at 15 ppm, with the highest CH4 degradation rates being detected in upland tundra top soils for both CH4 concentrations. Although a starting CH4 concentration of 15 ppm is significantly higher than the atmospheric concentration (1.7 ppm), it is in the range of concentrations that can be oxidized by high affinity methanotrophic bacterial communities but not by low affinity methanotrophic bacterial communities (Conrad, 1996). Moreover, for some upland tundra and acidic top soils, the CH4 concentration measured in the microcosm headspace at the end of the incubation period was below the atmospheric concentration. These results indicated that active methanotrophic bacterial populations with a high affinity for CH4 inhabit some of these high

Arctic soils and might, therefore, contribute to atmospheric CH4 uptake.

91 Interestingly, these results also suggest that, because of their capacity to efficiently degrade CH4 at low and high CH4 concentrations, some Arctic soils could play the double role of degrading the potentially high CH4 concentrations resulting from the melting of the permafrost and taking up CH4 at low concentrations from the atmosphere. However, further studies, including in situ

CH4 flux measurements, will be necessary to determine if CH4 oxidation at low and high CH4 concentrations is occurring under natural conditions.

Soil CH4 degradation rates at both low and high CH4 concentrations were significantly affected by the sampling date, with higher values detected for the 2008 samples (summer conditions) than for the 2009 samples (winter conditions). These results may reflect the seasonal effect on the activity of the methanotrophic bacterial population and the associated reduction of CH4 oxidation rates in winter samples reported in other studies (Abell et al., 2009; Henckel et al., 2000). Soil

CH4 degradation at a low CH4 concentration was also significantly higher in top soils than at a depth of 40 cm, while no significant differences were detected between sampling depths for degradation at a high CH4 concentration. Similarly, previous studies have shown that CH4 degradation at low concentrations was restricted to a thin subsurface layer of soil (Henckel et al., 2000; Kolb, 2009; Roslev et al., 1997). The results from this study indicated that the capacity of soils to oxidize

CH4 at low and high concentrations can be associated with the presence of specific groups of methanotrophic bacteria. High relative intensities of probes targeting putative atmospheric methane oxidizers from the USCγ genotype, as well as pmoA gene sequences related to these organisms, were detected in upland tundra top soils, in which high CH4 degradation rates were observed. The potential implication of the USCγ genotype for CH4 degradation in Arctic soils was further indicated by the detection of significant positive correlations between the soil CH4 oxidation rates at both low and high CH4 concentrations and the sum of the relative intensities of probes related to USCγ. Soil CH4 oxidation rates at low and high CH4 concentrations were also positively correlated to the sum of the relative intensities of probes targeting the tropical upland soil cluster 2 (TUSC2),

92 a genotype of putative atmospheric methane-oxidizing bacteria with pmoA genes related to the amoA gene of ammonia-oxdizing bacteria from a tropical upland soil (Knief et al., 2005). However, sequences related to this group of bacteria were not detected in any of the clone libraries from this study and, therefore, the significance of these results remain unclear.

5.5.3 Diversity of methanotrophic bacteria in high Arctic soils In this study, 15 different OTUs were detected at the species level in the 7 pmoA/amoA clone libraries, while positive signals for probes related to all the main groups of methanotrophic bacteria were detected by microarray analyses. Putative atmospheric methane-oxidizing bacteria related to the USCγ genotype formed the methanotrophic bacterial group with the highest diversity, with 5 different OTUs mainly detected in upland tundra top soil clone libraries. Three of these five OTUs had low similarity to previously published sequences and formed distinct clusters in the phylogenetic tree. Methane-oxidizing bacteria related to the USCα genotype formed three different OTUs. Interestingly, OTU 10, which includes the majority of the sequences related to the USCα genotype detected in this study, shared 100% amino acid sequence identity with sequences detected in a glacier forefield in Greenland (CBJ05749; unpublished data), but low identity (<93%) to sequences from other environments. Methanotrophic bacteria with pmoA gene sequences related to the amoA gene of ammonia-oxidizing bacteria were found to be important members of the methanotrophic bacterial communities in all of the high Arctic soils analysed using the PCR approach targeting the pmoA and amoA genes. Several probes related to different groups of amoA-related pmoA genes were detected in all of the upland tundra and wet meadow soils, and gene sequences related to these groups were also detected in the pmoA/amoA clone libraries from the corresponding soil types. They were divided into 4 different OTUs, all of them being more or less closely related (88 to 94% amino acid sequence identities) to Arctic methanotrophic bacterial isolates with highly divergent pmoA genes (Pacheco- Oliver et al., 2002). Although they were not fully characterized, analysis of the

93 16S rRNA genes from these bacterial isolates indicated that they are closely related to Type II methanotrophs (genera Methylosinus/Methylocystis), despite their highly divergent pmoA gene (Pacheco-Oliver et al., 2002). The pmoA genes from these methanotrophic bacterial isolates belong to a cluster referred to as “Cluster 1” (Kolb, 2009), which also includes several pmoA gene sequences from uncultured bacteria detected in forest soils with atmospheric methane-oxidizing capacities (Knief et al., 2006; Kolb et al., 2005). In this study, it was impossible to link the presence of these bacteria to high-affinity CH4 oxidation since they were abundant both in wet meadow soils and in upland tundra soil, which had low and high CH4 oxidation rates at a concentration of 15 ppm, respectively.

In addition to the different groups of putative CH4 oxidizers previously described, Type Ia methanotrophic bacteria (Methylobacter-Methylomonas- Methylomicrobium-Methylosarcina group) were also detected in high Arctic soils and were important members of the methanotrophic bacterial populations in acidic soils. Several probes targeting Type Ia methanotrophs were detected in most of the acidic soils and sequences related to bacteria from the genera Methylobacter and Methylosarcina represented more than 95% of the sequences found in the 40 cm deep acidic soil. Interestingly, a relatively small number of probes related to Type II methanotrophs were detected in our samples and generally had low signal intensities, while no sequence related to this group of methanotrophic bacteria was found in the clone libraries. Dominance of Type I over Type II methanotrophs in Arctic soils was also reported in other studies (Liebner et al., 2009; Martineau et al., 2010; Vecherskaya et al., 1993; Wagner et al., 2005; Yergeau et al., 2010). However, unlike what was found in some of these studies (Liebner et al., 2009; Martineau et al., 2010), our results indicated that methanotrophic bacterial diversity in Arctic soils is not limited to Type I methanotrophic bacteria and that, by looking at various soil types and at different sampling depths, a wider variety of methanotrophic bacterial groups can be found in these soils. To successfully amplify pmoA and/or amoA genes from all the soil samples analyzed in this study, two PCR approaches had to be used. The pmoA

94 genes from the acidic soils were successfully amplified using the two-step PCR targeting the pmoA gene and closely related genes, but not with the primers targeting the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries. The reverse situation was observed for the wet meadow soils, while the upland tundra soils could generally be amplified using both PCR approaches. The two different reverse primers used in these PCR approaches have different specificities (Stralis-Pavese et al., 2004), which explains why samples with highly divergent methanotrophic bacterial communities can be amplified using one of the reverse primers but not the other. Results from this study also indicated that the methanotrophic bacterial community structure detected for a single sample amplified using one or the other approach can be extremely different, as shown by the pmoA and pmoA/amoA microarray and clone library results from the upland tundra soils. The differences in community structure observed when using these two different reverse primers have also been described in other studies (Bourne et al., 2001; Knief et al., 2006). Although it has limited our capacity to fully compare the methanotrophic bacterial communities among the three soil types, the use of two different approaches to amplify the pmoA and/or amoA genes allowed for the detection of a broader diversity of methanotrophs than previously reported and to show for the first time the presence of atmospheric CH4 oxidizers in different Arctic soils.

5.6 Conclusions The presence of pmoA gene sequences related to the two main groups of methanotrophic bacteria associated with atmospheric CH4 oxidation in soils, which are formed of uncultured bacteria from the Gammaproteobacteria and Alphaproteobacteria classes (USCγ and USCα genotypes), were detected for the first time in Arctic soils, associated with soils with near neutral and acidic pH, respectively. Other groups of methanotrophic bacteria, including methanotrophs with pmoA gene sequences related to the amoA gene of ammonia-oxidizing bacteria and Type Ia methanotrophs, were also detected in these soils using microarray analyses and clone libraries of the pmoA/amoA genes. Upland tundra

95 top soils had the highest CH4 degradation rates at both low and high CH4 concentrations, and these rates were positively correlated to the presence of the methanotrophic bacterial genotype USCγ. The analysis of different Arctic soil types with highly distinctive characteristics suggested that methanotrophic bacterial diversity may be broader than previously described in the literature. These results also indicated that high Arctic soils may have the capacity to mitigate CH4 emissions from the melting permafrost and act as a sink for atmospheric CH4 through the activity of high-affinity methane-oxidizing bacteria.

However, further studies, including in situ CH4 flux measurements, will be necessary to determine if this activity is occurring under natural conditions and can play a significant role in mitigating the impact of CH4 in the context of global warming.

5.7 Acknowledgements The authors thank Yao Pan for the microarray analyses, the Canadian Polar Continental Shelf Project (PCSP) for their logistical support and McGill University’s High Arctic Research Station. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Program, Northern Supplements Program, and Special Research Opportunities IPY Program. CM was supported by NSERC and FQRNT postgraduate scholarships for the duration of this project. Additional funding for CM was provided by the Department of Indian and Northern Affairs - Northern Scientific Training Program (NSTP).

96 Connecting text Results from Chapter 5 indicated that methanotrophic bacterial communities in high Arctic soils differed depending on the soil type and associated soil characteristics, and that these differences can be linked to differences in the soil CH4 oxidation capacity. Knowing that bacterial communities implicated in CO2 production, another important process of the carbon cycle, can also be affected by the soil characteristics, a study in which microarrays and qPCR targeting 16S rRNA genes was used to characterize the bacterial communities in these soils was conducted. The microbiological data were linked to soil physico-chemical characteristics and CO2 production rates.

97 Chapter 6 - Differences in major bacterial taxa in three high Arctic soils with different carbon dioxide production rates

Christine Martineau1,2, Etienne Yergeau1,2, Lyle G. Whyte2 and Charles W. Greer1*

1National Research Council Canada, Biotechnology Research Institute, Montréal, QC, Canada 2Department of Natural Resource Sciences, McGill University, Ste. Anne de Bellevue, QC, Canada

Submitted to: ISME journal

6.1 Abstract

The melting of permafrost and its potential impact on greenhouse gas emissions is a major concern in the context of global warming. The fate of the carbon trapped in permafrost will largely depend on soil physico-chemical characteristics, among which are the quality and quantity of organic matter, nutrient availability and water content, and on microbial community composition. In this study, we used microarrays and real-time PCR (qPCR) targeting 16S rRNA genes to characterize the bacterial communities in three different soil types representative of various Arctic settings. The microbiological data were linked to soil physico-chemical characteristics and carbon dioxide (CO2) production rates. Microarray results indicated that soil characteristics, and especially the soil pH, were important parameters in structuring the bacterial communities at the genera/species levels. Shifts in community structure were also visible at the phyla/classes levels, with the soil CO2 production rate being positively correlated to the relative abundance of the Alphaproteobacteria, Bacteroidetes, and

Betaproteobacteria. These results indicated that CO2 production in Arctic might be related to the presence of specific groups of bacteria that have the capacity to actively degrade soil carbon, and not only on the environmental conditions.

98 6.2 Introduction Permafrost regions occupy approximately 23.9% of the exposed land area of the Northern Hemisphere (Zhang et al., 2008). In the past 100 years, the average temperatures in the Arctic regions have increased at almost twice the global average rate (McBean et al., 2005) and the melting of permafrost is one of the most important impacts of global warming on these high latitude environments. Theoretical modelling suggests that as much as 90% of the near- surface permafrost could thaw by the end of the 21st century (Lawrence and Slater, 2005). While it has been generally reported that 15% of the total soil organic carbon for the 0-100 cm depth is stocked in permafrost (Post et al., 1982), a recent estimate indicates that the northern permafrost region contains as much as 50% of the global organic carbon pool (Tarnocai et al., 2009). The presence of these large amounts of carbon in permafrost is raising serious concerns whether melting permafrost and the resulting increase in microbial activity, might be a source of extensive emissions of the greenhouse gases carbon dioxide (CO2) and methane (CH4) to the atmosphere. However, the fate of the carbon trapped in permafrost will largely depend on soil physico-chemical characteristics, among which are the quality and quantity of organic matter, nutrient availability and water content, and on the microbial community composition. It is, therefore, crucial to understand the links between microorganisms, soil characteristics and gaseous emissions in Arctic soils. Several soil parameters are known to play an important role in structuring soil bacterial communities. One of these parameters is the presence of a vegetative cover and the resulting input in organic matter. Vegetated soils harbour distinct bacterial communities as compared to non-vegetated soils (Thomson et al., 2010; Yergeau et al., 2007) and within vegetated soils, vegetation type can also influence the bacterial community structure (Wallenstein et al., 2007). Several bacterial phyla and classes were shown to react coherently to organic matter inputs and were, therefore, classified as copiotrophs or oligotrophs (Fierer et al., 2007). Oligotrophs grow slowly, are adapted to low carbon inputs and have a diversified metabolic capacity, while copiotrophs are fast growing and are adapted

99 to large inputs of carbon. Thus, copiotrophs are typically found in soils with high carbon turnover rates and oligotrophs in soils with low carbon turnover rates. The relative abundance of copiotrophic and oligotrophic bacteria in high Arctic soils might influence the rate of CO2 emissions from these soils and it is thus crucial to characterize these bacterial groups in typical Arctic soils. Soil parameters other than the quality and abundance of the carbon substrates provided by the vegetation cover might play an important role in structuring soil bacterial communities. Several studies have indicated that the soil pH was the main factor influencing bacterial community composition while other soil parameters had a minor impact. This strong effect of the pH on soil bacterial community structure was detected at the continent level (Fierer and Jackson, 2006; Lauber et al., 2009) and in various environments including arable soils (Baker et al., 2009; Rousk et al., 2010), wetland soils (Hartman et al., 2008) and the Arctic (Chu et al., 2010). Thus, soil characteristics like pH are likely to influence the composition of the bacterial community in Arctic soils, which could in turn influence CO2 emission rates.

In this paper we focus on the links between community structure and CO2 emissions in three largely different soil environments encountered in the Canadian high Arctic. Microarrays and real-time PCR (qPCR) targeting 16S rRNA genes were used to characterize the bacterial communities at two sampling depths within the active layer of the soil. The microbiological data were linked to soil physico- chemical characteristics and CO2 production rates. We found that pH was an important parameter in structuring bacterial communities in these soils at the genera/species levels. Shifts in community structure were also visible at the phyla/classes levels, with the soil CO2 production rate being positively correlated to the relative abundance of the Alphaproteobacteria, Bacteroidetes and

Betaproteobacteria. These results indicated that CO2 production in Arctic soils might be influenced by the presence of specific groups of bacteria that have the capacity to actively degrade the soil carbon and does not only depend on the environmental conditions.

100 6.3 Material and methods

6.3.1 Site description, soil characterization and CO2 production Soil samples were collected within a 2 km radius from the McGill Arctic Research Station (MARS), located 8 km inland at Expedition Fjord, Axel Heiberg Island, Nunavut, in the Canadian high Arctic (79.433N, 90.766W). Three different soil types with distinct characteristics soils were selected and grouped into acidic soils (AC; 2 sites in 2008, 1 site in 2009), upland tundra soils (UT; 2 sites in 2008, 1 site in 2009) and wet meadow soils (WM; 2 sites in 2008, 2 sites in 2009). Soil descriptions and general characteristics are presented in Table 6.1. Soil samples were aseptically collected in April 2008 and July 2009 from the top 15 cm and at a depth of 40 cm and were kept frozen at -20ºC until used. Soil pH was determined after vortexing a soil: water (1:1 mass: volume) suspension for 2 minutes. Soil nitrogen concentration was determined by Maxxam Analytics (St- Laurent, Québec) according to their standard operating method (Kjeldahl method). Organic matter content was determined by loss on ignition following combustion for 16 h at 375oC (Beaudoin, 2003) and the water content was determined at 105oC using a moisture analyzer (MA 30, Sartorius, Mississauga,

ON). The CO2 concentration in the soil samples was determined by extracting the

CO2 from the soil pore water solution using the protocol described for CH4 by Wagner et al. (2003), with some modifications. Briefly, 3 g of soil were added to a 20 mL vial containing 6 mL of saturated sodium chloride solution. The vial was crimp-sealed, vortexed for 30 s and incubated for one hour at 80°C to allow for the transfer of the CO2 from the soil solution to the headspace of the vial. The

CO2 production rate was determined by measuring the headspace CO2 concentration over time in 20 mL crimp-sealed vials containing 5 g of soil incubated at 10°C. The CO2 concentration in the headspace was determined by gas chromatography as described by Roy and Greer (2000). In situ CO2 flux measurements were performed on one representative site from each soil type in

July 2010 using the Li-Cor Li-8100 Automated Soil CO2 Flux System (Li-Cor Environmental, Lincoln, NE).

101 6.3.2 DNA extraction, purification and quantification DNA was co-extracted with RNA from 3 g of soil samples using the MoBio RNA PowerSoil Total RNA Isolation Kit and eluted from the purification column using the RNA PowerSoil DNA Elution Accessory Kit (MoBio Laboratories, Carlsbad, CA) following the manufacturer’s instructions. DNA extracts were purified using the OneStep PCR Inhibitor Removal Kit (Zymo Research, Orange, CA). DNA concentration in the purified extracts was determined using the PicoGreen® dsDNA quantitation assay (Invitrogen, Carlsbad, CA).

6.3.3 Real-time PCR Real-time PCR targeting the 16S rRNA gene of all bacteria and of six different bacterial phyla/classes (Acidobacteria, Actinobacteria, Alphaproteobacteria, Bacteroidetes, Betaproteobacteria, Firmicutes) were performed in 20 μl volumes containing 1-5 ng of template DNA using the QuantiTect SYBR Green PCR kit (Qiagen) on a Rotor-Gene 3000 apparatus (Corbett Life Science, Sydney, NSW, Australia) as previously described (Yergeau et al., 2009), with some modifications to the annealing temperatures, which were as follows: all bacteria, 53oC; Acidobacteria, 50oC; Actinobacteria, 60oC; Alphaproteobacteria, 55oC; Bacteroidetes, 60oC; Betaproteobacteria, 55oC; Firmicutes, 60oC. Standards were prepared from 10-fold dilutions of linearized plasmids containing the gene fragment of interest that was cloned from amplified soil DNA.

6.3.4 16S rRNA gene PCR amplification and microarray analysis The 16S rRNA genes were amplified from each DNA sample using the F1-R13 eubacterial 16S rRNA gene primers (Lane, 1991). PCRs were performed in 50 uL volumes containing 5-15 ng of template DNA, 0.5 μM of each oligonucleotide primer, 200 μM of each dNTP, 1 mM MgCl2, 5 μg of BSA, 5 μL of 10x buffer and 2.5 U of Taq polymerase (Amersham Biosciences). PCR amplification used a touchdown protocol: the annealing temperature was set to

102 65°C and decreased by 1°C at every cycle for 10 cycles, and then 20 additional cycles were performed at 55°C. After 5 min of initial denaturation at 94°C, each PCR cycle consisted of 45 sec at 94°C, 45 sec at the annealing temperature, and 1.5 min at 72°C. PCR products were purified using the PureLink™ PCR Purification Kit (Invitrogen Canada, Burlington, ON, Canada). A microarray platform targeting the 16S rRNA genes of bacteria and archaea found in cold environments consisting of 525 25-mer oligonucleotides targeting 159 taxa was used as previously described (Yergeau et al., 2009). To summarize, the purified 16S rRNA gene PCR products were chemically labeled with Cy5 and hybridized overnight onto the microarray. Details about microarray and probe design are available at the National Center for Biotechnology Information (NCBI) GEO database under the platform accession number GPL8953. The presence–absence of the different taxa targeted by the microarray was used as a community profile to evaluate the similarity between the soil samples.

6.3.5 Statistical analyses All statistical analyses were performed in R (version 2.9.0, The R Foundation for Statistical Computing). For analysis of variance (ANOVA) of the soil parameters and the qPCR data, normality was tested using the “shapiro.test” function. When necessary, data were log transformed to meet parametric ANOVA assumptions. ANOVA and post hoc Tukey HSD tests were then carried out using the “aov” and “TukeyHSD” functions, respectively. When transformations failed to normalize data, Kruskal-Wallis and associated multiple comparison tests were performed using “kruskal.test” and the “kruskalmc” functions of the “pgirmess” library, respectively. The effect of sampling depth was tested using the paired Wilcoxon signed rank test (“wilcox.test function”). For the microarray analysis, distance matrices were calculated using the ”vegdist” function of the ”vegan” library and principal coordinate analyses were carried out using the ”cmdscale” function based on the square root of 1-Jaccard. Pearson linear correlation analyses were performed using the “cor.test” function. Multiple regression was performed

103 with the “lm” function with forward and backward stepwise variable selection (“stepAIC” function).

6.4 Results 6.4.1 Soil characteristics Soil characteristics were not significantly affected by the sampling date (Spring 2008 or Summer 2009) and, therefore, results from both sampling dates were combined in subsequent statistical analyses. Sampling depth had a significant effect on nitrogen (p=0.037), water content (p=0.014), bacterial abundance (p=0.002) and pH (p=0.015) of the soils, with values generally being higher for nitrogen, water content and bacterial abundance but lower for the pH in the top soils than in the 40 cm deep soils (Table 6.2). The soil type had a significant effect on all of the soil parameters, but in some cases the effect was variable according to depth (Table 6.2). At a depth of 40 cm, no significant differences were observed between soil types for organic matter, nitrogen and water content, while significant differences were observed for these parameters in the top soils. The wet meadow top soils had higher organic matter, nitrogen and water content than the acidic soils, while the upland tundra soils had intermediate values (Table 6.2). Wet meadow soils from both sampling depths also had higher CO2 concentrations than the other soil types. Soil pH and bacterial abundance were significantly lower in acidic soils than in the other soil types at both sampling depths (Table 6.2). Significant correlations were detected between most of the soil characteristics measured in this study and the organic matter content of the soil. The nitrogen content (r=0.78, p<0.001), the water content (r=0.94, p<0.001), the

CO2 concentration (r=0.68, p<0.001) and the bacterial abundance (r=0.52, p=0.019) were all significantly positively correlated to the organic matter content. The pH was the only soil parameter that had no significant correlation with the organic matter content.

104 Table 6.1 - Soil description.

Number of Active layer Water table Water holding In situ CO2 Soil type Description a a b c sites depth depth capacity flux 2 2008/2009 cm cm % μmol/m /s In the flood plain of an acidic lake. Surface white salt encrustations top: 23.97 Acidic soil 2/1 60 40 0.395 ± 0.017 (calcium and magnesium sulfates). 40 cm: 25.55 Low vegetative cover.

Dry aerated soil; moderate top: 57.84 Upland tundra 2/1 60 >40 1.418 ± 0.012 vegetative cover. 40 cm: 35.04

Close to the runoff of a lake; highly top: 69.34 Wet meadow 2/2 vegetated. Partly flooded with water 40 <20 1.378 ± 0.057 40 cm: 21.59 in the summer. aInformation recorded in July 2009. bDetermined on one representative sample of each depth from the corresponding soil type collected in 2009. cDetermined on one representative site from each soil type in July 2010. Values are means of four replicate measurements performed at different time points on the same site. Standard error of the mean is presented.

Table 6.2 - Soil characteristics. Soil type organic matter pH total nitrogen Water content Bacterial abundance (107 CO2 concentration CO2 production (µmoL (%) (mg/kg) (%) copies of 16S/g soil) (µmoL CO2/g soil) CO2/g soil/day) top 40 top 40 top 40 top 40 top 40 top 40 top 40 Acidic soil 3.7a 4.7a 4.8a 4.8a 893a 1133a 21.1a 21.2a 8.56a 2.76a 0.72a 0.89a 0.057a 0.067a Upland tundra 10.7a 6.6a 6.4b 6.9b 2933b 2200a 27.1a 16.7a 50.9b 32.6b 1.63a 1.38a 0.624ab 0.238a Wet meadow 27.0b 5.5a 6.0b 7.0b 4575b 1458a 72.8b 22.3a 41.9b 24.9ab 4.20b 3.67b 0.968b 0.244a p-value 0.0023 0.462 0.0021 <0.001 0.0042 0.1431 0.0011 0.6093 0.0048 0.0215 0.0011 <0.001 0.0090 0.0841 Different letters within a column refer to significantly (P<0.05) different averages based upon Tukey–HSD test. top: value for the top soil. 40: value for the 40 cm deep soil.

105 6.4.2 16S rRNA gene microarray analysis A microarray targeting the 16S rRNA gene of microorganisms frequently found in cold environments consisting of 525 25-mer oligonucleotides targeting 159 bacterial and archaeal taxa was used as a “cold-adapted” community fingerprinting method. On average, 36 taxa were detected per individual sample, ranging from 20 to 53 taxa. The sampling date or depth did not have a significant effect on the number of taxa detected, while significant differences were detected between soil types (p<0.001). A lower number of taxa were detected in acidic soils (average of 25.7 taxa per soil sample) than in the upland tundra soils (average of 39.0 taxa per soil sample) or in the wet meadow soils (average of 42.0 taxa per soil sample). Principal coordinate analysis (PCoA) was used to produce ordination graphs in which the samples were positioned according to their similarity as assessed by the 16S rRNA gene microarrays (Figure 6.1A). The acidic soils were clearly separated from the upland tundra and wet meadow soils on the first axis. The second axis separated most of the samples collected in 2008 (winter, black symbols in Figure 6.1A) from the samples collected in 2009 (summer, white symbols in Figure 6.1A). No clear grouping of the samples according to their sampling depth was observed.

106 A

B

Figure 6.1 - Phylogenetic community composition of the acidic soils (AC), upland tundra soils (UT) and wet meadow soils (WM) from Axel Heiberg Island, in the Canadian high Arctic. (A) Principal coordinate analysis based on Jaccard similarity calculated from hybridization patterns of DNA extracted from 6 AC soil samples, 6 UT soil samples and 8 WM soil samples using a 16S rRNA gene microarray targeting microorganisms found in cold environments. Black symbols refer to samples collected in 2008 (winter); white symbols refer to samples collected in 2009 (summer). (b) Percentage of total bacterial abundance for different phyla or classes based on qPCR quantification of DNA extracted from 6 AC soil samples, 6 UT soil samples and 8 WM soil samples. Average values for each soil type and the associated standard error are presented. Different letters within a phylum/class refer to significantly different (P<0.05) averages based upon Tukey–HSD test. Acido: Acidobacteria; Actino, Actinobacteria; Alpha-: Alphaproteobacteria; Bactero: Bacteroidetes; Beta-: Betaproteobacteria

107 6.4.3 Real-time PCR The abundance of the phyla or classes targeted by 16S rRNA gene qPCR was not significantly affected by the sampling date and, therefore, the results from both sampling dates were combined in subsequent statistical analyses. Depth had a significant effect (p=0.0098) on the abundance of members from the Firmicutes only, with higher values in soils from 40 cm deep than in top soils. No other bacterial group was significantly affected by the sampling depth and, therefore, combined data for both depths are presented (Figure 6.1B). In the three soil types, members of the Actinobacteria were dominant, followed by the Acidobacteria and the Alphaproteobacteria (Figure 6.1B). The soil type had a significant effect on the relative abundance of the Bacteroidetes (p<0.001) and the Betaproteobacteria (p=0.0024), with higher values in the wet meadow soils than in the two other soil types (Figure 6.1B). The relative abundance of the Firmicutes was below 1% and was not significantly affected by the soil type (data not shown). The relative abundance of two of the bacterial groups analyzed, the Betaproteobacteria and the Bacteroidetes, was significantly correlated to the soil

CO2 concentration (r=0.570, p=0.0087 and r=0.607, p=0.0045, respectively). The relative abundance of the Actinobacteria was negatively correlated to the soil bacterial abundance (r=-0.593, p=0.0059). None of the other measured soil parameters were significantly correlated to the abundance of the different bacterial phyla/classes quantified.

6.4.4 Carbon dioxide production rate and in situ CO2 flux

The soil CO2 production rate as determined in microcosms was not significantly affected by the sampling date and, therefore, the results from both sampling dates were combined in subsequent statistical analyses. The sampling depth had a significant effect on the soil CO2 production rate (p=0.0098), with generally higher values in the top soils than in the 40 cm deep soils. The soil type also had a significant effect on the CO2 production rate of the top soils (p<0.001), with higher values in the wet meadow soils, intermediate values in the upland tundra soils and lower values in the acidic soils (Table 6.2). No significant

108 differences between soil types were observed at a depth of 40 cm. In situ CO2 flux was lower for the acidic soil than for the two other soil types, which had similar values (Table 6.1). The statistical significance of these differences could not be tested because the measurements were performed on a single site from each soil type.

The soil CO2 production rate was significantly correlated to several of the soil characteristics analyzed: organic matter content (r=0.599, p=0.0053), nitrogen content (r=0.745, p<0.001), water content (r=0.672, p=0.0012), pH (r=0.540, p=0.014), bacterial abundance (r=0.650, p=0.0019) and CO2 concentration

(r=0.477, p=0.033). The soil CO2 production rate was also positively correlated to the relative abundance of the Alphaproteobacteria (r=0.674, p=0.0011), the Bacteroidetes (r=0.575, p=0.0080) and the Betaproteobacteria (r=0.568, p=0.0090). Multiple regression with forward-stepwise selection was performed to identify the bacterial phyla/classes that most significantly influenced the soil CO2 production rate and to generate a simple model for CO2 production in Arctic aerobic soils. Four bacterial phyla/classes were selected and generated a significant model (adjusted R2=0.635, p<0.001) (Table 6.3). Positive coefficients were found for the Alphaproteobacteria and the Betaproteobacteria, while negative coefficients were found for the Acidobacteria and Actinobacteria (Table 6.3).

Table 6.3 - Multiple regression coefficients for CO2 production rates. Coefficient p-value Intercept 4.460 0.055 Alpha- 0.575 0.001 Beta- 0.400 0.083 Actinobacteria -0.116 0.008 Acidobacteria -0.256 0.032 Variables were selected using a forward stepwise procedure. Adjusted R2=0.635, p<0.001

109 6.5 Discussion

6.5.1 Links between CO2 production and bacterial community structure In the context of global warming, the presence of large amounts of carbon in permafrost is raising serious concerns about whether melting permafrost, and the resulting increase in microbial activity, might be a source of extensive emissions of CO2 to the atmosphere. In this study, we found that different high

Arctic soil types had different CO2 production capacities, based on the CO2 production rate in aerobic microcosm assays and in situ CO2 flux measurements, with lower values being detected in acidic soil (low vegetative cover) than in upland tundra soils or wet meadow soils (moderate and high vegetative covers).

The soil CO2 concentration followed similar trends, with lower values in the acidic soils, intermediate values in the upland tundra soils and higher values in the wet meadow soils. As expected, all these CO2-related values were positively correlated to the organic matter content of the soils and other related soil characteristics, but more striking was the correlation between these values and the relative abundance of specific bacterial phyla/classes. The relative abundance of the Bacteroidetes and the Betaproteobacteria in soils was found to be positively correlated to the soil CO2 concentration, while the relative abundance of these two bacterial groups and of the Alphaproteobacteria was found to be positively correlated to the soil CO2 production rate. Similar results were found in a study looking at the relative abundance of six bacterial phyla/classes in 71 different soils from a wide range of ecosystems in North America, with positive correlations between the relative abundance of the Bacteroidetes and Betaproteobacteria and the soil C mineralization rate but a negative correlation between the relative abundance of the Acidobacteria and the soil carbon mineralization rate being reported (Fierer et al., 2007). The authors suggested that, because of their higher occurrence in soils with high C availability and their positive response to carbon input, the Bacteroidetes and Betaproteobacteria possess attributes that can be associated with copiotrophic organisms, while the Acidobacteria possess characteristics related to oligotrophic organisms. The Actinobacteria are typically described as being oligotrophic (or r-strategists)

110 (Atlas and Bartha, 1997), while other studies have shown that the Alphaproteobacteria are more abundant in soils with higher organic matter input (Griffiths et al., 2006; Thomson et al., 2010). These high-level ecological characteristics are important in the context of permafrost thaw since, as we showed here, copiotrophs are normally associated with soils having a high CO2 production and oligotrophs are associated with soils having a low CO2 production.

Multiple regression analysis indicated that the soil CO2 production rate can be modeled relatively well by the relative abundance of different groups of bacteria. Indeed, the relative abundance of two of the four bacterial groups identified by the stepwise selection, the Acidobacteria and Actinobacteria

(oligotrophic bacteria), was negatively correlated to the soil CO2 production rate, while the relative abundance of the Alphaproteobacteria and the Betaproteobacteria (copiotrophic bacteria) was positively correlated to the soil

CO2 production rate. These results indicated that CO2 production in Arctic soils does not only depend on the environmental conditions or on soil characteristics like the organic matter content, but also largely on the presence of specific groups of bacteria that have the capacity to actively degrade the soil carbon. Although a restricted number of soils were included in our analysis, another potential implication of these findings is that the determination of the relative abundance of a few bacterial phyla/classes in Arctic soils by qPCR could be used to identify the soils that have a higher potential for CO2 production. However, a relatively large amount of variation was unexplained (36.5%) and other factors, in addition to the parameters entered in the model (like fungal abundance, etc.), could influence soil

CO2 production.

6.5.2 Links between soil characteristics and bacterial community structure Another aspect covered in the context of this study was the impact of the soil physico-chemical characteristics on the bacterial community structure. Three high Arctic soils from the same geographical location but having highly distinctive characteristics, mainly in terms of pH and vegetative cover, were chosen as representatives of various Arctic settings. The acidic soils had a low pH

111 and vegetative cover, the upland tundra soils had near neutral pH and moderate vegetative cover and the wet meadow soils had near neutral pH and high vegetative cover. The principal coordinate analysis of the microarray patterns associated with each sample clearly separated the acidic soils from the two other soil types in the ordination, while a significantly lower number of taxa were detected in this soil type. The acidic soils had several distinctive characteristics as compared to the two other soil types, but most of these differences were detected in the top soils only, while soils from 40 cm deep had more similar characteristics. Because samples from both sampling depths grouped together in the ordination, the soil pH, which was significantly lower in the acidic soils than in the other soil types at both sampling depths, appears as one parameter that could explain the grouping observed. Soil pH was previously described as the main factor influencing the bacterial community composition in soils from 29 different sites in the Arctic, as assessed by 16S rRNA pyrosequencing (Chu et al., 2010). The same trend was observed in other environments including arable soils (Baker et al., 2009; Rousk et al., 2010) and wetland soils (Hartman et al., 2008), but also across a wide range of ecosystems at the continental scale (Fierer and Jackson, 2006; Lauber et al., 2009). As previously suggested by Rousk et al. (2010), one hypothesis that could explain the importance of pH in structuring soil bacterial communities might be related to the narrow pH range at which most bacterial taxa can grow. The principal coordinate analysis of the microarray patterns associated with each sample separated most of the samples collected in 2008 (winter conditions) from the samples collected in 2009 (summer conditions) in the ordination. This might be indicative of a seasonal effect on the bacterial community due to the high variations of the air temperature between seasons observed in the high Arctic. Seasonal fluctuation of bacterial populations in soils has often been reported for cold environments (Lipson, 2007; Lipson and Schmidt, 2004; Zinger et al., 2009). The effect of soil pH and the seasonal shift observed by microarray analysis of the 16S rRNA genes at the genera/species levels were not detectable

112 when looking at the variation of the relative abundance of the 6 bacterial phyla/classes quantified by qPCR. Indeed, there was no significant correlation between the relative abundance of the 6 bacterial groups and the soil pH, and no significant differences of their relative abundances were detected between seasons. This might be related to the fact that the 16S rRNA gene microarray is a qualitative tool designed to target specific genera/species of bacteria (lower taxonomic levels), while the qPCR assays were designed to quantify bacterial phyla or classes (higher taxonomic levels). Similar results were found by Wallenstein et al. (2007) in Arctic tundra tussock and shrub soils, where vegetation was found to be the primary driver of microbial community composition at the phyla and subphyla levels but seasonal shifts were observed at finer taxonomic levels. As indicated by these authors, it is possible that season affects specific genera/species of bacteria, but that within each phylum/class, there are species or genera that are specially adapted to winter or summer conditions. Although this study included shallow and deep active layer soils that were significantly different in terms of nitrogen content, water content, bacterial abundance and pH, the sampling depth had no clear effect on the microarray patterns associated with each sample and no significant effect on the relative abundance of five of the six bacterial groups quantified by qPCR was detected. The Firmicutes was the only bacterial group for which the abundance was significantly affected by the sampling depth, with higher values at a depth of 40 cm than in the top soils. Using a metagenomic approach, Yergeau et al. (2010) found that, overall, the active layer soil and the 2 meter permafrost from the same borehole in the high Arctic had highly similar functional and phylogenetic community compositions, with the microorganisms present in the permafrost layer representing a subset of the microorganisms found in the active layer. Interestingly, they also found a higher abundance of the Firmicutes in the permafrost than in the active layer soil and suggested that this might be related to a better capacity of members of this bacterial group to endure the harsher conditions found in deeper soil zones (Yergeau et al., 2010).

113 6.6 Conclusions In this study, differences in community structure in different high Arctic soils were detected at the genera/species levels using microarrays of the 16S rRNA gene and appeared to be related to soil pH and seasonal changes. Shifts in community structure were also detected at the phyla/classes levels and were related to the soil CO2 production rate. Soils with a higher relative abundance of copiotrophic bacterial groups (Alphaproteobacteria, Betaproteobacteria and

Bacteroidetes) were also associated with higher CO2 production rates. These results indicated that CO2 production in Arctic soils might depend on the presence of specific bacterial groups that have the capacity to actively degrade soil carbon, and not only on the environmental conditions. The relatively higher abundance of copiotrophic bacterial taxa in high Arctic soils with high organic matter content might lead, upon warming, to a rapid increase in soil CO2 production. Extensive in situ measurements will be necessary to evaluate if this increase in CO2 production will lead to higher net CO2 fluxes from Arctic soils, as qualitative/quantitative modifications to the vegetation cover could act as a CO2 sink and reduce net CO2 fluxes. Furthermore, the soil water regime could potentially lead to anaerobic conditions under which CH4 would be the major greenhouse gas being produced upon permafrost thaw.

6.7 Acknowledgements

The authors thank Nadia C.S. Mykytczuk for in situ CO2 flux measurements, the Canadian Polar Continental Shelf Project (PCSP) for their logistical support and McGill University’s High Arctic Research Station. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Program, Northern Supplements Program, and Special Research Opportunities IPY Program. CM was supported by NSERC and FQRNT postgraduate scholarships for the duration of this project. Additional funding for CM was provided by the Department of Indian and Northern Affairs - Northern Scientific Training Program (NSTP).

114 Chapter 7 – General discussion and conclusions

General discussion and conclusions corresponding to the specific objectives of this study are presented in this section. A summary of the thesis is also presented.

7.1 Discussion and conclusions related to aerobic methane oxidation

7.1.1 Development of a protocol for the sensitive detection of DNA in CsCl density gradients

Stable isotope probing (SIP) of DNA is a powerful technique that allows for the identification of microorganisms implicated in the degradation of a specific substrate. In this thesis, an alternative protocol was presented in which ethidium bromide (EtBr) was replaced by SYBR safeTM to visualize DNA in cesium chloride (CsCl) density gradients for SIP assays. This protocol allowed for the detection of low DNA concentrations and was successfully applied for the detection of active methanotrophic bacteria in high Arctic soils. SYBR safeTM is a safe, sensitive and effective alternative to the use of EtBr in CsCl density gradients for DNA-SIP assays and it should, therefore, become widely used in further SIP studies.

7.1.2 Methane-oxidizing capacity of soils from the Canadian high Arctic Results from microcosm assays performed on samples collected in Eureka and Axel Heiberg Island indicated that soils from the Canadian high Arctic have the capacity to oxidize CH4 under various conditions. Methane oxidation was detected at 4°C, 10°C and room temperature (RT), with much higher rates being observed at RT than at 4°C. Although CH4 oxidation was observed without amendments, CH4 oxidation rates were strongly enhanced by the addition of nutrients, most probably because of the positive effect of nitrogen on methanotrophic bacterial growth in nitrogen-limited environments (Bodelier and Laanbroek, 2004), like high Arctic soils. Methane degradation at concentrations

115 as high as 10,000-50,000 ppm was detected for most soils tested, but some soils from Axel Heiberg Island, and more specifically well aerated upland tundra top soils, also had the capacity to degrade CH4 at a concentration of 15 ppm. Methane oxidation at this low concentration implies the presence in high Arctic soils of methanotrophic bacteria with a high affinity for CH4. The capacity of high Arctic soils to degrade CH4 at relatively low temperatures, without the addition of nutrients, and at both low and high CH4 concentrations, indicated that these soils might play an important role in mitigating CH4 emissions from the melting permafrost while acting as a sink for atmospheric CH4 through the activity of high-affinity methane-oxidizing bacteria. However, further studies, including in situ CH4 flux measurements, will be necessary to determine if this activity is occurring under natural conditions and can play a significant role in limiting CH4 emissions into the atmosphere under ambient conditions.

7.1.3 Diversity and community composition of the methanotrophs in different high Arctic soils

Previous work performed on soils from cold environments had found a low diversity of methanotrophic bacteria, with Type I methanotrophs from the genera Methylobacter and Methylomonas being generally dominant over Type II methanotrophs (Liebner et al., 2009; Vecherskaya et al., 1993; Wartiainen et al., 2003). In this work, a similar trend was observed when looking at three soils from Eureka with similar physico-chemical characteristics, in which active methanotrophic bacteria detected using DNA-SIP were mainly related to the Type I methanotroph Methylobacter tundripaludum. However, when looking at different soils from Axel Heiberg Island with highly distinctive physico-chemical characteristics (pH, organic matter content, nitrogen content, and water content), different groups of methanotrophic bacteria were detected. Putative atmospheric methane-oxidizing bacteria from the “upland soil cluster gamma” (USCγ) and “upland soil cluster alpha” (USCα) genotypes (Kolb, 2009) were reported for the first time in Arctic soils, while Type I methanotrophs and methanotrophic bacteria with pmoA gene sequences related to the amoA gene of ammonia-oxidizing

116 bacteria were also detected. These results suggested that methanotrophic bacterial diversity in soils from cold environments may be broader than previously described in the literature.

7.1.4 Links between the methanotrophic bacterial community composition, soil characteristics and CH4 oxidation capacity

The results from this research indicated that the methanotrophic bacterial community composition in high Arctic soils is affected by the soil characteristics. Using DNA-SIP, microarrays, and clone libraries, different groups of methanotrophic bacteria were detected in different soils from Eureka and Axel Heiberg Island with highly distinctive physico-chemical characteristics. The soil pH was the parameter that appeared to be the most important in structuring the methanotrophic bacterial communities in soils from Axel Heiberg Island, with the USCα and USCγ genotypes associated with acidic and neutral soils, respectively. In some soils, like the acidic soils from Axel Heiberg Island, the methanotrophic bacterial community structure was also affected by the sampling depth: the USCα genotype was dominant in the top soils and Type I methanotrophs were dominant at a depth of 40 cm. Because of their influence on the community structure, it appears from our results that the soil parameters can also influence the capacity of the soil to oxidize CH4. Thus, high rates of CH4 oxidation at both low and high

CH4 concentrations were important in upland tundra top soils, and these rates were positively correlated to the abundance of methanotrophs from the USCγ genotype. Further work should be performed in order to determine if the link between the presence of this genotype and high CH4 oxidation rates can be generalized to other Arctic soils.

7.2 Discussion and conclusions related to CO2 production

7.2.1 Carbon dioxide production capacity of different soils from the Canadian high Arctic

Although CO2 production was detected for all soils in this study, significant differences in the CO2 production rates, measured in microcosm assays

117 and in situ CO2 fluxes, were detected between acidic soils, upland tundra soils and wet meadow soils from Axel Heiberg Island. The acidic soils, characterized by a lower organic matter, nitrogen, and bacterial content than the two other soil types, also had the lowest CO2 concentrations, CO2 production rates and in situ CO2 fluxes. These results confirmed observations made in numerous studies indicating that CO2 production is influenced by several soil parameters (Mack et al., 2004; Reth et al., 2005; Thomson et al., 2010), which could, therefore, play an important role in determining the fate of the soil organic carbon in a warming climate.

7.2.2 Links between the bacterial community composition in different high Arctic soils and the soil characteristics

Members of the Actinobacteria, followed by the Acidobacteria and the Alphaproteobacteria, were found to be the dominant taxa in three distinct soils from the Canadian high Arctic. Actinobacteria are commonly found as dominant members of the bacterial community in Arctic soils (Steven et al., 2007; Vishnivetskaya et al., 2006; Yergeau et al., 2010) and, because members from this group have the capacity to degrade complex C compounds, they might play an important role in organic matter degradation in cold environments (Yergeau et al., 2010). The soil type appeared to have an important effect on the bacterial community structure in high Arctic soils. Thus, the relative abundance of two bacterial taxa, the Bacteroidetes and the Betaproteobacteria was significantly higher in the wet meadow soils than in the two other soil types. Microarray analyses also indicated an effect of the soil type on the bacterial community composition at the genera/species levels, with a highly distinctive bacterial community structure being associated with the acidic soils. As was the case for the methanotrophs, soil pH was an important parameter in structuring the bacterial community in these high Arctic soils. Soil pH is frequently reported as the main factor influencing the bacterial community composition in different environments (Baker et al., 2009; Fierer and Jackson, 2006; Hartman et al., 2008; Lauber et al.,

118 2009; Rousk et al., 2010) and the Arctic , and it appears that this also applies in Arctic soils (Chu et al., 2010).

7.2.3 Links between the bacterial community composition of high Arctic soils and CO2 production

In this study, the relative abundance of the Bacteroidetes and the

Betaproteobacteria in soils was found to be positively correlated to the soil CO2 concentration, while the relative abundance of these two bacterial groups and the

Alphaproteobacteria was found to be positively correlated to the soil CO2 production rate. These findings are in agrrement with results from other studies indicating that these three bacterial taxa have copiotrophic attributes (Fierer et al., 2007) and are relatively more abundant in environments with high carbon turnover rates (Fierer et al., 2007; Lipson, 2007; Thomson et al., 2010; Wallenstein et al., 2007). In addition, multiple regression analysis indicated that the soil CO2 production rate in high Arctic soils can be modeled relatively well using the relative abundance of different groups of bacteria. Indeed, the relative abundance of two bacterial groups, the Acidobacteria and Actinobacteria

(oligotrophic bacteria), was negatively correlated to the soil CO2 production rate, while the relative abundance of the Alphaproteobacteria and the Betaproteobacteria (copiotrophic bacteria) was positively correlated to the soil

CO2 production rate. Overall, the results from this work indicated that CO2 production in high Arctic soils is closely related to the presence of copiotrophic bacterial taxa which might, upon warming, lead to a rapid increase in soil CO2 production. Extensive in situ measurements will be necessary to evaluate if this increase in CO2 production will lead to higher net CO2 fluxes from Arctic soils, as qualitative/quantitative modifications to the vegetation cover could act as a CO2 sink and reduce net CO2 fluxes. Furthermore, the soil water regime could potentially lead to anaerobic conditions under which CH4 would be the major greenhouse gas being produced upon permafrost thaw.

119 7.3 Summary To get a better understanding of the links between soil micro-organisms, soil characteristics and carbon turnover in Arctic soils, the general objective of this thesis was to characterize the activity, diversity and community structure of the bacterial populations implicated in two aerobic microbially-driven processes of the carbon cycle, the aerobic oxidation of CH4 and the production of CO2, in different soils from the Canadian high Arctic. To achieve this objective, experiments were conducted on soils representative of different high Arctic settings from two different sites, Eureka (Ellesmere Island) and Axel Heiberg Island, both located in Nunavut, Canada. In the first part of this thesis (Chapters 3 and 4), a protocol for the sensitive detection of DNA in CsCl density gradients for SIP assays was developed and used to study active methanotrophic bacteria in soils from Eureka. In the second part of this thesis (Chapters 5 and 6), methane- oxidizing and carbon dioxide-producing bacteria were studied in three soils with highly distinctive physico-chemical characteristics and the microbiological data were linked to soil properties and CH4 oxidation or CO2 production capacities, respectively. The results of this research indicated that bacterial communities in high Arctic soils play an important role in two aerobic processes of the carbon cycle, CH4 oxidation and CO2 production. Methanotrophic bacteria and CH4 oxidation were detected in these soils and might play an important role in reducing CH4 emissions from the melting permafrost in the context of global warming, while contributing to the reduction of atmospheric CH4 concentrations through the activity of high-affinity methane-oxidizing bacteria. In addition, the relatively higher abundance of copiotrophic bacterial taxa in high Arctic soils with high organic matter content might lead, upon warming, to a rapid increase in soil

CO2 production. Further research is needed to assess the relevance of these findings under in situ conditions in a warming climate.

120 List of references

Abell GCJ, Stralis-Pavese N, Sessitsch A, Bodrossy L (2009). Grazing affects methanotroph activity and diversity in an alpine meadow soil. Environmental Microbiology Reports 1: 457-465.

Aislabie J, Saul D, Foght J (2006). Bioremediation of hydrocarbon-contaminated polar soils. Extremophiles 10: 171-179.

Allison SD, Wallenstein MD, Bradford MA (2010). Soil-carbon response to warming dependent on microbial physiology. Nature Geoscience 3: 336-340.

Andersson S, Nilsson SI (2001). Influence of pH and temperature on microbial activity, substrate availability of soil-solution bacteria and leaching of dissolved organic carbon in a mor humus. Soil Biology and Biochemistry 33: 1181-1191.

Angel R, Conrad R (2009). In situ measurement of methane fluxes and analysis of transcribed particulate methane monooxygenase in desert soils. Environmental Microbiology 11: 2598-2610.

Arneth A, Harrison SP, Zaehle S, Tsigaridis K, Menon S, Bartlein PJ et al (2010). Terrestrial biogeochemical feedbacks in the climate system. Nature Geoscience 3: 525-532.

Atlas R, Bartha R (1997). Microbial ecology: fundamentals and applications. Benjamin/Cummings Publishing: Menlo Park, CA.

Baani M, Liesack W (2008). Two isozymes of particulate methane monooxygenase with different methane oxidation kinetics are found in Methylocystis sp. strain SC2. Proceedings of the National Academy of Sciences of the United States of America 105: 10203-10208.

Baker GC, Smith JJ, Cowan DA (2003). Review and re-analysis of domain- specific 16S primers. Journal of Microbiological Methods 55: 541-555.

Baker KL, Langenheder S, Nicol GW, Ricketts D, Killham K, Campbell CD et al (2009). Environmental and spatial characterisation of bacterial community composition in soil to inform sampling strategies. Soil Biology and Biochemistry 41: 2292-2298.

Beal EJ, House CH, Orphan VJ (2009). Manganese- and iron-dependent marine methane oxidation. Science 325: 184-187.

Beaudoin A (2003). A comparison of two methods for estimating the organic content of sediments. Journal of Paleolimnology 29: 387-390.

121 Beller HR, Kane SR, Legler TC, Alvarez PJJ (2002). A real-time polymerase chain reaction method for monitoring anaerobic, hydrocarbon-degrading bacteria based on a catabolic gene. Environmental Science and Technology 36: 3977- 3984.

Bender M, Conrad R (1992). Kinetics of CH4 oxidation in oxic soils exposed to ambient air or high CH4 mixing ratios. FEMS Microbiology Letters 101: 261-269.

Bodelier PLE, Laanbroek HJ (2004). Nitrogen as a regulatory factor of methane oxidation in soils and sediments. FEMS Microbiology Ecology 47: 265-277.

Bodrossy L, Stralis-Pavese N, Murrell JC, Radajewski S, Weilharter A, Sessitsch A (2003). Development and validation of a diagnostic microbial microarray for methanotrophs. Environmental Microbiology 5: 566-582.

Bourne DG, McDonald IR, Murrell JC (2001). Comparison of pmoA PCR primer sets as tools for investigating methanotroph diversity in three Danish soils. Applied and Environmental Microbiology 67: 3802-3809.

Bousquet P, Ciais P, Miller JB, Dlugokencky EJ, Hauglustaine DA, Prigent C et al (2006). Contribution of anthropogenic and natural sources to atmospheric methane variability. Nature 443: 439-443.

Bowman J (2006). The Methanotrophs — The Families Methylococcaceae and Methylocystaceae. In: Dworkin M, Falkow S, Rosenberg E, Schleifer K-H and Stackebrandt E (eds). The prokaryotes, 3rd ed. Springer: New York. pp 266-289.

Bowman JP, McCammon SA, Skerratt JH (1997). Methylosphaera hansonii gen. nov., sp. nov., a psychrophilic, group I methanotroph from Antarctic marine- salinity, meromictic lakes. Microbiology-UK 143: 1451-1459.

Boyd ES, Skidmore M, Mitchell AC, Bakermans C, Peters JW Methanogenesis in subglacial sediments. Environmental Microbiology Reports 2: 685-692.

Cadisch G, Espana M, Causey R, Richter M, Shaw E, Morgan JAW et al (2005). Technical considerations for the use of 15N-DNA stable-isotope probing for functional microbial activity in soils. Rapid Communications in Mass Spectrometry 19: 1424-1428.

Campbell BJ, Polson SW, Hanson TE, Mack MC, Schuur EAG (2010). The effect of nutrient deposition on bacterial communities in Arctic tundra soil. Environmental Microbiology 12: 1842-1854.

Candrian U, Furrer B, Hofelein C, Luthy J (1991). Use of inosine-containing oligonucleotide primers for enzymatic amplification of different alleles of the gene coding for heat-stable toxin type I of enterotoxigenic Escherichia coli. Applied and Environmental Microbiology 57: 955-961.

122

Cariello NF, Keohavong P, Sanderson BJS, Thilly WG (1988). DNA damage produced by ethidium bromide staining and exposure to ultraviolet light. Nucleic Acids Research 16: 4157-4157.

Cavicchioli R (2006). Cold-adapted archaea. Nature Reviews Microbiology 4: 331-343.

Centre d'expertise environnementale du Québec (CEEQ) (1997). Méthode d’analyse - Sols - Dosage des hydrocarbures pétroliers C10 à C50. MA.410-HYD. 1.0. Ministère de l’environnement et de la faune du Québec.

Cebron A, Bodrossy L, Chen Y, Singer AC, Thompson IP, Prosser JI et al (2007a). Identity of active methanotrophs in landfill cover soil as revealed by DNA-stable isotope probing. FEMS Microbiology Ecology 62: 12-23.

Cebron A, Bodrossy L, Stralis-Pavese N, Singer AC, Thompson IP, Prosser JI et al (2007b). Nutrient amendments in soil DNA stable isotope probing experiments reduce the observed methanotroph diversity. Applied and Environmental Microbiology 73: 798-807.

Chapin FS, Shaver GR, Giblin AE, Nadelhoffer KJ, Laundre JA (1995). Responses of Arctic tundra to experimental and observed changes in climate. Ecology 76: 694-711.

Chen Y, Dumont MG, McNamara NP, Chamberlain PM, Bodrossy L, Stralis- Pavese N et al (2008a). Diversity of the active methanotrophic community in acidic peatlands as assessed by mRNA and SIP-PLFA analyses. Environmental Microbiology 10: 446-459.

Chen Y, Dumont MG, Neufeld JD, Bodrossy L, Stralis-Pavese N, McNamara NP et al (2008b). Revealing the uncultivated majority: combining DNA stable-isotope probing, multiple displacement amplification and metagenomic analyses of uncultivated Methylocystis in acidic peatlands. Environmental Microbiology 10: 2609-2622.

Chen Y, Murrell JC (2010). Ecology of aerobic methanotrophs and their role in methane cycling. In: Timmis KN (ed). Handbook of hydrocarbon and lipid microbiology. Springer Berlin Heidelberg. pp 3067-3076.

Chu H, Fierer N, Lauber CL, Caporaso JG, Knight R, Grogan P (2010). Soil bacterial diversity in the Arctic is not fundamentally different from that found in other biomes. Environmental Microbiology 12: 2998-3006.

Conrad R (1996). Soil microorganisms as controllers of atmospheric trace gases (H2, CO, CH4, OCS, N2O, and NO). Microbiological Reviews 60: 609-640.

123 Conrad R (2009). The global methane cycle: recent advances in understanding the microbial processes involved. Environmental Microbiology Reports 1: 285-292.

Costello AM, Auman AJ, Macalady JL, Scow KM, Lidstrom ME (2002). Estimation of methanotroph abundance in a freshwater lake sediment. Environmental Microbiology 4: 443-450.

Costello AM, Lidstrom ME (1999). Molecular characterization of functional and phylogenetic genes from natural populations of methanotrophs in lake sediments. Applied and Environmental Microbiology 65: 5066-5074.

Curry CL (2009). The consumption of atmospheric methane by soil in a simulated future climate. Biogeosciences 6: 2355-2367.

Davidson EA, Janssens IA (2006). Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440: 165-173.

Dedysh SN, Berestovskaya YY, Vasylieva LV, Belova SE, Khmelenina VN, Suzina NE et al (2004). Methylocella tundrae sp nov., a novel methanotrophic bacterium from acidic tundra peatlands. International Journal of Systematic and Evolutionary Microbiology 54: 151-156.

Dedysh SN, Khmelenina VN, Suzina NE, Trotsenko YA, Semrau JD, Liesack W et al (2002). Methylocapsa acidiphila gen. nov., sp nov., a novel methane- oxidizing and dinitrogen-fixing acidophilic bacterium from Sphagnum bog. International Journal of Systematic and Evolutionary Microbiology 52: 251-261.

Dedysh SN, Knief C, Dunfield PF (2005). Methylocella species are facultatively methanotrophic. Journal of Bacteriology 187: 4665-4670.

Dedysh SN, Liesack W, Khmelenina VN, Suzina NE, Trotsenko YA, Semrau JD et al (2000). Methylocella palustris gen. nov., sp nov., a new methane-oxidizing acidophilic bacterium from peat bogs, representing a novel subtype of serine- pathway methanotrophs. International Journal of Systematic and Evolutionary Microbiology 50: 955-969.

Dedysh SN, Panikov NS, Tiedje JM (1998). Acidophilic methanotrophic communities from sphagnum peat bogs. Applied and Environmental Microbiology 64: 922-929.

Degelmann DM, Borken W, Drake HL, Kolb S (2010). Different atmospheric methane-oxidizing communities in European Beech and Norway Spruce Soils. Applied and Environmental Microbiology 76: 3228-3235.

Deslippe JR, Egger KN, Henry GHR (2005). Impacts of warming and fertilization on nitrogen-fixing microbial communities in the Canadian high Arctic. FEMS Microbiology Ecology 53: 41-50.

124

Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS (1991). Touchdown PCR to circumvent spurious priming during gene amplification. Nucleic Acids Research 19: 4008-4008.

Dormann CF, Woodin SJ (2002). Climate change in the Arctic: using plant functional types in a meta-analysis of field experiments. Functional Ecology 16: 4-17.

Dumont MG, Murrell JC (2005). Stable isotope probing - linking microbial identity to function. Nature Reviews Microbiology 3: 499-504.

Dumont MG, Radajewski SM, Miguez CB, McDonald IR, Murrell JC (2006). Identification of a complete methane monooxygenase operon from soil by combining stable isotope probing and metagenomic analysis. Environmental Microbiology 8: 1240-1250.

Dunfield PF (2007). The soil methane sink. In: Reay D, Hewitt CN, Smith K and Grace J (eds). Greenhouse gas sinks. CABI Publishing: Wallingford, UK. pp 152- 170.

Dunfield PF, Belova SE, Vorob'ev AV, Cornish SL, Dedysh SN (2010). Methylocapsa aurea sp. nov., a facultatively methanotrophic bacterium possessing a particulate methane monooxygenase. International Journal of Systematic and Evolutionary Microbiology 60: 2659-2664.

Dunfield PF, Khmelenina VN, Suzina NE, Trotsenko YA, Dedysh SN (2003). Methylocella silvestris sp nov., a novel methanotroph isolated from an acidic forest cambisol. International Journal of Systematic and Evolutionary Microbiology 53: 1231-1239.

Dunfield PF, Yuryev A, Senin P, Smirnova AV, Stott MB, Hou S et al (2007). Methane oxidation by an extremely acidophilic bacterium of the phylum Verrucomicrobia. Nature 450: 879-882.

Dutaur L, Verchot LV (2007). A global inventory of the soil CH4 sink. Global Biogeochemical Cycles 21: GB4013.

Einola J-KM, Kettunen RH, Rintala JA (2007). Responses of methane oxidation to temperature and water content in cover soil of a boreal landfill. Soil Biology and Biochemistry 39: 1156-1164.

Eliasson PE, McMurtrie RE, Pepper DA, Strömgren M, Linder S, Ågren GI (2005). The response of heterotrophic CO2 flux to soil warming. Global Change Biology 11: 167-181.

125 Emanuele E, Markovitsi D, Millie P, Zakrzewska K (2005). UV spectra and excitation delocalization in DNA: Influence of the spectral width. Chemphyschem 6: 1387-1392.

Ettwig KF, Butler MK, Le Paslier D, Pelletier E, Mangenot S, Kuypers MMM et al (2010). Nitrite-driven anaerobic methane oxidation by oxygenic bacteria. Nature 464: 543-548.

Falkowski P, Scholes RJ, Boyle E, Canadell J, Canfield D, Elser J et al (2000). The global carbon cycle: A test of our knowledge of Earth as a system. Science 290: 291-296.

Fierer N, Bradford MA, Jackson RB (2007). Toward an ecological classification of soil bacteria. Ecology 88: 1354-1364.

Fierer N, Jackson RB (2006). The diversity and biogeography of soil bacterial communities. Proceedings of the National Academy of Sciences of the United States of America 103: 626-631.

Fortin N, Beaumier D, Lee K, Greer CW (2004). Soil washing improves the recovery of total community DNA from polluted and high organic content sediments. Journal of Microbiological Methods 56: 181-191.

Friedrich MW (2006). Stable-isotope probing of DNA: insights into the function of uncultivated microorganisms from isotopically labeled metagenomes. Current Opinion in Biotechnology 17: 59-66.

Fung I, John J, Lerner J, Matthews E, Prather M, Steele LP et al (1991). 3- Dimensional model synthesis of the global methane cycle. Journal of Geophysical Research-Atmospheres 96: 13033-13065.

Ganzert L, Jurgens G, Munster U, Wagner D (2007). Methanogenic communities in permafrost-affected soils of the Laptev Sea coast, Siberian Arctic, characterized by 16S rRNA gene fingerprints. FEMS Microbiology Ecology 59: 476-488.

Gilichinsky D (2002). Permafrost. In: Bitton G (ed). Encyclopedia of environmental microbiology. John Wiley & Sons: New York. pp 2367-2385.

Gilichinsky D, Vishnivetskaya T, Petrova M, Spirina E, Mamykin V, Rivkina E (2008). Bacteria in permafrost. In: Margesin R, Schinner F, Marx J-C and Gerday C (eds). Psychrophiles: from biodiversity to biotechnology. Springer Berlin Heidelberg: Berlin, Germany. pp 83-102.

Gounot A, Russel N (1999). Physiology of cold-adapted microorganisms. In: Margesin R and Schinner F (eds). Cold-adapted organsims. Ecology, physiology, enzymology and molecular biology. Springer: New York. pp 33-55.

126 Griffiths RI, Bailey MJ, McNamara NP, Whiteley AS (2006). The functions and components of the Sourhope soil microbiota. Applied Soil Ecology 33: 114-126.

Gründemann D, Schomig E (1996). Protection of DNA during preparative agarose gel electrophoresis against damage induced by ultraviolet light. Biotechniques 21: 898-903.

Hansen AA, Herbert RA, Mikkelsen K, Jensen LL, Kristoffersen T, Tiedje JM et al (2007). Viability, diversity and composition of the bacterial community in a high Arctic permafrost soil from Spitsbergen, Northern Norway. Environmental Microbiology 9: 2870-2884.

Hanson RS, Hanson TE (1996). Methanotrophic bacteria. Microbiological Reviews 60: 439-471.

Hartley IP, Hopkins DW, Garnett MH, Sommerkorn M, Wookey PA (2008). Soil microbial respiration in arctic soil does not acclimate to temperature. Ecology Letters 11: 1092-1100.

Hartman PS (1991). Transillumination can profoundly reduce transformation frequencies. Biotechniques 11: 747-748.

Hartman WH, Richardson CJ, Vilgalys R, Bruland GL (2008). Environmental and anthropogenic controls over bacterial communities in wetland soils. Proceedings of the National Academy of Sciences of the United States of America 105: 17842- 17847.

Hebraud M, Potier P (1999). Cold shock response and low temperature adaptation in psychrotrophic bacteria. Journal of Molecular Microbiology and Biotechnology 1: 211-219.

Henckel T, Jackel U, Schnell S, Conrad R (2000). Molecular analyses of novel methanotrophic communities in forest soil that oxidize atmospheric methane. Applied and Environmental Microbiology 66: 1801-1808.

Holmes AJ, Costello A, Lidstrom ME, Murrell JC (1995). Evidence that particulate methane monooxygenase and ammonia monooxygenase may be evolutionarily related. FEMS Microbiology Letters 132: 203-208.

Holmes AJ, Roslev P, McDonald IR, Iversen N, Henriksen K, Murrell JC (1999). Characterization of methanotrophic bacterial populations in soils showing atmospheric methane uptake. Applied and Environmental Microbiology 65: 3312- 3318.

Horz H-P, Rich V, Avrahami S, Bohannan BJM (2005). Methane-oxidizing bacteria in a California upland grassland soil: diversity and response to simulated global change. Applied and Environmental Microbiology 71: 2642-2652.

127

Huber T, Faulkner G, Hugenholtz P (2004). Bellerophon: a program to detect chimeric sequences in multiple sequence alignements. Bioinformatics 20: 2317– 2319.

Inagaki F, Nunoura T, Nakagawa S, Teske A, Lever M, Lauer A et al (2006). Biogeographical distribution and diversity of microbes in methane hydrate- bearing deep marine sediments on the Pacific Ocean Margin. Proceedings of the National Academy of Sciences of the United States of America 103: 2815-2820.

IPCC (2007). Climate change 2007: the physical science basis. Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press: Cambridge, 996pp.

Jonasson S, Michelsen A, Schmidt IK, Nielsen EV (1999). Responses in microbes and plants to changed temperature, nutrient, and light regimes in the Arctic. Ecology 80: 1828-1843.

Jugnia L-B, Ait-Benichou S, Fortin N, Cabral AR, Greer CW (2009). Diversity and dynamics of methanotrophs within an experimental landfill cover soil. Soil Science Society of America Journal 73: 1479-1487.

King GM, Schnell S (1994). Effect of increasing atmospheric methane concentration on ammonium inhibition of soil methane consumption. Nature 370: 282-284.

Knief C, Dunfield PF (2005). Response and adaptation of different methanotrophic bacteria to low methane mixing ratios. Environmental Microbiology 7: 1307-1317.

Knief C, Kolb S, Bodelier PLE, Lipski A, Dunfield PF (2006). The active methanotrophic community in hydromorphic soils changes in response to changing methane concentration. Environmental Microbiology 8: 321-333.

Knief C, Lipski A, Dunfield PF (2003). Diversity and activity of methanotrophic bacteria in different upland soils. Applied and Environmental Microbiology 69: 6703-6714.

Knief C, Vanitchung S, Harvey NW, Conrad R, Dunfield PF, Chidthaisong A (2005). Diversity of methanotrophic bacteria in tropical upland soils under different land uses. Applied and Environmental Microbiology 71: 3826-3831.

Knittel K, Boetius A (2009). Anaerobic oxidation of methane: progress with an unknown process. Annual Review of Microbiology 63: 311-334.

128 Kobabe S, Wagner D, Pfeiffer EM (2004). Characterisation of microbial community composition of a Siberian tundra soil by fluorescence in situ hybridisation. FEMS Microbiology Ecology 50: 13-23.

Kolb S (2009). The quest for atmospheric methane oxidizers in forest soils. Environmental Microbiology Reports 1: 336-346.

Kolb S, Knief C, Dunfield PF, Conrad R (2005). Abundance and activity of uncultured methanotrophic bacteria involved in the consumption of atmospheric methane in two forest soils. Environmental Microbiology 7: 1150-1161.

Kolb S, Knief C, Stubner S, Conrad R (2003). Quantitative detection of methanotrophs in soil by novel pmoA-targeted real-time PCR assays. Applied and Environmental Microbiology 69: 2423-2429.

Lane D (1991). 16S/23S rRNA sequencing. In: Stackebrandt E and Goodfellow M (eds). Nucleic acids techniques in bacterial systematics. John Wiley & Sons: West Sussex, United Kingdom. pp 115-175.

Lauber CL, Hamady M, Knight R, Fierer N (2009). Pyrosequencing-based assessment of soil pH as a predictor of soil bacterial community structure at the continental scale. Applied and Environmental Microbiology 75: 5111-5120.

Lawrence DM, Slater AG (2005). A projection of severe near-surface permafrost degradation during the 21st century. Geophysical Research Letters 32: L24401.

Liebner S, Rublack K, Stuehrmann T, Wagner D (2009). Diversity of aerobic methanotrophic bacteria in a permafrost active layer soil of the Lena Delta, Siberia. Microbial Ecology 57: 25-35.

Liebner S, Wagner D (2007). Abundance, distribution and potential activity of methane oxidizing bacteria in permafrost soils from the Lena Delta, Siberia. Environmental Microbiology 9: 107-117.

Lin J-L, Radajewski S, Eshinimaev BT, Trotsenko YA, McDonald IR, Murrell JC (2004). Molecular diversity of methanotrophs in Transbaikal soda lake sediments and identification of potentially active populations by stable isotope probing. Environmental Microbiology 6: 1049-1060.

Lipson DA (2007). Relationships between temperature responses and bacterial community structure along seasonal and altitudinal gradients. FEMS Microbiology Ecology 59: 418-427.

Lipson DA, Schmidt SK (2004). Seasonal changes in an alpine soil bacterial community in the Colorado Rocky Mountains. Applied and Environmental Microbiology 70: 2867-2879.

129 Luo Y, Wan S, Hui D, Wallace LL (2001). Acclimatization of soil respiration to warming in a tall grass prairie. Nature 413: 622-625.

MacDonald GJ (1990). Role of methane clathrates in past and future climates. Climatic Change 16: 247-281.

Mack MC, Schuur EAG, Bret-Harte MS, Shaver GR, Chapin FS (2004). Ecosystem carbon storage in Arctic tundra reduced by long-term nutrient fertilization. Nature 431: 440-443.

Martineau C, Whyte LG, Greer CW (2008). Development of a SYBR safeTM technique for the sensitive detection of DNA in cesium chloride density gradients for stable isotope probing assays. Journal of Microbiological Methods 73: 199- 202.

Martineau C, Whyte LG, Greer CW (2010). Stable isotope probing analysis of the diversity and activity of methanotrophic bacteria in soils from the Canadian high Arctic. Applied and Environmental Microbiology 76: 5773-5784.

Mastepanov M, Sigsgaard C, Dlugokencky EJ, Houweling S, Strom L, Tamstorf MP et al (2008). Large tundra methane burst during onset of freezing. Nature 456: 628-630.

McBean G, Alekseev G, Chen D, Førland E, Fyfe J, Groisman PY et al (2005). Arctic climate: past and present. In: Symon C, Arris L and Heal B (eds). Arctic climate impact assessment. Cambridge University Press: New-York. pp 21-60.

Melillo JM, Steudler PA, Aber JD, Newkirk K, Lux H, Bowles FP et al (2002). Soil warming and carbon-cycle feedbacks to the climate system. Science 298: 2173-2176.

Metje M, Frenzel P (2007). Methanogenesis and methanogenic pathways in a peat from subarctic permafrost. Environmental Microbiology 9: 954-964.

Mohanty SR, Bodelier PLE, Conrad R (2007). Effect of temperature on composition of the methanotrophic community in rice field and forest soil. FEMS Microbiology Ecology 62: 24-31.

Mohn WW, Stewart GR (2000). Limiting factors for hydrocarbon biodegradation at low temperature in Arctic soils. Soil Biology and Biochemistry 32: 1161-1172.

Morris SA, Radajewski S, Willison TW, Murrell JC (2002). Identification of the functionally active methanotroph population in a peat soil microcosm by stable- isotope probing. Applied and Environmental Microbiology 68: 1446-1453.

Muyzer G, Dewaal EC, Uitterlinden AG (1993). Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase

130 chain reaction-amplified genes-coding for 16S ribosomal RNA. Applied and Environmental Microbiology 59: 695-700.

Myneni RB, Keeling CD, Tucker CJ, Asrar G, Nemani RR (1997). Increased plant growth in the northern high latitudes from 1981 to 1991. Nature 386: 698- 702.

Neufeld J, Dumont M, Vohra J, Murrell J (2007a). Methodological considerations for the use of stable isotope probing in microbial ecology. Microbial Ecology 53: 435-442.

Neufeld JD, Mohn WW (2005). Unexpectedly high bacterial diversity in Arctic tundra relative to boreal forest soils, revealed by serial analysis of ribosomal sequence tags. Applied and Environmental Microbiology 71: 5710-5718.

Neufeld JD, Vohra J, Dumont MG, Lueders T, Manefield M, Friedrich MW et al (2007b). DNA stable-isotope probing. Nature Protocols 2: 860-866.

Niederberger TD, Perreault NN, Tille S, Lollar BS, Lacrampe-Couloume G, Andersen D et al (2010). Microbial characterization of a subzero, hypersaline methane seep in the Canadian high Arctic. The ISME journal 4: 1326-1339.

Noll M, Frenzel P, Conrad R (2008). Selective stimulation of type I methanotrophs in a rice paddy soil by urea fertilization revealed by RNA-based stable isotope probing. FEMS Microbiology Ecology 65: 125-132.

Omelchenko MV, Vasileva LV, Zavarzin GA, Saveleva ND, Lysenko AM, Mityushina LL et al (1996). A novel psychrophilic methanotroph of the genus Methylobacter. Microbiology 65: 339-343.

Omelchenko MV, Vasilyeva LV, Zavarzin GA (1993). Psychrophilic methanotroph from tundra soil. Current Microbiology 27: 255-259.

Op den Camp HJM, Islam T, Stott MB, Harhangi HR, Hynes A, Schouten S et al (2009). Environmental, genomic and taxonomic perspectives on methanotrophic Verrucomicrobia. Environmental Microbiology Reports 1: 293-306.

Pacheco-Oliver M, McDonald IR, Groleau D, Murrell JC, Miguez CB (2002). Detection of methanotrophs with highly divergent pmoA genes from Arctic soils. FEMS Microbiology Letters 209: 313-319.

Pawlowska M, Stepniewski W (2006). An influence of methane concentration on the methanotrophic activity of a model landfill cover. Ecological Engineering 26: 392-395.

131 Permafrost subcommittee (1988). Glossary of permafrost and related ground-ice terms. Associate Committee on Geotechnical Research, National Research Council of Canada, Ottawa.

Piao S, Ciais P, Friedlingstein P, Peylin P, Reichstein M, Luyssaert S et al (2008). Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature 451: 49-52.

Pietikäinen J, Pettersson M, Bååth E (2005). Comparison of temperature effects on soil respiration and bacterial and fungal growth rates. FEMS Microbiology Ecology 52: 49-58.

Popp TJ, Chanton JP, Whiting GJ, Grant N (2000). Evaluation of methane oxidation in the rhizosphere of a Carex dominated fen in northcentral Alberta, Canada. Biogeochemistry 51: 259-281.

Pospiech A, Neumann B (1995). A versatile quick-prep of genomic DNA from Gram-Positive bacteria. Trends in Genetics 11: 217-218.

Post WM, Emanuel WR, Zinke PJ, Stangenberger AG (1982). Soil carbon pools and world life zones. Nature 298: 156-159.

Radajewski S, Ineson P, Parekh NR, Murrell JC (2000). Stable-isotope probing as a tool in microbial ecology. Nature 403: 646-649.

Radajewski S, McDonald IR, Murrell JC (2004). Stable isotope probing of nucleic acids to identify active microbial populations. In: Kowalchuk GA, de Bruijn FJ, Head IM, Akkermans AD and van Elsas JD (eds). Molecular microbial ecology manual. Kluwer Academic Publishers: Dordrecht. pp 1661-1672.

Rahalkar M, Schink B (2007). Comparison of aerobic methanotrophic communities in littoral and profundal sediments of Lake Constance by a molecular approach. Applied and Environmental Microbiology 73: 4389-4394.

Reeburgh WS (2007). Oceanic methane biogeochemistry. Chemical Reviews 107: 486-513.

Reth S, Reichstein M, Falge E (2005). The effect of soil water content, soil temperature, soil pH-value and the root mass on soil CO2 efflux – A modified model. Plant and Soil 268: 21-33.

Rivkina E, Shcherbakova V, Laurinavichius K, Petrovskaya L, Krivushin K, Kraev G et al (2007). Biogeochemistry of methane and methanogenic archaea in permafrost. FEMS Microbiology Ecology 61: 1-15.

Rodrigues DF, Tiedje JM (2008). Coping with our cold planet. Applied and Environmental Microbiology 74: 1677-1686.

132

Roslev P, Iversen N (1999). Radioactive fingerprinting of microorganisms that oxidize atmospheric methane in different soils. Applied and Environmental Microbiology 65: 4064-4070.

Roslev P, Iversen N, Henriksen K (1997). Oxidation and assimilation of atmospheric methane by soil methane oxidizers. Applied and Environmental Microbiology 63: 874-880.

Rousk J, Baath E, Brookes PC, Lauber CL, Lozupone C, Caporaso JG et al (2010). Soil bacterial and fungal communities across a pH gradient in an arable soil. The ISME journal 4: 1340-1351.

Roy R, Greer CW (2000). Hexadecane mineralization and denitrification in two diesel fuel-contaminated soils. FEMS Microbiology Ecology 32: 17-23.

Rustad, Campbell, Marion, Norby, Mitchell, Hartley et al (2001). A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia 126: 543-562.

Schlesinger WH, Andrews JA (2000). Soil respiration and the global carbon cycle. Biogeochemistry 48: 7-20.

Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB et al (2009). Introducing mothur: open-source, platform-independent, community- supported software for describing and comparing microbial communities. Applied and Environmental Microbiology 75: 7537-7541.

Schuur EAG, Bockheim J, Canadell JG, Euskirchen E, Field CB, Goryachkin SV et al (2008). Vulnerability of permafrost carbon to climate change: implications for the global carbon cycle. BioScience 58: 701-714.

Schuur EAG, Vogel JG, Crummer KG, Lee H, Sickman JO, Osterkamp TE (2009). The effect of permafrost thaw on old carbon release and net carbon exchange from tundra. Nature 459: 556-559.

Shaw MR, Harte J (2001). Control of litter decomposition in a subalpine meadow- sagebrush steppe ecotone under climate change. Ecological Applications 11: 1206-1223.

Sheffield VC, Cox DR, Lerman LS, Myers RM (1989). Attachment of a 40-base- pair G + C-rich sequence (GC-clamp) to genomic DNA fragments by the polymerase chain reaction results in improved detection of single-base changes. Proceedings of the National Academy of Sciences of the United States of America 86: 232-236.

133 Simankova MV, Kotsyurbenko OR, Lueders T, Nozhevnikova AN, Wagner B, Conrad R et al (2003). Isolation and characterization of new strains of methanogens from cold terrestrial habitats. Systematic and Applied Microbiology 26: 312–318.

Singer VL, Lawlor TE, Yue S (1999). Comparison of SYBR® Green I nucleic acid gel stain mutagenicity and ethidium bromide mutagenicity in the Salmonella/mammalian microsome reverse mutation assay (Ames test). Mutation Research-Genetic Toxicology and Environmental Mutagenesis 439: 37-47.

Singleton DR, Furlong MA, Rathbun SL, Whitman WB (2001). Quantitative comparisons of 16S rRNA gene sequence libraries from environmental samples. Applied and Environmental Microbiology 67: 4374-4376.

Steudler PA, Bowden RD, Melillo JM, Aber JD (1989). Influence of nitrogen fertilization on methane uptake in temperate forest soils. Nature 341: 314-316.

Steven B, Briggs G, McKay CP, Pollard WH, Greer CW, Whyte LG (2007). Characterization of the microbial diversity in a permafrost sample from the Canadian high Arctic using culture-dependent and culture-independent methods. FEMS Microbiology Ecology 59: 513-523.

Steven B, Niederberger TD, Whyte LG (2009). Bacterial and archaeal diversity in permafrost. In: Margesin R (ed). Permafrost soils. Springer Berlin Heidelberg. pp 59-72.

Steven B, Pollard WH, Greer CW, Whyte LG (2008). Microbial diversity and activity through a permafrost/ground ice core profile from the Canadian high Arctic. Environmental Microbiology 10: 3388-3403.

Stoecker K, Bendinger B, Schöning B, Nielsen PH, Nielsen JL, Baranyi C et al (2006). Cohn’s Crenothrix is a filamentous methane oxidizer with an unusual methane monooxygenase. Proceedings of the National Academy of Sciences of the United States of America 103: 2363-2367.

Stralis-Pavese N, Sessitsch A, Weilharter A, Reichenauer T, Riesing J, Csontos J et al (2004). Optimization of diagnostic microarray for application in analysing landfill methanotroph communities under different plant covers. Environmental Microbiology 6: 347-363.

Tarnocai C, Campbell l (2006). Soils of the polar region. Encyclopedia of Soil Science: Second Edition: 1330-1333.

Tarnocai C, Canadell JG, Schuur EAG, Kuhry P, Mazhitova G, Zimov S (2009). Soil organic carbon pools in the northern circumpolar permafrost region. Global Biogeochemical Cycles 23: GB2023.

134 Thomson B, Ostle N, McNamara N, Bailey M, Whiteley A, Griffiths R (2010). Vegetation affects the relative abundances of dominant soil bacterial taxa and soil respiration rates in an upland grassland soil. Microbial Ecology 59: 335-343.

Trotsenko Y, Khmelenina V (2002). Biology of extremophilic and extremotolerant methanotrophs. Archives of Microbiology 177: 123-131.

Trotsenko YA, Khmelenina VN (2005). Aerobic methanotrophic bacteria of cold ecosystems. FEMS Microbiology Ecology 53: 15-26.

Tsubota J, Eshinimaev BT, Khmelenina VN, Trotsenko YA (2005). Methylothermus thermalis gen. nov., sp. nov., a novel moderately thermophilic obligate methanotroph from a hot spring in Japan. International Journal of Systematic and Evolutionary Microbiology 55: 1877-1884.

Tung HC, Bramall NE, Price PB (2005). Microbial origin of excess methane in glacial ice and implications for life on Mars. Proceedings of the National Academy of Sciences of the United States of America 102: 18292-18296.

Vecherskaya MS, Galchenko VF, Sokolova EN, Samarkin VA (1993). Activity and species composition of aerobic methanotrophic communities in tundra soils. Current Microbiology 27: 181-184.

Vishnivetskaya TA, Petrova MA, Urbance J, Ponder M, Moyer CL, Gilichinsky DA et al (2006). Bacterial community in ancient Siberian permafrost as characterized by culture and culture-independent methods. Astrobiology 6: 400- 414.

Wadham JL, Tranter M, Tulaczyk S, Sharp M (2008). Subglacial methanogenesis: A potential climatic amplifier? Global Biogeochemical Cycles 22: GB2021.

Wagner D, Gattinger A, Embacher A, Pfeiffer EM, Schloter M, Lipski A (2007). Methanogenic activity and biomass in Holocene permafrost deposits of the Lena Delta, Siberian Arctic and its implication for the global methane budget. Global Change Biology 13: 1089-1099.

Wagner D, Kobabe S, Liebner S (2009). Bacterial community structure and carbon turnover in permafrost-affected soils of the Lena Delta, northeastern Siberia. Canadian Journal of Microbiology 55: 73-83.

Wagner D, Kobabe S, Pfeiffer E-M, Hubberten H-W (2003). Microbial controls on methane fluxes from a polygonal tundra of the Lena Delta, Siberia. Permafrost and Periglacial Processes 14: 173-185.

Wagner D, Liebner S (2010). Methanogenesis in Arctic permafrost habitats. In: Timmis KN (ed). Handbook of hydrocarbon and lipid microbiology. Springer Berlin Heidelberg: Berlin, Germany. pp 655-663.

135

Wagner D, Lipski A, Embacher A, Gattinger A (2005). Methane fluxes in permafrost habitats of the Lena Delta: effects of microbial community structure and organic matter quality. Environmental Microbiology 7: 1582-1592.

Walker MD, Wahren CH, Hollister RD, Henry GHR, Ahlquist LE, Alatalo JM et al (2006). Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103: 1342-1346.

Wallenstein MD, McMahon S, Schimel J (2007). Bacterial and fungal community structure in Arctic tundra tussock and shrub soils. FEMS Microbiology Ecology 59: 428-435.

Wartiainen I, Hestnes AG, McDonald IR, Svenning MM (2006). Methylobacter tundripaludum sp nov., a methane-oxidizing bacterium from Arctic wetland soil on the Svalbard islands, Norway (78 degrees N). International Journal of Systematic and Evolutionary Microbiology 56: 109-113.

Wartiainen I, Hestnes AG, Svenning MM (2003). Methanotrophic diversity in high Arctic wetlands on the islands of Svalbard (Norway) - denaturing gradient gel electrophoresis analysis of soil DNA and enrichment cultures. Canadian Journal of Microbiology 49: 602-612.

Whalen SC, Reeburgh WS (1990). Consumption of atmospheric methane by tundra soils. Nature 346: 160-162.

Whiticar MJ (1999). Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161: 291-314.

Whittenbury R, Phillips KC, Wilkinson JF (1970). Enrichment, isolation and some properties of methane-utilizing bacteria. Journal of General Microbiology 61: 205-218.

Wise MG, McArthur JV, Shimkets LJ (1999). Methanotroph diversity in landfill soil: isolation of novel Type I and Type II methanotrophs whose presence was suggested by culture-independent 16S Ribosomal DNA analysis. Applied and Environmental Microbiology 65: 4887-4897.

Yergeau E, Arbour M, Brousseau R, Juck D, Lawrence JR, Masson L et al (2009). Microarray and real-time PCR analyses of the responses of high-Arctic soil cacteria to hydrocarbon pollution and bioremediation treatments. Applied and Environmental Microbiology 75: 6258-6267.

Yergeau E, Bokhorst S, Huiskes AH, Boschker HT, Aerts R, Kowalchuk GA (2007). Size and structure of bacterial, fungal and nematode communities along an Antarctic environmental gradient. FEMS Microbiology Ecology 59: 436-451.

136

Yergeau E, Hogues H, Whyte LG, Greer CW (2010). The functional potential of high Arctic permafrost revealed by metagenomic sequencing, qPCR and microarray analyses. The ISME journal 4: 1206 - 1214.

Zhang T, Barry RG, Knowles K, Heginbottom JA, Brown J (2008). Statistics and characteristics of permafrost and ground-ice distribution in the Northern Hemisphere. Polar Geography 31: 47 - 68.

Zhang W, Parker KM, Luo Y, Wan S, Wallace LL, Hu S (2005). Soil microbial responses to experimental warming and clipping in a tallgrass prairie. Global Change Biology 11: 266-277.

Zinger L, Shahnavaz B, Baptist F, Geremia RA, Choler P (2009). Microbial diversity in alpine tundra soils correlates with snow cover dynamics. The ISME journal 3: 850-859.

137

Appendices Microarray hybridization patterns

138 Probe AC1-top AC1-40 AC2-top AC2 40 Ac3-top AC3-40 UT1-top UT2-top UT3-top UT3-40 Mb_LW12-211 Mb_SL#3-300 BB51-299 Mb292 Mb282 Mb_URC278 LF1a-456 Mb_C11-403 Mb271 Type Ia MS1-440 Mm_RS311 Mm275 Mmb303 Mmb304 Mmb562 Mm229 LP20-644 DS2-220 USCγ USCG-225 USCG-225b JR3-505 501-375 MclS402 MclS394 MclS400 Type Ib Mcl404 Mcl408 JRC4-432 JHTY2-578 Mha-500 Mcy264 Mcy233 McyM309 Peat264 MsS314 Type II Msi263 Msi423 MsiT214 Msi269 Msi294 Msi232 NMcy1-247 NMsi1-469 RA14-299 USCα RA14-594 RA14-591 Wsh1-566 Wsh2-491 TUSC409 TUSC2 TUSC502

Figure A.1 - DNA microarray hybridization patterns of pmoA PCR products amplified using a two-step PCR approach targeting the pmoA and closely related genes from upland tundra soils (UT) and acidic soils (AC) from Axel Heiberg Island, in the Canadian high Arctic. Relative signal intensities are indicated as percentage of the individual hybridization potential of each probe, as follows: blue, below 2.5%; green, between 2.5-5%; yellow, between 5-20%; red, above 20%. In the identification of the samples, numbers 1 and 2 refer to samples from 2008 (winter) while numbers 3 and 4 refer to samples from 2009 (summer); “Top” and “40” indications refer to the sampling depth (within the top 15 cm or at a depth of 40 cm, respectively). USCα, upland soil cluster alpha; USCγ, upland soil cluster gamma; TUSC2, tropical upland soil cluster 2.

139 Probe UT1-top UT1-40 UT2-top UT2-40 UT3-top UT3-40 WM1-top WM1-40 WM2-top WM2-40 WM3-top WM3-40 WM4-top Mb292 Mmb303

Type Ia Mmb304 DS2-287 USCG-225 USCG-225b USCγ JR2-409 JR3-505 501-375 MclS402 MclS394 MclS400 Mcl404

Type Ib Mcl408 fw1-641 JHTY2-578 LW21-391 Mha-500 PmoC640 pmoC308 Mcy270 McyM309 Type II Peat264 MsS314 MsiT214 NMsi1-469 RA14-299 USCα RA14-594 RA14-591 AMO3-486 TUSC409

TUSC2 TUSC502 Pl6-306 NsNv207 NsNv363 SV308 SVrel583 Nit_rel471 Sed585 amoA‐ Sed422 related Nit_rel223 Nit_rel417 Nit_rel419 Nit_rel470 Nit_rel351 gp619

Figure A.2 - DNA microarray hybridization patterns of pmoA/amoA PCR products targeting the pmoA/amoA genes from methanotrophs, ammonia-oxidizing bacteria and homologous genes from environmental libraries and closely related genes from upland tundra soils (UT) and acidic soils (AC) from Axel Heiberg Island, in the Canadian high Arctic. Relative signal intensities are indicated as percentage of the individual hybridization potential of each probe, as follows: blue, below 2.5%; green, between 2.5-5%; yellow, between 5-20%; red, above 20%. In the identification of the samples, numbers 1 and 2 refer to samples from 2008 (winter) while numbers 3 and 4 refer to samples from 2009 (summer); “Top” and “40” indications refer to the sampling depth (within the top 15 cm or at a depth of 40 cm, respectively). USCα, upland soil cluster alpha; USCγ, upland soil cluster gamma; TUSC2, tropical upland soil cluster 2.

140