Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A multi-omic perspective on microbial mediated methane oxidation in the Saanich Inlet water column Mónica, Torres Beltrán 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2019_february_torres_beltran_monica.pdf [ 58.69MB ]
Metadata
JSON: 24-1.0374919.json
JSON-LD: 24-1.0374919-ld.json
RDF/XML (Pretty): 24-1.0374919-rdf.xml
RDF/JSON: 24-1.0374919-rdf.json
Turtle: 24-1.0374919-turtle.txt
N-Triples: 24-1.0374919-rdf-ntriples.txt
Original Record: 24-1.0374919-source.json
Full Text
24-1.0374919-fulltext.txt
Citation
24-1.0374919.ris

Full Text

A multi-omics perspective on microbial mediated methane oxidation in theSaanich Inlet water columnbyMo´nica Torres Beltra´nB.Sc. Oceanography, Universidad Auto´noma de Baja California, 2007M.Sc. Coastal Oceanography, Universidad Auto´noma de Baja California, 2009A THESIS SUBMITTED IN PARTIAL FULFILLMENTOF THE REQUIREMENTS FOR THE DEGREE OFDoctor of PhilosophyinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Microbiology and Immunology)The University of British Columbia(Vancouver)November 2018c© Mo´nica Torres Beltra´n, 2018The following individuals certify that they have read, and recommend to the Faculty of Graduate and Post-doctoral Studies for acceptance, the thesis entitled:A multi-omics perspective on microbial mediated methane oxidation in the Saanich Inlet wa-ter columnsubmitted by Mo´nica Torres Beltra´n in partial fulfillment of the requirements for the degree of Doctor ofPhilosophy in Microbiology and Immunology.Examining Committee:Steven J. Hallam, Microbiology and ImmunologySupervisorMaria T. MaldonadoUniversity ExaminerKristin OriansUniversity ExaminerSamantha B. JoyeExternal ExaminerAdditional Supervisory Committee Members:Thomas J. Beatty, Microbiology and ImmunologySupervisory Committee MemberWilliam W. Mohn,Microbiology and ImmunologySupervisory Committee MemberMartin Hirst, Microbiology and ImmunologySupervisory Committee MemberiiAbstractMicrobial communities play an integral role in the biogeochemical cycling of carbon, nitrogen and sul-fur throughout the biosphere. These communities interact, forming metabolic networks that change andadapt in response to availability of electron donors and acceptors. Oxygen minimum zones (OMZs) areregions of the ocean where oxygen (O2) is naturally depleted (<20 µM). In OMZs microbial communi-ties use alternative terminal electron acceptors such as nitrate, sulfate and carbon dioxide, resulting in fixednitrogen loss and production of greenhouse gases including methane (CH4). In this thesis, I explored mi-crobial community structure, dynamics and metabolic interactions as they relate to CH4 cycling in SaanichInlet, a seasonally anoxic fjord on the coast of British Columbia Canada that serves as a model ecosys-tem for studying microbial processes in OMZs. Leveraging decadal time series observations in SaanichInlet, I developed a geochemical dataset consisting of nutrient and gas measurements, coupled with multi-omic (DNA, RNA and protein) sequence information to chart microbial community structure and dynamicsalong defined redox gradients. I conducted methods optimization comparing in situ and on-ship samplingparadigms and used correlation analysis to infer putative microbial interaction networks in relation to wa-ter column CH4 oxidation. Methanotrophic bacteria in Saanich Inlet were identified associated with threeuncultivated Gammaproteobacteria clades termed OPU1, OPU3 and symbiont-related that partitioned in thewater column during periods of prolonged stratification. Water column distribution of the OPU3 clade wasfound to correlate with nitrite (NO2-). Based on these results, I conducted incubations with labelled CH4 andNO2- to test this correlation and constrain potential metabolic interactions between methanotrophs and otherone-carbon utilizing microorganisms under low O2 conditions. Using multi-omic information derived fromthese incubations I confirmed the role of OPU3 in coupling CH4 oxidation to NO2- reduction and uncoveredpotential metabolic interactions between OPU3 and other co-occurring microorganisms including Methy-lophilales, Planctomycetes and Bacteroidetes. Evidence for a communal function in CH4 oxidation expandsthe role of OPU3 in the global carbon budget and provides a conceptual foundation for the development ofnumerical models to predict CH4 flux from OMZs as they expand throughout the global ocean.iiiLay SummaryIn nature, microbes do not grow in isolation, they form communities of interacting cells, resulting in dis-tributed networks that drive nutrient and energy cycling. Oxygen plays a crucial role in the development ofthese networks. Marine oxygen minimum zones (OMZs) are currently expanding and intensifying due toclimate change. In OMZs, the use of alternative electron acceptors by microbes results in the production ofgreenhouse gases such as methane (CH4). In this thesis, I explored the structure and function of microbialcommunities in relation to CH4 cycling processes in a model ecosystem, Saanich Inlet, to study microbialprocesses in coastal and open ocean OMZs. Coupling multi-molecular (DNA, RNA and protein) sequenceinformation and geochemical parameter information, I identified microbial networks driving CH4 consump-tion under low oxygen conditions. Understanding how this greenhouse gas is cycled in expanding OMZswill help us to better predict climate change effects at a global scale.ivPrefaceThe work presented in this thesis was made possible by support of collaborators, former and current memberof the Hallam laboratory, and contractors as described in the following paragraphs.Chapter 2: The Saanich Inlet geochemical datasets are the results of the tireless effort of many post-doctoral fellows, students, technicians, and volunteers who think big about the microbial world. In chapter 2,I presented a compendium of time-series observations encompassing historical oxygen (O2) measurementsand more recent monthly (2006-2014) monitoring efforts, representing over 100 independent sampling ex-peditions. This chemical compendium partners with a compendium of multi-omic sequence informationfrom the Saanich Inlet water column detailing time-series microbial multi-omic datasets. Combined, thesecompendiums provide a community-driven framework for observing and predicting microbial communityresponses to changing levels of O2 deficiency extensible to open ocean OMZs.Dr. Steven Hallam designed and supervised the Saanich Inlet time-series project. Environmental chem-ical and multi-omic data collection was carried out aboard the RSV Strickland. Sample collection wasconducted with the valuable help and support of Captain Ken Brown and his crew. Sample collection for thegeneration of chemical, physical and multi-omic datasets was possible thanks to the extensive logistical sup-port and planning of several Chief Scientists. David Walsh, Elena Zaikova, Olena Shevchuk, Craig Mewis,Alyse K. Hawley, and I held the Chief Scientist position. Sea going technicians Chris Payne and LarysaPakhomova were keystones for sample collection, and CTD operation and calibration. Alyse K. Hawleyand David Capelle carried out dissolved gas measurements and quality control under the supervision of Dr.Philippe Tortell.Multi-omic datasets were the result of many hours of bench work carried out by many collaborators,technicians, students, and volunteers. Alyse K. Hawley, Melanie Scofield, Sam Kheirandish, AndreasMueller, Payal Sipahimalani, Olena Shevchuk, and I generated metagenomic and small subunit tag datasets.Sequencing was carried out at the Joint Genome Institute (JGI) and Ge´nome Que´bec Innovation Centre atthe McGill University. Metatranscriptomic datasets were generated following the protocol designed by Al-yse K. Hawley. Alyse K. Hawley and I carried out RNA extractions. Sequencing was carried out at the JGI.Alyse K. Hawley and Heather Brewer at the Environmental Molecular Science Laboratory (EMSL) and thePacific Northwest National Laboratory (PNNL) conducted Metaproteomic data including extractions andprotocol design.Portions of text, protocols and figures in chapter 2 were published in Nature Scientific Data as: Torres-Beltra´n, M., Hawley, A.K. et al. 2017. A compendium of geochemical information from the Saanich Inletwater column. Sci.Data. 4. doi:10.1038/sdata.2017.159 Additional manuscript fully detailing multi-omicvsequence data protocols was published in Nature Scientific Data as: Hawley, A.K., Torres-Beltra´n, M. et al.2017. A compendium of multi-omic sequence information from the Saanich Inlet water column. Sci.Data.4. doi:10.1038/sdata.2017.160.Chapter 3: I used small subunit ribosomal RNA (SSU rRNA) gene 454 sequencing data to compareand cross-calibrate in situ sampling devices such as McLane PPS with conventional bottle sampling methodsby testing bottle effects on microbial community composition, and potential activity when using differentfilter combinations and filtration methods. I used SSU rDNA and rRNA 454 pyrosequencing data generatedduring the SCOR Working Group “Microbial Community Responses to Ocean Deoxygenation” workshopheld in Vacnouver, B.C in July 2014. Collection of samples was carried out aboard the RSV Stricklandoperated by Captain Ken Brown. Chris Payne and Larysa Pakhomova conducted CTD deployment anddata processing. Dr. Virginia Edgcomb, Dr. Maria Pachiadaki, Andreas Mueller, Melanie Scofield, and Icollected and processed DNA and RNA samples on ship. Dr. Craig Taylor supervised sample collectionfrom in situ filtering systems. Andreas Mueller, Melanie Scofield and I extracted DNA and RNA samples.Kateryna Tyshchenko and I constructed cDNA libraries and performed PCR reactions of samples to besubmitted for 454 sequencing at Ge´nome Que´bec Innovation Centre at the McGill University. I conductedthe downstream analysis of sequences using QIIME, and all statistical analyses using R scripts I wrote. Igenerated all figures, and wrote the chapter under the supervision of Dr. Steven Hallam.A version of thischapter will be submitted to a peer reviewed journal.Chapter 4: I used geochemical information and SSU rRNA gene 454 sequencing data to address micro-bial eukaryote community dynamics over a 12-month period, uncovering significant correlations betweenparasitic dinoflagellates within Syndiniales and other eukaryotic taxa during months of peak water columnstratification. I used methods in this chapter to gain insight into the use of analytical and statistical tools thatlater I used to asses the methanotrophic community composition, dynamics, and metabolism as detailed inchapters 5 and 6. I used 454 sequencing data of the SSU rRNA gene generated from 159 DNA samples col-lected in the Saanich Inlet between May 2008 and April 2009. Collection of samples was carried out aboardthe RSV Strickland operated by Captain Ken Brown. Chris Payne and Larysa Pakhomova conducted CTDdeployment and data processing. Past members of the Hallam laboratory extracted DNA samples until April2009. I extracted DNA and RNA samples from July 2014. With assistance of Melanie Scofield and AndreasMueller, and Kateryna Tyshchenko, I performed PCR reactions of samples to be submitted for 454 sequenc-ing at Ge´nome Que´bec Innovation Centre at the McGill University. I conducted the downstream analysisof sequences using the QIIME software, and correlation analyses using CoNet. Taylor Sehein from theEdgcomb laboratory at Woods Hole Oceanographic Institution and Dr. Maria Pachiadaki from the BigelowLaboratory contributed greatly on data interpretation key for the chapter development. Taylor Sehein andI generated all figures. Dr. Virginia Edgcomb from Woods Hole Oceanographic Institution closely super-vised and provided constructive feedback on data analysis and manuscript production. Dr. Steven Hallamprovided constructive feedback on chapter and manuscript production and presentation.Text, protocols, and figures presented in this chapter were published as: Torres-Beltra´n, M., Sehein, T.et al. 2018. Protistan parasites along oxygen gradients in a seasonally anoxic fjord: a network approach toassessing potential host-parasite interactions. Deep Sea Res. II. doi.org/10.1016/j.dsr2.2017.12.026viChapter 5: I used SSU rRNA gene 454 sequencing data generated from 288 DNA samples collectedin the Saanich Inlet between May 2008 and July 2010. I surveyed the microbial community along seasonalredox gradients and focused on identifying microbial agents driving methane (CH4) oxidation. Collection ofsamples was carried out aboard the RSV Strickland operated by Captain Ken Brown. Chris Payne and LarysaPakhomova conducted CTD deployment and data processing. Past members of the Hallam laboratory ex-tracted DNA samples collected until April 2009. I extracted DNA samples from May 2009 to July 2010.With assistance of Melanie Scofield and Andreas Mueller, I performed PCR reactions of samples to be sub-mitted for 454 sequencing at Ge´nome Que´bec Innovation Centre at the McGill University. I conducted thedownstream analysis of sequences using the QIIME software, and conducted further phylogenetic assign-ment of methanotrophic taxa. Claire Stilwell constructed the particulate methane monooxygenase subunitβ (pmoA) gene libraries. Alyse K. Hawley extracted the DNA and RNA samples to generate the metage-nomic and metatranscriptomic datasets to be sequenced at JGI on the Illumna MiSeq platform. Maya Bhatiaprocessed the multi-omics datasets through MetaPathways, designed and built by Niels Hanson and KishoriKonwar. David Capelle carried out dissolved CH4 measurements and quality control. Evan Durno and Iconducted the correlation analyses using CoNet and R scripts (written by Evan Durno). For the publishedarticle I generated all figures and wrote the manuscript under the supervision of Dr. Steven Hallam.Text, protocols, and figures presented in chapter 5 were published as: Torres-Beltra´n, M. et al. 2016.Methanotrophic community dynamics in a seasonally anoxic fjord: Saanich Inlet, British Columbia. Front.Mar.Sci. doi.org/10.3389/fmars.2016.00268.Chapter 6: I used multi-omic sequence information (DNA, RNA and protein) derived from CH4 incu-bation experiments on Saanich Inlet O2 deficient waters to identify community-level interactions related toCH4 oxidation. David Capelle and I collected DNA, RNA and protein samples aboard the RSV Stricklandoperated by Captain Ken Brown. Chris Payne and Larysa Pakhomova conducted CTD deployment and dataprocessing. I extracted DNA samples and performed PCR reactions of samples to be submitted for 454sequencing at Ge´nome Que´bec Innovation Centre at the McGill University. I extracted RNA samples atEMSL under the supervision of Lye Meng Markillie. Lye Meng Markillie and Dr. Hugh Mitchell conductedthe RNA sequencing on the Ion Torrent platform and the initial quality control and processing of sequencesat EMSL. Kateryna Tyshchenko and I constructed cDNA libraries on remaining RNA and performed PCRreactions of samples to be submitted for 454 sequencing at Ge´nome Que´bec Innovation Centre at the McGillUniversity. I extracted protein samples at EMSL with the help and supervision of Heather Brewer. Gen-eration of peptide spectra was done by Heather Brewer, and spectra mapped to protein database by AngelaNorbeck and Samuel Purvine at EMSL. Dr. John Cliff and I conducted nanoSIMS quantification, andDr. Jim Moran conducted bulk isotopic incorporation. Connor Morgan-Lang assembled the metagenomicdatasets. I conducted the downstream analysis of sequences using the QIIME software and metagenomicanalysis using MetaPathways, designed by Niels Hanson and Kishori Konwar. I conducted statistical anal-yses using R scripts that I wrote. I generated all figures, and wrote the chapter under the supervision of Dr.Steven Hallam. A version of this chapter will be submitted to a peer reviewed journal.viiTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiLay Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiiList of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xviiiDedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Global ocean deoxygenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1.1 Marine oxygen minimum zones (OMZs) . . . . . . . . . . . . . . . . . . . . . . . 11.2 Methane in the ocean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Methanotrophs in OMZs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.3.1 Methanotrophy under water column O2 deficiency . . . . . . . . . . . . . . . . . . 51.3.2 Methanotrophic interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.3.3 Saanich Inlet is a model ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . 61.3.4 Thesis objectives and overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 A compendium of geochemical information from the Saanich Inlet water column . . . . . . . 102.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112.2.1 Environmental sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112.2.2 Chemical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12viii2.2.3 Data Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.3 Technical Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 Data Citation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.5 Conclusion and application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Sampling and processing methods impact microbial community structure and function . . . 183.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203.2.1 Environmental sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203.2.2 Workshop microbial biomass collection . . . . . . . . . . . . . . . . . . . . . . . . 203.2.3 Time-series microbial biomass collection . . . . . . . . . . . . . . . . . . . . . . . 223.2.4 Nucleic acid extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223.2.5 Small subunit ribosomal RNA sequencing . . . . . . . . . . . . . . . . . . . . . . . 233.2.6 Statistical analysis and data visualization . . . . . . . . . . . . . . . . . . . . . . . 233.2.7 Data deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.3.1 Water column conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.3.2 Benchmarking workshop and Saanich Inlet time-series results . . . . . . . . . . . . 253.3.3 Size-fractionation effects on community structure . . . . . . . . . . . . . . . . . . . 253.3.4 Size fractionation effects on indicator OTUs (DNA analyses) . . . . . . . . . . . . . 283.3.5 Size-fractionation effects on expressed OTUs within specific populations (rRNAanalyses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.4.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 Protistan parasites along water column oxygen gradients: a network approach to assessingpotential host-parasite interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.1 Environmental sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.2 Nucleic acid sampling and extraction . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.3 Small subunit ribosomal RNA and RNA gene sequencing and analysis . . . . . . . . 414.2.4 Statistical analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.2.5 Data deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.3.1 Water column conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.3.2 Eukaryotic community structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454.3.3 Exploring protistan co-occurrence patterns . . . . . . . . . . . . . . . . . . . . . . 454.3.4 Insight into potential Syndiniales parasitic interactions . . . . . . . . . . . . . . . . 484.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53ix4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 Methanotrophic community dynamics in Saanich Inlet . . . . . . . . . . . . . . . . . . . . . 595.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.2.1 Environmental sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.2.2 Nucleic acid sampling and extraction . . . . . . . . . . . . . . . . . . . . . . . . . 615.2.3 Small subunit ribosomal RNA gene sequencing and analysis . . . . . . . . . . . . . 625.2.4 pmoA gene libraries, metagenomic and metatranscriptomic sequencing and analysis 635.2.5 Phylogenetic inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.2.6 Statistical analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655.2.7 Data deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675.3.1 Water column conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675.3.2 Microbial community structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685.3.3 Methanotroph diversity and dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 685.3.4 PmoA diversity and expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735.3.5 Methanotroph niche partitioning and co-ocurrence patterns . . . . . . . . . . . . . . 735.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816 Community-level interactions support CH4 oxidation in Saanich Inlet O2-deficient watercolumn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846.2.1 Incubations implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846.2.2 Carbon incorporation measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 856.2.3 Nucleic acid and protein extractions . . . . . . . . . . . . . . . . . . . . . . . . . . 866.2.4 Nucleic acid tag sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876.2.5 Meta-genomic, transcriptomic and proteomic sequencing . . . . . . . . . . . . . . . 876.2.6 Tag sequence analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886.2.7 Multi-omic dataset analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896.2.8 Pathways network and metabolic model inference . . . . . . . . . . . . . . . . . . . 906.2.9 Data deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916.3.1 Water column conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916.3.2 Carbon incorporation into biomass . . . . . . . . . . . . . . . . . . . . . . . . . . 926.3.3 Microbial community composition . . . . . . . . . . . . . . . . . . . . . . . . . . 926.3.4 Insight into microbial community response to substrate additions . . . . . . . . . . . 966.3.5 PmoA and NirK diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98x6.4 Insight into microbial community response to substrates addition . . . . . . . . . . . . . . . 986.4.1 Elucidating microbial metabolic networks . . . . . . . . . . . . . . . . . . . . . . . 1026.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1067 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087.1 Standards of practice for sample collection and data analysis . . . . . . . . . . . . . . . . . 1087.2 Using co-occurrence network analysis to chart ecological interactions . . . . . . . . . . . . 1097.3 Integrative analysis of coupled biogeochemical processes . . . . . . . . . . . . . . . . . . . 1117.3.1 Incubation experiment limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127.4 The significance of this research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114A Sampling and processing methods impact microbial community structure and function . . . 141A.1 Supplementary results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141A.1.1 Saanich Inlet microbial community monitoring time-series . . . . . . . . . . . . . . 141A.1.2 Filtering conditions effect on microbial community diversity . . . . . . . . . . . . . 141B Protistan parasites along water column oxygen gradients: a network approach to assessingpotential host-parasite interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145B.1 Supplementary results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145B.1.1 Co-occurrence analysis and network . . . . . . . . . . . . . . . . . . . . . . . . . . 145C Methanotrophic community dynamics in Saanich Inlet . . . . . . . . . . . . . . . . . . . . . 153C.1 Supplementary methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153C.1.1 Network description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153C.1.2 Methanotroph sub-networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154C.1.3 Methanotroph interactions in sub-networks . . . . . . . . . . . . . . . . . . . . . . 154C.1.4 Statistical analyses considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . 154D Community-level interactions support methane oxidation in Saanicn Inlet oxygen-deficientwater column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165D.1 Supplementary Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165D.1.1 Co-occurrence metabolic network properties . . . . . . . . . . . . . . . . . . . . . 165E OPU3 genes distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172xiList of TablesTable 3.1 Biomass collection scheme for DNA and RNA in situ and on-ship samples . . . . . . . . 20Table 3.2 Microbial taxa with significant abundance differences among tested filtering conditions . 28Table 5.1 Multivariate regression statistics for methanotroph OTUs . . . . . . . . . . . . . . . . . 76Table 5.2 Negative binomial regression statistics for methanotroph OTUs . . . . . . . . . . . . . . 76Table 6.1 Key protein ratios across meta-transcriptomic and proteomic datasets . . . . . . . . . . . 100Table 6.2 Total number of transcripts and proteins affiliated with metabolically active taxa . . . . . 101Table A.1 Water column chemical properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142Table B.1 Indicator OTUs for summer stratification. . . . . . . . . . . . . . . . . . . . . . . . . . . 148Table B.1 Indicator OTUs for summer stratification (Continuation) . . . . . . . . . . . . . . . . . . 149Table B.3 List of Spring indicator OTUs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150Table B.4 List of Summer stratification indicator OTUs . . . . . . . . . . . . . . . . . . . . . . . . 151Table B.5 List of early Fall renewal indicator OTUs . . . . . . . . . . . . . . . . . . . . . . . . . . 152Table C.1 List of indicator OTUs for the oxic water column condition. . . . . . . . . . . . . . . . . 155Table C.2 List of indicator OTUs for the dysoxic-suboxic, and combined oxic-dysoxic-suboxic wa-ter column conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157Table C.3 List of indicator OTUs for the anoxic, and combined dysoxic-suboxic and anoxic watercolumn conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158Table C.4 Complete list of interactions between methanotroph and bacterial OTUs. . . . . . . . . . 159Table C.5 BLAST-based comparison for methanotroph sub-networks OTUs. . . . . . . . . . . . . . 160Table D.1 BLAST-based comparison for incubation subnetworks OTUs. . . . . . . . . . . . . . . . 170Table D.2 List of correlating pathways under dysoxic-suboxic water column conditions. . . . . . . . 171xiiList of FiguresFigure 1.1 Global OMZs distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Figure 1.2 Methane cycling processes in the ocean. . . . . . . . . . . . . . . . . . . . . . . . . . . 4Figure 1.3 Microbial co-occurrence networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Figure 1.4 Saanich Inlet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Figure 2.1 Geochemical data time series in the Saanich Inlet . . . . . . . . . . . . . . . . . . . . . 12Figure 2.2 Time series environmental parameters water column profiles. . . . . . . . . . . . . . . . 13Figure 2.3 Validation for environmental parameters. . . . . . . . . . . . . . . . . . . . . . . . . . 15Figure 3.1 SCOR workshop sampling scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Figure 3.2 Cluster analysis of Saanich Inlet time-series and SCOR workshop tag data. . . . . . . . 26Figure 3.3 Microbial community partitioning based filtration methods . . . . . . . . . . . . . . . . 27Figure 3.4 Microbial community abundance shifts in relation to filtration methods . . . . . . . . . 29Figure 3.5 Indicator OTUs associated with specific filtration methods . . . . . . . . . . . . . . . . 30Figure 3.6 Activity differences for abundant OTUs in relation to filtration methods . . . . . . . . . 32Figure 3.7 Activity differences for indicator OTUs in relation to filtration methods . . . . . . . . . 33Figure 4.1 Water column chemical parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Figure 4.2 Taxonomic breakdown of eukaryotic OTUs . . . . . . . . . . . . . . . . . . . . . . . . 46Figure 4.3 Eukaryotic community structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Figure 4.4 Co-occurrence network on SSU rDNA pyrotag protist data from May-August 2008. . . . 48Figure 4.5 Syndiniales OTUs interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Figure 4.6 Vertical distribution and abundance of Choreotrichia and Syndiniales OTUs. . . . . . . . 50Figure 4.7 Vertical distribution and abundance of Phaeocystis antarctica and Syndiniales OTUs. . . 51Figure 4.8 Vertical distribution and abundance of Maxillopoda and Syndiniales OTUs. . . . . . . . 52Figure 4.9 Distribution of Syndiniales and interacting Choreotrichia, Phaeocystis sp., and Maxil-lopoda OTUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Figure 5.1 Microbial community partitioning under changing levels of water column O2 deficiency 69Figure 5.2 Taxonomic composition of OTUs identified in SSU rRNA gene pyrotags between 2008-2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70xiiiFigure 5.3 Relative abundance and phylogenetic relationships between Type I and Type II methan-otroph OTUs in Saanich Inlet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Figure 5.4 Time-series observations for methanotrophic OTUs affiliated with OPU1, OPU3, andmethanotrophic symbiont groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Figure 5.5 Particulate methane monooxygenase subunit β (pmoA)phylogenetic tree . . . . . . . . . 75Figure 5.6 Co-occurrence patterns for methanotrophic and indicator OTUs from SSU rRNA genepyrotag datasets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Figure 6.1 Experimental workflow for CH4 incubation experiments carried out in February 2015 . . 85Figure 6.2 Labeled substrate incorporation into cellular biomass. . . . . . . . . . . . . . . . . . . . 92Figure 6.3 Microbial community partitioning based on depth and incubation treatments. . . . . . . 93Figure 6.4 Taxonomic composition of OTUs identified in SSU rDNA pyrotags among treatments. . 94Figure 6.5 Taxonomic composition of OTUs identified as potentially active taxa under incubationtreatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95Figure 6.6 Co-occurrence patterns for methanotrophic OTUs from SSU pyrotag datasets. . . . . . . 97Figure 6.7 Taxonomic and functional breakdown of transcripts related to key proteins for methaneand nitrogen cycling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99Figure 6.8 Taxonomic and functional breakdown of key proteins related to methane and nitrogencycling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103Figure 6.9 Co-occurrence patterns for dysoxic-suboxic metabolic pathways in the Saanich Inletwater column. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104Figure A.1 Non-metric multidimensional scaling plot for Saanich Inlet Time-Series SSU rDNApyrotag sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142Figure A.2 Shannon and alpha (α) diversity indexes for SSU rDNA pyrotags SCOR samples. . . . . 143Figure A.3 Active microbial community composition . . . . . . . . . . . . . . . . . . . . . . . . . 144Figure B.1 Vertical distribution overtime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147Figure C.1 Time-series chemical data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Figure C.2 Partitioning of methanotrophic OPU1, OPU3 and methanotrophic symbiont OTUs. . . . 161Figure C.3 Methane contour plot for gas concentration . . . . . . . . . . . . . . . . . . . . . . . . 162Figure C.4 Cumulative distribution of methanotroph OTUs throughout Saanich Inlet water column. 163Figure C.5 Distribution and abundance of methanotrophic bacteria in OMZs. . . . . . . . . . . . . 164Figure D.1 Time-series chemical data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166Figure D.2 Particulate methane monooxygenase subunit β (pmoA) phylogenetic tree. . . . . . . . . 167Figure D.3 Copper-containing nitrite reductase (nirK) phylogenetic tree. . . . . . . . . . . . . . . . 168Figure D.4 Difference in transcript abundance between 12C and 13C incubation treatments. . . . . . 169Figure D.5 Co-occurrence network of correlating pathways at dysoxic-suboxic water column con-ditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169xivFigure E.1 Global distribution of OPU3 pmoA and nirK genes. . . . . . . . . . . . . . . . . . . . . 173Figure E.2 Spatio-temporal distribution of OPU3 pmoA and nirK genes throughout Saanich Inletwater column. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174xvList of AbbreviationsAOM Anaerobic oxidation of methaneBLAST Basic local alignment search toolCTD Conductivity, temperature and depth measuring instrumentCO2 Carbon dioxideCH4 MethaneCH3OH MethanolCH2O FormaldehydeCHOO- FormateDNA Deoxyribonucleic acidcDNA Complementary DNArDNA Ribosomal DNAEMSL Environmental Molecular Sciences LaboratoryETNP Eastern Tropical North PacificETSP Eastern Tropical South PacificH2S Hydrogen sulfideH4SiO4 Silicic acidISA Indicator species analysisJGI Joint Genome InstituteMgCl2 Magnesium chlorideMALV Marine alveolateMMO Methane monooxygenaseMPP Masterflex peristaltic pumpN2 Dinitrogenxvin-damo Nitrite dependent anaerobic methane oxidationNH4+ AmmoniumNO3- NitrateNO2- NitriteN2O Nitrous oxideO2 OxygenOMZs Oxygen minimum zonesOPU Particulate methane monooxygenase encoding phylogenetic groupOTU Operational taxonomic unitPAR Photosynthetically active radiationPCR Polymerase chain reactionPNNL Pacific Northwest National LaboratoryPMO Particulate methane monooxygenasePO43- OrthophosphatePPS Phytoplankton samplerQIIME Quantitative Insights Into Microbial EcologyRNA Ribonucleic acidrRNA Ribosomal RNARuMP Ribulose monophosphateSCOR Scientific Committee on Oceanographic ResearchsMMO Soluble methane monooxygenaseSSU Ribosomal small subunitTEA Terminal electron acceptorxviiAcknowledgmentsFirst of all I would like to thank my supervisor Dr. Steven Hallam for his support and for always pushing meto do better. Thank you Dr. Hallam for your patience and generosity on life matters. I would like to thank mycommittee members, Dr. Thomas J. Beatty, Dr. William W. Mohn and Dr. Martin Hirst, for all their supportand input. I would like to thank past and present members of the Hallam lab for all the smiles. Thank youAlyse K. Hawley for your friendship and your continuous support both emotional and logistical. Thank youDavid Capelle,Maya Bhatia, Esther Gies, Elena Zaikova, Evan Durno, Aria Hahn, Ashley Arnold, KaterynaLevdokymenko for all your support and company over these years. Thank you Andreas Mueller and MelanieScofield for being the best team in the world and remind me everything is awesome!. I would like to thankChris Payne and Larysa Pakhomova for their help and support make possible our Saanich cruise. Thank youCaptain Ken Brown and his crew for their engaged effort on every cruise aboard the RSV Strickland. Thankyou to members of the Tortell, Crowe and Suttle Laboratories at UBC and members of the Varela Lab atUVic for logistical support.I would like to thank to my collaborators at the Joint Genome Institute for sequencing so many samples,and at Environmental Sciences Laboratory at Pacific Northwest National Labs for your support in trasncrip-tomics, proteomics, and stable isotope analysis none of this work would have been possible without you.Thank you to Consejo Nacional de Ciencia y Tecnologı´a (CONACyT) and the Tula Foundation for thefunding that made possible my degree.Muchas gracias mamita linda querida, gracias por todo tu esfuerzo y apoyo durante toda mi vida, graciaspor ayudarme a siempre levantarme y seguir adelante. Este trabajo tambie´n es tuyo. Te amo.Gracias a mis suegros por su incondicional apoyo en estos an˜os.Mi querida Citlali Ma´rquez, infinitas gracias amiga mı´a por todo tu apoyo, por tu compan˜ia y por laslargas platicas que hicieron ma´s fa´cil y llevaderos los an˜os del doctorado. Gracias por hacerme sentir enfamilia. Gracias mi querida Natalie Milla´n por tu compan˜ia sin importar la distancia, te quiero mucho amigamia!.Gracias a mi querida familia, Ricardo Mendoza y Marta Reta, por estar presentes en los momentos ma´sdifı´ciles.Mi amado Luis Malpica Cruz, mi Boo, mi papoy, sin ti esto simplemente no se hubiera logrado, literal-mente. Gracias por haber sido mis manos en muchos instantes, gracias por tu paciencia y apoyo en estosan˜os, gracias por regalarme perspectiva en momentos difı´ciles, gracias por darme la mano. Esta tesis, es tanmı´a como lo es tuya. Gracias por vivir esta etapa a mi lado y por haberla sobrevivido. Te amo con todo micorazo´n. Eres mi mejor amigo y el amor de mi vida.xviiiDedicationPara Luis y MarinaEn memoria de mi abueloxixChapter 1Introduction1.1 Global ocean deoxygenationThe ocean is changing. In addition to the tons of plastic polluting the ocean and the continued decline of fishstocks worldwide, unseen beneath the waves along with increasing acidification, another problem arises:the loss of oxygen (O2) from ocean waters. This O2 loss, or deoxygenation, is one of the most importantchanges occurring in an ocean increasingly modified by human activities (Breitburg et al., 2018).Greenhouse gas-driven global warming is likely the ultimate cause of this ongoing deoxygenation inmany parts of the open ocean (Bopp et al., 2013). Decreased O2 solubility due to a warming ocean isestimated to account for ∼15% of the total current global O2 loss and >50% of the O2 loss in the upper1000 m of the ocean (Helm et al., 2011; Schmidtko et al., 2017; Breitburg et al., 2018). Higher temperaturesalso raise metabolic rates, thus accelerating O2 consumption (Breitburg et al., 2018) and resulting in a spatialredistribution of available O2 (Brewer et al., 2017). For instance, as O2 levels decline, aerobic organismsshelter in more oxygenated waters, resulting in habitat compression accompanied by the diversion of energyinto microbial metabolism (Diaz and Rosenberg, 2008; Wright et al., 2012). Intensified stratification mayaccount for the remaining 85% of global ocean O2 loss by reducing O2 transport into the ocean interior andaffecting the nutrient supply controlling the production of organic matter and its subsequent sinking fromthe ocean surface (Breitburg et al., 2018). Sinking organic particles, or marine snow, provide a nucleationpoint for otherwise suboxic or anoxic processes adding to energy diversion towards microbial metabolismin oxygenated waters (Wright et al., 2012).As of 1960, global dissolved O2 concentration observations show ongoing regional changes in oceanicO2 (Whitney et al., 2007; Bograd et al., 2008; Stramma et al., 2008; Keeling et al., 2010; Stramma et al.,2010; Helm et al., 2011; Schmidtko et al., 2017), with an estimated 2% O2 loss (77 billion metric tons) overthe past 50 years (Schmidtko et al., 2017). As a consequence, open-ocean O2 minimum zones (OMZs) haveexpanded in area (4.5 million Km2, based on water with O2 concentrations <70 µM at 200 m depth) (Boppet al., 2002; Keeling et al., 2010; Keller et al., 2016; Schmidtko et al., 2017), and the volume with watercompletely devoid of O2 (anoxic) has more than quadrupled over the same period (Schmidtko et al., 2017).1.1.1 Marine oxygen minimum zones (OMZs)Marine oxygen minimum zones (OMZs) are widespread, naturally occurring water column features thatarise from the respiration of organic matter in subsurface waters with restricted circulation. Operationallydefined by O2 concentrations between 0 to 20 µM, the differential accumulation of nitrite (NO2-), andreduced sulphur compounds, OMZs currently constitute 1-7% of the global ocean volume (Fuenzalida et1O2  (μmol per kg water)Longitude (oE) Latitude (oN)12346578910806040200-20-40-60-80-150 -100 -50 0 50 100 150>200<314070351112131415 16Figure 1.1: Global OMZ distribution. The global distribution of OMZs includes: (1) the NortheasternSubarctic Pacific Ocean, (2) the Saanich Inlet, (3) the Hawaii Ocean Time-series, (4) the Guaymas Basin,(5) the Eastern Tropical North Pacific, (6) Costa Rica, (7) the Gulf of Mexico, (8) the Cariaco Basin, (9)Peruvian, (10) Eastern Tropical South Pacific , (11) Chilean, (12) the Baltic and (13) Black Seas, (14) theNamibian upwelling system, (15) Arabian Sea and (16) Bay of Bengal. Contour plot depicts the minimumoxygen (O2) concentrations (µmol per kg water) for different ocean regions. Figure was modified fromWright et al., 2012.al., 2009; Paulmier and Ruiz-Pino, 2009; Ulloa and Pantoja, 2009; Codispoti, 2010; Lam and Kuypers,2011; Ulloa et al., 2012; Wright et al., 2012). (Figure 1.1).As O2 levels decline, nutrients and energy are increasingly diverted away from higher trophic levelsinto microbial community metabolism (Diaz and Rosenberg, 2008; Wright et al., 2012), increasing nutrientand energy cycling through the use of alternative terminal electron acceptors (TEAs) (Diaz and Rosenberg,2008). As a result, OMZs are hotspots for the biogeochemical cycling of carbon, nitrogen and sulphurwith resulting feedback on nitrogen loss processes and greenhouse gases production including nitrous oxide(N2O) and methane (CH4) (Lam et al., 2009; Ward et al., 2009; Canfield et al., 2010; Lam and Kuypers,2011; Naqvi et al., 2010), which influence global warming (Lam et al., 2009; Ward et al., 2009). Both ofthese gases contribute to warming by increasing the amount of solar energy that is absorbed by the planetmeasured as radiative forcing in Watts per square meter. However, the global warming potential (GWP)of N2O and CH4 varies substantially from the most common greenhouse gas, carbon dioxide (CO2), by300- and 30-fold, respectively (based on one-hundred year atmospheric residence times) (IPCC, 2013).Current research efforts are defining the interaction networks underlying microbial metabolism in OMZsand generating new insights into coupled biogeochemical processes in the ocean driving the cycling ofclimate-active trace gases (Canfield et al., 2010; Hawlet et al., 2014; Louca et al., 2016; Tsementzi et al.,2016; Hawley et al., 2017a).1.2 Methane in the oceanMethane is the most abundant hydrocarbon in the atmosphere and the second most important climate ac-tive trace gas after CO2 (Solomon et al., 2007). In the ocean, CH4 is primarily produced in anoxic marine2sediments by methanogenic archaea (Kiene, 1991). However, the majority of the CH4 produced in marinesediments is oxidized by anaerobic methanotrophic (ANME) archaea. Microbial anaerobic oxidation ofCH4 (AOM) plays a role in CH4 regulation and reduces methane release from marine systems (Knittel andBoetius, 2009). Anaerobic oxidation of CH4 is well known to be coupled with sulfate reduction (sulfate-dependent AOM) in marine ecosystems (Knittel and Boetius, 2009, Milucka et al., 2012, McGlynn et al.,2015, Wegener et al., 2015), and recently it was demonstrated that AOM may be coupled with the dissim-ilatory reduction of metals, including Fe(III) and Mn(IV) (metal-dependent AOM, metal-AOM) (Ettwig etal., 2016, Fu et al., 2016, Scheller et al., 2016). As with sediment sources, new CH4 production in the watercolumn has been described. For example, CH4 concentrations in oxic surface waters are 5-75% supersatu-rated with respect to the atmosphere resulting in a “CH4 paradox” (Kiene, 1991; Karl et al., 2008). Oceanparticles, i.e. marine snow and fecal pellets, contain anoxic microenvironments and have been considereda potential mechanism for water column CH4 production and transport (Karl and Tilbrook, 1994). Further-more, recent molecular studies support the hypothesis that surface ocean CH4 is in part derived from thebreakdown of methylphosphonates produced by ammonia-oxidizing archaea (Metcalf et al., 2012). Pelagicaerobic or anaerobic methanotrophic bacteria may still oxidize any CH4 leaking into the water column. Forinstance, aerobic CH4 oxidation has been estimated to consume>50% of CH4 in the water column (Fung etal., 1991; Reeburgh et al., 1991) forming a final barrier and limiting its escape to the atmosphere (Reeburghet al., 1991; Blumenberg et al., 2007; Kessler et al., 2011; Heintz et al., 2012) (Fig. 1.2). Pelagic CH4 oxi-dation in marine environments is rarely quantified, but along the margins of an OMZ, where CH4 intersectstraces of O2, it could be a significant process (Mau et al., 2013) that likely has the greatest influence on theCH4 budget before its emission to the atmosphere (Reeburgh, 2007).Marine OMZs encompass large reservoirs of CH4 (Zhang et al., 2011; Pack et al., 2015). For instance,the Eastern Tropical North Pacific (ETNP) OMZ is both the largest OMZ (Paulmier and Ruiz-Pino, 2009)and the largest reservoir of oceanic CH4 in the world (Sansone et al., 2001; Reeburgh, 2007; Naqvi et al.,2010), potentially releasing ∼1 Tg CH4 yr-1 (Naqvi et al., 2010). In OMZs, O2 is consumed faster than itis resupplied, resulting in a layer of hypoxic waters surrounding a functionally anoxic core (Thamdrup etal., 2012) where CH4 accumulates (Wright et al., 2012). Thickening OMZs will likely move these largeCH4 pools closer to the ocean surface (Stramma et al., 2008; Keeling et al., 2010; Helm et al., 2011).Characterizing CH4 -consuming microbial populations in OMZs is critical for understanding greenhousegas and nutrient budgets under conditions of global warming and ocean deoxygenation.1.3 Methanotrophs in OMZsIn the water column, aerobic methanotrophic bacteria or methanotrophs utilize CH4 as the primary source ofcarbon and energy (Hanson and Hanson, 1996; Park et al., 2002; Tol et al., 2003; Murrell, 2010). Methan-otrophs harbor genes encoding the enzyme CH4 monooxygenase (MMO), which oxidizes CH4 to methanol(CH3OH). Methanol is subsequently converted to formaldehyde (CH2O) and ultimately to CO2 (CH2O isoxidized to formate (CHOO-) and CHOO- is oxidized to CO2), generating reducing power for biosynthesis.Carbon is assimilated into biomass at the oxidation level of formaldehyde via the Ribulose monophosphate(RuMP) or Serine cycles (Murrell, 2010). There are two structurally and biochemically distinct forms ofMMO, particulate (pMMO) and soluble CH4 monooxygenase (sMMO). Differential expression of the twoMMOs is controlled by the concentration of copper in the growth medium, with sMMO only produced atlow copper concentrations (<1 M) (Takeda and Tanaka, 1980; Hanson and Hanson, 1996). The pMMOis universal to all methanotrophs, while only a small subset harbors both sMMO and pMMO (Heyer etal., 2002; Murrell, 2010; Semrau et al., 2010). Therefore, functional gene probes targeting pMMO i.e.the particulate CH4 monooxygenase subunit β (pmoA) gene has been used to classify and identify novelmethanotrophs in different environments (McDonald and Murrell, 1997; Dunfield et al., 2007; Ettwig et al.,2010; Swan et al., 2011).3<31030100>250O2(mM) >250 100 30 10 <3 Figure 1.2: Methane cycling processes in the ocean. Methane (CH4) is produced in deep sediments bymethanogenic archaea (brown diamonds). It diffuses upwards and is consumed by the anaerobic oxidation ofCH4 (AMO), i.e. sulfate-dependent AOM, where sulphate reducers (green rods) use the hydrogen generatedmaintaining conditions that allow CH4 oxidation to proceed. Anoxic microenvironments in ocean particles,i.e. marine snow and fecal pellets (gray circles) and the breakdown of methylphosphonates (yellow stars)produced by ammonia-oxidizing archaea (orange and light green rods), have been considered a mechanismsfor CH4 production and transport into the water column. Methane released to the water column is availablefor aerobic oxidation by methanotrophs (pink rods).Although OMZs contribute mainly to global ocean CH4 cycling, information is scarce regarding thecomposition and activity of the microbes associated with CH4 cycling in these O2 deficient environments.Previous surveys have quantified CH4 oxidation rates in oxic (>90 µmol O2 kg-1) and anoxic (<1µmol O2kg-1) OMZ waters as well as in anoxic sediments (Ward et al., 1989; Ward and Kilpatrick, 1990; 1993).Results have indicated that sediment AOM occurred at a much slower rate than aerobic CH4 oxidation inthe water column (0.72-1.42 nmol L-1h-1 and 0.4-2 nmol L-1h-1, respectively) providing evidence of theimportance of aerobic CH4 oxidation as a second biological filter after sediment AOM (Ward et al., 1989;Ward and Kilpatrick, 1990; 1993). Subsequent small subunit ribosomal RNA (SSU rRNA) gene surveysfrom diverse OMZs, such as the Eastern Tropical South Pacific, the Namibian Upwelling and the Black Sea(Woebken et al., 2007; Stevens and Ulloa, 2008; Glaubitz et al., 2010), showed that canonical methanotrophswithin the Alpha and Gammaproteobacteria, are present albeit rare members of the microbial community,i.e. they are rare biosphere constituents (<0.01% of the total community).4Parallel efforts to describe the distribution, abundance and potential metabolic activity of the functionalgene pmoA identified pMMO-encoding phylogenetic groups (OPUs) OPU1 to OPU4 affiliated with Methy-lococcales in OMZs waters (Hayashi et al., 2007). For instance, OPU1 and OPU3 are commonly recoveredin molecular gene surveys in the open ocean and coastal OMZs (Hayashi et al., 2007; Tavormina et al., 2010;Tavormina et al., 2013; Knief, 2015; Torres-Beltra´n et al., 2016; Padilla et al., 2017) exhibiting differentialabundance and distribution patterns associated with water column O2 concentrations, i.e. OPU3 was ob-served to be more abundant under low O2 concentrations (<20 µM) (Tavormina et al., 2013). In addition tocanonical methanotrophs, more recent studies have identified a number of non-canonical microbial groupswith the potential to mediate CH4 cycling in the O2-deficient water column including bacteria affiliated withVerrucomicrobia (Dunfield et al., 2007), SAR324 within the Deltaproteobacteria (Swan et al., 2011) and theNC10 candidate division (Ettwig et al., 2010) expanding the phylogenetic range of potential CH4-oxidizingphenotypes.1.3.1 Methanotrophy under water column O2 deficiencyNitrite dependent anaerobic CH4 oxidation in NC10Studies of freshwater habitats have further linked AOM to the reduction of oxidized nitrogen compounds,including nitrate (NO3-) and NO2- (Raghoebarsing et al., 2006; Haroon et al., 2013). Recently discoveredNO2--reducing bacteria of the NC10 phylum couple CH4 oxidation to dinitrogen (N2) production through aunique NO2--dependent anaerobic CH4 oxidation (n-damo) pathway (Raghoebarsing et al., 2006; Ettwig etal., 2009). Characterized in the bacterium Candidatus Methylomirabilis oxyfera from freshwater sediments(Ettwig et al., 2010), the n-damo pathway reduces NO2- to nitric oxide (NO), which is then putativelydismutated into N2 and O2 gas, with O2 serving as the oxidant for intra-aerobic methanotrophy.Oxygen minimum zones provide potential niches for NO2--dependent anaerobic CH4 oxidation (n-damo) carried out by diverse bacteria linking CH4 oxidation to pathways of nitrogen loss under anoxicconditions. A recent study confirmed that OMZs harbor transcriptionally active bacteria affiliated with thecandidate division NC10. Observations suggest a niche for NC10 in nitrogen and CH4 cycling in OMZs(Padilla et al., 2016).Alternative n-damo in canonical methanotrophsRecent taxonomic and functional screening studies provide insight into methanotrophic community structureand activity. These studies suggest that some of the classic aerobic methanotroph species play an importantrole in nitrogen loss in OMZs as they thrive in O2 deficient environments by directly using NO3- or NO2- asterminal oxidants in CH4 oxidation pathways (Costa et al., 2000; Modin et al., 2007; Stein and Klotz, 2011;Beck et al., 2013; Hernandez et al., 2015). Evidence for the use of NO3- and NO2- in type I methanotrophsaffiliated with Methylococcales, has been found in both culture-dependent and -independent studies of di-verse low O2 environments (Kalyuzhnaya et al., 2013; Chistoserdova, 2015; Kits et al., 2015a,b; Danilovaet al., 2016). Furthermore, environmental expression of pmoCAB for group OPU3 was first demonstratedin a metatranscriptome from the Guaymas Basin OMZ (Lesniewski et al., 2012), and recently alternativemodes of CH4 oxidation by OPU3 haven been observed in the Costa Rica OMZ (Padilla et al., 2017). In ad-dition, genes mediating dissimilatory NO3- and NO2- reduction were identified in the OPU3 binned genome,and were found to be transcribed in conjunction with key enzymes catalyzing formaldehyde assimilation,suggesting partial denitrification linked to CH4 oxidation (Padilla et al., 2017).51.3.2 Methanotrophic interactionsIncubation studies enriched with CH4 revealed a marked increase of SSU rRNA gene sequences, whichsuggest cooperative metabolism between methanotrophic bacteria and potential microorganisms that utilizesingle-carbon compounds (Sauter et al., 2012). For instance, results derived from incubation experimentsusing sediment samples from Lake Washington showed a simultaneous response between Bacteroidetes,Methylophilales and canonical methanotrophs to CH4 addition over a range of O2 concentrations (15-75µM) (Beck et al., 2013; Hernandez et al., 2015). Although phylogenetic-based observations alone cannotexplain the underlying mechanisms of metabolite exchange, co-occurrence observations may shed light oncommunity-level interactions that support the metabolic requirements of methanotrophic agents in OMZs.Co-occurrence patterns and networksThe application of next-generation sequencing technologies revolutionized the field of microbial ecology(MacLean et al., 2009). It is now possible to study hundreds of samples of microbial communities simul-taneously and with great sequencing coverage (Hamady et al., 2008). This has allowed for co-occurrenceanalysis resulting in microbial association networks consisting of nodes and edges from which ecologicalrelationships between different microorganisms (Fig. 1.3), and between microorganisms and their environ-ment, can be inferred (Faust and Raes, 2012; Faust et al., 2015; Fuhrman et al., 2015). Co-occurrenceanalysis has been used to study the ecological interactions of microbes in lakes (Eiler et al., 2013; Peura etal., 2015), soils (Barbern et al., 2012), streams (Widder et al., 2014), the human microbiome (Faust et al.,2012), and the marine environment (Gilbert et al., 2011; Steele et al., 2011; Chow et al., 2014; Cram et al.,2015).Co-occurrence analysis has been increasingly explored with network inference techniques. A numberof methods are available to construct taxon co-occurrence networks from cross-sectional data (Faust andRaes, 2012), ranging from correlation combined with permutation tests (Barbern et al., 2012) and similarityassessments with hypergeometric distribution (Chaffron et al., 2010) to approaches dealing with compo-sitionality (Faust et al., 2012; Friedman and Alm, 2012), indirect edges (van den Bergh et al., 2012) andmultiple factors influencing taxon abundances (Faust et al., 2012; Trosvik et al., 2015). These networkinference techniques can be applied to construct dynamic models that are needed for a more comprehensiveunderstanding of the consequences of short- and long-term perturbations such as ocean deoxygenation onmicrobial metabolism, i.e. CH4 oxidation.As the activity and dynamics of CH4 oxidizing microbes in OMZs remain unconstrained, long-termmonitoring surveys that integrate methanotrophic activity and dynamics under changing levels of watercolumn O2 deficiency provide a promising environmental context to assess what constrains microbial con-trols of CH4 cycling. Understanding these constraints is an essential step in determining how biotic andabiotic factors merge to generate biological filters that reduce the flux of climate-active trace gases to theatmosphere.1.3.3 Saanich Inlet is a model ecosystemSaanich Inlet is a seasonally anoxic fjord on the coast of Vancouver Island, British Columbia, Canada(Carter, 1932; 1934; Herlinveaux, 1962; Anderson and Devol, 1973) (Fig. 1.4). Saanich Inlet water circu-lation is characteristic of an inverse estuary where a glacial sill at the mouth restricts exchange between thedeep basin and external waters for most of the year. Freshwater is supplied at the inlet mouth predominantlyby the Cowichan and Fraser Rivers, producing horizontal density differences that result in an inward flowin the surface layer and outward flow at depth (Herlinveaux, 1962; Gargett et al., 2003). During the springand summer months, high levels of primary productivity in surface waters and limited vertical mixing ofbasin waters below the sill result in anoxia and the accumulation of CH4, NH4+ and H2S (Lilley et al., 1982;6AbundanceNodesEdges-+Abundance+-A BFigure 1.3: Microbial co-occurrence networks. A) Model microbial co-occurrence network based onBray-Curtis correlations among OTUs. Nodes are depicted as gray dots whose size is relative to theirabundance and edges representing significant correlations (p < 0.001) are depicted as lines connectingnodes. B) Model hive panel network based on Bray-Curtis correlations among OTUs. Nodes are depictedas gray dots whose position on the axis is relative to their abundance and edges representing significantcorrelations (p < 0.001) are depicted as lines connecting nodes.Ward et al., 1989; Ward and Kilpatrick, 1990). In late summer and fall, neap tidal flows produce an influx ofdenser water from the northeastern subarctic Pacific Ocean (NESAP) that cascade over the sill, resulting invertical mixing and the re-supplying of deep basin waters with O2 and nutrients (Herlinveaux, 1962; Gargettet al., 2003). The recurring seasonal development of water column anoxia followed by deep-water renewalmakes the Saanich Inlet a model ecosystem for monitoring biogeochemical responses to changing levels ofwater-column O2 deficiency (Walsh et al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011; Wright etal., 2012).Saanich Inlet geochemical data time-seriesFor over four decades, Saanich Inlet has been the site of a number of important studies evaluating water-column chemistry, and microbial community structure, function and dynamics in relation to changing levelsof water-column O2 deficiency extensible to coastal and open ocean OMZs (Walsh et al., 2009; Zaikovaet al., 2010; Walsh and Hallam, 2011; Wright et al., 2012), including CH4 oxidation (Ward et al., 1989;Ward and Kilpatrick, 1993) and ongoing metagenomic and environmental monitoring surveys (Walsh et al.,2009; Zaikova et al., 2010). Resulting data from decadal sampling efforts is available as compendiums oftime-series observations encompassing historical O2 measurements (Herlinveaux, 1962; Lee et al., 1999)and more recent monthly geochemical data (detailed in chapter 2) and multi-omic (DNA, RNA and protein)sequence information (Hawley et al., 2017b; Torres-Beltra´n et al., 2017). In combination, these compendi-70km 3km 6km48.5°N48.6°N48.7°N 123.6°W  123.4°W  123.2°WS3British Columbia, CanadaFigure 1.4: Saanich Inlet. Saanich Inlet, on the east coast of Vancouver Island, indicating sampling stationS3.ums provide a community-driven framework for observing and predicting microbial community responsesto changing levels of O2 deficiency extensible to open ocean OMZs.Methane oxidation in Saanich InletMethane oxidation rate measurements, microbial community structure surveys, pmoA libraries and CH4incubation experiments have been previously carried out in Saanich Inlet. Results from CH4 oxidation ratemeasurements showed that during peak stratification, when subsurface (60-100 m) and deep (165-200 m)CH4 maximums are observed, methanotrophs inhabiting the oxic and dysoxic compartments of the watercolumn are expected to exhibit high CH4 oxidation rates (2 nmol L-1 h-1) (Ward et al., 1989).Interestingly, sequences affiliated with canonical methanotrophs such as Methylococcales and Methy-lomonas within the Gammaproteobacteria were recovered as rare biosphere components from SSU rRNAgene library datasets even during peak stratification (Zaikova et al., 2010). This low abundance of canonicalmethanotrophs has also been reported for open ocean OMZs (Woebken et al., 2007; Stevens and Ulloa,2008; Glaubitz et al., 2010). Moreover, pmoA libraries were dominated by a phylotype that clusters out-side of known canonical methanotrophs within the Alpha and Gammaproteobacteria (Stilwell, 2007). Morerecently a microcosm study enriched with CH4 revealed a marked increase of bacterial SSU rRNA gene se-quences affiliated with Methylophilales, Methylococcales, Methylophaga, Thiotrichales and Planctomycetes(Sauter et al., 2012).Previous surveys provide insight into microbial community members potentially associated with CH4oxidation in the Saanich Inlet water column. However, these surveys did not couple methanotrophic bacteriataxonomic affiliation, distribution and abundance throughout water-column O2 compartments. In addition,no linkages where made between specific community members to CH4 oxidation gene abundance, distri-bution or expression patterns, constituting the “Saanich Inlet CH4 oxidation conundrum”. Moreover, noanalysis was conducted comparing Saanich Inlet methanotrophic community composition and potential ac-tivity to observations reported in other coastal or open ocean OMZs.81.3.4 Thesis objectives and overviewIn this thesis, I addressed the “Saanich Inlet CH4 oxidation conundrum” detailed above using the geochemi-cal and sequence information available for Saanich Inlet with additional and novel experimental approaches.Results from this thesis provide a more coherent understanding of microbial community structure, functionand dynamics associated with CH4 oxidation in Saanich Inlet with important implications for understandingthe mechanisms controlling CH4 cycling throughout the global ocean.Chapter 2: A compendium of geochemical information from the Saanich Inlet water column.In Chapter 2, I detail the chemical and physical datasets used throughout this thesis that pair with SSUrDNA tag, metagenomic, metatranscriptomic and metaproteomic information. Chapter 2 aims to reinforcethe need for model systems and to provide more in-depth information on the Saanich Inlet time-series,while supporting the using model systems to understand coupled biogeochemical processes and ecologicalinteractions further.Chapter 3: Sampling and processing methods impact microbial community structure and function.In Chapter 3, I examine the effect of water sample collection and filtering methods on microbial com-munity structure and function using coupled SSU rDNA and rRNA 454 tag sequencing data. Chapter 3 aimsto highlight the need for standards of practice in model systems for extensible results and the interpretationof microbial community composition and biogeochemical cycling metabolism.Chapter 4: Protistan parasites along water column oxygen gradients: a network approach to assessingpotential host-parasite interactions.In Chapter 4, I present correlation analyses on SSU rDNA and rRNA 454 tag sequencing informationexamining the O2 effects on ecological interactions and energy flow in marine ecosystems. Chapter 4 aims todetermine host-parasite interactions along water-column O2 gradients and their potential effects on nutrientcycling in the Saanich Inlet water column.Chapter 5: Methanotrophic community dynamics in Saanich Inlet.In Chapter 5, I conduct correlation analyses on SSU rDNA tag sequencing information coupled withgeochemical parameters to identify the significant distribution patterns of methanotrophs and interactionsalong water-column O2 gradients. Chapter 5 aims to determine microbial community indicator groups andco-occurrence patterns associated with O2, CH4 and alternative terminal electron acceptors to gain insightinto community-level interactions for CH4 oxidation under water column O2 deficiency.Chapter 6: Community-level interactions support CH4 oxidation in the Saanich Inlet O2 -deficient watercolumn.In Chapter 6, I show CH4 incubation experiments with O2-deficient waters to generate multi-omic se-quencing information and identify community-level metabolic interactions for CH4 oxidation. Chapter 6aims to evaluate how gene expression and metabolism related to CH4 oxidation change under differentwater-column O2 and CH4 conditions, and to link metabolic potential to microbial community networkscarrying out CH4 oxidation under under water-column O2 deficiency.9Chapter 2A compendium of geochemical informationfrom the Saanich Inlet water column1This chapter details the methodologies and workflows for generating geochemical information includingphysical (temperature, salinity, density, irradiance, and fluorescence), chemical (PO43-, SiO2, NO3-, NO2-,NH4+, and H2S), dissolved gas (O2, CO2, N2, N2O, CH4), and biological (cell counts) parameter data.Moreover, this chapter reinforces the need for model systems and provide more in-depth information on thetime series ending with a call to use these systems to better understand coupled biogeochemical processesand ecological interactions.2.1 IntroductionMarine oxygen (O2) minimum zones (OMZs) are hotspots for the biogeochemical cycling of carbon, nitro-gen and sulphur with resulting feedback on nitrogen loss processes and climate active trace gas productionincluding nitrous oxide (N2O) and methane (CH4) (Lam et al., 2009; Ward et al., 2009; Canfield et al.,2010; Lam and Kuypers, 2011; Naqvi et al., 2010). The effects of climate change, including increasedstratification and reduced O2 solubility in warming waters are resulting in OMZ expansion and intensifica-tion (Arrigo, 2005; Whitney et al., 2007; Diaz and Rosenberg, 2008; Stramma et al., 2008; Paulmier andRuiz-Pino, 2009; Keeling et al., 2010; Schmidtko et al., 2017) reinforcing the need to monitor changes inwater column geochemistry in O2-deficient waters. Saanich Inlet is a seasonally anoxic fjord on the coast ofVancouver Island, British Columbia, Canada (Carter, 1932; 1934; Herlinveaux, 1962; Anderson and Devol,1973) where the recurring seasonal development of water column anoxia followed by deep water renewalmakes it a model ecosystem for monitoring biogeochemical responses to changing levels of water columnO2-deficiency (Walsh et al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011; Wright et al., 2012).Oceanographic surveys in OMZ waters rely on a standard suite of measurements including temperature,salinity, density and conductivity. Additional parameters including irradiance, used to measure water columnlight penetration, fluorescence used to monitor chlorophyll concentrations and dissolved gases including1A version of this chapter appears as Torres-Beltra´n, M., Hawley, A.K. et al. 2017. A compendium of geochemical informationfrom the Saanich Inlet water column. Sci.Data. 4. doi:10.1038/sdata.2017.15910O2 and carbon dioxide (CO2) provide information on primary production (Lewis et al., 1985; Kolber andFalkowski, 1993; Wright et al., 2012). Chemical measurements of phosphate (PO43-), silicic acid (SiO2),and nitrate (NO3-) are measured as essential nutrients supporting growth and cell division (Arrigo, 2005).Nitrite (NO2--) and ammonium (NH4+) are also measured to better constrain nitrogen cycling processes(Lam et al., 2009; Ward et al., 2009; Lam and Kuypers, 2011; Wright et al., 2012). Because some OMZs canbecome completely anoxic, hydrogen sulfide (H2S) concentrations can be used as an indicator for sulphatereduction driving chemoautotrophic metabolism (Canfield et al., 2010; Ulloa et al., 2012). Measurementsof N2O and CH4 can also be used to monitor potential climatological impacts of OMZ expansion (Monteiroet al., 2006; Lam et al., 2009; Ward et al., 2009; Canfield et al., 2010; Lam and Kuypers, 2011; Naqvi et al.,2010). Collectively, these measurements define geochemical gradients in OMZ water columns that shapethe conditions for coupled biogeochemical cycling.Here I present a compendium of time-series observations encompassing historical O2 measurements(Herlinveaux, 1962; Lee et al., 1999) (Fig. 2.1A) and more recent monthly monitoring efforts in Saanich In-let from 2006 through 2014, representing over 100 independent sampling expeditions (Fig. 2B). This com-pendium contains physical (temperature, salinity, density, irradiance, and fluorescence), chemical (PO43-,SiO2, NO3-, NO2-, NH4+, and H2S), dissolved gas (O2, CO2, N2, N2O, CH4), and biological (cell counts)parameter data (Fig. 2.1B and C) useful in comparing to other oceanographic time-series from the north-west Atlantic to Eastern Tropical Pacific through the Global Ocean Sampling expeditions (Sunagawa et al.,2015), the Hawaii and Tara Oceans (Karl and Church, 2014; Pesant et al., 2015) and Bermuda AtlanticTime-series (Steinberg et al., 2001) and in the development of biogeochemical models. In addition, thischemical compendium partners with a compendium of multi-omic sequence information from the SaanichInlet water column detailing time-series microbial multi-omic datasets (Hawley et al., 2017b). Combined,these compendiums provide a community-driven framework for observing and predicting microbial com-munity repsonses to changing levels of oxygen deficiency extensible to open ocean OMZs.2.2 MethodsTime-series monitoring in Saanich Inlet was conducted on a monthly basis aboard the MSV John Stricklandat station S3 (48o 35.500 N, 123o 30.300 W) as previously described (Zaikova et al., 2010). Water samplesfrom 16 high-resolution (HR) depths at station S3 (10, 20, 40, 60, 75, 80, 90, 97, 100, 110, 120, 135, 150,165, 185 and 200 m) spanning oxic (¿90 µmol O2 kg-1), dysoxic (90-20 µmolO2 kg-1), suboxic (20-1 µmolO2 kg-1) anoxic (¡1 µmol O2 kg-1) and sulfidic water column compartments (Wright et al., 2012) werecollected using Niskin or Go-Flow bottles for dissolved gasses: O2, CO2, CH4, Nitrogen gas (N2), N2O;nutrients: NO3-, NO2--, NH4+, SiO2, PO43-, H2S; and cell counts. Sampling methods for HR samples andadditional six large-volume depths (10, 100, 120, 135, 150 and 200 m) collected for time-series multi-omicsequence information analyses are published in an accompanying compendium (Hawley et al., 2017b).2.2.1 Environmental samplingHistorical dissolved O2 concentrations were obtained from station S3 by sampling with Niskin bottles at dis-crete depths and subsequently analyzing water samples using various modifications of the Winkler method11CD 20082006 2007 2009 2010 2011 2012 2013 2014Feb08 Jan09 Jan10 Jan11 Jan12 Jan13 Jan14>25010050155Oxygen (µM)<3100Feb06CTD dataNutrientsGasesB05010015020019531962197019801991200020102014Depth /mDepth /m050150200Data time pointsFigure 2.1: Geochemical data time series in the Saanich Inlet. A) Sampling station S3 in the SaanichInlet B) Historical sampling effort in Saanich Inlet depicted as O2 sampling points from 1953 to 2014. C)Oxygen concentration contour for CTD data (February 2008 onward), and points for 16 sampling depthsfor nutrients and gases. D) Sample inventory from February 2006 to October 2014 showing historical, CTDand nutrient datasets (solid black).(Herlinveaux, 1962; Carpenter, 1965; Lee et al., 1999) (Data Citation 1). Historical water column pro-files can also be accessed at the Ocean Sciences Data Inventory website hosted by the Institute of OceanSciences and Fisheries and Oceans Canada (http://www.pac.dfo-mpo.gc.ca/science/oceans/data-donnees/search-recherche/profiles-eng.asp). Samples collected from February 2006 to February 2008 were pro-cessed and analysed for dissolved gases and nutrients as first reported in Zaikova et al (Zaikova et al., 2010)(Fig. 2.2). Beginning on February 2008, a Sea-Bird SBE 25 CTD (conductivity, temperature and depth),with Sea-Bird SBE 43 dissolved O2 and Biospherical Instruments PAR sensors attached was used to measureconductivity, temperature, dissolved O2, PAR/Irradiance and fluorescence (Data Citation 1). To minimizethe effects of off-gassing, waters were collected in the following order; dissolved O2 for Winkler titration(from select depths for CTD calibration), dissolved gases (N2O and CH4), NH4+, H2S, nutrients, cell counts(Data Citation 1) and salinity (from selected depths for CTD calibration). A detailed seawater samplingvideo protocol can be found online (http://www.jove.com/video/1159/seawater-sampling-and-collection).2.2.2 Chemical DataCTD data analysisCTD data were downloaded, converted and pre-processed in the laboratory using the SeaBirdSeasoftsoftware. Downcast data of the deepest cast (200 m) was extracted and converted from ASCII format12O2 (μM)0 100 200 30050100150200Depth (m)0H2S (μM)0 10 2050100150200Depth (m)0NO3 (μM)0 10 20 3050100150200Depth (m)0Temperature (oC)8 10 12 1450100150200Depth (m)0Salinity (psu)50100150200Depth (m)020 28 3050100150200Depth (m)020 22 24Density ( θ)A B CD E FFigure 2.2: Time series environmental parameters water column profiles. Panel showing dot plots foroxygen (O2 ; blue), nitrate (NO3- ; green), hydrogen sulphide (H2S ; purple), temperature (oC ; red), salinity(psu ; black) and density (θ ; gray) measurements along the depth profile for samples taken from February2008 to October 2014 at Station S3 in Saanich Inlet.into a .cnv file for manual curation. Salinity and density were calculated using the Derive module withthe corrected conductivity measurements. Temperature and salinity were exported using an ITS-90 scale.Oxygen sensor measurements collected in millilitre per litre (ml/L) were converted to micromolar (µM)units (Data Citation 1). Discrete Winkler analyses from water samples spanning LV depths were used tocalibrate the CTD O2 measurements (Data Citation 1).Nitrate, Phosphate and Silicic acidFor each depth, sample water was filtered through a 0.2 µm acrodisc (Millipore) and used to rinse a 15ml tube three times before filling with 14 ml. Samples were stored on ice and later in the lab at -20o C for upto four months prior to analysis. A Bran Luebbe AutoAnalyser 3 using air-segmented continuous-flow andstandard colorimetric methods was used for analysis. In brief, nitrate (NO3-) was reduced to nitrite (NO2-)by a copper-cadmium reduction column. Nitrite was then quantified by a modified colorimetric assay (Arm-strong et al., 1967), reading sample absorbance at 550 nm. Orthophosphate (PO43-) was quantified basedon the colorimetric method for reduced phospho-molybdenum complex, reading samples absorbance at 880nm (Murphy and Riley, 1962). Silicic acid (H4SiO4) was quantified by reduction to a molybdenum blue13complex, reading sample absorbance at 820 nm. Oxalic acid was added to remove phosphate interference(Armstrong et al., 1967) (Data Citation 1).AmmoniumA fluorometric measurement protocol for ammonium (NH4+) analysis was carried out as previouslydescribed in Holmes et al for marine samples (Holmes et al., 1999). For each depth, glass amber scintillationbottles were rinsed three times, then filled to overflowing and capped immediately to minimize off-gassingof NH4+ and stored on ice for 1-3 hours before processing. A total of 5 ml of sample water was transferredto vials with 7.5 ml o-phthaldialdehyde (OPA; Sigma) in triplicate. Simultaneously, 7.5 ml of OPA wasadded to prepared NH4+ standard curve (0.025 10.0 µM NH4Cl) and stored at room temperature for up to4 hours. Fluorescence at 380ex/420emm was read using a Turner Designs TD-700 fluorometer (2006-2009)or Varioskan plate reader (2009-2014) in triplicate with 300 µl of sample or standard in a 96-well roundbottom plate (Corning) (Fig. 2.3) (Data Citation 1).NitriteThe protocol for NO2- analysis was carried out as previously described in Armstrong et al modifiedfor marine samples (Armstrong et al., 1967). For each depth, sample water was filtered through a 0.2 µmacrodisc (Millipore) and used to rinse a 15 ml tube three times before filling with 14 ml filtered samplewater and stored on ice for 1-3 hours before processing. A total of 2 ml of sample water was transferredto 4 ml plastic cuvettes in triplicate, and 100 µl sulphanilamide (Sigma) and 100 µl nicotinamide adeninedinucleotide (NAD; Sigma) were added. Simultaneously, reagents were added to prepared standards (0.0255.0 µM NaNO2-). Cuvettes were inverted and stored on ice for up to 4 hrs. Absorbance at 542 nm was readusing a Cary60 spectrometer (Fig. 2.3) (Data Citation 1).Hydrogen sulfideThe protocol for hydrogen sulfide (H2S) was carried out as previously described in Cline (Cline, 1969)modified for marine samples. For each depth, 10 ml sample water was collected directly into a 15 ml tubecontaining 200 µL 20% Zinc Acetate (Sigma) and stored on ice for 4-24 hours before processing. Sampleswere mixed prior transferring a total of 300 µL of sample into triplicate wells of a 96-well round or flat-bottom plate (Corning), and 6 µL Hach Reagent (Hach) 1 and 2 for sulphide assay were added to each well.After 5 min incubation, absorbance at 670 nm was read using a spectrophotometer (2008-2009) or Varioskanplate reader (2009-2014) (Data Citation 1).Cell countsFor each depth, 10 ml sample water was collected directly into a sterile 15 ml tube containing 1.1 mlof 37% formaldehyde and stored on ice. Back at the lab, samples were stored at 4o C for up to two daysprior to cell counting using a BD LSR II flow cytometer (2008- 2012) or MACS Quant Analyzer (20122014) based on the following protocols. For BD LSR II, a dye mixture was prepared by diluting 3 µL of theSYBR Green I (Invitrogen) dye in 1830 µL of sterile water. Six drops (Alignflow) alignment beads werethen added to this mixture. In a round-bottom polystyrene tubes, 25 µL of the dye mix was added to 475 µLof the water sample (in triplicates). The cells and beads were then counted using BD LSR II flow cytometer.For MAXSQuant, a dye mixture was prepared by diluting 240 µL of seawater sample with 10 µL of SYBRGreen I (Invitrogen) dye mix which contains 6 µm flow cytometry blue laser alignment beads (Alignflow),14100 200 300012345R2 = 0.99960NH4 (μmol)Fluoresence (380ex/420em)A0.05 0.15 0.25R2 = 0.99990123450NO2 (μmol)Absorbance (452nm)BR2 = 0.9990 2100.20.40.60.8Detector intensity x 106N2O (nmol)CFigure 2.3: Validation for environmental parameters. A-B) Typical standard curves for chemical param-eters ammonium (NH4+) and nitrite (NO2-), and C) gas concentration nitrous oxide (N2O).for calibration purposes. SYBR Green mix was prepared by diluting 4 µL of the dye in 1570 µL of sterilewater following an addition of 30 µL beads. Samples are prepared in triplicates in a 96-well flat bottomblack plate (Corning) and run on MACSQuant Analyzer (MiltenyiBiotec) (Data Citation 1).Dissolved GasesFor each depth, sample water was collected through silicon tubing (∼15 cm long and 1/4“ thick, pre-flushed for a few seconds with sample water) into a 30 or 60 ml borosilicate glass serum vial, overflowingthree times the volume and taking care to remove air bubbles from the tubing and vial during filling. Thevials were spiked with 50 µL saturated mercuric-chloride solution, then crimp-sealed with a butyl-rubberstopper and aluminium cap. Samples were stored in the dark at 4o C until processing. Dissolved gases wereanalysed using either headspace for CO2, CH4, N2 and N2O (2006-2009, samples stored for up to 2 years)or automated purge-and-trap for CH4 and N2O only (2009-2014, samples stored for <3 months) coupledwith gas chromatography-mass spectrometry (GC-MS) (Capelle et al., 2015) (Data Citation 1). Sampleswith >20% standard deviation between replicates were excluded to discard any long storage effects.2.2.3 Data RecordsData record 1 The Saanich Inlet O2 historical data (1953-2000) is accessible in comma-separated-valueformat file Historical O2 DATA.csv on the Dryad Digital (Data Citation 1).Data record 2 The Saanich Inlet time-series CTD data is accessible in comma-separated-value formatfile Saanich TimeSeries CTD DATA.csv on the Dryad Digital Repository (Data Citation 1).Data record 3 The Saanich Inlet time-series chemical data is accessible in comma-separated-value formatfile Saanich TimeSeries Chemical DATA.csv on the Dryad Digital Repository (Data Citation 1).Data record 4 The Saanich Inlet time-series Winkler O2 data is accessible in comma-separated-valueformat file Saanich TimeSeries Winkler DATA.csv on the Dryad Digital Repository (Data Citation 1).2.3 Technical ValidationData quality controlData in the Saanich Inlet time series was collected and processed by experienced scientists with extensive15training in the sampling methods and data processing steps described above. People interested in becomingpart of the scientific crew were invited to participate in training sessions with experienced scientists in thefield and laboratory to gain practical experience. Once in the field, trainees were carefully supervised duringsample collection for a minimum of 3 months for quality assurance. Following each cruise, the acting chiefscientist compiled all chemical and physical data collected and conducted initial quality controls, checkingfor outliers and verifying standard curves. Data were then entered into an in-house database along with fieldnotes and precise records of volumes of water filtered informing downstream analyses.CTD and chemical data validationThe SeaBird 43 dissolved O2 sensor was calibrated by Winkler O2 measurements (Winkler, 1888).Samples from selected depths were collected into Winkler glass Erlenmeyer flasks using latex tubing, over-flowing three times to ensure no air contamination. Oxygen concentration was determined using a Brinkmanautotitrator, routinely calibrated with a potassium iodide standard. Stability of CTD O2 measurements wasdetermined by comparing the high values with Winkler measurements, and low values with sulfidic profileswhere the sensors levels off. Where H2S is detected I consider O2 measurements to be 0 µM based on spon-taneous auto-oxidation reaction of H2S with O2. I have estimated our limit of detection for the automatedWinkler method at∼0.007 ml/L or∼0.3 µM. The SeaBird conductivity sensor was calibrated using salinitysamples collected at selected depths. Salinity glass bottles were rinsed 4 times and filled with water sample,stored at room temperature and analyzed within 4 months on a Guildline Portasal salinometer.For each cruise, standard curves for NH4+ and NO2- were prepared. Stock solutions and reagents for bothassays were freshly made every three months and stored in the dark at 4o C and were tested prior to beingused for analysis. Stock solution quality and assay validation was carried out using linear regression andcalculating the r squared value (r2 0.90) on the absorbance data (Fig. 2.3). Standard curve stock solutionsand reagents for H2S assay were evaluated every three months based on manufacturers instructions. I haveestimated our limit of detection for these assays to be 0.001 µM NH4+, 0.0006 µM NO2-, and 1.7 µM H2S.Samples for NO3-, PO43- and H4SiO4 were run in single measurements. Autoanalyzer estimated limit ofdetection for these measurements are 0.020 µM NO3-, 0.012 µM PO43- and 0.100 µM H4SiO4.Flow cytometry validationConcentration of flow cytometry (FL) alignment beads was determined by microscopy using a hemo-cytometer. Bead counts for each FL run were then used to calculate the volume of sample measured. Twoblanks were included in each FL run, and consisted of sterile water bead/dye solution with sterile water inplace of sample water, to ensure instrument cleanliness and optics function. Size gates were set to includebeads and bacterial and archaeal cell sizes and to reduce noise of any small particulate debris.Gas analysis validationA thorough review of the Purge and Trap GCMS (PT-GCMS) method validation has been previously de-scribed (Capelle et al., 2015). Standard curves were run at the start of each batch of 25 samples by injectingprecisely measured quantities of a standard gas mixture (CH4, N2O, CO2 and N2) calibrated against NationalOcean and Atmospheric Administration (NOAA) certified reference gas mixture. Single standards were alsomeasured every ∼2 hours (5 - 6 sample per run) to monitor instrument drift. The precision of CH4 and N2Omeasurements based on replicate measurements of air-equilibrated water samples was <4%. Accuracy16was confirmed by measuring dissolved N2O and CH4 in carefully prepared air-equilibrated, temperature-controlled Milli-Q water and comparing this to expected concentrations based on gas-solubility equations(Wiesenburg and Guinasso, 1979; Weiss and Price, 1980). Detection limits depend on the volume of samplebeing purged, and were 0.8 nM for CH4 and 0.5 nM for N2O for the samples analyzed in this time-series(2009-2014) (Fig 2.3). Samples were run in duplicate or triplicate to ensure reproducible readings. Therelative standard deviation between replicate samples was calculated and included in the output data. Theoutput data are also carefully inspected to ensure optimal instrument performance during sample analysisbefore being submitted to the database.2.4 Data Citation1. M. Torres-Beltran, A.K. Hawley et al. Dryad Digital Repository. http://dx.doi.org/10.561/dryad.nh035(2017).2.5 Conclusion and applicationThe combined use of geochemical and multi-omic sequence information have led to new insights into cou-pled biogeochemical cycling of carbon, nitrogen and sulfur between key microbial players and the develop-ment of a predictive ecosystem model describing the flow of multi-omic sequence information and processrates along O2 gradients (Hawley et al., 2014; Louca et al., 2016; Hawley et al., 2017a). Thus, time-seriesdata from Saanich Inlet provides a community-driven framework for observing and predicting microbialcommunity responses to ocean deoxygenation across multiple scales of biological organization.In this thesis, I used time-series geochemical data to correlate O2, CH4, NO3- and NO2- with methan-otrophic community members and predict community-level interactions for CH4 oxidation throughout O2gradients in the Saanich Inlet water column.17Chapter 3Sampling and processing methods impactmicrobial community structure andfunctionThis chapter reinforces the need for standards of practice in model systems for extensible results and focuseson the sample collection problem for multi-omics information. In addition, the methods used here form thebasis of microbial community composition and structure analyses carried out throughout chapters 4 to 6 inthe investigation of community-level interactions along water column O2 gradients.3.1 IntroductionCurrent research efforts are defining the interaction networks underlying microbial metabolism in oxygenminimum zones (OMZs) and generating new insights into coupled biogeochemical processes driving nutri-ent and energy flow among and between trophic levels on ecosystem scales (Hawley et al., 2014;Cram et al.,2015;Louca et al., 2016;Torres-Beltra´n et al., 2016a). However, marine microbial responses at the individ-ual, population and community levels to OMZ expansion, and the concomitant impact of these responses onglobal-scale nutrient and energy cycling remain poorly constrained due in part to inconsistent, and perhapsinadequate sampling methods that limit cross-scale comparisons between locations and may cloud our viewof in situ microbial processes.Over the past twenty years, oceanographic researchers have increasingly used multi-omic (DNA, RNA,protein and metabolites) methods to determine microbial community structure, function and activity in re-lation to physical, chemical and biological oceanographic processes. A wide range of sample collectionand processing methods have been used to generate these data sets without standardization (Stein et al.,1996;Fuhrman and Davis, 1997;Massana et al., 1997;Acinas et al., 1999;Crump et al., 1999;Murray et al.,1999;James et al., 2000;Moeseneder et al., 2001;LaMontagne and Holden, 2003;Venter et al., 2004;De-Long et al., 2006;Hewson and Fuhrman, 2006;Rusch et al., 2007;Waidner and Kirchman, 2007;Brown etal., 2009;Ghiglione et al., 2009;Walsh et al., 2009;Zaikova et al., 2009;Canfield et al., 2010;Gilbert et al.,2010;Hurwitz et al., 2013;Rodriguez-Mora et al., 2013;Smith et al., 2013;D’Ambrosio et al., 2014;Ganesh18et al., 2014;Parris et al., 2014;Brum et al., 2015;Ganesh et al., 2015;Orsi et al., 2015;Padilla et al., 2015;Pe-sant et al., 2015;Suzuki and Shimodaira, 2015). While large-scale patterns appear to be consistent betweenstudies with respect to major taxonomic groups and water column compartments, our ecological perspectiveis blurred by inconsistencies in marker gene selection and coverage.Recent reports have begun to evaluate potential biases in OMZ microbial community structure, functionand activity with emphasis on sample collection and filtration methods. Water column sampling typicallyinvolves the use of collection bottles (Niskin or GO-FLO) and on-ship filtration to concentrate microbialbiomass into two primary size fractions, a larger particle associated (>1-30 µm) and smaller free-living(<1-0.2 µm) fraction (Padilla et al., 2015). Size fractionation surveys conducted in the Eastern TropicalSouth Pacific (ETSP) and Eastern Tropical North Pacific (ETNP) showed differential microbial communitystructure and nitrogen cycling functional gene distribution and expression across size fractions (Ganesh etal., 2014;Ganesh et al., 2015). In addition to filter fractionation, Padilla and colleagues working in theManzanillo Mexico OMZ observed variation in microbial community structure based on filtered water vol-ume (Padilla et al., 2015). Most recently, a study in the Cariaco Basin observed that particles sinking ontimescales relevant to sample collection and filtration can influence microbial community structure in Niskinor GO-FLO bottles (Suter et al., 2016). A comparison of metatranscriptome data obtained from bathypelagicMediterranean Sea samples collected using Niskin bottles followed by shipboard filtration vs. filtration andfixation in situ found significant shifts in gene expression for particular groups of microorganisms (Edgcombet al., 2016). These findings reinforce the need for continued evaluation of the methods used for sample col-lection and processing in order to establish standards of practice that reduce collection bias and enable morerobust cross-scale comparisons. This is particularly relevant when conducting process rate measurements inwhich community structure variation due to bottle effects can result in potential rates that do not reflect ofin situ microbial activity (Stewart et al., 2012a).The Scientific Committee on Oceanographic Research (SCOR) initiated Working Group 144 MicrobialCommunity Responses to Ocean Deoxygenation to investigate and recommend community standards ofpractice for compatible multi-omic and process rate measurements in OMZs and other oxygen deficient wa-ters in order to facilitate and promote future cross-scale comparisons that more accurately reflect in situ mi-crobial community structure, function and activity (http://omz.microbiology.ubc.ca/page4/index.html). Theinaugural workshop of SCOR Working Group 144 was held in British Columbia Canada during the week ofJuly 14, 2014. During the workshop attendees participated in practical sampling and experimental activitiesin Saanich Inlet. During spring and summer months, restricted circulation and high levels of primary pro-duction lead to progressive deoxygenation and the accumulation of methane (CH4), ammonium (NH4+) andhydrogen sulfide (H2S) in deep basin waters. In late summer and fall, oxygenated nutrient rich waters flowinto the inlet from the Haro Strait renewing deep basin waters (Carter, 1932;1934;Herlinveaux, 1962;An-derson and Devol, 1973;Zaikova et al., 2010;Walsh and Hallam, 2011;Torres-Beltra´n et al., 2016b). Theseasonal pattern of water column anoxia and renewal makes the inlet a model ecosystem for evaluatingchanges in microbial community structure, function and activity in response to changing levels of water col-umn deoxygenation. Saanich Inlet is thus a tractable environment to test different water sample collectionand processing methods relevant to OMZs.19Table 3.1: Biomass collection scheme for DNA and RNA in situ and on-ship samples. Table shows thefiltering methods used for a given sampling condition (in situ vs. on-ship), including pre-filter size cutoff,biomass collection filter, filtered volume, and molecular sample obtained.Filtration typePre-filter size (μm)Collection filter size (μm)Volume (L) Molecular samplein situ 0.4 2 DNA, RNA2.51.50.50.252.51.50.50.25DNA, RNA*DNAOn ship0.42.70.220.22Experiments carried out during the workshop were designed to compare and cross-calibrate in situ sam-pling with conventional bottle sampling methods including the use of different filter combinations and sam-ple volumes. Here, I describe the effect of these parameters on microbial community structure and potentialactivity and discuss community standards development in relation to the mandate of the SCOR WorkingGroup 144.3.2 Methods3.2.1 Environmental samplingEnvironmental parameter data and molecular (DNA and RNA) collection methods used during the SCORworkshop were similar to those previously described (Zaikova et al., 2010; Walsh and Hallam, 2011; Torres-Beltra´n et al., 2016b). In brief, water sampling was conducted on aboard the MSV John Strickland at stationSI03 (48o 35.500 N, 123o 30.300 W) on July 16, 2014. Samples were collected using 12 L GoFLo bottleson a winch system for dissolved gases, nutrients and CTD with a PAR and O2 sensor attached was used tomeasure temperature, salinity, PAR/Irradiance, fluorescence, conductivity, density, and dissolved O2 at 165and 185 meter depth intervals spanning anoxic (<1µmol O2 kg-1) and sulfidic water column compartments(Fig. 3.1).3.2.2 Workshop microbial biomass collectionFor comparison to Niskin bottle sampling, water samples were collected and preserved in situ using aMcLane Phytoplankton Sampler (PPS) system deployed at 165 and 185 m depth intervals spanning anoxic(<1µmol O2 kg-1) and sulfidic water column compartments (Fig.3.1B). Sample volumes of 2 L were fil-tered onto 0.4 µm GFF membrane filters (47 mm diameter). Filter biomass was directly frozen and storedat -80 oC for downstream DNA and RNA extraction. On-ship samples were collected using Niskin bottles20Depth (m) 050100200Oxygen (µM) <3101002501502008 2009 2010 2011 2012 2013 2014 2015CTDNUTs + H2SDNARNAIn situ100Depth (m)0501502000 50 150 250Oxygen (μM)165m185mABCOn-ship(MPP)In situ (PPS)0.220.40.4Figure 3.1: SCOR workshop sampling schemeA) Saanich Inlet time-series CTD oxygen (O2) concen-tration contour (2008 -2015) showing seasonal water column stratification and deep-water renewal events.B) Sample inventory for SCOR workshop held in July 2014 includes CTD, nutrients (NO3-, NO2-, NH4+,PO4-3 and SiO2), hydrogen sulfide (H2S), and molecular on-ship and in situ data. Data shown in this surveyis depicted as solid symbols (in situ and on ship DNA and RNA (rectangle) collected at 165 (square) and 185(triangle) meters. C) Schematic model for filter combinations and filtration methods showing on-ship (MPP)filtration onto 0.4 µm pre-filters (green), filtration onto 0.22 µm filters with in-line pre-filtration using the0.4 µm pre-filters (yellow), and in situ filtration (PPS) onto 0.4 µm filters without pre-filtration (red).from 165 and 185 m as described above (Fig. 3.1B) and concentrated for DNA and RNA extractions witha MasterFlex peristaltic pump (MPP) (∼60 mL min-1) using different filter combinations (0.4 µm polycar-bonate or 2.7 µm GF/D pre-filters in-line with a 0.22 µm Sterivex polycarbonate filter cartridge) (Fig. 3.1C)and water volumes (250 ml, 500 ml, 1 L and 2.5 L) (Table 3.1). Filtered biomass on Sterivex filters waspreserved in 1.8 ml of sucrose lysis buffer (DNA analyses) or RNA later (RNA analyses) prior to storage at-80 oC. Pre-filters (0.4 µm only) were also preserved in 1.8 ml of lysis buffer (DNA) or RNA later (RNA)prior to storage at -80 oC.213.2.3 Time-series microbial biomass collectionTime-series samples were collected as previously described (Walsh et al., 2009; Zaikova et al., 2010; Haw-ley et al., 2017b; Torres-Beltra´n et al., 2016a). Briefly, large volume (10 L) samples were collected fromFebruary 2006 to February 2011 at six depths (10, 100, 120, 135, 150 and 200 m) and filtered with an in-line2.7 µm GDF glass fiber pre-filter onto a 0.22 µm Sterivex polycarbonate cartridge filter. Hi-resolution (2L) samples were collected from May 2008 to July 2010 from 16 depths (10 to 200 m) and filtered directlyonto a 0.22 µm Sterivex polycarbonate cartridge filter. All time-series samples were preserved in 1.8 ml ofsucrose lysis buffer (DNA) prior to storage at -80 oC.3.2.4 Nucleic acid extractionGenomic DNA was extracted from Sterivex filters as previously described (Zaikova et al., 2010; Hawleyet al., 2017b). Briefly, after defrosting Sterivex on ice, 100 µl lysozyme (0.125 mg ml-1; Sigma) and 20µl of RNAse (1 µl ml-1; ThermoFisher) were added and incubated at 37oC for 1 h with rotation followedby addition of 50 µl Proteinase K (Sigma) and 100 µl 20% SDS and incubated at 55 oC for 2 h withrotation. Lysate was removed by pushing through with a syringe into 15 mL falcon tube (Corning) andwith an additional rinse of 1 mL of lysis buffer. Filtrate was subject to chloroform extraction (Sigma) andthe aqueous layer was collected and loaded onto a 10K 15 ml Amicon filter cartridge (Millipore), washedthree times with TE buffer (pH 8.0) and concentrated to a final volume of between 150-400 µl. TotalDNA concentration was determined by PicoGreen assay (Life Technologies) and genomic DNA qualitydetermined by visualization on 0.8% agarose gel (overnight at 16V). Genomic DNA was extracted from 0.4and 2.7 µm pre-filters as follows. The filter was cut in half using sterile scissors. One half was minced intoin smaller pieces and used for DNA extraction while the remaining half was stored at -80 oC. Filter pieceswere transferred to a 15mL falcon tube followed by addition of 1.8 mL lysis buffer and 150 µL 20% SDS.In order to ensure biomass removal form filter, 3 and 2 mm zirconium beads were added for beat beadingusing a vortex mixer at maximum speed. Filters were shaken for 4 minutes in 2 minutes laps then subjectedto chloroform extraction and processed in the same way as described above for Sterivex filters.Total RNA was extracted from Sterivex filters using the mirVana Isolation kit (Ambion) (Shi et al., 2009;Stewart et al., 2010) protocol modified for sterivex filters (Hawley et al., 2017b). Briefly, after thawing thefilter cartridge on ice RNA later was removed by pushing through with a 3 ml syringe followed by rinsingwith an additional 1.8 mL of Ringer’s solution and incubated at room temperature for 20min with rotation.Ringer’s solution was evacuated with a 3 ml syringe followed by addition of 100 µl of 0.125 mg ml-1lysozyme and incubated at 37 oC for 30 min with rotation. Lysate was removed from the filter cartridgeand subjected to organic extraction following the mirVana kit protocol. DNA removal and clean up andpurification of total RNA were conducted following the TURBO DNA-free kit (ThermoFisher) and theRNeasy MinElute Cleanup kit (Qiagen) protocols respectively. Total RNA concentration was determinedby RiboGreen analysis (Life Technologies) prior to synthesize first strand cDNA using the SuperScript IIIFirst-Strand Synthesis System for RT-qPCR (Invitrogen) according to manufacturer instructions. Total RNAwas extracted from 0.4 and 2.7 µm filters as follows. The filter was cut in half using sterile scissors. Onehalf was minced into in smaller pieces and used for RNA extraction while the remaining half was stored at22-80 oC. Filter pieces were transferred to a 15 mL falcon tube followed by addition 1.8 ml MirVana LysisBuffer and 100 µl of 0.125 mg ml-1 lysozyme. In order to ensure biomass removal form filter, 3 and 2 mmzirconium beads were added for beat beading using a vortex mixer at maximum speed. Filters were shakenfor 4 minutes in 2 minutes laps followed by an incubation at 37 oC for 30 min with rotation, then processedin the same way as described above for Sterivex filters.3.2.5 Small subunit ribosomal RNA sequencingExtracted DNA and cDNA from 165 and 185 meter depth intervals (2.7, 0.4 and 0.22 µm) was used togenerate SSU rDNA and rRNA pyrotags with three domain resolution. PCR amplification procedures werecarried out as previously described (Hawley et al., 2017b). In brief, pyrotag libraries were generated by PCRamplification using multi-domain primers targeting the V6-V8 region of the SSU rRNA gene (Allers et al.,2013): 926F (5’-cct atc ccc tgt gtg cct tgg cag tct cag AAA CTY AAA KGA ATT GRC GG-3’) and 1392R(5’-cca tct cat ccc tgc gtg tct ccg act cag-<XXXXX>-ACG GGC GGT GTG TRC-3’). Primer sequenceswere modified by the addition of 454 A or B adapter sequences (lower case). In addition, the reverse primerincluded a 5 bp barcode designated <XXXXX> for multiplexing of samples during sequencing Twenty-five microliter PCR reactions were performed in triplicate and pooled to minimize PCR bias. Each reactioncontained between 1 and 10 ng of target DNA, 0.5 µl Taq DNA polymerase (Bioshop inc. Canada), 2.5µL Bioshop 10x buffer, 1.5 µl 25 mM Bioshop MgCl2, 2.5 µL 10 mM dNTPs (Agilent Technologies)and 0.5 µL 10 mM of each primer. The thermal cycler protocol started with an initial denaturation at 95oC for 3 minutes and then 25 cycles of 30 s at 95 oC, 45 s at 55 oC, 90 s at 72 oC and 45 s at 55 oC.Final extension at 72 oC for 10 min. PCR products were purified using the QiaQuick PCR purification kit(Qiagen), eluted elution buffer (25 µL), and quantified using PicoGreen assay (Life Technologies). SSUrDNA and rRNA amplicons were pooled at 100 ng for each sample. Emulsion PCR and sequencing ofthe PCR amplicons were sequenced on Roche 454 GS FLX Titanium at the Department of Energy JointGenome Institute (DOE-JGI), or the McGill University and Ge´nome Que´bec Innovation Center.A total of 1,027,601 small subunit ribosomal rDNA and rRNA pyrotags were processed together usingthe Quantitative Insights Into Microbial Ecology (QIIME) software package (Caporaso et al., 2010). Readswith length shorter than 200 bases, ambiguous bases, and homopolymer sequences were removed priorto chimera detection. Chimeras were detected and removed using chimera slayer provided in the QIIMEsoftware package. Sequences were then clustered into operational taxonomic units (OTUs) at 97% identityusing identity using UCLUST with average linkage algorithm. Prior to taxonomic assignment, singletonOTUs (OTUs represented by one read) were omitted, leaving 29,589 OTUs. Representative sequences fromeach non-singleton OTU were queried against the SILVA database release 111 using the BLAST algorithm(Altschul et al., 1990).3.2.6 Statistical analysis and data visualizationStatistical analyses were conducted using the R software package (RCoreTeam, 2013). Pyrotag datasetswere normalized to the total number of reads per sample. Hierarchical cluster analysis (HCA) and nonmetricmultidimensional scaling (NMDS) were conducted to identify community compositional profiles associated23with water column compartments using the pvclust (Suzuki and Shimodaira, 2015) and MASS (Venablesand Ripley, 2002) packages with Manhattan Distance measures, and statistical significance to the resultingclusters as bootstrap score distributions with 1,000 iterations and NMDS stress value <0.05.Diversity in-dexes (Shannon and alpha diversity) were calculated to identify changes in community structure based onfiltration parameters using the vegan (Oksanen et al., 2015) package. Non-parametric Friedman test blockdesign was conducted to test the significance of the volume variation on the community composition usingthe stats (RCoreTeam, 2013) package. In addition, one-way ANOVA was conducted to test the significanceof filtration combinations on taxa relative abundance using the ggpbur (Kassambara, 2017) package.Multi-level indicator species analysis (ISA) was conducted to identify OTUs specifically associated withdifferent filtering parameters based on groups resolved in HCA using the indicspecies package (De Caceresand Legendre, 2009). The ISA/multi-level pattern analysis calculates p values with Monte Carlo simulationsand returns indicator values (IV) and p-values with α <0.05. The IVs range between 0 and 1, whereindicator OTUs considered in the present chapter for further community analysis shown an IV >0.7 andp-value <0.001. ISA groups abundance was visualized as dot plots using the bubble.pl pearl script (http://hallam.microbiology.ubc.ca/LabResources/Software.html). Taxonomic distribution of identified OTUs wasvisualized using the ggplot2 (Wickham, 2009) package. The total SSU rRNA: rDNA ratios were calculatedfor the subset of matching samples (165 m 250 mL and 2.5 L on-ship 0.22 µm filters with in-line 0.4 µmpre-filtration, and 185 m 500 mL and 2.5 L on-ship 0.22 µm filters with in-line 0.4 µm pre-filtration) toaccount for variation in taxon abundance in the DNA pool (Frias-Lopez et al., 2008; Stewart et al., 2012b)and compared for a subset of microbial groups to explore how filtration parameters influence recovery ofpotentially active OTUs. I then selected OTUs based on ISA results and their shifts in abundance amongfiltering conditions.3.2.7 Data depositionSSU rDNA and rRNA pyrotag sequences reported in this chapter have been submitted to the The NationalCenter for Biotechnology Information (NCBI) under BioSample numbers: SAMN05392373 - SAMN05392466.3.3 Results3.3.1 Water column conditionsSamples were collected during a stratification period characteristic of summer months (June-August) inSaanich Inlet (Carter, 1932; 1934; Herlinveaux, 1962; Zaikova et al., 2010). Below 150 m, water columnCTD O2 concentrations were below <3 µM dissolved O2 consistent with water column anoxia during peakstratification (Fig 3.1A; Table A.1). In addition, increasing levels of H2S (13.95 µM) and NH4+ (6.1 µM)at 185 m were also observed indicating anoxic and sulfidic conditions in deep basin waters. NO3- and NO2-concentrations peaked at 150 m reaching 12 µM and 0.6 µM, respectively. Phosphate concentrations rangedbetween 4.5 and 5.8 µM from 150 to 185 m, and SiO2 concentration peaked at 185 m reaching 110 µM(Table A.1).243.3.2 Benchmarking workshop and Saanich Inlet time-series resultsI evaluated microbial community structure using 521 time-series samples traversing the Saanich Inlet watercolumn (Fig. A.1) and 29 samples collected during the SCOR workshop using rDNA pyrotag sequences tocompare and cross-calibrate in situ sampling with the McLane PPS system and bottle sampling methods.Non-metric multidimensional scaling indicated workshop samples clustered together primarily with high-resolution suboxic and anoxic samples from 165 to 200 m depth intervals (0.22 µm Sterivex filters withoutpre-filtration) collected during summer months (Fig. 3.2A). Similarities between time-series and workshopsamples provided an internal check on experimental design and a rationale for examining more granulardifferences between community structure and potential activity resulting from different filtration parameters.Workshop samples collected at 165 and 185 m depth intervals formed three groups in NMDS analysesassociated with on-ship 0.4 µm filters (group I), on-ship 0.22 µm filters with either 0.4 or 2.7 µm in-line pre-filtration (group II), and in situ 0.4 µm filters (group III) (Fig. 3.2B) were resolved based on NMDS analysis.For the most part, samples within groups partitioned by depth with small variation between sample volumesfrom the same depth. Consistent with NMDS, HCA resolved three groups (AU > 70, 1000 iterations)associated with on-ship 0.4 µm filters (group I), on-ship 0.22 µm filters with in-line 0.4 µm pre-filters(group II), and in situ filtration 0.4 µm filters (group III) (Fig. 3.3). Within each group, samples partitionedby depth and filtration volume. Friedman block test results indicated that differences in community structuredriven by sample volume were not significant.3.3.3 Size-fractionation effects on community structureBased on NMDS and HCA results, I focused on changes in OTU relative abundance and taxon identitybetween groups. Microbial community structure was primarily comprised of OTUs (>0.1% relative abun-dance) affiliated with ubiquitous OMZ taxa including Marine Group A, SAR11, SAR324, SUP05 (Field etal., 1997; Fuhrman and Davis, 1997; Brown and Donachie, 2007; Tripp et al., 2008; Lam et al., 2009; Walshet al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011; Wright et al., 2012) as well as Bacteroidetes,Desulfobacterales, and Euryarchaeota (Fig. 3.3B). Interestingly, several of these groups were not detected inon-ship 0.4 µm pre-filter samples but were recovered from in-line 0.22 µm filters (group II). These includedMarine Benthic Group E and Halobacteriales, SAR406 within the Marine Group A, Methylophilales withinthe Betaproteobacteria, Desulfuromonadales and Desulfarculales within the Deltaproteobacteria and SUP05within the Gammaproteobacteria. Conversely, Acidimicrobiales within the Actinobacteria, Cyanobacteria,Rhodobacterales within the Alphaproteobacteria, OM190 within the Planctomycetes, and eukaryotic phylaincluding Cnidaria and Arthropoda within the Opisthokonta and Diatoms within the Stramenopiles were de-tected in on-ship 0.4 µm filter samples but not recovered on in-line 0.22 µm filters or in situ 0.4 µm filters.Eukaryotic phyla affiliated with Alveolata were recovered on 0.22 µm filter samples with in-line 0.4 µmpre-filters (group I) and Phycisphaerales within the Planctomycetes were detected in in situ 0.4 µm filtersamples (group III), respectively (Fig. 3.3B).Filtration methods, including the use of different pre-filters and volumes, resulted in a significant sourceof variation (p <0.001) for the relative abundance of several bacterial phyla including Bacteroidetes, Defer-ribacteres, Alpha-, Delta- and Gammaproteobacteria, Planctomycetes, archaeal phyla including Thaumar-25NMDS 1NMDS 2−15 −10 −5 0 5 10 15−8−404−20 −10 0 10−20−1001020NMDS1NMDS2ABFigure 3.2: Cluster analysis of Saanich Inlet time-series and SCOR workshop tag data.Cluster analysisof small subunit ribosomal RNA (SSU rDNA) gene pyrotag data for the Saanich Inlet time-series (2006-2011) and SCOR workshop samples collected at 165 and 185 m. A) NMDS based on Manhattan distance(1000 iterations) of time-series and SCOR pyrotag data showing microbial community partitioning basedon oxygen gradients spanning suboxic (20-1 µmolO2 kg-1) and anoxic-sulfidic (<1 µmol O2 kg-1). B)NMDS based on Manhattan distance (1000 iterations) of SCOR pyrotag data showing microbial communitypartitioning based on filtering conditions. In both biplots samples are depicted by depth (165m = squareand 185m = triangle) and filter type (PPS in situ 0.4 µm (red), MPP on-ship 0.4 µm (green), 0.22 µm pre-filtered onto 0.4 µm filters(yellow), 0.22 µm pre-filtered onto 2.7 µm filters (blue), and time-series 0.22 µm(black)).26AlteromonadalesOceanospirillalesPseudomonadalesSUP05HalobacterialesMarine Benthic Group EThermoplasmatalesAcidimicrobialesOther BacteroidetesFlavobacterialesSAR406RhodobacteralesRhodospirillalesSAR11CiliophoraBurkholderialesMethylophilalesDesulfarculalesDesulfobacteralesDesulfuromonadalesSAR324OM190PhycisphaeralesArthropodaCnidariaDiatomeaEuryarchaeotaActinobacteriaBacteroidetesCyanobacteriaCandidate divisionProteobacteriaαδγβPlanctomycetesMarine Group AAlveolataOpisthokontaStramenopilesArchaeaBacteriaEukaryota4 8160.12Relative abundance (%)100989283 97 10090 9889100100 100100 10004080Manhattan DistanceABFigure 3.3: Microbial community partitioning based on filtration methods Top: Cluster analysis(AU>70, 1000 iterations) for 0.4 µm filter fractions. Samples are depicted with a colored bar by filtertype (PPS in situ 0.4 µm (red), and MPP on-ship 0.4 µm (green) and 0.22 µm pre-filtered onto 0.4 (yellow).Bottom: Abundant taxa (>0.1% relative abundance from total reads in sample) observed among filter com-binations for PPS in situ 0.4 µm (red), and MPP on-ship 0.4 µm (green) and 0.22 µm pre-filtered onto 0.4µm filters (yellow). The size of dots depicts the relative abundance for each taxa as indicated in key.27Table 3.2: Microbial taxa with significant abundance differences among tested filtering conditions.Significant p-values (p <0.001) obtained from one-way ANOVA testing for filtration methods effect onmicrobial relative abundance.Taxa p-valueThaumarchaeota 0.0000033Bacteroidetes 0.000051Deferribacteres 0.001α-proteobacteria 0.01δ-proteobacteria 2.1e-8γ-proteobacteria 0.0004Planctomycetes 4.1e-8Opisthokonta 0.01Rhizaria 0.01chaeota, and eukaryotic phyla including Opisthokonta and Rhizaria (Table 3.2 and Fig. 3.4). For example,the relative abundance of Bacteroidetes (Flavobacteriales), Alphaproteobacteria (SAR11, Rhodobacteralesand Rhodospirillales) and Opisthokonta (Maxillopoda) associated with on-ship 0.4 µm filters increased 5-fold compared to 0.22 µm filters with in-line 0.4 µm pre-filters and in situ 0.4 µm filters, while the relativeabundance of Deferribacteres, Deltaproteobacteria (SAR324, Desulfobacterales and Desulfarculales) andGammaproteobacteria (Oceanospirillales mainly affiliated with SUP05, Pseudomonadales and Alteromon-adales) associated with on-ship 0.4 µm filters decreased 5-fold compared to 0.22 µm filters with in-line 0.4µm pre-filters and in situ 0.4 µm filters (Fig. 3.4). Conversely, the relative abundance of Planctomycetes(Phycisphaerales and OM190) associated with in situ 0.4 µm filters increased 5-fold compared to on-ship0.22 µm filters with in-line 0.4 µm pre-filters and 0.4 µm filters (Fig. 3.4).Together these results indicate that filter selection and in-line positioning can introduce bias into micro-bial community structure data and reinforce the idea that filtration methods should be taken into considera-tion more carefully when interpreting microbial count data.3.3.4 Size fractionation effects on indicator OTUs (DNA analyses)To identify OTUs associated with specific filtration methods I conducted multi-level indicator species anal-ysis (ISA) on HCA groups I-III. As expected, resulting indicator OTUs varied with respect to filtrationmethods used (Fig. 3.5). The largest differences with respect to indicators were detected between in situand on-ship 0.4 µm filter samples. Indicator OTUs detected in on-ship 0.4 µm filter samples were mostlyaffiliated with bacterial phyla including Actinobacteria, Bacteroidetes, Cyanobacteria, Deferribacteres, Pro-teobacteria, and Verrucomicrobia, archaeal phyla including Euryarchaetoa, and eukaryotic phyla includingAlveolata, Opisthokonta, Rhizaria and Stramenopiles (Fig. 3.5). Indicator OTUs detected in in situ 0.4µm filter samples were mostly affiliated with bacterial phyla including Candidate divisions (WS3, OD1and BRC1), Chloroflexi, Deferribacteres, Firmicutes, Lentisphareae, Nitrospirae, Alpha-, Beta-, Delta- andGammaproteobacteria, and Planctomycetes, archaeal phyla including Euryarchaeota, and eukaryotic phylaincluding Alveolata, Excavata and Opisthokonta (Fig. 3.5). Indicator OTUs detected in 0.22 µm filter sam-28ThaumarchaeotaEuryarchaeotaActinobacteriaBacteroidetesCyanobacteriaexiCandidate divisionDeferribacteresFirmicutesLentisphaeraePlanctomycetesαβδγVerrucomicrobiaOther BacteriaAlveolataOpisthokontaOther EukaryotaRhizariaStranenopilesProteobacteriaArchaeaBacteriaEukaryota0 10 20 30 40 0 10 20 30 40Relative abundance (%)MPP0 10 20 30 40PPS**p-value*************Figure 3.4: Microbial community abundance shifts in relation to filtration methods. Community shiftsoccurred among and within filter combinations for PPS in situ 0.4 µm (red), and MPP on-ship 0.4 µm(green) and 0.22 µm pre-filtered onto 0.4 µm filters (yellow). The size of each box represents the average ofthe percentage of relative abundance throughout the water column over this period. For both plots, extendeddashed lines (whiskers) represent at the base the lower and upper quartiles (25% and 75%) and at the endthe minimum and maximum values encountered. The middle line represents the median. p-values forsignificant relative abundance shifts among filter groups are depicted as p<0.01 (*) and p<0.001(**) asidecorresponding taxa.29AcidimicrobialesHalobacterialesMarine Benthic Group EThermoplasmatalesBacteroidalesFlavobacterialesSphingobacterialesBRC1OD1WS3AnaerolinealesCaldilinealesCyanobacteriaSAR406ClostridialesLentisphaeriaNitrospiralesCaulobacteralesRhodobacteralesRhodospirillalesRickettsialesSphingomonadalesBurkholderialesBdellovibrionalesDesulfarculalesDesulfobacteralesDesulfovibrionalesDesulfuromonadalesMyxococcalesAlteromonadalesEnterobacterialesLegionellalesOceanospirillalesPseudomonadalesVibrionalesXanthomonadalesBrocadialesMSB−3A7MSBL9OM190PhycisphaeralesPlanctomycetalesPlanctomycetesOpitutaeCiliophoraAlveolataProtalveolataChlorophytaDiscicristataAscomycotaBasidiomycotaHolozoaArthropodaCercozoaRadiolariaBicosoecidaMAST−1MAST−3MAST−7MAST−8EuryarchaeotaActinobacteriaBacteroidetesCyanobacteriaCandidate divisionProteobacteriaαδγβPlanctomycetesVerrucomicrobiaDeferribacteresLentisphaeraeNitrospiraeFirmicutesArchaeplastidaAlveolataOpisthokontaStramenopilesRhizariaExcavataArchaeaBacteriaEukaryota48161Total number of OTUsMPP PPSFigure 3.5: Indicator OTUs associated with specific filtration methods. Indicator OTUs for filter groupsPPS in situ 0.4 µm (red), and MPP on-ship 0.4 µm (green) and 0.22 µm pre-filtered onto 0.4 µm filters(yellow). The size of dots depicts the total number of indicator OTUs affiliated to specific taxa.ples with in-line 0.4 µm pre-filter were mostly affiliated with bacterial phyla including Deferribacteres andBacteroidetes, and archaeal phyla including Euryarchaeota (Fig. 3.5).The differences observed between in situ and on-ship indicator OTUs reinforce the effect of size frac-tionation on microbial community structure and raise important questions about metabolic reconstructionefforts based solely on on-ship filtration methods.303.3.5 Size-fractionation effects on expressed OTUs within specific populations (rRNAanalyses)To further evaluate the impact of size fractionation on detection of active microbial groups I compared SSUrRNA:rDNA ratios of OTUs between 0.4 µm filters collected and preserved in situ vs. on-ship 0.22 µmfilters with in-line 0.4 µm pre-filters. I focused on OTUs exhibiting ratios>1 as a proxy for cellular activity(Blazewicz et al., 2013). Ratios for Candidate divisions, Desulfobacterales, SUP05, Phycisphaerae, andHalobacteria were highest in 0.4 µm in situ filter samples while SAR11, Rhodospirillales, Methylophilalesand Burkholderiales within Betaproteobacteria, and Verrucomicrobia and eukaryotic phyla affiliated withAlveolata, Opisthokonta, Rhizaria and Stramenopiles were highest in on-ship 0.22 µm filter samples within-line 0.4 µm pre-filtration (Fig. A.3).I detected OTUs affiliated with SUP05, Marine Group A, SAR11 and SAR324 that showed ratios >1,with differential expression between in situ and on-ship filters. For instance, a total of 4 SUP05 OTUs withratios ranging from 1-2, and 3 SAR324 OTUs with ratios equal to 3 were exclusively detected in 0.4 µm insitu filter samples (Fig. 3.6). I also observed 6 Marine Group A OTUs with ratios equal to 2 in on-ship 0.22µm filter samples with in-line 0.4 µm pre-filtration and 1 exclusively active in situ OTU (Fig. 3.6).Interestingly, I observed 8 SAR11 OTUs with ratios ranging from 1-3 exclusively in on-ship 0.22 µmfilter samples with in-line 0.4 µm pre-filtration (Fig. 3.6). Candidate divisions BCR1 and WS3, Delta-and Gammaproteobacteria, and Planctomycetes OTUs also manifested higher ratios in 0.4 µm in situ filtersamples than in on-ship samples (Fig. 3.7). These differences showed some depth specificity. For example,I observed BCR1 and WS3 OTUs with high ratio values at 165 m while Desulfovibrionales and Desulfarcu-lales within the Deltaproteobacteria had the highest ratio values at 185 m (Fig. 3.7). Similarly, most OTUsaffiliated with Planctomycetes (Phycisphaerae, OM190, Brocadiales and Plactomycetales) had the highestratio values at 185 m (Fig. 3.7). In contrast, Flavobacteriales within Bacteroidetes and Alphaproteobacteriahad higher ratio values in on-ship 0.22 µm filter samples with in-line 0.4 µm pre-filtration than 0.4 µm insitu sampled at both 165 and 185 m (Fig. 3.7).3.4 DiscussionIn this chapter I used SSU rDNA and rRNA count data generated during the SCOR Working Group 144Microbial Community Responses to Ocean Deoxygenation workshop to determine the effects of collectionand filtration methods on microbial community structure and potential activity in the anoxic water columnof Saanich Inlet. Observed differences in microbial community structure and potential activity associatedwith in situ versus on-ship size fractionation suggest potential sources of error when linking field processesto microbial agents based on genomic sequence information in isolation. In particular, in situ results de-tected several microbial groups implicated in the sulfur-cycle that are underrepresented in publicly availableamplicon and shotgun sequencing data sets (Hawley et al., 2017). Overall, results from this study provideuseful information on how different sampling methods can contribute to bias in experimental outcomes andreinforce the need for more integrated studies based on standardized sampling protocols that increasinglyincorporate in situ measurements.Microbial community shifts associated with size fractionation31Marine Group ASAR11SAR324SUP05rRNA:rDNAPPS MPP123250mL 2.5L 2.5L500mLFigure 3.6: Activity differences for abundant OTUs in relation to filtration methods. Small subunit(SSU) rRNA: rDNA ratio for abundant and ubiquitous taxa in OMZs i.e SUP05, SAR406, SAR11 andSAR324, observed at PPS in situ 0.4 µm (red) and MPP on-ship 0.22 µm pre-filtered onto 0.4 µm filters(yellow). The size of dots depicts ratio values for individual OTUs as indicated on plot size key.32BacteroidetesCandidate divisionProteobacteriaPlanctomycetesMaxillopodaαδγ250mL 2.5L 2.5L500mL102030FlavobacterialesBRC1WS3RhodobacteralesRhodospirillalesRickettsialesBdellovibrionalesDesulfarculalesDesulfuromonadalesOceanospirillalesLegionellalesXanthomonadalesVibrionalesPhycisphaeraeOM190BrocadialesPlanctomycetalesOpisthokonta1RNA:DNAA BPPS MPPFigure 3.7: Activity differences for indicator OTUs in relation to filtration methods. Small subunit(SSU) rRNA: rDNA ratio for indicator OTUs observed at PPS in situ 0.4 µm (red) and MPP on-ship 0.22µm pre-filtered onto 0.4 µm filters (yellow). The size of dots depicts ratio values for individual OTUs asindicated on plot size key.33Understanding how microorganisms interact within the ocean at different scales is integral to linkingmicrobial food webs to nutrient and energy flow processes (Azam and Malfatti, 2007). Particles play asalient role in structuring microbial community interactions and the interplay between “particle-associated”and “free-living” microbiota creates a dynamic metabolic network driving biogeochemical transformations(Smith et al.,et al., 1992;DeLong et al., 1993;Crump et al., 1999;Simon et al., 2002;Grossart, 2010;Ganeshet al., 2014). Previous observations from OMZ waters implicate particle maxima as hotspots for metaboliccoupling (Garfield et al., 1983;Naqvi et al., 1993;Whitmire et al., 2009;Ganesh et al., 2014). However,the definition of “particle-associated” versus “free-living” can sometimes seem arbitrary and the degree towhich microorganisms alternate between these two fractions in different water compartments is not firmlyestablished. Typically, anything >0.4 µm has been considered particle associated (Azam and Malfatti,2007) although most studies use a 0.2-1.6 or 2.7 µm cut-off to concentrate microbial biomass. In the currentchapter I compared 0.4 µm in situ filtration without a pre-filtration step to on-ship 0.22 µm filtration within-line 0.4 µm pre-filtration in order to evaluate “particle-associated” versus “free-living” fractions.The core microbial community detected in situ versus on-ship was similar to time-series observationsin suboxic-anoxic water column compartments (165-185 m) during summer months. However, at a moregranular OTU level important differences in community structure and potential activity for a number ofmicrobial groups were resolved that reinforce and expand on previous size fractionation studies in openocean OMZs. For example, Padilla and colleagues have shown that mode and magnitude of sampling biasdepends on filter type and pore size, particle load, and community complexity (Padilla et al., 2015). In thepresent chapter, sample volume had a non-significant effect (p >0.05) on microbial diversity although wiretime and filtration duration likely impacted particle size, as did potential bottle effects due to settling whenprocessing on-ship samples (Fig. A.2). Several studies have shown that particles can settle in samplingbottles on timescales relevant to on-ship processing (Gardner, 1977;Suter et al., 2016).With respect to size fractionation, community structure differences were driven by shifts in abundanceand activity of many known microorganisms. For example, as observed in the ETNP OMZ, OTUs af-filiated with Deferribacteres were enriched in the smaller size fraction (<0.4 µm) consistent with an au-totrophic lifestyle (Ganesh et al., 2014). Similarly, indicator OTUs affiliated with Bacteroidetes (Bac-teroidales and Flavobacteriales), Lentisphareae, Deltaproteobacteria (Myxococcales and Desulfobacterales),Planctomycetes, and Verrucomicrobia were enriched in the larger size fraction (>0.4 µm) as observed inboth ETNP and ETSP OMZs, consistent with attachment to sinking aggregates or zooplankton (Crumpet al., 1999;Simon et al., 2002;Eloe et al., 2011;Allen et al., 2012;Fuchsman et al., 2012;Ganesh et al.,2014;Padilla et al., 2015). These similarities transcended domain boundaries with eukaryotic phyla includ-ing Dinoflagellata (Alveolata), Radiolaria (Rhizaria) and Syndiniales (Stramenopiles) enriched in the 0.4µm filter samples (Guillou et al., 2008;Duret et al., 2015).With respect to in situ versus bottle collection methods, previous studies have identified changes incommunity gene expression profiles (Feike et al., 2012;Stewart et al., 2012a) and process rate measure-ments (Taylor and Doherty, 1990;Stewart et al., 2012a;Taylor et al., 2015;Edgcomb et al., 2016). Here Iidentified changes in microbial community structure that potentially explain variance in gene expression orprocess rates. For example, indicator OTUs detected in in situ samples indicate a previously unrecognized34role for sulfate-reducers and candidate divisions WS3, OD1 and BRC1 in the Saanich Inlet water column.Previous studies have implicated WS3 and OD1 metabolism in sulfur cycling and methanogen provisioning(Kirkpatrick et al., 2006;Wrighton et al., 2012). Similarly, Desulfobacterales and Desulfovibrionales, arecommonly underestimated in abundance in OMZs waters (Suter et al., 2016) as they are more prevalent onparticles that likely settle during on-ship processing.Working in the Cariaco Basin OMZ, Suter and colleagues provide a compelling description of bottlesettling rates (approximately 12 min for a 1mm particle to sink into below-spout space of an 8 L Niskin bottleor 18 min for a 12 L bottle) that can result in sampling bias (Suter et al., 2016). On-ship processing duringthe SCOR workshop took between 20-30 minutes. Shipboard processing times can often be even longerthan this. In addition, wire-time and turbulence associated with sampling moment and vibration (Suter etal., 2016), and filtration across the membrane (Duret et al., 2015) can impact particle size and stability inbottles e.g. production of smaller particles derived from larger aggregates. Based on sinking rates estimatedin the Cariaco Basin it is possible that observed community structure differences between in situ and on-shipsamples in Saanich Inlet could be explained by a combination of particle settling and turbulence associatedwith sample collection and filtration. This could affect our perception on the microbial metabolic networkwith respect to transient spatial interactions that are altered or disrupted using on-ship methods.Implications of size fractionation for inferring microbial activity in OMZsIn contrast to only examining rDNA sequences, combining those with analysis of rRNA sequences canprovide a robust proxy for past, and present or emerging cellular activities (Blazewicz et al., 2013) that caninform hypotheses related to life strategies and metabolic interactions within microbial communities (Leppand Schmidt, 1998;Barnard et al., 2013). Here, I considered rRNA:rDNA ratio values>1 as an indicator forpotentially active microbial community members. While, using ratios to infer activity at higher taxonomiclevels e.g. Phylum, Class, Order, can promote inconsistent results (Blazewicz et al., 2013;Ganesh et al.,2015), focusing on the OTU level can identify ecologically relevant populations with the potential to playintegral roles in nutrient and energy cycling within the ecosystem under study. For example, SUP05 hasbeen determined to be an abundant member of the Saanich Inlet microbial community comprising between20-30% of total bacteria at 165 and 185 m, respectively (Walsh et al., 2009; Zaikova et al., 2010; Walsh andHallam, 2011). I detected 85 OTUs affiliated with SUP05 based on rDNA sequences. However, only 4 hadrRNA:rDNA ratio values >1 indicating population level variation in potential activity.Consistent with previous observations in the ETSP using metatranscriptomic data (Padilla et al., 2015),examination of the taxonomic affiliation of indicator OTUs produced using in situ versus on-ship methodsidentified differences between the active microbial community in samples. Some candidate divisions recov-ered in in situ samples have not been previously well described in the Saanich Inlet water column basedon rDNA sequences due to their low abundance (<0.1%). Interestingly, the rRNA:rDNA ratios observedfor indicator WS3 and BCR1 OTUs (ratios equal to 7 and 2, respectively) were greater than those observedfor OTUs affiliated with ubiquitous and abundant taxa, including SUP05. Similar observations were madefor OTUs affiliated with Deltaproteobacteria, Chloroflexi, Firmicutes, Lentisphaera, Nitrospina and MarineGroup A, reinforcing the idea that multi-omic sequences and process rate measurements sets sourced fromon-ship samples have the potential to underestimate the contribution of some active microbial groups present35in the water column. Such groups may be sensitive to settling, turbulence or other factors including oxygenexposure, necessitating in situ sampling to reveal their contributions to the metabolic network.Hawley and colleagues used metaproteomics to develop a conceptual model of coupled carbon, nitrogenand sulfur cycling (Hawley et al., 2014). Louca and colleagues incorporated these ideas into a numericalmodel integrating multi-omic sequence and geochemical information to predict metabolic fluxes and growthyields under near steady-state conditions (Louca et al., 2016). Both conceptual and numerical models werebased on interactions between Thaurmarchaeota, SAR11, SUP05 and Planctomycetes (Walsh et al., 2009;Zaikova et al., 2010; Walsh and Hallam, 2011; Wright et al., 2012). Although, the metabolic potentialof WS3, OD1 and sulfate-reducing Deltaproteobacteria in the Saanich Inlet water column remains to bedetermined, the potential role of these groups at the nexus of sulfur cycling and methanogenesis (Kirkpatricket al., 2006) presents an opportunity for new hypothesis development and testing to integrate these groupsinto prevailing conceptual and numerical models for coupled biogeochemical cycling.3.4.1 ConclusionAs the research community transitions away from descriptive studies of marine microorganisms to morequantitative comparisons at ecosystem scales integrating multi-omic information with process rates andmodeling, the need for standards of practice that reduce sampling bias becomes increasingly important.Moreover, a general lack of consensus related to assignments to particle-associate versus free-living sizefractions further confounds robust cross-scale comparisons. Results from this chapter suggest that in situsampling approaches have the potential to limit many biases by providing a more authentic representation ofmicrobial activity than on-ship sampling methods. At the same time, consistent on-ship methods need to beestablished that limit bottle effects and harmonize filtration practices for effective cross-scale comparisons.Several promising devices such as the PPS (Edgcomb et al., 2016), Environmental Sample Processor (ESP)(Jones et al., 2008;Preston et al., 2009;Ottesen et al., 2011;Robidart et al., 2014), Automatic Flow InjectionSampler (AFIS) (Feike et al., 2012), and Clio (Jakuba et al., 2014) have been developed with the potential tosupport in situ sampling under a variety of operational scenarios. For example, recent studies with the ESPhave enabled dynamic intermittent sampling during light dark cycles in surface waters revealing conservedpatters of gene expression on ocean basin scales (Ottesen et al., 2014;Aylward et al., 2015). Althoughcommunity adoption of these new technologies remains in early stages due in part to accessibility, pricepoint, and operating constraints, these devices and their “descendents” likely reflect the future of microbialsampling in the ocean given their autonomous and programmable designs extensible to time series or eventresponse monitoring. Based on the information provided, I recommend a replicated study of different in situsampling technologies that incorporates multi-omic sequencing and measurements of process rates focusedon coupled carbon, nitrogen and sulfur cycling in coastal and open ocean OMZs.36Chapter 4Protistan parasites along water columnoxygen gradients: a network approach toassessing potential host-parasiteinteractions1This chapter introduce oxygen (O2) effects on ecological interactions and energy flow in aquatic ecosystems.Using methodologies from Chapter 3 for small subunit ribosomal (SSU) DNA and RNA tags I establishedeukaryotic community structure and interactions highlighting the potential impact parasitic interactions mayhave on population dynamics, extensible to nutrient cycling process, during seasonal water column stratifica-tion in Saanich Inlet. In addition, I explored correlation and network analyses, and visualization tools usingtime-series tag data that form the basis for community-level interactions analyses carried out in Chapters5 and 6. This chapter is the basis for interpreting microbial community structure, composition and inter-actions that serve as framework for constructing microbial community networks with potential metabolicimplications for nutrient cycling throughout water column O2 compartments.4.1 IntroductionAs oxygen (O2) levels decline, energy in oxygen minimum zones (OMZs) is increasingly diverted intomicrobial community metabolism with resulting feedback on metazoan organisms through habitat com-pression. This has a disproportionate impact on large multicellular eukaryotes that cannot permanentlyoccupy anoxic OMZ cores (Parris et al., 2014). Despite this tendency, some vertically-migrating crustacea,zooplankton, chaetognaths and fish are known to seek temporary refuge from predation in OMZ waters(Wishner et al., 1998; Escribano et al., 2009; Wishner et al., 2013). Moreover, a diverse community ofmicroeukaryotes and zooplankton can operate in OMZ waters where they help shape biogeochemistry and1A version of this chapter appears as Torres-Beltra´n, M., Sehein, T. et al. 2018. Protistan parasites along oxygen gra-dients in a seasonally anoxic fjord: a network approach to assessing potential host-parasite interactions. Deep Sea Res. II.doi.org/10.1016/j.dsr2.2017.12.02637ecology. Protists in particular, participate in wide-ranging interactions that couple metabolic processes andbehaviour to nutrient and energy flow patterns in the environment (Arstegui et al., 2009).Throughout marine water columns and sediments, protists shape pools of bioavailable carbon and othernutrients through production, grazing and symbiotic associations. Phagotrophic protists and viruses areconsidered as the main sources of mortality for marine organisms (Suttle, 2005; Arstegui et al., 2009).Organic carbon in the dissolved fraction is transferred to higher trophic levels (to Metazoa) in a heterotrophicfood chain involving bacteria, small and large nanoflagellates (2-20 µm), and larger flagellates and ciliates,termed the “microbial loop” (Azam et al., 1983; Taylor GT, 1986). The relative contributions of protistgrazing and viral lysis as top-down controls on marine bacterioplankton are still debated (e.g. (Pedrs-Ali etal., 2000; Cuevas and Morales, 2006; Chow et al., 2014), and undoubtedly vary depending on site-specificphysico-chemical conditions and the physiological state of individual populations. Diverse communities ofphagotrophic and parasitic protists have been described under low oxygen conditions, many of which exhibitputative symbiotic relationships with prokaryotes, and protist grazing has been shown to shape specific preypopulations along the redoxcline (Lin et al., 2007)(Edgcomb et al., 2011; Orsi et al., 2011, 2012; Wright etal., 2012; Parris et al. 2014; Jing et al. 2015).Parasitic protists are taxonomically diverse and a major source of mortality for other microbial eukary-otes as well as Metazoa (Chambouvet et al., 2008), likely making them important shapers of food webstructure (Jephcott et al., 2016). The Alveolata comprise one of the largest eukaryotic lineages, which in-cludes four protist groups, the Ciliophora, Dinoflagellata, Protalveolata, and Apicomplexa, the last two ofwhich are comprised primarily of parasites (Edgcomb, 2016). Exclusive marine parasitoid protists con-sistently recovered in high-throughput molecular studies were initially assigned to novel marine alveolate(MALV) Groups I and II within the Syndiniales (Lopez-Garcia et al., 2001; Moon-van der Staay et al.,2001; Massana et al., 2004b; Romari and Vaulot, 2004; Not et al., 2007; Edgcomb et al., 2011). GroupI consists of numerous undescribed species while Group II contains sequences affiliated with the genusAmoebophrya, including species influencing “red tide” blooms (Chambouvet et al., 2008). Over time theoriginal two groups have expanded to include five (Guillou et al., 2008) or eight (Richards and Bass, 2005)deeply diverging and abundant clades. For example, in euphotic zone samples from the Tara Oceans expedi-tion 36% of protistan operational taxonomic units (OTUs) were attributed to parasites, almost half of whichwere affiliated with a single MALV Group I clade (de Vargas et al., 2015)(de Vargas et al. 2015). Molecularsignatures of Syndiniales are almost universally retrieved from picoplankton (<3 µm) samples, althoughsome clades within these groups appear to be restricted to specific habitats (Guillou et al., 2008). Variationsin sequence abundances in ribosomal RNA- vs. DNA-based libraries, due to PCR amplification biases bytaxon-specific rDNA copy number, may indicate variability in the impact of parasitism in different habitats(Not et al., 2009).Infection of host cells by members of Syndiniales typically leads to host mortality. Known hosts in-clude other Syndiniales, photosynthetic dinoflagellates, ciliates, rhizarians and metazoans (Guillou et al.,2008; Brte et al., 2012; Massana et al., 2014). The most studied syndinian parasites belong to the speciescomplex, Amoebophrya. Amoebophrya infect new hosts by forming motile spores or weakly motile spores(dinospores) that penetrate the host cell membrane and travel to the host nucleus or cytoplasm to replicate38(Chambouvet et al., 2008). Over the course of several days, the trophont grows and ultimately rupturesthe host cell membrane, releasing the ephemeral, multinucleate, (often) multiflagellate parasite vermiforminto the water column, which subsequently fragments into individual dinospores (Chambouvet et al., 2008).In the case of some syndinian parasites (e.g. infection of tintinnid ciliates by Euduboscquella species), in-fection is thought to be a passive process, whereby the host ingests but does not digest the spores (Dolan,2013). Other syndinian trophonts, such as those formed by Euduboscquella species emerge through a moreelaborate process involving creation of a food vacuole encompassing the remains of the host cell, digestionof host remains, followed by serial nuclear and cytoplasmic divisions to form spores (Coats et al., 2012;Dolan, 2013). Increased syndinian parasite abundance during the decline of harmful algal blooms (i.e.Alexandruim catanella) suggest potential top-down control of phytoplankton species during bloom events(Bai et al., 2007; Velo-Surez et al., 2013; Choi et al., 2017). Studies of parasitism of other taxa suggestsyndinian parasites infect their hosts throughout the year, but studies of ciliates indicate that infections ofciliate hosts are most common in summer months when phytoplankton biomass is apparently sufficient tosupport elevated grazer densities (Cachon, 1964; Coats, 1989; Coats et al., 1994). Thus, Syndiniales mayhave the potential to contribute significantly to releases of organic carbon through cell lysis during bloomtermination events and may influence protist community diversity by exerting top-down control on selectspecies within a population.Fluorescence in situ hybridization (FISH) analyses has shown that Syndiniales alternate between twolife stages: the free-swimming, infective dinospore that can survive without a host for up to three days,and the multinuclear trophont phase within a host (Velo-Surez et al., 2013). Most infections lead to lysisof the host cell; however, one study of the Syndiniales Group II taxon Amoebophrya, revealed dinosporescould lie dormant within the resting stage of the dinoflagellate, Scrippsiella trochoidea and re-emergingsimultaneously when the host entered the vegetative state, suggesting potentially complex life histories thatcan impact protist populations at multiple life stages (Chambouvet et al., 2011). In molecular surveys ofprotistan diversity, Syndiniales are grouped taxonomically with the novel marine alveolate (MALV) lineagesI and II, and can form the majority of sequences from diverse marine habitats including coastal waters, openoceans, and stratified systems (Massana et al., 2004a; Not et al., 2007; Guillou et al., 2008). These parasitesare thought to exhibit top-down pressures on host populations within the eukaryotic community, such as,radiolarians, dinoflagellates, ciliates, and metazoa; however, there is a paucity of data on the range of hostspecies for different Syndiniales groups or on net ecosystem effects of parasitism by these taxa.Saanich Inlet’s seasonally stratified water column serves as a well-characterized model ecosystem forexamining how deoxygenation shapes microbial community population dynamics and interactions alongdefined redox gradients in the ocean. Stratification and deep-water renewal cycles are known to shapemicrobial community composition and activity, and to lead to seasonal blooms of not only various phyto-plankton (some of which are toxic species) but also metazoan species that serve as prey for commerciallyimportant fisheries in the area. A previous study indicated that protist populations occupy different nichesalong the redoxcline over a 12-month cycle including abundant Dinoflagellata OTUs (Orsi et al. 2012). Inoxic waters at 10m depth, Syndiniales OTUs represented ∼90% of all Dinoflagellata-affiliated sequencesin samples from all seasons (Orsi et al. 2012). At 200m depth an increase in Dinoflagellata OTUs (80%39affiliated with Syndiniales) was observed during deep water renewal periods (Orsi et al. 2012). While thismay reflect active infection of a deep-water bloom of non-phototrophic hosts, this may alternatively reflectdinospores released from sinking infected and lysed phototrophic hosts originating from the upper watercolumn as a responses to seasonal blooms.Here I explore eukaryotic small subunit ribosomal (SSU) RNA- and DNA-based pyrotag datasets gen-erated from monthly filtered samples collected at depths throughout the Saanich Inlet water column over a12-month period to elucidate changes in microbial eukaryote populations and eukaryote-eukaryote interac-tions that have the potential to influence OMZ biogeochemistry and ecology. I focus on population dynamicsof Syndiniales OTUs to better constrain their impact on organic matter release and bloom termination duringseasonal stratification and renewal.4.2 Methods4.2.1 Environmental samplingEnvironmental monitoring and sample collection were carried out monthly aboard the MSV John Stricklandat station SI03 (48o 35.500 N, 123o 30.300 W) in Saanich Inlet, B.C. From May 2008 to April 2009 and July2014 a total of 175 water samples were collected at 16 high-resolution depths (10, 20, 40, 60, 75, 80, 90, 97,100, 110, 120, 135, 150, 165, 185 and 200 m) spanning oxic (>90 µmol O2 kg-1), dysoxic (90-20µmol O2kg-1), suboxic (20-1 µmol O2 kg-1) anoxic (<1 µmol O2 kg-1) and sulfidic water column conditions (Wrightet al., 2012). Samples were processed and analyzed as previously described (Zaikova et al., 2010; Hawleyet al., 2017b; Torres-Beltra´n et al., 2017). Briefly, water samples were collected from Niskin or Go-Flowbottles for DNA and RNA (Hawley et al., 2017b). Conductivity, temperature, and depth were measuredusing a Seabird SBE19 CTD-device (Sea-Bird Electronics Inc., Bellevue USA), with a PAR and O2 sensorattached. An O2 probe attached to the CTD was also used to measure dissolved O2 throughout the watercolumn as previously described (Torres-Beltra´n et al., 2017).4.2.2 Nucleic acid sampling and extractionBiomass to generate SSU rDNA pyrotag datasets for microbial community composition profiling (2 L)was filtered directly onto a 0.22 µm Sterivex polycarbonate cartridge filter from high-resolution depthscollected from May 2008 to April 2009, and July 2014. Biomass to generate SSU rRNA pyrotag datasetsfor potentially active microbial community composition profiling analysis (2 L) was collected on July 2014at high-resolution depths and filtered directly onto a 0.22 µm Sterivex polycarbonate cartridge filter within20 minutes of shipboard sample collection.Genomic DNA was extracted from the Sterivex filters as previously described (Zaikova et al., 2010;Hawley et al., 2017b). Briefly, after defrosting Sterivex on ice, 100 µl lysozyme (0.125 mg ml-1; Sigma)and 20 µl of RNAse (1 µl ml-1; ThermoFisher) were added and incubated at 37oC for 1 h with rotationfollowed by addition of 50 µl Proteinase K (Sigma) and 100 µl 20% SDS and incubated at 55 oC for 2 hwith rotation. Lysate was removed by pushing through with a syringe into 15 mL falcon tube (Corning) and40with an additional rinse of 1 mL of lysis buffer. Filtrate was subject to chloroform extraction (Sigma) andthe aqueous layer was collected and loaded onto a 10K 15 ml Amicon filter cartridge (Millipore), washedthree times with TE buffer (pH 8.0) and concentrated to a final volume of between 150-400 µl. TotalDNA concentration was determined by PicoGreen assay (Life Technologies) and genomic DNA qualitydetermined by visualization on 0.8% agarose gel (overnight at 16V).Total RNA was extracted from Sterivex filters using the mirVana Isolation kit (Ambion) (Shi et al., 2009;Stewart et al., 2010) protocol modified for sterivex filters (Hawley et al., 2017b). Briefly, after thawing thefilter cartridge on ice RNA later was removed by pushing through with a 3 ml syringe followed by rinsingwith an additional 1.8 mL of Ringer’s solution and incubated at room temperature for 20min with rotation.Ringer’s solution was evacuated with a 3 ml syringe followed by addition of 100 µl of 0.125 mg ml-1lysozyme and incubated at 37 oC for 30 min with rotation. Lysate was removed from the filter cartridgeand subjected to organic extraction following the mirVana kit protocol. DNA removal and clean up andpurification of total RNA were conducted following the TURBO DNA-free kit (ThermoFisher) and theRNeasy MinElute Cleanup kit (Qiagen) protocols respectively. Total RNA concentration was determinedby RiboGreen analysis (Life Technologies) prior to synthesize first strand cDNA using the SuperScript IIIFirst-Strand Synthesis System for RT-qPCR (Invitrogen) according to manufacturer instructions.4.2.3 Small subunit ribosomal RNA and RNA gene sequencing and analysisAll pyrotag libraries were generated by PCR amplification using multi-domain primers targeting the V6-V8region of the SSU rRNA gene (Allers et al., 2013): 926F (5’-cct atc ccc tgt gtg cct tgg cag tct cag AAACTY AAA KGA ATT GRC GG-3’) and 1392R (5’-cca tct cat ccc tgc gtg tct ccg act cag-<XXXXX>-ACG GGC GGT GTG TRC-3’). Primer sequences were modified by the addition of 454 A or B adaptersequences (lower case). In addition, the reverse primer included a 5 bp barcode designated <XXXXX>for multiplexing of samples during sequencing. Twenty-five microliter PCR reactions were performed intriplicate and pooled to minimize PCR bias. Each reaction contained between 1 and 10 ng of target DNA,0.5 µl Taq DNA polymerase (Bioshop inc. Canada), 2.5 µL Bioshop 10 x buffer, 1.5 uL 25 mM BioshopMgCl2, 2.5 µL 10 mM dNTPs (Agilent Technologies) and 0.5 µL 10 mM of each primer. The thermalcycler protocol started with an initial denaturation at 95 oC for 3 minutes, followed by 25 cycles of 30s at 95 oC, 45 s at 55 oC, 90 s at 72 oC and 45 s at 55 oC, and a final extension at 72 oC for 10 min.PCR products were purified using the QiaQuick PCR purification kit (Qiagen), eluted elution buffer (25µL), and quantified using PicoGreen assay (Life Technologies). SSU rRNA and SSU rRNA gene (rDNA)amplicons were pooled at 100 ng for each sample. Emulsion PCR and sequencing of the PCR ampliconswere sequenced on Roche 454 GS FLX Titanium at the McGill University and Ge´nome Que´bec InnovationCenter.Pyrotag sequences from May 2008 to April 2009 were processed using the Quantitative Insights IntoMicrobial Ecology (QIIME) software package (Caporaso et al., 2010). To minimize the removal of falsepositive reads all 2,501,489 pyrotag sequences generated from the 159 samples were clustered together.Reads with a length shorter than 200 bases, ambiguous bases, and homopolymer sequences were removedprior to chimera detection. Chimeras were detected and removed using chimera slayer provided in the41QIIME software package. Sequences were then clustered into operational taxonomic units (OTUs) at 97%identity using uclust with average linkage algorithm. Prior to taxonomic assignment singleton OTUs (OTUsrepresented by one read) were omitted, leaving 251,337 protistan OTUs. Representative sequences fromeach non-singleton OTU were queried against the SILVA database (Pruesse et al., 2007) using BLAST(Altschul et al., 1990). Pyrotag sequences from July 2014 were recruited to the OTUs sequences from theMay 2008 to April 2009 dataset using BLAST (Altschul et al., 1990) with an identity threshold higher than98%. Only protistan OTUs showing the highest match (>98% identity) across the reference dataset wereselected to support their potential activity under defined water column conditions.4.2.4 Statistical analysesUnless otherwise indicated all statistical analyses were performed using the R software (RCoreTeam, 2013).Pyrotag datasets were Hellinger transformed (Legendre and Gallagher, 2001) using the vegan package tothe square root of observed total number of reads per sample. Hierarchical cluster analysis (HCA) wasconducted to identify groups associated with discrete water column conditions using the pvclust (Suzuki andShimodaira, 2015) package with Manhattan Distance measures, and statistical significance to the resultingclusters was computed as bootstrap score distributions with 1,000 iterations.Multi-level indicator species analysis (ISA) using the indicspecies package (De Caceres and Legendre,2009) was performed to identify OTUs specifically associated with different water column conditions de-fined by HCA that may be known by their parasite-host interaction. The ISA/multi-level pattern analysiscalculates p-values with Monte Carlo simulations and returns indicator values (IV) and p-values with α<0.05. The IVs range between 0 and 1, where indicator OTUs considered in the present chapter for furthercommunity analysis show an IV >0.6 and p-value <0.001.To generate a robust matrix of significant protistan co-occurrences between prevalent OTUs at watercolumn conditions that may represent parasite-host interactions, correlation coefficients from May 2008to April 2009 pyrotag samples were calculated using the Bray-Curtis and Spearmans rank correlations inthe CoNet software (Faust et al., 2012) with OTU abundance as count data as previously used for pyrotagdata (Torres-Beltra´n et al., 2016a). Prior analysis in CoNet, the OTU matrix derived from pyrotag taxonomicanalysis was transformed into presence-absence data to remove OTUs with less than 1/3 zero counts, leaving556 OTUs for all samples that were selected in the OTU abundance matrix for correlation analysis. Toconstruct ensemble networks, thresholds for Bray-Curtis and Spearmans rank correlations measures wereset to 0.6 correlation value as a pre-filter for OTU pairs, followed by computing edge scores only betweenhighly correlated OTU pairs. To assign statistical significance to the resulting scores, edge and correlationmeasure-specific permutation and bootstrap score distributions with 1,000 iterations each were computed.p-values were tail-adjusted so that low p-values correspond to co-presence and high p-values to exclusion,p-values on each final edge were corrected to q-values (cut-off of 0.05). Finally, only edges with at least twosupporting pieces of evidence, i.e. high correlation value on both correlation measures and q-value belowthreshold were retained. The final edges and nodes matrices were exported using Cytoscape 2.8.3 (Shannonet al., 2003). Nodes corresponded to individual OTUs and edges were defined by computed correlationsbetween corresponding OTU pairs. Edges were selected and exported based on specific paired OTUs such42as those affiliated with parasite-host taxonomic groups.To gain insight into potential parasite-host activity, SSU rRNA:rDNA ratios were calculated as previ-ously described for pyrosequencing data (Frias-Lopez et al., 2008; Stewart et al., 2012b). I selected OTUsfrom the July 2014 dataset based on their taxonomic affiliation (BLAST-based >98% similarity) with taxaof interest. In addition, only OTUs affiliated to indicators showing significant correlations were selected.4.2.5 Data depositionThe SSU rDNA and rRNA pyrotag sequences reported in this chapter have been submitted to the TheNational Center for Biotechnology Information (NCBI) under BioSample numbers: SAMN03387532 -SAMN03387915 and SAMN05392441 - SAMN05392453.4.3 Results4.3.1 Water column conditionsWater column properties were monitored through the progression of stratification and deep-water renewalover a one-year period (May 2008 to April 2009). Throughout the year the water column surface tempera-ture ranged between 7 to 16 oC during winter and summer months respectively, reaching an average of 9 oCat bottom waters. Salinity between surface and bottom waters ranged between 28 to 31 ppm throughout theyear, showing the lowest values at the surface during spring months. In addition, as previously describedfor OMZs (Wright et al., 2012), in this chapter I define water column conditions on the basis of O2 concen-tration ranges: oxic (>90 µM O2), dysoxic (90-20 µM O2), suboxic (20-1 µM O2) and anoxic (<1 µM)(Wright et al., 2012). Between May and August 2008, as water column stratification peaked, suboxic con-ditions intensified corresponding with the development of deep-water anoxia. The concentration of NO3-between the surface and 100m ranged between 5 to 20 µM, decreasing rapidly between 100 and 135m be-fore reaching a minimum of <1 µM in anoxic bottom waters (Fig. 4.1). The beginning of 2008 deep-waterrenewal occurred toward the end of September and continued through November. Initially, dissolved O2 wasobserved throughout the water column, although upwards shoaling of O2 and NO3- depleted bottom watersproduced an intermediate suboxic layer between 100 and 135 m. By October, dissolved O2 concentrationsbetween 150 and 200 m ranged between 14 and 28 µM consistent with complete oxygenation of the watercolumn. Over the same time interval, NO3- concentrations between surface and 100m ranged between 4to 29 µM, decreasing rapidly below 100 m before reaching a second maximum ranging between 21 to 24µM at 200 m (Fig. 4.1). In November, O2 and NO3- concentrations continued to increase above 100 andbelow 135 m with intervening depth intervals experiencing increased oxygen decline. In December watercolumn O2 deficiency intensified below 100 m as the water column become increasingly stratified throughApril 2009.43NO3O2SalinityTemperatureMay June JulyAugust September October10 20 30 10 20 30 10 20 3010 20 30 10 20 30 10 20 301010020010100200150 25050O2150 25050O2150 25050O2150 25050O2150 25050O2150 25050O2Figure 4.1: Water column chemical parameters. CTD and chemical data shown as monthly panels fortemperature (oC), Salinity (ppm), Oxygen (µM) and nitrate (µM) along the depth profile for samples takenfrom May to September 2008 at Station S3 in Saanich Inlet.444.3.2 Eukaryotic community structureTo explore eukaryotic community structure in the Saanich Inlet water column I analyzed SSU rDNA genepyrotag sequences from 159 samples collected at 16 depths ranging between 10 and 200 m over the timeinterval between May 2008 and April 2009. Overall, the opisthokont community was dominated by OTUsaffiliated with metazoan sequences within the phylum Arthropoda (34% total community) (Fig 4.2A). Protis-tan community composition was dominated by OTUs affiliated with taxonomic groups previously identifiedin the Saanich Inlet water column, using different identification approaches (Orsi et al., 2012) includingthe Alveolata (41%), Stramenopiles (10%), Hacrobia (3%) and Rhizaria (2%)(Fig 2A). Nineteen percentof Alveolata OTUs were affiliated with Dinoflagellata sequences within Syndiniales groups (I, II, III andV) (Fig 4.2B). Groups I (27%) and II (66%) dominated throughout the year showing peaks of abundanceduring the stratification period (May- August) in the inlet (Fig 4.3B) reaching up to 19% and 43% of relativeabundance, respectively, from the total protist sequences.To describe protistan community partitioning along water column O2 gradients I conducted hierarchicalcluster analysis using rDNA pyrotag profiles from each sample. Results revealed four major groups or clus-ters (AU>70, 1,000 iterations) associated with oxic (group I), dysoxic-suboxic (group II and III), and anoxic(group IV) water column conditions that were previously observed to occur with seasonal stratification anddeep water renewal events (Orsi et al., 2012) (Fig 4.3A). To further resolve OTUs occurring under specificwater column O2 conditions and to reveal protistan co-occurrence patterns between taxa i.e. Syndinialesand potential hosts, multi-level indicator species analysis (ISA) was conducted based on HCA groups (Ta-ble B.1-B.4). Indicator OTUs affiliated with Arthropoda, Cnidaria, Dinoflagellata and Stramenopiles werecharacteristic of the oxic water column compartment while indicator OTUs affiliated with Choanoflagel-latea, Colpodea, Cryptophyceae, Dinoflagellata, Picozoa, Prymnesiophyceae, Stramenopiles and Telonemiawere characteristic of the dysoxic-suboxic water column conditions (Fig 4.3B). Syndiniales OTUs were themost abundant indicator OTUs found in these water column conditions during peak stratification in sum-mer (56% of total indicator OTUs) (Fig 4.3C). In addition, OTUs affiliated with potential hosts such asthe ciliate subclass Choreotrichia and the metazoan class Maxillopoda, were also found as indicators for thedysoxic-suboxic water column conditions along with the Syndiniales OTUs. In comparison, indicator OTUsaffiliated with Dinoflagellata, Stramenopiles and Arthropoda were identified for the anoxic water columncompartment during the stratification period.4.3.3 Exploring protistan co-occurrence patternsTo explore community co-occurrence patterns focusing primarily on potential parasitic interactions betweenSyndiniales and protists throughout water column O2 gradients, I conducted a co-occurrence analysis basedon Bray-Curtis and Spearman correlations among OTUs. Resulting significant (Correlation Value >0.6, p<0.001) pairs were used to determine potential interactions i.e. parasitism dynamics over water columnstratification period. Correlation analysis resulted in a total of 6,273 significant pairs corresponding to co-occurrences between prevalent OTUs throughout water column conditions over time. Significant pairs wereobserved among different taxonomic groups suggesting a broad range of potential interactions resulting from325 unique source OTUs affiliated with Alveolata (3,765), Apusozoa (3), Archeaplastida (139), Excavata45SyndinialesGroup−V Group−IV Group−III Group−II Group−I0 20 40Relative abundance (%)A BDinophyceae CiliophoraOther Alveolata ArthropodaCnidariaAnnelidaOther MetazoaChoanoflagellidaStramenopilesBacillariophytaHacrobiaRhizariaOtherSyndinialesOpisthokontaMetazoaAlveolataDinophytaFigure 4.2: Taxonomic breakdown of eukaryotic OTUs. A) Krona chart showing the taxonomic com-position of eukaryotic OTUs found in the rDNA pyrotag datasets from May 2008 to April 2009. Layersrepresent hierarchical taxonomy from the upper (centre) to the lowest (outer) taxonomic level. B) Relativeabundance for Syndiniales OTUs divided by taxonomic groups found in pyrotag dataset. The size of eachbox represents the average of relative abundance (%) calculated from the total number of eukaryotic readsthroughout the water column over this period. Extended dashed lines (whiskers) represent at the base thelower and upper quartiles (25% and 75%) and at the end the minimum and maximum values encountered.The middle line represents the median.(250), Hacrobia (483), Opisthokonta (418), Rhizaria (224), and Stramenopiles (986).Syndiniales correlations corresponded to 63% of the total pairs observed, showing co-occurrence withOTUs affiliated with Alveolata (including Apicomplexa, Ciliophora, Dinoflagellata), Chlorophyta, Metazoaand a variety of Stramenopiles including MAST groups (Fig 4.4; Appendix B). The relative proportion oftaxa correlating with Syndiniales OTUs was similar between groups; however, the number of unique OTUsassociated with the different Syndiniales groups varied significantly. For instance, group II was the mosttaxonomically diverse assemblage, exhibited the most correlations with other protistan OTUs, while GroupV exhibited the fewest (Fig 4.5A). The occurrence of OTUs affiliated with amoebae, green algae, hapto-phytes, and Metazoa was exclusively statistically correlated with Syndiniales groups I and II (Fig 4.5B).Interestingly, significant positive correlations between Syndiniales and OTUs affiliated with taxa describedas potential hosts i.e. Choreotrichia, Phaeocystis sp. and Maxillopoda, were observed in different oxygenregimes. For instance, while Syndiniales (groups I-V) and Choreotrichia OTUs co-occurred throughout thewater column, Syndiniales groups II, III and V, and Chroreotrichia OTUs co-occurred mostly in dysoxic-suboxic (90-150 m) and anoxic (below 150 m) water column conditions as water column stratification inten-sified during summer months (Fig 4.6). Syndiniales-Choreotrichia co-occurrence patterns were supportedby ISA results showing these OTUs to be characteristic of dysoxic-suboxic water column conditions during4602040608592818781TrebouxiophyceaeTelonemiaSyndinialesStramenopiles−Group−9StramenopilesSpirotricheaPrymnesiophyceaePicozoaOligohymenophoreaMASTMamiellophyceaeLabyrinthuleaKatablepharidaceaeFilosa−ImbricateaFilosa−ChlorarachneaEuglenozoaEllobiopsidaeDinophytaDinophyceaeDictyochophyceaeCryptophyta−nucleomorphCryptophyceaeColpodeaChytridiomycotaBacillariophytaArthropodaApicomplexa0 >2010Number of indicator OTUsIIIIIIIVA BOxygen (μM)>250<3CSpring Fall (Renewal)Distance38%62% 56%44% 52%48%Figure 4.3: Eukaryotic community structure. A) Hierarchical clustering of eukaryotic pyrotag data dur-ing peak stratification based on Manhattan distance. Clusters are delimited by O2 concentration range rep-resented by number from I to IV: oxic = I (red), dysoxic- suboxic = II and III (green and blue, respectively),and anoxic =IV (purple). Bootstrap values (1000 iterations) are shown in gray. B) Indicator OTUs (Indicatorvalue 0.6, p =0.05, α < 0.01) for suboxic-dysoxic water column conditions during peak stratification. Barsdepict the total number of OTUs for each taxonomic group observed. Total Syndiniales indicator OTUs(n=54) are shown in dark gray. C) Distribution of Syndiniales indicator OTUs over three different watercolumn periods in the inlet (spring, summer stratification and fall deep-water renewal). Pie charts depict thepercentage of Syndiniales OTUs (dark gray) out of the total number of indicator OTUs for each period.the same time intervals (Table B.1). In addition, Syndiniales groups I and II, and Phaeocystis sp. OTUsco-occurred in oxic waters in May and become progressively associated with dysoxic-suboxic waters as thewater column became increasingly stratified in July and August (Fig 4.7). Syndiniales groups I and II, andMaxillopoda. OTUs exhibited a similar temporal co-occurrence pattern (Fig 4.8). Syndiniales-Maxillopodaco-occurrence patterns were supported by ISA results indicating that these OTUs were characteristic of oxicand dysoxic water column conditions during the same time intervals (Table B.1). Moreover, I observedthat OTUs affiliated with Syndiniales and the three target taxa were evenly distributed throughout the wa-ter column in September and October possibly reflecting vertical transport and mixing during deep waterrenewal.47OpisthokontaMetazoaArthropodaRhizariaStramenopilesDinophytaSyndinialesOther EukaryotaFigure 4.4: Co-occurrence network on SSU rDNA pyrotag protist data from May-August 2008. Co-occurrence network derived from Bray-Curtis and Spearman correlation measures on rDNA pyrotag protistdata from May-August 2008. Nodes depict OTUs and edges co-occurrence correlations. Syndiniales OTUsare highlighted in gray and other protist OTUs coloured as indicated in color key.4.3.4 Insight into potential Syndiniales parasitic interactionsGiven the observed co-occurrence patterns between rDNA of Syndiniales and Choreotrichia, Phaeocystissp. and Maxillopoda during the 2008 stratification period, I wanted to evaluate the potential activity of thesegroups under low oxygen conditions. I made this connection by comparing pyrotag SSU rDNA observationsand SSU rRNA: rDNA ratios from samples collected at the same water column depths during peak watercolumn stratification in July 2014.Consistent with observed patterns in 2008, rDNA observations in 2014 showed Syndiniales group Iand II OTUs were evenly distributed throughout the water column exhibiting highest relative abundance,10 and 38% respectively, in suboxic and anoxic water column conditions (Fig 4.9). Syndiniales group IIIOTUs distributed from surface (4% relative abundance) to the suboxic boundary of the water column, whileSyndiniales group V OTUs were primarily observed in dysoxic-suboxic water column conditions with arelative abundance <0.1% (Fig 4.9). Choreotrichia OTUs were also found evenly distributed throughoutthe water column and showed higher relative abundance values (∼6%) below the dysoxic boundary of thewater column and highest abundance in anoxic waters (16%) (Fig 9). Maxillopoda OTUs were found evenly48Group-IIGroup-III ChoreotrichiaMASTBacillariophytaPicozoaGroup-IIGroup-IGroup-VGroup-IIIChoreotrichiaMASTGroup-IIGroup-IChoreotrichiaMASTPicozoaAlveolataDinophytaStramenopilesHacrobiaSyndinialesGroup-IGroup-IIIMASTPicozoaBacillariophytaGroup I Group II Group III Group VDinophyceaeEuglenozoaRhizariaOpisthokontaOtherOTUs: 196 OTUs: 238 OTUs: 75 OTUs: 35OpisthokontaMetazoaArthropodaRhizariaStramenopilesDinophytaSyndinialesOther EukaryotaPicozoaChlorophytaHaptophythaAmoebophryaGroup I Group IIABFigure 4.5: Syndiniales OTUs interactions. A) Eukaryotic taxa diversity associated with each Syndinialesgroup. Krona charts depict total unique OTUs with significant interactions for each Syndiniales group(I, II, III and V). Layers represent hierarchical taxonomy from the upper (centre) to the lowest (outer)taxonomic level. Total number of unique OTUs is indicated at the bottom of each chart. B) Co-occurrencenetwork derived from Bray-Curtis and Spearman correlation measures on Sydiniales Group I and II derivedfrom rDNA pyrotag protist data from May-August 2008. Nodes depict OTUs and edges co-occurrencecorrelations. Syndiniales OTUs are highlighted in gray and other protist OTUs coloured as indicated incolor key.495010015020010May June July August September OctoberSyndiniales - Group ISyndiniales - Group II ChoreotrichiaDateDepth (m)Syndiniales - Group IIISyndiniales - Group VAbundance (log)0.5 1 1.5 2 2.550 150 250O2 (μM)Figure 4.6: Vertical distribution and abundance of Choreotrichia and Syndiniales OTUs Choreotrichiaand Syndiniales OTUs distribution through the water column from May-October 2008. Dots size depictsOTUs relative abundance in logarithmic scale.distributed throughout the water column with a relative abundance ∼0.45%, while Phaeocystis sp. OTUswere mostly distributed at dysoxic-suboxic water column conditions with a relative abundance∼0.01% (Fig4.9).Potential activity of these OTUs was determined by calculating SSU rRNA:rDNA ratios across depthintervals (10, 100, 120,135, 150, 165 and 185m) spanning the O2 concentration ranges from oxic to anoxic.In contrast with rDNA observations alone, SSU rRNA observations can provide a robust proxy for past,present or emerging cellular activities (Blazewicz et al., 2013). Moreover, monitoring SSU rRNA dynamicsover time can inform hypotheses related to life strategies within communities (Lepp and Schmidt, 1998;Barnard et al., 2013). I selected OTUs affiliated with indicator OTUs showing significant correlations, andconsidered values >1 as potentially occurring interactions among likely active OTUs in the same depth50Syndiniales - Group ISyndiniales - Group IIMay June July August September OctoberDate5010015020010Depth (m)Abundance (log)0.5 1 1.5 2 2.5Phaeocystis50 150 250O2 (μM)Figure 4.7: Vertical distribution and abundance of Phaeocystis antarctica and Syndiniales OTUsPhaeocystis antarctica and Syndiniales OTUs distribution through the water column from May-October2008. Bars depict OTUs relative abundance in logarithmic scale.51May June July August September October5010015020010Abundance (log)DateDepth (m)0.5 1 1.5 2 2.5Syndiniales - Group ISyndiniales - Group IIMaxillopoda50 150 250O2 (μM)Figure 4.8: Vertical distribution and abundance of Maxillopoda and Syndiniales OTUs Maxillopodaand Syndiniales OTUs distribution through the water column from May-October 2008. Dots size depictsOTUs relative abundance in logarithmic scale.interval. For instance, Syndiniales group II OTUs were active throughout the water column exhibiting thehighest ratio (9.3) at 120m corresponding to the suboxic water column compartment (Fig 4.8). Similarly,Choreotrichia, Phaeocystis and Maxillopoda OTUs were found throughout the water column but exhibitedthe highest ratios in suboxic waters. For instance, Choreotrichia OTUs showed the highest ratio (7.1) at135 m, while Phaeocystis and Maxillopoda OTUs at 150 m with ratio values equal to 14 and 8.6, respec-tively (Fig 4.8). Thus, it could be hypothesized that potential interactions among active OTUs could occurprimarily at dysoxic-suboxic water column conditions between the 100 and 150 m.521020406085100120135150165185ChroreotrichiaGroup−IGroup−IIGroup−IIIGroup−VMaxillopodaPhaeocystisDepth (m)50 150 250SyndinialesO2 (μM)4 8 16≤2Relative abundance (%)ChroreotrichiaGroup−IGroup−IIGroup−IIIGroup−VMaxillopodaPhaeocystisRatio value<1 >1SyndinialesDNA RNA:DNAFigure 4.9: Distribution of Syndiniales and interacting Choreotrichia, Phaeocystis sp., and Maxil-lopoda OTUs throughout the water column oxygen gradient during the stratification period of July 2014.Dissolved oxygen concentration (µM) profile is shown as a black sparkline on the left figure panel. OTUsvertical distribution is shown as dots which size depicts the relative abundance (%) of the OTUs affiliatedto each taxonomic group calculated out of the total number of reads in the SSU rDNA pyrotag dataset. TherRNA:rDNA ratio throughout the water column for the OTUs affiliated to each taxonomic group is shownas size and color constrained dots (gray <1 ; black >1).4.4 DiscussionThis chapter explores protistan diversity, and seasonal changes in abundance and distribution throughout thewater column in Saanich Inlet, a seasonally anoxic fjord that serves as a well-characterized model ecosystemfor examining how deoxygenation shapes microbial community population dynamics and interactions alongdefined redox gradients in the ocean. Observations primarily focused on identifying statistically significantcorrelations that hint to potential interactions between the protistan parasite Syndiniales and potential hosttaxa. Proposed interactions were based on correlational edges between OTUs obtained from Bray-Curtisand Spearmans rank correlations and reinforced by determining potential activity from SSU rRNA:rDNAratios for these groups based on the assumption that potential parasitic interactions are most likely to occurat depths where the host is likely to be active/alive. These results provide insights into the potential for par-53asitic interactions occurring under suboxic conditions during the summer stratification period with possibleimplications for carbon cycling and bloom termination dynamics.Water column seasonal stratification and deep-water renewal restructures microbial eukaryotic commu-nities that respond to changing O2 concentrations by altering their vertical distribution and activities (Parriset al., 2014;Duret et al., 2015;Jing et al., 2015). For instance, a previous study in Saanich Inlet highlightedseasonal differences in eukaryotic community structure correlated to water column O2 conditions (Orsi et al.2012). Similarly, the use of time resolved SSU rDNA pyrotag observations over this seasonal cycle allowedme to observe changes in abundance and distribution patterns of the eukaryote community along definedwater column O2 conditions. Although, the use of high-throughput sequencing data allowed detailed explo-ration of eukaryote diversity, I focused this analysis on Syndiniales, that comprise a diverse and abundant(43% during the summer stratification period) group of parasitic dinoflagellates that are responsive to watercolumn stratification and deep-water renewal, and potentially play a significant role in shaping microbialbiogeochemistry and ecology in diverse water column conditions. As an initial approach to select potentialOTU pairs that likely showed predator-prey or parasitic interactions, I used ISA and co-occurrence corre-lation analyses in complementary manner. For instance, Syndiniales OTUs exhibited concurrent patternswith three likely host taxa i.e. Choreotrichia, Phaeocystis sp. and Maxillopoda and were selected basedon statistical significance then further analyzed to provide baseline evidence for developing ecological hy-potheses on predator-prey or parasitic interactions that may be relevant in stratified and O2 deficient marineenvironments.Syndiniales interactions and the significance of peak stratificationThe initial co-occurrence correlation analysis considered the eukaryotic community throughout 12-month period and identified five Syndiniales OTUs that have statistically significant correlations with othercosmopolitan eukaryotic taxa including Stramenopiles from the MAST clades, picomonads, and other Syn-diniales (Table B.1). These taxa are small (2-8 µm) and not previously documented in the literature ashosts of Syndiniales; however, spatial and temporal trends of these interactions reveal Syndiniales OTUstrack the water column distribution of other microbial eukaryotes from surface water spring blooms to mid-waters as the inlet stratification strengthens in July (Table B.2-B.4). The abundance of Syndiniales OTUsthroughout the year, yet the absence of significant correlations detected in the 12-month analysis betweenSyndiniales and known host taxa, suggested a much broader host range. To reveal more ephemeral interac-tions between Syndiniales and common bloom-forming taxa (i.e. diatoms and dinoflagellates), I repeatedour co-occurrence correlation analysis by targeting the period May through August, encompassing the peakstratification period.Co-occurrence analysis of data collected during peak stratification revealed correlations among mem-bers of a diverse protistan community. Syndiniales OTUs were at the core of complex significant corre-lations with other protist taxa. The number of significant correlations for this period increased 10-foldand involved more diverse taxa, including several species of diatoms, the athecate dinoflagellate Gymno-dinium, the bloom-forming haptophyte Phaeocystis, and heterotrophs, including several species of ciliates,stramenopiles, groups of small flagellates, and Metazoa. Many eukaryotic OTUs showed significant corre-lations with multiple Syndiniales OTUs (Appendix B), suggesting co-occurrence might drive host plasticity54as a survival strategy that allows parasites to respond to host community shifts. Given the short-lived, free-swimming infective stage, I hypothesized co-occurrences between Syndiniales and other eukaryotes in theSaanich Inlet water column are more likely to reflect parasitism or predator-prey relationships than simpleco-occurrences under specific water column conditions. Indicator analyses were compared with results fromthe co-occurrence correlation analysis to gain insights about taxa that were uniquely abundant during thespring and summer months and that showed significant correlation with Syndiniales over the same timeperiod.Based on these criteria, the distributions of three taxa were examined in detail: Phaeocystis antarctica,which produced a bloom in 2008 providing an opportunity to investigate if parasitic dinoflagellates wereinvolved in the termination of the bloom; a ciliate OTU affiliated to Choreotrichia, an indicator taxon duringpeak stratification in micro-oxic waters; and a metazoan copepod taxon, an indicator OTU during peak strat-ification. Ciliates, phytoplankton, and copepod metazoa are all known to be potential hosts of Syndiniales(Guillou et al., 2008; Dolan, 2013; Massana et al., 2014). Observations of spatial and temporal distributionof OTUs affiliated to these taxa suggested that several types of parasitic Syndiniales could be potentiallyinfecting the same host, and/or in the case of heterotrophic protistan taxa, that the heterotrophs may bepotentially feeding on Syndiniales dinospores released from previously infected sinking cells. Nonethe-less, I acknowledge the possibility that a fraction of these OTUs may not be involved in predator/prey orparasitic relationships, and may simply co-occur as a community under specific environmental conditions.Furthermore, single-cell and microscopic analyses could be used in the future to verify specific interactionsproposed by the observations in this chapter.Interactions with ChoreotrichiaTintinnid ciliates belonging to the order Choreotrichida are ubiquitous in the marine plankton, particu-larly in coastal waters where their numbers are generally higher. Tintinnids play different roles in eukaryoticcommunities; they are preyed upon by planktonic zooplankton (Montagnes et al., 2010), as well as largerprotist taxa (Dolan, 2013). Also, as nanoplankton consumers, they have an important role in bloom ter-mination and/or as vectors for transferring bloom-associated toxins to higher trophic levels (Dolan, 2013).Furthermore, some tintinnid ciliates have been observed to serve as hosts for at least two different parasites,including Syndiniales, and even single cells may be co-infected (Dolan, 2013). In this chapter I observedciliates belonging to the subclass Choreotrichia were most abundant between 75 m and 150 m in August2008, and these OTUs were indicator OTUs for micro-oxic mid-waters during peak stratification. Corre-lation analysis results showed seven indicator Choreotrichia OTUs that were significantly correlated withSyndiniales groups (I-V) OTUs (Table B.3).However, the distribution of Syndiniales Group V OTUs most closely matched the distribution patternsof the OTUs affiliated with Choreotrichia throughout the water column between May and August (Fig 4.6).For example, in July 2008, the four Syndiniales groups and the ciliate subclass Choreotrichia were presentat the surface 10m and in micro-oxic waters between 85m and 120m (Fig. 4.6). while in August, onlySyndiniales Group V matched Choreotrichia abundance and distribution patterns in the mid- and deep-water samples (Fig 4.6). Analysis of SSU rRNA:rDNA ratios during peak stratification in 2014 indicatedpotential activity of Choreotrichia at 10 m, 97 m, 135 m, and 165 m depths in waters that ranged from55oxic to anoxic. Syndiniales Group II was also active at these depths, which may suggest either parasitismor grazing could take place under different water column O2 conditions. Interestingly, SSU rRNA: rDNAratios for Syndiniales Group I, III, and V indicated cells were not likely active in the upper oxic layer, whichsupports the hypothesis that Choreotrichia may be feeding on dinospores from these groups during peakstratification (Fig 4.9). The correlations between OTUs affiliated with Choreotrichia and the four knownSyndiniales groups in Saanich Inlet indicated by the correlation analysis suggest complex interactions couldoccur that are associated with season water column dynamics. Further single-cell analyses are required toreveal the extent to which parasitism versus grazing is contributing to these detected potential interactions.Interactions with Phaeocystis antarcticaPhaeocystis antarctica is a colony-forming haptophyte that contributes to primary production and dimethyl-sulfide formation (Kettle et al., 1999). The dimethylsulfide produced by bloom events can be oxidized tosulphuric acid in the atmosphere where it can function as nuclei for cloud condensation, impacting globalalbedo (Charlson et al., 1987). P. antarctica has a complex life cycle that includes colonial, bloom-formingaggregates, individual flagellates (microzoospores), and a recently discovered zygotic “benthic stage” (Gae-blerSchwarz et al., 2010).In this chapter, I observed OTUs affiliated with P. antarctica comprised 6% of the pyrotag library at 10m depth in May 2008, consistent with a late spring surface bloom. Based on the depth of detection of P.antarctica pyrotags in subsequent months, cells appeared to sink through the water column and signaturesof this taxon disappeared following deep water renewal in October 2008 (Fig 4.7). Four and eight OTUsaffiliated with Syndiniales groups I and II, respectively, were identified as having significant correlationswith this haptophyte, with OTUs from these two groups of Syndiniales tracking signatures of P. antarcticathrough the water column during peak stratification. SSU rRNA:rDNA ratio analysis indicated P. antarcticawas not potentially active at 10m, however its highest potential activity was observed at 150 m in July 2014as also reflected in its abundance based on rDNA observations alone. This suggests an ephemerally activebloom followed by sinking of likely infected cells and “benthic stage” zygotes (Fig 4.9). Following renewal(which occurred on October 14, 2008; Ocean Networks Canada VENUS node (http://www.oceannetworks.ca) the co-occurrence between Group I and II Syndiniales OTUs and P. antarctica was not detected, howeverthose two groups of Syndiniales appeared to correlate with different putative hosts in the surface waters, i.e.OTUs affiliated with Pseudocalanus copepods and maxillopods.Parasites of P. antarctica blooms have not been previously identified; however, based on our co-occurrencecorrelation analyses I hypothesize this haptophyte could host several strains of Group I and II Syndiniales,and that these two groups of parasites locate and infect new hosts when P. antarctica is absent from the watercolumn. Our observation hints at a testable hypothesis for future research is that parasitism of Phaeocystisblooms in Saanich Inlet could play a role in controlling the duration of phytoplankton blooms that play amajor role in spring primary production, and that influence dimethylsulfide synthesis.Interactions with copepodsCommon taxonomic groups of mesozooplankton in Saanich Inlet include calanoid copepods, chaetog-naths, euphausiids, cnidarians, and ctenophores. These taxa are critical for trophic transfer of carbon inmarine food webs (e.g. Jnasdttir et al., 2015) as mesozooplankton are predators of phytoplankton, het-56erotrophic protists, and bacteria, as well as prey items for larger metazoans including fish. Zooplanktonobserved from a time series collected off southern Vancouver Island undergo annual variations in abun-dance, leading to changes in overall biomass and relative species composition that can further influence thepelagic food web (Hargreaves et al., 1994; Beamish et al., 1997; Bertram et al., 2001). Copepods typicallyfound in shelf and slope waters off the coast of British Columbia have regionally specific ranges, drivenprimarily by water temperature, in addition to inter-annual shifts in species abundance. A few occurrencesof dinoflagellate syndinian parasites from marine alveolate Group II (MALV II) infecting copepods andcopepod eggs have been reported in the literature (Kimmerer and McKinnon, 1990; Skovgaard et al., 2005;Gmez et al., 2009); however, the overall ecological effects of these infections are not known.Co-occurrence correlation analysis shows that Metazoa exhibit significant correlations with Syndinialesin Saanich Inlet. While the observed correlations may imply Metazoa may be feeding directly on dinosporesor more likely, infected prey, these also suggest that Syndiniales OTUs could parasitize certain Metazoa. Iobserved OTUs affiliated with the genera Corycaeus and Oncaea were indicator OTUs for surface watersin May when prey taxa including Thlassiosira aestivalis, Chrysochromulina rotalis, Phaeocystis antarctica,Micromonas sp., and Florenciellales diatoms were abundant, suggesting higher metazoan grazing activityduring peak phytoplankton growth. Additionally, these copepods showed significant correlations with Syn-diniales from Groups I and II and, like the other taxa discussed, Syndiniales OTUs appears to track thegrazers throughout the water column (Fig. 4.8).The corresponding SSU rRNA:rDNA ratios for OTUs affiliated with the crustacean class Maxillopoda(copepods) indicated potential activity of this group of metazoans in the upper oxic layer during peak strati-fication in July 2014 (Fig. 4.9). Syndiniales Group II SSU rRNA:rDNA ratios indicated potential activity inthe water column at the same depths as the maxillopods, while Group I Syndiniales was likely inactive. TheRNA-based analyses thus suggest Syndiniales Group I OTUs are likely sinking cells, while Group II couldpotentially have parasitic or prey interactions with members of Maxillopoda below the surface. Given theapparent significance of the interactions between metazoan groups known to play important roles in marinefood webs, I consider it is important to understand the extent to which members of Syndiniales parasitizemetazoans and the impact that this may have on carbon and nutrient turnover parasitize metazoans.574.5 ConclusionIn this chapter I used molecular time series observations to detect interactions between groups of primarilyuncultured, parasitic dinoflagellates within the Syndiniales and other eukaryotes along defined oxygen gra-dients in Saanich Inlet. Co-occurrence correlation and indicator species analyses revealed the potential forsignificant interactions may occur between four known Syndiniales groups and different protistan and meta-zoan taxa including the stramenopile MAST clades not previously known to be parasitized by Syndiniales.I focused the analysis on several eukaryotic OTUs showing significant correlations with Syndiniales duringpeak stratification, including Phaeocystis antarctica, ciliates affiliated with Choreotrichia, and copepods.These observations provide baseline understanding on the potential host range of the major parasitic Syn-diniales groups that could infect key primary producing and heterotrophic populations along stratified watercolumns. Results presented here give insight into possible impacts these infections may have on populationdynamics, extensible to nutrient cycling processes, during seasonal water column stratification in SaanichInlet.58Chapter 5Methanotrophic community dynamics inSaanich Inlet1This chapter represents a unique correlation analysis coupling time-series geochemical and small subunitribosomal (SSU) RNA gene observations to chart spatial and temporal patterns of methanotrophic interac-tions. Here the trace gas problem in OMZs is reintroduced by focusing on the CH4 oxidation processes.Oxygen (O2) effects on ecological interactions and energy flow were reinforced by observed redox-drivenniche partitioning along changing water column redox gradients for methanotrophic bacteria and the po-tential use of novel metabolic strategies such as the use of alternative terminal electron acceptors for CH4oxidation. Correlation analyses revealed potential community-level interactions among methanotrophs andone-carbon compounds utilizing microorganisms that served as conceptual framework for Chapter 6.5.1 IntroductionReduced levels of dissolved O2 (<20 µM kg-1) enhances the use of alternative inorganic compounds aselectron acceptors for anaerobic respiration by microorganisms resulting in the production of climate activetrace gas such as carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) (Lam et al., 2009b; Ward etal., 2009). OMZs are the largest marine source of CH4 flux to the atmosphere (∼ 1 Tg CH4 yr-1) (Naqvi etal., 2010). Previous surveys have quantified CH4 oxidation rates in oxic (> 90 µmol O2 kg-1) and dysoxic-suboxic (<20 µmol O2 kg-1) OMZ waters as well as anoxic sediments. Although the process of anaerobicoxidation of CH4 (AOM) can consume ∼75% of CH4 in the sediment (Strous and Jetten, 2004; Knittel andBoetius, 2009), results indicated that anaerobic oxidation of CH4 (AOM) occurs at a much slower rate thanaerobic CH4 oxidation in the water column (0.72 nmol L-1h-1 and 2 nmol L-1h-1 respectively) (Ward et al.,1989; Ward and Kilpatrick, 1990; 1993). Aerobic CH4 oxidation has been estimated to consume >50% ofCH4 in the water column (Fung et al., 1991; Reeburgh et al., 1991a) likely having the largest influence onthe CH4 budget before emission to the atmosphere (Reeburgh, 2007). Thus, aerobic CH4 oxidation providesa second biological filter following sediment AOM that reduces CH4 flux (Ward et al., 1989; Ward and1A version of this chapter appears as Torres-Beltra´n, M. et al. 2016. Methanotrophic community dynamics in a seasonallyanoxic fjord: Saanich Inlet, British Columbia. Front. Mar.Sci. doi.org/10.3389/fmars.2016.00268 .59Kilpatrick, 1990; 1993).Efforts to understand microbial agents driving methane oxidation based on small subunit ribosomalRNA (SSU rRNA) gene surveys from diverse OMZs such as the Eastern Tropical South Pacific, the Namib-ian Upwelling and the Black Sea (Stevens and Ulloa, 2008; Glaubitz et al., 2010), indicate that canoni-cal methanotrophs within the alpha and gammaproteobacteria, are rare microbial community members i.e.<0.01% of the total microbial community. Parallel efforts to describe the distribution, abundance and po-tential metabolic activity of the functional gene particulate methane monooxygenase (pMMO) subunit β(pmoA), identified pMMO- encoding phylogenetic groups (OPUs), OPU1 to OPU4, affiliated with canoni-cal methanotrophic groups in OMZs waters (Hayashi et al., 2007; Tavormina et al., 2013). Phylogeneticallyaffiliated with Methylococcales OPU1 and OPU3 groups were initially observed in the Eastern Pacific OceanOMZ (Hayashi et al., 2007), and exhibit differential abundance and distribution patterns. OPU3 was moreabundant under low O2 concentrations in the Costa Rica OMZ water column (Tavormina et al., 2013), andexpression of pmoCAB for group OPU3 has recently been demonstrated in a metatranscriptome from theGuaymas Basin (Lesniewski et al., 2012). In addition to canonical methanotrophs, more recent studies haveidentified a number of non-canonical microbial groups (e.g. bacteria affiliated with the Verrucomicrobia(Dunfield et al., 2007), SAR324 within the deltaproteobacteria (Swan et al., 2011) and the NC10 candidatedivision (Ettwig et al., 2010)) with the potential to mediate CH4 cycling in OMZs. In both known and novelcases there is limited information on the dynamics and interspecific interactions of CH4 cycling microbesneeded to constrain their biological filtering capacity.Saanich Inlet is a seasonally anoxic fjord on the east coast of Vancouver Island British Columbia. Duringspring and summer months, restricted circulation and high levels of primary production lead to progressivedeoxygenation and the accumulation of CH4, ammonium (NH4+) and hydrogen sulfide (H2S) in the deepwaters of the Saanich Inlet basin. In late summer and fall, upwelling oxygenated nutrient rich ocean waterscascade into the inlet shoaling anoxic bottom waters upward and transforming the redox chemistry of thewater column. The recurring seasonal development of water column anoxia followed by deep water renewalmakes Saanich Inlet a model ecosystem for evaluating microbial community structure, function and dynam-ics in relation to changing levels of water column O2 deficiency extensible to coastal and open ocean OMZs(Walsh et al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011; Wright et al., 2012).Process rates and molecular surveys focused on CH4 oxidation have been previously conducted inSaanich Inlet during peak summer stratification. Process measurements indicated that CH4 oxidation rateswere highest near the oxic-anoxic interface (∼2 nmol L-1 d-1) (Ward et al., 1989). Subsequently, sequencesaffiliated with canonical methanotrophs such as Methylococcales within the gammaproteobacteria were re-covered as rare biosphere components based on full-length SSU rRNA gene sequences (Zaikova et al., 2010)and particulate methane monoxygenase subunit β (pmoA) libraries were dominated by a non-canonical phy-lotype (Stilwell, 2007). In the same study, anaerobic methane oxidizing archaea (ANME) were undetectablein the water column. The low abundance of canonical methanotrophs combined with measured CH4 oxi-dation rates in both oxycline and deep basin waters presents a “CH4 oxidation conundrum” (Ward et al.,1989; Ward and Kilpatrick, 1993; Zaikova et al., 2010). Here I attempt to constrain this conundrum us-ing taxonomic survey information over two years of seasonal stratification and renewal supplemented with60functional gene information for pmoA.5.2 Methods5.2.1 Environmental samplingEnvironmental monitoring and sample collection were carried out monthly aboard the MSV John Stricklandat station SI03 (48o 35.500 N, 123o 30.300 W) in Saanich Inlet, B.C. From February 2008 to July 2010 atotal of 288 water samples were collected at 16 high-resolution depths (10, 20, 40, 60, 75, 80, 90, 97, 100,110, 120, 135, 150, 165, 185 and 200 meters) spanning oxic (>90 µmol O2 kg-1), dysoxic (90-20 µmolO2 kg-1), suboxic (20-1 µmol O2 kg-1) anoxic (<1 µmol O2 kg1) and sulfidic water column compartments(Wright et al., 2012). Samples were processed and analyzed as previously reported for the Saanich Inlettime-series (Zaikova et al., 2010; Capelle et al., 2015; Torres-Beltra´n et al., 2016b). Briefly, water sampleswere collected from Niskin or Go-Flow bottles for dissolved O2 and CH4, nutrients (Nitrate (NO3-), Nitrite(NO2-), Ammonium (NH4), Silicon dioxide (SiO2), Phosphate (PO4), Hydrogen sulfide (H2S)), DNA, andRNA. In addition, conductivity, temperature, and depth were measured using a Seabird SBE19 CTD-device(Sea-Bird Electronics Inc., Bellevue USA), with a PAR and O2 sensor attached. CTD was also used tomeasure dissolved O2 throughout the water column. Dissolved gases and nutrient measurement protocolshave been previously reported for the Saanich Inlet time-series (Zaikova et al., 2010; Capelle et al., 2015;Torres-Beltra´n et al., 2016b), and data is available through Dryad Digital Repository (www.dryad.org).5.2.2 Nucleic acid sampling and extractionBiomass to generate full-length small subunit (SSU) rRNA gene clone library sequences and pmoA genelibrary sequences (Stilwell, 2007), and metagenomic datasets was collected on February 2006 and February2010 respectively, at six depths (10, 100, 120, 135, 150 and 200 m) and filtered with an in-line 2.7 µmGDF glass fibre pre-filter onto a 0.22 µm Sterivex polycarbonate cartridge filter. Biomass to generate SSUrRNA pyrotag datsets for microbial community composition profiling (2 L) and filtered directly onto a 0.22µm Sterivex polycarbonate cartridge filter from high-resolution depths collected between May 2008 to July2010. Biomass for RNA analysis (2 L) was collected on February 2010 at six depths (10, 100, 120, 135,150 and 200 m), and filtered with an in-line 2.7 µm GDF glass fibre pre-filter onto a 0.22 µm Sterivex filterwithin 20 min of shipboard sample collection.Environmental DNA was extracted from Sterivex filters as previously described (Wright et al., 2009;Zaikova et al., 2010). Briefly, after defrosting Sterivex on ice, 100 µl lysozyme (0.125 mg ml-1; Sigma)and 20 µl of RNAse (1 µl ml-1; ThermoFisher) were added and incubated at 37 oC for 1 h with rotationfollowed by addition of 50 µl Proteinase K (Sigma) and 100 µl 20% SDS and incubated at 55 oC for 2 hwith rotation. Lysate was removed by pushing through with a syringe into 15 mL falcon tube (Corning) andwith an additional rinse of 1 mL of lysis buffer. Filtrate was subject to chloroform extraction (Sigma) andthe aqueous layer was collected and loaded onto a 10K 15 ml Amicon filter cartridge (Millipore), washedthree times with TE buffer (pH 8.0) and concentrated to a final volume of between 150-400 µl. Total61DNA concentration was determined by PicoGreen assay (Life Technologies) and genomic DNA qualitydetermined by visualization on 0.8% agarose gel (overnight at 16V).Total RNA was extracted from Sterivex filters using the mirVana Isolation kit (Ambion) (Shi et al.,2009; Stewart et al., 2010) protocol modified for sterivex filters (Hawley et al., 2017b). Briefly, afterthawing the filter cartridge on ice RNA later was removed by pushing through with a 3 ml syringe followedby rinsing with an additional 1.8 mL of Ringers solution and incubated at room temperature for 20 minwith rotation. Ringers solution was evacuated with a 3 ml syringe followed by addition of 100 µl of 0.125mg ml-1 lysozyme and incubated at 37 oC for 30 min with rotation. Lysate was removed from the filtercartridge and subjected to organic extraction following the mirVana kit protocol. DNA removal and cleanup and purification of total RNA were conducted following the TURBO DNA-free kit (ThermoFisher) andthe RNeasy MinElute Cleanup kit (Qiagen) protocols respectively.5.2.3 Small subunit ribosomal RNA gene sequencing and analysisTo initially assess microbial community diversity in the Saanich Inlet water column full-length bacterialSSU rRNA gene sequences datasets were generated from the February 2006 LV samples as previouslydescribed (Zaikova et al., 2010). Briefly, DNA extracts from 10, 100, 120 and 200 m samples were amplifiedusing SSU rRNA primers targeting the bacterial domain: B27F (5’- AGAGTTTGATCCTGGCTCAG) andU1492R (5’-GGTTAC CTTAGTTACGACTT) under the following PCR conditions: 3 min at 94 oC followedby 35 cycles of 94 oC for 40 s, 55 oC for 1.5 min, 72 oC for 2 min and a final extension of 10 min at 72oC. Each 50 ml reaction contained 1 µl of template DNA, 1 µl each 10 mM forward and reverse primer,2.5 U Taq (Qiagen), 5 ml 10 mM deoxynucleotides, and 41.5 µl 1X Qiagen PCR Buffer. Clone libraryconstruction and screening. The SSU rRNA gene amplicons were visualized on 1% agarose gels in 1X TAEand purified using the MinElute PCR Purification Kit (Qiagen) according to the manufacturer’s instructions.Approximately 4 µl of each purified SSU rRNA gene product was cloned into a pCR4-TOPO vector usinga TOPO TA cloning kit for sequencing (Invitrogen) and transformed by chemical transformation into Mach-1-T1R cells according to the manufacturer’s instructions. Transformants were transferred to 96-well platescontaining 180 ml LBkan50 and 10% glycerol and grown overnight at 37 oC prior to storage at -80 oC.Cloned inserts were amplified directly from glycerol stocks with M13F (5’-GTAAAACGACGGCCAG) andM13R (5’- CAGGAAACAGCTATGAC) primers using the SSU rRNA gene PCR protocol. Bidirectionalend sequencing was performed on a Sanger platform at the Department of Energy Joint Genome Institute(DOE-JGI; Walnut Creek, CA).To survey microbial community structure and dynamics throughout the complete water column profileover time, DNA extracts from the samples from May 2008 to July 2010 were used to generate SSU rRNAgene pyrotag datasets. Pyrotag libraries were generated by PCR amplification using multi-domain primerstargeting the V6-V8 region of the SSU rRNA gene (Allers et al., 2013): 926F (5’-cct atc ccc tgt gtg ccttgg cag tct cag AAA CTY AAA KGA ATT GRC GG-3’) and 1392R (5’-cca tct cat ccc tgc gtg tct ccg actcag-<XXXXX>-ACG GGC GGT GTG TRC-3’). Primer sequences were modified by the addition of 454A or B adapter sequences (lower case). In addition, the reverse primer included a 5 bp barcode designated<XXXXX> for multiplexing of samples during sequencing. Twenty-five microliter PCR reactions were62performed in triplicate and pooled to minimize PCR bias. Each reaction contained between 1 and 10 ngof target DNA, 0.5 µl Taq DNA polymerase (Bioshop inc. Canada), 2.5 µL Bioshop 10 x buffer, 1.5 uL25 mM Bioshop MgCl2, 2.5 µL 10 mM dNTPs (Agilent Technologies) and 0.5 µL 10 mM of each primer.The thermal cycler protocol started with an initial denaturation at 95 oC for 3 minutes and then 25 cyclesof 30 s at 95 oC, 45 s at 55 oC, 90 s at 72oC and 45 s at 55 oC. Final extension at 72 oC for 10 min. PCRproducts were purified using the QiaQuick PCR purification kit (Qiagen), eluted elution buffer (25 µL), andquantified using PicoGreen assay (Life Technologies). SSU rRNA amplicons were pooled at 100 ng DNAfor each sample. Emulsion PCR and sequencing of the PCR amplicons were sequenced on Roche 454 GSFLX Titanium at the DOE-JGI, or the McGill University and Ge´nome Que´bec Innovation Center.Pyrotag sequences were processed using the Quantitative Insights Into Microbial Ecology (QIIME) soft-ware package (Caporaso et al., 2010). To minimize the removal of false positive reads all 3,985,489 pyrotagsequences generated from the 288 samples were clustered together. Reads with a length shorter than 200bases, ambiguous bases, and homopolymer sequences were removed prior to chimera detection. Chimeraswere detected and removed using chimera slayer provided in the QIIME software package. Sequences werethen clustered into operational taxonomic units (OTUs) at 97% identity using uclust with average linkagealgorithm. Prior to taxonomic assignment singleton OTUs (OTUs represented by one read) were omitted,leaving 69,051 OTUs. Representative sequences from each non-singleton OTU were queried against theSILVA database (Pruesse et al., 2007) and the Greengenes database (DeSantis et al., 2006) using BLAST(Altschul et al., 1990).5.2.4 pmoA gene libraries, metagenomic and metatranscriptomic sequencing and analysisExtracted total DNA from the February 2006 samples were used to identify bacterial particulate methanemonooxygenase (pMMO) subunit β (pmoA) presence in the Saanich Inlet water column. Briefly, pmoAsequences were PCR amplified (Stilwell, 2007) using gene-specific forward and reverse primers A189F (5’-GGNGACTGGGACTTCTGG) (Holmes et al., 1995) and mb661R (5’- CCGGMGCAACGTCYTTACC)(Costello and Lidstrom, 1999) and the following PCR profile: 30 cycles of 96 oC for 25 s, 54 oC for 45s, and 72 oC for 50 s. Each 50 µl reaction contained l µl of template DNA, 2 µl each 10 µM forwardand reverse primer, 2.5U Taq (Qiagen), 4 µl 10 mM deoxynucleotides, and 40.5 µl 1x Buffer (Qiagen-TaqPolymeraseKit). pmoA amplicons were purified using the MinElute PCRPurificationKit(Qiagen, CA),cloned into a pCR4-TOPO vector using a TOPO TA cloning kit for sequencing (Invitrogen, Carlsbad CA),and transformed by chemical transformation into Mach-1-T1R cells according to the manufacturer’s in-structions. Cloned inserts were amplified directly from glycerol stocks for fingerprint screening using thecommon 4-base cutter Rsa I (Invitrogen, CA). Restriction patterns were visually inspected and unique pat-terns selected for Sanger sequencing through the McGill University and Ge´nome Que´bec Innovation Centre(Montreal, Quebec, Canada) (Stilwell, 2007). Four libraries were constructed from 10, 100, 120 and 200 m,from which a total of 33 representative sequences were identified.Extracted total DNA (6 samples) and RNA (6 samples) from February 2010, corresponding to the 10,100, 120, 135, 150 and 200 m depth intervals, were used to generate metagenomic and metatranscriptomicdatasets at the DOE-JGI following the protocols for library production and sequencing and assembly previ-63ously described for the Saanich Inlet time-series (Hawley et al., 2016).A total of 6 assembled metagenomes and 6 assembled metatranscriptomes were analysed using MetaP-athways V2.5.1, an open source pipeline for predicting reactions and pathways using default settings (Kon-war et al., 2013) (https://github.com/hallamlab/metapathways2/wiki). For each gene, reads per kilobase permillion mapped (RPKM) was calculated as a proportion of the number of reads mapped to a sequence sec-tion, normalized for sequencing depth and ORF length (Konwar et al., 2015). RPKM values and relativeabundance to the total number of reads were used to describe the abundance of pmoA genes and transcripts.5.2.5 Phylogenetic inferenceTo generate a reference phylogenetic tree for methanotrophic bacteria, full-length SSU rRNA gene se-quences from Saanich Inlet, and diverse environmental reference sequences affiliated with methanotrophicbacteria were first aligned and compared. Full-length SSU rRNA gene sequences from Saanich Inlet wereedited manually using Sequencher software V4.1.2 (Gene Codes Corporation) and imported into the full-length SSU SILVA database (http://www.arb-silva.de) and aligned to the closest relative. Sequences affili-ated with known methanotrophs were extracted from the dataset. Methanotroph reference sequences fromSaanich Inlet in addition to 53 reference sequences for methanotrophic bacteria including cultured type Iand II methanotrophs, OMZ representatives, mussel symbionts and envioronmental clones were clusteredat 97% identity using mothur v.1.19.0 (Schloss et al., 2009). A total of 88 sequences affiliated with Methy-lococcales were recovered representing 1.3% from the total Saanich Inlet full-length SSU rRNA sequencesgenerated. Sequences clustered at 97% identity resolved into 11 distinct clusters, 5 of which containedmost of the identified sequences (93%). Representative sequences for the most abundant clusters revealed4 subgroups with phylogenetic similarity to environmental representatives of type I methanotrophs: Methy-lococcaceae (Mou et al., 2008), putative methanotrophic group OPU3 (Hayashi et al., 2007; Tavormina etal., 2010; Tavormina et al., 2013), environmental seafloor clones (Santelli et al., 2008), and methanotrophicsymbionts (Streams et al., 1997; Dubilier et al., 2008; Petersen and Dubilier, 2009). Representative se-quences for the most abundant (represented by more than one sequence) clusters were identified using theget.oturep command in mothur and were included in the phylogenetic tree. Representative sequences werealigned using the SSU SILVA database and imported into the ARB software (Ludwig et al., 2004) for treedistance matrix and alignment generation using the ARB parsimony tool. ARB sequences were exported toMesquite (V.2.0) and edited manually. A maximum likelihood phylogenetic tree was inferred by PHYML(Guindon et al., 2005) using an GTR model of nucleotide evolution where the parameter of the gammadistribution, the proportion of invariable sites and the transition/transversion ratio were estimated for eachdata set. The confidence of each node was determined by assembling a consensus tree of 1,000 bootstrapreplicates.To further resolve diversity of methanotrophic bacterial OTUs, I recruited pyrotag OTU sequences tofull-length SSU rRNA gene sequences described above. Pyrotag sequences affiliated with methanotrophsbased on BLAST-comparison in QIIME were re-clustered with SSU rRNA gene tree reference sequencesusing a 97% identity cut-off in mothur (Schloss et al., 2009). In addition, blastn was used to query repre-sentative pyrotag sequences from clusters against full-length SSU rRNA reference tree sequences. Only hits64with a perfect match across the full length of a query sequence were retrieved, and the number of pyrotagsmapping to all sequences in each cluster was summed. Clusters represented by one pyrotag sequence werenot used in downstream analyses. Representative pyrotag sequences were aligned in ARB to reference treesequences, imported to Mesquite for manual edition, and finally, included in the phylogenetic tree inferredby PHYML using the parameters detailed above.To generate a reference phylogenetic tree for PmoA, conceptually translated and annotated ORFs fromthe metagenomic and metatranscriptomic datasets were manually extracted from the functional annotationtable <ORF annotation table.txt> in the <results/annotation table> output directory. Sequences werealigned and compared to diverse environmental and reference PmoA sequences. A total of 60 PmoA se-quences, retrieved from metagenomic (34 sequences) and metatranscriptomic (26 sequences) datasets, wereclustered over a range of identity thresholds using the UClust algorithm (USEARCH V6.0) with 52 referencesequences including Saanich Inlet pmoA gene libraries, cultured and environmental sequences affiliated withType I and II methanotrophs, and novel pmoA phylotypes found within the SAR324 clade, Verrucomicrobiaand Candidate Methylomirabilis oxyfera NC10. Reference sequences also included ammonium monooxy-genase subunit α (AmoA). The 97% identity threshold was selected based on resolution of the OPUs andsymbiont groups. Cluster representative sequences were aligned using the Multiple Sequence Comparisonby Log- Expectation (MUSCLE) method (EMBL-EBI), and was manually curated in Mesquite. A maxi-mum likelihood phylogenetic tree was inferred by PHYML (Guindon et al., 2005) using an WAG model ofamino acid evolution where the parameter of the gamma distribution, the proportion of invariable sites andthe transition/transversion ratio were estimated. The confidence of each node was determined by assemblinga consensus tree of 1000 bootstrap replicates.5.2.6 Statistical analysesPyrotag datasets were normalized to the total number of reads per sample, and environmental parameter datawere transformed to the same order of magnitude so that each variable had equal weight. Hierarchical clusteranalysis (HCA) was conducted to identify groups associated with discrete water column compartments. Inaddition, OTUs were correlated using nonmetric multidimensional scaling (NMDS) with environmentalparameters. Hierarchical cluster and NMDS analyses of microbial community compositional profiles weredone using the pvclust (Suzuki and Shimodaira, 2015) and MASS (Venables and Ripley, 2002) packagesin the R software (RCoreTeam, 2013) with Manhattan Distance measures, and statistical significance tothe resulting clusters was computed as bootstrap score distributions with 1,000 iterations and NMDS stressvalue <0.05.Multi-level indicator species analysis (ISA) using the indicspecies package (De Caceres and Legendre,2009) in the R software (RCoreTeam, 2013) was performed to identify OTUs specifically associated withdifferent water column compartments defined by HCA. The ISA/multi-level pattern analysis calculates pvalues with Monte Carlo simulations and returns indicator values (IV) and p-values with α < 0.05. TheIVs range between 0 and 1, where indicator OTUs considered in the present chapter for further communityanalysis shown an IV > 0.7 and p-value < 0.001.A multivariate regression analysis (Fox and Weisberg, 2011) was conducted on time-series pyrotag data65to infer significant correlations between OTUs affiliated with methanotrophic bacteria while controlling forthe effect of depth. Analysis was conducted in the R environment (RCoreTeam, 2013). Parameter estimateswere calculated with least squares fit between relative abundances and depth measurements. Only represen-tative OTUs affiliated with methanotrophic bacteria were regressed. Statistical significance of correlationswas determined with bootstrapping (1,000 iterations), and using a Type-I error rate of 5%. Univariate re-gression analyses were conducted on time-series pyrotag data to infer correlations between OTUs affiliatedwith methanotrophic bacteria counts and environmental variables (O2, CH4, NO3-, NO2-, and H2S). Choiceof model per OTU was determined by AIC testing (Akaike, 1998), using a combination of negative binomialregression (Venables and Ripley, 2002) and zero-inflated negative binomial regression (Zeileis et al., 2008)employed due to many zeroed observations (average: 54%). Parameter estimates and statistical significancewere calculated in the R environment (RCoreTeam, 2013), using a Type-I error rate of 5%.Co-occurrence networks To generate a robust network emphasizing co-occurrences between prevalentOTUs in water column compartments defined by HCA rather than individual depth intervals, the Bray-Curtisand Spearman’s rank correlations were used. Correlation coefficients were calculated using CoNet (Faustet al., 2012) with OTU abundance as count data. First, an OTU matrix derived from the pyrotag taxonomicanalysis was transformed into presence-absence data to remove OTUs with less than 1/3 zero counts, leavinga matrix of 780 OTUs for all samples. Next, to construct ensemble networks, measure-specific thresholdsset to 0.6 were used as a pre-filter and edge scores were computed only between clade pairs. To assignstatistical significance to the resulting scores, edge and measure-specific permutation and bootstrap scoredistributions with 1,000 iterations each were computed. p-values were tail-adjusted so that low p-valuescorrespond to co-presence and high p-values to exclusion. After merging, p-values on each final edge werecorrected to q-values (cut-off of 0.05). The positivity or negativity of each relationship was determined byconsensus voting over all integrated data sources. Finally, only edges with at least two supporting pieces ofevidence were retained.The final edge matrix was visualized as a force directed network using Cytoscape 2.8.3 (Shannon etal., 2003). Network properties were calculated with the “Network Analysis” Plug-In. Nodes in the co-occurrence network corresponded to individual OTUs and edges were defined by computed correlationsbetween corresponding OTU pairs. The layout revealed distinct modules, which persisted after loweringthe correlation coefficient cut-off for edge creation to 0.90 reinforcing the robustness of the network. Edgesfrom modules were selected and visualized as sub-networks using the tool Hive Panel Explorer (https://github.com/hallamlab/HivePanelExplorer/wiki) (Perez, 2015). HivePlotter allows for edge selection basedon interactions within specific OTUs such as those affiliated with methanotrophic bacteria.5.2.7 Data depositionThe SSU rRNA gene sequences reported in this chapter have been submitted to the The National Centerfor Biotechnology Information (NCBI) under BioSample numbers: GQ346856, GQ349233, GQ347199,HQ163221, GQ350623, GQ349295. The SSU rRNA pyrotag sequences reported in this chapter have beensubmitted to the NCBI under BioSample numbers: SAMN03387532 - SAMN03387915. Metagenomesreported in this chapter have been submitted to the NCBI under BioSample numbers: SAMN05224436,66SAMN05224437, SAMN05224442, SAMN05224443, SAMN05224447, SAMN05224451. Metatranscrip-tomes reported in this chapter have been submitted to the NCBI under BioSample numbers: SAMN05238739,SAMN05238741, SAMN05238743, SAMN05238745, SAMN05238748, SAMN05238751.5.3 Results5.3.1 Water column conditionsI monitored changes in water column conditions corresponding to the progression of stratification (fromwinter through mid-summer) and deep water renewal (late summer into fall) events over a 2-year period(May 2008 to July 2010). As previously described for OMZs (Wright et al., 2012), in this chapter I definewater column compartments on the basis of O2 concentration ranges: oxic (>90 µM O2), dysoxic (90-20µM O2), suboxic (<20-1 µM O2) and anoxic (<1µM). Between May and August 2008, as water columnstratification peaked, suboxic conditions intensified. This intensification corresponded with increasing levelsof CH4 and hydrogen sulphide (H2S) below 150 m consistent with the development of deep-water anoxia.Two CH4 concentration peaks were observed, at the subsurface (20 100 m) ranging between the 20-120nM, and at in the deep basin increasing steadily below 150m to a maximum of 800 nM at 200 m in lateJuly. Over the same time interval, the concentration of H2S ranged between 2 to 8 µM in the anoxic bottomwaters (150-200m). The concentration of NO3- between the surface and 100m ranged between 5 to 20µM, decreasing rapidly between 100 and 135 m before reaching a minimum of <1 µM in anoxic bottomwaters. The concentration of NO2- between surface and 135 fm ranged between <0.1 to 0.25 µM reachinga maximum 0.3 µM in anoxic bottom waters (150-200 m) (Fig. C.1).The beginning of 2008 deep-water renewal occurred between the end of July and early September,continuing through November. During this time interval, oxygenated nutrient-rich waters flowed over thesill, displacing anoxic bottom waters upwards and disrupting the redox gradient established during springand summer months. Between September and October, dissolved O2 was observed throughout the watercolumn, although upwards shoaling O2 and NO3- depleted bottom waters produced an intermediate suboxiclayer between 100 and 135 m. Concomitantly, CH4 concentrations increased transiently in the suboxictransition zone (120-135 m) ranging between 180-700 nM while decreasing below 135 m to 0 nM in bottomwaters. In November, O2 and NO3- and NO2- concentrations continued to increase above 100 and below135m with intensification of water column O2 deficiency within intervening depth intervals (Fig. C.1).Beginning in December 2008 water column O2 deficiency intensified below 100 m consistent with strati-fication. However, in contrast to previous studies at Saanich Inlet, these observations indicate a significantlyweak renewal occurred in fall 2009 as indicated by no measurable increase in O2 or NO3- in deep basinwaters. This phenomenon extended O2 deficiency below 100 m resulting in CH4 accumulation through thesummer of 2010, and corresponded with anoxic conditions below 135 m (O2 and NO3- concentrations equalto 0 µM combined with high levels of H2S). Over this period, the concentration of CH4 between surface and100 m ranged between 20-500 nM increasing with depth to reach a maximum 1250 nM in anoxic bottomwaters in July 2010. The concentration of H2S ranged between 2-20 µM in the anoxic waters. Over thisperiod, the highest NO2- concentration (1.8 µM) was observed at 120 m in July 2010 (Fig.C.1).675.3.2 Microbial community structureTo identify microbial agents driving methane oxidation in the Saanich Inlet water column I analyzed SSUrRNA gene pyrotag sequences from 288 samples collected at 16 depths (10-200 m) over a two year timeperiod between May 2008 and July 2010. Results revealed consistent microbial community partitioningas commonly observed in stratified ecosystems where O2 deficiency is associated with redox-driven nichepartitioning (Alldredge and Cohen, 1987; Shanks and Reeder, 1993; Wright et al., 2012). For instance,hierarchical cluster analysis (HCA) resolved three major groups or clusters (AU > 70, 1,000 iterations)associated with oxic (group I), dysoxic-suboxic (group II), and anoxic (group III) water column conditions(Fig. 5.1A). Consistent with oxycline formation, O2 (R2 = 0.80) and NO3- (R2 = 0.63) were negativelycorrelated with groups II and III (Fig. 1B), while positively correlated with NH4 (R2 = 0.26) and H2S (R2 =0.52) (Fig. 5.1B).Overall, microbial community composition was dominated (relative abundance >1%) (Rapp and Gio-vannoni, 2003) by OTUs affiliated with ubiquitous and abundant taxonomic groups previously identifiedin marine O2 deficient environments (Field et al., 1997; Fuhrman and Davis, 1997; Brown and Donachie,2007; Tripp et al., 2008; Lam et al., 2009; Walsh et al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011;Wright et al., 2012) including SAR11, SAR324, Nitrospina, SUP05, Marine Group A and Methylophilaleswithin the bacterial domain, and Thaumarchaeota within the archaeal domain (Fig. 5.2). Consistent withprevious observations in OMZs (Stevens and Ulloa, 2008; Glaubitz et al., 2010), OTUs affiliated withMethylococcaceae and Methylomonas, both canonical type I methanotrophs, were identified among the rarebiosphere (relative abundance<1%) (Sogin et al., 2006). Additionally, OTUs affiliated with methylotrophicbacteria such as Methylobacteriaceae and Methylophaga, were also identified among the rare biosphere (Fig.5.3A).5.3.3 Methanotroph diversity and dynamicsMethanotroph diversity in Saanich Inlet was determined based on recruitment of representative OTU se-quences to full-length SSU rRNA gene reference sequences (full description of reference sequences usedin Methods). Initially, I identified all OTUs with a taxonomic assignment affiliated with methanotrophicbacteria using BLAST-based comparisons conducted in QIIME queried against the Silva and Greengenesreference databases. Sequences affiliated with these OTUs shared 90% identity with Methylococcales refer-ence sequences in Silva, and 90% identity with Methylococcaceae and Methylomonas reference sequencesin Greengenes. Subsequently, further analysis on representative methanotrophic OTUs was conducted us-ing full-length SSU rRNA gene sequences as fragment recruitment platforms. A total of 3,804 sequencesaffiliated with type I methanotrophs were recovered, representing 0.095% of the total pyrotag sequencesgenerated. Sequences clustered at 97% identity resolved into 66 distinct OTUs, 6 of which contained75% of total methanotroph sequences. Similar to full-length SSU sequences, pyrotag OTUs revealed 4subgroups with phylogenetic similarity to cultured and environmental representatives of type I methan-otrophs: Methylococcaceae (4%) (Mou et al., 2008), putative methanotrophic groups OPU1 (22.55%) andOPU3 (26.23%) (Hayashi et al., 2007; Tavormina et al., 2010; Tavormina et al., 2013), and methanotrophicsymbionts (46.89%) (Streams et al., 1997; Dubilier et al., 2008; Petersen and Dubilier, 2009) (Fig. 3).68Figure 5.1: Microbial community partitioning under changing levels of water column O2 deficiency.A) Hierarchical clustering of pyrotag data (May 2008 July 2010) based on Manhattan distance. Clusters aredelimited by O2 concentration range represented by number from I to III: oxic = I, dysoxic- suboxic = II andanoxic =III. Bootstrap values (1000 iterations) are shown in red. B) NMDS depicts microbial communitypartitioning along redox gradient showing correlation with environmental parameters. Samples are depictedin color dots according to O2 concentration.69SAR11RhodobacterSphingomonadalesOther MethylophilalesNitrospinaSAR324 AlteromonadalesSAR86SUP05Other Flavobacteriales Cytophaga OM1HasleaArctic96B-7SAR406CyanobacteriaBacteroidetesActinobacteriaMarine Group A α -proteobacteriaβ -proteobacteria δ -proteobacteria γ -proteobacteriaRelative abundance (%)0 20 40 60Marine Group IIMethanocaldococcaceaeThaumarchaeotaEuryarchaeotaCrenarcheota pISA1BacteriaArchaeaFigure 5.2: Taxonomic composition of OTUs identified in SSU rRNA gene pyrotags between 2008-2010. Abundant (>1% relative abundance) taxa found in pyrotag dataset. The size of each box represents theaverage of relative abundance (>0.01%) calculated from the total number of prokaryotic reads throughoutthe water column over this period. Extended dashed lines (whiskers) represent at the base the lower andupper quartiles (25% and 75%) and at the end the minimum and maximum values encountered. The middleline represents the median.The most abundant OTUs were related to putative methanotrophic OPU1 (OPU 01 = OTU55333; 16.9%),OPU3 (OPU3 01 = OTU6504; 8.4%), and methanotrophic symbionts (Symbiont 01 = OTU39693; 45.5%)(Fig.5.3B), comprising ∼70% of total methanotroph sequences found in the pyrotag datasets. Based on thisinformation I focused our analysis on these three groups.Population dynamics of OPU1 01, OPU3 01 and Symbiont 01 were determined throughout the watercolumn between May 2008 and July 2010. OPU1 01 and OPU3 01 were more abundant under dysoxic (<90 µM O2) and suboxic (<20 µM O2) conditions while Symbiont 01 was more abundant under suboxicand anoxic (<3 µM O2) conditions (Fig. 5.4A). Specifically, OPU1 01 showed two abundance peaks: 1)after the 2008 renewal (September November) reaching up to 0.24% relative abundance under suboxicconditions (150 and 165 m), and 2) during water column stratification (March and April 2009) reachingup to 0.15% relative abundance under oxic conditions (10 85 m). OPU3 01 peaked during stratificationperiods (July August 2008 and March April 2009) reaching up to 0.15% relative abundance under oxic anddysoxic conditions (10 97m), and during the extended stratification in July 2010 reaching up to 0.2% relativeabundance under suboxic conditions (120-135 m) (Fig. 5.4B). Symbiont 01 also showed two abundancepeaks: 1) during the 2008 stratification (May July) reaching up to 0.23% relative abundance, and 2) afterthe 2008 renewal (November to January 2009) reaching up to 0.6% relative abundance under oxic conditions(40-97 m), and 0.55% under suboxic anoxic (165 200 m) conditions (Fig. C.2). Cumulative abundancefor methanotroph OTUs throughout the water column indicated co-occurrence and peak abundance underdysoxic-suboxic water column conditions over time, suggesting that the highest CH4 oxidation activity is70carried out between these depth intervals in Saanich Inlet (Fig. C.2 and C.3).Observed OTU distribution patterns for OPU1 01, OPU3 01 and Symbiont 01 became increasinglycompartmentalized during the extended stratification period. For instance, OPU1 01 was observed un-der dysoxic conditions (90-110 m) and OPU3 01 under dysoxic-suboxic conditions (100-135 m), whileSymbiont 01 under suboxic-anoxic conditions (135-200 m) (Fig. 5.4). These observations point to redox-driven niche partitioning among methanotrophic bacteria with the potential to mediate CH4 oxidation indifferent water column compartments. Given these distribution patterns in relation to measured geochemi-cal profiles i.e OPU3 and NO2- (Fig. 5.4B), I hypothesized the use of alternative terminal electron acceptorsin the CH4 oxidation process.71Figure 5.3: Relative abundance and phylogenetic relationships between Type I and Type II methanotroph OTUs in Saanich Inlet. A) Rare(<1% relative abundance) methanotrophic and methylotrophic taxa found in SSU rRNA gene pyrotag dataset. The size of each box representsthe average of relative abundance (>0.01%) calculated from the total number of prokaryotic reads throughout the water column over this period.Extended dashed lines (whiskers) represent at the base the lower and upper quartiles (25% and 75%) and at the end the minimum and maximumvalues encountered. The middle line represents the median. B) Tree inferred using maximum likelihood implemented in PHYML. The percentage(>70%) of replicates in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. Referencesequences for lineages are shown in black. Representative sequences obtained from Saanich Inlet SSU rRNA gene libraries clustered at 97% similarityare shown in blue. OTUs representative sequences (97% similarity) from pyrotag datasets taxonomically identified as methanotrophs are shown ingreen. Sequences abundance is depicted by colored circles whose circumference indicated the total number of sequences (reads) within the cluster.725.3.4 PmoA diversity and expressionGiven the distribution of OPU1 01, OPU3 01 and Symbiont 01 OTUs in the Saanich Inlet water column Iwas interested in determining functional potential and activity of these groups. This was determined basedon the number of conceptually translated pmoA genes and transcripts found throughout the water columnin February 2010 and the recruitment of water column PmoA sequences to selected reference sequences(see Methods). Initially, I identified all sequences with a functional assignment affiliated with PmoA usingthe MetaPathways functional annotation table output. Similar to observations made for methanotroph OTUabundance, PmoA sequences recovered in the metagenomic and metatranscriptomic datasets were rare, rep-resenting<0.001% (34 sequences) and<0.018% (26 sequences) from the total number of predicted proteinsin the metagenomic and metatranscriptomic datasets, respectively. Representative sequences in pmoA clonelibraries were mostly related to OPU3 (Hayashi et al., 2007), and methanotrophic symbionts within thetype I methanotroph clade. However, all February 2010 metagenomic, and metatrasncriptomic PmoA rep-resentative sequences clustered at 97% similarity were related to OPU3 (Hayashi et al., 2007) (Fig. 5.5).Corresponding metatransciptome RPKM values for pmoA in February 2010 showed differential expressionacross the redox transition zone under dysoxic-suboxic conditions. Transcript expression, depicted as dotssized based on RPKM values (Fig. 5.5), was higher at 100m (mean RPKM = 124, SD = 12.10) than at120 (mean RPKM = 15.25, SD = 7.33) and 135m (mean RPKM = 17.05, SD = 0.58) potentially indicat-ing the boundaries for OPU-mediated CH4 oxidation in the Saanich Inlet under water column stratificationconditions (Fig. 5.5). v5.3.5 Methanotroph niche partitioning and co-ocurrence patternsTo better constrain niche partitioning among methanotrophic bacteria and the potential use of alternativeterminal electron acceptors, i.e NO3- and NO2-, I conducted multivariate linear and beta regression analyseson the time-series data. Multivariate regression allowed me to minimize the possible linear effect of depthon OTU distribution while beta regression allowed me to use environmental and relative abundance data tomodel variable correlations (Kieschnick and McCullough, 2003; Ferrari and Cribari-Neto, 2004). Multi-variate analysis showed significant positive correlation (p < 0.001) between OPU1 01 and OPU3 03 overstratification periods. However, negative correlation (p < 0.01) between OPU1 01 and symbiont 01 wasalso observed during these time periods. In addition, significant positive correlation (p < 0.0001) betweenOPU3 01 and Symbiont 01 was also observed during transitioning to stratification periods (Fig. C.2) (Ta-ble 5.1). Negative binomial regression results on the time-series data indicated OPU1 01 distribution wassignificantly and negatively correlated with O2 (p < 0.05) and CH4 (p= 0.001) and weakly correlated withNO3-, NO2- and H2S. OPU3 01 distribution was significantly and negatively correlated with CH4 (p= 0.001)and NO2- (p= 0.05), and weakly correlated with NO3- and H2S (Table 2, Fig. 4). Symboint 01 distributionwas weakly significantly and negatively correlated with CH4, H2S, NO3-and NO2- (p <0.001), and weaklycorrelated with O2 (Table 5.2).To identify characteristic OTUs occurring under specific water column oxygen conditions, multi-levelindicator species analysis (ISA) was conducted based on groups resolved in HCA. A co-occurrence networkwas then constructed to identify potential interactions with methanotrophic OTUs. Multi-level ISA iden-73Figure 5.4: Time-series observations for methanotrophic OTUs affiliated with OPU1, OPU3, andmethanotrophic symbiont groups. A) Methane contour plot for gas concentration (nM) data throughoutwater column from May 2008 to July 2010. Overlapped is shown the OTUs distribution mean trend through-out water column over time. B) Vertical distribution and relative abundance of methanotrophic OTUs overextended stratification period (June-July 2010). On the right, sparklines depict concentration trend for Oxy-gen (O2), Sulfide (H2S), Nitrate (NO3-) and Nitrite (NO2-).74Figure 5.5: Particulate methane monooxygenase subunit β (pmoA) phylogenetic tree. Topography wasinferred using maximum likelihood on PmoA and AmoA amino acid sequences from reference sequencesincluding pMMO-encoding groups OPU1 and OPU3, and symbionts. Bootstrap values (%) are based on100 replicates and are shown for branches with greater than 70% support. The scale bar represents 0.5substitutions per site. PmoA distribution throughout water column (100- 135 m) is depicted as dots whosesize represents the average RPKM value for February 2010 metatranscriptomic datasets.75Table 5.1: Multivariate regression statistics for methanotroph OTUs. Correlation values amongOPU1 01, OPU3 01 and Symbiont 01 are shown with corresponding pair p-values.Methanotroph OTUPair comparison Correlation p-valueOPU1_01 OPU3_01 0.1507 0.0012OPU1_01 Symbiont_01 -0.0992 0.0114OPU3_01 Symbiont_01 0.2025 0.00016Table 5.2: Negative binomial regression statistics for methanotroph OTUs. Correlation values (esti-mate) for environmental parameters (variable) and standard error are shown with z value indicating evidenceof true correlation. p-values for z (Pr(>z)) test the correlations significance.Methanotroph OTU Variable Estimate Std. Error z value Pr(>|z|) SignifO2 -0.00517 0.002331 -2.219 0.0265 <0.05CH4 -0.00626 0.00189 -3.313 0.000922 <0.001NO3 -0.00715 0.02036 -0.351 0.72558NO2 0.25372 0.32508 0.78 0.4351H2S 0.21858 0.12281 1.78 0.07511O2 0.00185 0.002623 0.707 0.4794CH4 -0.008607 0.00272 -3.165 0.00155 0.001NO3 -0.04083 0.0215 -1.898 0.05766NO2 -0.60275 0.30437 -1.98 0.0476 0.05H2S 0.6963 0.23799 2.926 0.0734O2 0.00138 0.00225 0.616 0.5382CH4 -0.00609 0.00145 -4.209 2.57 e-5 < 0.001NO3 -0.10017 0.02038 -4.911 9.04 e-7 <0.001NO2 -1.3045 0.33521 -3.892 9.95 e-5 <0.001H2S 0.30963 0.09272 3.339 0.0008 <0.001OPU1_01OPU3_01Symbiont_01tified OTUs affiliated with one carbon (C1) utilizing microorganisms including Methylophilales, Methy-lophaga, SAR324, Verrucomicrobia and Planctomycetes. SAR324 and Verrucomicrobia were indicatorsfor oxic and anoxic water column conditions while Methylophilales and Methylophaga were indicators forsuboxic and anoxic water column conditions. The indicator OTUs affiliated with Planctomycetes wereevenly distributed throughout the water column (Table. 5.1 and Table C.3). OPU1 01 was identified as anindicator species for dysoxic-suboxic water column conditions, while no OTUs affiliated with OPU3 01,or Symbiont 01 were identified as indicator species for a particular water column condition. However,OPU3 01 and Symbiont 01, as well as OTU20751, affiliated with the OPU3 group, were identified as indi-cators for combined oxic and dysoxic-suboxic conditions (Table C.2).Co-occurrence analysis based on Bray-Curtis and Spearman correlation values among OTUs, resulted ina microbial network (Fig. 5.6A) partitioned in modules similar to hierarchical clustered groups associated76with oxic (I), dysoxic-suboxic (II and III), and anoxic (IV) water column compartments (Appendix C). In-terestingly, significant positive correlations (CV>0.6, p <0.001), shown as sub-networks (Table C.5), wereobserved among OPU1 01, OPU3 01 and Symbiont 01 with indicator OTUs affiliated with potential C1utilizing microorganisms such as Methylophaga, Methylophilales, SAR324, Verrucomicrobia and Planc-tomycetes, and other ubiquitous OMZ microbes including representative taxa such as Marine Group A,Nitrospina, and SUP05 (Fig. 5.6B, Table S4). To provide further evidence for potential methanotroph inter-actions observed in sub-networks I compared microbial OTUs to SSU rRNA gene sequences recovered fromprevious CH4 microcosm experiments using Saanich Inlet suboxic waters (Sauter et al., 2012) (AppendixC). Most bacterial OTUs (83% of OTUs in sub-networks) were found affiliated (>80% identity) to micro-cosm SSU rRNA sequences that were enriched and active after CH4 addition, and related to Bacteroidetes,Marine Group A, Planctomycetes, Sphingomonadales, Nitrospina, Methylophilales, Methylophaga, SUP05and Verrucomicrobia (Table C.5) reinforcing co-occurrence network observations.5.4 DiscussionThis chapter charts methanotroph diversity, abundance and dynamics in Saanich Inlet, a seasonally anoxicfjord that serves as a model ecosystem for understanding microbial community responses to changing levelsof water column O2 deficiency. Observations encompass an atypical extended water column stratificationperiod in 2010 related to a relatively strong El Nin˜o event (Blunden et al., 2011). This extended stratificationperiod provided an opportunity to observe patterns of redox-driven niche partitioning among methanotrophiccommunity members (Fig. 5.4B), hypothesize the use of alternative electron acceptors i.e NO3- and NO2-for CH4 oxidation, and determine co-occurrence patterns between community members consistent withdifferential modes of metabolic coupling driving C1 metabolism along the redoxcline.Methanotrophic community composition was primarily comprised of rare OTUs affiliated with theOPU1, OPU3, and mussel symbionts. Interestingly, no OTUs affiliated with the NC10 phylum were recov-ered contrasting recent observations by Padilla and colleagues in the Eastern Tropical North Pacific OMZoff the coasts of northern Mexico and the Costa Rica OMZ (Padilla et al., 2016). The lack of detectionof NC10 could reflect differences in water column transport processes or geochemical conditions includingO2, NO2- or H2S concentrations. For instance, Padilla and colleagues suggested that NC10 distribution andabundance depended on persistent anoxic conditions and high CH4 concentrations (>1M) (Padilla et al.,2016). With the exception of NC10, methanotophic OTUs inhabiting Saanich Inlet were consistent withprevious observations in coastal OMZs (Hayashi et al., 2007; Tavormina et al., 2013), open ocean OMZs(Stevens and Ulloa, 2008; Glaubitz et al., 2010) and other O2 deficient marine environments (Elsaied et al.,2004; Dick and Tebo, 2010; Fuchsman et al., 2011; Kessler et al., 2011; Dick et al., 2013; Lke et al., 2016)(Fig. C.5) reinforcing the extensibility of Saanich Inlet time series as a model for understanding microbialcommunity dynamics under water column O2 deficiency.The time-series observations allowed me to observe dynamic abundance and distribution patterns thatchanged as a function of water column redox conditions as supported by multivariate regression analysis.For instance, OPU1, OPU3, and methanotrophic symbiont OTUs tended to co-occur during periods of deepwater renewal (Fig. C.2 and C.3). However, as the water column became increasingly stratified I observed77Figure 5.6: Co-occurrence patterns for methanotrophic and indicator OTUs from SSU rRNA genepyrotag datasets.A) Network based on significant (p <0.001) Bray-Curtis and Spearman correlation val-ues (>0.6) among OTUs that were present in at least 25% (n=72) of the total number of pyrotag samples(n=288). On top the oxygen gradient from oxic (>90 µM O2) to anoxic (<1 µM O2) is shown. B) Hivepanels for methanotrophic OPU1 and OPU3, and symbiont OTUs showing interactions with indicator OTUs(taxa coloured as shown in key). OTUs are distributed based on abundance (log transformed) on the threeaxes from less abundant located closer to the center, to more abundant located towards the end of each line.Nodes are depicted according to taxonomy as indicated in network. All interactions (edges) are shown assolid lines. Direct interactions for methanotrophic OTUs to indicators are shown as solid coloured lines inhive panels.78separation of these OTUs into distinct water column compartments consistent with redox-driven niche par-titioning. Our results expand on previous observations of OPU1 and OPU3 distributions in the Costa RicanOMZ water column where under suboxic conditions (O2 >7 µM) OPU1 was more abundant than OPU3.In comparison, OTUs related to methanotrophic symbionts were more abundant under suboxic conditionswhere O2 concentrations were under the detection limit (4 µM), and under anoxic sulphidic conditions,similar to observations made in other O2 deficient waters including deep-sea hydrothermal vents (Elsaied etal., 2004; Dick and Tebo, 2010).Regression analysis between OPU1, OPU3, methanotrophic symbiont OTUs, and geochemical datasuggested the potential use of alternative electron acceptors including NO3- and NO2-. This statistical ob-servation is consistent with previous enrichment and isolation studies focused on different methanotrophicgroups. For example, Nitrite driven anaerobic methane oxidation has been previously reported for membersof the NC10 candidate division. Although primarily identified in fresh water environments, a recent studyoff the coasts of northern Mexico and Costa Rica reported presence and activity of NC10 in pelagic OMZwaters (Padilla et al., 2016). Although no conclusive rate measurements were provided for methane oxi-dation by NC10, transcripts encoding nitric oxide (NO) reductase involved in NO dismutation to O2 weredetected in association with peak NO2- and CH4 concentrations. The potential for OPU or methanotrophicsymbionts to use NO3- or NO2- to drive CH4 oxidation under suboxic or anoxic conditions (e.g O2 lim-ited methane oxidation by facultative denitrifying aerobic methanotrophs) presents important stoichiometricconsiderations. In order for the reaction to occur at a CH4 concentration equal to 341 nM below 120 m, aminimum of 0.015 nmol NO3- and 0.042 nmol NO2- are required based on the CH4: NO3- and CH4: NO2-consumption ratio reported by Cuba and colleagues for batch reactors amended with CH4, NO3- and NO2-(Cuba et al., 2011). Given the average NO3- and NO2- concentrations of 6.5 and 0.2 µM respectively inSaanich Inlet, this coupled mechanism for CH4 oxidation and NO3-/NO2- reduction is permissive.The use of nitrogen species NO3- and NO2- to drive CH4 oxidation in type I methanotrophs has also beenindicated under bioreactor, microcosms and isolated culture conditions (Cuba et al., 2011; Hernandez et al.,2015; Kits et al., 2015). Cuba and colleagues observed increased CH4 loss in the batch reactors amendedwith NO3- (0.52 mol CH4 g-1 NO3-) or NO2- (0.17 mol CH4 g-1 NO2-) followed by community DGGEprofile indicating an enrichment in Methylomonas sp. SSU rRNA sequences in amended bioreactors (Cubaet al., 2011). Consistent with this observation, Kits and colleagues recently reported that Methylomonasdenitrificans strain FJG1T couples CH4 oxidation to NO3- reduction under O2 limiting anoxic conditions(<50 nM) resulting in N2O production (Kits et al., 2015). In addition, Hernandez and colleagues reportedMethylobacter sp. as a dominant methanotroph encoding respiratory nitrate and nitrite reductase genes un-der dysoxic or suboxic O2 conditions (15-75 µM) in Lake Washington microcosm experiment, indicatingpotential use of NO3- and NO2- as alternative electron acceptors under low O2 conditions (Hernandez etal., 2015). Given the absence of NC10 OTUs in our time series observations it will be of interest to deter-mine if convergent mechanisms of NO dismutation or NO3- reduction are used by OPU or methanotrophicsymbionts.Using time-series measurements, Capelle and colleagues have recently identified a persistent CH4 min-imum at 110 m near the oxic-anoxic interface (Capelle et al., 2017) correlating with the highest CH479oxidation rates (2 nmol L-1 d-1) observed in Saanich Inlet (Ward et al., 1989). These studies suggestedmethanotrophs were more abundant and/or metabolically active in the oxycline than in the upper watercolumn. Based on cumulative methanotroph abundance (2.1-2.8% relative abundance) under suboxic con-ditions (100-150m) it is possible to identify OPU1, OPU3 and methanotrophic symbionts as the primarydrivers of CH4 oxidation in the Saanich Inlet water column (Fig. C.4). Interestingly, CH4 oxidation ratesmeasured in Saanich Inlet are very similar to those reported in the Costa Rica OMZ (2.6 nmol d-1; four-fold nitrite driven anaerobic methane oxidation rate) (Padilla et al., 2016) where OPU1 and OPU3 were themost abundant methanotrophic groups identified. In support of this observation, I recovered pmoA ORFscorresponding to OPU3 during an extended stratification period in February 2010. Interestingly, only OPU3PmoA related sequences were found for this period, probably due to the higher abundance of this group.This was reflected in the relative abundance of OPU3 SSU rRNA sequences (up to 0.15%) when comparedwith those from OPU1 and methanotrophic symbiont groups (∼0.02% relative abundance) suggesting thepotential capability for OPU3 to thrive under extended O2 deficiency by using alternative electron accep-tors. Activity of this group was inferred based on recovery of pmoA transcripts affiliated with OPU3 underdysoxic-suboxic water column conditions throughout 100-135 m depth intervals (Fig. 5.5). Similar observa-tions in the Guaymas Basin indicated that pmoA transcripts were more abundant (∼0.32% from total KEGGannotations in metatranscriptome dataset) in dysoxic deep-sea hydrothermal plume samples (∼27 µM O2)and related to sequences retrieved from the Santa Monica Basin and North Fiji hydrothermal vent field, andhydrocarbon plumes from the Deepwater Horizon oil spill in the Gulf of Mexico (Lesniewski et al., 2012).These sequences form the widely distributed OPU3 group of monooxygenases from uncultivated organisms,which are thought to have important roles in the oxidation of methane in O2 deficient waters (Tavorminaet al., 2010). Based on this information I consider the potential activity of OPU3 under dysoxic-suboxicwater column conditions in Saanich Inlet. Based on the published CH4 oxidation rate (2 nmol L-1 d-1) andthe total volume of the dysoxic-suboxic waters between 100-135m (∼3.250x109 L) over 1 year in SaanichInlet, OPU3 have the potential to consume∼37.96 kg CH4 y-1. One kilogram of CH4 has a radiative forcingequal to 0.48 W m-2 with an atmospheric lifetime of 12 years, and a global warming potential (GWP) of23 (eight-fold CO2 GWP for 100 years) (IPCC, 2013). Thus, methanotrophs affiliated to OPU3 in SaanichInlet filter∼18.2 W m-2 equivalent to 450 years of radiative forcing that could be released to the atmosphereeach year. Although there is more to explore regarding the activity for OPU1 and methanotrophic symbiontsunder different water column O2 conditions, these results highlight the potential role of OPU3 as an impor-tant sink for CH4 along continental margins, and reinforce the extensibility of our time-series observationswith implications for modeling CH4 cycling in expanding OMZs.In addition to methanotroph redox-driven niche partitioning, co-occurrence patterns between C1 uti-lizing microorganisms were observed in the Saanich Inlet water column consistent with overlapping habi-tat, shared niche space or preference indicating potential metabolic interactions. In particular, I observedmethanotroph OTUs correlating with Bacteroidetes, Planctomycetes, and Methylophilales. These resultswere consistent with a previous microcosm study using Saanich Inlet waters amended with CH4 revealingmarked enrichment of SSU rRNA gene sequences affiliated with methanotrophs, Bacteroidetes, Plancto-mycetes, and Methylophilales (Sauter et al., 2012). Interestingly, cooperative metabolism between methan-80otrophic bacteria and potential C1 utilizing microorganisms affiliated with different Bacteroidetes and Methy-lophilales has been recently observed in incubation studies using sediment samples from Lake Washing-ton (Beck et al., 2013; Hernandez et al., 2015). Results indicated a simultaneous response between Bac-teroidetes, Methylophilales and canonical methanotrophs to CH4 addition over a range of O2 concentrations(15-75 µM) (Beck et al., 2013; Hernandez et al., 2015). Although phylogenetic-based observations alonecannot explain underlying mechanisms of metabolite exchange, our co-occurrence observations indicate thatCH4 oxidation in Saanich Inlet likely depend on community-level interactions that support the metabolic re-quirements of methanotrophic agents.5.5 ConclusionI used molecular time series observations in combination with geochemical information to determine themethanotrophic community composition and dynamics in Saanich Inlet revealing three rare OTUs affiliatedwith OPU1, OPU3, and methanotrophic symbiont groups that exhibited redox-driven niche partitioningalong changing water column redox gradients. Moreover, I resolved potential novel metabolic strategiesincluding the use of alternative terminal electron acceptors, and metabolic interactions between C1 utiliz-ing microorganisms supporting CH4 oxidation. Combined, these observations provide important baselineinformation on microbial agents that reduce the flux of climate active trace gases from ocean to atmosphereand support the potential role of OPU1, OPU3, and methanotrophic symbiont groups as a widely distributedpelagic sink for CH4 along continental margins. Looking forward, I recommend expanded use of multi-omic sequencing in combination with process rate measurements to determine coverage of CH4 oxidationpathways including all reactions implicated in biological CH4 transformation and C1 transfer. In addition,isotopic labeling and incubation coupled with gene expression studies should be conducted to link CH4 ox-idation pathways and process rates to specific microbial agents along defined water column redox gradientson regional and global scales to better constrain the CH4 filtering capacity of coastal and open ocean OMZs.81Chapter 6Community-level interactions support CH4oxidation in Saanich Inlet O2-deficientwater columnThis chapter is one of the first surveys that integrates multi-omic sequencing information to resolve community-level interactions for CH4 oxidation under O2 deficiency. Observations here support the use of NO2- for CH4oxidation by methanotrophs affiliated to OPU3, and provide evidence for coupled metabolic interactions be-tween methanotrophs and C1 utilizing microorganisms proposed by co-occurrence patterns in Chapter 5.For instance, CH4 oxidation coupled with NO2- reduction is ignited by OPU3 under low-O2 water-columnconditions, and following a fermentative metabolism OPU3 likely promotes community pathways for car-bon incorporation by releasing organic compounds that are used as carbon source by co-occurring taxa.This chapter expands on available information related to community-level metabolic interactions for gascycling that is relevant to models predicting microbial community responses to ocean deoxygenation and isextensible to biotechnological and engineering approaches using methanotrophic communities.6.1 IntroductionGlobal dissolved oxygen (O2) concentration observations since 1960 (Whitney et al., 2007; Bograd et al.,2008; Stramma et al., 2008; Keeling et al., 2010; Stramma et al., 2010; Helm et al., 2011; Schmidtko etal., 2017) show an ongoing regional decline in oceanic O2 concentrations, or deoxygenation, and a sub-sequent expansion of the mid-depth oxygen-minimum zones (OMZs) (Bopp et al., 2002; Keeling et al.,2010; Keller et al., 2016; Schmidtko et al., 2017). Ocean deoxygenation has potentially broad impacts onnutrient and greenhouse gases, i.e. CH4 cycling in the ocean (Worm et al., 2005; Diaz and Rosenberg, 2008;Vaquer-Sunyer and Duarte, 2008; Stramma et al., 2011). Marine OMZs encompass large reservoirs of CH4(Zhang et al., 2011; Pack et al., 2015). For instance, The Eastern Tropical North Pacific (ETNP) OMZ isboth the largest OMZ (Paulmier and Ruiz-Pino, 2009) and the largest reservoir of oceanic CH4 (Sansone etal., 2001; Reeburgh, 2007; Naqvi et al., 2010) in the world. In OMZs, CH4 accumulates in a functionallyanoxic core surrounded by a layer of hypoxic waters (Thamdrup et al., 2012;Wright et al., 2012). Under82a warming climate, the dissolution of O2 in seawater will decrease, whereas its consumption through res-piration will likely increase (Vzquez-Domnguez et al., 2007) and thermal stratification could become moreintense. Together, these biotic and abiotic changes will thicken OMZs, moving CH4 pools closer to the zoneof atmospheric exchange (Stramma et al., 2008; Keeling et al., 2010; Helm et al., 2011).Pelagic CH4 oxidation in marine environments is a rarely quantified process, but on the margins of anOMZ where CH4 intersects traces of O2, it could be a significant process (Mau et al., 2013). Pelagic aerobicor anaerobic CH4 oxidation processes form a final barrier preventing iCH4 escape to the atmosphere, i.e.aerobic CH4 oxidation has been estimated to consume >50% of CH4 in the water column (Fung et al.,1991; Reeburgh et al., 1991; Blumenberg et al., 2007; Kessler et al., 2011; Heintz et al., 2012). Previoustaxonomic and functional screening studies provide insight into methanotrophic community structure andactivity and suggest that at least some of the classic aerobic methanotroph species may be able to thrive in O2deficient environments by potentially utilizing alternative electron acceptors for their metabolism (Costa etal., 2000; Modin et al., 2007; Stein and Klotz, 2011; Beck et al., 2013; Hernandez et al., 2015). For instance,along continental margins two pMMO- encoding phylogenetic groups termed OPU1 and OPU3 (Hayashi etal., 2007) are commonly recovered in molecular gene surveys targeting aerobic methanotrophs in the OMZs(Hayashi et al., 2007; Tavormina et al., 2010; Tavormina et al., 2013; Knief, 2015; Torres-Beltra´n et al.,2016; Padilla et al., 2017). In addition, expression of pmoCAB for the OPU3 group was first demonstratedin a metatranscriptome from the Guaymas Basin (Lesniewski et al., 2012) and recently, alternative modes ofCH4 oxidation by OPU3 have been observed in the Costa Rica OMZ (Padilla et al., 2017). Genes mediatingdissimilatory nitrate (NO3-) and nitrite (NO2-) reduction were identified in the OPU3 binned genome andshown being transcribed in conjunction with key enzymes catalyzing formaldehyde assimilation, suggestingpartial denitrification linked to CH4 oxidation (Padilla et al., 2017). These observations provide importantbaseline information about methanotrophic agents that reduce the flux of CH4 from ocean to atmosphere.However, additional functional information is necessary to determine microbial community members andreactions implicated in community-level CH4 metabolism and derived carbon assimilation processes, underextended water-column O2 deficiency.Community-level CH4 metabolism is an emerging concept from the anaerobic CH4 oxidation processesthat depend on syntrophic associations due to either energetic constraints or the necessity for toxic inter-mediate removal (Haroon et al., 2013; Skennerton et al., 2017). Similarly, aerobic methanotrophs tend toform communities, and it is apparent that they share carbon from CH4 with other bacteria (Yu and Chis-toserdova, 2017). It has been suggested that compounds such as methanol, formate, acetate, succinate, andpossibly other organic acids released by methanotrophs, can support a broad range of microbes (Kalyuzh-naya et al., 2013); Modin, 2017; Tavormina et al., 2017). For instance, some of the species most com-monly co-occurring with Gammaproteobacterial methanotrophs are non-methane-utilizing methylotrophs,i.e. Methylophilaceae family and other non-methylotrophic bacteria affiliated with Burkholderiales andFlavobacteriales (Beck et al., 2013; Hernandez et al., 2015; Oshkin et al., 2015; Karwautz et al., 2018;Kumaresan et al., 2018). Analyses of both natural populations (Crevecoeur et al., 2015; Karwautz et al.,2018; Kumaresan et al., 2018), and manipulated laboratory microcosms (Beck et al., 2013; Hernandez etal., 2015; Oshkin et al., 2015) suggest that these partnerships in CH4 metabolism might not be random.83Although the communal nature of CH4 metabolism provides a new outlook on the environmental role ofthe methanotrophs as essential components of food webs driven by carbon from CH4, it is necessary todetermine how resilient these interactions are to low-O2 water-column conditions.Saanich Inlet is a seasonally anoxic fjord on the east coast of Vancouver Island British Columbia. Theseasonally stratified water column of Saanich Inlet serves as a well-characterized model ecosystem for exam-ining how deoxygenation shapes microbial community population dynamics and interactions along definedredox gradients in the ocean (Walsh et al., 2009; Zaikova et al., 2010; Walsh and Hallam, 2011; Wrightet al., 2012; Hawley et al., 2014; Louca et al., 2016; Torres-Beltra´n et al., 2016). Recent observationscharting methanotroph diversity, abundance and dynamics in Saanich Inlet, indicated CH4 oxidation underlow-O2 water-column conditions likely depend on community-level interactions between OPU3 and taxaaffiliated with Bacteroidetes, Planctomycetes, and Methylophilales (Torres-Beltra´n et al., 2016). In thischapter, I aim to resolve community-level interactions controlling CH4 oxidation occurring in the SaanichInlet O2-deficient water-column. I coupled incubation experiments and long-term monitoring surveys thatintegrate multi-omic information providing a promising environmental context to link CH4 oxidation path-ways to specific microbial agents along defined low-O2 water-column conditions. These observations in-dicate methanotrophs play an important role in global metabolic cycles beyond the CH4 cycle and provideevidence that expands our understanding of CH4 oxidation under low-O2 water-column conditions.6.2 Methods6.2.1 Incubations implementationTo explore microbial community-level interactions controlling CH4 oxidation in the Saanich Inlet watercolumn, I carried out an incubation experiment to test for CH4 oxidation coupled with NO2- reductionunder dysoxic-suboxic (<30- ∼3 µM O2) water-column conditions. The rational behind this approach wasbased on recent observations suggesting potential novel metabolic strategies including the use of alternativeterminal electron acceptors i.e. NO2- by methanotrophs affiliated to OPU3 under low-O2 water-columnconditions (Torres-Beltra´n et al., 2016; Padilla et al., 2017). I applied a multi-omic sequencing approach(small subunit (SSU) rDNA and rRNA pyrotags, metagenomic, metatranscriptomic and metaproteomic) toinfer metabolic interactions related to CH4 oxidation under low-O2 conditions (Fig. 6.1).Incubations were carried out on 12 L seawater samples collected on February 15, 2015, from two depths(100 and 150 m) selected based on O2 concentration spanning the upper and lower boundaries of the dysoxic-suboxic (<30- ∼3 µM O2) water-column conditions. For each depth, sample water was collected from two12 L Go-Flow bottles through silicon tubing (∼15 cm long and 1/4” thick, pre-flushed for a few secondswith sample water) into 12 L metallic bags (pre-flushed with Helium (He)). Bags were overfilled to removeany air bubbles from the tubing during filling and sealed right after. Five bags were filled at each depthcorresponding to an environmental control (background), an experimental control (no substrate addition),12CH4, 12CH4 + NO2- and 13CH4 + NO2-. Background samples for RNA (2 L) and protein (2 L) wereprocessed on-ship immediately after sampling while the rest were stored in the dark on ice until processingin the laboratory.8420015010010Depth (m)5012L12/13CH4 (+ NO2-)RNArDNAPyrotagsPyrotagsrRNAMetagenomicsMetatranscriptomicsMetaproteomicsBulk 13C + nanoSIMS  QIIMEQIIMEMetapathways Taxonomic and functional DBMap to DBTaxa and genes IDFunction linked to taxa and genesSubstrateincorporationMoleculeProteinExperimental approachBioinformatics Final outputO2NO3-NO2-H2SCH4D1D2Figure 6.1: Experimental workflow for CH4 incubation experiments carried out in February 2015.Water samples (12 L) for DNA, RNA and protein were taken at 100 (D1) and 150 m (D2) encompassingthe dysoxic-suboxic water-column compartments. Growth conditions included the addition of CH4 and12,13CH4, plus NO2-, and incubation for 72 h. A multi-omic approach (gray) including 16S pyrotag se-quencing, metagenomics, metatranscriptomics, metaproteomics coupled with bulk 13C measurements andnanoSIMS was carried out to obtain the final outputs (green), such as taxa, gene identification and func-tional information derived from substrate incorporation into cell biomass through the bioinformatics tools(orange). The environmental parameters profile is depicted as sparklines for oxygen (O2= light blue), nitrate(NO3-= orange), nitrite (NO2-= black), methane (CH4= red) and hydrogen sulfide (H2S= purple).Substrates were injected into bags through the airlock using a He-flushed glass syringe, and mixed usinga plate shaker to allow substrate dissolution in the sample. Substrates were added in a 10% increase of thehighest environmental concentration at each depth corresponding to 72 nM and 2 µM CH4, and 0.03 and0.1 µM NO2- for the 100 and 150 m, respectively. Water subsamples were taken before (t0; DNA (2L)) andafter 72 hour incubation (t72; DNA (4L), RNA (2L) and Protein (2L)). Subsamples were filtered onto a 0.22µm Sterivex filter as previously described for the Saanich Inlet time-series multi-omic datasets (Hawley etal., 2017b).Biomass to generate time-series metagenomic datasets was collected from June 2009 to August 2011 atsix depths (10, 100, 120, 135, 150 and 200 m) and filtered with an in-line 2.7 µm GDF glass fibre pre-filteronto a 0.22 µm Sterivex polycarbonate cartridge filter (Hawley et al., 2017b).6.2.2 Carbon incorporation measurementsTo first assess carbon incorporation into microbial community biomass I used 13CH4 samples and conductedtwo complementary isotopic approaches. I measured the carbon isotopic composition (δ 13C) of cell pellets(Bulk) and individual cells.Bulk δ 13C on cell material pellets was measured using an elemental analyzer (EA) from Costech An-alytical Technologies, Inc. (Valencia, CA) coupled to a Thermo Scientific (Bremen, Germany) Delta VPlus isotope ratio mass spectrometer (IRMS) at the Pacific Northwest National Laboratory (PNNL). Dueto the small size of the cell pellets, they were suspended in an aqueous buffer to facilitate their transferinto pre-weighed tin capsules (Costech Analytical Technologies, Inc.) used for EA-IRMS analysis. I driedthe samples then re-weighed the capsules to determine the total amount of biomass used for each analysis85(2-15 mg). The EA combustion reactor was loaded with cobaltic oxide and chromium oxide catalyst andmaintained at 1,020 oC while the combustion reactor was loaded with copper catalyst and maintained at650 oC. I used two in-house glutamic acid standards that were themselves calibrated against USGS 40 andUSGS 41 standards (δ 13C = -26.39 o/ooVPBD and 37.63 o/ooVPDB respectively) and applied a two-point,slope intercept correction to the data. All δ 13C results are referenced to Vienna Pee Dee Belemnite and arereported in delta notation:δ 13C = (RsampleRstandard−1)∗1000 (6.1)where Rsample is the 13C to 12C ratio of the measured sample and Rstandard is the 13C to 12C ratio of ViennePee Dee Belemnite (0.0112372). Total δ 13C parts per mil (o/oo) were obtained for each sample. Carbonincorporation (δ 13C) into individual cells was measured using a 50L Cameca. Cells were fixed onto 0.2 µmprecombusted polycarbonate filters (5 mL sample) and DAPI (1 µg ml-1 final concentration, Sigma) stainedfor field identification under the instrument microscope. Controls and labeled samples were measured for13C incorporation relative to the standard. δ 13C values of enriched cells (15-25 per field) were exported tothe R software package (RCoreTeam, 2013) for results visualization.6.2.3 Nucleic acid and protein extractionsGenomic DNA was extracted from Sterivex filters as previously described (Zaikova et al., 2010; Hawley etal., 2017b). Briefly, after thawing the filter cartridge on ice, 50 µl of 0.125 mg ml-1 lysozyme was addedand incubated at 37 oC for 1 h with rotation followed by addition of 50 µl Proteinase K (Sigma) and 100µl 20% SDS and incubated at 55 oC for 1 h with rotation. Lysate was removed by pushing through witha 3 ml syringe followed by rinsing with an additional 1 mL of sucrose lysis buffer. Filtrate was subjectto chloroform extraction and the aqueous layer was collected onto a 10K 15 ml Amicon filter cartridge,washed three times with TE buffer (pH 8.0) and concentrated to a final volume between 150-400 µl. TotalDNA concentration was determined by PicoGreen assay (Life Technologies) and genomic DNA qualitydetermined by visualization on 0.8% agarose gel run (16h/15V).Total RNA was extracted from Sterivex filters as previously described for the Saanich Inlet time-series(Hawley et al., 2017b). Briefly, after thawing the Sterivex cartridge on ice RNA later was removed by push-ing through with a 3 ml syringe followed by rinsing with an additional 1.8 mL of Ringer’s solution andincubated at room temperature for 20min with rotation. Ringer’s solution was removed by pushing throughwith a 3 ml syringe followed by adding 1.8 ml MirVana Lysis Buffer and 100 µl of 0.125 mg ml-1 lysozymeand incubated at 37 oC for 30 min with rotation. Lysate was removed from Sterivex and subjected to organicextraction following the mirVana kit protocol. DNA removal and clean up and purification of total RNAwere conducted following the TURBO DNA-free kit (Thermo Fisher Scientific) and the RNeasy MinEluteCleanup kit (Qiagen) protocols respectively. Total RNA concentration was determined by RiboGreen analy-sis (Life Technologies) prior to synthesize first strand cDNA using the SuperScript III First-Strand SynthesisSystem for RT-qPCR (Invitrogen) according to manufacturer instructions.Total protein was extracted in the Environmental Molecular Sciences Laboratory (EMSL) at PNNL from86Sterivex as previously described for the Saanich Inlet time-series proteomic datatsets (Hawlet et al., 2014;Hawley et al., 2017b). Briefly, after thawing Sterivex filters on ice, Bugbuster (Novagen) was added andincubated at room temperature for 20-30 min with rotation. Lysate was removed by extrusion and filterswere rinsed with 1 ml lysis buffer. Buffer exchange was carried out on combined lysate using AmiconUltra 10K (Millipore) with 100 mM NH4HCO3 a total of three times with a final volume between 200-500µl. Protein concentration was determined using the 2-(4-carboxyquinolin-2-yl) quinoline-4-carboxylic acid(Bicinchoninic acid or BCA) assay. Urea was added to a final concentration of 8 M and dithiothreitol addedto a final concentration of 5 mM and incubated at 60 oC for 30 min, followed by 10-fold dilution with 100mM NH4HCO3. Samples were then subject to trypsin digest at 37 oC for 6 h followed by C18 solid phaseextraction and strong cation exchange.6.2.4 Nucleic acid tag sequencingExtracted DNA and cDNA was used to generate small subunit (SSU) rDNA and rRNA pyrotags with three-domain resolution. PCR amplification procedures were carried out as previously described for the SaanichInlet time-series multi-omic data (Hawley et al., 2017b)., Pyrotag libraries were generated by PCR amplifi-cation using multi-domain primers targeting the V6-V8 region of the SSU rRNA gene (Allers et al., 2013):926F (5’-cct atc ccc tgt gtg cct tgg cag tct cag AAA CTY AAA KGA ATT GRC GG-3’) and 1392R (5’-ccatct cat ccc tgc gtg tct ccg act cag-<XXXXX>-ACG GGC GGT GTG TRC-3’). Primer sequences weremodified by the addition of 454 A or B adapter sequences (lower case). In addition, the reverse primerincluded a 5 bp barcode designated <XXXXX> for multiplexing of samples during sequencing.Twenty-five microliter PCR reactions were performed in triplicate and pooled to minimize PCR bias.Each reaction contained between 1 and 10 ng of target DNA, 0.5 µl Taq DNA polymerase (Bioshop inc.Canada), 2.5 µL Bioshop 10 x buffer, 1.5 µL 25 mM Bioshop MgCl2, 2.5 µL 10 mM dNTPs (Agilent Tech-nologies) and 0.5 µL 10 mM of each primer. The thermal cycler protocol started with an initial denaturationat 95 oC for 3 minutes and then 25 cycles of 30 s at 95 oC, 45 s at 55 oC, 90 s at 72 oC and 45 s at 55 oC.Final extension at 72 oC for 10 min. PCR products were purified using the QiaQuick PCR purification kit(Qiagen), eluted elution buffer (25 µL), and quantified using PicoGreen assay (Life Technologies). SSUrDNA and rRNA amplicons were pooled at 100 ng for each sample. Emulsion PCR and sequencing ofthe PCR amplicons were sequenced on Roche 454 GS FLX Titanium at the McGill University and GnomeQubec Innovation Center.6.2.5 Meta-genomic, transcriptomic and proteomic sequencingExtracted total DNA was used to generate Illumina paired-end metagenomic datasets at GeneWiz sequencedon the Illumina HiSeq platform. Extracted total DNA from time-series samples, were used to generatemetagenomic datasets at the DOE-JGI following the protocols for library production, sequencing and as-sembly previously described for the Saanich Inlet time-series (Hawley et al., 2017b). Extracted total RNAwas used to generate paired end metatranscriptomic datasets at EMSL-PNNL sequenced on the Ion PI Hi-Qplatform.Extracted total protein samples were used to generate metaproteomic datasets at EMSL-PNNL. Peptides87were analyzed by tandem MS (MS/MS), as previously described (Hawley et al., 2013; Hawley et al., 2014),using online capillary LC/MS/MS on a Thermo LTQ ion trap using data-dependent fragmentation. Detectedpeptides were identified from MS/MS using SEQUEST with a mass spectra generating function (MS-GF)cutoff value less than 10-11, corresponding to a false discovery rate of less than 2% (Kim et al., 2008).6.2.6 Tag sequence analysesIn order to relate community composition with potential activity, combined SSU rDNA and rRNA pyrotagsequences were analyzed using the Quantitative Insights Into Microbial Ecology (QIIME) software package(Caporaso et al., 2010). Reads with length shorter than 200 bases, ambiguous bases, and homopolymerruns were removed before to chimera detection. Chimeras were detected using the chimera slayer providedin the QIIME software package and removed before taxonomic analysis. A total of 564,694 non-chimericsequences were clustered at 97% identity into operational taxonomic units (OTUs). Prior to taxonomicassignment, singleton OTUs (OTUs represented by one read) were omitted (Engelbrektson et al., 2010)leaving 19,508 OTU sequences. Representative sequences from each non-singleton OTU were queriedagainst the SILVA database release 111 using the BLAST algorithm (Altschul et al., 1990). To furtherresolve the diversity of methanotrophic bacterial OTUs, I recruited pyrotag OTU sequences using a BLAST-comparison (>99% identity) to full-length SSU rRNA gene reference sequences including cultured andenvironmental Type I and II methanotrophs and pyrotag SSU rRNA gene sequences previously observed inthe Saanich Inlet water column (Torres-Beltra´n et al., 2016).The SSU rRNA:rDNA abundance ratio was calculated to account for variation in taxon abundance in theDNA pool (Frias-Lopez et al., 2008; Stewart et al., 2012b) and compared for a subset of microbial groups tounderstand how incubation treatments influence changes of potentially active OTUs across treatments. Sta-tistical analyses were conducted using the R software package (RCoreTeam, 2013). Pyrotag datasets werenormalized to the total number of reads per sample. Hierarchical cluster analysis (HCA) was conducted toidentify community compositional profiles associated with incubation treatments using the pvclust (Suzukiand Shimodaira, 2015) package with the Manhattan Distance measure, and statistical significance to the re-sulting clusters as bootstrap score distributions with 1,000 iterations. Further data curation and visualizationwere conducted using the dplyr (Wickham et al., 2015) and ggplot2 (Wickham, 2009) packages.To further frame my observations with the previously described Saanich Inlet methanotrophic co-occurrencenetworks (Torres-Beltra´n et al., 2016), I recruited incubations OTUs to time-series HR SSU rDNA pyrotagsequences (Torres-Beltra´n et al., 2016) using BLAST-comparison (>99% identity). The time-series refer-ence sequences included methanotrophic OTUs affiliated with the pMMO- encoding phylogenetic groupsOPU1 and OPU3, and methanotrophic symbionts that showed significant correlations with alternative ter-minal electron acceptors, i.e. NO2- and taxa involved in one-carbon (C1) metabolism under suboxic water-column conditions. Only hits with a perfect match of a query sequence were selected and mapped to thereference co-occurrence network to identify potential microbial hubs carrying out methane oxidation ig-nited by substrate addition. Nodes were selected and colored based on taxonomy in an already existingco-occurrence network (Torres-Beltra´n et al., 2016) using Cytoscape 2.8.3 (Shannon et al., 2003). Nodessize was depicted based on SSU rRNA:rDNA ratio values, showing only those OTUs with a ratio higher88than 1 that suggested potential activity under specific substrate conditions.6.2.7 Multi-omic dataset analysisIncubation metagenomic data were analyzed using MetaPathways V2.5.1, an open source pipeline for pre-dicting reactions and pathways using default settings (Konwar et al., 2013; https://github.com/hallamlab/metapathways2/wiki), as previously described for Saanich Inlet time-series samples (Torres-Beltra´n et al.,2016; Hawley et al., 2017a). Time-series metagenomic data was analysed using MetaPathways V2.5.1. Foreach gene, reads per kilobase per million mapped (RPKM) was calculated as a proportion of the number ofreads mapped to a sequence section, normalized for sequencing depth and open reading frame (ORF) length(Konwar et al., 2015). RPKM values were used to describe the abundance of genes. To cross-referenceincubation metagenomic datasets with the Saanich Inlet time-series metagenomic observations, conceptu-ally translated amino acid sequences of predicted ORFs were LAST+ compared to time-series ORFs se-quences. Sequence matches with higher than 70% identity were retrieved from the functional annotationtable <ORF annotation table.txt> in the <results/annotation table> output directory and used to identifyoccurrence and abundance of marker CH4 oxidation and NO2- reduction genes, i.e. particulate methanemonooxygenase subunit β (pmoA) and copper-containing nitrite reductase (nirK) in incubation samples.Metagenomic information from incubation samples was included into the time-series reference database tobe used for mapping and comparing levels of expression in metatranscriptomic and metaproteomic datasetsdescribed below.To incorporate PmoA sequences found in incubation datasets into a reference phylogenetic tree, concep-tually translated and annotated PmoA ORFs from the incubations metagenomic datasets were manually ex-tracted from the functional annotation table<ORF annotation table.txt> in the<results/annotation table>output directory. Sequences were aligned and compared to diverse environmental and reference PmoAsequences (including OPU3) previously used to generate a phylogenetic reference tree for PmoA in theSaanich Inlet (Torres-Beltra´n et al., 2016). A total of 104 PmoA sequences were clustered and representa-tive sequences aligned as previously described to infer a maximum likelihood phylogenetic tree that includedPmoA sequences observed in incubation samples. Similarly, to generate a phylogenetic reference tree forNirK, conceptually translated and annotated ORFs from the incubations metagenomic datasets were manu-ally extracted from the functional annotation table <ORF annotation table.txt> in the <results/annotationtable>Metapathways output directory. Sequences were aligned and compared to diverse environmental ref-erence NirK sequences. A total of 1,818 NirK sequences were clustered over a range of identity thresholdsusing the UClust algorithm (USEARCH V6.0) with 28 reference sequences, including cultured and envi-ronmental sequences affiliated with Bacteroidetes, Choloflexi, Chlorobi, Nitrospinaceae, Methylococcales(including OPU3), Planctomyces, and Thaumarchaeota. The 70% identity threshold was selected based onthe resolution of the taxonomic groups. Cluster representative sequences were aligned using the MultipleSequence Comparison by the Log- Expectation (MUSCLE) method (EMBL-EBI) and were manually cu-rated in Mesquite. A maximum likelihood phylogenetic tree was inferred using PHYML (Guindon et al.,2005) based on a WAG model of amino acid evolution where the parameter of the gamma distribution, theproportion of invariable sites and the transition/transversion ratio were estimated. The confidence of each89node was determined by assembling a consensus tree of 1000 bootstrap replicates.Metatranscriptomic data was checked for read quality and sequence expansion using the htseq-qa com-mand on the HTSeq python package (Anders et al., 2015). A total of 10,268,073 reads were aligned toreference metagenomic database using Torrent Mapping Alignment Program (TMAP) ((https://github.com/iontorrent/TS/tree/master/Analysis/TMAP)) using the map4 command with soft clipping enabled at bothends of the reads and the alignment output mode set to return a random alignment among those with the bestscore if multiple alignments receive the best score. Aligned reads were mapped to reference metagenomicdatabase using the htseq-count command in the HTSeq python package (Anders et al., 2015). Annotationwas converted to a suitable .gtf format using R costumed scripts. Expression analysis was performed us-ing the Bioconductor package DESeq2 (Anders and Huber, 2010). Downstream enrichment analysis wasperformed using the modified Fisher exact test (Hosack et al., 2003) using R costumed scripts.Proteomic data consisted of a total of 159,413 peptides. Peptides were searched against the metage-nomic reference database. Only peptides matched to protein sequences with a peptide prophet probability(PPP) score >0.95 were used in further analysis. A total of 26,828 unique proteins were identified. Tobetter characterize the metabolic processes within the microbial community for CH4 oxidation under in-cubations conditions, I surveyed the expression level of a given gene transcript and later the amount ofprotein produced in pseudo-quantitative manner. First, I calculated a normalised spectral abundance factor(NSAF)(Hawley et al., 2013; Hawlet et al., 2014). Second, I calculated the expression ratio by dividing theNSAF value of a given protein in a treatment (t72) over the control (t72). Ratio values higher than 1 wereconsidered as upregulated gene expression due to incubation conditions.To further resolve the diversity of PmoA and NirK sequences observed in the metatranscriptomic andproteomic datasets, I retrieved transcripts and peptides that were annotated as PmoA and NirK based onORF mapping to the metagenomic database and searched for their corresponding ORF in the gene clusteringanalysis conducted to build the phylogenetic trees. All PmoA and NirK transcripts and peptides found wereassigned to an ORF within a cluster represented in their corresponding phylogenetic tree.6.2.8 Pathways network and metabolic model inferenceTo generate a robust network emphasizing co-occurrences between prevalent pathways in the dysoxic-suboxic water-column compartment, Bray-Curtis and Spearman’s rank correlations were used on time-series metagenomic data. Correlation coefficients were calculated using CoNet (Faust et al., 2012) withpathway abundance as RPKM data. First, a pathways matrix was constructed using the environmental Path-ways/Genome Database (ePGDB) file <SAMPLE.pwy.txt> in the <results/pgdb> Metapathways outputdirectory for each time-series metagenomic sample. The matrix was transformed into presence-absencedata to remove pathways with less than 1/3 zero counts, leaving a matrix of 380 pathways for all samples.Next, to construct ensemble networks, measure-specific thresholds set to 0.6 were used as a pre-filter, andedge scores were computed only between clade pairs. To assign statistical significance to the resultingscores, edge and measure-specific permutation and bootstrap score distributions with 1,000 iterations eachwere computed. P-values were tail-adjusted so that low p-values correspond to co-presence and high p-values to exclusion. After merging, p-values on each final edge were corrected to q-values (cut-off of 0.05).90The positivity or negativity of each relationship was determined by consensus voting over all the integrateddata sources. Finally, only edges with at least two supporting pieces of evidence were retained.The final edge matrix was visualized as a force directed network using Cytoscape 2.8.3 (Shannon etal., 2003). Network properties were calculated with the “Network Analysis” Plug-In. Nodes in the co-occurrence network corresponded to individual pathways and edges were defined by computed correlationsbetween corresponding pathway pairs. The layout revealed distinct modules, which persisted after lower-ing the correlation coefficient cut-off for edge creation to 0.90 reinforcing the robustness of the network.Nodes and edges from modules were selected and visualized as sub-networks using the tool Hive PanelExplorer (https://github.com/hallamlab/HivePanelExplorer/wiki) (Perez, 2015). HivePlotter allows for edgeselection based on interactions within specific pathways, such as those affiliated with CH4 oxidation and C1metabolism. ORFs within each selected pathway in a sub-network were searched for their functional andtaxonomic affiliation with incubation multi-omic datasets. Pathways with ORFs showing functional repre-sentation in incubation experiments were colored in hive plots according to ORFs taxonomic affiliation.6.2.9 Data depositionThe SSU rDNA and rRNA pyrotag sequences reported in this chapter have been deposited in the NCBIunder the accession numbers: SRX2160172-SRX2160414. Metagenomes reported in this chapter havebeen deposited in the NCBI under the Biosample accession numbers: SAMN08283972- SAMN08283975.Metatranscriptomes reported in this chapter have been deposited in the NCBI under the Biosample ac-cession numbers: SAMN10425508-SAMN10425517. Metaproteomes reported in this chapter have beendeposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifierPXD011287 and 10.6019/PXD011287.6.3 Results6.3.1 Water column conditionsBeginning in February 2015- water column O2 deficiency intensified below 100 m and was consistent withthe onset of the stratification period in the Saanich Inlet. The beginning of stratification corresponded withthe intensification of suboxic conditions below 90 m and increasing levels of hydrogen sulphide (H2S) below150 m, which was consistent with the development of deep-water anoxia. I sampled two representativedepths corresponding to the upper (100 m; 29 O2 µM) and lower (150 m; 3.34 O2 µM) boundaries ofsuboxic water-column conditions. Methane concentrations ranged from 72 nM at 100 m to 2 µM at 150 m.Hydrogen sulphide was only detected below 150 m, peaking in concentration (26 µM) at the basin. Theconcentration of NO3- ranged from 27 µM at 100 to 0.055 µM at 150 m. Similarly, the concentration ofNO2- ranged from 0.03 µM at 100 m to 0.095 µM at 150 m.91−200 0 200400−500050010001500100mδ13C (ppm)δ15 N (ppm)−200 0 20040005001000150mδ15 N (ppm)δ13C (ppm)Environmental control bulk protein δ13C (ppm)Incubation bulk protein δ13C (ppm)nanoSIMS δ13C/δ15N counts ±1SDΔ156.1 Δ19.5Figure 6.2: Labeled substrate incorporation into cellular biomass. Cellular nanoSIMS δ 13C/δ 15N(ppm) counts for 100 and 150 m 13CH4 + NO2- incubation samples. Isotopic values are depicted as dots±1 SD. Environmental bulk protein δ 13C (ppm) is shown as a dotted red line while incubation sample bulkprotein δ 13C (ppm) is shown as a solid red line. Total sample δ 13C enrichment (∆) is shown between theenvironmental control and sample bulk protein values.6.3.2 Carbon incorporation into biomassTo evaluate carbon incorporation into community biomass derived from CH4 addition, I measured the iso-topic carbon composition (δ 13C) from cell material (pellets and individual cells). Bulk δ 13C measurementof cell pellets showed higher carbon incorporation in 100m (181.1 o/oo) samples than in environmental con-trol (100 m = -25 o/oo and 150 m = -31.2 o/oo) and 150 m (11.7 o/oo) samples (Fig. 6.2). In addition,individual cell δ 13C values were higher in 100 m samples (1.66- 26.09 o/oo) than 150 m samples (0.02- 11.4o/oo) (Fig. 6.2). Combined, δ 13C measurements suggest higher metabolic activity in 100 m samples withregard to carbon incorporation from CH4 oxidation.6.3.3 Microbial community compositionTo initially identify potential microbial agents carrying out CH4 oxidation coupled with NO2- reductionunder dysoxic-suboxic water column conditions, I analyzed incubation SSU rDNA pyrotag sequences in-cluding the environmental and experimental controls, and incubation treatments (12CH4, 12CH4 + NO2-and 13CH4 + NO2-) from 100 and 150 m. Results revealed microbial community partitioning associatedwith depth and incubation treatment. For instance, hierarchical cluster analysis (HCA) resolved two majorgroups, or clusters (AU > 70, 1,000 iterations), associated with 100 m (group I) and 150 m (group II), inwhich incubation treatment differentiate from controls (environmental and experimental) (Fig. 6.3).Overall, microbial community composition was primarily constituted by OTUs (average relative abun-dance >1% across samples) affiliated with ubiquitous and abundant taxonomic groups previously identified920102030405060Manhattan Distance877778 100989999767910099100100CH4NO2CH413CH4NO2Environmental controlExperimental control100m 150mDNARNA## Bootstrap valueFigure 6.3: Microbial community partitioning based on depth and incubation treatments. Hierarchicalclustering of pyrotag data (rDNA and rRNA) based on Manhattan distance. Clusters are delimited by depth(100 and 150 m). Samples are depicted in color dots according to incubation treatment. Bootstrap values(1000 iterations) are shown in red.in marine surveys, including the Saanich Inlet, (Field et al., 1997; Fuhrman and Davis, 1997; Brown andDonachie, 2007; Tripp et al., 2008; Lam et al., 2009; Walsh et al., 2009; Zaikova et al., 2010; Walshand Hallam, 2011; Wright et al., 2012) as Bacteroidetes, Chloroflexi, Rhodospirillaceae and Pelagibacter-aceae within the Alphaproteobacteria, Methylophilales within the Betaproteobacteria, Methylococcales andSUP05 within the Gammaproteobacteria, Plactomycetes, Marine Group A, and Thaumarchaeota (Fig. 6.4).Methanotroph diversity in incubation samples was determined based on the recruitment of representa-tive OTU sequences to full-length and pyrotag SSU rRNA gene reference sequences as previously described(Torres-Beltra´n et al., 2016). I identified 339 OTU sequences affiliated with Methylococcales (2% of thetotal rDNA pyrotag dataset) based on BLAST comparisons against the SILVA database (>97% identity)in the incubation samples. Subsequently, I conducted further analysis on representative methanotrophicOTUs affiliated with Methylococcales using full-length SSU rRNA gene sequences as fragment recruitmentplatforms (≥99% identity). In all, I identified 4 subgroups with phylogenetic similarity to OPU1 (48%)and OPU3 (37%) (Hayashi et al., 2007; Tavormina et al., 2010; Tavormina et al., 2013), methanotrophicsymbionts (7%) (Streams et al., 1997; Dubilier et al., 2008; Petersen and Dubilier, 2009) and type I methan-otrophs (8%) (Fig. D.1).93VibrionalesVerrucomicrobiaRhodospirillalesAcidimicrobialesRhodobacteraceaeBetaproteobacteriaHTCC2188EuryarchaeotaH−178LentisphaeraeOD1OM190CampylobacteralesWS3CrenarchaeotaAnaerolineaeFirmicutesOP11CandidatusScalinduaPhycisphaeraePirellulaceaeRickettsiaceaeDesulfobulbaceaeDesulfobacteraceaeChromatialesLegionellalesMethylococcalesSpirochaetesThaumarchaeotaThermoplasmataActinobacteriaBacteroidalesFlavobacteriaFlavobacterialesSphingobacterialesSAR202GemmatimonadetesPla3PlanctomycetaceaeRhodospirillaceaePelagibacteraceaeMethylophilalesDeltaproteobacteriaNitrospinaceaeSAR324GammaproteobacteriaAlteromonadalesOceanospirillalesSUP05ThiohalorhabdalesThiotrichalesOther BacteriaActinobacteriaBacteroidetesCandidate divisionα-proteobacteriaβ-proteobacteriaδ-proteobacteriaγ-proteobacteriaPlanctomycetesEuryarchaeotaCH4NO2CH413CH4NO2Environmental controlExperimental control1020401rDNA Relative abundance (%)100m 150mMarine Group AFigure 6.4: Taxonomic composition of OTUs identified in SSU rDNA pyrotags among treatments.Abundant (>1% relative abundance) taxa found in pyrotag datasets. The size of each dot represents taxarelative abundance calculated from the total number of prokaryotic reads in each treatment sample. Sampletreatments are depicted as symbols: environmental control = open square, experimental control = solidsquare, 12CH4 = open circle, 12CH4 + NO2- = gray circle, and 13CH4 + NO2- = solid circle.94TM6RhizobialesBurkholderialesOP9ZB3MyxococcalesVibrionalesVerrucomicrobiaAcidimicrobialesRhodobacteraceaeHTCC2188OD1OM190WS3AnaerolineaePhycisphaeraePirellulaceaeDesulfobacteraceaeChromatialesLegionellalesMethylococcalesThaumarchaeotaThermoplasmataBacteroidalesFlavobacteriaFlavobacterialesSAR202PlanctomycetaceaeMethylophilalesNitrospinaceaeSAR324AlteromonadalesOceanospirillalesSUP05ThiotrichalesBacteroidetesCandidate divisionα-proteobacteriaβ-proteobacteriaδ-proteobacteriaγ-proteobacteriaPlanctomycetesMarine Group AEuryarchaeota10x30x1x100m 150mCH4NO2CH413CH4NO2rRNA : rDNA SpirochaetesLentisphaeraeGemmatimonadetesThiohalorhabdalesFirmicutesCrenarchaeotaRhodospirillalesRhodospirillaceaeActinobacteriaActinobacteriaExperimental controlFigure 6.5: Taxonomic composition of OTUs identified as potentially active taxa under incubationtreatments. Active taxa (rRNA:rDNA >1) found in SSU pyrotag datasets. The size of each dot representsthe rRNA:rDNA ratio calculated from taxa relative abundance in rRNA data over taxa relative abundance inrDNA data. Sample treatments are depicted as symbols: environmental control = open square, experimentalcontrol = solid square, 12CH4 = open circle, 12CH4 + NO2- = gray circle, and 13CH4 + NO2- = solid circle.956.3.4 Insight into microbial community response to substrate additionsTo initially identify potential changes in microbial community activity responding to CH4 and NO2- addi-tion, I analyzed SSU rRNA pyrotag sequences from the experimental control, 12CH4, 12CH4 + NO2- and13CH4 + NO2- treatments from 100 and 150 m. Overall, specific taxa dominated the SSU rRNA pyro-tag datasets (average relative abundance >1%) exhibiting up to a 30-fold increase in SSU rRNA: rDNAwhen CH4 and NO2- were added. For instance, Methylophilales (10%), Nitrospinaceae (7%), Methylo-coccales (17%), Planctomycetes (affiliated with OM190 (2%), Pirellulaceae (5%) and Planctomycetaceae(4%)), Marine Group A (4%), Verrucomicrobia (4%), and Thaumarchaeota (12%) OTUs showed a 10-30fold increase in 100m CH4 and CH4 + NO2- treatments when compared to the experimental control (Fig.6.5). In contrast, Chloroflexi affiliated with SAR202 (3%), OP9 (2%), TM6 (3%), WS3 (6%) and ZB3 (8%)Candidate divisions, SAR324 (3%), Methylococcales (6%), SUP05 (27%), Planctomycetes affiliated withPhycisphaerae (3%) and Pirellulaceae (10%), Marine Group A (3%) and Thaumarchaeota (6%) showed a10-20 fold increase in 150 m CH4 and CH4 + NO2- treatments when compared to control (Fig. 6.5).To further identify activity patterns among methanotroph OTUs in the SSU rRNA datasets, I comparedMethylococcales OTUs in SSU rRNA pyrotag datasets to full-length and pyrotag SSU rRNA gene referencesequences as previously described (Torres-Beltra´n et al., 2016). I identified 2 subgroups (≥ 99% iden-tity) affiliated with OPU3 (average relative abundance 13%) (Hayashi et al., 2007; Tavormina et al., 2010;Tavormina et al., 2013) and methanotrophic symbionts (average relative abundance 4%) (Streams et al.,1997; Dubilier et al., 2008; Petersen and Dubilier, 2009) (Fig. D.1). Based on SSU rRNA: rDNA, OPU3and methanotrophic symbiont OTUs showed a 10-20-fold increase in 100m CH4 and CH4 + NO2- treatmentswhen compared to the experimental control (Fig. D.1).To provide primary evidence for potential methanotroph-community interactions occurring in incuba-tion samples that could be representative of environmental microbial assemblies, I first BLAST comparedmicrobial OTUs observed in incubations SSU rRNA pyrotag datasets to pyrotag sequences recovered froma previous methanotroph diversity survey carried out in Saanich Inlet waters (Torres-Beltra´n et al., 2016).A total of 57 OTUs were found affiliated (≥ 99% identity) to OTUs present in the network correspondingto 25% of total OTUs in dysoxic-suboxic and anoxic modules (Fig. 6.6A; Table D.2). These 57 OTUswere enriched and active (based on abundance and SSU rRNA:rDNA ratio values) on incubation treatmentssupporting our experimental community assemblies with environmental co-occurrence network observa-tions (Fig. 6.6A; Table D.1). Sub-networks were constructed highlighting active OTUs showing significantpositive correlations (CV>0.6, p<0.001) with OPU3 and methanotrophic symbiont OTUs (Fig. 6.6B).For instance, OTUs correlating with OPU3 were affiliated with Chloroflexi, Alphaproteobacteria, Nitro-spina, Marine Group A, Planctomycetes, Verrucomicrobia, and Thaumarchaeota (Fig. 6.7B). In contrast,OTUs correlating with methanotrophic symbionts were affiliated with Bacteroidetes, Anaerolineae withinChloroflexi, WS3 Candidate division, Alphaproteobacteria, Methylophilales, Nitrospina, SAR324, SUP05,Marine Group A, Planctomycetes, Verrucomicrobia and Thaumarchaeota (Fig. 6.6B).960-0.671.77-4 0.67-1.771-1.043.15-8 1.04-3.15OPU3 SymbiontCH4NO2CH413CH4NO2Experimental control α-proteobacteriaδ-proteobacteriaγ-proteobacteriaMethylococcalesMethylophilalesNitrospinaceaeSAR324SUP05VerrucomicrobiaBacteroidetes PlanctomycetesMarine Group AWS3Thaumarchaeota10x20x1xrRNA : rDNA OPU3SymbiontABFigure 6.6: Co-occurrence patterns for methanotrophic OTUs from SSU pyrotag datasets. A) Time-series network based on significant (p<0.001) Bray-Curtis and Spearman correlation values (>0.6) amongOTUs from SSU rRNA gene pyrotag datasets (May 2008 - July 2010). Network encompassing the sub-oxic module where nodes are depicted according to taxonomy as indicated in the legend, and all interactions(edges) are shown as solid lines. Representative networks for each treatment are shown as indicated with thesymbol on top (experimental control = solid square, 12CH4 = open circle), 12CH4 + NO2- = gray circle and13CH4 + NO2- = solid circle). Node size represents OTUs rRNA: rDNA cumulative values for each treat-ment. B) Hive panels for OPU3 and symbiont OTUs showing resilient interactions throughout incubationtreatments. OTUs are distributed based on average relative abundance (log transformed) on the three axesfrom low (located closer to the center) to high (located towards the end of each line). Nodes are depictedaccording to taxonomy as indicated in the network. All interactions (edges) are shown as solid lines.976.3.5 PmoA and NirK diversityTo gain insight into the microbial community carrying out CH4 oxidation coupled with NO2- reduction,I looked into the taxonomic affiliation of pmoA and nirK genes. Overall, pmoA sequences represented0.002% (170 ORF sequences) of the total number of predicted ORFs in the metagenomic datasets. AllpmoA representative sequences (clustered at 97% similarity) from incubation samples were related to OPU3(Hayashi et al., 2007) (Fig. D.2). In contrast, nirK sequences represented 0.021% (1818 ORF sequences) ofthe total of predicted ORFs in the metagenomic datasets. Overall, nirK sequences were distributed throughdifferent taxonomic groups including Archaea (40%), Nitrospina (20%), Bacteroidetes (7%), Chlorobi (3%),and OPU3 (7%), and environmental sequences from Arabian Sea and Black Sea OMZs (3%) (Fig. D.3).Approximately 20% of nirK sequences fell within singleton clusters affiliated with Bacteria (clustered at70% similarity).6.4 Insight into microbial community response to substrates additionTo further determine the metabolically active community associated with CH4 oxidation, CH4-derived car-bon fixation and nitrogen metabolism, I looked into expression patterns across meta-transcriptomic andproteomic datasets. Overall, transcripts and proteins related to CH4 oxidation, CH4-derived carbon fixa-tion, and nitrogen metabolism found in CH4 + NO2- treatments only showed higher expression with respectcontrols (environmental and experimental) and CH4 addition treatments (Table 6.1).Second, I determined the community composition in meta-transcriptomic and proteomic data by sum-ming transcripts and proteins affiliated to a given microbial group. I observed that transcripts (2% oftotal transcripts) and proteins (20% total proteins) were mostly affiliated taxonomically with Flavobac-teriales within Bacteroidetes, Chlorobi, Chloroflexi, Nitrospina, Rhodobacterales, Bradyrhizobiaceae andSAR 11 within the Alphaproteobacteria, Methylophilales within the Betaproteobacteria, Desulfobacteralesand SAR324 within the Deltaproteobacteria, Epsilonproteobacteria, Alteromonadaceae, Methylococcaceae,SUP05, OMG and SAR86 within the Gammaproteobacteria, Planctomycetaceae within the Planctomycetes,NC10, Verrucomicrobia and Thaumarchaeota (Table 6.2), supporting our metabolic inferences based onSSU rRNA: rDNA observations.Next, I evaluated in greater detail the abundance and taxonomic affiliation of key proteins involvedin CH4 oxidation, CH4-derived carbon fixation and nitrogen metabolism across metatranscriptomic andmetaproteomic datasets. To begin with, I observed some similarities with regard to the functional andtaxonomic composition between CH4 and CH4 + NO2- treatments, such as PmoA from OPU3, formate de-hydrogenase (Fdh) from Bradyrhizobiaceae and Methylobacteriaceae within the Alphaproteobacteria, andDesulfobacterales within the Deltaproteobacteria, and the serine hydroxymethyl transferase (SHMT) fromOPU3, and Bacteroidales within the Bacteroidetes (Fig. 6.7). In addition, I observed transcripts for key pro-teins from the Calvin-Benson (CBB) cycle, such as the ribulose bisphosphate carboxylase (RuBISCO) fromGammaproteobacteria, and ribulose phosphate 3-epimerase (RPE) from Methylophilales, and Flavobacteri-ales within the Bacteroidetes (Fig. 6.8). Further, I observed proteins related to the TCA cycle including thephosphoenolpyruvate carboxylase (PEP) from Flavobacteriales within the Bacteroidetes, and 2-oxoglutarateferredoxin oxidoreductase (OGOR) from Epsilonproteobacteria (Fig. 6.7). Additionally, I observed the ex-98MethylococcalesMethylophilalesNitrospinaceaeBacteroidetesα-proteobacteriaδ-proteobacteriaγ-proteobacteriaε-proteobacteriaPlanctomycetesThaumarchaeotaEuryarchaeotaNC10VerrucomicrobiaPmoAMxaFRPESHMTRuBISCOAmoAHAOOGORPGDNRXHzoANosZNarANifU150100CH4CH4 + NO2-02505007501000 0 200400600G3PDGPIFBAPEPPmoAXoxFMxaFRPESHMTFDHRuBISCOHAOOGORFHMDHSDHCSDHASdhBFabZPGDNRXHzoANosZNarANifU0500100015002000 0 50010001500PEPRmpAFBAFDHCountsCH4 oxidationMethanol oxidationFormaldehyde assimialtionFormate oxidationGlycolysisCBB cycleTCA cycleFermentive metabolismAmmonium oxidationNO2- reductionAnnamoxFunctional markerCH4 oxidationMethanol oxidationFormaldehyde assimialtionFormate oxidationGlycolysisCBB cycleTCA cycleFermentive metabolismHydroxylamine oxidationNO2- reductionAnnamoxNirKFigure 6.7: Taxonomic and functional breakdown of transcripts related to key proteins for methaneand nitrogen cycling. The transcripts related to key proteins for methane and 610 nitrogen cycling areas follows: PmoA, particulate methane monooxygenase subunit β ; MxaF, Ca-dependent methanol dehy-drogenase; XoxF, Lanthanide-dependent methanol dehydrogenase; SHMT, serine hydroxymethyl trans-ferase; RmpA, 3-hexulose-6-phosphate isomerase; FDH, formate dehydrogenase; GPI, glucose 6 phosphateisomerase; FBA, fructose bisphosphate aldolase; G3PD, glyceraldehyde-3-phosphate dehydrogenase; Ru-BISCO, ribulose bisphosphate carboxylase; RPE, ribulose phosphate 3-epimerase; PEP, phosphoenol pyru-vate carboxylase; SdhB, succinate dehydrogenase iron-sulfur subunit; MDH, malate dehydrogenase; FH,fumarate hydratase; OGOR, 2-oxoglutarate ferredoxin oxidoreductase; SDHA, succinate dehydrogenaseflavoprotein subunit; SDHC, succinate dehydrogenase cytochrome b556 subunit; PGD, 6-phosphogluconatedehydrogenase; FabZ, 3-hydroxyacyl ACP dehydratase; AmoA, ammonia monooxygenase subunit α; HAO,hydroxylamine reductase; NarA, nitrate reductase subunit α; NRX, nitrite oxidoreductase; NirK, nitrite re-ductase; NosZ, nitrous oxide reductase; NifU, nitrogen fixation protein; Hzo, hydrazine oxidoreductase.Function and taxonomy assignments were determined by the sequence identity of transcripts and peptidesto metagenomic reads. The total number of reads for each protein are depicted as bars coloured accordingto taxonomy as indicated in the color key.99Table 6.1: Key protein ratios across meta-transcriptomic and proteomic datasets. Summed ratio valuesof key proteins involved in CH4 oxidation, CH4-derived carbon fixation, and nitrogen metabolism. Values≥ 1 mean expression of a given transcript and protein was higher in samples than controls (environmentaland experimental). Lack of value (indicated as ’-’) means no transcript or protein was observed in controls.Marker protein CH412CH4 + NO2-13CH4 + NO2-CH412CH4 + NO2-13CH4 + NO2-CH412CH4 + NO2-13CH4 + NO2-CH412CH4 + NO2-13CH4 + NO2-PmoA; particulate methane monooxygenase subunit 4.88 6.28 7.32 12.03 15.48 18.04 1.21 7.81 6.02 1.09 3.05 3.50 MxaF; Ca-dependent methanol dehydrogenase 2.20 1.38 2.20 2.29 3.03 4.83 - 2.64 1.07 - - - XoxF; Lanthanide-dependent methanol dehydrogenase 1.36 6.84 7.52 2.75 4.37 4.80 - 1.70 1.77 - - - SHMT; serine hydroxymethyl transferase 2.61 4.43 4.41 1.43 3.64 3.62 - - - - - - RmpA; 3-hexulose-6- phosphate isomerase 7.53 9.53 11.61 2.86 7.12 8.67 - 2.50 1.53 - - - FDH; formate dehydrogenase 1.39 2.95 3.43 2.58 3.26 3.50 1.75 9.70 6.55 1.18 3.36 3.00 GPI; glucose 6 phosphate isomerase 1.05 2.22 1.96 2.45 3.64 4.83 - - - - - - FBA; fructose bisphosphate aldolase 1.05 2.22 2.94 1.72 1.82 2.42 - - - - - - G3PD; glyceraldehyde 3 phosphate dehydrogenase 1.05 2.77 1.10 1.17 1.82 1.41 - - - - - - RuBISCO; ribulose bisphosphate carboxylase 2.09 1.48 1.47 1.00 2.73 2.42 - - - - - - RPE; ribulose phosphate 3-epimerase 1.46 1.01 1.47 1.00 2.73 2.42 - - - - - - PEP; phosphoenol pyruvate carboxylase 1.19 3.88 3.31 1.72 2.55 2.17 - - - - - - SdhB; succinate dehydrogenase iron sulfur subunit 1.57 3.69 3.92 1.58 3.64 3.62 - - - - - - MDH; malate dehydrogenase 2.39 3.48 3.78 1.37 2.00 2.17 - - - - - - FH; fumarate hydratase 2.09 4.43 5.88 1.47 2.08 1.55 - - - - - - OGOR; 2-oxoglutarate ferredoxin oxidoreductase 1.05 2.77 2.20 1.15 1.21 1.61 - - - - - - SDHA; succinate dehydrogenase flavoprotein subunit 2.56 3.69 3.92 2.06 3.64 3.62 - - - - - - SDHC; succinate dehydrogenase cytochrome b556 subunit 2.51 1.11 1.29 2.21 1.82 1.21 - - - - - - PGD; 6 phosphogluconate dehydrogenase 1.25 1.77 1.76 1.72 2.43 2.42 - - - - - - FabZ; 3 hydroxyacyl ACP dehydratase 2.09 11.08 14.84 2.41 4.55 1.81 - - - - - - AmoA; ammonia monooxygenase subunit 4.88 3.10 3.92 2.00 2.32 2.42 - - - - - - HAO; hydroxylamine reductase 1.05 1.86 3.70 2.06 1.82 1.45 1.40 6.90 5.56 1.52 8.05 7.00 NarA; nitrate reductase subunit 3.28 5.10 4.43 1.21 2.73 2.56 - - - - - - NRX; nitrite oxidoreductase 1.39 4.43 4.41 1.12 7.25 8.27 1.77 3.26 3.30 1.36 6.76 4.20 NirK; nitrite reductase 1.39 4.43 4.41 1.82 6.87 6.64 5.70 10.35 9.19 2.82 14.04 11.20 NosZ; nitrous oxide reductase 6.27 5.76 8.23 6.44 5.92 8.45 - - - - - - NifU; nitrogen fixation protein 10.46 15.51 17.64 1.72 1.59 1.81 - - - - - - Hzo; hydrazine oxidoreductase 1.25 1.48 1.31 1.47 2.42 2.42 1.12 5.67 4.91 1.85 5.25 5.65Sample : Environmental ControlSample: Experimental ControlRNA ProteinSample : Environmental ControlSample: Experimental Controlpression of hydroxylamine oxidoreductase (HAO) from Planctomycetaceae and Candidate NC10, of nitriteoxidoreductase (NXR) and nitrate reductase (NarA) from Nitrospinaceae, as well as of nitrous oxide reduc-tase (NosZ) from Chlorobi within the Bacteroidetes (only from 100 m samples) (Fig. 6.7). I also observedthe expression of the nitrogen fixation protein (NifU) from Epsilonproteobacteria (only from 150 m sam-ples), and hydrazine oxidoreductase (Hzo) from Planctomycetaceae (Fig. 6.7).Unique expression for 100 m CH4 samples included transcripts related to the Glycolysis pathway, suchas glucose-6-phosphate isomerase (GPI) from Gammaproteobacteria (8.08 RPKM), fructose bisphosphatealdolase (FBA) from SAR11 (64.68 RPKM) within Alphaproteobacteria, SAR324 (97 RPKM) within theDeltaproteobacteria, Chlorobi (145.5 RPKM) and Flavobacteriales (16.17 RPKM) within the Bacteroidetes,and glyceraldehyde-3-phosphate dehydrogenase (G3PD) from Alteromonadaceae (41.8 RPKM), OM60100Table 6.2: Total number of transcripts and proteins affiliated with metabolically active taxa. Summedcount values for all transcripts and proteins which taxonomic affiliation was related to a given microbialgroup across treatments. Lack of value (indicated as ’-’) means no protein was observed.Transcripts ProteinsBacteroidetes Flavobacteriales 312 193Chlorobi 120 6Chloroflexi 360 24Nitrospina 7128 2728Rhodobacterales 2316 1673Bradyrhizobiaceae 960 720SAR11 2004 294Methylophilales 540 132Nitrosomonadales 240 -Desulfobacterales 1920 370SAR324 288 120-proteobacteria Campylobacterales 1680 888Alteromonadaceae 228 180Methylococcaceae 5916 5904SUP05 2648 2258OMG group 216 -SAR86 2184 10Planctomycetes Planctomycetaceae 672 128Candidatus Methylomirabilis oxyfera NC10 600 240Verrucomicrobia 120 -Archaea Thaumarchaeota 2400 625-proteobacteriaTotal number ofTaxonomy-proteobacteria-proteobacteriaclade (111.61 RPKM), and SAR92 (48.51 RPKM) within Gammaproteobacteria (Fig. 6.7). Additionally,the expression of the ammonia monooxygenase subunit α (AmoA) from Thaumarchaeota (22.63 RPKM)was observed. With respect to the 150 m CH4 sample, G3PD from Epsilonproteobacteria (168.5 RPKM)was observed particularly, as well as 6-phosphogluconate dehydrogenase (PGD) from OPU3 (5.26 RPKM).In addition, NarA from Planctomycetaceae (17.55 RPKM) was also uniquely observed in 150 m CH4 sample(Fig. 6.7).Differential expression patterns emerged based on O2 concentration (related to 100 and 150 m depthintervals) for the CH4 + NO2- treatments, with little variation between 12CH4 and 13CH4 substrates (Sup-plementary Information Fig. 4). Of note, was the occurrence of Ca-dependent (MxaF) and lanthanide-dependent (XoxF) methanol dehydrogenases from Methylophilales (157.2 RPKM) and OPU3 (113 RPKM),in 100 and 150 m samples, respectively (Fig.7). Also of note was the expression of 3-hexulose 6-phosphate(RmpA) from OPU3 (107 RPKM) in 150 m samples. Additionally, Fdh from Methylophilales (48.51RPKM) and Planctomycetaceae (205.6 RPKM) were observed in 100 m samples, while Fdh from Planc-tomycetaceae (802 PRKM), Epsilonproteobacteria (2.27 RPKM) and Methanomicrobia (178.86 RPKM)101within the Euryarchaeota were observed in 150 m samples (Fig. 6.7). In addition, FBA from Verrucomicro-bia (11.36 RPKM) was uniquely observed at 100 m. Interestingly, transcripts for the TCA cycle includingsuccinate dehydrogenase iron-sulfur subunit (SdhB), malate dehydrogenase (MDH), and fumarate hydratase(FH) (SdhB = 8.08 RPKM, MDH = 5.68 RPKM, and FH = 5.68 RPKM), and other fermentative enzymessuch as succinate dehydrogenase flavoprotein subunit (SDHA). The succinate dehydrogenase cytochromeb556 subunit (SDHC), 6-phosphogluconate dehydrogenase (PGD) and 3-hydroxyacyl ACP dehydratase(FabZ) were only observed at 150 m and were affiliated with OPU3 (SDHA=16.17 RPKM, SDHC=51.4RPKM, PGD= 15.65 RPKM, FabZ= 8.56 RPKM) (Fig. 6.7). The expression of NirK from OPU3 wasobserved at 100 and 150 m (125.5 and 5.68 RPKM, respectively) (Fig. 6.7).Protein expression was constrained to a very specific set of proteins and taxa within the pathways ex-plored. For instance, PmoA affiliated with OPU3 was expressed across all samples, showing higher NSAFvalues in CH4 + NO2- samples (1.64 and 0.918 NSAF at 100 and 150 m, respectively) (Fig. 6.8). MxaFand XoxF methanol dehydrogenases from Methylophilales and OPU3 were also expressed in CH4 + NO2-samples (0.473 and 0.123 NSAF at 100 and 150 m, respectively) (Fig. 6.8). Formate dehydrogenase fromGammaproteobacteria was expressed across all samples, showing the highest NSAF value (0.768) in 150 mCH4 + NO2- samples. Similarly, RmpA from OPU3 (0.107 NSAF) was expressed in 150 m CH4 + NO2-samples (Fig. 6.8). In addition, HAO from NC10 (0.674 NSAF) was only expressed in 150m CH4 + NO2-samples. Nitrite reductase (NirK) from OPU3 was only expressed (2.479 NSAF) in 100 m CH4 + NO2-samples (Fig. 6.8), while NXR from Nitrospinaceae was expressed across all samples, showing the high-est value (7.64 NSAF) in the 100 m CH4 + NO2- samples. Hydrazine oxidoreductase, HZO, affiliated withPlanctomycetaceae was expressed in 150 m samples showing the highest expression in CH4 + NO2- samples(0.258 NSAF) (Fig. 6.8).6.4.1 Elucidating microbial metabolic networksTo ultimately elucidate potential metabolic interactions between taxa, I converged taxonomic co-occurrencepatterns with metabolic network information. First, a co-occurrence analysis based on Bray-Curtis andSpearman correlation values among predicted pathways at the suboxic-dysoxic water-column compartment(between 100 and 150 m) over a two-year period (June 2009 to August 2011) resulted in a microbialmetabolic network (Supplementary Information S1). The co-occurrence analysis resulted in significantpositive correlations (CV>0.6, p<0.001) among pathways, including CH4 oxidation, methanol oxidationto formaldehyde, formaldehyde oxidation via the H4MPT pathway, formate oxidation, CO2 fixation via theTCA cycle, and ammonium and NO2- oxidation (Fig. 6.9; Table D.2).To provide further evidence on CH4 oxidation as a communal function under suboxic-dysoxic water-column conditions, I BLAST-compared (>90% identity) incubation transcripts to conceptually translatedORFs of key proteins within correlating pathways. I observed most ORFs showed identical functional andtaxonomic annotation to CH4 + NO2- transcripts from 150 m, except for AmoA from Thaumarchaeota thatmatched transcripts from the 100 m CH4 sample. For instance, PmoA was affiliated with OPU3, MxaF wasfound to be affiliated with Methylophilales, NXR was affiliated with Nitrospinaceae, Fdh was affiliated withGammaproteobacteria and Planctomycetes, and ShdB, MDH and FH within the TCA cycle were affiliated102MethylococcalesMethylophilalesNitrospinaceaeγ-proteobacteriaPlanctomycetesNC10PmoAMxaFNRXHzoA150100PmoAXoxFMxaFFDHHAONRXHzoARmpAFDHCountsCH4 oxidationMethanol oxidationFormate oxidationNO2- reductionAnnamoxFunctional markerCH4 oxidationMethanol oxidationFormaldehyde assimialtionFormate oxidationHydroxylamine oxidationNO2- reductionAnnamoxNirK010002000300040005000 0 25050075010000200040006000 0 50010001500CH4 CH4 + NO2- Figure 6.8: Taxonomic and functional breakdown of key proteins related to methane and nitrogen cy-cling. The key proteins related to methane and nitrogen cycling are as follows: PmoA, particulate methanemonooxygenase subunit β ; MxaF, Ca-dependent methanol dehydrogenase; XoxF, Lanthanide-dependentmethanol dehydrogenase; RmpA, 3-hexulose-6- phosphate isomerase; FDH, formate dehydrogenase; HAO,hydroxylamine reductase; NRX, nitrite oxidoreductase; NirK, nitrite reductase; NosZ, nitrous oxide re-ductase; HzoA, hydrazine oxidoreductase. Function and taxonomy assignments were determined by thesequence identity of transcripts and peptides to metagenomic reads. The total number of reads for eachprotein are depicted as bars coloured according to taxonomy as indicated in the color key.with OPU3 (Fig. 6.9).Based on ORFs and transcript pairs I elucidated a potential mechanism from which CH4 oxidationserves as carbon source for the microbial community under dysoxic-suboxic water column conditions. Isuggest methanotrophic OPU3 goes under a fermentative state due to O2 limitation and provides methanol tomethylotrophic Methylophilales. Formate may be released by OPU3 or Methylophilales, as a byproduct ofincomplete carbon assimilation, and could potentially feed the Gammaproteobacteria and Planctomycetes.Therefore, OPU3 could potentially uptake carbon via the TCA cycle. In addition, under O2 limiting condi-tions, it is likely Thaumarchaeota carries out ammonium oxidation, while Nitrospinaceae carries out NO2-oxidation (Fig. 6.9).6.5 DiscussionThis chapter charts methanotroph community composition and metabolic responses to CH4 and NO2- addi-tion under low-O2 conditions. The observation presented here encompass isotopic signatures demonstratingCH4-derived carbon incorporation into microbial biomass and multi-omic information describing commu-nity shifts in abundance and metabolism related to CH4 oxidation coupled with NO2- reduction. The obser-vations indicate that OPU3 metabolic activity supplies the microbial community with organic compoundsthat are carbon sources. In all, the results provide evidence supporting CH4 oxidation as community-level103-+RPKM (log)CH4CH3OHCH2OCHOOCO2TCA(1)(2)NO3NO2NONH4N2MethylococcaceaeMethylophilalesγ-proteobacteriaEuryarchaeotaNitrospinaceaeThaumarchaeotaPlanctomycetes0-7.957.95-79.0279.02-71512A BFigure 6.9: Co-occurrence patterns for dysoxic-suboxic metabolic pathways in the Saanich Inlet wa-ter column. A) Methane oxidation and carbon fixation via the TCA cycle, and Nitrogen cycling pathwaysreactions. Colored dots depict taxonomic affiliations for key proteins in a given reaction. B) Time-series net-work based on significant (p<0.001) Bray-Curtis and Spearman correlation values (>0.6) among predictedpathways from metagenomic datasets (June 2009-vAugust 2011). Hiveplot encompasses CH4 oxidationcorrelating pathways where nodes are depicted according to taxonomy, as indicated in the legend, and CH4oxidation interactions (edges) are shown as solid red lines. Nodes are distributed along the axis accordingto RPKM abundance.function knitting a functional metabolic community network around CH4 oxidation.Bulk 13C measurements suggest microbial communities from distinct depths incorporated carbon intoprotein differently, underlining potential differences in their metabolism due to O2 concentration. For in-stance, the community at 100 m showed higher carbon incorporation (156.1 %o increase from environmentalcontrol) than the community from 150 m (19.5 %o increase from the environmental control). In addition,I expected carbon incorporation to be derived from CH4 oxidation primarily by methanotrophs. However,methanotroph abundance after 72 h incubation period (2% average relative abundance from the total numberof prokaryotic reads) suggests only a small fraction of the oxidized CH4 was converted to biomass. Recentexperiments carried out in bioreactors showed that at low dissolved O2 concentrations, 40-50% CH4-derivedcarbon was extracellular acetate and formate, supporting the hypothesis that CH4 fermentation leads to littlebiomass synthesis (Kalyuzhnaya et al., 2013). In addition, an extended anaerobic starvation experiment onthe methanotrophic isolate strain WP 12 showed that cell biomass decreased during the cultivation periodand that metabolized carbon was recovered mainly as organic solutes in the starvation medium (Roslev andKing, 1995).Multi-omic information points to OPU3 as the main methanotrophic agent carrying out CH4 oxidationcoupled with NO2- reduction (PmoA transcripts and proteins affiliated with OPU3 showed 6-fold and 2-foldincreases in CH4 + NO2- treatments) under suboxic-dysoxic water column conditions in the Saanich Inlet.Recent environmental evidence based on time-series regression analyses (Torres-Beltra´n et al., 2016) andtranscriptional data (Padilla et al., 2017) from O2-deficient marine systems support OPU3 coupling CH4 ox-idation and NO2- reduction as a mechanism that thrives in the O2-deficient water columns. Furthermore, the104functional expression patterns observed for OPU3 in relation to O2 concentrations suggest different modesof carbon fixation by the methanotroph, derived from CH4 oxidation coupled with NO2- reduction. Forinstance, under suboxic water-column conditions (∼30 µM O2) OPU3 will likely incorporate carbon via theRuMP pathway. In contrast, under dysoxic water-column conditions (∼3 µM O2), OPU3 will likely incor-porate carbon via fermentative pathways. The presence of putative fermentation genes in methanotrophs islikely widespread (Kalyuzhnaya et al., 2013). In fact, the expression of fermentative genes, i.e. MDH andSDH by Methylomicrobium alcaliphilum 20Z, has been recently demonstrated under low-O2 growth condi-tions in a bioreactor experiment (Kalyuzhnaya et al., 2013), supporting the hypothesis that methanotrophsare capable of fermentation from CH4-derived formaldehyde, leading to the formation of end products.Combined, these metabolic features and plasticity might allow OPU3 to thrive under low-O2 water-columnconditions, despite the potential bioenergetic cost related to biomass synthesis (all cellular carbon is assimi-lated at the oxidation level of formaldehyde via RuMP cycle, while only∼ 15% via CO2 oxidation; (Hansonand Hanson, 1996).Co-occurrence patterns among metabolically active OTUs, identified based on SSU rRNA:rDNA ra-tios coupled with network analyses, suggest methanotroph partnerships with correlating community mem-bers might not be random (Sauter et al., 2012; Beck et al., 2013; Hernandez et al., 2015; Oshkin et al.,2015; Karwautz et al., 2018). I observed that the metabolically active community was primarily comprisedof OTUs affiliated with taxa, i.e. Methylococcaceae, Methylophilales, Bacteroidetes and Planctomycetesshowing a concomitant increase in transcript and protein abundance. These taxa have been described asco-occurring with methanotrophs as their abundance, based on SSU rRNA gene observations, commonlyincreases with CH4 addition under laboratory conditions (Sauter et al., 2012; Beck et al., 2013; Hernandez etal., 2015; Oshkin et al., 2015; Karwautz et al., 2018). Observed functional patterns supported co-occurrenceamong metabolically active OTUs hinting to communal CH4 metabolism based on methanotrophs feedingcommunity members on released compounds, i.e. methanol and formate due to fermentative metabolism(Kalyuzhnaya et al., 2013; Yu and Chistoserdova, 2017). For instance, differential expression of OPU3XoxF and Methylophilales MxaF indicates Methylophilales are possibly feeding on CH4-derived methanolfrom OPU3 when XoxF is not transcribed/translated into protein. In fact, the methanol cross-feeding mech-anism between methanotrophs and methylotrophs has been previously determined in coculture experimentsusing Lake Washington strains where methanol released by methanotrophic Methylobacter sp. supportedmethylotroph Methylonera sp. growth (Krause et al., 2017). In addition, concomitant with OPU3 ex-pression patterns, the expression of FDH from Methylophilales, Planctomycetes and Methanomicrobia wasobserved, suggesting these taxa may feed on formate released by methanotrophs. FDH expression by Methy-lophilales further supports the metabolic interdependence between methanotrophs and methylotrophs whenco-existing in culture conditions (Hernandez et al., 2015; Oshkin et al., 2015; Yu and Chistoserdova, 2017).Interestingly, the metabolic relationship between methanotrophs, methanogens and Planctomycetes may bebased on the evolution of genes for C1 transfer reactions between the oxidation levels of formaldehyde andformate (Chistoserdova et al., 2004). In particular, genomic information points to the use of formate asan electron donor by anammox Planctomycetes to aid in the Wood-Ljungdahl pathway for the reductionof carbon dioxide (Strous et al., 2006). Another taxa commonly co-occurring with methanotrophs in mi-105crocosm experiments is Bacteroidetes; particularly, as I observed here, Flavobacteriales (Hernandez et al.,2015; Oshkin et al., 2015; Karwautz et al., 2018). Genes expressed by Flavobacteriales affiliated with FBA,REP and PEP, suggest that these may be feeding on polymeric compounds derived from methanotrophmetabolism. Experimental results for CH4-fed microcosm experiments suggest the specific mechanismbehind the metabolic relationship between Bacteroidetes and methanotrophs relies on the former feed poly-meric substances produced and released by methanotrophs (Kalyuzhnaya et al., 2013; Yu and Chistoserdova,2017), i.e. biopolymers that serve as a source of carbon, energy, or reducing-power for methanotrophs inexceptional circumstances, such as under conditions of nutrient limitation (Strong et al., 2015). Combinedfunctional information and environmental pathway network analyses allowed me to connect the metabolicinteractions from specific taxa to the potential communal functions of CH4 oxidation, and to ultimatelypinpoint the potential metabolic mechanism by which CH4-derived carbon could be transferred to the com-munity in O2-deficient water columns (Fig. 6.9).According to the metabolic network I elucidated, ∼15% of the carbon incorporated by OPU3 could besynthesized into biomass (via TCA pathway under low-O2 conditions), explaining its low relative abun-dance in the Saanich Inlet waters (Torres-Beltra´n et al., 2016). The remaining ∼32kg C from CH4 couldpotentially be distributed to co-occurring partners as polymeric compounds released by the methanotroph.This rough estimation highlights the role of methanotrophs in O2-deficient systems beyond the CH4 cycleand indicates that future efforts should focus on assessing the functional capacity of methanotrophs through-out suboxic-dysoxic waters within coastal and open ocean regions worldwide to identify the environmentalfactors regulating their metabolism and community interactions.6.6 ConclusionIn the present chapter, I used experimental and environmental time-series multi-omic observations to de-termine methanotrophic community composition and activity in response to CH4 and NO2- additions underlow-O2 growth conditions, as well as to understand the nature of community-level interactions related toCH4 metabolism. Functional information revealed a communal response to CH4 oxidation ignited by OPU3coupling CH4 oxidation with NO2- reduction. OPU3 fermentative metabolism under low-O2 water-columnconditions likely shifts the community pathways of carbon incorporation by releasing organic compoundsas carbon sources for co-occurring taxa. In combination, these observations expand our understanding ofthe communal nature of CH4 metabolism (reviewed in Chistoserdova and Kalyuzhnaya, 2018) and provide anew outlook on the environmental role of the methanotrophs as essential components of microbial metabolicnetworks beyond the CH4 cycle. Although co-occurrence results suggested SAR324, MGA, SUP05, candi-date WS3 and Verrucomicrobia may play a part in communal CH4 oxidation, no functional information oncarbon or nitrogen metabolism from these taxa was observed concomitant to methanotroph response, exceptfor FBA from SAR324 and Verrucomicrobia. I encourage the use of tracing techniques, such as stable iso-tope probing coupled with RNA and protein analyses, to unveil the biochemical processes and mechanismsthat could explain the co-occurring patterns of these taxa with methanotrophs. In future research, I suggeststarting by addressing the biotic and abiotic factors involved in methanotrophic community co-occurrencepatterns as well as the mechanisms for metabolite exchange within methanotrophic communities across sys-106tems worldwide. I recommend future efforts focus on understanding how changing levels of O2 impact thedynamics among partnering groups and to assess the specific roles of microbes preventing CH4 emission.This information is necessary to better inform models that evaluate and predict system responses to oceandeoxygenation. Furthermore, knowledge gained from methanotrophic community partnerships will greatlyaid in overcoming synthetic community manipulation by providing the knowledge necessary to select a setof strains with characterized genomes and physiologies that can be used in ongoing biotechnological andengineering approaches aimed at methanotroph industrial applications and commercialization.107Chapter 7ConclusionsOxygen (O2) is fundamental to biological and biogeochemical processes in the ocean (Breitburg et al.,2018). The effects of O2-dependent nutrient-cycling processes are communicated by oceanic water masscirculation; as such, changes within oxygen minimum zones (OMZs) can influence microbial-driven bio-geochemical processes on regional and global scales. As OMZs expand and their upper boundaries advancetoward surface waters, OMZ microbial communities will shoal towards the upper water column. Many ofthe groups that play important roles in greenhouse gas cycling within OMZs may respond to habitat shifts bychanging their abundance, diversity or composition with potential downstream ecosystem function implica-tions. Microbial communities play significant roles driving ocean biogeochemistry (Falkowski et al., 2008;Beman and Carolan, 2013). Microbial community structure, in turn, is strongly influenced by the physicaland chemical environment (Margalef, 1968; Tozzi et al., 2004). As the ecology and biogeochemistry of theoceans are tightly interconnected, the response and resiliency of microbial communities to O2 loss may havefar-reaching effects on ocean biogeochemistry that could results in feedbacks to global warming driven bygreenhouse gases. Thus, in conjunction with rising ocean deoxygenation, there is the growing need to gen-erate more information to understand how O2 loss is altering microbial pathways and the rates of processesrelated to greenhouse gas cycling within the ocean water column.7.1 Standards of practice for sample collection and data analysisThe Saanich Inlet time-series data compendiums are community-driven research frameworks for observingand predicting microbial community responses to ocean deoxygenation across multiple scales of biologicalorganization. By generating more compatible geochemical and multi-omic sequence data, robust informa-tion may be generated across ecosystems that allows for more accurate global predictions. In combination,the geochemical and multi-omic sequence data are powerful tools with the potential to uncover tight rela-tionships among key microbial players related to the biogeochemical cycling of carbon, nitrogen and sulfur.Elucidating the metabolic networks entwined in these relationships is crucial for the comprehensive under-standing of community composition and function shifts due to O2 loss, and thus, the generation of moreaccurate predictive tools.Advances in multi-omic sequencing technology are enabling the study of microbial communities at108unprecedented scales (Hahn et al., 2016). Research community efforts on developing multi-omic time-seriesdata have greatly enriched our understanding of the metabolism of key taxa that directly impact nutrientcycling in OMZs. However, time-series data analyses encounter a major flaw; namely, the incompatiblecollection and filtering strategies between sampling efforts make it challenging to compare observationsamong systems and restrain the extensiveness of observations on a global scale. I consider developingtime-series datasets as reference databases and coupling in situ with on-ship sampling methods as a crucialstrategy to improve our understanding of microbial community metabolism across marine systems.In Chapter 3, I presented observations supporting the 2014 SCOR workshop effort to understand the ef-fect of collection methods on microbial ecology research and the interpretation of the microbial communitiesin O2-deficient water columns. Results using three-domain SSU rDNA and rRNA 454 tag sequencing datademonstrated shifts in microbial community composition, structure and function associated with collectionand filtration methods. Microbial community composition and structure showed considerable abundanceshifts for specific taxa that proved to be sensitive to collection time (in situ vs. on ship).For instance, Planc-tomycetes, Chloroflexi and Candidate divisions were only observed in in situ samples. Additionally, I pro-vided evidence that size fractionation and particle fragmentation due to filtration had an impact on specifictaxa abundance, i.e Bacteroidetes, Alphaproteobacteria and Opisthokonta increased 5-fold in abundance onlarge size filters (0.4 µm) compared with small size filters (0.22 µm). Coupling SSU rRNA and rDNA tagsequencing data, I demonstrated the impact of filtration methods on indicator OTUs providing insight intocommunity function differences between in situ vs. on ship samples. Of note, the identification of microbialtaxa within the uncultivable majority such as Candidate divisions (BCR1 and WS3), Desulfarculales, Desul-furomandales, and Phycisphaerae as these taxa are disregarded from the microbial community based onSSU rDNA studies due to their low abundance. However, these taxa may play important roles in sulfur cy-cling, carbon fixation and fermentation under O2-deficient water-column conditions. Bias against these raretaxa may lead to neglecting important localized processes and may potentially compromise biogeochemicalinterpretations (Suter et al., 2016) based on on-ship observations alone.Observations made in Chapter 3 elaborate on previous contributions provided by Ganesh et al., Padillaet al. and Suter et al., indicating the potential effects of sample collection methods in our understanding ofmicrobial community structure and function in OMZs. Combined, these observations highlight the impor-tance of considering the effects of biomass size-fractionation, filtered water volume, and collection timingin experimental design. Additionally, these findings indicate the need for establishing compatible moleculardata generation techniques that facilitate cross-scale comparisons and that more accurately assess in situ mi-crobial community composition and function. Furthermore, microbial community composition bias due tocollection methods may impact our understanding of microbial community function and consequently affectcurrent models estimations of biogeochemical cycling processes in OMZs worldwide that are undergoingocean deoxygenation.7.2 Using co-occurrence network analysis to chart ecological interactionsCorrelation analyses are powerful tools that can unveil environmental co-occurrence trends that predict avariety of microbial interactions. Hypotheses generated from correlation analyses can be tested by coupling109microscopy methods and diverse emerging technologies i.e cell sorting and high-throughput co-culturingcoupled with genomics to provide benchmark data for the evaluation and improvement of network inferenceapproaches. For instance, network inference combined with co-culture experiments can deliver the growthrates and interaction strengths required for mathematical simulations of the microbial community. Faust etal. thoroughly reviewed the technological advances and methodological innovations recently used in thefield of microbial modeling for the discovery of cooperative and competitive relationships between speciesand described how these techniques are opening the way towards global ecosystem network prediction andthe development of ecosystem-wide dynamic models. Such technological and methodological efforts arekey to developing comprehensive and realistic mathematical models required to better predict the effects ofmicrobial communities on nutrients cycling in expanding OMZs. Nowadays, microbial ecological modelingis essential for translating the highly complex, nonlinear and evolving systems of microbial communitiesto fundamental knowledge and a better understanding of our changing oceans. In addition to the analy-sis of time-series microbial community observations, the combination of community analyses in naturalenvironments under controlled conditions, i.e. micro-mesocosm and enrichment cultures, will allow us tounderstand the physiological and regulatory mechanisms at cellular level that ultimately control activity andaffect the dispersal of taxa driving the important cycling of greenhouse gases such as CH4.In this thesis, I used correlation methods to unravel O2-driven community-level interactions that affectnutrient and gas cycling in OMZs. First, in Chapter 4, I used correlation analysis to explore protistanparasitic interactions throughout defined O2 gradients in the Saanich Inlet water column. Bray-Curtis andSpearman’s rank correlations and indicator species analyses on SSU rDNA 454 tag sequencing data revealedpotential significant interactions occurring between four known Syndiniales groups and different protistanand metazoan taxa including Phaeocystis antarctica (ciliates affiliated with Choreotrichia) and copepodsduring periods of water-column stratification. In surveying host-parasite potential interactions, I providea baseline understanding of the potential host range of the major parasitic Syndiniales groups that couldinfect key primary producers and heterotrophic populations in stratified water columns. Furthermore, Igive insight into the possible impacts that these infections may have on population dynamics, extensible tonutrient cycling processes, during seasonal water-column stratification in the Saanich Inlet.In Chapter 5, I used different correlation methods to chart methanotroph dynamics over an atypical ex-tended water-column stratification period in 2010 related to a relatively strong El Nin˜o event. Correlationanalyses on SSU rDNA 454 tag sequencing data in combination with geochemical information allowed meto resolve potential novel metabolic strategies, including the use of alternative terminal electron acceptors,i.e. NO2- and metabolic interactions among C1-utilizing microorganisms supporting CH4 oxidation. Resultsderived from multivariate analysis allowed me to identify significant correlations among methanotrophicOTUs revealing redox-driven niche partitioning along changing water-column redox gradients among taxa,supporting the role of O2 in shaping microbial community structure and function. Furthermore, negativebinomial regressions enabled me to identify potential novel metabolic strategies, including the use of alter-native terminal electron acceptors such as NO2- by OPU3 (p < 0.05). In using a co-occurrence analysis,based on Bray-Curtis and Spearman correlation values among OTUs, I further resolved potential metabolicinteractions between OPU1, OPU3 and symbiont groups with Methylophaga, Methylophilales, SAR324,110Verrucomicrobia and Planctomycetes. In all, I provide important baseline information on microbial agentsthat reduce the flux of climate-active trace gases from ocean to atmosphere and support the potential roleof OPU3 as substantial pelagic sink for CH4 (18.2 W m-2, equivalent to 450 years of radiative forcing thatcould be released to the atmosphere each year) along continental margins.To date, the relevant abiotic and biotic environmental conditions shaping methanotrophic communi-ties and influencing their activity have been studied in detail (reviewed in Chistoserdova and Kalyuzhnaya,2018). However, the question regarding how the different factors act alone and in combination on the mem-bers of methanotrophic communities in different ecosystems remains unanswered. The correlation results Ipresented in Chapter 5 allowed me to provide evidence linking ecosystem function with community com-position and on the dependence of environmental parameters. I also contributed novel knowledge on theotherwise unknown niche differentiation and habitat preferences among methanotrophic OTUs relevant towidely-distributed marine clusters. Finally, results in Chapter 5 represent pioneering knowledge that pro-vides valuable insight into understanding the different responses of methanotrophs to low-O2 and nitrogenconcentrations relevant to CH4 cycling models in O2-deficient marine waters.Looking forward, I recommend the use of multi-omic sequencing information coupled with process ratemeasurements in order to determine the coverage and efficiency of CH4 oxidation and derived carbon fix-ation pathways. In addition, detailed incubation experiments using labeled substrates coupled with geneexpression studies should be conducted to link CH4 oxidation pathways and process rates to specific mi-crobial agents. Process rates measurements along defined water column redox gradients are particularlyimportant on regional and global scales to better constrain the CH4 filtering capacity of coastal and openocean OMZs.7.3 Integrative analysis of coupled biogeochemical processesCommunity-level CH4 metabolism has been previously suggested based on environmental and laboratorymicrobial community surveys (reviewed in Chistoserdova and Kalyuzhnaya, 2018). The observations Ipresented in Chapter 6 shed light on communal CH4 metabolism in nature and support a new outlook onthe environmental role of methanotrophs as essential components of food webs driven by CH4-derivedcarbon fixation. Combined, these observations support the premise that methanotrophs in nature may playimportant roles in global biogeochemical processes beyond the CH4 cycle. In Chapter 6, using multi-omic(DNA, RNA and protein) data generated from an incubation experiment on suboxic-dysoxic Saanich Inletwaters. I 1) pinpointed key taxa involved in CH4 oxidation coupled with NO2- reduction, 2) identifiedthe metabolic pathways driving carbon assimilation from CH4 under low O2 conditions, and 3) determinedmicrobial metabolic interactions among C1-utilizing microorganisms fueled by CH4 oxidation.Observations throughout multi-omic datasets supported and highlighted the potential role of OPU3 asmajor microbial agent carrying out CH4 oxidation coupled with NO2- reduction in the Saanich Inlet suboxic-dysoxic water-column compartments. Based on the observed metabolic potential of OPU3 for thriving underO2 deficiency, I assessed the global distribution and occurrence of pmoA and nirK genes affiliated withOPU3. I found these genes co-occur in dysoxic-suboxic O2 conditions throughout open ocean and coastalOMZs (Fig. E.1). The global distribution of these functional marker genes throughout dysoxic-suboxic111water OMZs indicates that OPU3 may represent a significant biological sink for CH4. However, futureefforts focusing on assessing the global functional expression of these genes are required in order to betterdetermine the environmental conditions regulating CH4 oxidation coupled with NO2- reduction in order tobetter estimate the metabolic efficiency of OPU3 and prevent CH4 loss to the atmosphere from currentlyexpanding OMZs. Padilla et al. (2017) provided first evidence on OPU3 pmoA and nirK expression underwater column oxygen deficiency in the Costa Rica OMZ. As a first approach summing to this effort, Ioverviewed the distribution, abundance, and functional expression of pmoA and nirK genes affiliated toOPU3 in the Saanich Inlet water column. I used metagenomic and metatranscriptomic time-series data fromJune 2009 to August 2011, and observed genes expression overlapped at water column stratification periodspeaking at suboxic water column conditions (between the 100-150 m) (Fig. E.2). I encourage future studiesto focus on determining OPU3’s NO2--dependent CH4 oxidation rates that could be used in mathematicalmodeling approaches.Open questions remain to be answered regarding the specific nature and rate of metabolites providedby methanotrophs to the community, as well as the potential benefits that methanotrophs may gain in re-turn. Potential metabolic interactions presented in Chapter 6 provide baseline knowledge that can be usedto further investigate the mechanisms of syntrophic interaction in aerobic CH4 oxidation under low-O2 con-ditions. The synthetic community manipulation approach combined with the already existing multi-omicdatabases is a promising tool to efficiently achieve the experimental validation of these metabolic inter-actions (reviewed in Chistoserdova and Kalyuzhnaya, 2018). Obtaining new insights into the communalfunction of CH4 oxidation would not only provide the necessary knowledge for predicting the activities ofmethylotrophs in environmental settings but would also enable an effective application of these organismsand their metabolism in industrial processes, i.e. metabolite synthesis or toxin bioremediation. The futureof synthetic methylotrophy has been recently discussed and reviewed by Chistoserdova and Kalyuzhnaya(2018) and highlights the promising and emerging potential methanotrophs have as platforms for biotech-nological applications.7.3.1 Incubation experiment limitationsI initially developed the mesocosm experiment based on the theoretical concept of stable isotope probing(SIP). I tried using RNA and protein SIP techniques as these have proven to be high-resolution approachesfor identifying active anaerobic and aerobic CH4-oxidizing bacteria with considerably low levels of substrateincorporation (2% of 13C incorporation) (Manefield et al. 2002; McDonald et al. 2005; reviewed in Seifertet al. 2012). Based on the available literature for SIP experiments and the expected relative small substrateincorporation required, I decided to limit the incubation period to 72 h in order to accurately link substrateincorporation to specific microbial agents. However, the low natural abundance and slow metabolic rates ofmethanotrophic bacteria in the Saanich Inlet water column did not allow for enough substrate incorporationwithin the incubation period, making RNA and protein SIP methodological unfeasible. This methodolog-ical flaw hindered the quantification of substrate incorporation rates. Based on these limitations, the datain Chapter 6 is restricted to an overview of related assimilation processes. I strongly suggest that futureefforts develop time-series incubation experiments with a longer terminal incubation time and the use of112replicates for the different growth conditions. Furthermore, I recommend the use of the multi-omics SIPapproach coupled with microscopy techniques, i.e. DNA/RNA-SIP with phylogeny-specific probes (fluo-resce in situ hybridization; FISH) and secondary ion mass spectrometry (SIMS). The proper developmentof specific probes for methanotrophic bacteria (OPU3 is currently possible considering the genomic infor-mation available) will increase the resolution and accuracy of linking phylogeny and function to specificmicroorganisms and will provide 1) process rates linked to specific taxa, and 2) further evidence on the na-ture and composition of methanotrophic consortia. Finally, I strongly suggest future research efforts shouldfocus on determining 1) the OPU3 NO2--dependent CH4 oxidation mechanism in nature and 2) the specificnature and rate of metabolites provided by methanotrophs, i.e. OPU3, to the community. These researchefforts will contribute to the better understanding of how microbial communities will likely thrive underocean deoxygenation conditions and prevent climate-active gases loss to the atmosphere.7.4 The significance of this researchThis thesis encompasses multi-omic time-series observations resolving complex microbial interactions andhighlights the use of correlation analysis and co-occurrence network approaches as novel tools to generatebaseline knowledge with the aim of developing hypotheses to address environmental questions related toactive-trace gas loss to the atmosphere in a time of climate change and ocean deoxygenation. Moreover,this thesis answers the need of the scientific community to develop standard workflows to survey the com-position, structure and function of microbial communities that can be scaled to global ocean genomic datasurveys. This scientific community effort will allow for better informed global models predicting microbialmetabolism responses associated with global ocean deoxygenation.Finally, this thesis constitutes an integral and comprehensive survey on methanotrophic communitystrategies that thrive under low-O2 water-column conditions in a model coastal OMZ. The results presentedhere on methanotroph interactions, both taxonomic and metabolic, represent fundamental environmentalknowledge that can be used in biotechnological and engineering approaches.113References• Acinas, S.G., Antn, J., and Rodrguez-Valera, F. (1999). Diversity of Free-Living and Attached Bacte-ria in Offshore Western Mediterranean Waters as Depicted by Analysis of Genes Encoding 16S rRNA.Applied and Environmental Microbiology 65, 514-522.• Akaike, H. (1998). ”A New Look at the Statistical Model Identification,” in Selected Papers of Hiro-tugu Akaike, eds. E. Parzen, K. Tanabe G. Kitagawa. (New York, NY: Springer New York), 215-222.• Alldredge, A.L., and Cohen, Y. (1987). Can Microscale Chemical Patches Persist in the Sea: Micro-electrode Study of Marine Snow, Fecal Pellets. Science 235, 689-691.• Allen, L.Z., Allen, E.E., Badger, J.H., Mccrow, J.P., Paulsen, I.T., Elbourne, L.D.H., Thiagarajan,M., Rusch, D.B., Nealson, K.H., Williamson, S.J., Venter, J.C., and Allen, A.E. (2012). Influence ofnutrients and currents on the genomic composition of microbes across an upwelling mosaic. ISME J6, 1403-1414.• Allers, E., Wright, J.J., Konwar, K.M., Howes, C.G., Beneze, E., Hallam, S.J., and Sullivan, M.B.(2013). Diversity and population structure of Marine Group A bacteria in the Northeast subarcticPacific Ocean. ISME J 7, 256-268.• Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990). Basic local alignmentsearch tool. Journal of Molecular Biology 215, 403-410.• Anders, S., and Huber, W. (2010). Differential expression analysis for sequence count data. GenomeBiology 11, R106.• Anders, S., Pyl, P.T., and Huber, W. (2015). HTSeq a Python framework to work with high-throughputsequencing data. Bioinformatics 31, 166-169.• Anderson, J.J., and Devol, A.H. (1973). Deep water renewal in Saanich Inlet, an intermittently anoxicbasin. Estuarine and Coastal Marine Science 1, 1-10.• Arstegui, J., Gasol, J.M., Duarte, C.M., and Herndld, G.J. (2009). Microbial oceanography of thedark ocean’s pelagic realm. Limnology and Oceanography 54, 1501-1529.114• Armstrong, F.a.J., Stearns, C.R., and Strickland, J.D.H. (1967). The measurement of upwelling andsubsequent biological process by means of the Technicon Autoanalyzer and associated equipment.Deep Sea Research and Oceanographic Abstracts 14, 381-389.• Arrigo, K.R. (2005). Marine microorganisms and global nutrient cycles. Nature 437, 349-355.• Azam, F., Fenchel, T., Field, J.G., Gray, J.S., Meyer-Reil, L.A., and Thingstad, F. (1983). The eco-logical role of water-column microbes in the sea. Marine ecology progress series. Oldendorf 10,257-263.• Azam, F., and Malfatti, F. (2007). Microbial structuring of marine ecosystems. Nat Rev Micro 5,966-966.• Bader, G.D., and Hogue, C.W.V. (2003). An automated method for finding molecular complexes inlarge protein interaction networks. BMC Bioinformatics 4, 2-2.• Bai, X., Adolf, J.E., Bachvaroff, T., Place, A.R., and Coats, D.W. (2007). The interplay between hosttoxins and parasitism by Amoebophrya. Harmful Algae 6, 670-678.• Barbern, A., Bates, S.T., Casamayor, E.O., and Fierer, N. (2012). Using network analysis to exploreco-occurrence patterns in soil microbial communities. The ISME Journal 6, 343-351.• Barnard, R.L., Osborne, C.A., and Firestone, M.K. (2013). Responses of soil bacterial and fungalcommunities to extreme desiccation and rewetting. The ISME Journal 7, 2229-2241.• Beamish, R.J., Neville, C.M., and Cass, A.J. (1997). Production of Fraser River sockeye salmon(Oncorhynchus nerka) in relation to decadal-scale changes in the climate and the ocean. CanadianJournal of Fisheries and Aquatic Sciences 54, 543-554.• Beck, D.a.C., Kalyuzhnaya, M.G., Malfatti, S., Tringe, Susannah g., Glavina Del Rio, T., Ivanova, N.,Lidstrom, Mary e., and Chistoserdova, L. (2013). A metagenomic insight into freshwater methane-utilizing communities and evidence for cooperation between the Methylococcaceae and the Methy-lophilaceae. PeerJ 1, e23.• Beman, J.M., and Carolan, M.T. (2013). Deoxygenation alters bacterial diversity and communitycomposition in the ocean’s largest oxygen minimum zone. Nature Communications 4, 2705.• Bertram, D.F., Mackas, D.L., and Mckinnell, S.M. (2001). The seasonal cycle revisited: interannualvariation and ecosystem consequences. Progress in Oceanography 49, 283-307.• Blazewicz, S.J., Barnard, R.L., Daly, R.A., and Firestone, M.K. (2013). Evaluating rRNA as anindicator of microbial activity in environmental communities: limitations and uses. The ISME Journal7, 2061-2068.• Blumenberg, M., Seifert, R., and Michaelis, W. (2007). Aerobic methanotrophy in the oxic-anoxictransition zone of the Black Sea water column. Organic Geochemistry 38, 84-91.115• Blunden, J., Arndt, D.S., and Baringer, M.O. (2011). State of the Climate in 2010. Bulletin of theAmerican Meteorological Society 92, S1-S236.• Bograd, S.J., Castro, C.G., Di Lorenzo, E., Palacios, D.M., Bailey, H., Gilly, W., and Chavez, F.P.(2008). Oxygen declines and the shoaling of the hypoxic boundary in the California Current. Geo-physical Research Letters 35.• Bopp, L., Le Qur, C., Heimann, M., Manning, A.C., and Monfray, P. (2002). Climate-induced oceanicoxygen fluxes: Implications for the contemporary carbon budget. Global Biogeochemical Cycles 16,6-1-6-13.• Brte, J., Krabberd, A.K., Dolven, J.K., Ose, R.F., Kristensen, T., Bjrklund, K.R., and Shalchian-Tabrizi, K. (2012). Radiolaria Associated with Large Diversity of Marine Alveolates. Protist 163,767-777.• Breitburg, D., Levin, L.A., Oschlies, A., Grgoire, M., Chavez, F.P., Conley, D.J., Garon, V., Gilbert,D., Gutirrez, D., Isensee, K., Jacinto, G.S., Limburg, K.E., Montes, I., Naqvi, S.W.A., Pitcher, G.C.,Rabalais, N.N., Roman, M.R., Rose, K.A., Seibel, B.A., Telszewski, M., Yasuhara, M., and Zhang, J.(2018). Declining oxygen in the global ocean and coastal waters. Science 359.• Brown, M.V., and Donachie, S.P. (2007). Evidence for tropical endemicity in the DeltaproteobacteriaMarine Group B/SAR324 bacterioplankton clade. Aquatic Microbial Ecology 46, 107-115.• Brown, M.V., Philip, G.K., Bunge, J.A., Smith, M.C., Bissett, A., Lauro, F.M., Fuhrman, J.A., andDonachie, S.P. (2009). Microbial community structure in the North Pacific ocean. ISME J 3, 1374-1386.• Brum, J.R., Ignacio-Espinoza, J.C., Roux, S., Doulcier, G., Acinas, S.G., Alberti, A., Chaffron, S.,Cruaud, C., De Vargas, C., Gasol, J.M., Gorsky, G., Gregory, A.C., Guidi, L., Hingamp, P., Iudicone,D., Not, F., Ogata, H., Pesant, S., Poulos, B.T., Schwenck, S.M., Speich, S., Dimier, C., Kandels-Lewis, S., Picheral, M., Searson, S., Bork, P., Bowler, C., Sunagawa, S., Wincker, P., Karsenti, E., andSullivan, M.B. (2015). Patterns and ecological drivers of ocean viral communities. Science 348.• Cachon, J. (1964 ). Contribution a l’etude des peridiniens parasites. . Cytologie, cycles evolutifs.Annales des Sciences Naturelles -Zoologie et Biologie Animale 6, 1-158.• Canfield, D.E., Stewart, F.J., Thamdrup, B., De Brabandere, L., Dalsgaard, T., Delong, E.F., Revs-bech, N.P., and Ulloa, O. (2010). A cryptic sulfur cycle in oxygen-minimum-zone waters off theChilean coast. Science 330, 1375-1378.• Capelle, D., Dacey, J., and Tortell, P.D. (2015). An automated, high-throughput method for accurateand precise measurements of dissolved nitrous-oxide and methane concentrations in natural seawaters.Limnology and Oceanography: Methods.116• Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., Fierer,N., Pena, A.G., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley, S.T., Knights, D., Koenig, J.E.,Ley, R.E., Lozupone, C.A., Mcdonald, D., Muegge, B.D., Pirrung, M., Reeder, J., Sevinsky, J.R.,Turnbaugh, P.J., Walters, W.A., Widmann, J., Yatsunenko, T., Zaneveld, J., and Knight, R. (2010).QIIME allows analysis of high-throughput community sequencing data. Nat Methods 7, 335-336.• Cardoso, J.C.S., Baquero, C., and Almeida, P.S. (Year). ”Probabilistic Estimation of Network Sizeand Diameter”, in: Dependable Computing, 2009. LADC ’09. Fourth Latin-American Symposiumon), 33-40.• Carpenter, J.H. (1965). The Chesapeake Bay Institute Technique for the Winkler Dissolved OxygenMethod. Limnology and Oceanography 10, 141-143.• Carter, N.M. (1932). The oceanography of the fjords of southern British Columbia. Fish. Res. Bd.Canada Prog. Rept. Pacific Coast Sta. 12, 7-11.• Carter, N.M. (1934). Physiography and oceanography of some British Columbia fjords. Proc. Fifth.Pacific Sci. Cong. 1, 721-733.• Chaffron, S., Rehrauer, H., Pernthaler, J., and Von Mering, C. (2010). A global network of coexistingmicrobes from environmental and whole-genome sequence data. Genome Research 20, 947-959.• Chambouvet, A., Alves-De-Souza, C., Cueff, V., Marie, D., Karpov, S., and Guillou, L. (2011).Interplay between the parasite Amoebophrya sp.(Alveolata) and the cyst formation of the red tidedinoflagellate Scrippsiella trochoidea. Protist 162, 637-649.• Chambouvet, A., Morin, P., Marie, D., and Guillou, L. (2008). Control of toxic marine dinoflagellateblooms by serial parasitic killers. Science 322, 1254-1257.• Charlson, R.J., Lovelock, J.E., Andreae, M.O., and Warren, S.G. (1987). Oceanic phytoplankton,atmospheric sulphur, cloud albedo and climate. Nature 326, 655-661.• Chistoserdova, L., Jenkins, C., Kalyuzhnaya, M.G., Marx, C.J., Lapidus, A., Vorholt, J.A., Staley,J.T., and Lidstrom, M.E. (2004). The Enigmatic Planctomycetes May Hold a Key to the Origins ofMethanogenesis and Methylotrophy. Molecular Biology and Evolution 21, 1234-1241.• Chistoserdova, L., and Kalyuzhnaya, M.G. (2018). Current Trends in Methylotrophy. Trends inMicrobiology 26, 703-714.• Choi, C.J., Brosnahan, M.L., Sehein, T.R., Anderson, D.M., and Erdner, D.L. (2017). Insights intothe loss factors of phytoplankton blooms: The role of cell mortality in the decline of two inshoreAlexandrium blooms. Limnology and Oceanography 62, 1742-1753.• Chow, C.-E.T., Kim, D.Y., Sachdeva, R., Caron, D.A., and Fuhrman, J.A. (2014). Top-down controlson bacterial community structure: microbial network analysis of bacteria, T4-like viruses and protists.ISME J 8, 816-829.117• Cline, J.D. (1969). Spectrophotometric determination of hydrogen sulfide in natural waters. Limnol-ogy and Oceanography 14, 454-458.• Coats, D.W., And Heisler, J.J. (1989). Spatial and temporal occurrence of the parasitic dinoflagellateDubusquella cachoni and its tintinnine host Eutintinnus pectinis in Chesapeake Bay. Marine Biology101, 401-409.• Coats, D.W., Bachvaroff, T.R., and Delwiche, C.F. (2012). Revision of the family Duboscquellidaewith a description of Euduboscquella crenulata n. gen., n. sp. (Dinoflagellata, Syndinea), an intra-cellular parasite of the cilliate Favella panamensis Kofoid Campbell, 1929. Journal of EukaryoticMicrobiology 59, 1-11.• Coats, D.W., Bockstahler, K.R., Berg, G.M., and Sniezek, J.H. (1994). Dinoflagellate infections ofFavella panamensis from two North American estuaries. Marine Biology 119, 105-113.• Codispoti, L.A. (2010). Interesting Times for Marine N2O. Science 327, 1339.• Costa, C., Dijkema, C., Friedrich, M., Garca-Encina, P., Fernndez-Polanco, F., and Stams, A.J.M.(2000). Denitrification with methane as electron donor in oxygen-limited bioreactors. Applied Mi-crobiology and Biotechnology 53, 754-762.• Costello, A.M., and Lidstrom, M.E. (1999). Molecular Characterization of Functional and Phyloge-netic Genes from Natural Populations of Methanotrophs in Lake Sediments. Applied and Environ-mental Microbiology 65, 5066-5074.• Cottrell, M.T., and Kirchman, D.L. (2000). Natural Assemblages of Marine Proteobacteria and Mem-bers of the Cytophaga-Flavobacter Cluster Consuming Low- and High-Molecular-Weight DissolvedOrganic Matter. Applied and Environmental Microbiology 66, 1692-1697.• Cram, J.A., Xia, L.C., Needham, D.M., Sachdeva, R., Sun, F., and Fuhrman, J.A. (2015). Cross-depthanalysis of marine bacterial networks suggests downward propagation of temporal changes. ISME J9, 2573-2586.• Crump, B.C., Armbrust, E.V., and Baross, J.A. (1999). Phylogenetic Analysis of Particle-Attachedand Free-Living Bacterial Communities in the Columbia River, Its Estuary, and the Adjacent CoastalOcean. Applied and Environmental Microbiology 65, 3192-3204.• Cuba, R.M.F., Duarte, I.C., Saavedra, N.K., Varesche, M.B.A., and Foresti, E. (2011). Denitrificationcoupled with methane anoxic oxidation and microbial community involved identification. BrazilianArchives of Biology and Technology 54, 173-182.• Cuevas, L.A., and Morales, C.E. (2006). Nanoheterotroph grazing on bacteria and cyanobacteria inoxic and suboxic waters in coastal upwelling areas off northern Chile. Journal of Plankton Research28, 385-397.118• D’ambrosio, L., Ziervogel, K., Macgregor, B., Teske, A., and Arnosti, C. (2014). Composition andenzymatic function of particle-associated and free-living bacteria: a coastal/offshore comparison. TheISME Journal 8, 2167-2179.• De Caceres, M., and Legendre, P. (2009). Associations between species and groups of sites: indicesand statistical inference. Ecology. http://sites.google.com/site/miqueldecaceres/.• De Vargas, C., Audic, S., Henry, N., Decelle, J., Mahe, F., Logares, R., Lara, E., Berney, C., Bescot,N.L., Probert, I., Carmichael, M., Poulain, J., Romac, S., Colin, S., Aury, J.-M., Bittner, L., Chaffron,S., Dunthorn, M., Engelen, S., Flegontova, O., Guidi, L., Horak, A., Jaillon, O., Lima-Mendez, G.,Lukes, J., Malviya, S., Morard, R., Mulot, M., Scalco, E., Siano, R., Vincent, F., Zingone, A., Dimier,C., Picheral, M., Searson, S., Kandels-Lewis, S., Coordinators, T.O., Acinas, S.G., Bork, P., Bowler,C., Gorsky, G., Grimsley, N., Hingamp, P., Iudicone, D., Not, F., Ogata, H., Pesant, S., Raes, J.,Sieracki, M.E., Speich, S., Stemmann, L., Sunagawa, S., Weissenbach, J., Wincker, P., and Karsenti,E. (2015). Eukaryotic planktonic diversity in the sunlit ocean. Science 348.• Delong, E.F., Franks, D.G., and Alldredge, A.L. (1993). Phylogenetic diversity of aggregate-attachedvs. free-living marine bacterial assemblages. Limnology and Oceanography 38, 924-934.• Delong, E.F., Preston, C.M., Mincer, T., Rich, V., Hallam, S.J., Frigaard, N.-U., Martinez, A., Sulli-van, M.B., Edwards, R., Brito, B.R., Chisholm, S.W., and Karl, D.M. (2006). Community GenomicsAmong Stratified Microbial Assemblages in the Ocean 039;s Interior. Science 311, 496.• Desantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., Huber, T., Dalevi, D.,Hu, P., and Andersen, G.L. (2006). Greengenes, a Chimera-Checked 16S rRNA Gene Database andWorkbench Compatible with ARB. Applied and Environmental Microbiology 72, 5069-5072.• Diaz, R.J., and Rosenberg, R. (2008). Spreading dead zones and consequences for marine ecosystems.Science 321, 926-929.• Dick, G.J., Anantharaman, K., Baker, B.J., Li, M., Reed, D.C., and Sheik, C.S. (2013). The microbi-ology of deep-sea hydrothermal vent plumes: ecological and biogeographic linkages to seafloor andwater column habitats. Frontiers in Microbiology 4.• Dick, G.J., and Tebo, B.M. (2010). Microbial diversity and biogeochemistry of the Guaymas Basindeep-sea hydrothermal plume. Environmental Microbiology 12, 1334-1347.• Dolan, J.R., Montagnes, D.J.S,, Agathgam S,., Coats, D.W.,, Stoecker, D.K. Eds. (2013). The Biologyand Ecology of Tintinnid Ciliates, Models for Maroine Plankton. . Wiley-Blackwell, Hoboken, NJ,USA.• Dubilier, N., Bergin, C., and Lott, C. (2008). Symbiotic diversity in marine animals: the art ofharnessing chemosynthesis. Nat Rev Micro 6, 725-740.119• Dunfield, P.F., Yuryev, A., Senin, P., Smirnova, A.V., Stott, M.B., Hou, S., Ly, B., Saw, J.H., Zhou,Z., Ren, Y., Wang, J., Mountain, B.W., Crowe, M.A., Weatherby, T.M., Bodelier, P.L.E., Liesack, W.,Feng, L., Wang, L., and Alam, M. (2007). Methane oxidation by an extremely acidophilic bacteriumof the phylum Verrucomicrobia. Nature 450, 879-882.• Dunne, J.A., Williams, R.J., and Martinez, N.D. (2002). Food-web structure and network theory: Therole of connectance and size. Proceedings of the National Academy of Sciences 99, 12917-12922.• Duret, M.T., Pachiadaki, M.G., Stewart, F.J., Sarode, N., Christaki, U., Monchy, S., Srivastava, A., andEdgcomb, V.P. (2015). Size-fractionated diversity of eukaryotic microbial communities in the EasternTropical North Pacific oxygen minimum zone. FEMS Microbiology Ecology 91, fiv037-fiv037.• Edgcomb, V., Orsi, W., Bunge, J., Jeon, S., Christen, R., Leslin, C., Holder, M., Taylor, G.T., Suarez,P., Varela, R., and Epstein, S. (2011). Protistan microbial observatory in the Cariaco Basin, Caribbean.I. Pyrosequencing vs Sanger insights into species richness. The ISME Journal 5, 1344-1356.• Edgcomb, V.P. (2016). Marine protist associations and environmental impacts across trophic levels inthe twilight zone and below. Current Opinion in Microbiology 31, 169-175.• Edgcomb, V.P., Taylor, C., Pachiadaki, M.G., Honjo, S., Engstrom, I., and Yakimov, M. (2016).Comparison of Niskin vs. in situ approaches for analysis of gene expression in deep MediterraneanSea water samples. Deep Sea Research Part II: Topical Studies in Oceanography 129, 213-222.• Eiler, A., Drakare, S., Bertilsson, S., Pernthaler, J., Peura, S., Rofner, C., Simek, K., Yang, Y., Zna-chor, P., and Lindstrm, E.S. (2013). Unveiling Distribution Patterns of Freshwater Phytoplankton bya Next Generation Sequencing Based Approach. PLOS ONE 8, e53516.• Eloe, E.A., Shulse, C.N., Fadrosh, D.W., Williamson, S.J., Allen, E.E., and Bartlett, D.H. (2011).Compositional differences in particle-associated and free-living microbial assemblages from an ex-treme deep-ocean environment. Environmental Microbiology Reports 3, 449-458.• Elsaied, H.E., Hayashi, T., and Naganuma, T. (2004). Molecular Analysis of Deep-Sea HydrothermalVent Aerobic Methanotrophs by Targeting Genes of 16S rRNA and Particulate Methane Monooxyge-nase. Marine Biotechnology 6, 503-509.• Engelbrektson, A., Kunin, V., Wrighton, K.C., Zvenigorodsky, N., Chen, F., Ochman, H., and Hugen-holtz, P. (2010). Experimental factors affecting PCR-based estimates of microbial species richnessand evenness. ISME J 4, 642-647.• Escribano, R., Hidalgo, P., and Krautz, C. (2009). Zooplankton associated with the oxygen minimumzone system in the northern upwelling region of Chile during March 2000. Deep Sea Research PartII: Topical Studies in Oceanography 56, 1083-1094.• Ettwig, K.F., Butler, M.K., Le Paslier, D., Pelletier, E., Mangenot, S., Kuypers, M.M.M., Schreiber,F., Dutilh, B.E., Zedelius, J., De Beer, D., Gloerich, J., Wessels, H.J.C.T., Van Alen, T., Luesken, F.,120Wu, M.L., Van De Pas-Schoonen, K.T., Op Den Camp, H.J.M., Janssen-Megens, E.M., Francoijs, K.-J., Stunnenberg, H., Weissenbach, J., Jetten, M.S.M., and Strous, M. (2010). Nitrite-driven anaerobicmethane oxidation by oxygenic bacteria. Nature 464, 543-548.• Ettwig,K.F., B. Zhu, D. Speth, J.T. Keltjens, M.S. Jetten, B. Kartal. (2016). Archaea catalyze iron-dependent anaerobic oxidation of methane. Proc. Natl. Acad. Sci. U. S. A., 113,12792-12796.• Faust, K., Lima-Mendez, G., Lerat, J.-S., Sathirapongsasuti, J.F., Knight, R., Huttenhower, C., Lenaerts,T., and Raes, J. (2015). Cross-biome comparison of microbial association networks. Frontiers in Mi-crobiology 6, 1200.• Faust, K., and Raes, J. (2012). Microbial interactions: from networks to models. Nat Rev Micro 10,538-550.• Faust, K., Sathirapongsasuti, J.F., Izard, J., Segata, N., Gevers, D., Raes, J., and Huttenhower, C.(2012). Microbial Co-occurrence Relationships in the Human Microbiome. PLoS ComputationalBiology 8, e1002606.• Feike, J., Jurgens, K., Hollibaugh, J.T., Kruger, S., Jost, G., and Labrenz, M. (2012). Measuringunbiased metatranscriptomics in suboxic waters of the central Baltic Sea using a new in situ fixationsystem. ISME J 6, 461-470.• Ferrari, S., and Cribari-Neto, F. (2004). Beta Regression for Modelling Rates and Proportions. Journalof Applied Statistics 31, 799-815.• Field, K.G., Gordon, D., Wright, T., Rapp, M., Urback, E., Vergin, K., and Giovannoni, S.J. (1997).Diversity and depth-specific distribution of SAR11 cluster rRNA genes from marine planktonic bac-teria. Applied and Environmental Microbiology 63, 63-70.• Fox, J., and Weisberg, S. (2011). Multivariate linear models in R. An R Companion to AppliedRegression 2nd Ed.• Frias-Lopez, J., Shi, Y., Tyson, G.W., Coleman, M.L., Schuster, S.C., Chisholm, S.W., and Delong,E.F. (2008). Microbial community gene expression in ocean surface waters. Proceedings of theNational Academy of Sciences of the United States of America 105, 3805-3810.• Friedman, J., and Alm, E.J. (2012). Inferring Correlation Networks from Genomic Survey Data.PLOS Computational Biology 8, e1002687.• Fu,L., S.W. Li, Z.W. Ding, J. Ding, Y.Z. Lu, R.J. Zeng. (2016). Iron reduction in the DAMO/She-wanella oneidensis MR-1 coculture system and the fate of Fe(II). Water Res., 88,808-815.• Fuchsman, C.A., Kirkpatrick, J.B., Brazelton, W.J., Murray, J.W., and Staley, J.T. (2011). Metabolicstrategies of free-living and aggregate-associated bacterial communities inferred from biologic andchemical profiles in the Black Sea suboxic zone. FEMS Microbiology Ecology 78, 586-603.121• Fuchsman, C.A., Staley, J.T., Oakley, B.B., Kirkpatrick, J.B., and Murray, J.W. (2012). Free-livingand aggregate-associated Planctomycetes in the Black Sea. FEMS Microbiology Ecology 80, 402-416.• Fuenzalida, R., Schneider, W., Garcs-Vargas, J., Bravo, L., and Lange, C. (2009). Vertical and hori-zontal extension of the oxygen minimum zone in the eastern South Pacific Ocean. Deep Sea ResearchPart II: Topical Studies in Oceanography 56, 992-1003.• Fuhrman, J.A., Cram, J.A., and Needham, D.M. (2015). Marine microbial community dynamics andtheir ecological interpretation. Nature Reviews Microbiology 13, 133.• Fuhrman, J.A., and Davis, A.A. (1997). Widespread Archaea and novel bacteria from the deep sea asshown by 16S rRNA gene sequences. Oceanographic Literature Review 44, 1025.• Fukunaga, Y., Kurahashi, M., Sakiyama, Y., Ohuchi, M., Yokota, A., and Harayama, S. (2009).Phycisphaera mikurensis gen. nov., sp. nov., isolated from a marine alga, and proposal of Phycis-phaeraceae fam. nov., Phycisphaerales ord. nov. and Phycisphaerae classis nov. in the phylumPlanctomycetes. The Journal of General and Applied Microbiology 55, 267-275.• Fung, I., John, J., Lerner, J., Matthews, E., Prather, M., Steele, L.P., and Fraser, P.J. (1991). Three-dimensional model synthesis of the global methane cycle.• Gaebler-Schwarz, S., Davidson, A., Assmy, P., Chen, J., Henjes, J., Nthig, E.M., Lunau, M., andMedlin, L.K. (2010). A new cell stage in the haploid-diploid life cycle of the colony forming hapto-phyte Phaeocystis antartica and its ecological implications. Journal of Phycology 46, 1006-1016.• Ganesh, S., Bristow, L.A., Larsen, M., Sarode, N., Thamdrup, B., and Stewart, F.J. (2015). Size-fraction partitioning of community gene transcription and nitrogen metabolism in a marine oxygenminimum zone. ISME J 9, 2682-2696.• Ganesh, S., Parris, D.J., Delong, E.F., and Stewart, F.J. (2014). Metagenomic analysis of size-fractionated picoplankton in a marine oxygen minimum zone. ISME J 8, 187-211.• Gardner, W.D. (1977). Incomplete extraction of rapidly settling particles from water samplers1. Lim-nology and Oceanography 22, 764-768.• Garfield, P.C., Packard, T.T., Friederich, G.E., and Codispoti, L.A. (1983). A subsurface particlemaximum layer and enhanced microbial activity in the secondary nitrite maximum of the northeasterntropical Pacific Ocean. Journal of Marine Research 41, 747-768.• Gargett, A.E., Stucchi, D., and Whitney, F. (2003). Physical processes associated with high primaryproduction in Saanich Inlet, British Columbia. Estuarine, Coastal and Shelf Science 56, 1141-1156.• Ghiglione, J.-F., Conan, P., and Pujo-Pay, M. (2009). Diversity of total and active free-living vs.particle-attached bacteria in the euphotic zone of the NW Mediterranean Sea. FEMS MicrobiologyLetters 299, 9-21.122• Gies, E.A., Konwar, K.M., Beatty, J.T., and Hallam, S.J. (2014). Illuminating Microbial Dark Matterin Meromictic Sakinaw Lake. Applied and Environmental Microbiology 80, 6807-6818.• Gilbert, J.A., Meyer, F., Schriml, L., Joint, I.R., Mhling, M., and Field, D. (2010). Metagenomes andmetatranscriptomes from the L4 long-term coastal monitoring station in the Western English Channel.Standards in Genomic Sciences 3, 183-193.• Gilbert, J.A., Steele, J.A., Caporaso, J.G., Steinbrck, L., Reeder, J., Temperton, B., Huse, S., Mchardy,A.C., Knight, R., Joint, I., Somerfield, P., Fuhrman, J.A., and Field, D. (2011). Defining seasonalmarine microbial community dynamics. The Isme Journal 6, 298.• Glaubitz, S., Labrenz, M., Jost, G., and Jrgens, K. (2010). Diversity of active chemolithoautotrophicprokaryotes in the sulfidic zone of a Black Sea pelagic redoxcline as determined by rRNA-based stableisotope probing. FEMS Microbiology Ecology 74, 32-41.• Gmez, F., Moreira, D., and Lpez-Garca, P. (2009). Life cycle and molecular phylogeny of the di-noflagellates Chytriodinium and Dissodinium, ectoparasites of copepod eggs. European journal ofprotistology 45, 260-270.• Grossart, H.-P. (2010). Ecological consequences of bacterioplankton lifestyles: changes in conceptsare needed. Environmental Microbiology Reports 2, 706-714.• Guillou, L., Viprey, M., Chambouvet, A., Welsh, R., Kirkham, A., Massana, R., Scanlan, D.J., andWorden, A. (2008). Widespread occurrence and genetic diversity of marine parasitoids belonging toSyndiniales (Alveolata). Environmental Microbiology 10, 3349-3365.• Guindon, S., Lethiec, F., Duroux, P., and Gascuel, O. (2005). PHYML Online: a web server for fastmaximum likelihood-based phylogenetic inference. Nucleic Acids Research 33, W557-W559.• Hahn, A.S., Konwar, K.M., Louca, S., Hanson, N.W., and Hallam, S.J. (2016). The informationscience of microbial ecology. Curr Opin Microbiol 31, 209-216.• Hamady, M., Walker, J.J., Harris, J.K., Gold, N.J., and Knight, R. (2008). Error-correcting barcodedprimers allow hundreds of samples to be pyrosequenced in multiplex. Nature methods 5, 235-237.• Hanson, R.S., and Hanson, T.E. (1996). Methanotrophic bacteria. Microbiological Reviews 60, 439-471.• Hargreaves, N., Ware, D., and Mcfarlane, G. (1994). Return of Pacific sardine (Sardinops sagax) tothe British Columbia coast in 1992. Canadian Journal of Fisheries and Aquatic Sciences 51, 460-463.• Haroon, M.F. (2013). Anaerobic oxidation of methane coupled to nitrate reduction in a novel archaeallineage. Nature. 500, 567570123• Hawley, A.K., Brewer, H.M., Norbeck, A.D., Pasa-Tolic, L., and Hallam, S.J. (2014). Metaproteomicsreveals differential modes of metabolic coupling among ubiquitous oxygen minimum zone microbes.Proc Natl Acad Sci U S A 111, 11395-11400.• Hawley, A.K., Nobu, M.K., Wright, J.J., Durno, W.E., Morgan-Lang, C., Sage, B., Schwientek, P.,Swan, B.K., Rinke, C., Torres-Beltra´n, M., Mewis, K., Liu, W.-T., Stepanauskas, R., Woyke, T., andHallam, S.J. (2017a). Diverse Marinimicrobia bacteria may mediate coupled biogeochemical cyclesalong eco-thermodynamic gradients. Nature Communications 8, 1507.• Hawley, A.K., Torres-Beltra´n, M., Zaikova, E., Walsh, D.A., Mueller, A., Scofield, M., Kheirandish,S., Payne, C., Pakhomova, L., Bhatia, M., Shevchuk, O., Gies, E.A., Fairley, D., Malfatti, S.A.,Norbeck, A.D., Brewer, H.M., Pasa-Tolic, L., Del Rio, T.G., Suttle, C.A., Tringe, S., and Hallam, S.J.(2017b). A compendium of multi-omic sequence information from the Saanich Inlet water column.Scientific Data 4, 170160.• Hayashi, T., Obata, H., Gamo, T., Sano, Y., and Naganuma, T. (2007). Distribution and phylogeneticcharacteristics of the genes encoding enzymes relevant to methane oxidation in oxygen minimumzones of the Eastern Pacific Ocean. Res. J. Environ. Sci. 1, 275-284.• Heintz, M.B., Mau, S., and Valentine, D.L. (2012). Physical control on methanotrophic potential inwaters of the Santa Monica Basin, Southern California. Limnology and Oceanography 57, 420-432.• Helm, K.P., Bindoff, N.L., and Church, J.A. (2011). Observed decreases in oxygen content of theglobal ocean. Geophysical Research Letters 38,.• Herlinveaux, R.H. (1962). Oceanography of Saanich Inlet in Vancouver Island, British Columbia.Journal of the Fisheries Research Board of Canada 19, 1-37.• Hernandez, M.E., Beck, D.a.C., Lidstrom, M.E., and Chistoserdova, L. (2015). Oxygen availabil-ity is a major factor in determining the composition of microbial communities involved in methaneoxidation. PeerJ 3, e801.• Hewson, I., and Fuhrman, J.A. (2006). Improved Strategy for Comparing Microbial AssemblageFingerprints. Microbial Ecology 51, 147-153.• Heyer, J., Galchenko, V.F., and Dunfield, P.F. (2002). Molecular phylogeny of type II methane-oxidizing bacteria isolated from various environments. Microbiology 148, 2831-2846.• Holmes, A.J., Costello, A., Lidstrom, M.E., and Murrell, J.C. (1995). Evidence that participatemethane monooxygenase and ammonia monooxygenase may be evolutionarily related. FEMS Mi-crobiology Letters 132, 203-208.• Holmes, R.M., Aminot, A., Krouel, R., Hooker, B.A., and Peterson, B.J. (1999). A simple andprecise method for measuring ammonium in marine and freshwater ecosystems. Can J Fish Aquat Sci56, 1801-1808.124• Hosack, D.A., Dennis, G., Sherman, B.T., Lane, H.C., and Lempicki, R.A. (2003). Identifying bio-logical themes within lists of genes with EASE. Genome Biology 4, R70.• Hurwitz, B.L., Deng, L., Poulos, B.T., and Sullivan, M.B. (2013). Evaluation of methods to concen-trate and purify ocean virus communities through comparative, replicated metagenomics. Environ-mental Microbiology 15, 1428-1440.• Ipcc (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I tothe Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UnitedKingdom and New York, NY, USA: Cambridge University Press.• James, T.H., Patricia, S.W., and Michael, C.M. (2000). Similarity of particle-associated and free-living bacterial communities in northern San Francisco Bay, California. Aquatic Microbial Ecology21, 103-114.• Jephcott, T.G., Alves-De-Souza, C., Gleason, F.H., Van Ogtrop, F.F., Sime-Ngando, T., Karpov, S.A.,and Guillou, L. (2016). Ecological impacts of parasitic chytrids, syndiniales and perkinsids on popu-lations of marine photosynthetic dinoflagellates. Fungal Ecology 19, 47-58.• Jing, H., Rocke, E., Kong, L., Xia, X., Liu, H., and Landry, M.R. (2015). Protist communities in amarine oxygen minimum zone off Costa Rica by 454 pyrosequencing. Biogeosciences Discuss. 2015,13483-13509.• Jnasdttir, S.H., Visser, A.W., Richardson, K., and Heath, M.R. (2015). Seasonal copepod lipid pumppromotes carbon sequestration in the deep North Atlantic. Proceedings of the National Academy ofSciences 112, 12122-12126.• Kalyuzhnaya, M.G., Yang, S., Rozova, O.N., Smalley, N.E., Clubb, J., Lamb, A., Gowda, G.a.N.,Raftery, D., Fu, Y., Bringel, F., Vuilleumier, S., Beck, D.a.C., Trotsenko, Y.A., Khmelenina, V.N., andLidstrom, M.E. (2013). Highly efficient methane biocatalysis revealed in a methanotrophic bacterium.Nature Communications 4, 2785.• Karl, D.M., Beversdorf, L., Bjorkman, K.M., Church, M.J., Martinez, A., and Delong, E.F. (2008).Aerobic production of methane in the sea. Nature Geosci 1, 473-478.• Karl, D.M., and Church, M.J. (2014). Microbial oceanography and the Hawaii Ocean Time-seriesprogramme. Nat Rev Microbiol 12, 699-713.• Karl, D.M., and Tilbrook, B.D. (1994). Production and transport of methane in oceanic particulateorganic matter. Nature 368, 732-734.• Karwautz, C., Kus, G., Stckl, M., Neu, T.R., and Lueders, T. (2018). Microbial megacities fueled bymethane oxidation in a mineral spring cave. The ISME Journal 12, 87-100.• Kassambara, A. (2017). ggpubr: ’ggplot2’ Based Publication Ready Plots. . R package version 0.1.5.125• Keeling, R.E., Kortzinger, A., and Gruber, N. (2010). Ocean deoxygenation in a warming world. AnnRev Mar Sci 2, 199-229.• Keller, D.P., Kriest, I., Koeve, W., and Oschlies, A. (2016). Southern Ocean biological impacts onglobal ocean oxygen. Geophysical Research Letters 43, 6469-6477.• Kessler, J.D., Valentine, D.L., Redmond, M.C., Du, M., Chan, E.W., Mendes, S.D., Quiroz, E.W.,Villanueva, C.J., Shusta, S.S., Werra, L.M., Yvon-Lewis, S.A., and Weber, T.C. (2011). A PersistentOxygen Anomaly Reveals the Fate of Spilled Methane in the Deep Gulf of Mexico. Science 331,312-315.• Kettle, A., Andreae, M., Amouroux, D., Andreae, T., Bates, T., Berresheim, H., Bingemer, H., Boni-forti, R., Curran, M., and Ditullio, G. (1999). A global database of sea surface dimethylsulfide (DMS)measurements and a procedure to predict sea surface DMS as a function of latitude, longitude, andmonth. Global Biogeochemical Cycles 13, 399-444.• Kiene, R.P. (1991). Production and consumption of methane in aquatic systems. American Societyfor Microbiology.• Kieschnick, R., and Mccullough, B.D. (2003). Regression analysis of variates observed on (0, 1):percentages, proportions and fractions. Statistical Modelling 3, 193-213.• Kim, S., Gupta, N., and Pevzner, P.A. (2008). Spectral Probabilities and Generating Functions ofTandem Mass Spectra: A Strike against Decoy Databases. Journal of Proteome Research 7, 3354-3363.• Kimmerer, W., and Mckinnon, A. (1990). High mortality in a copepod population caused by a para-sitic dinoflagellate. Marine Biology 107, 449-452.• Kirkpatrick, J., Oakley, B., Fuchsman, C., Srinivasan, S., Staley, J.T., and Murray, J.W. (2006). Diver-sity and Distribution of Planctomycetes and Related Bacteria in the Suboxic Zone of the Black Sea.Applied and Environmental Microbiology 72, 3079-3083.• Kits, K.D., Klotz, M.G., and Stein, L.Y. (2015). Methane oxidation coupled to nitrate reductionunder hypoxia by the Gammaproteobacterium Methylomonas denitrificans, sp. nov. type strain FJG1.Environmental Microbiology 17, 3219-3232.• Knief, C. (2015). Diversity and Habitat Preferences of Cultivated and Uncultivated Aerobic Methan-otrophic Bacteria Evaluated Based on pmoA as Molecular Marker. Frontiers in Microbiology 6, 1346.• Knittel, K., and Boetius, A. (2009). Anaerobic Oxidation of Methane: Progress with an UnknownProcess. Annual Review of Microbiology 63, 311-334.• Kolber, Z., and Falkowski, P.G. (1993). Use of active fluorescence to estimate phytoplankton photo-synthesis in situ. Limnology and Oceanography 38, 1646-1665.126• Konwar, K.M., Hanson, N.W., Bhatia, M.P., Kim, D., Wu, S.-J., Hahn, A.S., Morgan-Lang, C., Che-ung, H.K., and Hallam, S.J. (2015). MetaPathways v2.5: quantitative functional, taxonomic andusability improvements. Bioinformatics 31, 3345-3347.• Konwar, K.M., Hanson, N.W., Page, A.P., and Hallam, S.J. (2013). MetaPathways: a modular pipelinefor constructing pathway/genome databases from environmental sequence information. BMC Bioin-formatics 14, 202.• Krause, S.M.B., Johnson, T., Samadhi Karunaratne, Y., Fu, Y., Beck, D.a.C., Chistoserdova, L., andLidstrom, M.E. (2017). Lanthanide-dependent cross-feeding of methane-derived carbon is linked bymicrobial community interactions. Proceedings of the National Academy of Sciences 114, 358.• Krzywinski, M., Birol, I., Jones, S.J., and Marra, M.A. (2012). Hive plots rational approach to visu-alizing networks. Briefings in Bioinformatics 13, 627-644.• Lam, P., and Kuypers, M.M.M. (2011). Microbial Nitrogen Cycling Processes in Oxygen MinimumZones. Annual Review of Marine Science 3, 317-345.• Lam, P., and Kuypers, M.M. (2009). Revising the nitrogen cycle in the Peruvian oxygen minimumzone. Proc Natl Acad Sci U S A 106, 4752-4757.• Lamontagne, M.G., and Holden, P.A. (2003). Comparison of free-living and particle-associated bac-terial communities in a coastal Lagoon. Microbial Ecology 46, 228-237.• Latora, V., and Marchiori, M. (2009). Is the Boston Subway a Small-World Network• Lee, B.-S., Bullister, J.L., and Whitney, F.A. (1999). Chlorofluorocarbon CFC-11 and carbon tetra-chloride removal in Saanich Inlet, an intermittently anoxic basin. Marine Chemistry 66, 171-185.• Legendre, P., and Gallagher, E.D. (2001). Ecologically meaningful transformations for ordination ofspecies data. Oecologia 129, 271-280.• Lepp, P.W., and Schmidt, T.M. (1998). Nucleic acid content of Synechococcus spp. during growth incontinuous light and light/dark cycles. Archives of Microbiology 170, 201-207.• Lesniewski, R.A., Jain, S., Anantharaman, K., Schloss, P.D., and Dick, G.J. (2012). The meta-transcriptome of a deep-sea hydrothermal plume is dominated by water column methanotrophs andlithotrophs. The ISME Journal 6, 2257-2268.• Lewis, M.R., Warnock, R.E., and Platt, T. (1985). Absorption and photosynthetic action spectra fornatural phytoplankton populations: Implications for production in the open ocean1. Limnology andOceanography 30, 794-806.• Li, M., Wang, J., and Chen, J. (Year). ”A Fast Agglomerate Algorithm for Mining Functional Modulesin Protein Interaction Networks”, in: 2008 International Conference on BioMedical Engineering andInformatics), 3-7.127• Lilley, M.D., Baross, J.A., and Gordon, L.I. (1982). Dissolved hydrogen and methane in SaanichInlet, British Columbia. Deep Sea Research Part A. Oceanographic Research Papers 29, 1471-1484.• Lin, X., Scranton, M.I., Varela, R., Chistoserdov, A., and Taylor, G.T. (2007). Compositional re-sponses of bacterial communities to redox gradients and grazing in the anoxic Cariaco Basin. AquaticMicrobial Ecology 47, 57-72.• Llorens-Mares, T., Yooseph, S., Goll, J., Hoffman, J., Vila-Costa, M., Borrego, C.M., Dupont, C.L.,and Casamayor, E.O. (2015). Connecting biodiversity and potential functional role in modern euxinicenvironments by microbial metagenomics. ISME J 9, 1648-1661.• Lopez-Garcia, P., Rodriguez-Valera, F., Pedros-Alio, C., and Moreira, D. (2001). Unexpected diver-sity of small eukaryotes in deep-sea Antarctic plankton. Nature 409, 603-607.• Louca, S., Hawley, A.K., Katsev, S., Torres-Beltra´n, M., Bhatia, M.P., Kheirandish, S., Michiels,C.C., Capelle, D., Lavik, G., Doebeli, M., Crowe, S.A., and Hallam, S.J. (2016). Integrating biogeo-chemistry with multiomic sequence information in a model oxygen minimum zone. Proceedings ofthe National Academy of Sciences 113, E5925-E5933.• Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar, Buchner, A., Lai, T.,Steppi, S., Jobb, G., Frster, W., Brettske, I., Gerber, S., Ginhart, A.W., Gross, O., Grumann, S.,Hermann, S., Jost, R., Knig, A., Liss, T., Lmann, R., May, M., Nonhoff, B., Reichel, B., Strehlow,R., Stamatakis, A., Stuckmann, N., Vilbig, A., Lenke, M., Ludwig, T., Bode, A., and Schleifer, K.H.(2004). ARB: a software environment for sequence data. Nucleic Acids Research 32, 1363-1371.• Lke, C., Speth, D.R., Kox, M.a.R., Villanueva, L., and Jetten, M.S.M. (2016). Metagenomic analysisof nitrogen and methane cycling in the Arabian Sea oxygen minimum zone. PeerJ 4, e1924.• Maclean, D., Jones, J.D., and Studholme, D.J. (2009). Application of ’next-generation’ sequencingtechnologies to microbial genetics. Nature Reviews Microbiology 7, 260.• Marcy, Y., Ouverney, C., Bik, E.M., Lsekann, T., Ivanova, N., Martin, H.G., Szeto, E., Platt, D.,Hugenholtz, P., Relman, D.A., and Quake, S.R. (2007). Dissecting biological ’dark matter” withsingle-cell genetic analysis of rare and uncultivated TM7 microbes from the human mouth. Proceed-ings of the National Academy of Sciences of the United States of America 104, 11889-11894.• Massana, R., Balagu, V., Guillou, L., and Pedros-Alio, C. (2004a). Picoeukaryotic diversity in anoligotrophic coastal site studied by molecular and culturing approaches. FEMS microbiology ecology50, 231-243.• Massana, R., Castresana, J., Balagu, V., Guillou, L., Romari, K., Groisillier, A., Valentin, K., andPedrs-Ali, C. (2004b). Phylogenetic and ecological analysis of novel marine stramenopiles. Appliedand environmental microbiology 70, 3528-3534.128• Massana, R., Del Campo, J., Sieracki, M.E., Audic, S., and Logares, R. (2014). Exploring the uncul-tured microeukaryote majority in the oceans: reevaluation of ribogroups within stramenopiles. ISMEJ 8, 854-866.• Massana, R., Murray, A.E., Preston, C.M., and Delong, E.F. (1997). Vertical distribution and phy-logenetic characterization of marine planktonic Archaea in the Santa Barbara Channel. Applied andEnvironmental Microbiology 63, 50-56.• Mau, S., Blees, J., Helmke, E., Niemann, H., and Damm, E. (2013). Vertical distribution of methaneoxidation and methanotrophic response to elevated methane concentrations in stratified waters of theArctic fjord Storfjorden (Svalbard, Norway). Biogeosciences 10, 6267-6278.• Mcdonald, I.R., and Murrell, J.C. (1997). The particulate methane monooxygenase gene pmoA andits use as a functional gene probe for methanotrophs. FEMS Microbiology Letters 156, 205-210.• Metcalf, W.W., Griffin, B.M., Cicchillo, R.M., Gao, J., Janga, S.C., Cooke, H.A., Circello, B.T.,Evans, B.S., Martens-Habbena, W., Stahl, D.A., and Van Der Donk, W.A. (2012). Synthesis ofmethylphosphonic acid by marine microbes: a source for methane in the aerobic ocean. Science(New York, N.Y.) 337, 1104-1107.• Milucka.G. Ferdelman, L. Polerecky, D. Franzke, G. Wegener, M. Schmid, et al. (2012).Zero-valentsulphur is a key intermediate in marine methane oxidation Nature, 491,541-546.• Modin, O., Fukushi, K., and Yamamoto, K. (2007). Denitrification with methane as external carbonsource. Water Research 41, 2726-2738.• Moeseneder, M.M., Winter, C., and Herndl, G.J. (2001). Horizontal and vertical complexity of at-tached and free-living bacteria of the eastern Mediterranean Sea, determined by 16S rDNA and 16SrRNA fingerprints. Limnology and Oceanography 46, 95-107.• Montagnes, D.J.S., Dower, J.F., and Figueiredo, G.M. (2010). The Protozooplankton-IchthyoplanktonTrophic Link: An Overlooked Aspect of Aquatic Food Webs1. Journal of Eukaryotic Microbiology57, 223-228.• Monteiro, P.M.S., Van Der Plas, A., Mohrholz, V., Mabille, E., Pascall, A., and Joubert, W. (2006).Variability of natural hypoxia and methane in a coastal upwelling system: Oceanic physics or shelfbiology. Geophysical Research Letters 33,.• Moon-Van Der Staay, S.Y., De Wachter, R., and Vaulot, D. (2001). Oceanic 18S rDNA sequencesfrom picoplankton reveal unsuspected eukaryotic diversity. Nature 409, 607-610.• Mou, X., Sun, S., Edwards, R.A., Hodson, R.E., and Moran, M.A. (2008). Bacterial carbon processingby generalist species in the coastal ocean. Nature 451, 708-711.• Murphy, J., and Riley, J.P. (1962). A modified single solution method for the determination of phos-phate in natural waters. Analytica Chimica Acta 27, 31-36.129• Murray, A.E., Blakis, A., Massana, R., Strawzewski, S., Passow, U., Alldredge, A., and Delong, E.F.(1999). A time series assessment of planktonic archaeal variability in the Santa Barbara Channel.Aquatic Microbial Ecology 20, 129-145.• Murrell, J.C. (2010). ”The Aerobic Methane Oxidizing Bacteria (Methanotrophs),” in Handbook ofHydrocarbon and Lipid Microbiology, ed. K. Timmis. Springer Berlin Heidelberg), 1953-1966.• McGlynn G.L. Chadwick, C.P. Kempes, V.J. Orphan. (2015).Single cell activity reveals direct elec-tron transfer in methanotrophic consortia. Nature, 526,531-535.• Naqvi, S.W.A., Bange, H.W., Faras, L., Monteiro, P.M.S., Scranton, M.I., and Zhang, J. (2010).Marine hypoxia/anoxia as a source of CH4 and N2O. Biogeosciences 7, 2159-2190.• Naqvi, S.W.A., Kumar, M.D., Narvekar, P.V., De Sousa, S.N., George, M.D., and D’silva, C. (1993).An intermediate nepheloid layer associated with high microbial metabolic rates and denitrification inthe northwest Indian Ocean. Journal of Geophysical Research: Oceans 98, 16469-16479.• Not, F., Campo, J.D., Balague, V., Vargas, C.D., and Massana, R. (2009). New insights into thediversity of marine picoeukaryotes. PLoS ONE 4.• Not, F., Valentin, K., Romari, K., Lovejoy, C., Massana, R., Tbe, K., Vaulot, D., and Medlin, L.K.(2007). Picobiliphytes: a marine picoplanktonic algal group with unknown affinities to other eukary-otes. science 315, 253-255.• Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., R.Minchin, P., O’hara, R.B., Simpson, G.L.,Solymos, P., Henry, M., Stevens, H., and Wagner, H. (2015). vegan: Community Ecology Package.http://CRAN.R-project.org/package=vegan R package version 2.3-1.• Orsi, W., Song, Y.C., Hallam, S., and Edgcomb, V. (2012). Effect of oxygen minimum zone formationon communities of marine protists. The ISME Journal 6, 1586-1601.• Orsi, W.D., Smith, J.M., Wilcox, H.M., Swalwell, J.E., Carini, P., Worden, A.Z., and Santoro, A.E.(2015). Ecophysiology of uncultivated marine euryarchaea is linked to particulate organic matter.ISME J 9, 1747-1763.• Oshkin, I.Y., Beck, D.a.C., Lamb, A.E., Tchesnokova, V., Benuska, G., Mctaggart, T.L., Kalyuzhnaya,M.G., Dedysh, S.N., Lidstrom, M.E., and Chistoserdova, L. (2015). Methane-fed microbial micro-cosms show differential community dynamics and pinpoint taxa involved in communal response. TheISME Journal 9, 1119-1129.• Pack, M.A., Heintz, M.B., Reeburgh, W.S., Trumbore, S.E., Valentine, D.L., Xu, X., and Druffel,E.R.M. (2015). Methane oxidation in the eastern tropical North Pacific Ocean water column. Journalof Geophysical Research: Biogeosciences 120, 1078-1092.130• Padilla, C.C., Bertagnolli, A.D., Bristow, L.A., Sarode, N., Glass, J.B., Thamdrup, B., and Stew-art, F.J. (2017). Metagenomic Binning Recovers a Transcriptionally Active GammaproteobacteriumLinking Methanotrophy to Partial Denitrification in an Anoxic Oxygen Minimum Zone. Frontiers inMarine Science 4, 23.• Padilla, C.C., Bristow, L.A., Sarode, N., Garcia-Robledo, E., Gomez Ramirez, E., Benson, C.R.,Bourbonnais, A., Altabet, M.A., Girguis, P.R., Thamdrup, B., and Stewart, F.J. (2016). NC10 bacteriain marine oxygen minimum zones. ISME J.• Padilla, C.C., Ganesh, S., Gantt, S., Huhman, A., Parris, D.J., Sarode, N., and Stewart, F.J. (2015).Standard filtration practices may significantly distort planktonic microbial diversity estimates. Fron-tiers in Microbiology 6, 547.• Park, S., Brown, K.W., and Thomas, J.C. (2002). The effect of various environmental and designparameters on methane oxidation in a model biofilter. Waste Management Research 20, 434-444.• Parris, D.J., Ganesh, S., Edgcomb, V.P., Delong, E.F., and Stewart, F.J. (2014). Microbial eukaryotediversity in the marine oxygen minimum zone off northern Chile. Frontiers in Microbiology 5, 543.• Paulmier, A., and Ruiz-Pino, D. (2009). Oxygen minimum zones (OMZs) in the modern ocean.Progress in Oceanography 80, 113-128.• Pedrs-Ali, C., Caldern-Paz, J.I., and Gasol, J.M. (2000). Comparative analysis shows that bacterivory,not viral lysis, controls the abundance of heterotrophic prokaryotic plankton. FEMS MicrobiologyEcology 32, 157-165.• Perez, S. (2015). Exploring microbial community structure and resilience through visualization andanalysis of microbial co-occurence networks. The University of British Columbia.• Pesant, S., Not, F., Picheral, M., Kandels-Lewis, S., Le Bescot, N., Gorsky, G., Iudicone, D., Karsenti,E., Speich, S., Troubl, R., Dimier, C., and Searson, S. (2015). Open science resources for the discov-ery and analysis of Tara Oceans data. Scientific Data 2, 150023.• Petersen, J.M., and Dubilier, N. (2009). Methanotrophic symbioses in marine invertebrates. Environ-mental Microbiology Reports 1, 319-335.• Peura, S., Bertilsson, S., Jones, R.I., and Eiler, A. (2015). Resistant Microbial Cooccurrence PatternsInferred by Network Topology. Applied and Environmental Microbiology 81, 2090.• Proulx, S.R., Promislow, D.E.L., and Phillips, P.C. (2005). Network thinking in ecology and evolu-tion. Trends in Ecology Evolution 20, 345-353.• Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., Peplies, J., and Glckner, F.O. (2007).SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequencedata compatible with ARB. Nucleic Acids Research 35, 7188-7196.131• Rapp, M.S., and Giovannoni, S.J. (2003). The Uncultured Microbial Majority. Annual Review ofMicrobiology 57, 369-394.• Rcoreteam (2013). R: A language and environment for statistical computing. R Foundation for Sta-tistical Computing, Vienna, Austria. URL http://www.R-project.org/.• Reeburgh, W.S. (2007). Oceanic Methane Biogeochemistry. Chemical Reviews 107, 486-513.• Reeburgh, W.S., Ward, B.B., Whalen, S.C., Sandbeck, K.A., Kilpatrick, K.A., and Kerkhof, L.J.(1991). Black Sea methane geochemistry. Deep Sea Research Part A. Oceanographic Research Papers38, Supplement 2, S1189-S1210.• Richards, T.A., and Bass, D. (2005). Molecular screening of free-living microbial eukaryotes: diver-sity and distribution using a meta-analysis. Current Opinion in Microbiology 8, 240-252.• Rinke, C., Schwientek, P., Sczyrba, A., Ivanova, N.N., Anderson, I.J., Cheng, J.-F., Darling, A.,Malfatti, S., Swan, B.K., Gies, E.A., Dodsworth, J.A., Hedlund, B.P., Tsiamis, G., Sievert, S.M., Liu,W.-T., Eisen, J.A., Hallam, S.J., Kyrpides, N.C., Stepanauskas, R., Rubin, E.M., Hugenholtz, P., andWoyke, T. (2013). Insights into the phylogeny and coding potential of microbial dark matter. Nature499, 431-437.• Rodriguez-Mora, M.J., Scranton, M.I., Taylor, G.T., and Chistoserdov, A.Y. (2013). Bacterial com-munity composition in a large marine anoxic basin: a Cariaco Basin time-series survey. FEMS Mi-crobiology Ecology 84, 625-639.• Romari, K., and Vaulot, D. (2004). Composition and temporal variability of picoeukaryote communi-ties at a coastal site of the English Channel from 18S rDNA sequences. Limnology and Oceanography49, 784-798.• Roslev, P., and King, G.M. (1995). Aerobic and Anaerobic Starvation Metabolism in MethanotrophicBacteria. Applied and Environmental Microbiology 61, 1563-1570.• Rusch, D.B., Halpern, A.L., Sutton, G., Heidelberg, K.B., Williamson, S., Yooseph, S., Wu, D., Eisen,J.A., Hoffman, J.M., Remington, K., Beeson, K., Tran, B., Smith, H., Baden-Tillson, H., Stewart,C., Thorpe, J., Freeman, J., Andrews-Pfannkoch, C., Venter, J.E., Li, K., Kravitz, S., Heidelberg,J.F., Utterback, T., Rogers, Y.-H., Falcn, L.I., Souza, V., Bonilla-Rosso, G., Eguiarte, L.E., Karl,D.M., Sathyendranath, S., Platt, T., Bermingham, E., Gallardo, V., Tamayo-Castillo, G., Ferrari, M.R.,Strausberg, R.L., Nealson, K., Friedman, R., Frazier, M., and Venter, J.C. (2007). The SorcererII Global Ocean Sampling Expedition: Northwest Atlantic through Eastern Tropical Pacific. PLoSBiology 5, e77.• Salcher, M.M., Posch, T., and Pernthaler, J. (2013). In situ substrate preferences of abundant bacteri-oplankton populations in a prealpine freshwater lake. The ISME Journal 7, 896-907.132• Sansone, F.J., Popp, B.N., Gasc, A., Graham, A.W., and Rust, T.M. (2001). Highly elevated methanein the eastern tropical North Pacific and associated isotopically enriched fluxes to the atmosphere.Geophysical Research Letters 28, 4567-4570.• Santelli, C.M., Orcutt, B.N., Banning, E., Bach, W., Moyer, C.L., Sogin, M.L., Staudigel, H., andEdwards, K.J. (2008). Abundance and diversity of microbial life in ocean crust. Nature 453, 653-656.• Sauter, L.M., Latypova, E., Smalley, N.E., Lidstrom, M.E., Hallam, S., and Kalyuzhnaya, M.G.(2012). Methanotrophic communities of Saanich Inlet: A microcosm perspective. Systematic andApplied Microbiology 35, 198-203.• Scheller,S., H. Yu, G.L. Chadwick, S.E. McGlynn, V.J. Orphan. (2016). Artificial electron acceptorsdecouple archaeal methane oxidation from sulfate reduction Science, 351,703-707.• Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski, R.A.,Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Van Horn, D.J.,and Weber, C.F. (2009). Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities. Applied and Environ-mental Microbiology 75, 7537-7541.• Schmidtko, S., Stramma, L., and Visbeck, M. (2017). Decline in global oceanic oxygen content duringthe past five decades. Nature 542, 335-339.• Semrau, J.D., Dispirito, A.A., and Yoon, S. (2010). Methanotrophs and copper. FEMS MicrobiologyReviews 34, 496-531.• Shanks, A.L., and Reeder, M.L. (1993). Reducing microzones and sulfide production in marine snow.Mar.Ecol.Prog.Ser 96, 43-47.• Shannon, P., Markiel, A., Ozier, O., Baliga, N.S., Wang, J.T., Ramage, D., Amin, N., Schwikowski,B., and Ideker, T. (2003). Cytoscape: A Software Environment for Integrated Models of BiomolecularInteraction Networks. Genome Research 13, 2498-2504.• Shi, Y., Tyson, G.W., and Delong, E.F. (2009). Metatranscriptomics reveals unique microbial smallRNAs in the ocean/’s water column. Nature 459, 266-269.• Simon, M., Grossart, H.-P., Schweitzer, B., and Ploug, H. (2002). Microbial ecology of organicaggregates in aquatic ecosystems. Aquatic Microbial Ecology 28, 175-211.• Skennerton, C.T. (2017) Methane-fueled syntrophy through extracellular electron transfer:uncoveringthe genomic traits conserved within diverse bacterial partners of anaerobic methanotrophic Archaea.mBio 8, e00530-17.• Skovgaard, A., Massana, R., Balague, V., and Saiz, E. (2005). Phylogenetic position of the copepod-infesting parasite Syndinium turbo (Dinoflagellata, Syndinea). Protist 156, 413-423.133• Smith, D.C., Simon, M., Alldredge, A.L., and Azam, F. (1992). Intense hydrolytic enzyme activityon marine aggregates and implications for rapid particle dissolution. Nature 359, 139-142.• Smith, M.W., Zeigler Allen, L., Allen, A.E., Herfort, L., and Simon, H.M. (2013). Contrasting ge-nomic properties of free-living and particle-attached microbial assemblages within a coastal ecosys-tem. Frontiers in Microbiology 4, 120.• Sogin, M.L., Morrison, H.G., Huber, J.A., Welch, D.M., Huse, S.M., Neal, P.R., Arrieta, J.M., andHerndl, G.J. (2006). Microbial diversity in the deep sea and the underexplored ’rare biosphere’.Proceedings of the National Academy of Sciences 103, 12115-12120.• Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., and Miller, H.L.(2007). Contribution of Working Group I to the Fourth Assessment Report of the IntergovernmentalPanel on Climate Change. Cambridge University Press.• Steele, J.A., Countway, P.D., Xia, L., Vigil, P.D., Beman, J.M., Kim, D.Y., Chow, C.-E.T., Sachdeva,R., Jones, A.C., Schwalbach, M.S., Rose, J.M., Hewson, I., Patel, A., Sun, F., Caron, D.A., andFuhrman, J.A. (2011). Marine bacterial, archaeal and protistan association networks reveal ecologicallinkages. The Isme Journal 5, 1414.• Stein, J.L., Marsh, T.L., Wu, K.Y., Shizuya, H., and Delong, E.F. (1996). Characterization of uncul-tivated prokaryotes: isolation and analysis of a 40-kilobase-pair genome fragment from a planktonicmarine archaeon. Journal of Bacteriology 178, 591-599.• Stein, L.Y., and Klotz, M.G. (2011). Nitrifying and denitrifying pathways of methanotrophic bacteria.Biochemical Society Transactions 39, 1826.• Steinberg, D.K., Carlson, C.A., Bates, N.R., Johnson, R.J., Michaels, A.F., and Knap, A.H. (2001).Overview of the US JGOFS Bermuda Atlantic Time-series Study (BATS): a decade-scale look atocean biology and biogeochemistry. Deep Sea Research Part II: Topical Studies in Oceanography 48,1405-1447.• Stevens, H., and Ulloa, O. (2008). Bacterial diversity in the oxygen minimum zone of the easterntropical South Pacific. Environmental Microbiology 10, 1244-1259.• Stewart, F.J., Dalsgaard, T., Young, C.R., Thamdrup, B., Revsbech, N.P., Ulloa, O., Canfield, D.E.,and Delong, E.F. (2012a). Experimental Incubations Elicit Profound Changes in Community Tran-scription in OMZ Bacterioplankton. PLOS ONE 7, e37118.• Stewart, F.J., Ottesen, E.A., and Delong, E.F. (2010). Development and quantitative analyses of auniversal rRNA-subtraction protocol for microbial metatranscriptomics. ISME J 4, 896-907.• Stewart, F.J., Ulloa, O., and Delong, E.F. (2012b). Microbial metatranscriptomics in a permanentmarine oxygen minimum zone. Environmental Microbiology 14, 23-40.134• Stilwell, C.P. (2007). Microbial community dynamics in Saanich Inlet: Insights into population struc-ture and methane cycling in a seasonally anoxic fjord. Thesis. Master of Science. The University ofBritish Columbia.• Stramma, L., Johnson, G.C., Sprintall, J., and Mohrholz, V. (2008). Expanding Oxygen-MinimumZones in the Tropical Oceans. Science 320, 655.• Stramma, L., Prince, E.D., Schmidtko, S., Luo, J., Hoolihan, J.P., Visbeck, M., Wallace, D.W.R.,Brandt, P., and Krtzinger, A. (2011). Expansion of oxygen minimum zones may reduce availablehabitat for tropical pelagic fishes. Nature Climate Change 2, 33.• Stramma, L., Schmidtko, S., Levin, L.A., and Johnson, G.C. (2010). Ocean oxygen minima expan-sions and their biological impacts. Deep Sea Research Part I: Oceanographic Research Papers 57,587-595.• Streams, E.M., Fisher, R.C., and Fiala-Mdioni, A. (1997). Methanotrophic symbiont location and fateof carbon incorporated from methane in a hydrocarbon seep mussel. Marine Biology 129, 465-476.• Strong, P.J., Xie, S., and Clarke, W.P. (2015). Methane as a Resource: Can the Methanotrophs AddValue. Environmental Science Technology 49, 4001-4018.• Strous, M., and Jetten, M.S.M. (2004). Anaerobic Oxidation of Methane and Ammonium. AnnualReview of Microbiology 58, 99-117.• Strous, M., Pelletier, E., Mangenot, S., Rattei, T., Lehner, A., Taylor, M.W., Horn, M., Daims, H.,Bartol-Mavel, D., Wincker, P., Barbe, V., Fonknechten, N., Vallenet, D., Segurens, B., Schenowitz-Truong, C., Mdigue, C., Collingro, A., Snel, B., Dutilh, B.E., Op Den Camp, H.J.M., Van Der Drift,C., Cirpus, I., Van De Pas-Schoonen, K.T., Harhangi, H.R., Van Niftrik, L., Schmid, M., Keltjens,J., Van De Vossenberg, J., Kartal, B., Meier, H., Frishman, D., Huynen, M.A., Mewes, H.-W., Weis-senbach, J., Jetten, M.S.M., Wagner, M., and Le Paslier, D. (2006). Deciphering the evolution andmetabolism of an anammox bacterium from a community genome. Nature 440, 790.• Sunagawa, S., Coelho, L.P., Chaffron, S., Kultima, J.R., Labadie, K., Salazar, G., Djahanschiri, B.,Zeller, G., Mende, D.R., Alberti, A., Cornejo-Castillo, F.M., Costea, P.I., Cruaud, C., D’ovidio, F.,Engelen, S., Ferrera, I., Gasol, J.M., Guidi, L., Hildebrand, F., Kokoszka, F., Lepoivre, C., Lima-Mendez, G., Poulain, J., Poulos, B.T., Royo-Llonch, M., Sarmento, H., Vieira-Silva, S., Dimier, C.,Picheral, M., Searson, S., Kandels-Lewis, S., Tara Oceans, C., Bowler, C., De Vargas, C., Gorsky,G., Grimsley, N., Hingamp, P., Iudicone, D., Jaillon, O., Not, F., Ogata, H., Pesant, S., Speich, S.,Stemmann, L., Sullivan, M.B., Weissenbach, J., Wincker, P., Karsenti, E., Raes, J., Acinas, S.G., andBork, P. (2015). Structure and function of the global ocean microbiome. Science 348.• Suter, E.A., Scranton, M.I., Chow, S., Stinton, D., Medina Faull, L., and Taylor, G.T. (2016). Niskinbottle sample collection aliases microbial community composition and biogeochemical interpretation.Limnology and Oceanography,.135• Suttle, C.A. (2005). Viruses in the sea. Nature 437, 356-361.• Suzuki, R., and Shimodaira, H. (2015). pvclust: Hierarchical Clustering with P-Values via MultiscaleBootstrap Resampling. R package version 2.0-0. http://CRAN.R-project.org/package=pvclust.• Swan, B.K., Martinez-Garcia, M., Preston, C.M., Sczyrba, A., Woyke, T., Lamy, D., Reinthaler,T., Poulton, N.J., Masland, E.D.P., Gomez, M.L., Sieracki, M.E., Delong, E.F., Herndl, G.J., andStepanauskas, R. (2011). Potential for Chemolithoautotrophy Among Ubiquitous Bacteria Lineagesin the Dark Ocean. Science 333, 1296-1300.• Takeda, K., and Tanaka, K. (1980). Ultrastructure of intracytoplasmic membranes of Methanomonasmargaritae cells grown under different conditions. Antonie van Leeuwenhoek 46, 15-25.• Tavormina, P.L., Ussler, W., Joye, S.B., Harrison, B.K., and Orphan, V.J. (2010). Distributions ofputative aerobic methanotrophs in diverse pelagic marine environments. ISME J 4, 700-710.• Tavormina, P.L., Ussler, W., Steele, J.A., Connon, S.A., Klotz, M.G., and Orphan, V.J. (2013). Abun-dance and distribution of diverse membrane-bound monooxygenase (Cu-MMO) genes within theCosta Rica oxygen minimum zone. Environmental Microbiology Reports 5, 414-423.• Taylor, C.D., and Doherty, K.W. (1990). Submersible Incubation Device (SID), autonomous instru-mentation for the in situ measurement of primary production and other microbial rate processes. DeepSea Research Part A. Oceanographic Research Papers 37, 343-358.• Taylor, C.D., Edgcomb, V.P., Doherty, K.W., Engstrom, I., Shanahan, T., Pachiadaki, M.G., Molyneaux,S.J., and Honjo, S. (2015). Fixation filter, device for the rapid in situ preservation of particulate sam-ples. Deep Sea Research Part I: Oceanographic Research Papers 96, 69-79.• Taylor Gt, K.D., Pace Ml (1986). Impact of bacteria and zooflagellates on the composition of sinkingparticles: An in situ experiment. . Mar. Ecol. Prog. Ser. 29, 141-155.• Thamdrup, B., Dalsgaard, T., and Revsbech, N.P. (2012). Widespread functional anoxia in the oxygenminimum zone of the Eastern South Pacific. Deep Sea Research Part I: Oceanographic ResearchPapers 65, 36-45.• Tol, R.S.J., Heintz, R.J., and Lammers, P.E.M. (2003). Methane Emission Reduction: An Applicationof FUND. Climatic Change 57, 71-98.• Torres-Beltra´n, M., Hawley, A., Capelle, D., Bhatia, M., Durno, E., Tortell, P., and Hallam, S. (2016).Methanotrophic community dynamics in a seasonally anoxic fjord: Saanich Inlet, British Columbia.Frontiers in Marine Science 3, 268.• Torres-Beltra´n, M., Hawley, A.K., Capelle, D., Zaikova, E., Walsh, D.A., Mueller, A., Scofield, M.,Payne, C., Pakhomova, L., Kheirandish, S., Finke, J., Bhatia, M., Shevchuk, O., Gies, E.A., Fairley,D., Michiels, C., Suttle, C.A., Whitney, F., Crowe, S.A., Tortell, P.D., and Hallam, S.J. (2017). A136compendium of geochemical information from the Saanich Inlet water column. Scientific Data 4,170159.• Tripp, H.J., Kitner, J.B., Schwalbach, M.S., Dacey, J.W.H., Wilhelm, L.J., and Giovannoni, S.J.(2008). SAR11 marine bacteria require exogenous reduced sulphur for growth. Nature 452, 741-744.• Trosvik, P., De Muinck, E.J., and Stenseth, N.C. (2015). Biotic interactions and temporal dynamicsof the human gastrointestinal microbiota. The ISME Journal 9, 533-541.• Tsementzi, D., Wu, J., Deutsch, S., Nath, S., Rodriguez-R, L.M., Burns, A.S., Ranjan, P., Sarode, N.,Malmstrom, R.R., Padilla, C.C., Stone, B.K., Bristow, L.A., Larsen, M., Glass, J.B., Thamdrup, B.,Woyke, T., Konstantinidis, K.T., and Stewart, F.J. (2016). SAR11 bacteria linked to ocean anoxia andnitrogen loss. Nature 536, 179-183.• Ulloa, O., Canfield, D.E., Delong, E.F., Letelier, R.M., and Stewart, F.J. (2012). Microbial oceanog-raphy of anoxic oxygen minimum zones. Proceedings of the National Academy of Sciences 109,15996-16003.• Ulloa, O., and Pantoja, S. (2009). The oxygen minimum zone of the eastern South Pacific. Deep SeaResearch Part II: Topical Studies in Oceanography 56, 987-991.• Van Den Bergh, M.R., Biesbroek, G., Rossen, J.W.A., De Steenhuijsen Piters, W.a.A., Bosch, A.a.T.M.,Van Gils, E.J.M., Wang, X., Boonacker, C.W.B., Veenhoven, R.H., Bruin, J.P., Bogaert, D., andSanders, E.a.M. (2012). Associations between Pathogens in the Upper Respiratory Tract of YoungChildren: Interplay between Viruses and Bacteria. PLOS ONE 7, e47711.• Vaquer-Sunyer, R., and Duarte, C.M. (2008). Thresholds of hypoxia for marine biodiversity. Pro-ceedings of the National Academy of Sciences 105, 15452.• Vzquez-Domnguez, E., Vaqu, D., and Gasol, J.M. (2007). Ocean warming enhances respiration andcarbon demand of coastal microbial plankton. Global Change Biology 13, 1327-1334.• Velo-Surez, L., Brosnahan, M.L., Anderson, D.M., and Mcgillicuddy Jr, D.J. (2013). A quantitativeassessment of the role of the parasite Amoebophrya in the termination of Alexandrium fundyenseblooms within a small coastal embayment. PloS one 8, e81150.• Venables, W.N., and Ripley, B.D. (2002). Modern Applied Statistics with S. Fourth Edition. Springer,New York. ISBN 0-387-95457-0.• Venter, J.C., Remington, K., Heidelberg, J.F., Halpern, A.L., Rusch, D., Eisen, J.A., Wu, D., Paulsen,I., Nelson, K.E., Nelson, W., Fouts, D.E., Levy, S., Knap, A.H., Lomas, M.W., Nealson, K., White, O.,Peterson, J., Hoffman, J., Parsons, R., Baden-Tillson, H., Pfannkoch, C., Rogers, Y.-H., and Smith,H.O. (2004). Environmental Genome Shotgun Sequencing of the Sargasso Sea. Science 304, 66.137• Waidner, L.A., and Kirchman, D.L. (2007). Aerobic Anoxygenic Phototrophic Bacteria Attached toParticles in Turbid Waters of the Delaware and Chesapeake Estuaries. Applied and EnvironmentalMicrobiology 73, 3936-3944.• Walsh, D.A., and Hallam, S.J. (2011). ”Bacterial Community Strucuture and Dynamics ina a Season-ally Anoxic Fjord: Saanich Inlet, British Columbia,” in Handbook of Molecular Microbial EcologyII: Metagenomics in Different Habitats, ed. F.J. De Brujin. Wiley-Blackwell), 253-267.• Walsh, D.A., Zaikova, E., Howes, C.G., Song, Y.C., Wright, J.J., Tringe, S.G., Tortell, P.D., andHallam, S.J. (2009). Metagenome of a versatile chemolithoautotroph from expanding oceanic deadzones. Science 326, 578-582.• Wang, J., Zhong, J., Chen, G., Li, M., Wu, F.X., and Pan, Y. (2015). ClusterViz: A Cytoscape APPfor Cluster Analysis of Biological Network. IEEE/ACM Transactions on Computational Biology andBioinformatics 12, 815-822.• Ward, B.B., Devol, A.H., Rich, J.J., Chang, B.X., Bulow, S.E., Naik, H., Pratihary, A., and Jayakumar,A. (2009). Denitrification as the dominant nitrogen loss process in the Arabian Sea. Nature 461, 78-81.• Ward, B.B., and Kilpatrick, K.A. (1990). Relationship between substrate concentration and oxidationof ammonium and methane in a stratified water column. Continental Shelf Research 10, 1193-1208.• Ward, B.B., and Kilpatrick, K.A. (1993). Methane oxidation associated with mid-depth methanemaxima in the Southern California Bight. Continental Shelf Research 13, 1111-1122.• Ward, B.B., Kilpatrick, K.A., Wopat, A.E., Minnich, E.C., and Lidstrom, M.E. (1989). Methaneoxidation in Saanich inlet during summer stratification. Continental Shelf Research 9, 65-75.• Wegener,G., V. Krukenberg, D. Riedel, H.E. Tegetmeyer, A. Boetius. (2015). Intercellular wiringenables electron transfer between methanotrophic archaea and bacteria Nature, 526,587-590.• Weiss, R.F., and Price, B.A. (1980). Nitrous oxide solubility in water and seawater. Marine Chemistry8, 347-359.• Whitmire, A.L., Letelier, R.M., Villagrn, V., and Ulloa, O. (2009). Autonomous observations of invivo fluorescence and particle backscattering in an oceanic oxygen minimum zone. Optics Express17, 21992-22004.• Whitney, F.A., Freeland, H.J., and Robert, M. (2007). Persistently declining oxygen levels in theinterior waters of the eastern subarctic Pacific. Progress in Oceanography 75, 179-199.• Wickham, H. (2009). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New Yorkhttp://ggplot2.org.138• Wickham, H., Francois, R., and (2015). dplyr: A Grammar of Data Manipulation. http://CRAN.R-project.org/package=dply R package version 0.4.3.• Widder, S., Besemer, K., Singer, G.A., Ceola, S., Bertuzzo, E., Quince, C., Sloan, W.T., Rinaldo, A.,and Battin, T.J. (2014). Fluvial network organization imprints on microbial co-occurrence networks.Proceedings of the National Academy of Sciences 111, 12799.• Wiesenburg, D.A., and Guinasso, N.L. (1979). Equilibrium solubilities of methane, carbon monoxide,and hydrogen in water and sea water. Journal of Chemical Engineering Data 24, 356-360.• Winkler, L.W. (1888). Die Bestimmung des im Wasser gelsten Sauerstoffes. Berichte der deutschenchemischen Gesellschaft 21, 2843-2854.• Wishner, K.F., Gowing, M.M., and Gelfman, C. (1998). Mesozooplankton biomass in the upper1000m in the Arabian Sea: overall seasonal and geographic patterns, and relationship to oxygengradients. Deep Sea Research Part II: Topical Studies in Oceanography 45, 2405-2432.• Wishner, K.F., Outram, D.M., Seibel, B.A., Daly, K.L., and Williams, R.L. (2013). Zooplankton inthe eastern tropical north Pacific: Boundary effects of oxygen minimum zone expansion. Deep SeaResearch Part I: Oceanographic Research Papers 79, 122-140.• Woebken, D., Fuchs, B.M., Kuypers, M.M.M., and Amann, R. (2007). Potential Interactions ofParticle-Associated Anammox Bacteria with Bacterial and Archaeal Partners in the Namibian Up-welling System. Applied and Environmental Microbiology 73, 4648-4657.• Worm, B., Sandow, M., Oschlies, A., Lotze, H.K., and Myers, R.A. (2005). Global Patterns ofPredator Diversity in the Open Oceans. Science 309, 1365.• Wright, J.J., Konwar, K.M., and Hallam, S.J. (2012). Microbial ecology of expanding oxygen mini-mum zones. Nat Rev Microbiol 10, 381-394.• Wright, J.J., Lee, S., Zaikova, E., Walsh, D.A., and Hallam, S.J. (2009). DNA Extraction from 0.22mu;M Sterivex Filters and Cesium Chloride Density Gradient Centrifugation. e1352.• Wrighton, K.C., Thomas, B.C., Sharon, I., Miller, C.S., Castelle, C.J., Verberkmoes, N.C., Wilkins,M.J., Hettich, R.L., Lipton, M.S., Williams, K.H., Long, P.E., and Banfield, J.F. (2012). Fermentation,Hydrogen, and Sulfur Metabolism in Multiple Uncultivated Bacterial Phyla. Science 337, 1661.• Yu, Z., and Chistoserdova, L. (2017). Communal Metabolism of Methane and the Rare Earth ElementSwitch. Journal of Bacteriology 199, e00328-00317.• Zaikova, E., Hawley, A., Walsh, D.A., and Hallam, S.J. (2009). Seawater Sampling and Collection.e1159.139• Zaikova, E., Walsh, D.A., Stilwell, C.P., Mohn, W.W., Tortell, P.D., and Hallam, S.J. (2010). Mi-crobial community dynamics in a seasonally anoxic fjord: Saanich Inlet, British Columbia. EnvironMicrobiol 12, 172-191.• Zeileis, A., Kleiber, C., and Jackman, S. (2008). Regression Models for Count Data in R. Journal ofStatistical Software 27.• Zhang, X., Hester, K.C., Ussler, W., Walz, P.M., Peltzer, E.T., and Brewer, P.G. (2011). In situRaman-based measurements of high dissolved methane concentrations in hydrate-rich ocean sedi-ments. Geophysical Research Letters 38,.140Appendix ASampling and processing methods impactmicrobial community structure andfunctionA.1 Supplementary resultsA.1.1 Saanich Inlet microbial community monitoring time-seriesPyrotag data has been used to describe Saanich Inlet microbial community composition over time andthroughout water column O2 gradients. Results derived from analyzing 16S rDNA pyrotag data from sam-ples taken over 2006 to 2011 revealed consistent microbial community partitioning into discrete clusters as-sociated with an oxygen gradient throughout the oxic, dysoxic/suboxic and anoxic water column conditions(Supplementary Fig 1). This pattern is commonly observed in stratified ecosystems where O2 deficiency isassociated with redox-driven niche partitioning among and between microorganisms(Alldredge and Cohen,1987; Shanks and Reeder, 1993; Wright et al., 2012). In addition, microbial community structure analysishave revealed among the abundant biosphere (relative abundance >1%)(Rapp and Giovannoni, 2003), op-erational taxonomic units (OTUs) affiliated with SAR11, SAR324, Nitrospina, SUP05, and Marine GroupA within the bacterial domain, and Thaumarchaeota within the archaeal domain dominated most composi-tional profiles (Zaikova et al., 2010; Walsh and Hallam, 2011). Because of their abundance and distinctivemetabolism these taxa have been previously described in marine O2- deficient environments as commu-nity key groups (Field et al., 1997; Fuhrman and Davis, 1997; Rapp and Giovannoni, 2003; Brown andDonachie, 2007; Tripp et al., 2008; Lam et al., 2009; Walsh et al., 2009; Zaikova et al., 2010; Walsh andHallam, 2011).A.1.2 Filtering conditions effect on microbial community diversityTo investigate the effect of filtering conditions, both filter type and volume, on community’s diversity, Ievaluated the alpha and Shannon indices for 29 SSU rDNA pyrotag samples from 165 and 185m. Diversity141Figure A.1: Non-metric multidimensional scaling plot for Saanich Inlet Time-Series SSU rDNA py-rotag samples (2006 -2011) based on Manhattan distance (1000 iterations). Oxygen gradient from oxic(>250 µM; red) to anoxic (<3 µM; purple) is shown on top and pyrotag samples are depicted as gray dots.Table A.1: Water column chemical properties during SCOR workshop on July14th, 2014. Oxygen(O2), Phosphate (PO4-3), Silicic acid (SiO2), Nitrate (NO3-), Nitrite (NO2-), Ammonium (NH4)+) and Hy-drogen Sulfide (H2S) concentrations (µM) for 150, 165 and 185 m depth intervals. **Cadmium columnclosed150 3.803 4.576 86.007 11.952 0.639165 2.355 5.116 90.789 2.311 0185 1.923 5.89 108.296 NaN** 0Depth (m) PO-34 (μM) SiO2 (μM) NO-2 (μM)O2 (μM) NO-3 (μM)values changed as function of filtering timing (in situ vs. on ship) and filter size. Overall 0.4 µm in situsamples showed higher alpha diversity values (>150) than those collected from bottles and filtered on ship,regardless the depth. However, Shannon diversity values were evenly distributed among samples, rangingfrom 4 to 5, for 0.4 and 2.7 µm pre-filtered samples regardless the depth and filtering timing (SupplementaryInformation Fig 2). Filtered water volume showed to have a different effect on pre-filtered 0.22 µm samplesbased on depth except. For instance, samples from 165m with <1.5 L water pre-filtered onto 0.4 µmshowed higher diversity values than those with greater volume filtered, and samples from 185m with <2Lwater filtered onto 2.7 µm showed lower diversity than those with greater volume filtered.1422L2L2L2L1.5L250ml500ml1.5L 2.5L250ml500ml20L1.5L2.5L250ml1L2.5L250ml500ml20L1.5L2.5L500ml1L2.5L250ml500ml20L20L345100 200 300α Shannon2.72.7 -> 0.22In Situ 0.40.40.4 -> 0.2165m185m500ml2.5L500ml250ml1L250ml500ml1.5L500ml1L2.5L250ml2L2L3.43.63.84.04.250 100 1502.5LA BFigure A.2: Shannon and alpha (α) diversity indexes for SSU rDNA pyrotags SCOR samples. Filteringconditions used are depicted as shown in color key for in situ from 0.4µm in situ (red), on ship 0.4µm pre-filter (green) and pre-filtered (0.4 µm = yellow; 2.7 µm = blue) and in laboratory 2.7 µm (black) samples.143HalobacterialesThermoplasmatalesAnaerolinealesOther BacteroidetesCandidate DivisionChlorobialesSAR406FlavobacterialesRhodobacteralesRhodospirillalesRickettsialesSAR11BurkholderialesMethylophilalesBdellovibrionalesDesulfarculalesDesulfobacteralesDesulfuromonadalesSAR324OceanospirillalesOther PhycisphaeraePlanctomycetalesSpirochaetaceaeVerrucomicrobiaEuryarchaeotaBacteroidetesSUP05ChlorobiMarine Group APlanctomycetesVerrucomicrobiaSpirochaetesαδγβProteobacteriaCiliophoraOther AlveolataOther MetazoaHolozoaArthropodaAlveolataOpisthokontaStramenopilesRhizariaEukaryotaSyndinialesArchaeaBacteriaTotal number of OTUs1552PPS MPPFigure A.3: Active microbial community composition based on rRNA: rDNA ratio (>1) for in situ 0.4µmfilter (PPS; red) and on-ship 0.22µm pre-filtered onto 0.4mm filter (MPP; yellow). The size of dots depictsthe total number of OTUs affiliated to specific taxa.144Appendix BProtistan parasites along water columnoxygen gradients: a network approach toassessing potential host-parasiteinteractions1B.1 Supplementary resultsB.1.1 Co-occurrence analysis and networkTo determine potential interactions between OTUs throughout the water column over the stratification period(May-August) in Saanich Inlet, a co-occurrence network was constructed using both Bray-Curtis and Spear-man correlation measures. All statistically significant correlations among OTUs resulting after permutationsand bootstrap score distributions were included in the downstream analyses. The final edges matrix was vi-sualized as a force directed network using Cytoscape 2.8.3 (Shannon et al., 2003). Each node representsan OTU and each edge a statistically significant positive correlation indicating co-occurrence. The resultingnetwork contains 325 nodes, connected by 6,273 edges. Average node degree (mean edges per node) (Proulxet al., 2005) was 3, the average path length (the expected distance between two connected nodes) (Latoraand Marchiori, 2009) was 2.73, and the network diameter (longest path between two nodes) (Cardoso et al.,2009) was equal to 8. The clustering coefficient (connectedness of a nodes neighbour) (Proulx et al., 2005)was 0.518, and connectance (proportion of all possible links realized) (Dunne et al., 2002) was 0.05. To de-scribe the potential interactions occurring among Syndiniales OTUs and protists, OTUs were taxonomicallyidentified in the Cytoscape network. Nodes corresponding to Syndiniales and their pairs were highlightedto summarize interactions. Based on nodes OTU number correlation data was extracted and exported for1A version of this appendix appears as supplementary information in Torres-Beltra´n, M., Sehein, T. et al. 2018. Protistanparasites along oxygen gradients in a seasonally anoxic fjord: a network approach to assessing potential host-parasite interactions.Deep Sea Res. II. doi.org/10.1016/j.dsr2.2017.12.026145co-occurrence description and visualization.146Figure B.1: Vertical distribution overtime for Syndiniales, Picobiliphyta, and Stramenopiles MAST1 and 3 OTUs. Top: Bars depict theabundance of Syndiniales OTUs found to correlate with Picobiliphyta, and Stramenopiles MAST1 and 3 throughout the water column from May2008 to April 2009. Bottom: Stacked bars depict the abundance (log transformed) for the taxa OTUs found to correlate with Syndiniales throughoutthe water column from May 2008 to April 2009.147Table B.1: Indicator OTUs for summer stratification suboxic-dysoxic water column conditions fromMay to August 2008. The OTU Id number, and significant (p =0.05, α < 0.01) indicator value are shown.Selected OTUs potentially showing a parasite-type interaction are highlighted.OTU Id numberIndicator value   p value  value Taxonomy15413 0.995 0.005 ** Eukaryota;Hacrobia;Haptophyta;Prymnesiophyceae;Prymnesiophyceae;Prymnesiophyceae;Phaeocystis;Phaeocystis+antarctica4362 0.979 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-4;Amoebophrya;Amoebophrya+sp13270 0.978 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-1;MAST-1A;MAST-1A;MAST-1A+sp12385 0.969 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1+sp8298 0.965 0.005 ** Eukaryota;Alveolata;Alveolata;Ellobiopsidae;Ellobiopsidae;Ellobiopsidae;Thalassomyces;Thalassomyces+fagei41856 0.962 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp20554 0.948 0.005 ** Eukaryota;Rhizaria;Cercozoa;Filosa-Thecofilosea;Cryomonadida;Protaspa-lineage;Protaspa-lineage;Protaspa-lineage+sp42840 0.945 0.005 ** Eukaryota;Hacrobia;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta+sp26650 0.94 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp2962 0.94 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-22;Dino-Group-II-Clade-22;Dino-Group-II-Clade-22+sp41598 0.935 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2+sp35162 0.935 0.005 ** Eukaryota;Hacrobia;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta+sp8780 0.931 0.005 ** Eukaryota;Hacrobia;Haptophyta;Prymnesiophyceae;Prymnesiales;Chrysochromulinaceae;Chrysochromulina;Chrysochromulina+rotalis35489 0.928 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-1;MAST-1B;MAST-1B;MAST-1B+sp7561 0.927 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp43589 0.924 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyta;Dinophyta;Dinophyta;Dinophyta;Dinophyta+sp26582 0.923 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp2274 0.916 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-4-6-7-8-9-10-11;MAST-7;MAST-7;MAST-7+sp26380 0.915 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II;Dino-Group-II;Dino-Group-II+sp27885 0.915 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-1;MAST-1C;MAST-1C;MAST-1C+sp18459 0.905 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp38500 0.901 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-3-12;MAST-3;MAST-3;MAST-3+sp10557 0.898 0.005 ** Eukaryota;Alveolata;Ciliophora;Spirotrichea;Choreotrichia;Choreotrichia;Choreotrichia;Choreotrichia+sp50356 0.898 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1+sp42773 0.897 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_H;Stephanoecidae_Group_H;Stephanoecidae_Group_H+sp10202 0.896 0.005 ** Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-2;Telonemia-Group-2;Telonemia-Group-2+sp17737 0.892 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-8;Dino-Group-II-Clade-8;Dino-Group-II-Clade-8+sp7847 0.881 0.005 ** Eukaryota;Archaeplastida;Chlorophyta;Mamiellophyceae;Mamiellales;Mamiellaceae;Micromonas;Micromonas+sp23562 0.878 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp9437 0.878 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;StramenopilesX;StramenopilesX+sp10460 0.87 0.005 ** Eukaryota;Hacrobia;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta+sp658 0.869 0.005 ** Eukaryota;Archaeplastida;Chlorophyta;Mamiellophyceae;Mamiellales;Bathycoccaceae;Bathycoccus;Bathycoccus+prasinos25851 0.864 0.005 ** Eukaryota;Alveolata;Apicomplexa;Apicomplexa;Gregarines;Cephaloidophoroidea;Cephaloidophoroidea;Cephaloidophoroidea+sp17547 0.862 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-V;Dino-Group-V;Dino-Group-V;Dino-Group-V+sp31392 0.86 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;StramenopilesX;StramenopilesX+sp22164 0.858 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp33553 0.851 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-20;Dino-Group-II-Clade-20;Dino-Group-II-Clade-20+sp10761 0.845 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4+sp49179 0.843 0.005 ** Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-1;Telonemia-Group-1;Telonemia-Group-1+sp30577 0.838 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-3-12;MAST-3;MAST-3;MAST-3+sp27111 0.832 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-16;marine;marine+metagenome18049 0.83 0.005 ** Eukaryota;Rhizaria;Cercozoa;Filosa-Chlorarachnea;Filosa-Chlorarachnea;LC104-lineage;LC104-lineage;LC104-lineage+sp25511 0.827 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp16229 0.823 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp48960 0.822 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp40167 0.822 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-29;Dino-Group-II-Clade-29;Dino-Group-II-Clade-29+sp4155 0.819 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp28241 0.815 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-1;MAST-1C;MAST-1C;MAST-1C+sp23310 0.805 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4+sp43345 0.801 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4+sp13554 0.791 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp32070 0.789 0.025 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-3;Dino-Group-II-Clade-3;Dino-Group-II-Clade-3+sp27690 0.788 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp44219 0.788 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp39085 0.788 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II;Dino-Group-II;Dino-Group-II+sp16672 0.784 0.005 ** Eukaryota;Alveolata;Ciliophora;Spirotrichea;Choreotrichia;Undellidae;Undella;Undella+marsupialis40010 0.784 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp50251 0.783 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5+sp7286 0.782 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2+sp49329 0.778 0.01 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Gymnodinium;Gymnodinium+sp36622 0.774 0.01 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp8102 0.77 0.005 ** Eukaryota;Alveolata;Ciliophora;Spirotrichea;Choreotrichia;Choreotrichia;Choreotrichia;Choreotrichia+sp19736 0.769 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp34794 0.768 0.02 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-12;Dino-Group-II-Clade-12;Dino-Group-II-Clade-12+sp7259 0.766 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp47404 0.764 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp13489 0.759 0.025 * Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dissodinium;Dissodinium+pseudolunula10911 0.753 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-6;Dino-Group-I-Clade-6;Dino-Group-I-Clade-6+sp50898 0.753 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6+sp50219 0.749 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5+sp8473 0.746 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-30;Dino-Group-II-Clade-30;Dino-Group-II-Clade-30+sp15504 0.744 0.005 ** Eukaryota;Excavata;Discoba;Euglenozoa;Diplonemea;Diplonemea;Diplonemea;Diplonemea+sp31377 0.735 0.005 ** Eukaryota;Hacrobia;Katablepharidophyta;Katablepharidaceae;Katablepharidales;Katablepharidales;Katablepharidales;Katablepharidales+sp38481 0.73 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp21594 0.726 0.005 ** Eukaryota;Archaeplastida;Chlorophyta;Trebouxiophyceae;Chlorellales;Chlorellales;Chlorellales;Chlorellales+sp26361 0.725 0.01 ** Eukaryota;Alveolata;Ciliophora;Colpodea;Colpodea-1;Colpodea-1;Colpodea-1;Colpodea-1+sp16728 0.721 0.02 * Eukaryota;Alveolata;Ciliophora;Oligohymenophorea;Scuticociliatia;Scuticociliatia-1;Scuticociliatia-1;Scuticociliatia-1+sp35326 0.719 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5+sp1611 0.717 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1;Dino-Group-I-Clade-1+sp14399 0.717 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-36;Dino-Group-II-Clade-36;Dino-Group-II-Clade-36+sp10016 0.717 0.005 ** Eukaryota;Excavata;Discoba;Euglenozoa;Diplonemea;Diplonemea;Diplonemea;Diplonemea+sp7059 0.714 0.015 * Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp45285 0.708 0.015 * Eukaryota;Alveolata;Ciliophora;Colpodea;Colpodea-1;Colpodea-1;Colpodea-1;Colpodea-1+sp24068 0.702 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-13;Dino-Group-II-Clade-13;Dino-Group-II-Clade-13+sp148Table B.1: Indicator OTUs for summer stratification suboxic-dysoxic water column conditions fromMay to August 2008. The OTU Id number, and significant (p =0.05, α < 0.01) indicator value are shown.Selected OTUs potentially showing a parasite-type interaction are highlighted. (Continuation)OTU Id numberIndicator value   p value  value Taxonomy26382 0.697 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp14519 0.697 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp50042 0.678 0.005 ** Eukaryota;Stramenopiles;Stramenopiles;Labyrinthulea;Labyrinthulales;Labyrinthulaceae;Labyrinthulaceae;Labyrinthulaceae+sp5605 0.678 0.005 ** Organelle;nucleomorph-Archaeplastida;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Teleaulax;Teleaulax+amphioxeia2745 0.675 0.02 * Eukaryota;Rhizaria;Cercozoa;Filosa-Chlorarachnea;Filosa-Chlorarachnea;NPK2-lineage;NPK2-lineage;NPK2-lineage+sp43154 0.668 0.015 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5;Dino-Group-I-Clade-5+sp11111 0.665 0.01 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp17402 0.662 0.015 * Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp26536 0.661 0.035 * Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-2;Telonemia-Group-2;Telonemia-Group-2+sp13400 0.658 0.02 * Eukaryota;Alveolata;Ciliophora;Oligohymenophorea;Scuticociliatia;Scuticociliatia-1;Scuticociliatia-1;Scuticociliatia-1+sp1893 0.654 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II;Dino-Group-II;Dino-Group-II+sp1973 0.65 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp27308 0.645 0.01 ** Eukaryota;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;Stramenopiles;StramenopilesX;StramenopilesX+sp29105 0.641 0.025 * Eukaryota;Opisthokonta;Fungi;Chytridiomycota;Chytridiomycotina;Chytridiomycetes;Spizellomycetales-and-Rhizophlyctidales;Spizellomycetales-and-Rhizophlyctidales+sp14478 0.637 0.02 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2;Dino-Group-I-Clade-2+sp503 0.637 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27+sp27014 0.637 0.015 * Eukaryota;Hacrobia;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta;Picobiliphyta+sp3056 0.637 0.005 ** Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp2939 0.637 0.015 * Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Radial-centric-basal-Coscinodiscophyceae;Corethron;Corethron+hystrix21902 0.63 0.015 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-20;Dino-Group-II-Clade-20;Dino-Group-II-Clade-20+sp4066 0.629 0.02 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp9442 0.629 0.03 * Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-2;Telonemia-Group-2;Telonemia-Group-2+sp19028 0.615 0.015 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-3;Amoebophrya;Amoebophrya+sp6127 0.615 0.025 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5+sp42296 0.615 0.035 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6+sp13243 0.615 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6+sp41630 0.615 0.035 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp6998 0.615 0.02 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-8;Dino-Group-II-Clade-8;Dino-Group-II-Clade-8+sp26680 0.615 0.035 * Eukaryota;Hacrobia;Cryptophyta;Cryptophyceae;Cryptophyceae;Cryptomonadales;Cryptomonadales;Cryptomonadales+sp18473 0.615 0.015 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp5983 0.615 0.025 * Eukaryota;Rhizaria;Cercozoa;Filosa-Imbricatea;Filosa-Imbricatea;Novel-clade-2;Novel-clade-2;Novel-clade-2+sp40970 0.615 0.02 * Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-4-6-7-8-9-10-11;MAST-6;MAST-6;MAST-6+sp29572 0.604 0.045 * Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Craspedida;Monosigidae_Group_M;Monosigidae_Group_M;Monosigidae_Group_M+sp47466 0.602 0.035 * Eukaryota;Alveolata;Dinophyta;Dinophyta;Dinophyta;Dinophyta;Dinophyta;Dinophyta+sp16021 0.593 0.03 * Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp9651 0.593 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp47939 0.593 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-14;Dino-Group-II-Clade-14;Dino-Group-II-Clade-14+sp43319 0.593 0.015 * Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp23147 0.593 0.02 * Eukaryota;Rhizaria;Cercozoa;Filosa-Thecofilosea;Cryomonadida;Protaspa-lineage;Protaspa-lineage;Protaspa-lineage+sp37245 0.593 0.04 * Eukaryota;Stramenopiles;Stramenopiles;Dictyochophyceae;Dictyochophyceae;Florenciellales;Florenciellales;Florenciellales+sp7889 0.593 0.035 * Eukaryota;Stramenopiles;Stramenopiles;Stramenopiles-Group-9;Stramenopiles-Group-9;Stramenopiles-Group-9;Stramenopiles-Group-9;Stramenopiles-Group-9+sp20006 0.593 0.02 * Organelle;nucleomorph-Archaeplastida;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Cryptophyta-nucleomorph;Falcomonas;Falcomonas+daucoides18789 0.58 0.035 * Eukaryota;Hacrobia;Haptophyta;Prymnesiophyceae;Prymnesiophyceae;Prymnesiophyceae;Prymnesiophyceae;Prymnesiophyceae+sp27024 0.569 0.025 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-26;Dino-Group-II-Clade-26;Dino-Group-II-Clade-26+sp50929 0.569 0.035 * Eukaryota;Opisthokonta;Choanoflagellida;Choanoflagellatea;Acanthoecida;Stephanoecidae_Group_D;Stephanoecidae_Group_D;Stephanoecidae_Group_D+sp149Table B.3: List of Spring indicator OTUs (May-June 2008). Table shows the OTU Id number, indicatorvalue (IV > 0.6, p-value=0.05 and αvalue <0.01), and BLAST-based taxonomic assignment.OTU Id numberIndicator value   p value value Taxonomy38633 1 0.005 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp41739 0.933 0.01 ** Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Polar-centric-Mediophyceae;Polar-centric-Mediophyceae+sp51135 0.894 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp47657 0.894 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp20157 0.894 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp17656 0.894 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp44881 0.894 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-14;Dino-Group-II-Clade-14;Dino-Group-II-Clade-14+sp47889 0.894 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp45315 0.876 0.01 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp0 0.871 0.005 ** Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp20422 0.859 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II;Dino-Group-II;Dino-Group-II+sp4066 0.851 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp9651 0.814 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp23960 0.804 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7;Dino-Group-II-Clade-7+sp50488 0.775 0.015 * Eukaryota;Opisthokonta;Metazoa;Cnidaria;Cnidaria;Hydrozoa;Hydrozoa;Hydrozoa+sp277 0.775 0.01 ** Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Chaetoceros;Chaetoceros+decipiens50514 0.775 0.02 * Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-2;Telonemia-Group-2;Telonemia-Group-2+sp26360 0.775 0.015 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp1973 0.773 0.025 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp17143 0.759 0.01 ** Eukaryota;Stramenopiles;Stramenopiles;Labyrinthulea;Labyrinthulales;Labyrinthulaceae;Labyrinthulaceae;Labyrinthulaceae+sp35395 0.757 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp19800 0.748 0.02 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4+sp25638 0.737 0.015 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp50019 0.737 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp44615 0.733 0.02 * Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Skeletonema;Skeletonema+sp33748 0.728 0.025 * Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Polar-centric-Coscinodiscophyceae;Polar-centric-Coscinodiscophyceae+sp27766 0.724 0.015 * Eukaryota;Alveolata;Ciliophora;Oligohymenophorea;Scuticociliatia;Scuticociliatia;Scuticociliatia;Scuticociliatia+sp33594 0.722 0.025 * Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Polar-centric-Coscinodiscophyceae;Polar-centric-Coscinodiscophyceae+sp16264 0.717 0.03 * Eukaryota;Hacrobia;Katablepharidophyta;Katablepharidaceae;Katablepharidales;Katablepharidales;Katablepharis;Katablepharis+japonica43348 0.697 0.025 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-3;Dino-Group-I-Clade-3;Dino-Group-I-Clade-3+sp39795 0.689 0.04 * Eukaryota;Stramenopiles;Stramenopiles;MAST;MAST-3-12;MAST-12;MAST-12;MAST-12+sp48230 0.679 0.035 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11;Dino-Group-II-Clade-10-and-11+sp3694 0.663 0.03 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-22;Dino-Group-II-Clade-22;Dino-Group-II-Clade-22+sp8473 0.661 0.045 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-30;Dino-Group-II-Clade-30;Dino-Group-II-Clade-30+sp28694 0.632 0.03 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp28607 0.632 0.03 * Eukaryota;Stramenopiles;Stramenopiles;Bacillariophyta;Bacillariophyta;Polar-centric-Mediophyceae;Polar-centric-Mediophyceae;Polar-centric-Mediophyceae+sp15495 0.632 0.03 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp22207 0.632 0.03 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp33168 0.632 0.03 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp25560 0.632 0.04 * Eukaryota;Stramenopiles;Stramenopiles;Stramenopiles-Group-7;Stramenopiles-Group-7;Stramenopiles-Group-7;Stramenopiles-Group-7;Stramenopiles-Group-7+spOTU Id numberIndicator value   p value value Taxonomy9442 0.869 0.005 ** Eukaryota;Hacrobia;Telonemia;Telonemia;Telonemia;Telonemia-Group-2;Telonemia-Group-2;Telonemia-Group-2+sp16838 0.866 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1;Dino-Group-II-Clade-1+sp21162 0.829 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27+sp35088 0.777 0.01 ** Eukaryota;Excavata;Discoba;Euglenozoa;Diplonemea;Diplonemea;Diplonemea;Diplonemea+sp18473 0.776 0.03 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp46903 0.744 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp37276 0.743 0.015 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5;Dino-Group-II-Clade-5+sp43833 0.727 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-I;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4;Dino-Group-I-Clade-4+sp24591 0.726 0.02 * Eukaryota;Excavata;Discoba;Euglenozoa;Kinetoplastida;Neobodonid;Neobodo;Neobodo+designis34542 0.702 0.05 * Eukaryota;Opisthokonta;Metazoa;Arthropoda;Crustacea;Maxillopoda;Maxillopoda;Maxillopoda+sp2667 0.667 0.03 * Eukaryota;Opisthokonta;Fungi;Ascomycota;Pezizomycotina;Eurotiomycetes;Aspergillus;Aspergillus+spOTU Id numberIndicator value   p value value Taxonomy36180 0.925 0.015 * Eukaryota;Opisthokonta;Metazoa;Cnidaria;Cnidaria;Hydrozoa;Hydrozoa;Hydrozoa+sp46532 0.92 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-III;Dino-Group-III;Dino-Group-III;Dino-Group-III+sp23562 0.919 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp4223 0.899 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-31;Dino-Group-II-Clade-31;Dino-Group-II-Clade-31+sp503 0.894 0.005 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27;Dino-Group-II-Clade-27+sp22164 0.854 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp36622 0.831 0.015 * Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp23147 0.796 0.005 ** Eukaryota;Rhizaria;Cercozoa;Filosa-Thecofilosea;Cryomonadida;Protaspa-lineage;Protaspa-lineage;Protaspa-lineage+sp42296 0.786 0.035 * Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-II;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6;Dino-Group-II-Clade-6+sp47404 0.784 0.005 ** Eukaryota;Alveolata;Dinophyta;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae;Dinophyceae+sp38851 0.784 0.01 ** Eukaryota;Alveolata;Dinophyta;Syndiniales;Dino-Group-III;Dino-Group-III;Dino-Group-III;Dino-Group-III+sp43154 0.782 0.01 ** Eukaryota;Al