UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Distribution and dynamics of biogenic sulfur in the northeast Subarctic Pacific : insights from new and… Herr, Alysia Elizabeth 2018

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

Item Metadata


24-ubc_2019_february_herr_alysia.pdf [ 4.42MB ]
JSON: 24-1.0375753.json
JSON-LD: 24-1.0375753-ld.json
RDF/XML (Pretty): 24-1.0375753-rdf.xml
RDF/JSON: 24-1.0375753-rdf.json
Turtle: 24-1.0375753-turtle.txt
N-Triples: 24-1.0375753-rdf-ntriples.txt
Original Record: 24-1.0375753-source.json
Full Text

Full Text

 Distribution and dynamics of biogenic sulfur in the northeast Subarctic Pacific:  insights from new and refined analytical techniques  by  Alysia Elizabeth Herr  B.S., Western Washington University, 2012  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Oceanography)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2018  © Alysia Elizabeth Herr, 2018       ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  Distribution and dynamics of biogenic sulfur in the northeast Subarctic Pacific: insights from new and refined analytical techniques  submitted by Alysia Herr in partial fulfillment of the requirements for the degree of Master of Science in Oceanography  Examining Committee: Philippe Tortell Supervisor  Sean Crowe Supervisory Committee Member  Allan Bertram Supervisory Committee Member  Additional Examiner   Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member   iii Abstract  The northeast subarctic Pacific (NESAP) is a globally important source of the climate-active gas dimethylsulfide (DMS), yet the processes driving DMS variability across this region are poorly understood.  This thesis aims to provide insight into the distribution and cycling of DMS and related sulfur compounds dimethylsulfoxide (DMSO) and dimethylsulfoniopropionate (DMSP) by examining new concentration data, together with biological cycling rates and related oceanographic variables. Chapter 2 examines the distribution of DMS at various spatial scales across contrasting oceanographic regimes of the NESAP.  We present a new data set of high spatial resolution DMS measurements across hydrographic frontal zones, together with key environmental variables and biological rate measurements.  We combine these new data with existing observations to produce a revised summertime DMS climatology for the NESAP.  Our results suggest the presence of two distinct DMS cycling regimes corresponding to microphytoplankton-dominated waters along the continental shelf, and nanoplankton-dominated transitional waters. In all areas, DMS consumption appeared to be an important control on concentration gradients, with higher DMS consumption rate constants associated with lower DMS concentrations.  Based on our compiled observations, we estimated that this region emits 0.30 Tg of sulfur to the atmosphere during the summer season. Chapter 3 presents results from two cruises examining DMSO distributions and cycling across the NESAP.  We measured DMSO concentrations and turnover rates across a range of hydrographic regions, and quantified rates of DMSO reduction, DMSP cleavage and DMS oxidation.  Our results show high concentrations and rapid turnover rates of DMSO across the NESAP.  Across our survey, DMSO reduction exceeded DMSPd cleavage at nearly all stations, while the rates of DMSO reduction exceeded those of DMS oxidation at four stations where both   iv these rates were measured.  These results suggest that DMSO reduction was an important net source of DMS.  A Lagrangian survey showed a significant decrease in DMSO concentrations during periods of peak irradiance, in conjunction with markers of oxidative stress.  Our findings highlight the significant contribution of DMSO to DMS production in the NESAP, and its potential physiological importance as an anti-oxidant in phytoplankton assemblages.       v Lay Summary  The trace gas dimethylsulfide (DMS) plays important roles in marine microbial communities, acts as an olfactory foraging cue for marine vertebrates, and has been linked to global climate through the production of atmospheric aerosols.  This thesis examines the environmental and biological factors that drive the distribution, loss and production of this compound in the subarctic Northeast Pacific (NESAP): a high-nutrient, low-chlorophyll region that exhibits particularly high DMS concentrations.  The research discussed here shows evidence for two distinct DMS cycling regimes across the NESAP region, and points towards the importance of dimethylsulfoxide (DMSO) in driving DMS concentrations.     vi Preface  All data analysis, research questions, and writing contained herein are my own, conducted with the guidance and contribution of Dr. Philippe Tortell.  Dimethylsulfide (DMS) concentration data described in Chapter 2 were collected using a membrane inlet mass spectrometer (MIMS) between the years 2004 – 2017 by several past and present members and affiliates of the Tortell laboratory, including Dr. Philippe Tortell, Nina Nemcek, Dr. Nina Schuback, Robert Izett, Dr. Anissa Merzouk, and Dr. Elizabeth Asher.  Additional ancillary thermos-salinograph (TSG) and conductivity, temperature, and depth (CTD) data were provided by the Institute of Ocean Sciences (IOS) and the scientists aboard the University Oceanographic Laboratory System (UNOLS) vessel R/V Oceanus, operated by Oregon State University.  Further ancillary data including station-based measurements of dissolved and total dimethylsulfoniopropionate (DMSPd/t), bacterial productivity, chlorophyll, and DMS/P loss rates were collected by members of the laboratory of Dr. Ronald Kiene of University of South Alabama.  Primary productivity was measured by Tortell lab post-doctoral fellow (PDF) Dr. Nina Schuback, and underway optical chl-a measurements were collected by Tortell lab PDF Dr. William Burt.  Samples for phytoplankton pigment analysis were collected by Dr. William Burt and Dr. Bethan Jones.  Chapter 2 has been submitted for publication to a scientific journal.   The data presented in Chapter 3 were collected using the OSSCAR system, which was initially developed by Dr. Elizabeth Asher, Dr. Philippe Tortell and Dr. John Dacey (Woods Hole Oceanographic Institute), with coding and hardware modifications made by Tereza Jarníková.  I made further extensive modifications to hardware and coding, and operated the instrument.  The stable isotope tracer method was developed by co-author Dr. John Dacey, with modifications by Dr. Elizabeth Asher and myself.  The transition to a new chemical-ionization   vii mass spectrometer for this work was heavily facilitated by Dr. Ross McCulloch.  Stable isotope tracer experiments were conducted by myself with the help of Dr. Ross McCulloch and Melissa Beaulac.  Measurements of photosynthetic parameters were made by Dr. Nina Schuback.  A version of Chapter 3 is currently in preparation for submission to a scientific journal.    viii Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ....................................................................................................................... viii List of Tables ............................................................................................................................... xii List of Figures ............................................................................................................................. xiii List of Abbreviations ...................................................................................................................xv Acknowledgements .................................................................................................................. xviii Dedication .....................................................................................................................................xx Chapter 1: Introduction ................................................................................................................1 1.1 Ecology and physiology of DMS, DMSP, and DMSO .................................................. 1 1.2 Distribution and cycling of DMS, DMSP and DMSO ................................................... 2 1.3 Thesis overview .............................................................................................................. 5 Chapter 2: Patterns and drivers of dimethylsulfide concentration in the northeast Subarctic Pacific across multiple spatial and temporal scales ...................................................6 2.1 Introduction ..................................................................................................................... 6 2.2 Methods......................................................................................................................... 10 2.2.1 Data overview ....................................................................................................... 10 2.2.2 New high-resolution data sets ............................................................................... 10 Underway ship-board measurements ................................................................ 10 Station-based measurements ............................................................................. 12 2.2.3 Compilation of published data .............................................................................. 14   ix MIMS data sets ................................................................................................. 14 PMEL data extraction ....................................................................................... 14 2.2.4 Ancillary measurements ........................................................................................ 15 2.2.5 Data binning and province assignment ................................................................. 16 2.2.6 Statistical analysis and empirical algorithms ........................................................ 17 2.3 Results ........................................................................................................................... 18 2.3.1 Oceanographic conditions in the NESAP during summer 2016-2017 ................. 19 2.3.2 DMS distributions (2016 & 2017) ........................................................................ 20 2.3.3 Detailed surveys of DMS across hydrographic frontal zones ............................... 21 Transect 1 .......................................................................................................... 21 Transect 2 .......................................................................................................... 23 Transect 3 .......................................................................................................... 25 2.3.4 Regional DMS distribution – comparisons of 2016 and 2017 observations with past studies ............................................................................................................................ 26 2.3.5 Correlations and algorithm testing ........................................................................ 28 2.4 Discussion ..................................................................................................................... 29 2.4.1 Contrasting cycling regimes within the NESAP ................................................... 29 2.4.2 The importance of phytoplankton assemblage composition ................................. 30 2.4.3 The effect of DMS/P consumption rates on DMS distribution ............................. 33 2.4.4 Insights from merged data set ............................................................................... 35 2.4.5 Biogeochemical provinces .................................................................................... 36 2.4.6 Correlation with environmental variables ............................................................. 37 2.4.7 Algorithm performance ......................................................................................... 38 2.5 Conclusion .................................................................................................................... 39   x 2.6 Acknowledgements ....................................................................................................... 40 Chapter 3: Spatial and temporal dynamics of dimethylsulfoxide in the northeast Subarctic Pacific and potential contributions to phytoplankton physiology and regional sulfur cycling........................................................................................................................................................56 3.1 Introduction ................................................................................................................... 56 3.2 Methods......................................................................................................................... 60 3.2.1 Study area .............................................................................................................. 60 3.2.2 Concentration measurements ................................................................................ 61 3.2.3 Rate measurements ............................................................................................... 63 3.2.4 Diel cycle measurements using Lagrangian drifters ............................................. 65 3.2.5 Ancillary measurements ........................................................................................ 65 3.3 Results ........................................................................................................................... 66 3.3.1 Oceanographic conditions ..................................................................................... 66 3.3.2 DMS/O concentrations .......................................................................................... 67 3.3.3 Rate experiments ................................................................................................... 68 3.3.4 Diel cycling of DMSO .......................................................................................... 70 3.4 Discussion ..................................................................................................................... 72 3.4.1 DMSO concentration in the NESAP ..................................................................... 72 3.4.2 Relationship between DMS and DMSO concentrations ....................................... 73 3.4.3 The role and impacts of DMSO reduction ............................................................ 75 3.4.4 Diel cycling ........................................................................................................... 78 3.5 Conclusion and outlook ................................................................................................ 81 3.6 Acknowledgements ....................................................................................................... 82 Chapter 4: Conclusion .................................................................................................................95   xi 4.1 Future work ................................................................................................................... 96 References .....................................................................................................................................98    xii List of Tables   Table 2.1 Summary of DMS data included in this study .............................................................. 41 Table 2.2 Summertime DMS data coverage across the NESAP region and within Longhurst provinces ....................................................................................................................................... 42 Table 2.3 Mean DMS concentrations, sea-air fluxes and total summertime DMS flux for the PMEL data set, and the updated data base used in this study. ...................................................... 43 Table 2.4 Pearson’s correlation coefficients between DMS concentrations and other oceanographic variables ................................................................................................................ 44 Table 2.5 Pearson's correlation coefficients and root mean square errors between observed DMS concentrations and empirical predictions derived from the SD02, VS07 and W07 algorithms ... 45 Table 3.1 Correlation table for station-based data ........................................................................ 83 Table 3.2 Simultaneous measurements of DMSO reduction and DMS oxidation and effect to the DMS pool for 4 stations measured during the Lagrangian drifter survey. ................................... 84 Table 3.3 Comparison of data from current and previous studies in which in situ DMSO reduction was measured. ............................................................................................................... 85    xiii List of Figures  Figure 2.1 Cruise tracks and discrete sampling stations for the July 2016 cruise and August 2017 cruise ............................................................................................................................................. 46 Figure 2.2 Spatial distribution of DMS and related oceanographic parameters during the July 2016 and August 2017 cruises ...................................................................................................... 47 Figure 2.3 DMS concentrations during the July 2016 and August 2017 cruises as measured by MIMS and gas chromatograph. ..................................................................................................... 48 Figure 2.4 Measurements of DMS, DMSPd, DMS/P consumption rate constants, and related oceanographic parameters along the T1 transect .......................................................................... 49 Figure 2.5 Line plot of sea surface height anomaly and observed DMS concentrations along the T1 transect ..................................................................................................................................... 50 Figure 2.6 Measurements of DMS, DMSPd, DMS/P consumption rate constants, and related oceanographic parameters along the T2 transect. ......................................................................... 51 Figure 2.7 Measurements of DMS, DMSPd, DMS/P consumption rate constants, and related oceanographic parameters along the T3 transect .......................................................................... 52 Figure 2.8 Spatial distribution of summertime DMS measurements from MIMS and the PMEL data sets ......................................................................................................................................... 53 Figure 2.9 Latitudinal distribution of data containing bins and average DMS concentration for PMEL and combined data sets ...................................................................................................... 54 Figure 2.10  Summertime DMS concentrations, DMS sea-air fluxes, and DMS:chl-a ratios  binned to 1˚ x 1˚ spatial resolution ............................................................................................... 55 Figure 3.1 Study area showing chl-a concentrations from AquaMODIS satellite and cruise tracks from May/June 2017 and August 2017 cruises ............................................................................. 86   xiv Figure 3.2 DMS concentrations, DMSO concentrations, and DMSO:DMS during the May/June 2017 cruise .................................................................................................................................... 87 Figure 3.3 DMS concentrations, DMSO concentrations, and DMSO:DMS during the August 2017 cruise .................................................................................................................................... 88 Figure 3.4 Relationship between DMSO and DMS concentrations across the full study area .... 89 Figure 3.5 DMSP cleavage and DMSO reduction rate constants, rates, and DMSPd/DMSO concentrations ............................................................................................................................... 90 Figure 3.6 Spatial distribution of measured DMSO reduction rates ............................................. 91 Figure 3.7 Relationship between bacterial productivity and rate of DMSO reduction ................ 92 Figure 3.8 Residual DMS/O concentrations and related variables measured during a Lagrangian drift survey conducted in Oregon coastal upwelling waters ......................................................... 93 Figure 3.9 Relationship between DMSO concentrations and non-photochemical quenching during a Lagrangian survey .......................................................................................................... 94    xv List of Abbreviations 13C: stable carbon isotope with atomic mass of 13 35S: radioisotope of sulfur with atomic mass of 35 ALSK: Alaska coastal downwelling province chl-a: chlorophyll-a pigment CCAL: coastal California current province CTD: oceanography instrument to measure vertical profiles of temperature, salinity and depth d: day D-3: stable isotope compound containing three deuterium atoms instead of hydrogen D-6: stable isotope compound containing six deuterium atoms instead of hydrogen DFO: Department of Fisheries and Oceans Canada DMOP: dimethyloxosulfoniopropionate DMS: dimethyl sulfide DMS/O/P: dimethyl sulfide DMSO: dimethyl sulfoxide DMSOd: dissolved dimethyl sulfoxide DMSOp: particulate dimethyl sulfoxide DMSOt: total dimethyl sulfoxide (dissolved and particulate) DMSOP: dimethylsulfoxonium propionate DMSP: dimethylsulfoniopropionate DMSPd: dissolved dimethylsulfoniopropionate DMSPp: particulate dimethylsulfoniopropionate DMSPt: total dimethylsulfoniopropionate (dissolved and particulate) FPD: flame photometric detector   xvi GC: gas chromatography HNLC: high nutrient, low chlorophyll HPLC: high performance liquid chromatography Io: sea surface irradiance level IOS: Institute of Ocean Sciences k: rate constant kd: extinction coefficient km: kilometer L: liter L11: Lana 2011 DMS climatology LED: light-emitting diode LPA11: August 2011 Line P cruise M: molar MIMS: membrane inlet mass spectrometry MLD: mixed layer depth N2: nitrogen gas N2O: nitrous oxide NaOH: sodium hydroxide NCP: net community production NO3: nitrate nM: nanomolar nmol: nanomoles NESAP: Northeast subarctic Pacific NPQ: non-photochemical quenching   xvii O16: July 2016 cruise O17: August 2017 cruise OSSCAR: organic sulfur sequential chemical analysis robot PAR: photosynthetically active radiation PFPD: pulse flame photometric detector PIC: particulate inorganic carbon PMEL: Pacific Marine Environmental Laboratory PSAE: Pacific subarctic gyres – east province PSII: photo system II PTR: proton-transfer reaction psu: practical salinity units RMSE: root mean square error ROS: reactive oxygen species S: sulfur SD02: algorithm from Simo and Dachs (2002) SSHA: sea surface height anomaly SRD: solar radiative dose SST: sea surface temperature UNOLS: University-National Oceanographic Laboratory System UV: ultra-violet radiation VIJ04: June 2004 Vancouver Island cruise VIJ10: June 2010 Vancouver Island cruise VS07: Algorithm from Vallina and Simo (2007) WCAC10: July/August 2010 West Coast Acidification Cruise    xviii Acknowledgements I would like to express my gratitude to the many people who guided, supported, and encouraged me throughout the course of this degree.  Thanks, first and foremost, to my advisor Philippe Tortell for allowing me the freedom to develop my own research, while providing the guidance needed to keep it on track.  Thanks also to John Dacey for a thoroughly enjoyable introduction to the world of custom-built scientific instruments and for inspiring conversations on science, art, cats, and everything in between.  To Ross McCullough, thank you for your guidance – scientific and otherwise – and for making the most exciting aspects of my research logistically possible.  Thanks to Dr. Allan Bertram and Dr. Sean Crowe for their revision of this thesis. I would also like to thank the many colleagues, collaborators and shipmates that I’ve had the pleasure of working with.  Thanks to Ron Kiene for your all-encompassing expertise, Susan Evans and the Micheal Murphy lab for a primer on enzyme isolation, Erin McParland, Tara Williams, Kait Esson, Ali Rellinger, Bethan Jones, Chris Paine, Maureen Soon, Jade Shiller, and the captains and crews of the CCGS John P Tully and R/V Oceanus.  Thanks to my lab mates and friends at UBC, whose contributions exceed that which I can list here.  To Georgia for slicing rocks with me (and being my roommate), Tereza for introducing me to OSSCAR and en plein air coffee, Nina for science-peripheral adventures, Robert for solving problems before anyone else noticed they existed, Anna and Ania for levity and mental support, Maite, Will, Sarah, Lindsay, Dave (both of them), Chen, Iselle, Kang, Cara, Yuanji, Jingxuan, Melissa, and Zarah.   Thanks to the coffeeshops of East Van for acting as my de facto office, and for everyone at Kickstand Community Bikes for being my community outside of academia.  Thanks also to   xix the captain and crew of the tall ship Lady Washington (circa 2004) for my introduction to the sea and to bioluminescence.   Finally, I would like to thank to my partner and my parents for their ongoing support.  Erik, thank you for having fun with me, even when I didn’t schedule time for it.  Thanks to my mom for being objectively wonderful at all times, and my dad for sharing his excitement about scientific research well before I could understand what he was talking about.  It’s a joy to now be able to share that excitement together.     xx Dedication To Dad and to Josa   1 Chapter 1: Introduction  Oceanic emissions of the biogenic trace gas dimethylsulfide (DMS) serve as the largest natural source of sulfur to the atmosphere (Bates et al, 1992, Gondwe et al., 2003) and contribute to the growth and formation of marine aerosol particles.  These particles can influence the Earth’s radiative balance though direct backscattering of radiation and the formation of cloud condensation nuclei, with potentially important impacts on global and regional climate.  Further, the flux of sulfur from the ocean to the atmosphere serves to link marine sulfur pools to atmospheric and terrestrial systems, making DMS and related compounds essential components of global sulfur cycle.  Due to the significance of this proposed climate link, DMS, and the related compounds dimethylsulfoniopropionate (DMSP) and dimethylsulfoxide (DMSO) are among the best studied organic molecules in the world’s ocean.  This extensive research has revealed numerous physiological and ecological roles for these compounds, as well as complex biogeochemical cycling processes (Bullock et al., 2017; Simó, 2004; Stefels et al., 2007).  Yet despite this body of research, the environmental and biological factors controlling DMS distribution remain poorly understood.  This thesis aims to address this gap in knowledge by investigating spatial patterns and potential drivers of DMS, DMSO, and DMSP distribution in the northeast Subarctic Pacific, a globally important DMS “hot spot”.  Below, I briefly discuss the ecological and physiological roles of these sulfur molecules, the various biotic and abiotic processes that contribute to their accumulation, and the rationale for studying sulfur dynamics in the northeast Subarctic Pacific.    1.1 Ecology and physiology of DMS, DMSP, and DMSO DMS, DMSP, and DMSO play a number of roles at both cellular and ecosystems levels.  While the majority of research has focused on DMSP and DMS, all three compounds have been   2 shown to play important roles in microbial food webs as carbon and sulfur substrates for bacteria (de Bont et al., 1981; Kiene et al., 2000; Vila‐Costa et al., 2006).  DMS and DMSP have further been recognized as olfactory foraging cues for plankton and marine vertebrates, thereby driving multi-species feeding aggregations, shaping marine food webs and influencing ocean carbon cycles (Johnson et al., 2016; Savoca, 2018; Seymour et al., 2010).  On a cellular level, the algal metabolite DMSP has been implicated in a number physiological roles.  Among a number of potential cellular functions,  it has been suggested that DMSP may act as an anti-herbivory defense mechanism (Fredrickson and Strom, 2009; Wolfe et al., 1997), a cryoprotectant (Karsten et al., 1996), an osmolyte (Dickson and Kirst, 1987), an overflow mechanism for excess fixed carbon (Stefels, 2000), and as protection against oxidative stress induced by salinity changes, nutrient limitation, and excess solar radiation (Sunda et al., 2002). This anti-oxidant function has received considerable attention, with related DMSP breakdown products DMS, DMSO, methanesulfinic acid, and acrylate implicated as antioxidants, serving to scavenge reactive oxygen species (ROS).  In support of this hypothesis, numerous studies have demonstrated increased intracellular DMSP and/or cellular DMSP lysis to DMS in conjunction with increased oxidative stress (Bucciarelli et al., 2013; Sunda et al., 2007; Wolfe et al., 2002).  More recently, DMS production through DMSO reduction has also been shown to be upregulated in phytoplankton under oxidative stress conditions (Spiese and Tatarkov, 2014).  1.2 Distribution and cycling of DMS, DMSP and DMSO The biogeochemical cycling of DMS and related compounds involves the interplay of many biotic and abiotic processes.  Phytoplankton-derived DMSP is generally considered the primary precursor of both DMS and DMSO.  Intracellular DMSP concentrations can vary by up   3 to six orders of magnitude across phytoplankton taxa (Caruana and Malin, 2014), and within a single species, cellular production can fluctuate by more than a factor of 10 depending on environmental conditions (Bucciarelli et al., 2013; Bucciarelli and Sunda, 2003).  Thus, total DMSP concentration in surface water is highly dependent on not only phytoplankton biomass, but also taxonomic composition and physiological status.  DMSP is released in small amounts by actively growing phytoplankton, and in greater quantities by cells undergoing senescence or exposed to zooplankton grazing or viral lysis (Dacey and Wakeham, 1986; Hill et al., 1998; Matrai and Keller, 1993).  Once released, dissolved DMSP (DMSPd) may be either used by bacteria to satisfy sulfur requirements via the demethylation pathway, or enzymatically cleaved to DMS and acrylate.  DMSP remaining in the particulate pool may also undergo cleavage by phytoplankton species containing the DMSP lyase enzyme.  The proportion of DMSP resulting in DMS production is generally low, resulting in DMS yields between ~1 – 40 % (Lizotte et al., 2017; Simó and Pedrós-Alió, 1999; Stefels et al., 2007).  Despite these low yields, DMSP cleavage is typically considered to be the primary source of DMS in the surface ocean.  The production and fate of DMSO is significantly less understood.  It is believed that the primary source of this compound is the bacterial and photochemical oxidation of DMS, although there is some evidence that DMSO may also be produced intracellularly through the oxidation of DMSP and DMS by ROS (Sunda et al. 2002, Spiese et al. 2009).  DMSO loss pathways include bacterial consumption (de Bont et al., 1981; Tyssebotn et al., 2017), biological oxidation to methanesulfinic acid (Sunda et al., 2002), and biological reduction to DMS by bacteria and phytoplankton (Spiese et al., 2009; Zinder and Brock, 1978).  Due to the two-way inter-conversion between DMS and DMSO, it is difficult to determine whether each compound acts primarily as a source or sink to the other.  However, recent studies have demonstrated extremely high DMSO concentration and reduction rates in Antarctic waters (Asher et al., 2011), raising   4 the question of how important this pathway is to DMS production.  DMSO is a quantitively important sulfur pool, with dissolved concentrations often exceeding that of DMSPd.  As such, DMS production through DMSO reduction may constitute an important source of DMS on ecologically-relevant timescales, particularly if DMSO is produced directly through DMSP oxidation and/or algal exudation (rather than from DMS oxidation).  DMS removal processes are somewhat better constrained than those driving DMS production.  These processes include biological DMS consumption, photo-oxidation to DMSO, and sea-air flux.  Sea-air flux generally accounts for only a small fraction of DMS removal  (<10%; Archer et al., 2002; Bates et al., 1994; Kloster, 2006)), and is dependent primarily on wind speed and temperature and salinity-dependent gas solubility.  By comparison, microbially-mediated processes and photo-oxidation serve as important sinks, with photo-oxidation accounting for >90% of DMS loss in some systems (Hatton, 2002), and microbial loss dominating in others (del Valle et al., 2009). While understanding of the DMS/P/O cycle is improving, significant variation in concentrations and rate processes across regions and seasons makes correlation of these compounds to more easily measured environmental variables elusive.  However, a global data base of over 47,000 data points administered by Pacific Marine Environmental Laboratory (PMEL; http://saga.pmel.noaa.gov/dms/) has facilitated compilation of regional observations to estimate global sea-air DMS fluxes and examine large-scale spatial and temporal variability.  Lana et al. (2011) utilized these data to construct a global climatology of surface ocean DMS concentrations and sea–air fluxes.  Through this work and others, researchers have identified several regions where high DMS concentrations (>10 nM) can persist for months.  These regions include the Southern Ocean and the northeast Subarctic Pacific (NESAP), which is the focus of research in this thesis.   5  1.3 Thesis overview The research chapters of this thesis examine the distribution patterns, drivers, and rates of processes governing DMS, DMSP and DMSO concentrations in the NESAP.  Chapter 2 examines the spatial distribution of DMS at various spatial scales across contrasting oceanographic regimes in the NESAP.  Specifically, I present a new data set of high spatial resolution DMS measurements across hydrographic frontal zones along the British Columbia continental shelf, together with key environmental variables, biological rate measurements, and phytoplankton taxonomy.  I combine these new data with existing regional observations to produce a revised summertime DMS climatology for the NESAP, yielding a broader context for our sub-mesoscale process studies.  In this chapter, I explore the importance of high and low-DMSP producing phytoplankton groups to DMS accumulation across contrasting regions.  Chapter 3 examines the relatively understudied molecule DMSO.  I map the distribution of DMSO concentrations across the NESAP, and examine diel cycling of this compound using a Lagrangian approach where water masses are tracked by following an in situ drifter.  I also provide direct comparisons of DMSO reduction and DMSP cleavage, and offer insight into their relative contribution to DMS production.  Together, these two chapters help address principal challenges in understanding DMS dynamics by providing important data regarding spatial and temporal distributions of DMS/P/O concentrations and turnover rates across the NESAP region.       6 Chapter 2: Patterns and drivers of dimethylsulfide concentration in the northeast Subarctic Pacific across multiple spatial and temporal scales  2.1  Introduction Spurred by a proposed role in climate regulation as a source of cloud-condensation nuclei and back-scattering aerosols, the biogenic trace gas dimethylsulfide (DMS) and related organic sulfur compounds dimethylsulfonioproprionate (DMSP) and dimethyl sulfoxide (DMSO) have been studied for more than four decades (Lovelock et al. 1972; Charlson et al. 1987).  This body of research has revealed complex sulfur biogeochemical cycling in the oceans, and important physiological and ecological roles for these molecules (Simó 2004; Stefels et al. 2007).  DMSP and DMS have been shown to play an essential function in marine microbial systems as sources of carbon and sulfur (Kiene et al. 2000; Reisch et al. 2011).  These molecules also act as olfactory foraging cues for numerous species of birds, fish, marine invertebrates and mammals (Seymour et al. 2010; Johnson et al. 2016), thereby driving interactions both within and beyond the marine microbial food web.  The ecological, chemical and climatological significance of DMS and related compounds has stimulated significant effort to understand the surface ocean distribution of these molecules and the underlying factors driving their variability.   The Pacific Marine Environmental Laboratory (PMEL) has compiled a database of over 47,000 discrete DMS measurements.  Lana et al. (2011) utilized these data to construct a global climatology of surface ocean DMS concentrations and sea–air fluxes, providing broad-scale understanding of oceanic distribution patterns.  The global mean DMS concentration is estimated to be approximately 2 nM, but the climatology reveals several regional ‘hot-spots’ of elevated DMS accumulation, including the Southern Ocean and northeast Subarctic Pacific (NESAP),   7 where surface ocean DMS concentrations 5–10-fold higher than the mean oceanic value are commonly observed.  Although large-scale global patterns derived from the climatology are likely robust, a fuller understanding of spatial and temporal patterns of regional DMS variability is constrained by the relatively poor spatial and temporal coverage of existing measurements.    The NESAP, defined here as the region bounded by 44.5˚ N and 61˚ N latitude and 180˚ W and 120˚ W, exhibits consistently high summertime DMS concentrations in both open ocean and coastal regions, with maxima of ~20 nM observed during the late summer season (Wong et al. 2005; Asher et al. 2011, 2017; Steiner et al. 2012).  This oceanic region is also characterized by strong spatial heterogeneity of environmental characteristics. High-productivity coastal upwelling regions transition to iron-limited high nutrient low chlorophyll (HNLC) waters offshore (Boyd and Harrison 1999; Boyd et al. 2004).  Seasonally varying surface currents, fresh water inputs, coastal upwelling and recurrent formation of westward-propagating mesoscale eddies result in semi-permanent and transient hydrographic frontal zones, impacting regional marine biodiversity and productivity (Crawford et al. 2005; Whitney et al. 2005; Ribalet et al. 2010).  This spatial heterogeneity makes it challenging to quantify DMS distributions from discrete ship-based sampling, and complicates region-wide generalizations of DMS dynamics. Recent work has highlighted differences in the distribution of DMS and related compounds across distinct domains of the NESAP, particularly in offshore and coastal regions (Wong et al. 2005; Asher et al. 2011, 2017; Steiner et al. 2012).  The HNLC offshore region was identified by L11 as an area of high DMS concentrations and sea–air fluxes. Results from in situ observations (Wong et al. 2005; Levasseur et al. 2006; Merzouk et al. 2006; Asher et al. 2011) and numerical models (Steiner et al. 2012) suggest that elevated DMS concentrations in these open ocean waters are driven by the presence of high DMS/P producing phytoplankton taxa, such as prymnesiophytes and dinoflagellates, and the effects of mixed layer stratification and Fe-  8 limitation, which may act to increase DMS/P production as a means to offset oxidative stress (Sunda et al. 2002).  A low particulate carbon to organic sulfur ratio in the HNLC regime further influences bacterial DMSP metabolism, resulting in increased DMS-yield from DMSP metabolism (Merzouk et al. 2006; Royer et al. 2010).  In the physically dynamic coastal waters of the NESAP, high DMS concentrations likely result, in part, from seasonal coastal upwelling, which drives high phytoplankton biomass accumulation. Indeed, recent work (Asher et al. 2017) has demonstrated an enhancement of DMS accumulation following upwelling events, consistent with previously observed high DMS/P concentrations in upwelling regions (Hatton et al. 1998; Zindler et al. 2012; Wu et al. 2017).  Increased DMS concentrations in the post-upwelling bloom phase may result from nitrogen limitation, increased grazing pressure (which releases DMSP into the dissolved pool), oxidative stress associated with shoaling mixed layers, and a phytoplankton community shift towards high DMSP-producing species (Nemcek et al. 2008; Franklin et al. 2009). New advances in sensor technology over the past decade have begun to significantly expand DMS data coverage in a number of ocean regions.  These fine scale measurements reveal novel features and highlight the apparent influence of oceanographic frontal zones in driving fine-scale DMS distribution patterns (Locarnini et al. 1998; Tortell 2005a; Nemcek et al. 2008; Jarníková et al. 2018).  In previous work (Asher et al. 2017), we have documented sharp transitions in DMS concentrations across salinity frontal zones in nearshore NESAP waters.  This earlier work did not include corresponding measurements of DMS/P turnover rates, limiting mechanistic interpretation of the observed spatial patterns.  To our knowledge, there has been no systematic evaluation of the processes driving fine-scale DMS variability across frontal zones.  Such a study requires high resolution concentration measurements together with assessments of biological productivity and DMS/P turnover rates.   9 In this article, we present a new data set of DMS/P concentrations across coastal and open ocean waters of the Subarctic Pacific, from the northern Gulf of Alaska to the Oregon coast.  Using a suite of measurements collected during two summer cruises (2016–2017), we document regional-scale features, and characterize sub-mesoscale DMS structure across hydrographic frontal zones in on-shelf and transition regions.  Using real-time ship-board measurements, we were able to select contrasting sites across frontal zones for more extensive sampling and analysis, allowing us to probe underlying rate processes in adjacent areas of contrasting DMS/P concentrations and surface water hydrography.  We combined our new data set with existing observations from our own group and from the existing PMEL database to produce a new summertime DMS climatology for the NESAP.  This updated climatology enables us to better constrain the summertime distribution of DMS in the NESAP, identifying persistent DMS ‘hot spots’, and exploring correlations between DMS concentration and other biotic and abiotic variables.  We use our compiled data set to evaluate various empirical algorithms predicting DMS concentrations and sea–air fluxes across the NESAP.  Our results yield new insights into the spatial patterns and potential drivers of summertime NESAP DMS distribution across various spatial scales in a globally important oceanic region.    10 2.2 Methods 2.2.1 Data overview In this study, we combined new data from two recent oceanographic expeditions with existing observations derived from several decades of compiled DMS measurements in the NESAP.  Ancillary data (e.g. environmental and biological variables) were obtained from a number of sources (both ship-based measurements, remote sensing and blended data products) to help interpret DMS distribution patterns. The various data sets are described below.   2.2.2 New high-resolution data sets Underway ship-board measurements Field sampling was conducted on board the University–National Oceanographic Laboratory System (UNOLS) vessel R/V Oceanus during July of 2016 and August of 2017 (O16, O17, respectively).  Our cruise tracks included offshore, coastal and transitional waters throughout the Gulf of Alaska (Fig. 2.1).  We define the coastal regime as those waters with bottom depths shallower than 2000 m, as per Asher et al. (2011).  We utilized real-time DMS measurements (see below) and NASA satellite ocean color imagery (AquaMODIS) to guide our cruise track, enabling us to identify areas with high concentrations of DMS and strong spatial gradients in surface water phytoplankton biomass and hydrography (sea surface temperature and salinity).  During O16 we also conducted detailed surveys of three hydrographic frontal zones that exhibited sharp DMS concentration gradients.  One of these surveys (T1; Fig. 2.1) was located in the coastal-open ocean transition near Dixon Entrance north of Haida Gwaii (formerly the Queen Charlotte Islands), while the other two transects were located along the British Columbia continental shelf (T2: Hecate Strait and T3: La Perouse Bank; Fig. 2.1).  After an initial survey to examine frontal structure, stations were selected for depth-resolved sampling to   11 cover the gradients present across the frontal zone.  The O17 cruise covered a similar area as O16.  Although we did not perform detailed transect surveys on this second cruise, we did sample waters near T1–T3. High resolution surface water DMS measurements were conducted using membrane inlet mass spectrometry (MIMS) following published methods (Tortell 2005b; Nemcek et al. 2008).  The MIMS system, utilizing the ship’s underway seawater flow through system (~5 m intake depth), allows for high-frequency sampling (2–3 times per minute), yielding a spatial resolution of ~150–200 m at normal ship speeds (8–10 kts).  During these cruises, DMS concentrations were also measured in discrete water samples collected at 5 m depth using a purge-and-trap system connected to a gas chromatograph equipped with a flame-photometric detector (FPD-GC) (Kiene and Service 1991).  These discrete measurements were used to assess the accuracy of MIMS-based measurements.  We found good agreement between methods, with a mean absolute error of 0.93 nM between the two instruments across the full range of measured concentrations (Fig. 2.2). High resolution DMS measurements were paired with ancillary underway data and rate measurements to examine potential drivers of spatial variation.   A ship board thermosalinograph was used to measure sea surface temperature (SST) and salinity at high spatial resolution (SBE 45 and SBE 38 for salinity and temperature, respectively).  Chlorophyll-a (chl-a) concentration was measured using a WET labs ACS absorbance/attenuation meter, based on the absorption line height at 676 nm (Bricaud et al. 1995; Roesler and Barnard 2013; Burt et al. 2018).  These chl-a concentrations were further used to derive an estimate of phytoplankton size and taxonomic distributions, based on the empirical algorithm of Hirata et al. (2011).  Zeng et al. (2018) have recently demonstrated the efficacy of this algorithm in our region as validated against HPLC-pigment-derived phytoplankton taxonomy data (detailed below), and we used the empirical   12 coefficients for this model tuned specifically for the NESAP.  MIMS was also used to determine the ratio of oxygen and argon concentrations relative to atmospheric saturation.  The resulting biological oxygen saturation term, ∆O2/Ar, can be used to calculate net community productivity (NCP) from the air–sea gas exchange of O2 (Kaiser et al. 2005).  We used the calculation approach of Reuer et al. (2007) to compute NCP from our ∆O2/Ar measurements.  We note that some of these estimates, particularly in regions of active upwelling, are likely biased by the entrainment of O2 under-saturated water into the mixed layer.  While this effect can be accounted for using N2O measurements (Izett et al. 2018), we do not have these data available for our cruises.  Our derived NCP estimates thus likely represent under-estimates, and we have removed all negative NCP values.  Nonetheless, the general spatial patterns we observed in NCP are likely to be robust. Station-based measurements We measured a suite of variables at selected sampling stations along the cruise track.  All water for ancillary measurements was taken from 5 m depth, collected using Niskin bottles.  A Seabird CTD probe (Seabird 911plus) was deployed at each station to obtain depth profiles of hydrographic features over the upper 200 m of the water column.  A density difference criterion of 0.05 kg m-3 was used to calculate mixed layer depths.  DMS loss and DMSP consumption rates were measured using the radio-labeled 35S methods outlined by Kiene and Linn (2000).  Briefly, 35S-labeled DMSPd or DMS were added to samples at non-perturbing concentrations (<1 % of ambient levels).  Samples were incubated in the dark at surface water temperatures for <1 h (35S-DMSP) or <7 h (35S-DMS).  The rate constant for DMSPd turnover was determined by measuring the disappearance of 35S-DMSP from the dissolved (< 0.2 µm) pool.  The rate constants for DMS loss were determined by   13 measuring the accumulation of dissolved, non-volatile 35S transformation products of the volatile 35S-DMS tracer.   Consumption rates (nmol L-1 d-1) were calculated by multiplying in situ DMS or DMSPd concentrations by the measured rate constant (kDMS or kDMSPd respectively).   Primary productivity was measured using 24 h 14C uptake incubations, following the method outlined by Schuback et al. (2015).  Incubation bottles were held in a deck-board incubator plumbed with continuously flowing seawater to achieve in situ temperature.  The light intensity was adjusted to ~ 30 % surface irradiance enriched in blue light using neutral density screening in combination with blue photographic film (LEE filters: #209 and CT blue maximum transmission at approx. 460 nm).  Light levels in the tank were measured with a ULM-500 light meter equipped with a 4π-sensor (Walz).  Bacterial production was measured using the tritiated leucine method (Smith and Azam 1992).  Station samples were also analyzed for total and dissolved DMSP (DMSPt and DMSPd) using the previously described NaOH cleavage and small-volume gravity drip filtration method (Dacey and Blough 1987; Kiene and Slezak 2006).   We obtained discrete estimates of phytoplankton assemblage composition using diagnostic pigment analysis (DPA) based on measurements of photosynthetic pigments using HPLC analysis.  For these measurements, 1 L sampled were collected on GF/F filters (nominal pore size ~ 0.7 µm), flash frozen in liquid nitrogen and stored frozen until analysis in the laboratory.  The DPA method was originally developed by Vidussi et al. (2001), and subsequently refined (Uitz et al. 2006; Hirata et al. 2008; Brewin et al. 2010) to more accurately capture phytoplankton type and size class.  The estimation formulas here are those of Hirata et al. (2011).  Percent contribution to phytoplankton assemblage was assessed for three size classes (micro-, nano-, and pico-) and several taxonomic groups, including diatoms, dinoflagellates and prymnesiophytes.      14 2.2.3 Compilation of published data To provide a broader regional spatial context for our observations, we combined discrete DMS measurements from the PMEL data archive with high spatial resolution DMS measurements made using MIMS since the early 2000s.  Table 1 provides dates and spatial domains of the cruises, along with relevant literature citations.  Note that some of the DMS data included in this compilation have not been previously published.  All of our compiled MIMS data have been made available on the PMEL database. MIMS data sets MIMS-based observations included in this study are derived from 11 cruises conducted between 2004 and 2017, primarily aboard the Canadian Coast Guard Ship John P. Tully as part of ongoing time-series monitoring programs conducted by the Department of Fisheries and Oceans Canada (DFO).  Only summertime data (defined here as June, July and August) falling within the NESAP region (44.5˚–61˚ N, 180˚–120˚ W) were included in this compilation.  All measurements were binned to a temporal sampling resolution of 1 minute. The cruises VIJ04, VIJ10, WCAC10, LPA11, O16 and O17 include paired NCP values obtained from MIMS measurements, using the ∆O2/Ar-based method described above.  All DMS data points are paired with shipboard sea surface salinity and SST. PMEL data extraction We accessed the PMEL data base (http://saga.pmel.noaa.gov/dms/) on 6 December, 2017 to extract observations from June, July and August in the NESAP region defined above.  Our selection criteria yielded 3236 data points between 1984 and 2003.  These observations were relatively evenly distributed between the three months, but were biased spatially, with a   15 preponderance of data derived from on-shelf waters off the coast of Alaska (see Fig. 2.8b).  As with MIMS data, each data point includes paired sea surface salinity and SST measurements.  2.2.4 Ancillary measurements Ancillary oceanographic data were used to contextualize DMS spatial distributions, examine potential correlations to environmental variables and evaluate the performance of several empirical algorithms predicting DMS concentrations.  In many cases, ancillary variables of interest (e.g. chl-a) were not reported in conjunction with DMS data, and we thus utilized a number of remote sensing data products, as described below.  Remotely-sensed parameters were linearly interpolated to the spatial resolution of ship-based DMS observations. AquaMODIS satellite data were used to obtain information on photosynthetically available radiation (PAR), chl-a (OCI algorithm), calcite (Gordon et al. 2001; Balch et al. 2005) and diffuse attenuation coefficients (Werdell and Bailey 2005).  For these data products, we extracted level 3 gridded data from http://oceancolor.gsfc.nasa.gov/cgi/l3 at 9 km resolution.  Monthly means for chl-a, calcite and kd were utilized to maximize spatial coverage, whereas 8 day average PAR data were used.  AquaMODIS chlorophyll data were also used to calculate sea surface nitrate (SSN) using a North Pacific-specific algorithm based on chl-a and SST (Goes et al. 2000).  AquaMODIS data are only available starting in July of 2002, whereas most of the PMEL data set in this region is from sampling prior to 2003.  For earlier observations, we used chl-a data from the SeaWiFS satellite. We obtained information on sea-surface height anomalies (SSHA) using gridded data sets (5 day, 0.17° x 0.17° resolution) obtained from ftp://podaac-ftp.jpl.nasa.gov/allData/merged_alt/L4/cdr_grid_interim.  This level 4 satellite product is derived from various sensors, and data are not available before 1992.  Mixed layer depths at a monthly,   16 1˚ resolution were obtained from the China Second Institute of Oceanography (CSIO) ftp://data.argo.org.cn/pub/ARGO/BOA_Argo/.  These data are based on gridded Argo float data produced using the Barnes method, and are available for the years 2004–present (Li et al. 2017).  Due to limitations in Argo operational depths, data are largely absent from waters shallower than 2000 m. We calculated sea–air DMS fluxes from DMS concentration data and surface wind-speeds using the gas transfer parameterization of Sweeney et al. (2007) and the Schmidt number formulation of Saltzman et al. (1993).  Wind speed data for flux calculations were obtained from the NCEP/NCAR reanalysis dataset (https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.pressure.html) at a 2.5° daily resolution.  Following previous studies, we assume negligible atmospheric DMS concentrations for our calculations, leading to a potential (though likely small) overestimate of the sea–air flux.  For purposes of comparison to fluxes, we calculated DMS column burden along transects by multiplying DMS concentration and average mixed layer depth.    2.2.5 Data binning and province assignment High resolution, underway measurements may introduce sampling biases due the large number of data points collected. For example, a ship holding station will increase spatial data density at a particular location, and the large number of observations can exert a disproportionate influence on derived mean values.  To address this, all measurements in the data set were assigned to 1° spatial bins, in which all observations for a given year were averaged.  The resulting yearly data grids were then averaged to create long-term gridded means.  This technique effectively assigns equal weight to each year of measurements in a given grid cell.  Both DMS and paired ancillary parameters were binned using this method.   17 Following the approach of L11, data grid cells were assigned to Longhurst’s Biogeochemical Provinces to examine patterns across different regimes within the greater NESAP (Longhurst 2007).  Three primary provinces fall within the domain of our study region: California Upwelling Coastal Province (CCAL), Alaska Downwelling Coastal Province (ALSK), and Pacific Subarctic Gyres Province – East (PSAE) (Fig. 2.8).  The CCAL province as defined by Longhurst extends south to 16.5° N.  Hereafter, all references to the CCAL refer to the portion of this province above 44.5° N latitude.  Province boundary designations were obtained from www.marineregions.com (accessed October 2017), and the MATLAB native inpolygon.m function was used to assign grid cells to individual provinces.  Any grid cell either inside or on the edge of boundaries was assigned to a particular province.  As such, some data cells (37 out of 216 total) are assigned to multiple provinces.  Average summer DMS concentrations and flux measurements were computed for each province.  For comparison to L11, we recalculated the average summertime DMS concentration and flux in the three study provinces using only the PMEL data utilized by Lana et al. (2011).  Data were first binned using the year-weighted method described above.       2.2.6 Statistical analysis and empirical algorithms We used our compiled data set to examine broad-scale relationships between DMS and other oceanographic variables.  For this analysis, data were log-transformed to overcome non-normal distributions, and the strength of pair-wise relationships was assessed by computing Pearson’s correlation coefficients.  Correlations were applied to 1˚ x 1˚ binned data both within individual provinces and across the entire NESAP. We also used several existing empirical algorithms to reconstruct DMS fields at a 1˚ x 1˚ resolution from various environmental predictor variables, comparing the accuracy of the   18 resultant products against our binned DMS observations.  The algorithms tested in this study include that of Simó and Dachs (2002), Vallina and Simó (2007), and Watanabe et al. (2007) (hereafter, SD02, VS07, and W07, respectively).   Both SD02 and VS07 used global data bases to develop their algorithms.  SD02 relates DMS to chl-a:MLD, with chl-a values > 15 µg L-1 removed prior to analysis. VS07 relates DMS concentration to solar radiative dose (SRD).  This term, as defined by the authors, is based on light extinction coefficients (kd), sea surface irradiance (I0), and mixed layer depth.  Due to the large areal extent of the study area, we used AquaMODIS derived PAR in lieu of the station-based I0 measurements used by the authors.  Similarly, strong variation in kd in coastal vs. open ocean waters is expected.  We thus modified the author’s approach and used satellite derived kd (based on a chlorophyll-dependent algorithm; Werdell and Bailey 2005) rather than a fixed coefficient.  W07 uses data specific to the North Pacific and relates DMS to SST, SSN and latitude.  Recognizing the utility of re-parameterizing proposed algorithms for specific areas, we tested all algorithms using both published linear coefficients, and coefficients derived specifically for the NESAP observations using a least-squares approach to determine best fit to our data set.   2.3 Results We begin by presenting an overview of our new DMS measurements and ancillary data from the 2016–2017 summer cruises, highlighting DMS distributions and the presence of distinct surface water properties across different parts of our transect.  We then provide a detailed description of DMS dynamics across several hydrographic frontal zones, discussing the potential role of various processes in driving these gradients. Finally, we construct an updated summertime climatology for this region, compiling our new measurements with existing DMS observations from across the NESAP to examine large-scale patterns in DMS distributions, and   19 correlations with other oceanographic variables.  The potential role of these variables in driving DMS distributions in the NESAP, and the need for additional process studies is addressed in the discussion.    2.3.1 Oceanographic conditions in the NESAP during summer 2016-2017 Our 2016 and 2017 cruises surveyed oceanographic regimes from offshore HNLC regions to productive coastal upwelling zones.  As indicated by AquaMODIS satellite imagery, chl-a concentrations exhibited strong gradients across the oceanic-coastal transition in both 2016 and 2017 (Fig. 2.1).  Coastal waters showed elevated chl-a, with maximum values of 50 µg L-1 and 18 µg L-1 in 2016 and 2017, respectively. In both years, highest chl-a values were observed in waters with shallow mixed-layer depths (<10 m) along the La Perouse Bank (Fig. 2.1).  In the off-shelf regions, chl-a concentrations appeared uniformly low in 2016, although significant cloud cover limited the availability of satellite imagery.  By comparison, we observed generally higher chl-a concentrations in offshore waters in 2017.  Most notably, our cruise track passed through a likely coccolithophore bloom in the northern Gulf of Alaska, where a large apparent calcite signal (~2 mmol PIC m-3) was detected in AquaMODIS imagery.  Patterns in NCP were generally similar to those of chl-a, with elevated production in coastal waters (Fig. 2.2c).  In both years, we observed NCP on La Perouse Bank exceeding 100 mmol O2 m-2 d-1 (Fig. 2.2c, inset).   Coastal regions exhibited generally fresher surface waters and shallower mixed layer depths, except for several regions of enhanced vertical mixing associated with upwelling.  This coastal upwelling signature was apparent in elevated salinity and decreased temperature of surface waters, and also through the presence of negative sea surface height anomalies (Fig. 2.1c,d). Small-scale regional heterogeneity in coastal regions was apparent in both years, with salinity and temperature exhibiting sharp gradients over the continental shelf, associated with   20 riverine input and complex mixing processes.  By comparison, oceanic surface waters showed less spatial heterogeneity, and were generally more saline, with deeper mixed layers (Fig. 2.2b).  The sea-surface height anomaly field indicated the presence of several Sitka and Haida eddies in both years (Fig. 2.1c,d), enhancing mesoscale variability through the transport of coastal water offshore.   Using the approach of Hirata et al. (2011) and Zeng et al. (2018), we derived high resolution estimates of phytoplankton assemblage composition from our underway chl-a measurements.  This approach revealed a predominance of phytoplankton in the micro- size class (>20 µm) in coastal waters (Fig. 2.2d), with an average of 50 % of chl-a attributable to microphytoplankton.  In contrast, off-shelf waters showed greater diversity in phytoplankton composition.  In these waters, microphytoplankton accounted for ~25 % of total chl-a, while the pico- and nano- size classes accounted for ~30 % and ~40 %, respectively (Fig. 2.2e,f).  2.3.2 DMS distributions (2016 & 2017) Across our study region, surface water DMS concentrations ranged from <1–24 nM in 2016 and <1–18 nM in 2017 (Fig. 2.2a, Fig. 2.3).  The average DMS concentrations showed very small (though statistically significant) difference between years (2.8 nM median, 1.0 – 5.7 interquartile range in 2016 and 2.6 nM median, 1.8 – 4.0 interquartile range in 2017; p<0.001 using Wilcoxon rank sum test).  We observed a number of localized DMS ‘hot spots’ in regions of elevated chl-a and NCP.  In both years, these localized high DMS regions were particularly evident in the vicinity of the highly productive La Perouse Bank (Fig. 2.2a, inset).  We also observed several areas where strong DMS gradients co-occurred with salinity fronts.  These areas include the T1–T3 transects survey in O16, detailed below.  Despite associations between DMS concentration and several variables in some localized regions, we only observed weak   21 correlations between DMS and other measured variables across the full cruise tracks.  During O16, DMS concentrations were most strongly correlated to NCP, with a Pearson’s coefficient of r=0.42 (p<0.001).  This relationship was substantially weaker in O17 (r=0.29, p<0.001).  2.3.3 Detailed surveys of DMS across hydrographic frontal zones During the O16 cruise, we sampled along three repeated transects to map DMS distributions near hydrographic frontal zones.  All three transects showed significant gradients in salinity, chl-a and DMS/P concentrations, as well as in the metabolic activity of phytoplankton and bacteria (Fig. 2.4, 2.6–2.7).  While DMS concentrations appeared to co-vary with salinity and chl-a across these frontal zones, the strength and direction of these relationships were not consistent across the three transects.  We discuss each transect in detail below. Transect 1 T1 was located west of Dixon Entrance (Fig. 2.1) in waters influenced by the Alaska Current and coastal water masses.  Offshore waters along this transect were more saline and colder than those on the shelf.  The area exhibited DMS concentrations up to 10 nM in off-shelf, saline waters (Fig. 2.4).  At the shelf break (approx. 134.4˚ W, indicated on Fig. 2.4 by dotted line), we measured a sharp drop in salinity and corresponding decrease in DMS concentrations, with concentrations remaining below ~3 nM over the most coastal parts of the transect.  Across the entire T1 transect, DMS concentrations displayed a striking fine-scale coherence to salinity (r=0.80, p<0.001; Fig. 2.4a, b).  A significant positive correlation was also observed with SSHA (r=0.51, p<0.001), indicating a potential influence of westward-propagating Haida eddies.  Fig. 2.5 shows a line plot of SSHA measurements from the approximate time of T1 sampling,   22 overlaid by DMS concentrations.  Despite differences in spatial resolution, the coherence between DMS concentrations and mesoscale oceanographic features can be seen in this figure.   The lower salinity coastal waters were characterized by elevated chl-a concentrations (Fig. 2.4c), resulting in a negative correlation between DMS concentrations and chlorophyll (r=-0.47, p<0.001).  Figure 2.4c shows the estimated percent abundance of prymnesiophytes and diatoms as derived from HPLC-based DPA-analysis.  Although HPLC samples are not available for all of the coastal waters we sampled, results obtained from the empirical algorithm of Hirata et al. (2011) suggest a shift in phytoplankton assemblage composition between smaller size classes in offshore waters to a microphytoplankton-dominated community in on-shelf waters.  DMS exhibited relatively weak, though statistically significant (p<0.001) positive correlations with the derived abundance of nano- and picophytoplankton size-classes (r=0.55 and r=0.38, respectively), and a negative correlation with the relative abundance of microphytoplankton (r=-0.53).  In support of this result, HPLC measurements revealed a strong positive relationship between DMS concentration and relative abundance of prymnesiophytes (r=0.88, p=0.002), and a negative correlation to diatom abundance (r=-0.70, p=0.036).  Moreover, we observed a strong positive correlation between DMS and DMSPt:chl-a (r=0.80, p=0.003) suggesting higher cellular DMSP concentrations in phytoplankton assemblages in the off-shelf regions of this transect.  Overall, results from this transect demonstrate a transition from high DMS concentrations in the lower productivity, nano-plankton dominated offshore waters, to low DMS concentrations in higher productivity, diatom-dominated nearshore region. Rate constants (d-1) for biological consumption of both DMS and DMSPd were higher in the on-shelf region (although the on-shelf/off-shelf difference was only statistically significant for kDMSPd), signifying faster removal of DMS/P from coastal surface waters with lower DMS concentrations.  For DMS and DMSPd respectively, loss constants averaged 1.15 ± 0.3 d-1 and   23 88.2 ± 13.9 d-1 onshore, as compared to 0.66 ± 0.045 d-1 and 39.6 ± 1.45 d-1 in offshore stations (Fig. 2.4e).  Net primary productivity and bacterial productivity were also higher, on average, in the low DMS coastal waters, but these differences were not statistically significant.  These results suggest that enhanced microbial activity and relatively higher DMS/P consumption rate constants played a role in maintaining lower concentrations of these compounds in nearshore waters.   We calculated the mixed layer DMS burden by multiplying concentration and average mixed layer depth (13 m).  Biological DMS loss integrated over the mixed layer averaged 22 µmol m-2 d-1, sufficient for daily removal of 47 % of the DMS burden.  By comparison, derived sea–air flux estimates across the transect exhibited a mean value of 13 µmol DMS m-2 d-1.  Using this value, we found that DMS loss to sea–air flux could on average remove 26 % of mixed layer DMS burden daily, given no new production.  Due to a relatively homogenous wind field over the area of our sampling transect, the sea–air fluxes were tightly correlated to DMS concentrations, such that the lower DMS concentrations in nearshore regions cannot be explained by greater rates of ventilation to the atmosphere. Transect 2 The second sampling transect, T2, was located in the coastal waters of Hecate Strait situated on the continental shelf (Fig. 2.1).  Sea surface temperatures along this transect showed small variation (standard deviation ~0.5˚ C), with the lowest temperature waters located mid-transect in areas of highest chl-a.  Mixed layer depths ranged from 10–15 m, and DMS concentrations ranged from < 0.5 nM to nearly 20 nM (Fig. 2.6).  In contrast to our observations for T1, DMS concentrations exhibited negative correlations to both salinity (r=-0.69, p<0.001; Fig. 2.6b) and SSHA (r=-0.81, p<0.001) in this area.  Despite the small absolute change in   24 salinity across this transect (< 0.4 psu), we observed strong coherence of DMS to salinity structure.  In contrast to T1, however, DMS concentrations were not significantly correlated to chl-a (Fig. 2.6c).  Despite lack of correlation to chl-a, DMS did exhibit significant, though weak, positive correlations with estimates of relative microphytoplankton abundance (r=0.22, p<0.001), and stronger negative correlations with the abundance of pico- and nano- size classes (T2: r=-0.47, r=-0.45; p<0.001; Fig. 2.6c).  In support of this observation, HPLC-pigment data revealed a strong positive relationship between DMS concentration and relative abundance of diatoms derived (r=0.88, p=0.001), and a negative correlation between DMS and prymnesiophyte abundance (r=-0.77, p=0.010).  These correlations thus suggest diatoms as an important source of DMS, a result that is opposite to that observed for T1. Unlike bulk chl-a concentrations, we found that primary productivity showed a strong positive correlation with DMS along T2 (r=0.90, p=0.037), although this result is based on only four data points.  Bacterial productivity was also significantly higher in the high DMS waters, although this variable was even more sparsely sampled along the transect, and we cannot infer any meaningful statistical association with DMS (Fig. 2.6f).  As with T1, both kDMSPd and kDMS were higher in the low-DMS portions of the transect.  Across the entire transect, consumption values ranged from 0.51 to 1.29 d-1 for kDMS and 28.8 to 49.5 d-1 for kDMSPd (Fig. 2.6e).  This result suggests microbial consumption as potential driver of DMS distributions, with higher DMS/P consumption rate constants in waters with lower DMS concentrations.   Integrated biological DMS loss was significantly higher than that of T1, with an average 78 µmol m-2 d-1 (equivalent to removal of 87 % of the DMS burden per day).  By comparison, DMS sea-air flux across the transect was low, with a mean value of 2.9 µmol m-2 d-1.  This flux was sufficient to remove only ~6 % of mixed layer DMS burden daily.  As with T1, the spatial pattern of sea-air flux was tightly correlated with DMS concentrations, and can thus not explain   25 the observed DMS distribution patterns (i.e. higher sea-air flux in regions of high DMS concentrations).  We thus conclude that DMS turn-over along this transect was dominated by biological processes. Transect 3 T3 was located in the highly productive coastal waters of La Perouse Bank, along the continental shelf of the west coast of Vancouver Island (Fig. 2.1).  Mixed layer depths ranged from 8–12 m, with the shallowest values found in fresher, salinity-stratified inshore waters influenced by riverine input.  Sea surface temperature was lower in these fresher waters, although it varied little over the transect (standard deviation < 1˚ C).  With respect to other measured variables, DMS behaved similarly to the coastal T2 transect (Fig. 2.7a).  We observed a negative correlation between DMS and salinity (r=-0.78, p<0.001; Fig. 2.7b).  We also found elevated chl-a in the low salinity waters, although there was only a weak positive correlation between chl-a and DMS (r=0.25, p<0.001) across the transect (Fig. 2.7c).  A significant negative correlation was found between DMS and SSHA (r=-0.75, p<0.001).   Microphytoplankton dominated the low-salinity, high-DMS waters of the transect, with a shift towards smaller cells observed in the more saline waters farther offshore (Fig. 2.7c).  Similar to T2, we found a significant positive correlation between DMS and microphytoplankton (r=0.90, p<0.001), and a negative correlation between DMS and phytoplankton of the nano- and pico- size class (r=-0.77, r=-0.75; p<0.001).  In support of this observation, HPLC-pigment data showed a strong positive relationship between DMS concentration and relative abundance of diatoms (r=0.86, p=0.003), and a negative correlation with prymnesiophyte abundance (r=-0.75, p=0.019).  A negative relationship was also observed between DMSPt:chl-a and DMS (r=-0.88, p=0.002) (Fig. 2.7d).  In contrast to T1, high DMS coincided with regions of lower cellular   26 DMSP concentrations among phytoplankton, consistent with the dominance of diatoms in the high DMS portions of this transect.  Along the T3 transect, DMS showed a positive association with primary productivity and bacterial productivity, though these relationships are based on very few sampling points.  It is noteworthy that the bacterial productivity measured along T3 was higher than anywhere else along the cruise track, with production more than 5-fold greater than the cruise-wide average.  Values of kDMS ranged from 0.8–2.7 d-1 across the transect.  As with T1 and T2, kDMS was higher in low-DMS regions of T3.  In contrast, kDMSPd values along T3 increased in parallel with DMS concentrations (higher rate constants in higher DMS waters).  DMSP loss constants ranged from 38.6 to 92.1 d-1 (Fig. 2.7e).  The highest DMSP loss constant translates into a derived turnover time of just 16 minutes.   Biological DMS loss integrated over the mixed layer was sufficiently high to remove >100 % of the DMS burden daily (~47 µmol m-2 d-1).  Sea–air fluxes were a minor loss term by comparison (4.9 µmol m-2 d-1), and were sufficient to remove only ~12 % of the mixed layer DMS burden.  Due to low removal rates and relative homogeneity of wind speed fields, sea–air flux cannot be invoked to explain the spatial distribution of DMS across this transect.    2.3.4 Regional DMS distribution – comparisons of 2016 and 2017 observations with past studies To explore potential regional-scale relationships between DMS concentrations and other environmental variables, we combined our new DMS data with measurements collected over the past three decades, including previously unpublished high-resolution MIMS data.   The addition of new measurements to the existing PMEL data set substantially increases spatial and temporal coverage in the NESAP.  When data were binned to 1˚ x 1˚ resolution, coverage was increased   27 by ~20 % in the CCAL and ALSK Longhurst provinces, and 14 % in the PSAE, with the overall addition of 90 data-containing grid cells (Table 2).  As shown in Fig. 2.8, our measurements primarily increase data coverage in waters below 57˚ N.  These regions were previously under-sampled in the PMEL data set utilized by L11, which was strongly biased to measurements near the coast of Alaska.  Figure 2.9a further illustrates the latitudinal shift in data coverage with the inclusion of additional MIMS data.  As shown in Fig. 2.9b, average derived DMS concentrations across latitudinal bands at the north and south extremes of our study area remain similar to those derived from the PMEL data set utilized by L11.  However, in the region between 50˚ N and 54˚ N, where there were few observations in the PMEL database, our compiled data show mean concentrations as much as 4.5 nM (~40 %) lower than those calculated using PMEL data alone.   Table 3 shows the change in province-wide average DMS concentration, sea–air fluxes, and total summertime DMS flux based on our updated analysis.  Relative to our revised estimates, DMS concentration and flux derived using only the PMEL data were lower in the CCAL and higher in both the PSAE and ALSK provinces.  The most pronounced difference was that of sea–air flux in the PSAE, where estimated sea–air fluxes decreased by 4.5 µmol m-2 d-1 (20 %).  Despite these regional differences, the total summer DMS flux across the NESAP differed by only 6.5 % between our compiled data set (0.30 Tg S) and the PMEL data set utilized by L11 (0.32 Tg S).   Our compiled data set provides greater confidence in DMS concentrations and sea–air fluxes across the NESAP, and enables us to better constrain spatial patterns.  Figure 2.10 shows binned average summertime DMS concentration across the region, as well as the derived sea–air DMS fluxes.  The highest concentrations are observed in ALSK, where coastal waters contain maximum DMS concentrations exceeding 20 nM.  A persistent region of elevated DMS concentrations is also evident in mid-PSAE oceanic region, with concentrations greater than 10   28 nM.  Sea–air DMS fluxes show a similar spatial distribution as DMS concentrations, with maximum values of >100 µmol m-2 d-1.  We also calculated and DMS:chl-a ratio for binned data (Fig. 2.10c), showing generally higher ratios in offshore NESAP waters.  2.3.5 Correlations and algorithm testing Using our new data compilation, we examined the relationship between DMS concentrations and a suite of oceanographic variables across the NESAP.  Table 4 shows both NESAP-wide and province-specific correlations derived from this analysis.  While many correlations are weak or not statistically significant, some patterns do emerge, particularly in the offshore waters of the PSAE domain.  No single variable explains a large portion of the DMS variation in this province, but statistically significant correlations exist between DMS and chl-a and calcite (r=0.45 and r=0.50, respectively).  We also found a negative relationship between DMS and SSHA (r=-0.47).  For the ALSK province, we found weak inverse correlations between DMS and SST (r=-0.32) and water depth (r=-0.34).  Significant positive correlations between DMS and derived surface NO3 concentrations, PAR, and chl-a are also observed (r=0.30, r=0.41, and r=0.34 respectively).  In contrast to other provinces, we observed a statistically significant correlation between DMS and NCP in the CCAL province (r=0.43).  The lack of other significant correlations in the province may, in part, reflect the lower number of data points obtained for this region.   Moving beyond simple pairwise correlations, multi-variate empirical algorithms provide an additional approach to assess the potential drivers of regional DMS dynamics.  We evaluated the ability of four previously published algorithms to reproduce patterns in the DMS observations.  In order to obtain the best possible results, we modified the original equations, using a least squares method to obtain the best-fit coefficients for our data set.  We evaluated the   29 algorithm outputs against observations using Pearson’s correlation coefficients and root mean square errors (RMSE).  As shown in Table 5, model performance was generally low in all cases, with correlation coefficients less than 0.62 and RMSE values ranging from 1.2 to 81.6 nM.  No single model performed best in all provinces.  For example, while the customized SD02 model was the best performing in the CCAL province (r=0.62, and RMSE=1.6 nM), this model performed poorly elsewhere.  The customized VS07 (with coefficients tuned to the NESAP data) showed the best overall performance across the entire NESAP region.  Yet, even this model showed only weak correlation between predicted and observed DMS values (r=0.31).  Interestingly, the original linear coefficients for this model yielded DMS concentrations that were inversely correlated to the measured values.  In no case did models using original linear coefficients outperform those using recalculated coefficients.    2.4 Discussion Our results provide new information on the fine-scale and regional patterns of DMS distributions across the NESAP.  Our ship-board observations document sub-mesoscale variability in DMS concentration across hydrographic frontal zones, with associated process measurements providing insight into potential driving factors.  By combining these new data with more than three decades of DMS measurements, we are able to improve data coverage for the NESAP to examine larger-scale spatial patterns and provide a more robust regional climatology to evaluate potential empirical predictive algorithms.  2.4.1 Contrasting cycling regimes within the NESAP A number of studies have documented differences in DMS dynamics across oceanographic regimes in the NESAP (Royer et al. 2010, Asher et al. 2011, 2017).  These   30 regional differences result from complex ecosystem and environmental interactions, and limit broad-scale prediction of DMS concentrations and sea–air fluxes (Galí et al. 2018).  Taxonomic composition of phytoplankton assemblages has been identified as a main driver of DMS distribution patterns.  For example, dinoflagellates and prymnesiophytes typically have elevated DMS production, associated with greater intracellular concentrations of DMSP and, in some cases, high activity of DMSP lyase (the enzyme that cleaves DMSP to DMS and acrylate) (Keller 1989; Steinke et al. 2002; Wolfe et al. 2002; Curson et al. 2018).  In contrast, bloom-forming diatom species have typically lower intracellular DMSP levels (Keller 1989).  However, nutrient limitation has been shown to significantly increase diatom DMS/P production (Bucciarelli and Sunda 2003; Sunda et al. 2007; Harada et al. 2009).  Thus, the accumulation of DMS in the water column depends on both the composition of phytoplankton assemblages and their physiological state.  Other factors, including zooplankton grazing and the metabolic demands of heterotrophic bacteria are also important (Levasseur et al. 1996, Kiene and Linn 2000, Merzouk et al. 2006, Asher et al. 2017).  Below, we discuss the potential factors driving high DMS concentrations along three frontal zones exhibiting sharp DMS concentration gradients.  Specifically, we contrast the nanophytoplankton dominated T1 transect with the diatom-dominated coastal T2 and T3 transects, examining the environmental and biological conditions that may have led to the different DMS accumulation patterns in these areas.    2.4.2 The importance of phytoplankton assemblage composition The T1 transect, located in the southern-most portion of the ALSK province, spanned 5˚ of longitude from deep (>3000 m) offshore waters, across the shelf break into nearshore waters over the continental shelf.  These oceanographic regimes were separated by strong hydrographic frontal features in the vicinity of the shelf break.  The negative correlation between DMS and   31 chl-a along this transect demonstrates that DMS accumulation did not directly scale with bulk phytoplankton biomass.  Rather, our results suggest that DMS concentrations were likely influenced by phytoplankton assemblage composition, with the highest DMS concentrations associated with the greatest relative proportion of prymnesiophytes (Fig. 2.4c) and the highest DMSPt:chl-a (Fig. 2.4d).  Similar relationships have been documented in numerous studies focusing on offshore waters of the NESAP.  In these areas, elevated DMS concentrations are often attributed to a preponderance of high-DMSP phytoplankton taxa.  Comparison of T2 and T3 with T1 show that the association of elevated DMS with prymnesiophyte dominance and high DMSPt:chl-a ratios did not hold across our entire survey region.  As has been observed in previous studies, we measured generally low DMSPt:chl-a ratios in the diatom-dominated coastal waters of T2 and T3 (Fig. 2.6d, 2.7d).  Yet, DMS concentrations measured in these waters were extremely high, at times exceeding 20 nM (Fig. 2.2a).  Unlike the T1 transect, DMS concentrations along T2 and T3 increased with decreasing DMSPt:chl-a ratios, and were strongly correlated with diatom abundance. One potential explanation for the difference between T1 and T2/T3 may relate to the different location of these sampling regions.  The T1 transect sits along the transition between offshore and inshore water, where different nutrient regimes control phytoplankton productivity.  Inshore waters over the continental shelf are typically limited by macronutrients, whereas offshore waters transition into iron-limitation (Boyd and Harrison 1999).  At the boundary between these regimes, mixing of water masses through horizontal advection can stimulate phytoplankton productivity (Lam and Letters 2008).   Ribalet et al. (2010) observed an active community of nanoplankton in the transitional waters, and attributed this to the stimulation of (often high-DMSP yielding) oceanic phytoplankton by water mass mixing, at the boundary of macro- and micro-nutrient rich waters.  Eddy formation and associated transport and mixing may   32 also play a role in stimulating productivity in transition waters (Johnson et al. 2005; Whitney et al. 2005).  In support of this, we observed the highest DMS associated with positive SSHA along T1 (Fig. 2.5). In contrast to the transition waters, nearshore waters over the continental shelf are typically dominated by low DMSP-producing diatoms.  Elevated DMS in these diatom-rich waters may reflect a combination of high absolute biomass and an upregulation of DMSP production observed under nutrient stress (Bucciarelli and Sunda 2003; Sunda et al. 2007; Hockin et al. 2012; Bucciarelli et al. 2013).  A meta-analysis by McParland and Levine (in revision) reported an average 12-fold upregulation of intracellular DMSP production under nutrient-stress conditions among phytoplankton typically considered low-producers.  By comparison, high DMSP producers only showed an average 1.4-fold upregulation.   In coastal waters, seasonal upwelling may drive high phytoplankton biomass accumulation and increased DMS production in the late-bloom phase, when stratified surface layers are exposed to higher mean light intensities (due to more shallow mixing) and become depleted of nutrients (Zindler et al. 2012).  These environmental conditions would act to increase cellular oxidative stress, thus promoting the production of DMS/P as part of a response mechanism (Sunda et al. 2002).  The results of Asher et al. (2017) demonstrating high DMS concentrations in post-upwelling waters support this idea.  Measurements of SSHA in coastal regions can provide a signature for recent upwelling.  The combined effect of wind-induced seasonal water transport offshore and the presence of high density (cold and saline) upwelled water acts to depress sea surface height relative to annual means (Smith 1974; Tabata et al. 1986; Strub and James 1995; Saraceno et al. 2008; Venegas et al. 2008).  Negative relationships between DMS concentrations and SSHA were observed in both the T2 and T3 transects, suggesting an association between DMS and upwelling events.     33 Additional ecosystem processes may influence DMS accumulation in surface waters.  In particular, zooplankton grazing and viral infection may increase DMS concentrations, due to the release of cellular DMSP in phytoplankton during sloppy feeding and cellular lysis.  Both of these factors are density-dependent, and thus likely to become more significant with higher phytoplankton cell densities in the late bloom phase.  Unfortunately, we do not have measurements to address these processes directly, but the elevated DMSPd concentrations along T3 (~7 nM) may reflect viral and zooplankton mediated loss of particulate DMSP into the dissolved pool. Taken together, our results support previous studies showing the importance of DMSP-rich species in driving high DMS concentration in offshore waters of the NESAP.  In coastal waters, it appears that diatom-dominated phytoplankton assemblages can also support elevated DMS accumulation, particularly under high biomass conditions during the late bloom phase.     2.4.3 The effect of DMS/P consumption rates on DMS distribution DMS consumption rates constants across our study area can be translated to biological DMS turnover times ranging from 9 h to 2.5 d (average of 25 h).  By comparison, turnover times calculated from sea–air flux removal rates averaged 6.1 d across this area.  While these measures do not encompass all loss processes, biological consumption and sea–air flux alone are sufficient to quickly erase a DMS accumulation signature in the mixed layer.  Thus, DMS concentrations measured here appear to be reflective of short-term production and consumption processes.  Across our study area, biological DMS removal rate constants (d-1) were inversely related to DMS concentrations (r=-0.55, p=0.03), with lower kDMS in waters with elevated DMS.   The relationship may reflect a time-lag of bacterial response to increased DMS concentrations.  Results from previous studies have shown that bacterial consumption of DMS increases some   34 time after a rapid rise in DMS, resulting in consumption rate constants that are relatively low when DMS concentrations are initially high.  As consumption rate constants increase, DMS concentrations decrease (Zubkov et al. 2004; del Valle et al. 2009).  These results, along with the observed positive correlation between DMS and bacterial activity (r=0.53, p=0.03), suggest that microbial consumption is an important control on DMS accumulation, regardless of phytoplankton community assemblage.  In contrast to DMS rate constants (d-1), water column DMS consumption rates (nM d-1) showed a positive correlation with DMS concentrations (r=0.65, p=0.01).  This result is not unexpected, as consumption rates are the product of rate constants and in situ concentrations.  Regardless, the positive correlation between DMS loss rates and concentrations suggests that microbial consumption may not be sufficient to offset new DMS production.  Previous studies have examined the impact of DMS loss and production in driving distributions,  demonstrating, in some cases, a correlation between DMS concentrations and microbial consumption and production rates (Wolfe and Kiene 1993; Zubkov et al. 2002 Merzouk et al. 2006, Vila-Costa et al. 2008).  Our observed relationship between DMS and kDMS and bacterial activity may reflect the preponderance of on-shelf stations measured for DMS consumption in our survey (10 out of 16 stations), and significantly higher rates of bacterial metabolism in onshore waters (5.51 ± 2.0 vs 0.73 ± 0.20 nM Leucine uptake d-1 for on- and off-shelf stations, respectively).  In contrast to biological loss, turnover time due to sea-air flux showed no correlation to DMS concentrations. Recent studies in the NESAP have estimated that photo-oxidation may account for 20–70 % of gross DMS removal in the NESAP (Asher et al. 2017), and it possible that this process is particularly important in offshore waters.  Bouillon and Miller (2004) found that quantum yields of DMS oxidation in the NESAP correlated well to nitrate concentrations, suggesting that this pathway is particularly relevant in the HNLC region where excess macronutrients persist   35 throughout the summer.  Thus, the role of biological DMS consumption on influencing total DMS concentrations may be more important in the generally low nitrate coastal waters.   No correlation was found between DMSPd loss rates or loss rate constants and DMS concentrations in our study.  This lack of correlation may be due, in part, to variation in DMSPd loss pathways.  The DMS yield of DMSP metabolism can vary significantly depending on metabolic needs of bacteria present, and relative abundance of phytoplankton with DMSP lyase activity (Yoch 2002).  In the NESAP, a low carbon to organic sulfur ratio in the HNLC regime results in increased DMS-yield from DMSP metabolism, whereas onshore DMS-yield is relatively lower (Merzouk et al. 2006; Royer et al. 2010).  Further, variation in DMS loss processes may obscure a relationship between DMSPd cleavage and DMS concentrations, as high loss terms may disproportionately impact net DMS production. Regardless, correlation between DMSPd and DMS concentrations (r=0.51, p=0.01) indicate that this compound is likely an important DMS precursor, as widely noted in literature.  We are currently investigating, in greater detail, the patterns of DMS and DMSPd consumption from our O16 and O17 cruises (Kiene et al., in prep).  2.4.4 Insights from merged data set Our merged data set, binned to 1˚ x 1˚ spatial resolution, builds on the L11 climatology to further constrain summertime DMS distributions across the NESAP region.  Despite an overall ~20 % increase in data-containing bins, and the inclusion of data from seven additional years, we see only small changes in the derived climatological DMS concentrations and sea–air fluxes when compared to the PMEL data set used by L11 (Table 3).  Our new observations thus support the validity of the L11 climatology in the NESAP region, providing further confidence in the apparent distribution patterns, and a greater spatial footprint for the climatological field.  A   36 significant result of our analysis is the presence of high DMS:chl-a in offshore waters (Fig. 2.10c), building on previous reports of higher DMSP:chl-a concentrations in offshore NESAP waters, and highlighting the importance of prymnesiophytes and other DMSP-rich phytoplankton taxa in driving DMS accumulation in offshore NESAP waters.  2.4.5 Biogeochemical provinces When examining results from our 1˚ binned data set, a separation of the NESAP into on- and off-shelf regimes does not capture the biogeochemical complexities of the region.  Ecological provinces, as defined by Longhurst (2007), define regions with coherent seasonal trends in physical processes, which give rise to similar biological and chemical characteristics.  The use of Longhurst’s biogeochemical provinces may thus be a more suitable (though still imperfect) approach to examine large-scale and long-term differences in DMS cycling across the region. Work by Reygondeau et al. (2013) has demonstrated the potential for shifts in province boundaries over time, including decrease of coastal province size during El Nino periods, and a general shore-ward shift of ALSK boundaries during summer months.  A model-based classification of marine ecosystems in the North Pacific by Gregr and Bodtker (2007) divides our study region into six domains that show little similarity to Longhurst provinces.  It is difficult to say which of these classification schemes is most appropriate for examination of DMS dynamics.  However, we follow the approach of L11 and others in using Longhurst’s provinces to examine regional cycling differences.  While we acknowledge these provinces provide only a crude measure of biogeochemical regimes, they remain a best-approximation without delving into more complicated time-resolved ecological province models (Reygondeau et al. 2013).  Further, the use of these provinces allowed us to directly compare our results with those of L11.  Going   37 forward, it may be useful to examine DMS dynamics in sub-regions defined with a number of different metrics.    2.4.6 Correlation with environmental variables Our analysis shows that no single variable can explain an appreciable amount of variability in DMS concentrations across the NESAP.  This result is consistent with previous global and regional studies (Kettle et al. 1999; Vézina 2004; Lana et al. 2011).  Nonetheless, an examination of the differing relationships between DMS concentrations and other environmental variables provides insight into potential underlying factors driving DMS distribution (Table 4).  For example, although we found a moderately strong significant positive correlation between DMS and chl-a in the largely HNLC PSAE province, no relationship is observed between these variables in the CCAL province.  As noted above and confirmed in several previous studies, the phytoplankton community structure in the offshore PSAE region consists largely of small, DMSP-rich species (Booth et al. 1993; Suzuki et al. 2002; Royer et al. 2010; Steiner et al. 2012), and large blooms are infrequent.  Indeed, the average binned chl-a concentration in this province is < 1 µg L-1.  As such, modest increases chl-a likely reflects a stimulation of this high DMSP-producing community.  The positive correlation with calcite (an indicator of high-DMSP producing coccolithophores) supports this idea.    The relationship between chl-a and DMS is more complicated in the CCAL.  High productivity in coastal upwelling zones results in a strong onshore/offshore trend in average chl-a concentrations. Yet, no such trend is observed in DMS concentrations.  This may be due, in part, to the sensitivity of DMS concentrations to phytoplankton assemblage composition and bloom dynamics.  High phytoplankton biomass alone will not result in elevated DMS.  Rather,   38 elevated DMS concentrations may occur as a response to conditions of late-bloom nutrient stress, as discussed above.   Factors driving observed DMS distribution patterns in the ALSK province are more difficult to surmise.  DMS is notably high in the cold, productive waters adjacent to the Alaskan Peninsula.  This is affirmed by a weak negative correlation between DMS and SST, and the positive correlation between DMS and chl-a.  Given that this portion of the province is known to experience localized summer upwelling, it is possible that high DMS in the regions simply reflects elevated productivity.    2.4.7 Algorithm performance Our results suggest that no single empirical algorithm is likely to perform well in predicting DMS distributions across the subarctic Pacific.  Perhaps the most interesting result was the negative correlation between measured and modeled results using the VS07 algorithm.  This algorithm predicts DMS concentrations from solar radiative dose, a term that measures depth-integrated exposure to sunlight, with the assumption that increases in SRD are accompanied by increases in DMS due to UV-induced oxidative stress (Vallina and Simó 2007).   However, it is also possible that elevated SRD can also lead to a decrease in surface water DMS concentrations through DMS photo-oxidation.  As observed in previous studies, photo-oxidation in the NESAP may account for up to 70 % of gross DMS removal, and rates are positively correlated with nitrate concentrations (Bouillon and Miller 2004; Asher et al. 2017).  Thus, in the high-nitrate NESAP, SRD may serve primarily to remove DMS from surface waters, rather than stimulate DMS production.  This result, as well as the regionally-varying performance of other tested algorithms, underlines the need for tuning and selecting models best suited for a given area and season.  In the absence of a single robust predictive algorithm for DMS concentrations in the   39 NESAP, it will be important to improve mechanistic understanding of DMS/P dynamics, merging field-based process studies with prognostic numerical models (e.g. Aumont et al. 2002, Clainche et al. 2004, Steiner et al. 2012, Wang et al. 2015, Hayashida et al. 2016).  2.5 Conclusion This study examines distribution and cycling of DMS across the NESAP at various spatial scales.  Our results affirm the importance of high-DMSP producers (i.e. prymnesiophytes) to DMS accumulation in offshore waters, while also demonstrating the importance of diatom-dominated assemblages in driving DMS distribution in coastal upwelling regions.  We further highlight the importance of metabolic rate processes and provide evidence for the importance of DMS consumption on concentration gradients at a fine-scale.  On the short spatial scales covered by our transect surveys, we observed strong correlations between DMS concentrations and other variables (i.e. SSHA, salinity).  Over regional scales, however, we only observed weak statistical relationships.  All predictive algorithms we tested showed poor performance in predicting DMS concentrations across the NESAP region, although performance was improved through the use of regionally-tuned coefficients.  Our compiled data set further support the importance of the NESAP as a global DMS ‘hot spot’, with patterns of DMS concentrations and sea–air fluxes similar to those observed in Lana et al.’s 2011 climatology.  Given the significance of the NESAP in global oceanic DMS emissions, future studies should seek to improve mechanistic understanding of the factors driving DMS accumulation in this region, with the aim of predicting climate-dependent changes over the coming decades.    40 2.6 Acknowledgements We wish to thank many individuals involved in data collection and logistical aspects of the cruises presented here, including scientists from the Institute of Ocean Sciences, the captain and crew of the R/V Oceanus and the CCGS John P. Tully, and members Tortell, Kiene, Levine and Hatton laboratory groups.  We also thank T. Ahlvin for GIS support.  Support for this work was provided from the US National Science Foundation (Grant #1436344), and from the Natural Sciences and Engineering Research Council of Canada.   41  Table 2.1 Summary of DMS data included in this study.  With the exception of the PMEL data, all measurements are derived from membrane inlet mass spectrometry (MIMS).   Cruise  abbreviation Vessel affiliation;  cruise name and number Sampling dates Areal extent Provinces included No. data  points References VIJ04 DFO; Central Coast BioChemical Study; 2004-24 12–19 Aug   2004 48˚ N - 52˚ N 131 ˚ W - 123˚ W ALSK, CCAL 1913 (Nemcek et al. 2008)  LPJ07 DFO; Line P; 2007-13 1–16 Jun  2007 47˚ N - 55˚ N 146 ˚ W - 123˚ W ALSK, CCAL, PSAE 21478 (Asher et al. 2011)  LPA07 DFO; Line P; 2007-15  16–30 Jun  2007 48˚ N - 54˚ N 146 ˚ W - 123˚ W ALSK, CCAL, PSAE 16418 (Asher et al. 2011)  LPJ08 DFO; Line P; 2008-26 1–15 Jun 2008 48˚N - 52˚ N 146 ˚ W - 123˚ W ALSK, CCAL, PSAE 15304 (Asher et al. 2011)  LPA08 DFO; Line P; 2008-27 14–30 Aug 2008 48˚N - 52˚ N 146 ˚ W - 123˚ W ALSK, CCAL, PSAE 20881 (Asher et al. 2011)  VIJ10 DFO; La Perouse; 2010-12 1–4 Jun 2010 48˚N - 52˚ N 130 ˚ W - 123˚ W ALSK, CCAL 4551 (Tortell et al. 2012) WCAC10 DFO; Ocean Acidification; 2010-36 22 Jul–15 Aug  2010 47˚N - 57˚ N 138 ˚ W - 123˚ W ALSK, CCAL, PSAE 25167 (Asher et al. 2017) LPA11  DFO; Line P; 2011-27 19–28 Aug 2011 48˚ N - 51˚ N 146 ˚ W - 126˚ W CCAL, PSAE 10802 (Asher et al. 2017) LPA14 DFO; Line P and Strait of Georgia; 2014-19 29–31 Aug  2014 50˚ N - 51˚ N 145˚ W - 134˚ W PSAE 2560  (Asher et al. 2015)  O16 UNOLS; Resolving DMS 1: OC-1607A 12–27 Jul  2016 45˚ N - 56˚ N 143 ˚ W - 124˚ W ALSK, CCAL, PSAE 18712 Previously unpublished O17 UNOLS; Resolving DMS II: OC-1708A 12–27 Aug  2017 47˚ N - 57˚ N 146 ˚ W - 126˚ W ALSK, CCAL, PSAE 10015 Previously unpublished PMEL various various,  1984–2004 45˚ N - 61˚ N 167 ˚ W - 124˚ W ALSK, CCAL, PSAE 3236 Various   42  Table 2.2 Summertime DMS data coverage across the NESAP region and within Longhurst provinces.   Values indicate the number of data-containing 1˚ x 1˚ spatial bins out of the total number of bins within the given area.  The left column represents the coverage for the PMEL data set (as utilized by L11) and the right column represents the updated data set containing both PMEL measurements and MIM-based DMS concentration measurements.  Province Name PMEL This Study CCAL 30/75 (40.0 %) 45/75 (60.0 %) ALSK 61/119 (51.3 %) 83/119 (69.8 %) PSAE 5/430 (12.8 %) 114/430 (26.5 %) Total 126/1140 (11.1 %) 216/1140 (19.0 %)      43 Table 2.3 Mean DMS concentrations, sea-air fluxes and total summertime DMS flux for the PMEL data set utilized by L11, and the updated data base used in this study.       PMEL   This Study Province Name DMS (nM) DMS Flux (µmol m-2 d-1) Total summer DMS flux (Tg S) DMS (nM) DMS Flux (µmol m-2 d-1) Total summer DMS flux (Tg S) CCAL 4.0 ± 0.5 4.4 ± 0.95 0.01 4.6 ± 0.4 6.3 ± 0.7 0.02 ALSK 8.9 ± 1.1 16.4 ± 4.0 0.06 7.5 ± 0.9 14.4 ± 3.0 0.05 PSAE 8.9 ± 0.7 21.0 ± 4.0 0.38 6.5 ± 0.4 16.5 ± 2.2 0.30 Total 7.2 ± 0.5 12.7 ± 2.0 0.32 6.2 ± 0.3 12.2 ± 1.4 0.30      44 Table 2.4 Pearson’s correlation coefficients between DMS concentrations and other oceanographic variables binned to 1˚ spatial resolution.  DMS data were derived from our combined PMEL and MIMS data set, variables derived from in-situ and satellite-based data.  N represents the number of data pairs available for each correlation calculation.  * indicates significance of p<0.05.  Variable Whole region CCAL ALSK PSAE Salinity r = -0.04 N = 223 r = 0.24 N = 31 r = -0.04 N = 83 r = 0.07 N = 102 SST r = -0.01 N = 248 r = -0.17 N = 44 r = -0.32* N = 83 r = 0.18 N = 114 Chlorophyll-a r = 0.17* N = 207 r = -0.11 N = 31 r = 0.34* N = 79 r = 0.45* N = 99 Calcite r = 0.12 N = 205 r = -0.08 N = 30 r = -0.01 N = 83 r = 0.50* N = 99 PAR r = 0.04 N = 212 r = -0.28 N = 32 r = 0.41* N =52 r = 0.19 N = 91 Depth r = -0.05 N = 201 r = 0.20 N = 45 r = -0.34* N = 12 r = -0.02 N = 96 MLD r = -0.14 N = 98 r = 0.14 N = 21 r = -0.06 N = 11 r = -0.18 N = 70 SSN r = 0.01 N = 207 r = 0.14 N = 31 r = 0.30* N = 79 r = -0.18 N = 99 SSHA r = -0.20* N = 207 r = -0.34 N = 30 r = -0.05 N = 80 r = -0.47* N = 102 NCP r = 0.22* N = 91 r = 0.43* N = 26 r = 0.05 N = 25 r = 0.29 N = 37 Wind r = 0.17* N = 249 r = -0.06 N = 45 r = 0.08 N = 83 r = 0.29* N = 114      45 Table 2.5 Pearson’s correlation coefficients and root mean square errors (nmol L-1) between observed DMS concentrations and empirical predictions derived from the SD02, VS07 and W07 algorithms, using both published coefficients (original) and coefficients derived specifically for our NESAP observations using a least-squares approach (custom).  Algorithm performance is shown for full NESAP region, as well as the three Longhurst biogeographical provinces within our study area.  * indicates significance of p<0.05.  Province SD02 original SD02 custom VS07 original VS07 custom W07 original W07 custom Whole region r = 0.05 RMSE = 3.77 r = 0.08 RMSE = 3.03 r = -0.31* RMSE = 4.95 r = 0.31* RMSE = 2.63 r = -0.08 RMSE = 67.1 r = 0.17* RMSE = 5.86 CCAL r = 0.04 RMSE = 3.42 r = 0.62* RMSE = 1.61 r = -0.23 RMSE = 4.54 r = 0.23 RMSE = 1.20 r = -0.17 RMSE = 81.6 r = 0.27 RMSE = 2.04 ALSK r = 0.16 RMSE = 2.37 r = 0.12 RMSE = 2.07 r = -0.10 RMSE = 3.43 r = 0.10 RMSE = 2.09 r = -0.20 RMSE = 47.5 r = 0.53* RMSE = 7.19 PSAE r = 0.09 RMSE = 3.97 r = 0.23 RMSE = 2.94 r = -0.39* RMSE = 5.28 r = 0.39* RMSE = 2.81 r = -0.01 RMSE = 20.6 r = 0.44* RMSE = 4.59    46  Figure 2.1 Cruise tracks and discrete sampling stations (red circles) for the July 2016 (O16) cruise (a, c) and August 2017 (O17) cruise (b, d).  Panels (a) and (b) show chl-a concentration (log scale), derived from AquaMODIS satellite, and averaged over the duration of the respective cruise.  Panels (c) and (d) show average sea surface height anomaly (SSHA). Panel (a) shows the location of the T1-T3 transects surveyed during the 2016, whereas panel (b) shows the geographic location of locations of interest.  The grey line represents the coastal-oceanic boundary, defined here as the 2000 m isobath.     T2T1T3Hecate StraitDixon EntranceHecate StraitLa Perouse Bank  47    Figure 2.2 Spatial distribution of DMS (a), salinity (b), net community production (C; note log scale), and micro-, nano-, and picophytoplankton relative abundance (d-f) during the O16 cruise July of 2016 and the O17 cruise August of 2017.  Color scaling on the maps are adjusted to ensure readability and best illustrate spatial patterns.  Some data values are higher than the maximum scale of the color bar.  The inset box shows the La Perouse Bank region, as marked by the red circle. The grey line represents the coastal-open ocean boundary (2000 m isobath).       48    Figure 2.3 DMS concentrations during the O16 cruise in July of 2016 (a) and the O17 cruise in August of 2017 (b) as measured by membrane inlet mass spectrometry (MIMS, continuous black line) and a purge-and-trap sampling system connected to a gas chromatograph equipped with a flame-photometric detector (grey symbols).  Mean absolute error was 0.93 nM for all paired measurements between the two instruments.   1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 0 5101525DMS (nM)// //MIMSFPD-GC500 1000 1500 2000 2500 3000 3500 4000 4500Distance along cruise track (km)051015DMS (nM)MIMSFPD-GC(a)(b)  49  Figure 2.4 MIMS-based DMS concentration measurements and station-based DMSPd measurements (a), salinity and bathymetry (b), chl-a and HPLC-based station estimates of diatom and prymnesiophyte as defined % contribution to total assemblage (c), DMSPt:chl-a ratios (d), DMS/P consumption rate constants (e), and bacterial and primary productivity rates (f) along the T1 transect west of Dixon Entrance during July 2016 (O16 cruise).   Shaded regions represent standard deviation of repeated measurements across the transect.  The vertical dotted line in panel (b) indicates the approximate shelf break (2000 m), at 134.4˚ W. (a)(b)(c)(d)(e)(f)  50   Figure 2.5 Line plot of sea surface height anomaly (SSHA) on 15 July, 2016 and observed DMS concentrations between 14 July and 16 July, 2016 along T1.  DMS along the T1 transect is highest in those areas influenced by positive SSHA values.   01234567891011DMS (nM) (m)MIMS DMSSSHA  51   Figure 2.6 As for Fig. 2.4, but for the T2 transect.  (a)(b)(c)(d)(e)(f)  52  Figure 2.7 As for Figs. 2.4 and 2.6, but for the T3 transect.    (a)(b)(c)(d)(e)(f)  53    Figure 2.8 Spatial distribution of summertime DMS measurements from MIMS (a; 2004-2017) and the PMEL (b; 1984 - 2004) data set.  Black lines represent boundaries of Longhurst biogeographical provinces, with province names show in panel (a).      54   Figure 2.9 Latitudinal distribution of data containing bins (a) and average DMS concentration (b) for PMEL (dashed line) and combined (PMEL and MIMS) data sets.   40 45 50 55 60 65Latitude (°N) 0 51015202530Number of data containing binsThis studyPMEL40 45 50 55 60 65Latitude (°N)246810121416Average DMS concentration (nM)This studyPMEL(a) (b)  55   Figure 2.10  Summertime DMS concentrations (a), DMS sea-air fluxes (b), and DMS:chl-a ratios (c) binned to 1˚ x 1˚ spatial resolution.  These maps were derived using our combined PMEL/MIMS data set (1984–2017; June, July and August).  Black lines correspond to boundaries of Longhurst biogeochemical provinces.  Maximum values (47 nM, 180 µmol m-2 d-1, and 47 nmol µg-1 for panels a, b, and c, respectively) exceed the bounds of the colorbars.  Maximum values for DMS and DMS flux occur in the waters south of the Alaska Peninsula, whereas maximum DMS:chl-a occurs mid PSAE.   56 Chapter 3: Spatial and temporal dynamics of dimethylsulfoxide in the northeast Subarctic Pacific and potential contributions to phytoplankton physiology and regional sulfur cycling  3.1 Introduction Dimethylsulfoxide (DMSO) is an abundant, though poorly understood, organic sulfur compound found throughout the surface ocean.  DMSO, and the related compounds dimethylsulfide (DMS) and dimethylsulfoniopropionate (DMSP), are essential components of the marine sulfur cycle, playing important roles in marine microbial systems (Simo, 2004; Stefels et al., 2007) and trophic webs (Johnson et al., 2016; Savoca, 2018; Seymour et al., 2010).  Oceanic DMS emissions have been linked to global climate through the production of atmospheric aerosols (Charlson et al., 1987; Leaitch et al., 2013; McCoy et al., 2015; Quinn and Bates, 2011; Tesdal et al., 2015).  DMSP has been shown to be an important carbon and sulfur source for marine bacteria (Bullock et al., 2017; Kiene et al., 2000; Reisch et al., 2011), and is believed to serve a number of physiological functions in phytoplankton, including protection from grazing (Fredrickson and Strom, 2009; Wolfe et al., 1997) and osmotic and oxidative stress (Dickson and Kirst, 1987; Lyon et al., 2016; Sunda et al., 2002).  Relative to DMSP and DMS, much less information is available on the metabolic functions or oceanographic distribution of DMSO.   To date, the study of DMSO cycling in marine systems has been limited, to a large extent, by analytical capabilities.  DMSO measurements are typically based on an indirect method, in which DMSO is converted to DMS prior to GC-based analysis of discrete samples (Kiene and Gerard, 1994; Simó et al., 1996).  These methods can yield robust and reproducible   57 DMSO measurements, yet they are laborious and time-consuming, and thus not amenable to high through-put analysis methods designed for DMS measurements (Royer et al., 2016; Tortell, 2005; Zhang and Chen, 2015).  Our group recently developed the Organic Sulfur Sequential Chemical Analysis Robot (OSSCAR) (Asher et al., 2015; Jarníková et al., 2018) to provide automated measurements of DMS, DMSP and DMSO in oceanic waters, using a sample processing system that includes DMSO reductase conversion of DMSO to DMS (Hatton et al., 1994).  We have also recently developed a stable isotope tracer method to examine the cycling of DMS, DMSP and DMSO (Asher et al., 2017a; 2011; 2017b), using specific isotope labelling patterns to simultaneously quantify multiple reactions in a single seawater sample.  Recent results obtained with this method suggest that DMSO is actively cycled in the mixed layer of the Southern Ocean and Subarctic Pacific, and can be a significant source of DMS.   In this article, we build on our recent work, presenting new results from two research cruises examining DMSO distributions and cycling across the northeastern subarctic Pacific (NESAP).  Our goal was to use the OSSCAR system to examine regional-scale patterns in DMSO concentrations, and to employ isotope tracer studies to identify dominant DMSO consumption and production pathways across a range of hydrographic regimes.  In particular, we aimed to investigate the influence of DMSO on DMS concentrations and compare these values directly to rates of DMS yield from DMSP cleavage.  We also employed a Lagrangian drifter approach to examine the diel cycle of DMSO concentrations and turn-over rates.  By pairing our data with measurements of phytoplankton community composition and other ancillary oceanographic variables, we examined potential ecological and environmental controls on DMS/O cycling in the NESAP.  Below, we provide context for our work by briefly reviewing key features of the oceanic DMSO cycle.   58  While global DMSO measurements are sparse, the available data show that dissolved concentrations typically exceed those of dissolved DMS and DMSP in surface waters (Hatton et al., 1996; 1999; 2004; Simo et al., 1997), and correlate strongly with DMS.  Particularly high DMSO concentrations (>100 nM) are observed in DMS ‘hot-spot’ regions in the Arctic, Antarctic, coastal equatorial Pacific, and North Sea (Asher et al., 2017b; Hatton et al., 2004; Jarníková et al., 2018; Speeckaert et al., 2018).  A number of studies have demonstrated the widespread existence of a large particulate DMSO pool (DMSOp) (Lavoie et al., 2015; Lee and Mora, 1999; Riseman and DiTullio, 2004; Simó and Vila-Costa, 2006), with intracellular DMSO concentrations in the millimolar range found in a variety of phytoplankton species (Bucciarelli et al., 2013; Hatton and Wilson, 2007; Spiese et al., 2009). Given the high permeability of DMSO through cellular membranes (Narula, 1967; Spiese et al., 2009; Tanaka et al., 2001), diffusive loss of intracellular DMSO from phytoplankton may be an important, though poorly quantified, source of dissolved DMSO.   Additionally, a recent study has identified the compound dimethylsulfoxonium propionate (DMSOP) as a previously unknown DMSO source (Thume et al., 2018).  DMSOP production was demonstrated in both phytoplankton and bacteria, as a likely oxidative by-product of DMSP.  The oxidation of DMSOP by bacteria and phytoplankton yields DMSO, both intracellularly and in the dissolved pool (Thume et al., 2018).  Wet deposition from the atmosphere constitutes another potential DMSO source (Kiene and Gerard, 1994), though it is unlikely to be quantitatively important (Sciare et al., 1998).  Rather, the primary source of DMSO is believed to be the bacterial and photochemical oxidation of DMS, which acts as both a significant source of DMSO and an important loss term in the DMS budget (Brimblecombe and Shooter, 1986; del Valle et al., 2007; 2009).  The relative contribution of biotic and abiotic oxidation appears highly variable, but both processes have been   59 shown to be potentially significant (Hatton, 2002; Jian et al., 2017; Kieber et al., 1996; Levine et al., 2015; Toole et al., 2004; del Valle et al., 2009).   DMSO consumption pathways remain poorly quantified, although a number of possible DMSO sinks have been suggested, including bacterial consumption (de Bont et al., 1981; Tyssebotn et al., 2017), export via sinking particles (Hatton et al., 1996), biological oxidation to methanesulfinic acid (Sunda et al., 2002) and biological reduction to DMS (Zinder and Brock, 1978).  To date, no study has attempted to determine the relative importance of these pathways within a single sample.  DMSO reduction is of particular interest, as this mechanism provides a two-way inter-conversion between DMS and DMSO, through a coupled oxidation-reduction pathway that could act to buffer DMS concentrations.  DMSO reduction activity is widespread among bacteria and phytoplankton, both in laboratory cultures and natural assemblages (Fuse et al., 1995; González et al., 1999; Griebler and Slezak, 2001; Spiese et al., 2009; Vila‐Costa et al., 2006).  As with DMSP-lyase activity, DMSO reduction activity in phytoplankton appears to be linked to oxidative stress, with increased activity observed in nutrient-limited laboratory cultures of Thalassiosira pseudonana (Spiese and Tatarkov, 2014). This finding points towards the potential role of DMSO reduction in the cellular oxidative stress response.   To our knowledge, only three studies have directly measured marine DMSO reduction rates in situ (Asher et al., 2017a; 2011; 2017b).  Asher et al. (2011) demonstrated extremely high DMSO concentrations and reduction rates in sea ice brines, ice-covered seawater, and polynya waters of the Antarctic sea-ice zone, with DMSO reduction rates exceeding those of DMSP cleavage.  By comparison, DMSO reduction rates in the northeast subarctic Pacific (Asher et al., 2017a) were less than 10% of DMS yield from DMSP cleavage.  In these prior studies, DMS oxidation to DMSO was not measured, and the net impact of DMSO reduction on the DMS pool could not be quantified.  In light of the discovery of the DMSOP pool and increasing evidence of   60 DMSO production by phytoplankton, the potential importance of DMSO as a net source of DMS must be re-examined. The relatively high concentrations of DMSO present in surface seawater, together with the reported prevalence of DMSO reduction, motivates our work.   In this paper, we leverage our OSSCAR system and stable isotope tracer methodology to contribute new information on DMSO concentrations and cycling in the NESAP, with a particular emphasis on the influence of DMSO on DMS concentrations.  We quantified DMSO reduction rates across the NESAP, and compared these with simultaneous measurements of DMSP cleavage and DMS oxidation rates to investigate the relative importance of DMSO reduction to the DMS cycle in this region.  By combining ship-track observations with a Lagrangian drifter survey, we were able to explicitly examine temporal and spatial components of observed variability in sulfur concentrations and rate processes.    3.2 Methods 3.2.1 Study area We conducted field sampling on two oceanographic cruises in the Subarctic NE Pacific, a region of elevated DMS concentrations (Lana et al., 2011; Steiner et al., 2012).  To date, only three studies have measured DMSO concentrations in this region (Asher et al., 2017a; 2015; Bates et al., 1994) and one study has provided preliminary observations of DMSO reduction rates (Asher et al., 2017a).  Coastal waters adjacent to Vancouver Island and Haida Gwaii (formerly known as the Queen Charlotte Islands) were surveyed between 26th May and 3rd June 2017 aboard the Canadian Coast Guard Ship John P. Tully, in conjunction with the La Perouse monitoring program run by Fisheries and Oceans Canada.  We also surveyed open ocean and coastal waters from the northern Gulf of Alaska to the Oregon coast aboard the UNOLS vessel R/V Oceanus between 11th – 27th August 2017.  During this latter cruise, we had the opportunity to conduct a   61 Lagrangian Drifter survey in upwelling continental shelf waters along the Oregon coast (Fig. 3.1).  3.2.2 Concentration measurements We measured surface water DMS, DMSO and DMSP concentrations using the organic sulfur chemical analysis robot (OSSCAR).  This system provides automated underway measurements of these compounds at relatively high spatial resolution, with continuous measurements every ~40 minutes along a cruise track.  OSSCAR has been deployed on several previous cruises (Asher et al., 2015; Jarníková et al., 2018), and detailed methodology of this system is described elsewhere (Asher et al., 2015).  Briefly, the system consists of a custom-built sample-handling module, a purge-and-trap system to extract and concentrate gases from seawater samples, and a gas chromatograph equipped with a capillary column (Restek SS MXT, 15 m) and a pulsed-flame photometric detector (PFPD).  Unfiltered seawater samples (5 mL) were collected from the ship’s underway supply (nominal sampling depth ~ 5 – 7 m) and analyzed sequentially for DMS, DMSO and DMSP.  The samples were first sparged with ultra-high purity N2 gas (99.999%), and the DMS concentrated at onto a stainless-steel trap held at room temperature and loaded with Carbopack X (Sigma; graphitized carbon black adsorbent material).  The analytes collected in the trap were subsequently desorbed by rapid electrical heating, and flushed onto a capillary column for separation and detection by PFPD.  After DMS analysis, DMSO in the remaining water was analyzed following enzymatic conversation to DMS using an automated version of the DMSO reductase method (Hatton et al., 1994).  The DMSO to DMS reduction reaction was catalyzed by exposure to LED lights for 20 minutes, and the resulting DMS measured as outlined above.  It is important to note that our DMSO analysis does not involve a NaOH addition step, which has previously been shown to artificially increase DMSO   62 measurements in unfiltered samples, possibly via the cleavage of either DMSP-derived dimethyloxosulfoniopropionate (DMOP) or DMSOP to DMSO (Brinkley, 2008; Lavoie et al., 2015; Spiese et al., 2009; Thume et al., 2018).  Following DMSO measurement, DMSP was cleaved to DMS and acrylate through addition of 5 M NaOH, using the method of Dacey and Blough (1987), with the resulting DMS measured as described above.  After each analysis, the system completed a rinsing cycle, prior to initiating the subsequent measurement cycle.  A 6-point calibration curve was performed daily to correct for instrument drift.   As we did not pre-filter seawater samples, DMS, DMSP, and DMSO measurements reflect potential contributions from the dissolved and particulate pools.  With the NaOH hydrolysis step, DMSP concentrations measured here are assumed to reflect something close to total concentrations (i.e. combined dissolved and particulate pools).  By comparison, the enzyme-based DMSO analysis method does not disrupt cellular membranes (beyond potential damage due to gas sparging), and our DMSO concentrations may thus somewhat underestimate the total (DMSOt) pool.  However, given the rapid equilibration between intracellular and extracellular DMSO (first order rate constant of 0.4 – 0.5 s-1; Spiese et al. 2009) we expect any underestimation to be minor, as conversion of extracellular DMSO to DMS by the enzyme would lower the extracellular DMSO concentration, promoting efflux from the intracellular pool.  Cellular exposure to high light levels may pose an additional source of error to DMSO concentration estimates, by promoting cellular DMS production from DMSP, which would be measured as DMSO during subsequent analysis.  However, this source of error has not been noted in previous papers utilizing the DMSO reductase method, and is expected to be minor. During the August 2017 cruise, concentrations of DMS, total DMSP (DMSPt) and dissolved DMSP (DMSPd) were also measured in discrete water samples collected from Niskin bottles at most sampling stations.  DMSPd samples were collected using the small-volume   63 gravity drip filtration method (Kiene and Slezak, 2006), and DMS concentrations measured using a purge-and-trap system connected to a gas chromatograph equipped with a flame-photometric detector (FPD-GC) (Kiene and Service, 1991).  DMS measurements obtained using this method showed good agreement to OSSCAR measurements, with a root mean squared error (RMSE) of 0.67 nM over a measured concentration range of >1 to 26 nM.   3.2.3 Rate measurements Measurements of net DMS production, gross DMS loss, DMS production through DMSP cleavage and DMS production through DMSO reduction were made using a stable isotope tracer addition method, previously described in detail by Asher et al. (2017a; 2011; 2017b).  This method offers the benefit of simultaneous quantification of these four measured rate processes within a single experimental treatment.  We conducted a total of 5 rate measurements in British Columbia coastal waters in May-June 2017.  During August 2017, 26 further rate measurements were conducted in both on- and off-shore waters of the northeast subarctic Pacific, including 12 experiments during a Lagrangian drift survey, described below. Seawater for rate experiments was collected from 5m depth with Niskin bottles, and analyzed for initial DMS, DMSO and DMSPd concentrations, when possible, using the OSSCAR system or discrete FPD-GC analysis.  Sample water was then spiked with D-3 deuterated DMS (CDN Isotopes, 99.9% purity), D-6 deuterated DMSP (produced using the method of Challenger and Simpson, 1948, from D-6 DMS Sigma Aldrich, 99% purity), and D-6 deuterated, 13C-labeled DMSO (ISOTEC, 99% purity) at tracer level concentrations (10% ambient) when initial concentrations were immediately available.  When logistical constraints prevented the measurement of initial concentrations (19 out of 31 experiments), we added tracers to a final concentration of 0.5 nM.  Sample water was gently homogenized, and 1 or 3 L   64 dispensed into UV-transparent FEP bags (Welch Fluorocarbon).  Bags were incubated in a deck-board seawater tank maintained close to in-situ surface temperature using flowing seawater.  Light levels were reduced to ~30% surface irradiance using either blue photographic film (LEE filters: #202; May-June 2017), or two layers of neutral density screening (August 2017).  During the May-June cruise, the incubator lid remained on, effectively blocking exposure of the samples to UV.  As such, rates calculated during this cruise do not reflect UV-dependent processes, including DMS photolysis.  During our August 2017 experiments, the tank lid was left open to allow for UV exposure.  We note, however, that UV intensity was reduced compared to ambient levels, due to UV blockage from the sides of the incubator.   Over a period of ~2 – 5 hours, we collected 5 mL subsamples from each sample bag every 30-90 minutes using a glass syringe with a Luer Lock port to minimize gas exchange.  The concentration of isotopically labelled DMS species was determined using a SCIEX API 3200-series triple quadrapole mass spectrometer equipped with an atmospheric pressure chemical ionization source (McCulloch et al., in prep).  DMS, and each of its isotopically labelled analogues were found to ionize exclusively through the proton transfer (PT) mechanism, yielding abundant [M+H]+ ions.  As a result, gases were measured at mass to charge ratio (m/z) 1 unit higher than the native state.  The instrument was operated in multiple reaction monitoring (MRM) mode, for high analyte selectivity.  We measured the change in m/z 63 signal of natural DMS (to measure net DMS production), labeled D-3 DMS (m/z 65; to measure gross DMS consumption), D-6 labeled DMS (m/z 69) resulting from cleavage of labeled DMSP, and m/z 71 DMS resulting from reduction of D-6 13C labeled DMSO.  Rate constants were calculated as pseudo-first order reactions, as per Asher et al. (2017a; 2017b).  Uncertainty estimates were derived from the standard error of linear fits to the data using the MATLAB linfit.m function (Glover et al., 2011).     65 For a sub-set of stations, DMSO production from DMS oxidation was measured by tracking the appearance of DMSO-derived m/z 66 from the conversion of D-3 DMS.  Samples from these experiments were frozen at -80 ˚C for ~7 months prior to subsequent laboratory analysis of the labelled DMSO pool (Hatton et al., 2012; Kiene and Gerard, 1994).  Samples were thawed in the dark at room temperature, sparged to remove endogenous DMS and enzymatically reduced prior to analysis.      3.2.4 Diel cycle measurements using Lagrangian drifters We conducted a Lagrangian drifter survey during our August 2017 cruise in the coastal and offshore waters of central Oregon.  Sea surface temperature, chlorophyll and sea level anomaly data derived from AquaMODIS and AVISO multi-satellite based platforms were used to guide our ship track in order to identify distinct water masses of interest.  The drifter was deployed in cold, saline shelf waters off the Oregon coast for a period of ~80 hours.   Water samples for incubation experiments and ancillary measurements (described below) were collected every 6 hours (00:00, 06:00, 12:00, and 18:00), for a total of 14 sampling time points (L1-L14).  Isotope tracer experiments were conducted at 12 of the 14 time points.  Of these 14 experiments, 4 experiments were further analyzed for isotopic changes in the DMSO pool (time points L6, L8, L12, and L14).    3.2.5 Ancillary measurements To provide an oceanographic context for our measurements, we collected ancillary data from the ship’s underway sensors and at discrete sampling stations.  Salinity and sea surface temperature (SST) were measured using a thermosalinograph connected to the ship’s underway seawater supply (Sea-Bird SBE-21 on the R/V John P. Tully, and SBE 45 and SBE 38 for salinity   66 and temperature respectively, aboard the R/V Oceanus).  Chlorophyll-a (chl-a) concentrations were determined optically using a WET labs ACS absorbance/attenuation meter, based on the measurement of particulate absorption at 676 nm (Bricaud et al., 1995; Burt et al., 2018; Roesler and Barnard, 2013).  Particulate inorganic carbon (PIC) was derived from AquaMODIS satellite data (Balch et al., 2005; Gordon et al., 2001).  At selected stations during the August 2017 cruise, bacterial production was measured using the tritiated leucine method (Smith and Azam, 1992), and primary productivity was measured using 24 h 14C uptake incubations, following the method of Schuback et al.(2015).  We used HPLC measurement of phytoplankton pigment concentrations to infer phytoplankton assemblage composition using diagnostic pigment analysis (DPA) following the methods described in Zeng et al. (2018), based on the initial work of Hirata et al. (2011).  During the drifter survey, measurements of various photosynthetic parameters including non-photochemical quenching (NPQ) were conducted every 2 hours for a 48 hour period using a Soliense fast repetition rate fluorometer (FRRF) following the protocols described by Schuback et al. (2015).  A separate FRRF was used to collect high resolution underway measurements of maximum quantum efficiency of photosystem II, as estimated by Fv/Fm (Roháček, 2002).  3.3 Results 3.3.1 Oceanographic conditions Our study region encompassed two distinct oceanographic regimes: high-nutrient low-chlorophyll (HNLC) waters offshore that experience summer-time iron limitation, and higher productivity coastal regions influenced by seasonally-variable currents, freshwater input and coastal upwelling (Boyd et al., 2004; Harris et al., 2009; Harrison et al., 1999; Taylor and Haigh, 1996).  Strong coastal upwelling was apparent along the Oregon coast in August 2017, as   67 evidenced by high salinity, low temperature waters, and negative sea surface height anomalies (Venegas et al., 2008).  Chl-a concentrations across our study area showed typical distributions for the region, with higher values observed in coastal regions (mean 1.9 µg L-1) compared to offshore waters (mean 0.37 µg L-1; Fig. 3.1).  During both cruises, we observed areas of elevated chl-a concentrations adjacent to the Brooks Peninsula and La Perouse Bank along the west coast of Vancouver Island, with maximum values of 18 µg L-1.  Phytoplankton assemblages at on-shelf stations were generally diatom-dominated, except at stations with nitrate concentrations less than ~0.5 µM.  By comparison, off-shelf stations were generally dominated by nano- and pico-sized phytoplankton.  During August, 2017, a large apparent coccolithophore bloom (calcite ~0.002 mol PIC/m3) was observed in the central Gulf of Alaska, with HPLC-based estimates of phytoplankton assemblage composition showing prymnesiophyte abundance to be greater than 25% in this region.  More information on the hydrography and phytoplankton assemblage composition across our survey region is presented in Herr et al. (in review).    3.3.2 DMS/O concentrations The spatial distribution of DMSO during our surveys was similar to that of DMS, with both of these compounds exhibiting significant heterogeneity, particularly in coastal waters.  Across our full cruise tracks, DMS and DMSO concentrations ranged from <1 to 14 nM (DMS) and <1 to 25 nM (DMSO) in May-June (Fig. 3.2), and <1 to 26 nM and <1 to 183 nM, respectively, in August (Fig. 3.3). The highest DMSO concentrations (greater than the 95th percentile, 42 nM) were found in the central Gulf of Alaska, in the region of an apparent coccolithophore bloom, where DMS concentrations were ~20 nM.  As shown in figure 3.4, at concentrations below 20 nM, DMS and DMSO showed a linear relationship, with DMSO concentrations equal to or slightly higher than DMS ([DMS]=1.194*[DMSO]+0.6372).    68 However, at higher DMSO concentrations, we observed a significant decoupling between DMSO and DMS, resulting in a non-linear relationship between these two compounds across the full range of concentrations observed (Fig. 3.4).  Overall, we were able to model the DMS – DMSO relationship using a two-term exponential curve that explained 84% of the observed variability:  [DMS]=21.3*e5.53E-5*[DMSO]-20.5*e-0.0414*[DMSO]  Across both cruises, the mean DMSO:DMS ratio was 1.48 ± 1.43, with an overall range of 0.07 to 12.8 (Fig. 3.2, 3.3).  We found a positive correlation between the DMSO:DMS ratio and the DMSPt:chl-a ratio (r=0.51, p<0.001, n = 308), suggesting higher concentrations of DMSO relative to DMS in areas where the phytoplankton assemblage had elevated intracellular DMSP concentrations.  3.3.3 Rate experiments Across our study area, DMSO reduction rate constants ranged from <1 d-1 to 6.5 d-1,   and were highest in on-shelf (<2000 m depth) waters.  As shown in figure 3.5a, we consistently observed DMSO reduction rate constants (kDMSOred) that were similar to or higher than DMSPd cleavage rate constants in the experimental incubations (kDMSPcleav).  For the experiments conducted in August, average kDMSOred was 2.33 ± 1.67 d-1, as compared to average kDMSPcleav of 1.54 ± 1.55 d-1.  A similar pattern of higher kDMSOred was observed during May/June (5.34 ± 0.84 d-1 and 4.39 ± 2.37 d-1, respectively), although average rate constants for both DMSO reduction and DMSP cleavage were generally higher at this time.  These higher overall rate constants during the May cruise may be partially explained by the use of a closed incubator during these   69 experiments and the resulting lack of UV-dependent DMS oxidation in samples.  Regardless, results from both cruises demonstrated the ubiquity of DMSO reduction throughout this region, and its quantitative significance relative to DMSP cleavage.   Rates of DMSO reduction and DMSP cleavage (nM d-1) were calculated by multiplying measured rate constants by in situ concentration of DMSO and DMSPd for those stations where we had reliable concentration measurements.  Due to the significantly larger DMSO pool, DMSO reduction exceeded DMSPd cleavage at nearly all stations (Fig. 3.5b).  DMSO reduction rates were higher in on-shelf waters (Fig. 3.6), ranging from below detection to ~145 nM d-1, with an average of 12.2 ± 27.7 nM d-1 during our August cruise, and 14.4 ± 4.26 nM d-1 during May/June.  By comparison, DMSPd cleavage rates averaged 1.88 ± 2.42 nM d-1 for experiments conducted in August.  DMSPd cleavage rates were not calculated for the May/June cruise due to the lack of reliable in situ concentration data.   Rates of DMSO reduction were strongly correlated to bacterial production (r=0.96, p<0.001; n = 11) (Fig. 3.7).  Bacterial production was also positively correlated to the corresponding rate constants (d-1) for DMSO reduction (r=0.62, p=0.02).  These strength of these correlations may be due, in part, to the high bacterial production observed at station O7 (14.3 nM Leu d-1), where the highest cruise-wide DMS consumption rate was also observed.  However, significant positive correlations remain when this high value is removed.  DMSO reduction rate constants showed a negative correlation with DMSPt:chl-a ratios (r=-0.58, p=0.003) and prymnesiophyte relative abundance (r=-0.43, p=0.048), and a positive correlation with relative abundance of microphytoplankton and dinoflagellates (r=0.43, p=0.048; 0.48, p=0.023, respectively) (Table 1).  In contrast to the relationships with bacterial production and taxonomic group abundance, we found that DMSO reduction rate constants did not correlate to concentrations of DMS, DMSPd, DMSPt, DMSO, or primary productivity.  Similarly, DMS/P/O   70 concentrations were not significantly correlated to rate constants of DMS net production, DMSP cleavage or DMS consumption.   At 4 time points during the drifter survey (see below), we measured the rate of total DMS oxidation (light-dependent and biological) by tracking the production of D-3 deuterated DMSO derived from an added D-3 DMS tracer.  We compared these rates with those of DMSO reduction measured at the same sampling time point to assess whether DMSO acted as a net source or net sink to the DMS pool.  In these samples, we observed a strong correlation between rate constants of DMS oxidation and DMSO reduction (r=0.99, p=0.011, n = 4) suggesting a tight coupling between these processes.  DMS oxidation rate constants (kDMSox) ranged from 0.81 d-1 to 4.08 d-1, with associated rates ranging from 2.77 nM d-1 to 9.56 nM d-1.  By comparison kDMSOred values ranged from 1.67 d-1 to 3.93 d-1, with associated rates of 3.27 nM d-1 to 14.1 nM d-1.  In all cases, the rates of DMSO reduction exceeded those of DMS oxidation (Table 2), suggesting DMSO reduction as a net source of DMS.  The magnitude of this source ranged from 0.50 nM d-1 to 7.18 nM d-1.  3.3.4 Diel cycling of DMSO We deployed a Lagrangian drifter for ~80 hours in relatively cold (12.0 ˚C), saline (32.6 psu) upwelling coastal waters west of Waldport, OR, approximately 40 km offshore (Fig 3.1).  Surface waters at the drifter deployment site were characterized by excess nutrients (8.4 µM nitrate, 0.79 µM phosphate and 10.0 µM silicic acid concentrations), and an average chl-a concentration of 1.13 µg L-1 (Fig 3.8c).  Through the duration of drifter deployment, we observed evidence for diel cycles in a number of biogeochemical properties, including chl-a concentrations and photosynthetic efficiency (Fv/Fm).  The phytoplankton community varied   71 over the course of the study, with diatom relative abundance increasing from ~40 to 70%, and nanoplankton relative abundance decreasing from ~45 to 20%.   The Lagrangian drifter survey provided an opportunity to examine diel processes influencing DMSO and DMS concentrations, superimposed on longer time-scale processes associated with environmental forcing (e.g. surface warming).  During the ~80 hours of Lagrangian measurements, both DMS and DMSO showed an overall increase in concentrations (Fig 3.3).  In order to better isolate a diel signature from our data, we removed the multi-day linear trend using the detrend function in MATLAB.  The linearly detrended residuals of DMS and DMSO concentrations are shown in figure 3.8 (note that non-detrended observations are shown in Fig. 3.3), more clearly highlighting apparent diel cycles.   Most notably, we observed a decrease in DMSO concentrations relative to DMS during periods of peak irradiance levels near mid-day.  These decreases in DMSO concentrations occurred in conjunction with daily minima in photosynthetic efficiency (Fv/Fm), and showed a striking inverse pattern to values of non-photochemical quenching (NPQ; Fig. 3.8a).  Indeed, for observations collected over the entire drifter period, we observed a statistically significant linear relationship between detrended DMSO concentrations and NPQ (r = -0.65, p<0.001; Fig. 3.9), and a negative correlation with Fv/Fm (r = 0.43, p<0.001).  In contrast to DMSO, little significant diel cycling of DMS concentrations was apparent, although DMS appeared to decrease with increasing PAR during day 3 and 4 of the drifter survey (Fig. 3.8b).   DMSO reduction rates varied significantly over this course of the drifter deployment, but did not show a strong diel pattern (note that we lack mid-nighttime measurements of DMSO reduction in 2 of the 3 nights surveyed; Fig. 3.8a).   Changes in DMS oxidation showed patterns very similar to those of DMSO reduction rate constants (Fig. 3.8b), with no apparent diel cycling observed.   72  3.4 Discussion The lack of widespread seawater DMSO concentration data and associated rate measurements limits our understanding of this compound’s role in the marine sulfur cycle.  The results presented here, derived from a recently developed analytical system and stable isotope tracer technique, reveal high DMSO concentrations and reduction rates across a number of hydrographic regimes in the NESAP.  Notably, we found that rates of DMS production from DMSO reduction were comparable to those of DMSP cleavage, challenging the assumption that DMSP cleavage is the predominant source of DMS.  Below we discuss the potential drivers of observed DMSO spatial distribution and diel trends, the influence of phytoplankton community assemblage on DMSO concentrations and reduction rates, and the impact of DMSO reduction on DMS concentrations.    3.4.1 DMSO concentration in the NESAP To date, few studies have examined DMSO concentrations in the NESAP, and there are thus limited data available for comparative purposes.  Approximately 90% of DMSO measurements presented here fall within the range of previous observations (Asher et al., 2017a; 2015; Bates et al., 1994).  However, the maximum concentrations we observed in a localized coccolithophore bloom in the northern Gulf of Alaska (> 100 nM) greatly exceed previously reported values for our study region.  There is, however, precedent for such elevated DMSO concentrations in other marine systems, including previous reports from the Arctic (Bouillon et al., 2002), southwest Pacific coastal (Lee and de Mora, 1996), equatorial Pacific (Hatton et al., 1998), North Sea (Speeckaert et al., 2018), and Antarctic (Asher et al., 2017b).  It is thus possible that such high values in our study region have simply not been observed previously due to sparse   73 and infrequent sampling of DMSO.  As discussed below, there are several factors that could act to promote such elevated DMSO concentrations across various parts of the NESAP.    3.4.2 Relationship between DMS and DMSO concentrations Our results show a positive correlation between DMS and DMSO, confirming previous observations at local (Asher et al., 2015; Kiene et al., 2007) and global (Hatton et al., 2004) scales.  While DMS and DMSO concentrations showed a strong linear relationship at concentrations up to ~ 20 nM, we observed a decoupling of the DMS – DMSO relationship at high DMSO concentrations (Fig. 3.4).  In previous studies, high concentrations of DMSO relative to DMS have been attributed to elevated bacterial oxidation of DMS into DMSO (Speeckaert et al., 2018).  In our study, the majority of high (> 40 nM) DMSO concentrations were observed in a region of elevated calcite in the northern Gulf of Alaska in August of 2017.  Although chl-a values within this area were generally below the study-wide mean (1.3 µg L-1), levels of organic matter may have been elevated due to detrital carbon originating from the bloom, stimulating bacterial metabolism of reduced sulfur compounds.  Potential iron limitation may also explain the high DMSO:DMS ratios in the northern Gulf of Alaska.  Previous work by Kinsey et al. (2015) demonstrated a 6.9-fold increase in cell-normalized DMSOd concentrations in iron-limited cultures of Phaeocystis antarctica, as compared to iron-replete controls, with similar increases in cellular DMSP and acrylate levels.  The authors proposed that elevated DMSO was due to increased scavenging of reactive oxygen species (ROS) by DMS and DMSP, resulting in increased intra-cellular DMSO production.  In support of this idea, excess surface nitrate concentrations (~7 µM) and low photosynthetic efficiency (Fv/Fm ~0.28) in the northern Gulf of Alaska are consistent with iron limitation in the region of the high DMSO:DMS ratios.     74 The strong DMSO correlation with DMSPd and DMSPt across our survey region (r=0.81, p<0.001; r=0.74, p<0.001, respectively; Table 1) suggests that phytoplankton may be important contributors to seawater DMSO concentrations, through intracellular production and rapid efflux, or exudation and subsequent oxidation of DMSOP.  Due to varying intracellular concentrations of DMSO and DMSO reduction activity among phytoplankton taxa, differences in phytoplankton community assemblage composition would also be expected to play an important role in determining DMSO – DMS dynamics.  Indeed, we observed a positive correlation between the DMSO:DMS and DMSPt:chl-a ratios (r=0.51, p<0.001, n = 308).  While this relationship was largely driven by extreme DMSO:DMS values within the coccolithophore bloom, a statistically significant positive correlation remains even if these values are removed.  A taxonomic influence on DMSO concentrations and cycling may reflect the interacting roles of DMSO and DMSP in phytoplankton.  Several authors have proposed that intracellular DMSO is produced through the oxidation of DMS, DMSP and DMSOP by ROS, as part of a cellular antioxidant system (Spiese et al., 2009; Sunda et al., 2002; Thume et al., 2018).  Under this assumption, DMSP and its oxidative by-products (DMS, DMSOP, DMSO and methanesulfinic acid) act to sequentially scavenge ROS, with DMS acting as a significantly more active ROS scavenger than either DMSP or DMSO (Spiese and Tatarkov, 2014; Sunda et al., 2002).  Indeed, Spiese et al. (2009) suggested that cellular oxidation of DMS by ROS may provide a signal of oxidative stress, activating transcription of genes involved in DMSO reduction, which would act to regenerate DMS.  Such DMS-DMSO cycling may function as an effective anti-oxidant system, and be particularly important in those phytoplankton with lower DMSP lyase activity and hence lower capacity to generate DMS from DMSP.  We might therefore expect to see a functional trade-off between DMSO reductase and DMSP lyase activity within a phytoplankton assemblage, with DMSP-rich assemblages preferentially utilizing a   75 DMSP-lyase to generate DMS, as opposed to DMSO reduction.  Lower rates of DMSO reduction would, in turn, contribute to elevated high DMSO:DMS ratios, as observed in the vicinity of the high calcite signal in the northern Gulf of Alaska.  In support of this theory, we observed a negative correlation between DMSO reduction rate constants and DMSPt:chl-a ratios (r=-0.66, p=0.003) and prymnesiophyte abundance (r=-0.43, p=0.048).   The relative importance of phytoplankton and bacteria to DMSO reduction is currently unknown, though it has been typically assumed that the majority of DMSP-lyase and DMSO-reductase activity is attributable to bacteria (Spiese et al. 2009).  Our results suggest that phytoplankton may play an important (if perhaps indirect) role in DMSO cycling in systems where bacterial activity is low.  In the coccolithophore bloom, for example, high DMS and DMSO concentrations were observed, while bacterial activity was ~30% that of the cruise-wide average.  This result highlights the potential importance of phytoplankton in determining DMS and DMSO dynamics in this system.  Future studies employing tracer measurements in size-fractionated seawater samples could be used to examine the contribution of different trophic groups to DMSO cycling.  3.4.3 The role and impacts of DMSO reduction  In 13 of 15 experiments where both DMSP cleavage and DMSO reduction rates were measured (Fig. 3.5b), we found that DMSO reduction exceeded rates of DMSP cleavage.  The high DMSO reduction rates were attributable to both elevated rate constants and high DMSO concentrations relative to DMSPd.  This result suggests that DMSO reduction was a major DMS production term across our survey region.  Further experiments, tracking the appearance of labeled DMS in the DMSO pool, enabled us to measure both reduction of DMSO and oxidation of DMS, and thus quantify the net contribution of DMSO reduction to the DMS pool.  To our   76 knowledge, this is the first direct comparison of in situ measurements of DMS oxidation and DMSO reduction.  Results from these experiments (derived from a relatively small number of samples) suggest that DMSO reduction is a net contributor to the DMS pool, with reduction rates exceeding those of DMS oxidation in all samples.  The net gain to the DMS pool ranged from 0.5 nM d-1 to 7.18 nM d-1 (Table 2).  By comparison, rates of DMS production from DMSPd cleavage across the same four sampling time points averaged 2.20 nM d-1.  These results challenge the commonly held assumption of DMSPd cleavage as the predominant source of DMS.   We note that the use of DMSOt (rather than DMSOd) in these calculations may result in an overestimate of DMSO reduction rates.  However, the high membrane permeability of DMSO relative to DMSP (Spiese et al. 2009) means that isotopically labeled DMSOd is likely available to phytoplankton, and thus calculated rate constants likely represent community-wide reduction rates.  In contrast, cleavage rates of DMSPd measured here are likely restricted to bacterial processes acting on the dissolved DMSP pool, and do not include any particulate (cellular) DMSP cleavage performed by phytoplankton.   Previous laboratory studies have demonstrated DMSO reduction to DMS in cultures of natural marine bacterial communities (Vila‐Costa et al., 2006), and it is now apparent that DMSO reductase activity is widespread among bacteria. Indeed, methods have been developed using DMSO reduction (and subsequent production of DMS) as a proxy for bacterial activity in many different ecosystems (Alef and Kleiner, 1989; Griebler and Slezak, 2001; Saari and Martikainen, 2003; Sklorz and Binert, 1994; Sparling and Searle, 1993).  In support of this result, we found a positive correlation between DMSO reduction rates and bacterial activity (Fig. 3.7), suggesting a major role for bacteria in DMSO reduction.   We note, however, that DMSO reduction has also been observed in all phytoplankton thus far examined (13 species to date),   77 including those with no DMSP-lyase activity (Fuse et al., 1995; Spiese et al., 2009).  Spiese et al. (2009) found that among 6 phytoplankton species tested, ~64-97% of total intracellular phytoplankton DMS production could be attributed to DMSO reduction.  While the active production and exudation of DMSO from phytoplankton has yet to be measured directly, high intracellular DMSO levels and particle-associated dissolved DMSO production suggest that seawater DMSO is not derived from extracellular DMS oxidation alone (Brimblecombe and Shooter, 1986; del Valle et al., 2007a; 2009).   Further, new results have demonstrated the production and exudation of DMSOP, which is readily metabolized to DMSO.  Given the high apparent potential for both bacterial and phytoplankton to reduce DMSO, it is reasonable to expect that DMSO reduction acts as an important net source of DMS.  In comparison to laboratory studies, direct field-based measurements of DMSO reduction are scarce.  To our knowledge, results from this study provide only the fourth data set of oceanic DMSO reduction to DMS (Table 3).  Previous measurements of in situ DMSO reduction in the Antarctic and NESAP have yielded rates ranging from 2.36 – 150 nM d-1.  Reduction rates were particularly high in sea-ice brines and polynya waters of the Ross Sea, where plankton are likely to experience significant oxidative stress (Asher et al., 2011).  Our measurements fall within the range of those previously reported for Antarctic waters, but they are higher than previously reported values from the NESAP region (Table 3).  For example, Asher et al. (2017b) found mean DMSO reduction rates in the NESAP to be approximately 50% lower than the values reported here.  The associated DMSO reduction rate constants measured by Asher et al. (2017b) averaged 0.07 d-1, as compared to our average of 2.82 d-1.  While the difference in rates between the two studies is surprising given similar study timing and location, the lower rates observed by Asher et al. may be due in part to differences in the plankton communities sampled.  The samples analyzed by Asher et al. were dominated (~ 80% of chl-a) by dinoflagellates and   78 prymnesiophytes, which are known to contain high intracellular DMSP and DMSP lyase levels.  By comparison, we observed much lower representation of these groups in our samples (20% of chl-a).  As discussed above, there may be a trade-off between DMSP lyase activity and DMSO reductase, possibly explaining the lower DMSO reduction rates observed by Asher et al.   Analytical differences between our study and that of Asher et al. may also explain some of the differences in derived DMSO reduction rates.  While the overall experimental procedure was similar in our work and that of Asher et al., our recent acquisition of a high-sensitivity mass spectrometer has increased measurement accuracy and precision, and thus provides increased confidence in the resulting measurements.  We also note that Asher et al. conducted experiments at only 5 stations, as compared to the 31 stations sampled during our cruises.  Further measurements are thus needed to better characterize the variability in potential DMSO reduction rates across the NESAP and other oceanic regions.     3.4.4 Diel cycling DMSO removal and cycling has generally been assumed to be a slow process (Kiene and Gerard, 1994; del Valle et al., 2007b), and very few studies to date have examined variability of DMSO concentrations and turn-over rates on timescales ranging from hours to days.  However, given the importance of photo-oxidation and UV-induced oxidative stress on DMS/O dynamics, diel changes in rates and concentrations are expected.  DMSO concentrations may increase during daylight hours due to enhanced DMS photooxidation and/or higher rates of intracellular DMS and DMSP oxidation by UV-stressed phytoplankton.  At the same time, DMSO loss (through biological reduction and/or further oxidation to methanesulfinic acid) may also be enhanced.  It is thus unclear whether DMSO concentrations should increase or decrease during daylight hours.  Moreover, DMSO yields from photolysis vary substantially (from 10 – 99%)   79 (Hatton, 2002; Kieber et al., 1996; Toole et al., 2004) across oceanographic regions, complicating the relationship between DMSO production and UV dose.  Previous studies documenting diel changes in DMS concentrations and cycling have shown a similar interplay of counteracting factors: UV exposure may stimulate DMS production and inhibit bacterial consumption, while also increasing photochemical oxidation of DMS (Royer et al., 2016; Yang et al., 2013).  Imbalances in these loss and production terms can thus drive diel concentration changes, although the pattern varies depending on environmental and biological conditions.   The Lagrangian survey used in this study minimized the confounding influence of spatial variability, allowing us to better understand the importance of temporal and light-dependent effects on DMSO cycling in the NESAP.  We observed clear diel cycling of DMSO during our drifter deployment, with lower concentrations of DMSO observed during mid-day periods of high PAR, when low photosynthetic efficiency (decreased Fv/Fm) and increased non-photochemical quenching (NPQ) indicated photo-inhibitory responses of phytoplankton in surface waters.   The mid-day decrease in DMSO concentrations appeared to become amplified over the course of our survey, in conjunction with decreasing minima in Fv/Fm.  In contrast, only slight evidence of diel effects on DMS concentrations were observed during the Lagrangian survey, with some decrease in concentration apparent during daylight hours during days 3 and 4 of our diel sampling.   It has been shown that DMSO reduction (and subsequent production of DMS) in phytoplankton can increase under conditions of oxidative stress (Spiese and Tatarkov, 2014).  Accordingly, UV-induced stress may elevate DMSO reduction rates, and this could provide an explanation for our observed daytime decrease in DMSO levels.  However, diel cycling of these rates was not apparent in our study, and no concurrent daytime increase in DMS was observed.  Likewise, we observed no evidence of direct DMS oxidation that would explain the changes   80 observed in DMSO concentrations.  While photochemical oxidation has been demonstrated in the laboratory (Bardouki et al., 2002; Barnes et al., 2006), previous studies have shown no evidence of DMSO photolysis in seawater (Toole et al., 2004), and we did not directly measure this process.   Biological oxidation of DMSO is a further possibility to explain the daytime decrease in DMSO concentrations.  Given the high membrane-permeability of DMSO, this compound may be actively or passively taken up by phytoplankton (or produced intracellularly through DMS oxidation) and oxidized to methanesulfinic acid through scavenging of ROS.  This process may serve to mitigate photo-inhibitory stress and would remove DMSO from the water column without a concomitant increase in DMS.  Although biological oxidation processes also affect DMS, a relative lack of daytime DMS loss may be accounted for by the compensating effects of DMSO reduction.  Indeed, rates of DMS oxidation and DMSO reduction were extremely well-coupled during this survey (r=0.99, p=0.011, n=4).    Several lines of evidence suggest a biological regulation of DMSO uptake / oxidation that is coupled to a photo-physiological response.  High irradiance levels lead to a build-up of excess energy in phytoplankton photosystem II (PSII) and proton accumulation in the intra-thylakoid space.  The presence of this strong pH gradient enhances potentially damaging intracellular production of ROS, and also triggers non-photochemical quenching (NPQ) processes, which serve to thermally dissipate excess energy.  Under these conditions, PSII quantum efficiency is decreased by NPQ processes and ROS-induced photodamage (Demmig-Adams et al., 2014; Pospíšil, 2016).  Thus, measurements of NPQ and PSII efficiency (Fv/Fm) serve as indicators of oxidative stress within cells (Pintó-Marijuan and Munné-Bosch, 2014).  During the drifter deployment, maxima in NPQ and minima in Fv/Fm occurred during mid-late afternoon, coincident with the observed decrease in DMSO concentrations.  Both Fv/Fm and NPQ showed statistically significant correlations with DMSO (r = 0.43, p<0.001, and r = -0.65, p<0.001,   81 respectively; Fig 3.9).  To our knowledge, this is a novel result, potentially linking DMSO cycling to cellular photo-physiology.  The role of cellular ROS-scavenging by DMSO has been previously proposed to explain variation in DMSPp:DMSOp ratios in phytoplankton (Vila‐Costa et al., 2006), DMSO concentrations in coral endosymbionts (Gardner et al., 2017) and negative correlations between DMSO and chl-a (Lee et al., 1999).  Our results provide the first direct set of observations linking DMSO concentrations with several proxies of oxidative stress.  Further laboratory studies will be needed to explore this potential relationship more closely.  3.5 Conclusion and outlook Our work highlights the significant contribution of DMSO to DMS production in the NESAP, and its potential physiological importance in phytoplankton assemblages.  We provide new high-resolution DMSO measurements across the NESAP, highlighting variations in DMSO:DMS ratios across this region.  Our results suggest that DMSO concentrations and cycling processes may be influenced by phytoplankton taxonomy due to the interacting roles of DMSO and DMSP in cellular metabolism.  We also demonstrate high rates of DMSO reduction across a number of distinct hydrographic regions, and provide the first direct comparison of DMS oxidation and DMSO reduction.  These findings suggest that net DMS production rates from DMSO are comparable to those of DMSP cleavage.  Further, our Lagrangian studies demonstrate that DMSO is actively cycled more rapidly than previously assumed, and may play an important physiological role among phytoplankton as an antioxidant under conditions of potential UV and light stress. Additional high spatial resolution concentration measurements of DMSO are needed to better characterize spatial and temporal variability of this compound.  Further, additional studies comparing rates of DMS oxidation and DMSO reduction across various ecosystems are needed   82 to determine whether DMSO consistently acts as a net source of DMS, as suggested by this work.  If, indeed, DMSO proves to be a significant source of DMS, improved understanding of the role of DMSO reduction, and examination of the relative contribution of phytoplankton and bacteria to this process, will be important in building models predicting DMS concentrations.  Likewise, direct measurements of DMSO production from phytoplankton are needed to confirm a potential algal source, and to quantify the contribution of DMSOP oxidation to the DMSO pool.  Finally, an exploration into the prevalence of DMSO reduction and oxidation processes among different phytoplankton communities under various environmental conditions will further serve to clarify the physiological role that this compound plays in the physiological ecology of marine primary producers.  3.6 Acknowledgements We wish to thank members of the Tortell, Kiene, and Levine laboratory groups for data contributions, insights, and assistance in the logistical aspects of the cruises presented here.  We would also like to thank the captain and crew of the R/V Oceanus and the CCGS John P. Tully, and the scientists from the Institute of Ocean Sciences.  Support for this work was provided from the US National Science Foundation (Grant #1436344), and from the Natural Sciences and Engineering Research Council of Canada.        83 Table 3.1 Correlation table for station based data.  *** denotes significance p≤0.001 **denotes significance p≤0.01 *denotes significance p≤0.05   DMS DMSO DMSPd  DMSPt DMSPt:chl-a  Bacterial  productivity DMSO  reduction kDMSOreduction DMSP  cleavage KDMSPcleavage DMS  loss KDMSloss DMSO 0.96***            DMSPd 0.72*** 0.81***           DMSPt 0.78*** 0.74*** 0.81***          DMSPt:Chl-a 0.01 0.21 0.33 -0.16         Bacterial productivity 0.17 0.02 -0.31 -0.14 0.39        DMSO reduction 0.50** 0.36 -0.11 0.11 -0.31 0.96***       kDMSOreduction 0.04 -0.13 -0.18 0.20 -0.58** 0.62** 0.55**      DMSP cleavage 0.19 0.19 0.38 0.43 0.24 0.00 0.15 0.31     KDMSPcleavage 0.09 0.07 0.07 0.08 0.22 0.16 0.21 0.26 0.91***    DMS loss 0.41** 0.27 -0.04 0.29 -0.18 0.81*** 0.83*** 0.55** 0.39 0.33   KDMSloss -0.11 -0.35 0.04 0.31 0.10 0.16 0.22 0.39 0.62** 0.42* 0.69***       84 Table 3.2 Simultaneous measurements of DMSO reduction and DMS oxidation and effect to the DMS pool for 4 stations measured during the Lagrangian drifter survey.  Station DMSO reduction (nM/d) DMS oxidation (nM/d) Reduction – oxidation  (net gain to DMS pool; nM/d) L6 3.27 ± 1.15 2.77 ± 1.70 0.50 ± 2.05 L8 10.31 ± 1.57 9.56 ± 6.95 0.74 ± 7.12 L12 L14 10.47 ± 6.49 14.09 ± 1.43 3.29 ± 3.05 7.22 ± 1.97 7.18 ± 7.16 6.87 ± 2.43      85 Table 3.3 Comparison of data from current and previous studies in which in situ DMSO reduction was measured.   Location Dates Mean DMSO reduction rate (nM d-1) Mean DMSO reduction rate constant (d-1) Mean DMSP cleavage rate (nM d-1) Mean DMSP cleavage rate constant (d-1) Asher et al. 2011 Ross Sea, Antarctica (sea-ice brine)  Dec 2010 – Jan 2011 150 ± 22 Not reported 37 ± 6 Not reported Asher et al. 2011 Ross Sea, Antarctica (polynya waters)  Dec 2010 – Jan 2011 34 ± 16 Not reported 8.4 ± 1.1 Not reported Asher et al. 2017a Palmer Station, Antarctica (coastal waters)  Oct 2012 – Mar 2013 2.36 ± 1.61 1.6 ± 0.57 5.55 ± 3.28 2.7 ± 0.50 Asher et al. 2017b Northeast Subarctic Pacific  Aug – Sept 2011 5.67 ± 2.38 0.07 ± 0.05 1.55 ± 1.04 1.38  ± 0.14 Present study Northeast Subarctic Pacific May 2017; Aug 2017 12.60 ± 4.56 2.82 ± 0.34 2.60 ± 0.84 2.07 ± 0.36      86  Figure 3.1 Study area showing chl-a concentrations from AquaMODIS satellite and cruise tracks.  Data from the August 2017 cruise aboard the R/V Oceanus is shown in the main panel, while data from the May/June 2017 cruise is show in the inset.  Symbols represent the location of discrete sampling stations, while the box represents the area in which Lagrangian drifter sampling was conducted.  Grey lines represent the 100m and 2000m isobaths.     87  Figure 3.2 DMS concentrations (a), DMSO concentrations, and DMSO:DMS (b) during the May/June 2017 cruise.    88  Figure 3.3 DMS concentrations (a), DMSO concentrations, and DMSO:DMS (b) during the August 2017 cruise.  Note scale break in panel (a).  The grey patch labeled LG1 denotes the period of Lagrangian drifter survey.    89  Figure 3.4 Relationship between DMSO and DMS concentrations across the full survey region.  The curved line represents a two-term exponential fit to the data: [DMS]=21.3*e5.53E-5*[DMSO]-20.5*e-0.0414*[DMSO]; r2 = 0.84.  The dashed line shows a linear fit to the data (for values under 20nM): [DMS]=1.194*[DMSO]+0.6372; r2 =. 0.59. The grey dotted line represents a 1:1 relationship.    90  Figure 3.5 DMSP cleavage and DMSO reduction rate constants (a), rates (b), and DMSPd/DMSO concentrations for which both sets of data exist (c). Dotted line denotes separation between two cruises, with La Perouse date below the line.  Station identifiers are along the y-axis.  Black capped lines show error.     91  Figure 3.6 Spatial distribution of measured DMSO reduction rates. Squares denote location of experimental stations during May of 2017 and circles denote location of experimental station during August 2017. The red square denotes the location of the inset map.   92  Figure 3.7 Relationship between bacterial productivity and rate of DMSO reduction.     93  Figure 3.8 Detrended DMS/O concentrations and related variables measured during a Lagrangian drift survey conducted in Oregon coastal upwelling waters.  Panel (a) shows detrended DMSO concentrations, Fv/Fm, NPQ and DMSO reduction rate constants. Panel (b) shows DMS concentrations, DMS oxidation rate constants and PAR.  Panel (c) shows chl-a concentrations.  Shaded areas indicate night-time.  The short-duration dark period on 8/21 (thin grey bar) was a total solar eclipse.    94  Figure 3.9 Relationship between DMSO concentrations and non-photochemical quenching (NPQ) during the Lagrangian survey (r2=0.42).  Note that linearly detrended values of DMSO concentrations were used for this analysis.     95 Chapter 4: Conclusion  The aim of this thesis was to identify and constrain the environmental and biological processes driving DMS, DMSP and DMSO accumulation in the surface waters of the subarctic Northeast Pacific.  In Chapter 2, I examined the distribution and cycling of DMS at various spatial scales.  I combined high-resolution transects of DMS concentration with numerous other biological and environmental measurements to determine drivers of DMS distribution at sub-mesoscales across hydrographic frontal regions.  I used these new observations, along with more than 150,000 existing DMS concentration measurements to create an updated summertime climatology of DMS in the NESAP, providing context for these sub-mesoscale studies.  Results from this work demonstrate the importance of diatom-dominated assemblages to DMS accumulation in coastal upwelling regions, while re-affirming the importance of prymnesiophytes and other high DMSP-producers in driving high DMS concentrations in offshore waters.  Rate measurement experiments using isotope tracers highlight the importance of DMS consumption to distribution patterns across these transects, while strong correlations between DMS and other variables (i.e., SSHA and salinity) demonstrate the utility of these measures in predicting DMS concentration at a fine-scales.  At a regional scales, however, correlations between DMS and other oceanographic variables were weak, and we were unable to produce an empirical algorithm with good predictive performance.  Regardless, our updated climatology of summertime DMS concentrations and sea-air fluxes provides improved confidence in long-term average DMS distribution in the NESAP region, and further supports the importance of the NESAP as a global DMS ‘hot spot’.  In Chapter 3, I focused on the distribution, cycling, and potential physiological roles of DMSO.  I examined DMSO distributions at high-resolution across the NESAP, provided   96 measurements of DMSO reduction across distinct hydrographic regimes, and used data from Lagrangian drifter studies to examine diel variability in DMSO concentrations in response to solar irradiance cycles.  This work helps present a clearer picture of DMS/P/O dynamics in the NESAP region, and suggests a potential role for DMSO in a cellular anti-oxidant system.  The new high-resolution DMSO measurements provide much needed data to better characterize spatial and temporal variability of this compound across the NESAP region.  I further documented variations in DMSO:DMS ratios across different phytoplankton assemblages, with results suggesting that DMSO concentrations and cycling processes may be influenced by phytoplankton taxonomy due to the interacting roles of DMSO and DMSP in cellular metabolism.  I also demonstrated high DMSO reduction rates across distinct oceanographic regimes, and provided the first direct comparison of DMS oxidation and DMSO reduction.  These results show that net DMS production from DMSO is comparable to that from DMSP cleavage, challenging traditional assumptions of DMSP as the primary source of DMS.  Results from a Lagrangian approach suggest rapid biological cycling of DMSO over diel cycles, and provide new evidence supporting an antioxidant role for this molecule.    4.1 Future work The data presented in this thesis substantially add to existing measurements of DMS and DMSO in the NESAP, although spatial and temporal gaps remain.  Our automated methods for DMS (MIMS) and DMS/P/O (OSSCAR) systems provide the opportunity for continued collection of high-resolution concentration data.  Improved characterization of spatial and temporal variability of these compounds, in conjunction with cycling rate measurements, will improve mechanistic understanding of the factors driving DMS accumulation, thereby improving model parameters and enhancing our understanding of the impact of DMS on regional and global   97 climate.  Future field studies should focus on high sulfur regions including the NESAP and Southern Ocean, with a particular focus on physically heterogenous coastal waters, where high levels of DMS are frequently observed.   Much additional work is needed to fully understand the role of DMSO both as a potential driver of DMS accumulation and as a physiologically important molecule.  Going forward, it will be important to compare rates of DMS oxidation and DMSO reduction across varying marine ecosystems in order to determine whether DMSO consistently acts as a net source of DMS, as suggested by the work presented in chapter 3 of this thesis.  Laboratory experiments may also be useful in determining the primary factors (e.g., UV) that influence relative magnitude of these two rates.  Additionally, the sources of DMSO in marine surface waters must be better constrained, and the potential production of this compound by phytoplankton confirmed to understand whether DMSO reduction constitutes a net source of DMS.   Research examining the relative contribution of phytoplankton and bacteria to total community DMSO reduction is needed to clarify the importance of these groups across differing marine systems.  This can be accomplished, in part, by the comparison of DMSO reduction rates among axenic cultures of bacteria and phytoplankton isolates.  Finally, more work is needed to determine the importance of DMSO reduction and oxidation processes under various environmental conditions, in order to clarify the physiological role of this compound in marine primary producers.  Changes in these rates under varying light and nutrient conditions should be examined in a laboratory environment.     98 References Alef, K. and Kleiner, D.: Rapid and sensitive determination of microbial activity in soils and in soil aggregates by dimethylsulfoxide reduction, Biol Fert Soils, 8(4), doi:10.1007/BF00263168 , 1989.   Asher, E. C., Dacey, J. W., Mills, M. M., Arrigo, K. R. and Tortell, P. D.: High concentrations and turnover rates of DMS, DMSP and DMSO in Antarctic sea ice, Geophys Res Lett, 38(23), doi:10.1029/2011GL049712 , 2011.  Asher, E. C., Merzouk, A., and Tortell, P. D.: Fine-scale spatial and temporal variability of surface water dimethylsufide (DMS) concentrations and sea–air fluxes in the NE Subarctic Pacific, Mar. Chem., 126, 63–75, doi:10.1016/j.marchem.2011.03.009, 2011.  Asher, E. C., Dacey, J. W., Jarniková, T. and Tortell, P. D.: Measurement of DMS, DMSO, and DMSP in natural waters by automated sequential chemical analysis, Limnol Oceanogr Methods, 13(9), 451–462, doi:10.1002/lom3.10039 , 2015.  Asher, E., Dacey, J. W., Ianson, D., Peña, A., and Tortell, P.D.: Concentrations and cycling of DMS, DMSP, and DMSO in coastal and offshore waters of the Subarctic Pacific during summer, 2010‐2011, J. Geophys. Res.-Oceans, 122(4), 3269–3286, doi:10.1002/2016JC012465, 2017a.  Asher, E. C., Dacey, J. W., Stukel, M., Long, M. C. and Tortell, P. D.: Processes driving seasonal variability in DMS, DMSP, and DMSO concentrations and turnover in coastal Antarctic waters, Limnol Oceanogr, 62(1), 104–124, doi:10.1002/lno.10379 , 2017b.  Aumont, O., Belviso, S., and Monfray, P.: Dimethylsulfoniopropionate (DMSP) and dimethylsulfide (DMS) sea surface distributions simulated from a global three‐dimensional ocean carbon cycle model, J. Geophys. Res.-Oceans, 107, 3029, doi:10.1029/1999JC000111, 2002.    99 Balch, W. M., Gordon, H. R., Bowler, B. C., Drapeau, D. T., and Booth, E. S.: Calcium carbonate measurements in the surface global ocean based on Moderate‐Resolution Imaging Spectroradiometer data, J. Geophys. Res.-Oceans, 110, C07001, doi:10.1029/2004JC002560, 2005.  Bardouki, H., Barcellos da Rosa, M. and Mihalopoulos, N.: Kinetics and mechanism of the oxidation of dimethylsulfoxide (DMSO) and methanesulfinate (MSI−) by OH radicals in aqueous medium, Atmos Environ, 36, 4627–4634, doi:10.1016/s1352-2310(02)00460-0 , 2002.  Barnes, I., Hjorth, J. and Mihalopoulos, N.: Dimethyl sulfide and dimethyl sulfoxide and their oxidation in the atmosphere, Chem Rev, 106, 940–975, doi:10.1021/cr020529+ , 2006.  Bates, T. S., Kiene, R. P., Wolfe, G. V., Matrai, P. A., Chavez, F. P., Buck, K. R., Blomquist, B. W. and Cuhel, R. L.: The cycling of sulfur in surface seawater of the northeast Pacific, J Geophys Res-Oceans, 99(C4), 7835–7843, doi:10.1029/93JC02782 , 1994.  de Bont, J. A., Van Dijken, J. P. and Harder, W.: Dimethyl sulphoxide and dimethyl sulphide as a carbon, sulphur and energy source for growth of Hyphomicrobium S, Microbiology, 127, 315–323, doi:10.1099/00221287-127-2-315 , 1981.  Booth, B., Lewin, J., and Postel, J.: Temporal variation in the structure of autotrophic and heterotrophic communities in the subarctic Pacific, Prog. Oceanogr., 32, 57–99, doi:10.1016/0079-6611(93)90009-3, 1993.  Bouillon, R.-C., Lee, P. A., de Mora, S. J., Levasseur, M. and Lovejoy, C.: Vernal distribution of dimethylsulphide, dimethylsulphoniopropionate, and dimethylsulphoxide in the North Water in 1998, Deep-Sea Res Pt II, 49(22–23), 5171–5189, doi:10.1016/S0967-0645(02)00184-4 , 2002.   Bouillon, R.-C., and Miller, W. L.: Determination of apparent quantum yield spectra of DMS photo-degradation in an in situ iron-induced Northeast Pacific Ocean bloom, Geophys. Res. Lett., 31, L06310, doi:10.1029/2004GL019536, 2004.    100 Boyd, P.J. and Harrison, P.J.: Phytoplankton dynamics in the NE subarctic Pacific, Deep-sea Res. Pt. I., 46, 2405–2432, doi:10.1016/S0967-0645(99)00069-7, 1999.  Boyd, P. W., Law, C. S., Wong, C. S., Nojiri, Y., Tsuda, A., Levasseur, M., Takeda, S., Rivkin, R., Harrison, P. J., Strzepek, R., Gower, J., McKay, M., Abraham, E., Arychuk, M., Barwell-Clarke, J., Crawford, W., Crawford, D., Hale, M., Harada, K., Johnson, K., Kiyosawa, H., Kudo, I., Marchetti, A., Miller, W., Needoba, J., Nishioka, J., Ogawa, H., Page, J., Robert, M., Saito, H., Sastri, A., Sherry, N., Soutar, T., Sutherland, N., Taira, Y., Whitney, F., Wong, S.-K. E., and Yoshimura, T.: The decline and fate of an iron-induced subarctic phytoplankton bloom, Nature 428, 549–553, doi:10.1038/nature02412, 2004.  Brewin, R., Sathyendranath, S., Hirata, T., Lavender, S., Barciela, R., and Hardman-Mountford, N.: A three-component model of phytoplankton size class for the Atlantic Ocean, Ecol. Model., 221, 1472–1483, doi:10.1016/j.ecolmodel.2010.02.014, 2010.  Bricaud, A., Babin, M., Morel, A. and Claustre, H.: Variability in the chlorophyll‐specific absorption coefficients of natural phytoplankton: Analysis and parameterization, J. Geophys. Res.-Oceans, 100, 13321–13332, doi:10.1029/95JC00463, 1995.  Brimblecombe, P. and Shooter, D.: Photo-oxidation of dimethylsulphide in aqueous solution, Mar Chem, 19, 343–353, 1986.   Brinkley, J.: Quantification of dimethylsulfoxide in seawater through its reduction to dimethylsulfide: A method comparison (Master’s thesis), 2008.   Bucciarelli, E., and Sunda, W.: Influence of CO2, nitrate, phosphate, and silicate limitation on intracellular dimethylsulfoniopropionate in batch cultures of the coastal diatom Thalassiosira pseudonana, Limnol. Oceanogr., 48, 2256–2265, doi:10.4319/lo.2003.48.6.2256, 2003.  Bucciarelli, E., Ridame, C., Sunda, W., Dimier-Hugueney, C., Cheize, M., and Belviso, S.: Increased intracellular concentrations of DMSP and DMSO in iron‐limited oceanic   101 phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum, Limnol. Oceanogr., 58, 1667–1679, doi:10.4319/lo.2013.58.5.1667, 2013.  Bullock, H. A., Luo, H. and Whitman, W. B.: Evolution of Dimethylsulfoniopropionate Metabolism in Marine Phytoplankton and Bacteria, Front Microbiol, 8, doi:10.3389/fmicb.2017.00637 , 2017.   Burt, W., Westberry, T., Behrenfeld, M., Zeng, C., Izett, R., Tortell, P.: Carbon: Chlorophyll Ratios and Net Primary Productivity of Subarctic Pacific Surface Waters Derived From Autonomous Shipboard Sensors, Global Biogeochem. Cy., 32, 267–288, doi:10.1002/2017GB005783, 2018.  Caruana, A. and Malin, G.: The variability in DMSP content and DMSP lyase activity in marine dinoflagellates, Prog Oceanogr, 120, 410–424, doi:10.1016/j.pocean.2013.10.014 , 2014.   Challenger, F. and Simpson, M.: 320. Studies on biological methylation. Part XII. A precursor of the dimethyl sulphide evolved by Polysiphonia fastigiata. dimethyl-2-carboxyethylsulphonium hydroxide and its salts, J Chem Soc, 1591–1597, doi:10.1039/JR9480001591 , 1948.   Charlson, R. J., Lovelock, J. E., Andreae, M. O. and Warren, S. G.: Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate, Nature, 326, 655–661, doi:10.1038/326655a0, 1987.  Clainche, Y. L., Levasseur, M., Vézina, A., Dacey, J. W. H., and Saucier, F. J.: Behaviour of the ocean DMS (P) pools in the Sargasso Sea viewed in a coupled physical-biogeochemical ocean model, Can. J. Fish. Aquat. Sci., 61, 788–803, doi:10.1139/F04-027, 2004.  Crawford, W., Brickley, P., Peterson, T., and Thomas, A.: Impact of Haida Eddies on chlorophyll distribution in the Eastern Gulf of Alaska, Deep-Sea Res. Pt. II, 52, 975–989, doi:10.1016/j.dsr2.2005.02.011, 2005.    102 Curson, A., Williams, B., Pinchbeck, B., Sims, L., Martínez, A., Rivera, P., Kumaresan, D., Mercadé E., Spurgin, L., Carrión O., Moxon S., Cattolico, R., Kuzhiumparambil, U., Guagliardo, P., Clode, P., Raina, J.-B., and Todd, J.: DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton, Nat. Microbiol., 3, 430–439, doi:10.1038/s41564-018-0119-5, 2018.  Dacey, J. W. and Wakeham, S. G.: Oceanic Dimethylsulfide: Production During Zooplankton Grazing on Phytoplankton, Science, 233(4770), 1314–1316, doi:10.1126/science.233.4770.1314, 1986.  Dacey, J. W., and Blough, N. V.: Hydroxide decomposition of dimethylsulfoniopropionate to form dimethylsulfide, Geophys. Res. Lett., 14, 1246–1249, doi:10.1029/GL014i012p01246, 1987.  Demmig-Adams, B., Garab, G., Adams, W. and Govindjee: Non-photochemical quenching and energy dissipation in plants, algae and cyanobacteria, Springer, 2014.   Dickson, D. M. and Kirst, G. O.: Osmotic adjustment in marine eukaryotic algae: the role of inorganic ions, quaternary ammonium, tertiary sulphonium and carbohydrate solutes, New Phytol, 106, 645–655, doi:10.1111/j.1469-8137.1987.tb00165.x , 1987.   Franklin, D. J., Poulton, A. J., Steinke, M., Young, J., Peeken, I., and Malin, G.: Dimethylsulphide, DMSP-lyase activity and microplankton community structure inside and outside of the Mauritanian upwelling, Prog. Oceanogr., 83, 134–142, doi:10.1016/j.pocean.2009.07.011, 2009.  Fredrickson, K. A. and Strom, S. L.: The algal osmolyte DMSP as a microzooplankton grazing deterrent in laboratory and field studies, J of Plankton Res, 31(2), 135–152, doi:10.1093/plankt/fbn112 , 2009.   Fuse, H., Takimura, O., Kamimura, K., Murakami, K., Yamaoka, Y. and Murooka, Y.:   103 Transformation of Dimethyl Sulfide and Related Compounds by Cultures and Cell Extracts of Marine Phytoplankton, Biosci Biotechnology Biochem, 59(9), 1773–1775, doi:10.1271/bbb.59.1773 , 1995.   Galí, M., Levasseur, M., Devred, E., Simó, R., and Babin, M.: Sea-surface dimethylsulfide (DMS) concentration from satellite data at global and regional scales, Biogeosciences, 15, 3497–3519, doi:10.5194/bg-2018-18, 2018.  Gardner, S. G., Raina, J.-B., Ralph, P. J. and Petrou, K.: Reactive oxygen species (ROS) and dimethylated sulphur compounds in coral explants under acute thermal stress, J Exp Biol, 220(10), doi:10.1242/jeb.153049 , 2017.   Gondwe, M., Krol, M., Gieskes, W., Klaasen, W. and de Baar, H.: The contribution of ocean‐leaving DMS to the global atmospheric burdens of DMS, MSA, SO2, and NSS SO4=, Global Biogeochem Cy, 17(2), doi:10.1029/2002GB001937 , 2003.  Glover, D., Jenkins, W. and Doney, S.: Modeling methods for marine science, 2011.   Goes, J., Saino, T., Oaku, H., Ishizaka, J., Wong, C., and Nojiri, Y: Basin scale estimates of sea surface nitrate and new production from remotely sensed sea surface temperature and chlorophyll, Geophys. Res. Lett., 27, 1263–1266, doi:10.1029/1999GL002353, 2000.  González, J., Kiene, R. and Moran, M.: Transformation of Sulfur Compounds by an Abundant Lineage of Marine Bacteria in the α-Subclass of the Class Proteobacteria, Appl Env Microb , 1999.   Gordon, H. R., Boynton, G. C., Balch, W. M., Groom, S. B., Harbour, D. S., and Smyth, T. J.: Retrieval of coccolithophore calcite concentration from SeaWiFS imagery, Geophys. Res. Lett., 28, 1587–1590, doi:10.1029/2000GL012025, 2001.    104 Gregr, E., and Bodtker, K.: Adaptive classification of marine ecosystems: Identifying biologically meaningful regions in the marine environment, Deep-Sea Res. Pt. I, 54, 385–402, doi:10.1016/j.dsr.2006.11.004, 2007.  Griebler, C. and Slezak, D.: Microbial Activity in Aquatic Environments Measured by Dimethyl Sulfoxide Reduction and Intercomparison with Commonly Used Methods, Appl and Env Microbb, 67(1), 100–109, doi:10.1128/AEM.67.1.100-109.2001 , 2001.  Harada, H., Vila-Costa, M., Cebrian, J., and Kiene, R.: Effects of UV radiation and nitrate limitation on the production of biogenic sulfur compounds by marine phytoplankton, Aquat. Bot., 90, 37–42, doi:10.1016/j.aquabot.2008.05.004, 2009.  Harris, S. L., Varela, D. E., Whitney, F. W. and Harrison, P. J.: Nutrient and phytoplankton dynamics off the west coast of Vancouver Island during the 1997/98 ENSO event, Deep-Sea Res Pt II, 56(24), 2487–2502, 2009.  Harrison, P. J., Boyd, P. W., Varela, D. E., Takeda, S., Shiomoto, A. and Odate, T.: Comparison of factors controlling phytoplankton productivity in the NE and NW subarctic Pacific gyres, Prog Oceanogr, 43, 205–234, doi:10.1016/S0079-6611(99)00015-4 , 1999.   Hatton, A. D., Malin, G., McEwan, A. G. and Liss, P. S.: Determination of Dimethyl Sulfoxide in aqueous solution by an enzyme-linked method, Anal Chem, 66(22), 4093–4096, doi:10.1021/ac00094a036 , 1994.  Hatton, A., Malin, G. and Turner: DMSO: A Significant Compound in the Biogeochemical Cycle of DMS, Plenum Press, 1996.  Hatton, A. D., Turner, S. M., Malin, G., and Liss, P. S.: Dimethylsulphoxide and other biogenic sulphur compounds in the Galapagos Plume, Deep-Sea Res. Pt. II, 45, 1043–1053, doi:10.1016/S0967-0645(98)00017-4, 1998.    105 Hatton, A. D., Malin, G. and Liss, P. S.: Distribution of biogenic sulphur compounds during and just after the southwest monsoon in the Arabian Sea, Deep-Sea Res Pt II, 46, 617–632, 1999.  Hatton, A.: Influence of photochemistry on the marine biogeochemical cycle of dimethylsulphide in the northern North Sea, Deep-Sea Res Pt II, 49, 3039–3052, 2002.   Hatton, A. D., Darroch, L. and Malin, G.: The role of dimethylsulphoxide in the marine biogeochemical cycle of dimethylsulphide, Oceanography and Marine Biology: An Annual Review, 42, 29–56, doi:10.1201/9780203507810 , 2004.  Hatton, A. D. and Wilson, S. T.: Particulate dimethylsulphoxide and dimethylsulphoniopropionate in phytoplankton cultures and Scottish coastal waters, Aquat Sci, 69, 330–340, doi:10.1007/s00027-007-0891-4 , 2007.   Hatton, A. D., Shenoy, D., Hart, M. C., Mogg, A. and Green, D. H.: Metabolism of DMSP, DMS and DMSO by the cultivable bacterial community associated with the DMSP-producing dinoflagellate Scrippsiella trochoidea, Biogeochemistry, 110(1–3), 131–146, doi:10.1007/s10533-012-9702-7 , 2012.   Hayashida, H., Steiner, N., Monohan, A., Galindo, V., Lizotte, M., and Levasseur, M.: Implications of sea-ice biogeochemistry for oceanic production and emissions of dimethyl sulfide in the Arctic, Biogeosciences, 14, 3129–3155, 2017.  Hill, R., White, Cottrell, M. and Dacey, J.: Virus-mediated total release of dimethylsulfoniopropionate from marine phytoplankton: a potential climate process, Aquat Microb Ecol, 14, 1–6, doi:10.3354/ame014001 , 1998.  Hirata, T., Aiken, J., Hardman-Mountford, N., Smyth, T. J., and Barlow, R. G.: An absorption model to determine phytoplankton size classes from satellite ocean colour, Remote Sens. Environ., 112, 3153–3159, doi:10.1016/j.rse.2008.03.011, 2008.    106 Hirata, T., Hardman-Mountford, N. J., Brewin, R. J., Aiken, J., Barlow, R., Suzuki, K., Isada, T., Howell, E., Hashioka, T., Noguchi-Aita, M., and Yamanaka, Y.: Synoptic relationships between surface Chlorophyll-a and diagnostic pigments specific to phytoplankton functional types, Biogeosciences, 8, 311–327, doi:10.5194/bg-8-311-2011, 2011.  Hockin, N. L., Mock, T., Mulholland, F., Kopriva, S., and Malin, G.: The response of diatom central carbon metabolism to nitrogen starvation is different from that of green algae and higher plants, Plant Physiol., 158, 299–312, doi:10.1104/pp.111.184333, 2012.  Izett, R. W., Manning, C. C., Hamme, R. C., and Tortell, P. D.: Refined Estimates of Net Community Production in the Subarctic Northeast Pacific Derived From ΔO2/Ar Measurements With N2O‐Based Corrections for Vertical Mixing, Global Biogeochem. Cy., 32, 326–350, doi:10.1002/2017GB005792, 2018.  Jarníková, T., Dacey, J., Lizotte, M., Levasseur, M., and Tortell, P.: The distribution of methylated sulfur compounds, DMS and DMSP, in Canadian subarctic and Arctic marine waters during summer 2015, Biogeosciences, 15, 2449–2465, doi:10.5194/bg-15-2449-2018, 2018.  Jian, S., Zhang, H., Zhang, J. and Yang, G.: Spatiotemporal distribution characteristics and environmental control factors of biogenic dimethylated sulfur compounds in the East China Sea during spring and autumn, Limnol Oceanogr, S280–S298, doi:10.1002/lno.10737 , 2017.   Johnson, K., Miller, L., Sutherland, N., and Wong, C. S.: Iron transport by mesoscale Haida eddies in the Gulf of Alaska, Deep-Sea Res. Pt. II, 52, 933–953, doi:10.1016/j.dsr2.2004.08.017, 2005.  Johnson, W., Soule, M., and Kujawinski, E.: Evidence for quorum sensing and differential metabolite production by a marine bacterium in response to DMSP, ISME J., 10, 2304–2316, doi:10.1038/ismej.2016.6, 2016.    107 Kaiser, J., Reuer, M. K., Barnett, B., and Bender, M. L.: Marine productivity estimates from continuous O2/Ar ratio measurements by membrane inlet mass spectrometry, Geophys. Res. Lett., 32, L19605, doi:10.1029/2005GL023459, 2005.  Karsten, U., Kück, K., Vogt, C. and Kirst, G.: Dimethylsulfoniopropionate production in phototrophic organisms and its physiological functions as a cryoprotectant, Springer, doi:10.1007/978-1-4613-0377-0_13 , 1996.  Keller, M. D.: Dimethyl Sulfide Production and Marine Phytoplankton: The Importance of Species Composition and Cell Size, Biol. Oceanogr., 6, 375–382, doi:10.1080/01965581.1988.10749540, 1989.  Kettle, A. J., Andreae, M. O., Amouroux, D., Andreae, T. W., Bates, T. S., Berresheim, H., Bingemer, H., Boniforti, R., Curran, M. A., DiTullio, G. R., Helas, G., Jones, G. B., Keller, M. D., Kiene, R. P., Leck, C., Levasseur, M., Malin, G., Maspero, M., Matrai, P., McTaggart, A. R., Mihalopoulos, N., Nguyen, B. C., Novo, A., Putaud, J. P., Rapsomanikis, S., Roberts, G., Schebeske, G., Sharma S., Simó, R., Staubes, R., Turner, S., and Uher, G.: A global database of sea surface dimethylsulfide (DMS) measurements and a procedure to predict sea surface DMS as a function of latitude, longitude, and month, Global Biogeochem. Cy., 13, 399–444, doi:10.1029/1999GB900004, 1999.  Kieber, D. J., Jiao, J., Kiene, R. P. and Bates, T. S.: Impact of dimethylsulfide photochemistry on methyl sulfur cycling in the equatorial Pacific Ocean, J of Geophys Res-Oceans, 101(C2), 3715–3722, doi:10.1029/95JC03624 , 1996.  Kiene, R., and Service, S.: Decomposition of dissolved DMSP and DMS in estuarine waters: dependence on temperature and substrate concentration., Mar. Ecol. Prog. Ser., 76, 1–11, doi:10.3354/meps076001, 1991.  Kiene, R. P. and Gerard, G.: Determination of trace levels of dimethylsulfoxide (DMSO) in seawater and rainwater, Mar Chem, 47(1), 1–12, doi:10.1016/0304-4203(94)90009-4 , 1994.    108 Kiene, R. P., and Linn, L. J.: Distribution and turnover of dissolved DMSP and its relationship with bacterial production and dimethylsulfide in the Gulf of Mexico, Limnol. Oceanogr., 45, 849–861, doi:10.4319/lo.2000.45.4.0849, 2000.  Kiene, R. P., Linn, L. J., and Bruton, J. A.: New and important roles for DMSP in marine microbial communities, J. Sea. Res., 43, 209–224, doi:10.1016/S1385-1101(00)00023-X, 2000.  Kiene, R. P., and Slezak, D.: Low dissolved DMSP concentrations in seawater revealed by small‐volume gravity filtration and dialysis sampling, Limnol. Oceanogr.-Meth., 4, 80–95, doi:10.4319/lom.2006.4.80, 2006.  Kiene, R. P., Kieber, D. J., Slezak, D., Toole, D. A., del Valle, D. A., Bisgrove, J., Brinkley, J. and Rellinger, A.: Distribution and cycling of dimethylsulfide, dimethylsulfoniopropionate, and dimethylsulfoxide during spring and early summer in the Southern Ocean south of New Zealand, Aquat Sci, 69(3), 305–319, doi:10.1007/s00027-007-0892-3 , 2007.   Kinsey, J., Kieber, D. and Neale, P.: Effects of iron limitation and UV radiation on Phaeocystis antarctica growth and dimethylsulfoniopropionate, dimethylsulfoxide and acrylate concentrations, Environ Chem, 13(2), 195–211, doi:10.1071/EN14275 , 2015.  Lam, P. J., and Letters, J.K.: The continental margin is a key source of iron to the HNLC North Pacific Ocean, Geophys. Res. Lett., 35, L07608, doi:10.1029/2008GL033294, 2008.  Lana, A., Bell, T. G., Simó, R., Vallina, S. M., Ballabrera‐Poy, J., Kettle, A. J., Dachs, J., Bopp, L., Saltzman, E. S., Stefels, J., Johnson, J. E.,  and Liss, P. S.: An updated climatology of surface dimethlysulfide concentrations and emission fluxes in the global ocean, Global Biogeochem. Cy., 25, GB1004, doi:10.1029/2010GB003850, 2011.  Lavoie, M., Levasseur, M. and Sunda, W. G.: A steady-state physiological model for intracellular dimethylsulfoxide in marine phytoplankton, Environ Chem, 13(2), 212–219, doi:10.1071/en14221 , 2015.   109  Leaitch, W.R., Sharma, S., Huang, L., Toom-Sauntry, D., Chivulescu, A., Macdonald, A.M., Salzen, K. von ., Pierce, J.R., Bertram, A.K., Schroder, J.C., Shantz, N.C., Chang, R.Y.-W. and Norman, A.-L.: Dimethyl sulfide control of the clean summertime Arctic aerosol and cloud, Elem Sci Anth, 1, p.000017, 2013  Lee, P., Haase, R., de Mora, S., Chanut, -P J and Gosselin, M.: Dimethylsulfoxide (DMSO) and related sulfur compounds in the Saguenay Fjord, Quebec, Canadian Journal of Fisheries and Aquat Sci, 56(9), 1631–1638, doi:10.1139/f99-094 , 1999.  Lee, P. A. and de Mora, S. J.: DMSP, DMS and DMSO Concentrations and Temporal Trends in Marine Surface Waters at Leigh, New Zealand, in Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds, pp. 391–404, Plenum Press., 1996.  Lee, P. A. and de Mora, S. J.: Intracellular dimethylsulfpoxide (DMSO) in unicellular marine algae: speculations on its origin and possible biological role, Journal of Phycology, 35(1), 8–18, doi:10.1046/j.1529-8817.1999.3510008.x , 1999.  Levasseur, M., Michaud, S., Egge, J., Cantin, G., Nejstgaard, J. C., Sanders, R., Fernandez, E., Solberg, P. T., Heimdal, B., and Gosselin, M.: Production of DMSP and DMS during a mesocosm study of an Emiliania huxleyi bloom: influence of bacteria and Calanus finmarchicus grazing, Mar. Biol., 126, 609–618, doi:10.1007/BF00351328, 1996.  Levasseur, M., Scarratt, M. G., Michaud, S., Merzouk, A., Wong, C., Arychuk, M., Richardson, W., Rivkin, R. B., Hale, M., Wong, E., Marchetti, A., and Kiyosawa, H.: DMSP and DMS dynamics during a mesoscale iron fertilization experiment in the Northeast Pacific—Part I: Temporal and vertical distributions, Deep-Sea Res., Pt. II 53, 2353–2369, doi:10.1016/j.dsr2.2006.05.023, 2006.  Levine, N. M., Toole, D. A., Neeley, A., Bates, N. R., Doney, S. C. and Dacey, J. W.: Revising upper-ocean sulfur dynamics near Bermuda: new lessons from 3 years of concentration and rate   110 measurements, Environ Chem, 13(2), 302–313, doi:10.1071/EN15045 , 2015.   Li, H., Liu, Z., Xu, J., Wu, X., Chaohui, S., Lu, S., and Minjie, C.: Manual of Global Ocean Argo gridded data set (BOA-Argo) (Version 2017), 2017.  Locarnini, S., Turner, S. and Liss, P.: The distribution of dimethylsulfide, DMS, and dimethylsulfoniopropionate, DSMP, in waters off the Western Coast of Ireland, Cont. Shelf Res., 18, 1455–1473, doi:10.1016/S0278-4343(98)00035-1, 1998.  Longhurst, A. R.: Ecological Geography of the Sea, 2nd ed. Elsevier, doi:10.1016/B978-0-12-455521-1.X5000-1, 2007.  Lovelock, J., Maggs, R., and Rasmussen, R.: Atmospheric dimethyl sulphide and the natural sulphur cycle, Nature, 237, 452–453, doi:10.1038/237452a0. 1972.  Lyon, B. R., Lee, P., Janech, M. and DiTullio, G.: Role of dimethylsulfoniopropionate as an osmoprotectant following gradual salinity shifts in the sea-ice diatom Fragilariopsis cylindrus, Environ Chem, 13, 181–194, doi:10.1071/EN14269 , 2016.  Matrai, P. A. and Keller, M. D.: Dimethylsulfide in a large-scale coccolithophore bloom in the Gulf of Maine, Cont Shelf Res, 13(8–9), 831–843, doi:10.1016/0278-4343(93)90012-m , 1993.  McCoy, D. T., Burrows, S. M., Wood, R., Grosvenor, D. P., Elliott, S. M., Ma, P.-L., Rasch, P. J. and Hartmann, D. L.: Natural aerosols explain seasonal and spatial patterns of Southern Ocean cloud albedo, Science Advances, 1(6), e1500157, doi:10.1126/sciadv.1500157 , 2015.  McParland, E. and Levine, N.: The role of differential DMSP regulation and community composition in predicting variability of global surface DMSP concentrations, Limnol. Oceanogr., in revision.  Merzouk, A., Levasseur, M., Scarratt, M. G., Michaud, S., Rivkin, R. B., Hale, M. S., Kiene, R. P., Price, N. M., and Li, W.: DMSP and DMS dynamics during a mesoscale iron fertilization   111 experiment in the Northeast Pacific–Part II: Biological cycling, Deep-Sea Res. Pt. II, 53, 2370–2383, doi:10.1016/j.dsr2.2006.05.022, 2006.  Narula, P. N.: The comparative penetrant‐carrier action of dimethyl sulfoxide and ethyl alcohol in vivo, Ann NY Acad Sci, 141(1), 277–278, doi:10.1111/j.1749-6632.1967.tb34890.x , 1967.  Nemcek, N., Ianson, D., and Tortell, P.: A high‐resolution survey of DMS, CO2, and O2/Ar distributions in productive coastal waters, Global Biogeochem. Cy., 22, GB2009, doi:10.1029/2006GB002879, 2008.  Pintó-Marijuan, M. and Munné-Bosch, S.: Photo-oxidative stress markers as a measure of abiotic stress-induced leaf senescence: advantages and limitations, J Exp Bot, 65(14), 3845–3857, doi:10.1093/jxb/eru086 , 2014.  Pospíšil, P.: Production of Reactive Oxygen Species by Photosystem II as a Response to Light and Temperature Stress, Front Plant Sci, 7, 1950, doi:10.3389/fpls.2016.01950 , 2016.  Quinn, P. and Bates, T.: The case against climate regulation via oceanic phytoplankton sulphur emissions, Nature, 480(7375), 51–56, doi:10.1038/nature10580 , 2011.  Reisch, C. R., Stoudemayer, M. J., Varaljay, V. A., Amster, I. J., Moran, M. A., and Whitman, W.: Novel pathway for assimilation of dimethylsulphoniopropionate widespread in marine bacteria, Nature, 473, 208–211, doi:10.1038/nature10078, 2011.  Reuer, M., Barnett, B., Bender, M., Falkowski, P., and Hendricks, M.: New estimates of Southern Ocean biological production rates from O2/Ar ratios and the triple isotope composition of O2, Deep-Sea Res. Pt. I, 54, 951–974, doi:10.1016/j.dsr.2007.02.007, 2007.  Reygondeau, G., Longhurst, A., Martinez, E., Beaugrand, G., Antoine, D., and Maury, O.: Dynamic biogeochemical provinces in the global ocean, Global Biogeochem. Cy., 27, 1046–1058, doi:10.1002/gbc.20089, 2013.    112 Ribalet, F., Marchetti, A., Hubbard, K., Brown, K., Durkin, C., Morales, R., Robert, M., Swalwell, J., Tortell, P., and Armbrust, V.: Unveiling a phytoplankton hotspot at a narrow boundary between coastal and offshore waters, P. Natl. A. Sci. USA, 107, 16571–16576, doi:10.1073/pnas.1005638107, 2010.  Riseman, S. and DiTullio, G.: Particulate dimethylsulfoniopropionate and dimethylsulfoxide in relation to iron availability and algal community structure in the Peru Upwelling System, Can J Fish Aquat Sci, 61, 721-735, 2004.  Roesler, C., and Barnard, A.: Optical proxy for phytoplankton biomass in the absence of photophysiology: Rethinking the absorption line height, Methods in Oceanography, 7, 79–94, doi:10.1016/j.mio.2013.12.003, 2013.  Roháček, K.: Chlorophyll fluorescence parameters: the definitions, photosynthetic meaning, and mutual relationships, Photosynthetica, 40(1), 13–29, doi:10.1023/A:1020125719386 , 2002.   Royer, S.-J., Levasseur, M., Lizotte, M., Arychuk, M., Scarratt, M., Wong, C., Lovejoy, C., Robert, M., Johnson, K., Peña, A., Michaud, S., and Kiene, R.:. Microbial dimethylsulfoniopropionate (DMSP) dynamics along a natural iron gradient in the northeast subarctic Pacific, Limnol. Oceanogr., 55, 1614–1626, doi:10.4319/lo.2010.55.4.1614, 2010.  Royer, S.-J., Galí, M., Mahajan, A., Ross, O., Pérez, G., Saltzman, E. and Simó, R.: A high-resolution time-depth view of dimethylsulphide cycling in the surface sea, Sci Rep-UK, 6(1), 32325, doi:10.1038/srep32325 , 2016.  Saari, A. and Martikainen, P. J.: Dimethyl sulphoxide (DMSO) and dimethyl sulphide (DMS) as inhibitors of methane oxidation in forest soil, Soil Biol and Biochem, 35(3), 383–389, 2003.  Saltzman, E. S., King, D. B., Holmen, K., and Leck, C.: Experimental determination of the diffusion coefficient of dimethylsulfide in water, J. Geophys. Res., 98, 16481–16468, doi:10.1029/93JC01858, 1993.    113 Saraceno, M., Strub, P. T., and Kosro, P. M.: Estimates of sea surface height and near‐surface alongshore coastal currents from combinations of altimeters and tide gauges, J. Geophys. Res.-Oceans, 113, C11013, doi:10.1029/2008JC004756, 2008.  Savoca, M.: Chemoattraction to dimethyl sulfide links the sulfur, iron, and carbon cycles in high-latitude oceans, Biogeochemistry, 138, 1–21, doi:10.1007/s10533-018-0433-2 , 2018.  Schuback, N., Schallenberg, C., Duckham, C., Maldonado, M. and Tortell, P.: Interacting Effects of Light and Iron Availability on the Coupling of Photosynthetic Electron Transport and CO2-Assimilation in Marine Phytoplankton, Plos One, 10: e0133235, doi:10.1371/journal.pone.0133235, 2015.  Sciare, J., Baboukas, E., Hancy, R. and of atmospheric …, M. N.: Seasonal variation of dimethylsulfoxide in rainwater at Amsterdam Island in the Southern Indian Ocean: Implications on the biogenic sulfur cycle, J Atmos Chem, 30, 229-240, doi:10.1023/A:1005983625947 , 1998.  Seymour, J. R., Simó, R., Ahmed, T., and Stocker, R.: Chemoattraction to Dimethylsulfoniopropionate Throughout the Marine Microbial Food Web, Science, 329, 342–345, doi:10.1126/science.1188418, 2010.  Simó, R., Grimalt, J. O. and Albaigés, J.: Sequential Method for the Field Determination of Nanomolar Concentrations of Dimethyl Sulfoxide in Natural Waters, Anal Chem, 68(9), 1493–1498, doi:10.1021/AC9510907 , 1996.  Simó, R., Grimalt, J. and Albaigés, J.: Dissolved dimethylsulphide, dimethylsulphoniopropionate and dimethylsulphoxide in western Mediterranean waters, Deep-Sea Res Pt II, 44(3-4), 929-950, 1997.  Simó, R., and Dachs, J.: Global ocean emission of dimethylsulfide predicted from biogeophysical data, Global Biogeochem. Cy., 16, 1078, doi:10.1029/2001GB001829, 2002.    114 Simó, R.: From cells to globe: approaching the dynamics of DMS (P) in the ocean at multiple scales, Can. J. Fish. Aquat. Sci., 61, 673–684, doi:10.1139/f04-030, 2004.  Simó, R. and Vila-Costa, M.: Ubiquity of algal dimethylsulfoxide in the surface ocean: Geographic and temporal distribution patterns, Mar Chem, 100(1–2), 136–146, doi:10.1016/j.marchem.2005.11.006 , 2006.  Sklorz, M. and Binert, J.: Determination of microbial activity in activated sewage sludge by dimethyl sulphoxide reduction, Environ Sci and Pollut R, 1(3), 140–145, doi:10.1007/BF02986936 , 1994.   Smith, D., and Azam, F.: A simple, economical method for measuring bacterial protein synthesis rates in seawater using 3H-leucine, Marine Microbial Food Webs, 6, 107–114, 1992.  Smith, R. L.: A description of current, wind, and sea level variations during coastal upwelling off the Oregon coast, July–August 1972, J. Geophys. Res., 79, doi:10.1029/JC079i003p00435, 1974.  Sparling, G. P. and Searle, P. L.: Dimethyl sulphoxide reduction as a sensitive indicator of microbial activity in soil: The relationship with microbial biomass and mineralization of nitrogen and sulphur, Soil Biol Biochem, 25(2), 251–256, doi:10.1016/0038-0717(93)90035-A , 1993.   Speeckaert, G., Borges, A. V., Champenois, W., Royer, C. and Gypens, N.: Annual cycle of dimethylsulfoniopropionate (DMSP) and dimethylsulfoxide (DMSO) related to phytoplankton succession in the Southern North Sea, Sci Total Environ, 622, 362–372, doi:10.1016/j.scitotenv.2017.11.359 , 2018.  Spiese, C. E., Kieber, D. J., Nomura, C. T. and Kiene, R. P.: Reduction of dimethylsulfoxide to dimethylsulfide by marine phytoplankton, Limnol Oceanogr, 54(2), 560–570, doi:10.4319/lo.2009.54.2.0560 , 2009.   Spiese, C. and Tatarkov, E.: Dimethylsulfoxide reduction activity is linked to nutrient stress in Thalassiosira pseudonana NCMA 1335, Mar Ecol Prog Ser, 507, 31-38, 2014.   115  Stefels, J.: Physiological aspects of the production and conversion of DMSP in marine algae and higher plants, J Sea Res, 43(3–4), 183–197, doi:10.1016/S1385-1101(00)00030-7 , 2000.  Stefels, J., Steinke, M., Turner, S., Malin, G., and Belviso, S.: Environmental constraints on the production and removal of the climatically active gas dimethylsulphide (DMS) and implications for ecosystem modelling, Biogeochemistry, 83, 245–275, doi:10.1007/s10533-007-9091-5, 2007.  Steiner, N., Robert, M., Arychuk, M., Levasseur, M., Merzouk, A., Peña, A., Richardson, W. and Tortell, P.: Evaluating DMS measurements and model results in the Northeast subarctic Pacific from 1996–2010, Biogeochemistry, 110, 269–285, doi:10.1007/s10533-011-9669-9, 2012.  Steinke, M., Malin, G., Archer, S. D., Burkill, P. H., and Liss, P. S.: DMS production in a coccolithophorid bloom: evidence for the importance of dinoflagellate DMSP lyases, Aquat. Microb. Ecol., 26, 259–270, doi:10.3354/ame026259, 2002.  Strub, T., and James, C.: The large-scale summer circulation of the California Current, Geophys. Res. Lett., 22, 207–210, doi:10.1029/94GL03011, 1995.  Sunda, W., Kieber, D. J., Kiene, R. P., and Huntsman, S.: An antioxidant function for DMSP and DMS in marine algae, Nature, 418, 317-320, doi:10.1038/nature00851, 2002.  Sunda, W., Hardison, R., Kiene, R., and Bucciarelli, E.: The effect of nitrogen limitation on cellular DMSP and DMS release in marine phytoplankton: climate feedback implications, Aquat. Sci., 69, 341–351, doi:10.1007/s00027-007-0887-0, 2007.  Suzuki,  K., Minami, C. , Liu, H., and Saino, T.: Temporal and spatial patterns of chemotaxonomic algal pigments in the subarctic Pacific and the Bering Sea during the early summer of 1999, Deep-Sea Res. Part II, 49, 5685–5704, doi:10.1016/s0967-0645(02)00218-7, 2002.    116 Sweeney, C., Gloor, E., Jacobson, A. R., Key, R. M., McKinley, G., Sarmiento, J. L., and Wanninkhof, R.: Constraining global air‐sea gas exchange for CO2 with recent bomb 14C measurements, Global Biogeochem. Cy., 21, GB2015, doi:10.1029/2006gb002784, 2007.  Tabata, S., Thomas, B. and Ramsden, D.: Annual and interannual variability of steric sea level along line P in the northeast Pacific Ocean, J. Phys. Oceanogr, 16, 1378–1398, doi:10.1175/1520-0485(1986)016<1378:AAIVOS>2.0.CO;2, 1986.  Tanaka, J. Y., Walsh, J. R., Diller, K. R., Brand, J. J. and Aggarwal, S. J.: Algae Permeability to Me2SO from −3 to 23°C, Cryobiology, 42(4), 286–300, doi:10.1006/cryo.2001.2334 , 2001.   Taylor, F. J. R. and Haigh, R.: Spatial and temporal distributions of microplankton during the summers of 1992–1993 in Barkley Sound, British Columbia, with emphasis on harmful species, Can J Fish Aquat Sci, 53(10), 2310–2322, doi:10.1139/cjfas-53-10-2310 , 1996.  Tesdal, J.-E., Christian, J., Monahan, A. and von Salzen, K.: Sensitivity of modelled sulfate radiative forcing to DMS concentration and air-sea flux formulation, Atmos Chem Phys Disc, 15(17), 23931–23968, doi:10.5194/acpd-15-23931-2015 , 2015.  Thume, K., Gebser, B., Chen, L., Meyer, N. and Nature, K. D.: The metabolite dimethylsulfoxonium propionate extends the marine organosulfur cycle, Nature Letter, doi:10.1038/s41586-018-0675-0 , 2018.  Toole, D., Kieber, D., Kiene, R., White, E., Bisgrove, J., del Valle, D. and Slezak, D.: High dimethylsulfide photolysis rates in nitrate‐rich Antarctic waters, Geophys Rese Lett, 31(11), doi:10.1029/2004GL019863 , 2004.   Tortell, P. D.: Dissolved gas measurements in oceanic waters made by membrane inlet mass spectrometry, Limnol. Oceanogr.-Meth., 3, 24–37, doi:10.4319/lom.2005.3.24, 2005a.  Tortell, P. D.: Small‐scale heterogeneity of dissolved gas concentrations in marine continental shelf waters, Geochem. Geophy. Geosy., 6, Q11M04, doi:10.1029/2005GC000953, 2005b.   117  Tyssebotn, I. B., Kinsey, J. D., Kieber, D. J., Kiene, R. P., Rellinger, A. N. and Motard‐Côté, J.: Concentrations, biological uptake, and respiration of dissolved acrylate and dimethylsulfoxide in the northern Gulf of Mexico, Limnol Oceanogr, doi:10.1002/lno.10495 , 2017.  Uitz, J., Claustre, H., Morel, A., and Hooker, S. B.: Vertical distribution of phytoplankton communities in open ocean: An assessment based on surface chlorophyll, J. Geophys. Res.-Oceans, 111, C08005, doi:10.1029/2005JC003207, 2006.  del Valle, D. A., Kieber, D. J. and Kiene, R. P.: Depth-dependent fate of biologically-consumed dimethylsulfide in the Sargasso Sea, Mar Chem, 103, 197–208, 2007a.  del Valle, D. A., Kieber, D. J., Bisgrove, J. B. and Kiene, R. P.: Light‐stimulated production of dissolved DMSO by a particle‐associated process in the Ross Sea, Antarctica, Limnol  Oceanogr, 52(6), 2456–2466, doi:10.4319/lo.2007.52.6.2456 , 2007b.   del Valle, D. A., Kieber, D. J., Toole, D. A., Brinkley, J., and Kiene, R. P.: Biological consumption of dimethylsulfide (DMS) and its importance in DMS dynamics in the Ross Sea, Antarctica, Limnol. Oceanogr., 54, 785–798, doi:10.4319/lo.2009.54.3.0785, 2009.  del Valle, D. A., Kieber, D. J., Toole, D. A., Bisgrove, J. and Kiene, R. P.: Dissolved DMSO production via biological and photochemical oxidation of dissolved DMS in the Ross Sea, Antarctica, Deep-Sea Res Pt I, 56, 166–177, 2009.   Vallina, S., and Simó, R.: Strong relationship between DMS and the solar radiation dose over the global surface ocean, Science 315, 506–508, doi:10.1126/science.1133680, 2007.  Venegas, R. M., Strub, P. T., Beier, E., Letelier, R., Thomas, A. C., Cowles, T., James, C., Soto-Mardones, L., and Cabrera, C.: Satellite‐derived variability in chlorophyll, wind stress, sea surface height, and temperature in the northern California Current System, J. Geophys. Res.-Oceans 113, C03015, doi:10.1029/2007JC004481, 2008.    118 Vézina, A.: Ecosystem modelling of the cycling of marine dimethylsulfide: a review of current approaches and of the potential for extrapolation to global scales, Can. J. Fish. Aquat. Sci., 61, 845–856, doi:10.1139/f04-025, 2004.  Vidussi, F., Claustre, H., Manca, B. B., Luchetta, A., and Marty, J.-C.: Phytoplankton pigment distribution in relation to upper thermocline circulation in the eastern Mediterranean Sea during winter, J. Geophys. Res.-Oceans, 106, doi:10.1029/1999JC000308, 19939–19956, 2001.  Vila‐Costa, M., Valle, D. A., González, J. M., Slezak, D., Kiene, R. P., Sánchez, O. and Simó, R.: Phylogenetic identification and metabolism of marine dimethylsulfide‐consuming bacteria, Environ Microbiol, 8(12), 2189–2200, doi:10.1111/j.1462-2920.2006.01102.x , 2006.  Vila-Costa, M., Kiene, R., and Simó, R.: Seasonal variability of the dynamics of dimethylated sulfur compounds in a coastal northwest Mediterranean site, Limnol. Oceanogr., 53, doi:10.4319/lo.2008.53.1.0198, 198–211, 2008.  Wang, S., Elliott, S., Maltrud, M., and Cameron-Smith, P.: Influence of explicit Phaeocystis parameterizations on the global distribution of marine dimethyl sulfide, J. of Geophys. Res.-Biogeo., 120, 2158–2177, doi:10.1002/2015JG003017, 2015. Watanabe, Y. W., Yoshinari, H., Sakamoto, A., Nakano, Y., Kasamatsu, N., Midorikawa, T. and Ono, T.: Reconstruction of sea surface dimethylsulfide in the North Pacific during 1970s to 2000s, Mar. Chem., 103, 347–358, doi:10.1016/j.marchem.2006.10.004, 2007. Werdell, J., and Bailey, S.: An improved in-situ bio-optical data set for ocean color algorithm development and satellite data product validation, Remote Sens. Environ., 98, 122–140, doi:10.1016/j.rse.2005.07.001, 2005.  Whitney, F. A., Crawford, W. R., and Harrison, P. J.: Physical processes that enhance nutrient transport and primary productivity in the coastal and open ocean of the subarctic NE Pacific, Deep-Sea Res. Pt. II, 52, 681–706, doi:10.1016/j.dsr2.2004.12.023, 2005.    119 Wolfe, G. V., and Kiene, R.P.: Radioisotope and chemical inhibitor measurements of dimethyl sulfide consumption rates and kinetics in estuarine waters, Mar. Ecol. Prog. Ser., 99, 261–269, doi:10.3354/meps099261, 1993.  Wolfe, G., Steinke and Nature, K. G.: Grazing-activated chemical defence in a unicellular marine alga, Nature, 387, 894-897, doi:10.1038/43168 , 1997.  Wolfe, G. V., Strom, S. L., Holmes, J. L., Radzio, T. and Olson, B. M.: Dimethylsulfiopropionate cleavage by marine phytoplankton in response to mechanical, chemical, or dark stress, J. Phycol., 38, 948–960, doi:10.1046/j.1529-8817.2002.t01-1-01100.x, 2002.  Wong, C., Wong, S., Richardson, W., Ith, G., Arychuk, M., and Page, J. Temporal and spatial distribution of dimethylsulfide in the subarctic northeast Pacific Ocean: a high‐nutrient–low‐chlorophyll region, Tellus 57B, 317–331, doi:10.1111/j.1600-0889.2005.00156.x, 2005.  Wu, X., Li, P., Liu, C., Zhang, H., Yang, G., Zhang, S., and Zhu, M.: Biogeochemistry of Dimethylsulfide, Dimethylsulfoniopropionate, and Acrylic Acid in the Changjiang Estuary and the East China Sea, J. Geophys. Res.-Oceans, 122, 10,245–10,261, doi:10.1002/2017JC013265, 2017.  Yang, M., Archer, S., Blomquist, B., Ho, D., Lance, V. and Torres, R.: Lagrangian evolution of DMS during the Southern Ocean gas exchange experiment: The effects of vertical mixing and biological community shift, J Geophys Res-Oceans, 118(12), 6774–6790, doi:10.1002/2013JC009329 , 2013.  Yoch, D. C.: Dimethylsulfoniopropionate: its sources, role in the marine food web, and biological degradation to dimethylsulfide, Appl. Environ. Microbiol., 68, 5804–5815, doi:10.1128/AEM.68.12.5804-5815.2002, 2002.  Zeng, C., Rosengard, S. Z., Burt, W., Peña, A., Nemcek, N., Zeng, T., Arrigo, K. R., and Tortell, P. D.: Optically-derived estimates of phytoplankton size class and taxonomic group biomass in   120 the Eastern Subarctic Pacific Ocean, Deep-Sea Res. Pt. I, 136, 107–118, doi:10.1016/j.dsr.2018.04.001, 2018.  Zhang, M. and Chen, L.: Continuous underway measurements of dimethyl sulfide in seawater by purge and trap gas chromatography coupled with pulsed flame photometric detection, Mar Chem, 174, 67–72, 2015.   Zinder, S. and Brock, T.: Dimethyl sulphoxide reduction by micro-organisms, J Gen Microb, 105, 335-342, 1978.  Zindler, C., Peeken, I., and Marandino, C.: Environmental control on the variability of DMS and DMSP in the Mauritanian upwelling region, Biogeosciences, 9, 1041–1051, doi:10.5194/bg-9-1041-2012, 2012.  Zubkov, M. V., Fuchs, B. M., Archer, S. D., and Kiene, R. P.: Rapid turnover of dissolved DMS and DMSP by defined bacterioplankton communities in the stratified euphotic zone of the North Sea, Deep-Sea Res. Pt. II, 49, 3017–3038, doi:10.1016/S0967-0645(02)00069-3, 2002.  Zubkov, M., Linn, L. J., Amann, R., and Kiene, R. P. Temporal patterns of biological dimethylsulfide (DMS) consumption during laboratory-induced phytoplankton bloom cycles, Mar. Ecol. Prog. Ser., 271, 77–86, doi:10.3354/meps271077, 2004.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items