Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The dynamic regulation of DMSP production in marine phytoplankton
(USC Thesis Other)
The dynamic regulation of DMSP production in marine phytoplankton
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
The dynamic regulation of DMSP production in marine phytoplankton Erin L. McParland A Dissertation Presented to the Faculty of the USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirement for the Degree DOCTOR OF PHILOSOPHY MARINE BIOLOGY AND BIOLOGICAL OCEANOGRAPHY AUGUST 2019 ! ! 1! Approved by Advisory Committee: Naomi M. Levine (Chair) Eric A. Webb David Hutchins Seth John ! ! 2! Table of Contents Table of Contents ............................................................................................................................2 Dedication .......................................................................................................................................3 Acknowledgements .........................................................................................................................4 Dissertation Abstract .......................................................................................................................6 Dissertation Introduction ................................................................................................................7 Introduction References ................................................................................................................14 Chapter One: The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations .....................................................21 Main text............................................................................................................................22 Supplementary information...............................................................................................39 Chapter Two: Evidence for differential regulation of dimethylsulfoniopropionate (DMSP) production by Emiliania huxleyi and Thalassiosira oceanica in contrasting growth stress conditions ......................................................................................................................................65 Abstract .............................................................................................................................65 Introduction .......................................................................................................................66 Methods .............................................................................................................................68 Results................................................................................................................................71 Discussion and Conclusion................................................................................................76 Acknowledgements............................................................................................................80 Figures and Tables ............................................................................................................81 Supplemental Figures and Tables......................................................................................88 Supplemental Note.............................................................................................................97 References .........................................................................................................................98 Chapter Three: Using DMSP synthesis genes as marker genes for high and low DMSP producers to predict in situ DMSP production ............................................................................................105 Abstract ...........................................................................................................................105 Introduction .....................................................................................................................106 Methods ...........................................................................................................................108 Results .............................................................................................................................114 Discussion and Conclusion .............................................................................................123 Acknowledgements..........................................................................................................127 Figures and Tables ..........................................................................................................128 Supplemental Figures and Tables....................................................................................137 References .......................................................................................................................142 Dissertation Conclusion ..............................................................................................................149 Conclusion References ................................................................................................................155 ! ! 3! Dedication This thesis is dedicated to Dr. Ronald P. Kiene, an amazing mentor and friend. Thank you for sharing with me your meticulous attention to detail in the laboratory and your love of science, fishing and baseball. ! ! 4! Acknowledgements My dissertation is a result of many years of hard work that was supported by a wonderful support base in my life. I first thank my advisor Naomi Levine. I could not have asked for a better role model both in and outside of the Allan Hancock building. She was a constant source of encouragement and advice, from which I learned how to actually ‘do’ science. She helped me become an expert on the DMS(P) cycle, while also reminding me to maintain a broader perspective of the ocean, a perspective with which I will build my future career on. Words cannot express how grateful I am for her time and support. I thank both my qualifying and dissertation committee members: Eric Webb, Dave Hutchins, Seth John, Dave Caron and Sergio Sañudo-Wilhemy. Thank you all for sharing your own knowledge base and perspectives which significantly contributed to my research and providing extensive professional guidance. Additionally, I thank all of the MEB and Earth Sciences faculty for informal support over the years, particularly Jim Moffett, Doug Capone, Suzanne Edmands, Feixue Fu, Karla Heidelberg, Will Berelson, Frank Corsetti and Doug Hammond. I would also like to thank my two previous mentors, Claudia Benitez-Nelson and Ron Kiene, without whom I likely would not have pursued this PhD. I would like to recognize the exceptional experience of being part of the Levine lab thanks to its amazing members (both past and present): Xiao Liu, Elizabeth Teel, Sara Rivero-Calle, Xuening Wen, Meagan He, Emily Zakem, Nathan Walworth, Noele Norris, Zaiss Bowman and Ryan Reynolds. All of my labmates were kind enough to listen to my presentations on DMSP and provided incredibly insightful advice. As well, I would not have been able to survive all of my labwork without the help of some outstanding undergraduate students: Anna Wright, Alexandra Koops, Emily Vainstein and Kristin Art. During my thesis, I also learned a lot from some wonderful co-authors, contributors and chief scientists: Stephanie Dutkiewicz, Oliver Jahn, Harriet Alexander, Kate Mackey, Justin Seymour, Bonnie Laverock, Mak Satio and Phil Tortelle. Endless hours in the laboratory, at sea and in front of the computer would not have been possible without the love of my family and friends. I first thank my (almost!) husband Pieter- Ewald Share. You were an endless fountain of support, advice, fun and love while working on your own PhD simultaneously. I also thank my parents and my sister for never questioning my abilities and supporting my love for the ocean since the beginning. My sweet grandparents, and wonderful aunts, uncles and cousins were also sources of support and fun times during this PhD. ! ! 5! I also greatly leaned on a great base of friends, both new and old, and thank them all for listening to stories about labwork and my “babies” (aka phytoplankton cultures). I particularly thank my peers who not only made graduate school more fun, but also significantly contributed to my professional development by sharing their own experiences: Joyce Yager, Kirstin Washington, Paige Connell, Sarah Hu, Chris Suffridge, Yubin Raut, Josh Kling, Laura Zinke, Mike Lee, Danie Monteverde, and Paulina Piñedo. Finally I would like to acknowledge my funding sources: The National Science and Engineering Graduate fellowship, the Gerald Bakus graduate fellowship, USC Dornsife, the Australian Research Council and the Rose Hills Foundation. ! ! 6! Dissertation Abstract Dimethylsulfoniopropionate (DMSP) is a labile sulfur and carbon metabolite that significantly contributes to both the cycling of marine dissolved organic carbon and the balance of Earth’s albedo. DMSP is produced by the majority of eukaryotic marine phytoplankton and by many prokaryotes, but despite decades of research, the cellular mechanism and environmental drivers of DMSP production remain unknown. My thesis confirms that the cellular mechanism of DMSP is differentiated by the cellular concentrations of DMSP in different producers, where high DMSP producers (e.g. dinoflagellates and haptophytes) constitutively produce DMSP and low DMSP producers (e.g. cyanobacteria and diatoms) actively regulate DMSP production in response to environmental stress. However, with natural community experiments and global model predictions, my thesis demonstrates that variability of in situ DMSP production is driven by the biomass of high producers. My thesis highlights the potential for predicting in situ DMSP concentrations with a high DMSP producer marker gene and demonstrates the importance of accurately capturing the sub-dominant community for prediction of DMSP, or other similar metabolites produced by a small fraction of the marine microbial community. Insight into the differential regulation of DMSP by HiDPs and LoDPs presented here should dramatically shift understanding of in situ DMSP cycling. Previous work assumes a universal mechanism of DMSP in all producers, but this thesis clearly demonstrates the importance of differential regulation across DMSP producer taxonomy which should be considered when resolving the significance of DMSP in carbon cycling and climate regulation. 7" Dissertation Introduction Motivation Oceans cover 70% of the Earth’s surface and play a critical role in regulating climate on Earth and provide vital resources of socio-economic wealth (IPCC 2013). During the new epoch of the Anthropocene, human activities have altered the physical and chemical landscape of the oceans and forced adaption of all marine organisms to these rapid changes, and will continue to do so without mitigation of human emissions (Gruber 2011; IPCC 2013, 2018). Microscopic marine organisms also directly impact Earth’s climate and understanding the impact of a rapidly changing ocean on these processes is critical (Cavicchioli et al. 2019). Though marine phytoplankton biomass only amounts to ~1% of that of global plants on land, these photosynthetic microorganisms are responsible for fixing half of global CO2 and producing half of global oxygen (Behrenfeld 2014). Before turning over on a relatively short timescale, marine phytoplankton consume and produce a suite of metabolites that play an active role in the function of the microbial ecosystem (Kujawinski 2011; Moran 2015; Moran et al. 2016), which are also subject to future ocean changes (Arandia-Gorostidi et al. 2017). Dimethylsulfide (DMS) and its algal precursor, dimethylsulfoniopropionate (DMSP), are organic sulfur metabolites that not only play an active role in marine microbial interactions and carbon cycling, but also have the potential to impact Earth’s climate (Charlson et al. 1987; Kiene et al. 2000; Lana et al. 2011; Johnson et al. 2016; Seymour et al. 2017). Understanding the regulation of DMSP production is critical for predicting how the DMS(P) cycle will change in the Anthropocene. Despite more than 30 years of research, the cellular role and environmental controls of DMSP have remained enigmatic. My thesis significantly improves understanding of DMSP production and strongly argues that the regulation of DMSP is controlled by two different mechanisms which are invoked at different scales and within different microbial communities. Introduction DMS(P) are produced and cycled by the microbial loop in the surface ocean. DMS is estimated to be a significant source of natural sulfur to the atmosphere contributing ~28.1 Tg of sulfur to the atmosphere annually (Lana et al. 2011; Mahajan et al. 2015). In 1987, the DMS(P) cycle was hypothesized to play an important role in regulating the Earth’s heat balance through a negative feedback loop. The ‘CLAW hypothesis’ proposed that increased sea surface temperatures 8" would increase DMS production by phytoplankton. Once ventilated to the atmosphere, DMS is oxidized into sulfate aerosols which serve as cloud condensation nuclei (CCN) promoting the formation of clouds and reflecting incoming solar radiation, thereby cooling sea surface temperatures (Charlson et al. 1987). Despite over 30 years of research, the potential for DMS(P) to act as a negative climate feedback loop is still unknown (Quinn and Bates 2011). However, DMS does significantly contribute to the balance of radiative budgets in atmospheric models, though the total contribution spans a range of 0.05 - 2 W·m -2 due to unconstrained atmospheric dynamics (Gabric et al. 2003; Gunson et al. 2006; Thomas et al. 2010; Woodhouse et al. 2013). Uncertainty in the mechanisms driving DMS(P) production, and therefore the magnitude of the response to changes in climate, as well as non-local effects of DMS derived CCN due to tropospheric entrainment into the marine boundary layer (Quinn et al. 2017) still allow for a potential significant effect of DMS on Earth’s radiative budget. DMSP, the primary precursor of DMS, is an important metabolite for many marine phototrophic species and is actively cycled through the organic carbon and sulfur cycles in the upper ocean (Kiene and Linn 2000; Ksionzek et al. 2016). DMSP can account for up to 11% of phytoplankton cellular carbon (Stefels et al. 2007) and, in both coastal and open ocean regimes, DMSP production can account for up to 5% of total primary production (Galí et al. 2013; Levine et al. 2015). DMSP is produced by a wide range of phototrophic taxa in the major eukaryotic supergroups rhizaria, chromalveolate, and archaeplastid, as well as marine cyanobacteria and heterotrophic bacteria (reviewed in McParland and Levine 2019). DMSP intracellular concentrations across these taxa vary by several orders of magnitude, ranging from <0.1 mM to 400 mM. Based on the seminal study of Keller et al. (1989), phytoplankton have conventionally been divided into two functional groups: high DMSP producers (HiDPs) which are typically dinoflagellates and haptophytes with intracellular concentrations >> 50 mM DMSP, and low DMSP producers (LoDPs) which are typically diatoms with intracellular concentrations <<50mM. The more recently described DMSP producing heterotrophic bacteria and phototrophic cyanobacteria typically have cellular DMSP concentrations that are an order of magnitude less than diatoms (Bucciarelli et al. 2013; Curson et al. 2017). While some producers, mostly HiDPs, contain lyases that directly cleave DMSP to DMS (Steinke et al. 1998; Caruana and Malin 2014), the majority of DMSP produced by phytoplankton is released into the dissolved phase via active exudation, cell lysis and grazing (Stefels et al. 2007). 9" Dissolved DMSP (DMSPd) is rapidly remineralized by the heterotrophic community where an estimated 60% of all surface ocean bacterioplankton are capable of degrading dissolved DMSP (Moran et al. 2012). DMSP consumption is estimated to supply 1 to 13% of the bacterial carbon demand and up to 100% of the sulfur demand (Kiene et al. 2000; Tripp et al. 2008). Bacteria degrade DMSP via one of two enzymatically mediated pathways. The demethylation pathway provides reduced carbon and sulfur for oxidation and is estimated to account for 80% of all DMSPd loss (Kiene et al. 2000). The DMSP cleavage pathway produces a 3-carbon compound (3- hydroypropionate or acrylate) and DMS, which is believed to diffuse out of the cell (Kiene et al. 2000; Moran et al. 2012). DMS can be degraded by a small group of heterotrophic ‘specialists’ (Simo 2004; Vila-Costa et al. 2006). Alternatively, DMS can be abiotically oxidized via photochemical reactions or ventilated to the atmosphere (Kieber et al. 1996). Together, all of these pathways result in only a small percentage (1-5%) of all phytoplankton-produced DMSP being ventilated to the atmosphere as DMS (Bates 1994; Simó et al. 2002). The DMS(P) cycle constitutes a significant recycling of organic carbon and sulfur in the global surface ocean and plays a role in regulating Earth’s albedo, but the cellular mechanism and environmental drivers of DMSP production remain unknown. DMSP synthesis The majority of DMSP producers are believed to synthesize DMSP from methionine through a transamination pathway, first proposed 20 years ago in a green macro-alga (Gage et al. 1997; Stefels 2000). One prokaryotic and two eukaryotic DMSP synthesis genes were very recently discovered and are functionally ratified (Curson et al. 2017, 2018; Kageyama et al. 2018). All three genes encode for a methyltransferase enzyme that catalyzes the third step of DMSP synthesis. While heterotrophic bacteria upregulate the transcription of dsyB and cellular DMSP concentrations under environmental stress, the genes encoding the first two steps of synthesis are not upregulated under the same stressors (Curson et al. 2017; Kageyama et al. 2018). The first two synthesis steps seem to be widespread in marine microbes, but only DMSP producers carry the gene encoding for the third step of DMSP synthesis (Curson et al. 2017). Interestingly, dsyB knock-out mutants in the same conditions did not exhibit changes in growth suggesting that DMSP synthesis is not a critical stress mechanism of these heterotrophic bacteria. dsyB prevalence is 10" primarily confined to the order Rhodobacterales and is found in ~0.5% of TARA metagenomes (Curson et al. 2017). The eukaryotic homologue of dsyB, DSYB, is monophyletic across eukaryotes. DMSP synthesis is proposed to have originated in prokaryotes and was then transferred to eukaryotes through endosymbiotic events or horizontal gene transfer (Curson et al. 2018). DSYB however was not present in all transcriptomes of known eukaryotic DMSP producers. TpMT2 is another eukaryotic DMSP synthesis gene with low homology with dsyB and was found in three specific diatoms that are missing DSYB (Kageyama et al. 2018). While DSYB transcription correlated with cellular concentrations of DMSP, DSYB transcription in salinity and nitrogen stress varied between species. Two haptophytes and a dinoflagellate containing DSYB did not increase transcription under N stress, but a diatom with DSYB did, though, transcription in these LoDPs is 8-fold lower than the HiDPs. TpMT2 transcription and cellular DMSP concentrations are also upregulated under nitrogen stress (Kageyama et al. 2018) Hypothesized functions of DMSP production Many hypotheses have been proposed as to the cellular physiology behind DMSP production including an osmolyte, a cryoprotectant, a signaling molecule, an overflow mechanism and an antioxidant (Karsten et al. 1996; Stefels and Leeuwe 1998; Stefels 2000; Sunda et al. 2002; Seymour et al. 2010). Experimental results are often conflicting, and DMSP likely plays multiple roles in the cell or different roles for different functional groups (Stefels et al. 2007; Archer et al. 2010; Bucciarelli et al. 2013). Below I briefly review the current state of knowledge (prior to my dissertation research) on the proposed cellular function for DMSP. Osmolyte and/or cryoprotectant: The biochemical structure of DMSP classifies the compound as an osmolyte (Stefels 2000). While DMSP does not appear to be upregulated quickly enough to counteract osmotic shocks, it may still serve as a compatible solute that is stored in preparation for changes of osmolarity and/or stabilizes proteins and enzymes during osmotic stress (Nishiguchi and Somero 1992; Karsten et al. 1996; Stefels 2000). Complementary to the hypothesis that the primary cellular function of DMSP is as an osmolyte is the hypothesis that, under nitrogen limitation, DMSP is substituted for high nitrogen content osmolytes such as glycine betaine (Andreae 1986). DMSP has also been shown to stabilize protein activity at low temperatures 11" (Karsten et al. 1996; Stefels 2000). However, the majority of this research was conducted with enzyme isolates and while DMSP may function as a cryoprotectant in polar ice algae, this is not the primary cellular function of DMSP for the majority of phytoplankton taxa that do not experience sub-zero temperatures (Stefels et al. 2012; Lyon et al. 2016). Signaling molecule: DMSP was proposed to behave as a grazing deterrent by Strom et al. (2003). This hypothesis was disproven with microfluidics experiments that demonstrated DMSP is actually a powerful chemoattractant for microzooplankton grazers, bacterial grazers and non- DMSP producing phytoplankton (Seymour et al. 2010). The elevated concentrations of DMSP confined to the phycosphere is key to these relationships (Seymour et al. 2017). The addition of high DMSP concentrations to bulk seawater in earlier studies likely obscured micro-scale processes by diffusion of the signaling molecule and subsequent masking of prey position, and therefore inadvertently decreased grazing. Recent studies suggest that DMSP serves as a signaling molecule for heterotrophic bacteria that could potentially maintain important phyto-heterotroph relationships (Seyedsayamdost et al. 2011; Johnson et al. 2016). Seyedsayamdost et al. (2011) proposed that DMSP mediates a “Jekyll and Hyde” relationship between the DMSP producer Emiliania huxleyi and an associated heterotroph. Specifically, the heterotroph uses DMSP as a carbon and sulfur source when E. huxleyi is growing exponentially but switches to using DMSP to synthesize roseobacticides (which promote algal death) when E. huxleyi reaches stationary phase. Additionally, when grown on DMSP as a sole carbon source, the heterotroph Ruegeria pomeroyi actively upregulated production of quorum sensing molecules and decreased intracellular glutamine, a phenotype indicating that DMSP signaled a phytoplankton bloom to the heterotroph (Johnson et al. 2016). Previous work investigating DMSP synthesis focuses on the cellular mechanism of the phototrophic producer alone. However, this recent evidence suggests that DMSP synthesis may be regulated in order to be traded with associated heterotrophs as an easily assimilated source of carbon and sulfur in return for heterotrophic byproducts (e.g. B12). Antioxidant and/or overflow mechanism: The two hypotheses for the physiological function of DMSP currently favored by the DMS(P) community are an overflow mechanism and an antioxidant. It was proposed that DMSP serves as an overflow mechanism by balancing intracellular sulfur and nitrogen flows during unbalanced growth conditions caused by 12" environmental stressors. By dissipating excess carbon and freeing nitrogen from methionine for synthesis of other amino acids, the cell is able to adapt its enzyme system to new stressful conditions while also dissipating excess sulfur and carbon (Karsten et al. 1996; Stefels 2000). The hypothesized overflow mechanism would require regulation of DMSP degradation or removal. As DMSP is a zwitterion, removal is dependent on either cleavage by the DMSP lyase (which has only been observed in high DMSP producers) or active transport, which would require a costly transporter protein (Stefels 2000; Archer et al. 2010). DMSP and its breakdown products, DMS, acrylate and dimethylsulfoxide (DMSO), have been shown to effectively scavenge free radicals leading to the hypothesis that DMSP functions as a part of an ‘antioxidant cascade’ to remove harmful reactive oxygen species (ROS) produced when a cell experiences environmental stress (Sunda et al. 2002). A recently discovered metabolite, dimethylsulfoxonium-propionate (DMSOP), may also serve as a sink for ROS by oxidation of DMSP to DMSOP (Thume et al. 2018). Despite the well-documented ability of these compounds to readily react with ROS, direct evidence for the physiological functioning of the DMSP antioxidant system in marine phytoplankton is generally lacking with inconsistent results from previous studies (e.g. Keller et al. 1999 vs. Bucciarelli et al. 2003). Environmental controls of DMSP production In situ DMSP production is decoupled from total chl-a in the majority of the ocean (Kettle et al. 1999). Instead, in situ DMSP concentrations are a function of the DMSP producer community abundance and the cellular DMSP concentrations of that community (Masotti et al. 2010). In high biomass regions where haptophytes and dinoflagellates are abundant and blooms of coccolithophores, DMSP concentrations are correlated to the biomass of these HiDPs (Turner et al. 1988; Malin et al. 1993; Scarratt et al. 2002; Speeckaert et al. 2018). However, DMSP concentrations throughout the oligotrophic ocean have proven more difficult to predict (Vogt et al. 2010; Belviso et al. 2012). Previous prognostic ecosystem models that include a DMSP module have predicted DMSP production with two phytoplankton types, a DMSP producer assigned a constant intracellular DMSP concentration and non-DMSP producer, though some incorporated an upregulation of cellular DMSP concentrations as a function of light or nutrient status (Archer et al. 2009; Le Clainche et al. 2010; Vogt et al. 2010; Polimene et al. 2011). However, these methods consistently misrepresent the seasonal dynamics and absolute concentrations of 13" particulate DMSP in the oligotrophic ocean (Masotti et al. 2016). A remote-sensing based estimate of in situ DMSP showed global DMSP patterns could only be reproduced using two different fits for low biomass regions and high biomass regions, where DMSP was predicted as a function of stratification stress for the low biomass regions (Galí et al. 2015). A complicating factor is that measurements of in situ DMSP are greatly biased towards higher biomass regions making predictions of DMSP for low biomass regions more difficult to validate. In the largest database of DMS(P) measurements (Kettle et al. 1999; Lana et al. 2012), there are ~100 DMSP data points reported between 20˚- 30 ˚N compared to >1300 data points reported between 50˚- 60˚N (Galí et al. 2015). Thesis motivation The goal of my thesis is to improve predictions of how DMSP supply to the microbial loop and to the atmosphere as DMS will change in the future ocean by addressing the overarching question of what are the environmental drivers of DMSP production, and at what scales do these drivers matter? In this dissertation, I combined culture and field experiments with ‘omic and numerical modeling techniques to quantify the different drivers of DMSP production at various scales. Specifically, this thesis demonstrates that abundance of HiDPs determines variability of in situ DMSP production (Chapter 1). This thesis also quantifies the differential regulation of DMSP production in HiDP and LoDPs, suggesting that at a cellular level the DMSP mechanism is also taxonomically dependent (Chapter 2). Finally, this thesis highlights the importance of accurately capturing sub-dominant communities, which are currently poorly estimated in global ecosystem models and satellite algorithms. This thesis demonstrates the potential for using HiDP marker genes to quantify HiDP abundance and predict in situ DMSP concentrations (Chapter 3). Ultimately, my thesis defined two different scales at which DMSP production should be considered and significantly moved the field’s understanding of DMSP drivers forward. This thesis is composed of three chapters: the first is published in Limnology and Oceanography, and the second and third are written as manuscripts that were submitted after the defense. Therefore, I refer to the work conducted in the chapters in the context of “we”. 14" Introduction References Andreae, M. 1986. The ocean as a source of atmospheric sulphur compounds. In The role of air- sea exchange in geochemical cycling. Edited by P. Buat-Merard. Reidel, New York. Pp. 331-362. Arandia-Gorostidi, N., P. K. Weber, L. Alonso-Sáez, X. A. G. Morán, and X. Mayali. 2017. Elevated temperature increases carbon and nitrogen fluxes between phytoplankton and heterotrophic bacteria through physical attachment. ISME J. 11: 641–650. doi:10.1038/ismej.2016.156 Archer, S. D., D. G. Cummings, C. a. Llewellyn, and J. R. Fishwick. 2009. Phytoplankton taxa, irradiance and nutrient availability determine the seasonal cycle of DMSP in temperate shelf seas. Mar. Ecol. Prog. Ser. 394: 111–124. doi:10.3354/meps08284 Archer, S. D., M. Ragni, R. Webster, R. L. Airs, and R. J. Geider. 2010. Dimethylsulfoniopropionate and dimethylsulfide production in response to photoinhibition in Emiliania huxleyi. Limnol. Oceanogr. 55: 1579–1589. doi:10.4319/lo.2010.55.4.1579 Bates, T. S. 1994. The cycling of sulfur in surface seawater of the northeast Pacific. J. Geophys. Res. 99: 7835–7843. doi:10.1029/93JC02782 Behrenfeld, M. 2014. Climate-mediated dance of the plankton. Nat. Clim. Chang. 4: 880–887. Belviso, S., I. Masotti, A. Tagliabue, and others. 2012. DMS dynamics in the most oligotrophic subtropical zones of the global ocean. Biogeochemistry 110: 215–241. doi:10.1007/s10533- 011-9648-1 Bucciarelli, E., C. Ridame, W. G. Sunda, C. Dimier-Hugueney, M. Cheize, and S. Belviso. 2013. Increased intracellular concentrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum. Limnol. Oceanogr. 58: 1667–1679. doi:10.4319/lo.2013.58.5.1667 Bucciarelli, E., W. G. Sunda, and P. Nicolas. 2003. Influence of CO2, nitrate, phosphate, and silicate limitation on intracellular DMSP of the coastal diatom Thalassiosira pseudonana in batch cultures. Limnol. Oceanogr. 48: 2256–2265. Caruana, A. M. N., and G. Malin. 2014. 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 Cavicchioli, R., W. J. Ripple, K. N. Timmis, and others. 2019. Scientists’ warning to humanity: 15" microorganisms and climate change. Nat. Rev. Microbiol. doi:10.1038/s41579-019-0222-5 Charlson, R., J. Lovelock, M. Andrae, and S. Warren. 1987. Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Nature 326: 655–661. Le Clainche, Y., A. Vezina, M. Levasseur, and others. 2010. A first appraisal of prognostic ocean DMS models and prospects for their use in climate models. Global Biogeochem. Cycles 24: 1–13. doi:10.1029/2009GB003721 Curson, A. R. J., J. Liu, A. Bermejo Martínez, and others. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nat. Microbiol. 2: 1–9. doi:10.1038/nmicrobiol.2017.9 Curson, A., B. Williams, B. Pinchbeck, and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat. Microbiol. 4: 430– 439. doi:10.1038/s41564-018-0119-5 Gabric, A. J., R. Cropp, T. Hirst, and H. Marchant. 2003. The sensitivity of dimethyl sulfide production to simulated climate change in the Eastern Antarctic Southern Ocean. Tellus B 55: 966–981. doi:10.1034/j.1600-0889.2003.00077.x Gage, D., D. Rhodes, K. Nolte, W. A. Hicks, T. Leustek, A. J. L. Cooper, and A. Hanson. 1997. A new route for synthesis of in marine algae. Nat. Lett. 387: 891–894. Galí, M., E. Devred, M. Levasseur, S. J. Royer, and M. Babin. 2015. A remote sensing algorithm for planktonic dimethylsulfoniopropionate (DMSP) and an analysis of global patterns. Remote Sens. Environ. 171: 171–184. doi:10.1016/j.rse.2015.10.012 Galí, M., R. Simó, G. L. Pérez, C. Ruiz-González, H. Sarmento, S. J. Royer, A. Fuentes-Lema, and J. M. Gasol. 2013. Differential response of planktonic primary, bacterial, and dimethylsulfide production rates to static vs. dynamic light exposure in upper mixed-layer summer sea waters. Biogeosciences 10: 7983–7998. doi:10.5194/bg-10-7983-2013 Gruber, N. 2011. Warming up, turning sour, losing breath: ocean biogeochemistry under global change. Philos. Trans. A. Math. Phys. Eng. Sci. 369: 1980–96. doi:10.1098/rsta.2011.0003 Gunson, J. R., S. A. Spall, T. R. Anderson, A. Jones, I. J. Totterdell, and M. J. Woodage. 2006. Climate sensitivity to ocean dimethylsulphide emissions. Geophys. Res. Lett. 33: 2–5. doi:10.1029/2005GL024982 IPCC. 2013. Climate Change 2013: The physical science basis. Contribution of working group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, T.F. 16" Stocker, D. Qin, G.-K. Plattner, et al. [eds.]. Cambridge University Press. IPCC. 2018. Summary for policymakers, In Masson-Delmotte, V.P. Zhai, H.-O. Pörtner, et al. [eds.], Global Warming of 1.5C. An IPCC Special Report on the impacts of global warming of 1.5C above pre-industrial levels and related global greenhouse gas emission pathways, in the cntext of strengthening the global response to the threat of climate change, su. World Meteorological Organization. Johnson, W. M., M. C. K. Soule, and E. B. Kujawinski. 2016. Evidence for quorum sensing and differential metabolite production by the marine heterotroph, Ruegeria pomeroyi, in response to DMSP. ISME J. 10: 2304–2316. doi:10.1038/ismej.2016.6 Kageyama, H., Y. Tanaka, A. Shibata, R. Waditee-Sirisattha, and T. Takabe. 2018. Dimethylsulfoniopropionate biosynthesis in a diatom Thalassiosira pseudonana: Identification of a gene encoding MTHB-methyltransferase. Arch. Biochem. Biophys. 645: 100–106. doi:10.1016/j.abb.2018.03.019 Karsten, U., K. Kück, C. Vogt, and G. O. Kirst. 1996. Dimethylsulfoniopropionate production in phototrophic organisms and its physiological functions as a cryoprotectant, p. 143–153. In Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds. Keller, M., W. Bellows, and R. Guillard. 1989. Dimethyl sulfide production in marine phytoplankton, p. 167–182. In Biogenic sulfur in the environment. Keller, M. D., R. P. Kiene, P. A. Matrai, and W. K. Bellows. 1999. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. I. Batch cultures. Mar. Biol. 135: 237–248. doi:10.1007/s002270050621 Kettle, A. J., M. O. Andreae, D. Amouroux, and others. 1999. 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. Cycles 13: 399–444. Kieber, D. J., J. Jiao, R. P. Kiene, and T. S. Bates. 1996. Impact of dimethylsulfide photochemistry on methyl sulfur cycling in the equatorial Pacific Ocean. J. Geophys. Res. Ocean. 101: 3715–3722. doi:10.1029/95JC03624 Kiene, R., and L. Linn. 2000. The fate of dissolved dimethylsulfoniopropionate ( DMSP ) in seawater": Tracer studies using 35 S-DMSP. Geochimi 64: 2797–2810. Kiene, R. P., L. J. Linn, and J. A. Bruton. 2000. New and important roles for DMSP in marine microbial communities. J. Sea Res. 43: 209–224. 17" Ksionzek, K. B., K. B. Ksionzek, O. J. Lechtenfeld, S. L. Mccallister, P. Schmitt-kopplin, J. K. Geuer, W. Geibert, and B. P. Koch. 2016. Dissolved organic sulfur in the ocean": Biogeochemistry of a petagram inventory. Science (80-. ). 354: 456–459. doi:10.1126/science.aaf7796 Kujawinski, E. B. 2011. The Impact of Microbial Metabolism on Marine Dissolved Organic Matter. Ann. Rev. Mar. Sci. 3: 567–599. doi:10.1146/annurev-marine-120308-081003 Lana, a., T. G. Bell, R. Simó, and others. 2011. An updated climatology of surface dimethlysulfide concentrations and emission fluxes in the global ocean. Global Biogeochem. Cycles 25: n/a-n/a. doi:10.1029/2010GB003850 Lana, A., R. Simó, S. M. Vallina, and J. Dachs. 2012. Re-examination of global emerging patterns of ocean DMS concentration. Biogeochemistry 110: 173–182. doi:10.1007/s10533- 011-9677-9 Levine, N. M., D. A. Toole, A. Neeley, N. R. Bates, S. C. Doney, and J. W. H. Dacey. 2015. Revising upper-ocean sulfur dynamics near Bermuda: New lessons from 3 years of concentration and rate measurements. Environ. Chem. 13: 302–313. doi:10.1071/EN15045 Lyon, B. R., J. M. Bennett-Mintz, P. A. Lee, M. G. Janech, and G. R. Ditullio. 2016. 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 Mahajan, A. S., S. Fadnavis, M. a Thomas, L. Pozzoli, S. Gupta, S. Royer, A. Saiz-lopez, and R. Simó. 2015. Quantifying the impacts ofanupdated global dimethyl sulfide climatology on cloud microphysics and aerosol radiative forcing. J. Geophys. Res. Atmos. 120: 2524–2536. doi:10.1002/2014JD022687.Received Malin, G., S. Turner, P. Liss, P. Holligan, and D. Harbour. 1993. Dimethylsulphide and dimethylsulphoniopropionate in the Northeast Atlantic during the summer coccolithophore bloom. Deep. Res. Part I 40: 1487–1508. doi:10.1016/0967-0637(93)90125-M Masotti, I., S. Belviso, S. Alvain, and others. 2010. Spatial and temporal variability of the dimethylsulfide to chlorophyll ratio in the surface ocean: An assessment based on phytoplankton group dominance determined from space. Biogeosciences 7: 3215–3237. doi:10.5194/bg-7-3215-2010 Masotti, I., S. Belviso, L. Bopp, A. Tagliabue, and E. Bucciarelli. 2016. Effects of light and 18" phosphorus on summer DMS dynamics in subtropical waters using a global ocean biogeochemical model. Environ. Chem. 13: 379–389. doi:10.1071/EN14265 McParland, E. L., and N. M. Levine. 2019. The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations. Limnol. Oceanogr. 64: 757–773. doi:10.1002/lno.11076 Moran, M. A. 2015. The global ocean microbiome. Science (80-. ). 350. doi:10.1126/science.aac8455 Moran, M. A., E. B. Kujawinski, A. Stubbins, and others. 2016. Deciphering ocean carbon in a changing world. Proc. Natl. Acad. Sci. 113: 3143–3151. doi:10.1073/pnas.1514645113 Moran, M. A., C. R. Reisch, R. P. Kiene, and W. B. Whitman. 2012. Genomic Insights into Bacterial DMSP Transformations. Ann. Rev. Mar. Sci. 4: 523–542. doi:10.1146/annurev- marine-120710-100827 Nishiguchi, M. K., and G. N. Somero. 1992. Temperature- and concentration-dependence of compatibility of the organic osmolyte β-dimethylsulfoniopropionate. Cryobiology 29: 118– 124. doi:10.1016/0011-2240(92)90011-P Polimene, L., S. D. Archer, M. Butenschön, and J. I. Allen. 2011. A mechanistic explanation of the Sargasso Sea DMS “summer paradox.” Biogeochemistry 110: 243–255. doi:10.1007/s10533-011-9674-z Quinn, P. K., D. J. Coffman, J. E. Johnson, L. M. Upchurch, and T. S. Bates. 2017. Small fraction of marine cloud condensation nuclei made up of sea spray aerosol. Nat. Geosci. 10: 674–679. doi:10.1038/ngeo3003 Scarratt, M. G., M. Levasseur, S. Michaud, G. Cantin, M. Gosselin, and S. J. De Mora. 2002. Influence of phytoplankton taxonomic profile on the distribution of dimethylsulfide and dimethylsulfoniopropionate in the northwest Atlantic. Mar. Ecol. Prog. Ser. 244: 49–61. doi:10.3354/meps244049 Seyedsayamdost, M. R., R. J. Case, R. Kolter, and J. Clardy. 2011. The Jekyll-and-Hyde chemistry of Phaeobacter gallaeciensis. Nat. Chem. 3: 331–5. doi:10.1038/nchem.1002 Seymour, J. R., S. A. Amin, J. B. Raina, and R. Stocker. 2017. Zooming in on the phycosphere: The ecological interface for phytoplankton-bacteria relationships. Nat. Microbiol. 2: 1–12. doi:10.1038/nmicrobiol.2017.65 Seymour, J. R., R. Simó, T. Ahmed, and R. Stocker. 2010. Chemoattraction to 19" Dimethylsulfoniopropionate throughout the Marine Microbial Food Web. Science (80-. ). 329: 342–345. doi:10.1126/science.1188418 Simo, R. 2004. 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 Simó, R., S. D. Archer, C. Pedrós-Alió, L. Gilpin, and C. E. Stelfox-Widdicombe. 2002. Coupled dynamics of dimethylsulfoniopropionate and dimethylsulfide cycling and the microbial food web in surface waters of the North Atlantic. Limnol. Oceanogr. 47: 53–61. doi:10.4319/lo.2002.47.1.0053 Speeckaert, G., A. V. Borges, W. Champenois, C. Royer, and N. Gypens. 2018. Annual cycle of dimethylsulfoniopropionate (DMSP) and dimethylsulfoxide (DMSO) related to phytoplankton succession in the Southern North Sea. Sci. Total Environ. 622–623: 362– 372. doi:10.1016/j.scitotenv.2017.11.359 Stefels, J. 2000. Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43: 183–197. Stefels, J., G. Carnat, J. W. H. Dacey, T. Goossens, J. T. M. Elzenga, and J.-L. Tison. 2012. The analysis of dimethylsulfide and dimethylsulfoniopropionate in sea ice: Dry-crushing and melting using stable isotope additions. Mar. Chem. 128–129: 34–43. doi:10.1016/j.marchem.2011.09.007 Stefels, J., and M. A. Van Leeuwe. 1998. Effects of iron and light stres on the biochemical composition of antarctic Phaeocystis sp. I. Intracellular DMSP concentrations. J. Phycol. 34: 486–495. Stefels, J., M. Steinke, S. Turner, G. Malin, and S. Belviso. 2007. 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 Steinke, M., G. V Wolfe, and G. Kirst. 1998. Partial characterisation of dimethylsulfonio- propionate (DMSP) lyase isozymes in 6 strains of Emiliania huxleyi. Mar. Ecol. Prog. Ser. 175: 215–225. Strom, S., G. Wolfe, A. Slajer, S. Lambert, and J. Clough. 2003. Chemical defense in the microplankton II: Inhibition of protist feeding by ß-dimethylsulfoniopropionate (DMSP). Limnol. Oceanogr. 48: 230–237. doi:10.4319/lo.2003.48.1.0230 20" Sunda, W., D. J. Kieber, R. P. Kiene, and S. Huntsman. 2002. An antioxidant function for DMSP and DMS in marine algae. Nature 418: 317–320. Thomas, M. A., P. Suntharalingam, L. Pozzoli, S. Rast, A. Devasthale, S. Kloster, J. Feichter, and T. M. Lenton. 2010. Quantification of DMS aerosol-cloud-climate interactions using the ECHAM5-HAMMOZ model in a current climate scenario. Atmos. Chem. Phys. 10: 7425–7438. doi:10.5194/acp-10-7425-2010 Thume, K., B. Gebser, L. Chen, N. Meyer, D. J. Kieber, and G. Pohnert. 2018. The metabolite dimethylsulfoxonium propionate extends the marine organosulfur cycle. Nature. doi:10.1038/s41586-018-0675-0 Tripp, H. J., J. B. Kitner, M. S. Schwalbach, J. W. H. Dacey, L. J. Wilhelm, and S. J. Giovannoni. 2008. SAR11 marine bacteria require exogenous reduced sulphur for growth. Nature 452: 741–744. doi:10.1038/nature06776 Turner, S. M., G. Malin, P. S. Liss, D. S. Harbour, and P. M. Holligan. 1988. The seasonal variation of dimethylsulfide and dimethylsulfoniopropionate concentrations in nearshore waters. Limnol. Oceanogr. 33: 364–375. doi:10.4319/lo.1988.33.3.0364 Vila-Costa, M., R. Simó, H. Harada, J. M. Gasol, D. Slezak, and R. P. Kiene. 2006. Dimethylsulfoniopropionate uptake by marine phytoplankton. Science 314: 652–654. doi:10.1126/science.1131043 Vogt, M., S. M. Vallina, E. T. Buitenhuis, L. Bopp, and C. Le Quéré. 2010. Simulating dimethylsulphide seasonality with the Dynamic Green Ocean Model PlankTOM5. J. Geophys. Res. Ocean. 115: 1–21. doi:10.1029/2009JC005529 Woodhouse, M. T., G. W. Mann, K. S. Carslaw, and O. Boucher. 2013. Sensitivity of cloud condensation nuclei to regional changes in dimethyl-sulphide emissions. Atmos. Chem. Phys. 13: 2723–2733. doi:10.5194/acp-13-2723-2013 21# Chapter One: The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations Chapter one was previously published in Limnology and Oceanography November 26, 2018. The following is the published version of the Main text and Supplementary information. Limnol. Oceanogr. 9999, 2018, 1–17 © 2018 The Authors.Limnology and Oceanographypublished by WileyPeriodicals,Inc. on behalf of Association for the Sciences of Limnology and Oceanography. doi: 10.1002/lno.11076 The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations Erin L. McParland , * Naomi M. Levine Department of Marine and Environmental Biology, University of Southern California, Los Angeles, California Abstract Dimethylsulfoniopropionate (DMSP) is an important labile component of the marine dissolved organic mat- ter pool that is produced by the majority of eukaryotic marine phytoplankton and by many prokaryotes. Despite decades of research, the contribution of different environmental drivers of DMSP production to regional and seasonal variability remains unknown. A synthesis of the current state-of-knowledge suggested that approx- imately half of confirmed DMSP producers are low producers (intracellular DMSP < 50 mM). Low DMSP pro- ducers (LoDPs; e.g., diatoms) were shown to strongly regulate intracellular DMSP concentrations (~ 16-fold change) as a predictable function of nutrient stress. By comparison, high DMSP producers (HiDPs; e.g., coccolithophores) showed very little response (~ 1.5-fold change). To assess the importance of differential DMSP production by low and high producers, DMSP concentrations were predicted for two time-series sites (a high- and low-productivity site)and for theglobal ocean by explicitly incorporatingboth communitycompo- sition and mechanistic nutrient stress. Despite large, predictable intracellular DMSPchanges, low producers con- tributed less than 5% to global DMSP. This indicates that, while variations in DMSP production by low producers could be important for predicting microbial interactions and low producer physiology, it is not neces- sary for predictingglobalDMSPconcentrations.Ouranalysissuggeststhat communitycomposition, particularly HiDP biomass, is the dominant driver of variability in in situ DMSP concentrations, even in low-productivity regions where high producers are typically the subdominant group. Accurate predictions of in situ DMSP con- centrations require improved representation of subdominant community dynamics in ecosystem models and remote-sensing algorithms. The marine dissolved organic matter (DOM) reservoir stores an equivalent amount of carbon as the atmosphere and fuels microbial life in the upper ocean (Moran et al. 2016). The labile fraction of DOM is rapidly turned over in the surface ocean on timescales of minutes to weeks (Carlson and Hansell 2014). This is especially true for compounds containing nitro- gen, phosphorus, and sulfur that are preferentially reminera- lized in the upper ocean to support microbial growth (Bronk et al. 1994; Dyhrman et al. 2007; Ksionzek et al. 2016). Dimethylsulfoniopropionate (DMSP) is one of the few identi- fiedcompoundscomprisingtherapidlycycled,labiledissolved organic sulfur pool (Ksionzek et al. 2016). Dissolved DMSP has been shown to turn over multiple times a day and supply up to 13% of the bacterial carbon and up to 100% of the bacterial sulfur demands (Kiene et al. 2000; Tripp et al. 2008; Levine et al. 2015). In addition, one of the by-products of DMSP deg- radation, dimethylsulfide (DMS), is a climatically active trace gas that serves as an important source of cloud condensation nuclei in the remote marine boundary layer (Charlson et al. 1987; Quinn et al. 2017) and significantly contributes to Earth’s radiative budget (e.g., Thomas et al. 2010). Both DMS and DMSP have been shown to act as important infochem- icals in microbial interactions at the microscale (Seymour et al. 2010; Johnson et al. 2016). Yet, the mechanisms driving cellular DMSP regulation and variations in in situ particulate DMSP (DMSPp) concentrations are still not fully understood. A seminal work by Keller et al. (1989) provided the first analysis of the diversity of DMSP producers by surveying 123 marine phytoplankton cultures from the Center for Cul- ture of Marine Phytoplankton (CCMP, now the National Cen- ter for Marine Algae) grown under nutrient-replete conditions. Based on this study, phytoplankton have conventionally been divided into two functional groups: high DMSP producers (HiDPs) with intracellular concentrations > 100 mM DMSP and low DMSP producers (LoDPs) with intracellular *Correspondence: mcparlan@usc.edu Additional Supporting Information may be found in the online version of this article. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 1 concentrations < 50 mM DMSP. In the proceeding decades, many hypotheses for the physiological function of DMSP have been proposed, including an osmolyte, a cryoprotectant, a ballasting mechanism, a signaling molecule, an overflow mechanism, and an antioxidant (Karsten et al. 1996; Stefels and Van Leeuwe 1998; Stefels 2000; Sunda et al. 2002; Sey- mour et al. 2010; Lavoie et al. 2015). A unifying aspect of these hypotheses is that all mechanisms propose that DMSP synthesis will be upregulated under different types of cellular stresses (e.g., changes in osmotic pressure or increases in reac- tive oxygen species production). More than 50 studies have quantified the effect of a wide array of environmental stressors on DMSP production using phytoplankton monocultures (e.g., nitrate limitation, low pH, and UV stress). However, these studies often show conflicting results (e.g., van Rijssel and Gieskes 2002; Archer et al. 2010) suggesting that DMSP may play multiple roles in the cell or different roles for differ- ent phytoplankton species (Archer et al. 2010; Bucciarelli et al. 2013). Exactly what those different roles are and how they vary by functional group (e.g., HiDPs vs. LoDPs) remains unclear. In a review article, Stefels et al. (2007) hypothesized that HiDPs may not significantly change cellular DMSP con- centrations in response to nutrient stress, while LoDPs appeared to respond with large changes in cellular DMSP. However, to date, there has not been a synthesis of the pub- lished literature to test this hypothesis. In situ DMSPp concentrations are a function of both the abundance of DMSP producers and changes in cellular DMSP in response to environmental conditions. Previous work has primarily focused on HiDPs, particularly Phaeocystis and Emiliania huxleyi, based on the assumption that HiDPs will dominate DMSPp production due to high-intracellular con- centrations (Stefels 2000). This relationship appears to hold true for highly productive regions with classic spring blooms (e.g., North Atlantic and Southern Ocean) where several previ- ous studies have observed a significant relationship between in situ DMSPp and HiDP biomass (Turner et al. 1988; Malin et al. 1993; Scarratt et al. 2002; Speeckaert et al. 2018). However, seasonal changes in DMSPp concentrations in low productivity, oligotrophic regions have been more difficult to predict (e.g., Masotti et al. 2016). The assumption that DMSPp production by LoDPs is not environmentally relevant is based on intracellular quotas measured in replete laboratory (i.e., nonstressed) conditions (Keller et al. 1989). More recent studies have shown that, under environmental stress (e.g., iron limitation), LoDPs significantly upregulate intracel- lular DMSP to concentrations that approach those of HiDPs (e.g., Bucciarelli et al. 2013). This suggests that LoDPs, which include many important primary producers such as diatoms and cyanobacteria, may contribute significantly to DMSPp cycling, particularly in oligotrophic regions where environ- mental stress is highest (Behrenfeld et al. 1993, 2005). Both numerical ecosystem models and remote-sensing algo- rithms have been used previously to estimate in situ DMSPp. Prognostic ecosystem models that incorporate DMSP modules (e.g., Belviso et al. 2012) predict in situ DMSPp using literature values of DMSP : carbon ratios for a small number of phyto- plankton types. These models tend to both underestimate DMSPp in the oligotrophic oceans and struggle to capture observed seasonal dynamics (e.g., Vogt et al. 2010). A remote- sensing-based estimate of in situ total DMSP (DMSPt) success- fully reproduced global DMSPt patterns over a wide range of biomass, but required two different fits to represent mixed and stratified regimes, which implicitly incorporate differences in light and nutrient status (Galí et al. 2015). Validation of DMSP predictions is difficult, particularly for low-biomass regions, as measurements of in situ DMSP are greatly biased toward higher biomass regions: in the PMEL DMS database (Kettle et al. 1999; Lana et al. 2012), there are ~ 100 DMSPt data points reported between 20 ! N and 30 ! N compared to > 1300 data points reported between 50 ! N and 60 ! N (Galí etal.2015, 2018). To provide mechanistic insight into the drivers of variability in DMSPp concentrations, we identified over-arching trends in the regulation of intracellular DMSP in response to nutrient stress using previously published monoculture physiology stud- ies supplemented with new measurements. Here, we focused on nutrient stress responses as the current body of literature does not allow for comparison of HiDP and LoDP DMSP regulation under other stressors. The experimentally derived equations parameterizing the relationship between intracellular DMSP of LoDPs and HiDPs and nutrient stress were then used to mecha- nistically predict DMSPp as afunction of bothcommunity com- position and environmental stress at two time-series sites and globally using a biogeochemical ecosystem model. LoDPs were shown to strongly upregulate intracellular DMSP as a predict- able function of growth limitation due to nutrient stress, while HiDPs exhibited a seemingly constitutive production of DMSP. We show that the magnitude of change in intracellular DMSP due to nutrient stress plays a minor role in determining in situ DMSPp, and that community composition, particularly HiDP biomass, is the driving mechanism of variability inseasonal and regional differencesofin situ DMSPp. Methods Diversity of DMSP producers The known diversity of DMSP producers was assessed using previously published studies that confirmed DMSP production (Supporting Information Table S1). These studies used indirect measurements of DMSP through derivatization to DMS and gas chromatography (GC) detection. We only included measure- ments from monoculture studies as in situ estimates of species specific DMSP concentrations (e.g., Archer et al. 2011) could be biased by the uptake of dissolved DMSP which can vary signifi- cantly byspecies (Vila-Costaetal.2006; Lavoieetal.2018). We supplemented the published data by quantifying DMSP in monoculturesof 20additional phytoplanktonstrains, includ- ing: cyanobacteria (Crocosphaera watsonii [WH003], Crocosphaera McParland and Levine Driving mechanisms of global surface DMSP 2 watsonii [WH005], Trichodesmium erythraeum [strains IMS101 and 2175], Trichodesmium thiebautii [strain GBR], Synechocystis sp. [PCC6803], 11 isolates of marine Synechococcus), a single cell Rhizarian (Bigelowiella natans [CCMP2755]), a Pavlophyte (Exanthemachrysis sp. isolated from the San Pedro Ocean Time- series [33 ! 33 0 N, 118 ! 24 0 W] in 2017), and diatoms (Fragilariopsis cylindrus and Pseudo-nitzschia subcurvata isolated from the ice edge in McMurdo Sound [77.62 ! S, 165.47 ! E] in the Ross Sea, Antarctica in 2013) (Supporting Information Table S1). DMSPt was quantified on a custom Shimadzu 2014 GC using a flame photometricdetectorandaChromosil330(Supelco)packedcol- umn through either a direct headspace injection using Hamil- ton gas-tight needles or a purge and trap method using a liquid nitrogen cold trap. DMSP was measured as its derivative, DMS, producedviaalkalinehydrolysiswith5 NNaOH(Kiene andSer- vice 1991). DMSP was also quantified directly for Trichodesmium erythraeum (strain IMS101, previously measured by Bucciarelli et al. 2013 with GC) and Trichodesmium thiebautii (strain GBR) using triple quadrupole mass spectrometry at the Woods Hole Oceanographic Institution (Kido Soule et al. 2015) and for Syne- chocystis sp. using high performance liquid chromatography (HPLC) with a modified protocol described by Gorham (1984) (SupportingInformationTableS1). A tree of all confirmed DMSP producers was built with 16S and 18S phylogeny using Geneious version R11 (http://www. geneious.com, Kearse et al. 2012). 16S and 18S sequences were extracted from the NCBI and had a minimum of 1200 base- pairs. The sequences were aligned in Geneious with the Multi- ple Alignment Tool. The eukaryotic tree was grouped by the five currently accepted major supergroups (Burki 2014) and the prokaryotic tree was grouped by either phylum or class (cyano- bacteria, alphaproteobacteria, and gammaproteobacteria). To conserve space, some DMSP producers were omitted from the eukaryotic tree but are included in Supporting Information Table S1. Strain sequences identical to those in the studies were used whenever possible. If a strain number from an older study was altered or not available, 18S sequences of the same genus were used to represent the species in the tree. All reported pro- karyotic species were included if 16S sequences were available. We differentiate between HiDPs and LoDPs with intracellular DMSP concentrations (normalized to cell volume) using a 50 mM cut-off. If only cellular quotas were reported, we used a cut-offof1 fmol DMSP cell −1 , whichrepresentsthe lowerrange of prymnesiophytes (Supporting Information Table S1). Nor- malization per volume is highly preferable (Keller 1989) and cell volumes should be measured in the future for accurate assignment ofHiDP and LoDPfunctionaltypes. Intracellular DMSP regulation as a function of nutrient stress All known previous studies that report the effect of nutri- ent limitation (CO 2 , N, Fe, P, Si, or stationary/senescence phase) on cellular DMSP were assessed. For each study, DMSP values were extracted from the text as reported by the authors. If values were not reported in the text, data was extracted from figures using the MathWorks GRABIT GUI program. To allow for intercomparison, only studies reporting intracellular DMSP concentrations in mmol DMSP cell L −1 (mM) were included in the analysis (n = 11) (Supporting Information Table S2). For each study, a fold change due to environmental stress was calculated as: FC q,k ¼ DMSPstressed q,k DMSPreplete q,k ð1Þ whereFC q,k isthefoldchangeforphytoplanktonstrainqunder nutrient limitation k. DMSP stressed q,k is intracellular DMSP measuredundernutrientlimitation,andDMSP replete q,k isintra- cellular DMSP measured under nutrient replete (nonstressed) conditions. For the majority of studies, intracellular DMSP concentrations were quantified during mid-exponential growth under replete (DMSP replete q,k ) and nutrient-limited (DMSP stressed q,k ) conditions. If intracellular DMSP was mea- suredacrossabatchgrowthcurve(e.g.,Franklinetal.2010),the observed value from the late exponential or stationary phase was used for DMSP stressed q,k and the observed value from mid-exponential growth was used for DMSP replete q,k . For each FC q,k , the impact of nutrient limitation on growth was calculatedas: γ q,k ¼ μ q,k μmax q,k ð2Þ where γ q,k is growth limitation for phytoplankton strain q due to nutrient stress k, μ q,k is the reported growth rate under nutrient limitation, μmax q,k is the reported growth rate under nutrient replete conditions. γ q,k ranges from 1 (unstressed) to 0 (completely inhibited). The relationship between intracellular DMSP and growth limitation was derived for each study. For LoDPs, there were a total of 11 fits (Thalassiosira pseudonana [n = 7], Thalassiosira oceanica [n = 2], Skeletonema marinoi [n = 1], Trichodesmium ery- thraeum [n = 1]) (Supporting Information Fig. S1). When LoDP intracellular DMSP was measured for three or more signifi- cantly different growth rates (n = 5), a sigmoidal function was fit to the data: I q,k ¼p1 q + p2 q −p1 q 1+e p3q−γ q,k ðÞ*p4q ð3Þ where I q,k is intracellular DMSP (mM), γ q,k is growth limita- tion, and p1 q –p4 q are bestfit parameter values for phytoplank- ton strainq determined using a Levenberg-Marquardt nonlinear regression. This form was chosen as it best approximated the observed response curves with R 2 > 0.9 for the parameter fits. McParland and Levine Driving mechanisms of global surface DMSP 3 If only two significantly different growth rates were reported for LoDPs (n = 6), a linearfit was used: I q,k ¼p1 q *γ q,k +p2 q ð4Þ where p1 q is the slope fit for phytoplankton strain q and p2 q is the y-intercept. All HiDPs (n = 9) were assigned a linear fit (coccolitho- phores [n = 6], Phaeocystis antarctica [n = 2], and Amphidinium carterae [n = 1]) (Supporting Information Fig. S2). For those strains with intracellular DMSP measured at more than two significantly different growth rates (n = 5), R 2 values ranged from 0.05 to 0.96. Modelingin situ DMSPpconcentrations To assess the impact of nutrient stress on global DMSPp prediction, we used output from a global biogeochemical- ecosystem model. The version of the MIT ecosystem model (DARWIN) used in this study was based on Dutkiewicz et al. (2015) and contained 51 plankton types incorporating both functional and size diversity. Specifically, the ecosystem model included 35 phytoplankton types: picoprokaryotes (2 size classes), picoeukaryotes (2 size classes), coccolitho- phores (5 size classes), diazotrophs (5 size classes), diatoms (11 size classes), and mixotrophic dinoflagellates (10 size clas- ses). The phytoplankton were grazed by 16 size classes of zoo- plankton. The size classes within functional groups followed the allometric parameterizations used in Ward et al. (2012). The ecosystem was embedded in a three dimensional physical model with 18 km resolution constrained by observations (Menemenlis et al. 2008). DMSPp production was calculated for all 35 model phyto- plankton types as a function of nutrient stress (Eq. 3–8) using 3-d average surface nutrient and biomass fields from the year 2000. The year 2000 was chosen as a representative “generic” year with no significant El Niño or La Niña effects. The pico- prokaryotes, diazotrophs, and diatoms were classified as LoDPs. The picoeukaryotes, coccolithophores, and dinoflagellates were classified as HiDPs. Growth limitation due to nutrient stress for each phytoplankton type (j) for each model grid cell was calcu- lated as the most limiting nutrient (i) after Dutkiewicz etal. (2015): γ j ¼min N limji !" ð5Þ Limitation by PO 4 , Si, and Fe was calculated following Michaelis-Menten formulation: N limji ¼ N i N i +κN ij ð6Þ where N i is the nutrient concentration and κN ij is the half- saturation constant for nutrient i for phytoplankton type j. Nitrogen limitation in the model takes the form: N limji ¼ NO 3 +NO 2 NO 3 +NO 2 +κ inj e −ψNH4 + NH 4 NH 4 +κ nh4j ð7Þ where κ inj is the half-saturation constant of inorganic nitrogen (NO 3 + NO 2 ), κ nh4j is the half-saturation constant of NH 4 , and ψ reflects thefixed nitrogen uptake inhibition by ammonia. Intracellular DMSP concentrations were then calculated for each phytoplankton type in each model grid cell using the derived equations for the relationship between intracellular DMSP and growth limitation (Eqs. 2–4) and the modeled growth limitation (Eq. 5). For functional groups with multiple strain specific relationships (picoeukaryotes, diatoms, cocco- lithophores, and dinoflagellates; Supporting Information Figs. S1, S2), the intracellular DMSP concentration was calcu- lated for each relationship and then the average predicted intracellular concentration was used. As there were no pub- lished studies on intracellular DMSP under nutrient limitation for the picoprokaryotes, Synechococcus, and Prochlorococcus, we used the response reported for Trichodesmium (an organism with very different ecological importance but similar intracel- lular DMSP concentrations) (Supporting Information Fig. S1). In DARWIN, nutrient half-saturation constants varied as a function of cell size such that larger cells experienced more nutrient limitation than smaller cells within the same group. This yielded a range of intracellular DMSP quotas within each functional group. Finally, water column DMSPp (nM) was cal- culated for each model grid cell as: DMSPp j ¼I j *vol j * 1 Q j *X j ð8Þ where I j is the intracellular DMSP for phytoplankton type j, vol j is the assigned biovolume (μm 3 ) for j, Q j is the assigned carbon quota for j, and X j is the modeled biomass of j (mmol Cm −3 ). Values for half saturation constants, cell volume, and cellular carbon quota of the 35 phytoplankton types are pro- vided in Supporting Information Table S3. GlobalDMSPdatabase The global distribution of DMSPp estimated using DARWIN output was compared against observed surface DMSP concen- trations (< 10 m) from the PMEL database (https://saga.pmel. noaa.gov/dms/). Due to the paucity of DMSPp observations, we relied upon DMSPt measurements to validate the global patterns predicted by the model. DMSPt is a reasonable proxy for DMSPp global patterns as DMSPp typically comprises > 90% of the DMSPt pool (Kiene and Slezak 2006). We excluded studies that followed an acidification protocol for samples with a high abundance of Phaeocystis (n = 145) (creat- ing significant overestimation, del Valle et al. 2011) and sam- ples collected in harbors, estuaries, or with water column depths < 200 m (n = 207) (Galí et al. 2015). We also added a new dataset (n = 13) from a transect in the equatorial western Pacific in 2015, an under-sampled area for DMSP McParland and Levine Driving mechanisms of global surface DMSP 4 measurements. After curating the dataset, there were a total of 4210 DMSPt measurements. Time-series in situ data The methodology described above for estimating in situ DMSPp based on community composition and nutrient stress response curves was used to calculate DMSPp at two time- series sites with well-characterized seasonal DMSPp dynamics: the Bermuda Atlantic Time series Site (BATS) (25.9 ! N, 58.7 ! W) in the subtropical North Atlantic from August 2007 to September 2008, and the Antarctica LTER Palmer Station B (Palmer) (64.78 ! S, 64.07 ! W) from November 2005 to February 2006. BATS is a low productivity, oligotrophic site while Palmer is a high-productivity site with a classic spring/summer bloom. HPLC pigments, chlorophyll a (Chl a), and particulate organic carbon (POC) measurements corresponding to the days DMSPp was measured were extracted from respective databases (http://batsftp.bios.edu/, accessed 18 December 2017, https://oceaninformatics.ucsd.edu/datazoo/catalogs/ pallter/, accessed 17 March 2018). Equation 8 was used to predict DMSPp at both time-series sites. X j (total carbon biomass of functional group j) was esti- mated from HPLC pigment concentrations. At BATS, the HPLC algorithms of Andersenet al. (1996) that have beenvali- dated for the BATS site were used to estimate group specific Chl a concentrations of prochlorophytes and diatoms (assigned to be LoDPs) and prymnesiophytes, pelagophytes, and dinoflagellates (assigned to be HiDPs). At Palmer, the site- Cyanobacteria Rhizaria Purple Sulfur Bacteria Purple Non-Sulfur Bacteria Alphaproteobacteria Diatoms Stramenopiles Prymnesiophytes Dinoflagellates Chlorophytes and Prasinophytes 16S 18S Fig. 1. Tree of representative prokaryotic (left) and eukaryotic (right) DMSP producers built with 16S and 18S phylogeny. The prokaryotic producers are grouped by functional groups, while the eukaryotic producers are grouped by the major eukaryotic supergroups. Blue text represents LoDP’s (intracellular DMSP < 50 mM) and red text represents HiDP’s (intracellular DMSP > 50 mM). McParland and Levine Driving mechanisms of global surface DMSP 5 specific HPLC CHEMTAX ratios (Kozlowski et al. 2011) were used to estimate Chl a of flagellates and diatoms (assigned to be LoDPs) and prasinophytes, cryptophytes, and Phaeocystis (assigned to be HiDPs). Group-specific Chl a was converted to carbon using in situ measured POC : Chl a ratios. As group specific POC : Chl a ratios were not available, the median POC : Chl a at Palmer of 146 was used. At BATS, seasonality of POC : Chl a was found to be important for accurate DMSPp prediction and therefore seasonally varying POC : Chl a based on observed values were used (see Supporting Information Fig. S3 for details). Group specific carbon was then converted to biovolume using the biovolume : carbon quota ratio for the corresponding functional group in the DARWIN model (see Supporting Information Table S4 for constants). Finally, DMSPp was predicted (Eq. 8) using the estimated bio- volume m −3 for each group (vol j * 1 Qj *X j ) and the modeled intracellular DMSP of each group (I j in Eq. 8) calculated using the nutrient stress for each site estimated by DARWIN. Pre- dicted DMSPp production by each group was summed and concentrations were compared against monthly measured DMSPpconcentrations during an entire seasonal cycle atBATS (Levine et al. 2015) and against approximately biweekly DMSPp measurements during the Antarctic field season (November–February) at Palmer (Herrmann et al. 2012). Results Diversity of DMSP producers A compilation of all published studies and our own mea- surements demonstrated a much greater diversity in DMSP producers than previously assumed (Keller et al. 1989), with over 50% of these DMSP producers classified as LoDPs (n = 113 of 216 total) (Fig. 1, Supporting Information Table S1). The major eukaryotic supergroups Rhizaria, Chro- malveolate, and Archaeplastid all have many confirmed DMSP producing representatives. The absence of representatives from the unikont and excavate supergroups does not imply these groups lack the ability to produce DMSP, but rather a lack of measurements for these groups. Though often over- looked, a diverse array of prokaryotic groups has also been shown to produce DMSP, including cyanobacteria, purple sul- fur bacteria, purple non-sulfur bacteria, and alphaproteobac- teria (Karsten et al. 1996; Curson et al. 2017) (Fig. 1). Critically, four key photosynthetic marine cyanobacteria (Synechococcus, Prochlorococcus, Trichodesmium, and Croco- sphaera) have all been shown to produce DMSP (Corn et al. 1996; Bucciarelli et al. 2013; this study), albeit with extremely low-intracellular concentrations. The diversity of DMSP producers highlighted by this compilation suggests a potentially deep rooted evolution of the DMSP synthesis path- way as DMSP is found in many prokaryotic phyla belonging to ancient lineages (Fig. 1). This is consistent with the phylog- eny of the recently identified prokaryotic DMSP synthesis gene (dysB) and eukaryotic DMSP synthesis gene (DYSB) (Curson et al. 2017, 2018). Meta-analysis To identify over-arching trends in DMSP regulation by LoDPs and HiDPs in response to environmental stress, we ana- lyzed the previously reported response of 19 different strains of phytoplankton to seven different types of nutrient limita- tion (Supporting Information Table S2). Comparing across previous studies is complicated due to fundamental differ- ences in methodologies: strains studied, growth conditions, and normalization factor. In particular, the use of different biomass normalizations can create very different results even within the same study as nutrient stress affects multiple com- ponents of cellular physiology, not just cellular DMSP regula- tion. For example, Thalassiosira oceanica grown under iron limitation showed a 12-fold increase in DMSP when normal- ized to cell volume, a 40-fold increase when normalized to Chl a, but only a 1.2-fold increase when normalized to cell carbon (Bucciarelli et al. 2013). To remove this effect, only studies reporting intracellular DMSP (normalized to cell vol- ume) were included in our analysis. In addition, to facilitate comparison across studies, we analyzed results in terms of growth limitation (γ, nutrient stressed growth rate relative to replete growth rate) and the fold change in DMSP production (nutrient stressed intracellular DMSP concentrations relative to replete concentrations). Despite large differences in cultur- ing techniques and nutrient limitations, similar responses were seen across the seven different types of nutrient limita- tions with consistent differences observed between HiDPs and LoDPs (Table 1). The average intracellular DMSP concentration for HiDPs under significant limiting conditions of γ < 0.5 (253! 120 mM DMSP) was not significantly different than the average concentration under replete conditions (224! 127 mM DMSP, t-test p > 0.1). The high variability in the average intracellular DMSP for HiDPs (Fig. 2) was primarily driven by the large range of intracellular DMSP concentrations for 10 coccolithophore species (range of 174–715 mM DMSP) reported by Franklin et al. (2010). The average intracellular DMSP concentration for LoDPs under the same limiting con- ditions (23! 16 mM DMSP) was significantly higher than the average intracellular DMSP under replete conditions (8! 14 mM DMSP, t-test p = 0.004) (Fig. 2). While intracellu- lar DMSP concentrations of LoDPs were much lower than those of HiDPs, the LoDPs’ intracellular DMSP concentrations under nutrient stress begin to approach those of the HiDPs, with a maximum of 67 mM intracellular DMSP reported for Skeletonema marinoi (Spielmeyer and Pohnert 2012) (Fig. 2). LoDPs showed an average fold change in intracellular DMSP of 16! 19 in response to significant nutrient stress (γ < 0.5) (Fig. 2; Table 1), with a maximum fold change of 73 by Thalassiosira pseudonana in response to CO 2 limitation (Sunda et al. 2002). In contrast, the HiDPs showed both a McParland and Levine Driving mechanisms of global surface DMSP 6 smaller fold change and a less variable response of 1.4! 0.5, with a maximum fold change of 2.6 by Emiliania huxleyi in response to CO 2 limitation (Sunda et al. 2002) (Fig. 2; Table 1). Differences in experimental design and degrees of nutrient limitation employed across these studies most likely contributed significantly to the highly variable fold changes for LoDPs. Despite this variability, we found the strong upre- gulation of intracellular DMSP by LoDPs to be highly predict- able as a function of γ for every study and strain tested (n = 11) (Fig. 3, Supporting Information Fig. S1). In contrast, the low variability in HiDP fold changes, despite similar differ- ences in experimental design, strongly supports a lack of a DMSP response by HiDPs to nutrient stress. Unlike LoDPs, a predictable increase in intracellular DMSP as a function of γ was not found for HiDPs (n = 9), and some studies even showed a slight decrease of intracellular DMSP with decreasing γ (Fig. 3, Supporting Information Fig. S2). The aim of this study was to identify overarching trends in DMSP regulation by HiDPs and LoDPs in response to nutrient stressors. However, for completeness, an analysis of the avail- able data for the response to temperature, light, and global change stressors are included in the Supporting Information S1. The response of HiDPs to non-nutrient stressors was simi- lar to the nutrient stress response. Further studies quantifying the response of LoDPs to non-nutrient stressors are needed to understand how these stressors might impact DMSP produc- tion by this group and how this differs from the HiDP response. Predicted DMSPp in high- and low-productivity regions To assess the contribution of a nutrient stress response to variability in in situ DMSPp concentrations and to estimate the potential contribution of LoDPs to in situ DMSPp, the observed relationship between growth limitation and intracel- lular DMSP was used to predict surface DMSPp at two time- series sites. Good agreement was observed at both Palmer and BATS (Fig. 4) between in situ predicted and observed DMSPp temporal patterns (R 2 = 0.8, p < 0.001 at Palmer, R 2 = 0.4, Table 1. The average and maximum fold changes for HiDPs and LoDPs in response to six different nutrient limitations (n = number of data points analyzed) with a significant reduction in growth limitation due to nutrient stress (γ < 0.5). Nutrient n Average fold change Maximum fold change HiDP LoDP HiDP LoDP HiDP LoDP All 23 24 1! 0.5 16! 19 —— NO 3 78 1! 0.3 14! 2 2 29 Fe 6 8 1! 0.3 13! 15 2 45 CO 2 75 2! 0.5 37! 31 3 73 Si — 1 —— — 7 PO 4 — 1 —— — 3 k/50 1 —— — 2 — Stationary 2 1 1! 0.1 2! 0.4 2 3 HiDP LoDP Intracellular DMSP (mM) 0 100 200 300 400 500 HiDP LoDP Fold Change 0 20 40 60 80 (a) (b) Fig. 2. (a) Intracellular DMSP reported for HiDP’s and LoDP’s when γ < 0.5. Black dashed line represents the 50 mM boundary between HiDP’s and LoDP’s. (b) Fold change in HiDP’s and LoDP’s intracellular DMSP under γ < 0.5 relative to replete growth. Black dashed line of 1 represents no response to nutrient limitation. McParland and Levine Driving mechanisms of global surface DMSP 7 p = 0.02 at BATS, Supporting Information Table S5). Assump- tions for predicting biomass from HPLC pigments (see “Methods” section) and the use of average intracellular DMSP fits most likely contributed to divergences between predicted and measured DMSPp. Good agreement was also observed between our mechanistic prediction of DMSPp and the empir- ical algorithm from Galí et al. (2015) at Palmer (R 2 = 0.95, p < 0.0001) (Supporting Information Fig. S4). At BATS, while both the DARWIN and Galí et al. predictions showed similar root mean squared error when compared against in situ obser- vations, there was no relationship between the two predic- tions or between Galí et al. and the in situ measurements (SupportingInformationTable S5). Theability of theDARWIN prediction to better match important temporal features of DMSPp at BATS relative to Galí et al. (Supporting Information Fig. S4 and Table S5) suggests that explicit incorporation of nutrient stress and/or community composition is important for predicting in situ DMSPp variability in low-productivity, oligotrophic regions. Asexpected,thehigh-productivitysite(Palmer)nevershowed significant growth limitation in the DARWIN model and γ was always predicted to be > 0.9. As a result, intracellular DMSP was alwaysataminimumforallphytoplanktontypesandvariability ininsituDMSPpwasdrivenbyshiftsincommunitycomposition. DuringtheprimaryDMSPpmaximum(predictedtobe>400nM), acryptophyte(HiDP)bloomaccountedfor46%oftotalbiomass andwasestimatedtohaveproduced77%oftotalDMSPp(Fig.4). During the secondary maxima (16–30 January 2006), the LoDPs (diatoms andflagellates) dominated more than 90% of the total biomass, and yet, the HiDPs were still estimated to produce an average66%ofthetotalDMSPpduringthisperiod. In contrast, at the low-productivity site (BATS), average γ for all phytoplankton types varied from a maximum of 0.7 during the winter to a community averaged minimum of 0.04 during the summer in DARWIN. This resulted in upregulation of DMSP by both LoDPs and HiDPs during the summer. How- ever, the large upregulation of LoDP intracellular DMSP due to nutrient stress was not enough to result in a significant contri- bution to in situ DMSPp. For example, prochlorophytes (LoDPs) dominated the community (an average 53% of total biomass throughout the year, and maximum of 87%), but were only estimated to contribute a maximum 7% of total DMSPp (Fig. 4). Diatoms experienced the most extreme growth limitation (γ ~ 0.01) but never contributed more than 1% of total DMSPp. Instead, similar to Palmer, the HiDPs at BATS (pelagophytes, prymnesiophytes, and dinoflagellates) were estimated to dominate DMSPp production, accounting for 97% of the total seasonal DMSPp produced. These two examples suggest that our single set of equations incorporat- ing differential nutrient response for HiDPs and LoDPs may allow us to accurately predict in situ DMSPp for very different oceanographic regions. Furthermore, these results suggest that HiDP biomass dominates DMSPp production, even when HiDPs are the subdominant population. To confirm the importance of community composition in determining in situ DMSPp, we also estimated DMSPp for the 0 0.2 0.4 0.6 0.8 1 Intracellular DMSP (mM ) 100 200 300 400 High DMSP producers Amphidinium carterae (R 2 =0.05) Phaeocystis antarctica (R 2 =0.97) Emiliania huxleyi (R 2 =0.23 ) 50 (a) 40 30 20 10 0 γ 0 0.2 0.4 0.6 0.8 1 (b) Low DMSP producers No Growth (γ = 0) No limitation (γ = 1) Increasing nutrient stress γ No Growth (γ = 0) (γ = 1) Increasing nutrient stress 50 Skeletonema marinoi (R 2 =0.97) Thalassiosira oceanica (R 2 =0.99) Thalassiosira pseudonana (R 2 =0.98) Intracellular DMSP (mM ) No limitation Fig. 3. Response curves of intracellular DMSP as a function of growth limitation due to nutrient stress for (a) LoDP’s and (b) HiDP’s. Three representative examples are shown here. All response curves are shown in Supporting Information Fig. S1 and S2. McParland and Levine Driving mechanisms of global surface DMSP 8 two sites using total Chl a (Supporting Information Fig. S4). As has been shown previously (e.g., Toole and Siegel 2004), the Chl a based estimate was able to reproduce the majority of the variability observed in the high-productivity region, but was unable to capture the dynamics in the low-productivity region. Indeed, total Chl a at Palmer was significantly corre- lated with observed DMSPp, while there was no relationship at BATS (Fig. 5). We propose that the linear relationship between DMSPp and Chl a in high-productivity regions is a result of Chl a being a good proxy for HiDP biomass in these regions (R 2 = 0.6, p < 0.001 for Chl a vs. HiDP cellular biovo- lume at Palmer). In fact, a correlation was observed between DMSPp and HiDP cellular biovolume concentrations (μL cell L −1 ) for both sites (Fig. 5). At BATS, the weaker relationship was expected due to uncertainties in our assumptions (i.e., C : Chl a ratios, biovolume, and cellular carbon quota) that introduce greater uncertainty in the estimate of HiDP cel- lular biovolume at low-biomass concentrations than at Palmer where total biomass is an order of magnitude greater (note x- axes of Fig. 5). Despite an order of magnitude difference in the observed range of DMSP concentrations, the slopes of the lin- ear relationship between DMSPp and HiDP cellular biovolume at both time-series sites were not statistically different (p = 0.2, Z-test statistic) and were equivalent to typical HiDP intracellu- lar DMSP concentrations (Fig. 5). This provides evidence that a single set of equations for predicting DMSPp based on HiDP cellular biovolume is justified even in contrasting high- and low-productivity regions. Furthermore, potential stress mecha- nisms, including γ at Palmer and BATS (Fig. 5) and upper mixed layer UV light dose at BATS (Supporting Information Fig. S5), were not significantly correlated with in situ DMSPp. This again suggests that environmental stress plays a secondary role to community composition in driving in situ DMSPp variability. ModeledglobalDMSPp Spatial and temporal differences in the contribution of nutrient stress and HiDPs and LoDPs to global patterns of in situ DMSPp were assessed using the DARWIN model. The DARWIN-based estimates of DMSPp captured the critical spa- tial and temporal features in observed in situ DMSPt concen- trations (Fig. 6, Supporting Information Fig. S6). Specifically, the model estimates captured the observed elevated DMSPp concentrations in the high latitude spring and upwelling regions, and the low DMSPp concentrations in oligotrophic gyres (Fig. 6), as well as temporal features in oceanographic provinces (Supporting Information Fig. S6). A significant lin- ear relationship was observed between DARWIN predicted DMSPp and in situ measured DMSPt observations binned by HiDP community composition as predicted by DARWIN (Fig. 7). However, predicted DMSPp was both underestimated relative to observations with a mean relative bias of−53% and showed significantly lower variability (Figs. 6, 7). Differences between the observed and modeled predictions of DMSP were expected as interannual variability in the tim- ing and spatial location of blooms and shifts in community composition are not captured by the DARWIN model output for a single average year. In high-productivity regions, the model-observation differences were attributed to a sampling bias in the observational dataset toward spring bloom mea- surements (Galí et al. 2018), spatial averaging or an underesti- mate of the magnitude of spring blooms in the model, and mismatches in community composition between the model and in situ data. Blooms are often patchy and driven by fine- ND J F M A 0 100 200 300 400 DMSPp (nM) 0 5 10 15 20 25 30 BATS JF M A M J JA S O N D Palmer (a) (b) Prochlorophytes Diatoms Pelagophytes Prymnesiophytes Dinoflagellates Flagellates Diatoms Cryptophytes Phaeocystis Prasinophytes DMSPp (nM) Fig. 4. Water column predicted DMSPp at (a) Palmer and (b) BATS using HPLC pigment biomass. Bars represent HiDP phytoplankton types (red) and LoDP phytoplankton types (blue) contribution to total DMSPp. Solid line and black dots represent DMSPp measured in situ. Error bars at BATS are reported standard deviation from Levine et al. (2015). Error estimate was not available for the entire time series at Palmer, but of those available (January–March), standard error was consistently < 3% (Herrmann et al. 2012). McParland and Levine Driving mechanisms of global surface DMSP 9 scale processes (e.g., Mahadevan et al. 2012) that are typically not captured by large-scale models (Hashioka et al. 2013). Spe- cifically, DARWIN predicted DMSPp represents the “mean state” (averaged over 3-d and 18 km) and therefore does not capture fine-scale variability and elevated Chl a that are observed during blooms. Finally, the lack of a Phaeocystis func- tional group in DARWIN, particularly in the Antarctic where diatoms(LoDPs) compensatefor this ecologicalniche, contrib- uted substantially to DMSPp underestimation in the model (Wang et al. 2015). Underestimates in oligotrophic regions are primarily attributed to an underestimate of HiDP biomass when these groups were the subdominant phytoplankton community. In DARWIN, cyanobacteria (picoprokaryotes and diazotrophs) dominated phytoplankton biomass (~ 50–100%) in oligotrophic regions (Supporting Information Fig. S7) and HiDP phytoplankton types appeared to be underestimated based on a comparison to HPLC pigments at BATS. As observed at the Palmer and BATS time-series sites, HiDP biomass is the primary determinant of in situ DMSPp concentrations in all oceanographic regions (Fig. 7). Specifi- cally, the biomass ofthe 17HiDP phytoplanktontypes contrib- uted 96% of the predicted in situ DMSPp concentrations. In areas of high DMSPp production, mainly upwelling regions and frontal zones, coccolithophores and dinoflagellates domi- nated DMSPp production. In oligotrophic regions, the picoeu- karyotes were the most abundant HiDP and so dominated DMSPp production (e.g., at BATS in DARWIN, picoeukaryotes produced 81% of the total DMSPp in August). The importance of picoeukaryotes for oligotrophic DMSPp production in our predictions supports the findings of Galí et al. (2015), which also suggested an assignment of HiDP to the picoeukaryote functional group despite having very few measurements of cel- lular DMSP (Supporting Information Table S1) for this impor- tant group (Massana 2011). The modeled relationship between HiDP community composition and DMSPp concentration was alsoobservedintheinsitudatasetwheresampleswithahigher fraction of HiDP (predicted by DARWIN) had higher measured DMSPt thansampleswithalowerfractionHiDP (Fig.7). 10 15 20 25 0 50 100 150 200 250 300 350 400 0.02 0.04 0.06 0.08 0.1 0.12 0.14 10 12 14 16 18 20 22 24 26 0 0.5 50 100 150 200 250 300 350 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 8 10 12 14 16 18 20 22 24 26 Growth limitation 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 HiDP cellular biovolume (µL cell L -1 ) Chl a (µg L -1 ) 5 0 Chl a (µg L -1 ) 1 1.5 2 HiDP cellular biovolume (µL cell L -1 ) Palmer BATS BATS DMSP (nM) DMSP (nM) DMSP (nM) DMSP (nM) m = 176±32 y-int = 45±16 R 2 = 0.8 p < 0.001 m = 127±61 y-int = 9±3 R 2 = 0.3 p = 0.07 Palmer m = 10±2 y-int = 33±18 R 2 = 0.6 p < 0.001 (a) (b) (c) (d) Fig. 5. In situ measured DMSPp (nM) vs. Chl a (μgL −1 ) at Palmer (a) and BATS (b). m is the slope, y-int is the y intercept, R 2 is the R-square statistic and p is the p-value from a least squares linear regression. There was no significant relationship at BATS. In situ measured DMSPp (nM) vs. cumulative HiDP cellular biovolume concentrations (μL cell L −1 ) at Palmer (c) and BATS (d) colored by growth limitation due to nutrient stress (γ). The sum of all HiDP car- bon biomass (estimated from HPLC pigments, see “Methods” section) was converted using a cellular carbon quota of 6.6 × 10 −6 mmol C cell −1 and a biovolume of 46 μm 3 (phytoplankton type 6 in DARWIN; Supporting Information Table S3). McParland and Levine Driving mechanisms of global surface DMSP 10 LoDP mean intracellular DMSP concentrations were greatly upregulated in the oligotrophic oceans due to nutrient stress as expected, 56-fold, 10-fold, and 62-fold for picoprokaryotes, dia- zotrophs, and diatoms, respectively. HiDPs upregulated intra- cellular DMSP in these regions as well, but never as drastically as the LoDPs with a maximum 1.4-fold, 1.6-fold, and 1.6-fold upregulation by picoeukaryotes, coccolithophores, and dinofla- gellates,respectively.Despiteasmallresponsetonutrientstress, HiDPs dominated the increase in DMSPp inventory due to nutrient stress. Specifically, 76% of the global increase in DMSPp due to nutrient stress was associated with the small upregulation of HiDP intracellular DMSP. Including nutrient stress in modeled DMSPp resulted in an average 1.8 nM increase in in situ DMSPp (a 28% increase) relative to estimates using only community composition (i.e., assigning HiDPs and LoDPs stagnant replete laboratory intracellular quotas as done previously; Le Clainche et al. 2010) (Fig. 8). LoDP intracellular DMSP was always ata maximum inthe oligotrophicocean, but the magnitude of this upregulation was not great enough to compensate for the inherently lower intracellular DMSP. Thus, LoDPs did not contribute significantly to predicted DMSPp (annual average 4%). In fact, the highest contribution of DMSPp by LoDPs was found not in the nutrient limited oligo- trophic oceans, but in the regions where these phytoplankton typesweremost abundant. Of the LoDPs, diatoms had the most potential to contribute to in situ DMSPp due to their higher intracellular DMSP con- tent relative to the other LoDP phytoplankton types (Supporting Information Fig. S1), their larger size, which made them more nutrient limited in the model, and their ability to achieve higher biomass values (blooms) in the model. How- ever, the maximum contribution by diatoms to total DMSPp was 2 nM in the equatorial upwelling region (or 5.9 nM if only the response for Skeletonema marinoi was used; Fig. 3). While these results suggest that environmental (nutrient) stress is not the first-order mechanism for predicting trends in global DMSPp, it was important for accurately representing the contribution by LoDPs. Without incorporating this mech- anistic upregulation (Fig. 3), contribution to global DMSPp by LoDPs would be much less (annual average 1.8%). DMSP (nM) 10 20 30 40 50 0 -80 -60 -40 -20 0 20 40 60 80 -150 -100 -50 0 50 100 150 -80 -60 -40 -20 0 20 40 60 80 -150 -100 -50 0 50 100 150 -80 -60 -40 -20 0 20 40 60 80 -150 -100 -50 0 50 100 150 -80 -60 -40 -20 0 20 40 60 80 -150 -100 -50 0 50 100 150 Winter Spring Summer Fall (a) (b) (c) (d) Fig. 6. Seasonal mean predicted DMSPp for (a) winter (December–February), (b) spring (March–May), (c) summer (June–August), (d) fall (September– November). Overlay of points represent measured DMSPt from the curated PMEL database from the same months. Note the maximum 50 nM of the scale bar is much lower than the maximum observed in situ DMSPt (see Fig. 7). The scale was capped to facilitate comparison of spatial and temporal trends between model and observations. McParland and Levine Driving mechanisms of global surface DMSP 11 Discussion The high intracellular DMSP of HiDPs with little regulation suggests constant, constitutive production by this group. By contrast, the active regulation of intracellular DMSP in response to nutrient limitation by LoDPs appears to be an acute stress response (Fig. 2), but the mechanism (overflow, antioxidant, or signaling molecule) remains to be confirmed. These contrasting dynamics suggest different physiological roles for DMSP in LoDPs and HiDPs. For example, the strong regulation and low-intracellular concentrations of DMSP in LoDPs are not consistent with DMSP being a major osmolyte for this group. Specifically, we estimate that DMSP could only contribute 0.004–5% of total osmolarity in LoDPs (see Sup- porting Information S2 for calculation details). On the other hand, intracellular DMSP in HiDPs could account for 14–34% of total osmolarity. Similarly, substitution of DMSP for high nitrogen content osmolytes under nitrogen limitation (e.g., glycine betaine; Keller et al. 2004; Bertrand and Allen 2012), could only alleviate 0.6% and 7% of LoDP and HiDP’s cellular nitrogen requirements, respectively (see Supporting Information S2 for calculation details). These calculations sug- gest that, even under nutrient limitation, regulation of DMSP could only serve a significant role in osmotic balance for HiDPs. However, despite a lack of regulation, DMSP could still be serving to protect HiDP cells against environmental stressors, as high intracellular DMSP concentrations in HiDPs may not require the drastic upregulation seen in LoDPs to deal with radical quenching (Sunda et al. 2002; Lavoie et al. 2016). Different physiological roles for DMSP in LoDPs and HiDPs are consistent with recent genomic insight which suggests there may be multiple eukaryotic DMSP synthesis genes. Of the 53 species with DYSB hits in Curson et al. (2018), 46 are either known or related to HiDPs and none of the confirmed LoDPs examined in this study appear to harbor DYSB. The LoDP gene may be related to a candidate protein recently described for DMSP synthesis in Thalassiosira pseudonana In Situ DMSPt (nM) 0 100 120 140 160 Predicted DMSPp (nM) 0 10 20 30 40 50 200 300 400 500 600 700 800 900 1000 1100 % HiDP of Total Community 10 20 30 40 50 60 70 80 20 60 80 40 m = 0.8 y-int = 6 R 2 = 0.3 p-value < 0.001 RMSE = 13nM MRB = -53% Fig. 7. ComparisonofallinsituDMSPtobservationsandcorrespondingDARWINpredictedDMSPpforthesamemonthandlocation.Samplesarecoloredby thepercentofthetotalcommunitycomprisedofHiDPsasestimatedbyDARWIN.SampleswerebinnedbypredictedHiDPcommunitycomposition(0–80%in incrementsof10%)andthemedianpredictedDMSPpandinsituDMSPtperbinisshown(blackdots).Errorbarsrepresent25 th and75 th quartilesofbinned data. Graydotsrepresent insituDMSPtobservations for Antarctica(< 60 ! S) thatwerenotincluded inthe binningasDARWINdoesnot include a Phaeocystis functional type (see main text for details). Reported statistics reflect the significant linear relationship found between the binned in situ and predicted DMSP values.Dashedlinesrepresentthe95%confidenceintervalforthepredictedslopeandy-intercept.Notethebreakinthex-axisat175 nM(1SDabovethemean ofallinsituDMSPtobservations).Themajorityofobservationsarefoundwithintherangeof0–175 nMDMSPt(n = 3824observationsof4201total). -80 -60 -40 -20 0 20 40 60 80 -150 -100 -50 0 50 100 150 70 10 30 50 Percent Increase in DMSPp (%) Fig. 8. Global annual mean percent increase in DMSPp due to incorpo- ration of the differential regulation of the nutrient stress mechanism in addition to community composition. Gray regions represent areas where predicted DMSPp < 1 nM and therefore created inflated percent increases due to nutrient stress. McParland and Levine Driving mechanisms of global surface DMSP 12 (TpMMT), which exhibited high homology with two putative methyltransferases in the LoDPs Phaeodactylum tricornutum and Thalassiosira oceanica (Kageyama et al. 2018). As these three diatoms are currently the only algae known to contain homologs of TpMMT, it is unknown if TpMMT is environ- mentally relevant or if there are additional DMSP synthesis genes that have yet to be identified. While an intriguing hypothesis, further studies are needed to understand if differ- ent DMSP synthesis genes explain the differential regulation of DMSP by LoDPs and HiDPs and to determine the DMSP synthesis genes of other organisms not predicted to contain dysB, DYSB, or TpMMT (e.g., Trichodesmium). Future studies incorporating direct measurements of DMSP synthesis (Stefels et al. 2009; Archer et al. 2017) with genomic measurements (Curson et al. 2017, 2018) will help decipher the physiological mechanisms behind differential LoDP and HiDP regulation. While constraining the physiological mechanisms of DMSP for different phytoplankton types is critical for understanding the ecological roles of DMSP, simply differentiating between the two different types of DMSP producers was sufficient for reproducing the majority of observed spatial and temporal dynamics of in situ DMSP concentrations. In particular, our calculations captured the observed decoupling between Chl a and DMSPp in a low-productivity region (Fig. 4), which pre- vious models struggle to replicate. The model presented here provides two important improvements on our ability to pre- dict global DMSPp. First, we apply a single set of equations to multiple oceanographic regimes and do not rely on the defini- tion of different biogeographical regimes or thresholds for a stress response. This is especially critical for future predictions as these thresholds might change with shifts in the climate. Second, we use the observed responses from a large number of laboratory monoculture experiments to model intracellular DMSP regulation due to nutrient stress in multiple phyto- plankton types. Most previous models either used fixed DMSP : carbon ratios or a single relationship between cellular DMSP and environmental stress (light, nutrients, and/or tem- perature) for all phytoplankton types (Le Clainche et al. 2010; Vogt et al. 2010). By including a mechanistic representation of the differen- tial regulation of DMSP in response to nutrient stress, we were able to quantitatively evaluate the relative importance of nutrient stress in determining variability in DMSPp concentra- tions (Fig. 8). We demonstrate that nutrient stress most likely plays a minor role in determining temporal and spatial pat- terns of in situ DMSPp concentrations (Fig. 5). Other environ- mental stressors have been shown to impact DMSP production through upregulation (e.g., temperature stress) or downregulation (e.g., acute UV stress) but were not included in our calculations due to insufficient data. A survey of the current state of knowledge on the impact of these other stressors suggests the response is comparable to that observed for nutrients (Supporting Information S1), and including these other stressors would not alter our conclusion that community composition is the primary driver of variability in DMSPp concentrations. The dominant role of community composition in predict- ing DMSPp found in this study may appear contradictory to previous studies that required environmental stress parameter- izations to predict in situ DMSP in oligotrophic regions. How- ever, we hypothesize that the impact of environmental stress implicitly accounted for shifts in community composition. For example, many studies used light stress to improve in situ DMSP predictions in oligotrophic regions. However, seasonal changes in light stress co-occur with community composition shifts (e.g., Polimene et al. (2012)) and thus implicitly incor- porate this dominant mechanism. Similarly, the remote-sens- ing-based algorithms developed by Galí et al. (2015) divide DMSPt production into two different regime types (stratified and mixed water columns), which also implicitly included shifts in community composition. These methodologies were overall successful, but an explicit representation of HiDP and LoDP abundance used here provides a mechanistic representa- tion of these dynamics and improved predictions of temporal dynamics in oligotrophic regions. This work highlights that the primary challenge for robust predictions ofin situ DMSP is accurately capturingthe dynam- ics of the subdominant phytoplankton community, specifi- cally that of the HiDP phytoplankton types. This is a significant challenge both for numerical ecosystem models and for remote-sensing algorithms which have been shown to successfully capture the dominant phytoplankton type (e.g., Alvain et al. 2008; Dutkiewicz et al. 2015) but struggle to accurately predict the biomass of groups that make up a small fraction of the community. This is particularly problematic in the DARWIN oligotrophic and polar (> 60 ! ) communities, where the dominant DMSP producers are outcompeted. Slight changes in HiDP biomass for these regions would result in large changes in watercolumn DMSP. Forexample, to produce a water column of 15 nM DMSPp, a drop of seawater (1 μL) would require 15 small (48 μm 3 ) diatom cells (γ = 0.01, 22 mM intracellular DMSP), but less than 1 cell of the same size coccolithophore (γ = 0.03, 354 mM intracellular DMSP). The large range of intracellular DMSP content within species of the same genus (e.g., Franklin et al. 2010, range from 174 mM DMSP to 715 mM DMSP) further complicates this estimate as species or even strain level shifts could signifi- cantly impact in situ DMSP. While the impact of the subdominant HiDP community on in situ DMSP production has been observed before, there has been no direct assessment of its role as a driving mechanism in DMSP predictions in global ecosystem models or satellite algorithms. Masotti et al. (2010) encountered similar difficul- ties with poorly resolved subdominant communities when investigating global trends in DMS. The authors found that certain regions dominated by LoDP phytoplankton types according to the PhySAT algorithm were associated with unex- pectedly high, in situ measurements of DMS : Chl a ratios and McParland and Levine Driving mechanisms of global surface DMSP 13 concluded that these ratios must be driven by the subdomi- nant HiDP groups. Here, our calculations provide direct sup- port for this finding. Phytoplankton types designed to represent DMSPp dynamics would greatly improve the predic- tion power of DMS(P) in global models (Stefels et al. 2007; Galí et al. 2015), but only if the model was able to accurately capture the coexistence of these groups when they make up a small fraction of the biomass. Similarly, this analysis suggests that improved remote-sensing-based algorithms that robustly estimate the subdominant community will significantly improve our ability to predict DMSPp using algorithms similar to Galí et al. (2015). Conclusion The ability to synthesize DMSP is widespread throughout the tree of life and includes both eukaryotes and prokaryotes (Fig. 1). Furthermore, our analysis suggests that LoDPs are most common and might outnumber HiDPs. Our meta- analysis provides evidence for the hypothesis presented by Stefels et al. (2007) that there are two distinct mechanisms for DMSP regulation. Specifically, HiDPs appear to maintain con- stitutive DMSP production while LoDPs actively regulate intra- cellular DMSP in response to nutrient stress. We hypothesize that different DMSP synthesis genes in the two groups may encode these differences. Using the observed upregulation of DMSP in response to nutrient stress, we tested the hypothesis that LoDPs might contribute significantly to in situ DMSP (Bucciarelli et al. 2013). We show that LoDPs cannot produce environ- mentally relevant DMSPp concentrations, even in regions of very high productivity (e.g., Palmer) or regions of extreme nutrient stress (e.g., oligotrophic regions). While less environ- mentally relevant on a global scale, the maintenance of the genetic machinery for DMSP synthesis by LoDPs suggests that DMSP serves an important physiological function in these cells. In particular, significant changes in DMSP due to nutri- ent stress in oligotrophic environments indicate that DMSP may be required for survival if limited resources are being used to upregulate DMSP synthesis. DMSP is known to play an important role in microscale organic carbon and sulfur cycling within the phycosphere and could serve as a signaling mole- cule in the open oceans (Seymour et al. 2010, 2017; Johnson et al. 2016). It is plausible that the environmental significance of DMSP production by LoDPs is confined to a much smaller scale within the phycosphere, where LoDPs “trade” DMSP for limiting nutrients with associated heterotrophic bacteria (Amin et al. 2015; Seymour et al. 2017). A similar relationship has been observed for a different organosulfur compound (2,3-dihydroxypropane-1-sulfonate) both in monocultures and in situ (Durham et al. 2015, 2017), but remains to be observed for DMSP. The inherently higher intracellular DMSP of HiDPs, particu- larly coccolithophores, dinoflagellates, and Phaeocystis, dominates global DMSPp production even when these groups are the subdominant community. This study highlights the importance of accurately representing the subdominant community in order to accurately capture in situ DMSPp dynamics.Even with the 35 phytoplankton types in DARWIN, over-estimation of the biomass of the dominant phytoplank- ton type added uncertainty to DMSPp predictions. Future work to better capture the dynamics of subdominant commu- nities will be critical for improving predictions of both DMSPp dynamics and, most likely, other biogeochemically important metabolites whose cycling is primarily driven by a small frac- tion of the community. References Alvain, S., C. Moulin, Y. Dandonneau, and H. Loisel. 2008. Sea- sonal distribution and succession of dominant phytoplank- ton groups in the global ocean: A satellite view. Global Biogeochem.Cycles22:1–15. doi:10.1029/2007GB003154 Amin, S. A., and others. 2015. Interaction and signalling between acosmopolitan phytoplanktonand associatedbac- teria. Nature522: 98–101. doi:10.1038/nature14488 Andersen, R., R. Bidigare, M. Keller, and M. Latasa. 1996. A comparison of HPLC pigment signature and electron microscopic observations for oligotrophic waters of the North Atlantic and Pacific Oceans. Deep-Sea Research II43: 517–537. doi:10.1016/0967-0645(95)00095-X Archer, S. D., M. Ragni, R. Webster, R. L. Airs, and R. J. Geider. 2010. Dimethylsulfoniopropionate and dimethylsulfide production in response to photoinhibition in Emiliania huxleyi. Limnol. Oceanogr. 55: 1579–1589. doi:10.4319/ lo.2010.55.4.1579 Archer, S. D., G. A. Tarran, J. A. Stephens, L. J. Butcher, and S. A. Kimmance. 2011. Combining cell sorting with gas chromatography to determine phytoplankton group- specific intracellular dimethylsulphoniopropionate. Aquat. Microb. Ecol.62: 109–121. doi:10.3354/ame01464 Archer, S. D., J. Stefels, R. L. Airs, T. Lawson, T. J. Smyth, A. P. Rees, and R. J. Geider. 2017. Limitation of dimethylsulfo- niopropionate synthesis at high irradiance in natural phy- toplankton communities of the tropical Atlantic. Limnol. Oceanogr.63: 227–242. doi:10.1002/lno.10625 Behrenfeld,M.,J.Hardy,H.Gucinski,A.Hanneman,H.Lee,andA. Wones. 1993. Effects of ultraviolet-B radiationonprimarypro- duction along latitudinal transects in the South PacificOcean. Mar. Environ. Res. 35:349–363. doi:10.1016/0141-1136(93) 90102-6 Behrenfeld, M. J., E. Boss, D. A. Siegel, and D. M. Shea. 2005. Carbon-based ocean productivity and phytoplankton phys- iology from space. Global Biogeochem. Cycles 19:1–14. doi:10.1029/2004GB002299 Belviso, S., and others. 2012. DMS dynamics in the most oli- gotrophic subtropical zones of the global ocean. Biogeo- chemistry110: 215–241. doi:10.1007/s10533-011-9648-1 McParland and Levine Driving mechanisms of global surface DMSP 14 Bertrand,E.M.,andA.E.Allen.2012.InfluenceofvitaminBaux- otrophy on nitrogen metabolism in eukaryotic phytoplank- ton.Front.Microbiol.3:1–16.doi:10.3389/fmicb.2012.00375 Bronk, D. A., P. Glibert, and B. B. Ward. 1994. Nitrogen uptake, dissolved organic nitrogen release, and new production. Sci- ence265: 1843–1846. doi:10.1126/science.265.5180.1843 Bucciarelli, E., C. Ridame, W. G. Sunda, C. Dimier-Hugueney, M. Cheize, and S. Belviso. 2013. Increased intracellular con- centrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium ery- thraeum. Limnol. Oceanogr. 58: 1667–1679. doi:10.4319/ lo.2013.58.5.1667 Burki, F. 2014. The eukaryotic tree of life from a global phylo- genomic perspective. Cold Spring Harb. Perspect. Biol. 6: 1–17. doi:10.1101/cshperspect.a016147 Carlson, C. A., and D. A. Hansell. 2014. DOM sources, sinks, reactivity, and budgets, p. 65–126. In D. Hansell and C. Carlson [eds.], Biogeochemistry of marine dissolved organic matter, 2nd ed. Elsevier Inc. doi:10.1016/B978-0-12- 405940-5.00003-0 Charlson, R., J. Lovelock, M. Andrae, and S. Warren. 1987. Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Nature326: 655–661. doi:10.1038/326655a0 Corn, M., S. Belviso, F. Partensky, N. Simon, and U. Christaki. 1996. Origin and importance of picoplanktonic DMSP. p. 191–201. In R. P. Kiene, P. T. Visscher, M. D. Keller, G. O. Kirst [eds.], Biological and environmental chemistry of DMSP and related sulfonium compounds. Springer. Curson, A. R., and others. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nat. Microbiol. 2:1–9. doi: 10.1038/nmicrobiol.2017.9 Curson, A. R., and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phyto- plankton. Nat. Microbiol. 3: 430–439. doi:10.1038/ s41564-018-0119-5 del Valle, A., D. Slezak, C. M. Smith, A. N. Rellinger, D. J. Kieber, and R. P. Kiene. 2011. Effect of acidification on preservation of DMSP in seawater and phytoplankton cultures : Evidence for rapid loss and cleavage of DMSP in samples containing. Mar.Chem.124:57–67. doi:10.1016/j.marchem.2010.12.002 Durham, B. P., and others. 2015. Cryptic carbon and sulfur cycling between surface ocean plankton. Proc. Natl. Acad. Sci. USA112: 453–457. doi:10.1073/pnas.1413137112 Durham, B. P., S. P. Dearth, S. Sharma, S. A. Amin, C. B. Smith, S. R. Campagna, E. V. Armbrust, and M. A. Moran. 2017. Recognition cascade and metabolite transfer in a marine bacteria-phytoplankton model system. Environ. Microbiol.19: 3500–3513. doi:10.1111/1462-2920.13834 Dutkiewicz, S., A. E. Hickman, O. Jahn, W. W. Gregg, C. B. Mouw, and M. J. Follows. 2015. Capturing optically impor- tant constituents and properties in a marine biogeochemi- cal and ecosystem model. Biogeosciences 12: 4447–4481. doi:10.5194/bg-12-4447-2015 Dyhrman, S., J. Ammerman, and B. Van Mooy. 2007. Microbes and the marine phosphorus cycle. Oceanography 20: 110–116. doi:10.5670/oceanog.2007.54 Franklin, D. J., M. Steinke, J. Young, I. Probert, and G. Malin. 2010. Dimethylsulphoniopropionate (DMSP), DMSP lyase activity (DLA) and dimethylsulphide (DMS) in 10 species of coccolithophore. Mar. Ecol. Prog. Ser. 410: 13–23. doi: 10.3354/meps08596 Galí, M., E. Devred, M. Levasseur, S. J. Royer, and M. Babin. 2015. A remote sensing algorithm for planktonic dimethyl- sulfoniopropionate (DMSP) and an analysis of global pat- terns. Remote Sens. Environ. 171: 171–184. doi:10.1016/j. rse.2015.10.012 Galí, M., M. Levasseur, E. Devred, R. Simó, and M. Babin. 2018. Sea-surface dimethylsulfide (DMS) concentration from satellite data at global and regional scales. Biogeos- ciences15: 3497–3519. doi:10.5194/bg-2018-18 Gorham, J. 1984. Separation of plant betaines and their sul- phur analogues by cation-exchange high-performance liq- uid chromatography. J. Chromatogr. 287: 345–351. doi: 10.1016/S0021-9673(01)87710-4 Hashioka, T., and others. 2013. Phytoplankton competition during the spring bloom in four plankton functional type models. Biogeosciences 10: 6833–6850. doi:10.5194/bg- 10-6833-2013 Herrmann, M., and others. 2012. Diagnostic modeling of dimethylsulfide production in coastal water west of the Antarctic Peninsula. Cont. Shelf Res. 32: 96–109. doi: 10.1016/j.csr.2011.10.017 Johnson, W. M., M. C. K. Soule, and E. B. Kujawinski. 2016. Evidence for quorum sensing and differential metabolite production by the marine heterotroph, Ruegeria pomeroyi, in response to DMSP. ISME J. 10: 2304–2316. doi:10.1038/ ismej.2016.6 Kageyama, H., Y. Tanaka, A. Shibata, R. Waditee-Sirisattha, and T. Takabe. 2018. Dimethylsulfoniopropionate biosynthesis in a diatom Thalassiosira pseudonana: Identification of a gene encoding MTHB-methyltransferase. Arch. Biochem. Biophys. 645: 100–106.doi:10.1016/j.abb.2018.03.019 Karsten,U.,K.Kück,C.Vogt,andG.O.Kirst.1996.Dimethylsul- foniopropionate production in phototrophic organisms and itsphysiologicalfunctionsasacryoprotectant,p.143–153.In R.P.Kiene,P.T.Visscher,M.D.Keller,G.O.Kirst[eds.],Bio- logicalandenvironmentalchemistryofDMSPandrelatedsul- foniumcompounds.Springer. Kearse, M., and others. 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization andanalysisofsequencedata.Bioinformatics28:1647–1649. Keller,M.D.1989.Dimethylsulfideproductionandmarinephy- toplankton: The importance of species composition and cell size. Biol. Oceanogr.6: 375–382. doi:10.1080/01965581. 1988.10749540 Keller, M. D., W. K. Bellows, and R. R. L. Guillard. 1989. Dimethyl sulfide production in marine phytoplankton, McParland and Levine Driving mechanisms of global surface DMSP 15 p. 167–182. In E. S. Saltzman and W. J. Cooper [eds.], Bio- genic sulfur in the environment. American Chemical Society. Keller, M. D.,P.A.Matrai,R.P.Kiene,andW. K.Bellows. 2004. Responses of coastal phytoplankton populations to nitrogen additions: Dynamics of cell-associated dimethylsulfoniopro- pionate (DMSP), glycine betaine (GBT), and homarine. Can. J. Fish.Aquat.Sci.61:685–699.doi:10.1139/f04-058 Kettle, A. J., and others. 1999. 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. Cycles 13: 399–444. doi: 10.1029/1999GB900004 Kido Soule, M. C., K. Longnecker, W. M. Johnson, and E. B. Kujawinski. 2015. Environmental metabolomics: Analytical strategies. Mar. Chem. 177: 374–387. doi:10.1016/j. marchem.2015.06.029 Kiene, R. P., and S. K. Service. 1991. Decomposition of dis- solved DMSP and DMS in estuarine waters: Dependence on temperature and substrate concentration. Mar. Ecol. Prog. Ser.76:1–11. doi:10.3354/meps076001 Kiene, R. P., L. J. Linn, and J. A. Bruton. 2000. New and impor- tant roles for DMSP in marine microbial communities. J. Sea Res. 43: 209–224. doi:10.1016/S1385-1101(00) 00023-X Kiene, R. P., and D. Slezak. 2006. Low dissolved DMSP concen- trations in seawater revealed by small volume gravity filtra- tion and dialysis sampling. Limnol. Oceanogr.: Methods 4: 80–95. doi:10.4319/lom.2006.4.80 Kozlowski, W. A., D. Deutschman, I. Garibotti, C. Trees, and M. Vernet. 2011. An evaluation of the application of CHEMTAX to Antarctic coastal pigment data. Deep-Sea Res. Part I Oceanogr. Res. Pap. 58: 350–364. doi:10.1016/j. dsr.2011.01.008 Ksionzek, K. B., K. B. Ksionzek, O. J. Lechtenfeld, S. L. Mccallister, P. Schmitt-kopplin, J. K. Geuer, W. Geibert, and B. P. Koch. 2016. Dissolved organic sulfur in the ocean: Biogeochemistry of a petagram inventory. Science 354: 456–459. doi:10.1126/science.aaf7796 Lana, A., R. Simó, S. M. Vallina, and J. Dachs. 2012. Re- examination of global emerging patterns of ocean DMS con- centration. Biogeochemistry 110: 173–182. doi:10.1007/ s10533-011-9677-9 Lavoie, M., M. Levasseur, and M. Babin. 2015. Testing the potential ballast role for dimethylsulfoniopropionate in marine phytoplankton: A modeling study. J. Plankton Res. 37: 699–711. doi:10.1093/plankt/fbv050 Lavoie, M., M. Levasseur, and W. G. Sunda. 2016. A steady- state physiological model for intracellular dimethylsulfox- ide in marine phytoplankton. Environ. Chem. 13: 212–219. doi:10.1071/EN14221 Lavoie,M.,J.C.Waller,R.P.Kiene,andM.Levasseur.2018.Polar marine diatoms likely take up a small fraction of dissolved dimethylsulfoniopropionate relative to bacteria in oligotrophic environments. Aquat. Microb. Ecol. 81: 213–218.doi:10.3354/ame01871 Le Clainche, Y., and others. 2010. A first appraisal of prognos- tic ocean DMS models and prospects for their use in cli- mate models. Global Biogeochem. Cycles 24:1–13. doi: 10.1029/2009GB003721 Levine, N. M., D. A. Toole, A. Neeley, N. R. Bates, S. C. Doney, and J. W. H. Dacey. 2015. Revising upper-ocean sulfur dynamics near Bermuda: New lessons from 3 years of con- centration and rate measurements. Environ. Chem. 13: 302–313. doi:10.1071/EN15045 Mahadevan, A., E. D’Asaro, C. Lee, and M. J. Perry. 2012. Eddy-driven stratification initiates North Atlantic spring phytoplankton blooms. Science 337: 54–58. doi:10.1126/ science.1218740 Malin, G., S. Turner, P. Liss, P. Holligan, and D. Harbour. 1993. Dimethylsulphide and dimethylsulphoniopropionate in the Northeast Atlantic during the summer coccolitho- phore bloom. Deep-Sea Res. Part I Oceanogr. Res. Pap. 40: 1487–1508. doi:10.1016/0967-0637(93)90125-M Masotti, I., and others. 2010. Spatial and temporal variability of the dimethylsulfide to chlorophyll ratio in the surface ocean: Anassessment based on phytoplanktongroupdomi- nance determined from space. Biogeosciences 7: 3215–3237. doi:10.5194/bg-7-3215-2010 Masotti, I., S. Belviso, L. Bopp, A. Tagliabue, and E. Bucciarelli. 2016. Effects of light and phosphorus on summer DMS dynamicsin subtropicalwaters using a globaloceanbiogeo- chemical model. Environ. Chem. 13: 379–389. doi: 10.1071/EN14265 Massana, R. 2011. Eukaryotic picoplankton in surface oceans. Annu. Rev. Microbiol. 65: 91–110. doi:10.1146/annurev- micro-090110-102903 Menemenlis, D., J.-M. Campin, P. Heimbach, C. Hill, T. Lee, A. Nguyen, M. P. Schodlok, and H. Zhang. 2008. ECCO2: High resolution global ocean and sea ice data synthesis. Mercat. Ocean Q. Newsl.31: 13–21. Moran, M. A., and others. 2016. Deciphering ocean carbon in a changing world. Proc. Natl. Acad. Sci. USA 113: 3143–3151. doi:10.1073/pnas.1514645113 Polimene, L., S. Archer, M. Butenschön, and J. Icarus Allen. 2012. A mechanistic explanation of the Sargasso Sea DMS “summer paradox”. Biogeochemistry 110: 243–255. doi: 10.1007/s10533-011-9674-z Quinn, P. K., D. J. Coffman, J. E. Johnson, L. M. Upchurch, and T. S. Bates. 2017. Small fraction of marine cloud con- densation nuclei made up of sea spray aerosol. Nat. Geosci. 10: 674–679. doi:10.1038/ngeo3003 Scarratt, M. G., M. Levasseur, S. Michaud, G. Cantin, M. Gosselin, and S. J. De Mora. 2002. Influence of phytoplank- ton taxonomic profile on the distribution of dimethylsul- fide and dimethylsulfoniopropionate in the Northwest Atlantic. Mar. Ecol. Prog. Ser. 244: 49–61. doi:10.3354/ meps244049 McParland and Levine Driving mechanisms of global surface DMSP 16 Seymour, J. R., R. Simó, T. Ahmed, and R. Stocker. 2010. Che- moattraction to dimethylsulfoniopropionate throughout the marine microbial food web. Science 329: 342–345. doi: 10.1126/science.1188418 Seymour, J. R., S. A. Amin, J. B. Raina, and R. Stocker. 2017. Zooming in on the phycosphere: The ecological interface for phytoplankton-bacteria relationships. Nat. Microbiol. 2: 1–12. doi:10.1038/nmicrobiol.2017.65 Speeckaert, G., A. V. Borges, W. Champenois, C. Royer, and N. Gypens. 2018. Annual cycle of dimethylsulfoniopropionate (DMSP)anddimethylsulfoxide(DMSO)relatedtophytoplank- ton succession in the Southern North Sea. Sci.Total Environ. 622–623:362–372.doi:10.1016/j.scitotenv.2017.11.359 Spielmeyer, A., and G. Pohnert. 2012. Daytime, growth phase and nitrate availability dependent variations of dimethylsulfoniopropionate in batch cultures of the dia- tom Skeletonema marinoi. J. Exp. Mar. Biol. Ecol. 413: 121–130. doi:10.1016/j.jembe.2011.12.004 Stefels, J. 2000. Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43: 183–197. doi:10.1016/S1385-1101(00) 00030-7 Stefels, J., and M. A. Van Leeuwe. 1998. Effects of iron and light stress on the biochemical composition of antarctic Phaeocystis sp.I.IntracellularDMSPconcentrations.J.Phycol.34:486–495. doi:10.1046/j.1529-8817.1998.340486.x Stefels, J., M. Steinke, S. Turner, G. Malin, and S. Belviso. 2007. Environmental constraints on the production and removal of the climatically active gas dimethylsulphide (DMS) and implications for ecosystem modelling. Biogeo- chemistry83: 245–275. doi:10.1007/s10533-007-9091-5 Stefels, J., J. W. H. Dacey, and J. T. M. Elzenga. 2009. In vivo DMSP-biosynthesis measurements using stable-isotope incorporation and proton-transfer-reaction mass spectrom- etry (PTR-MS). Limnol. Oceanogr.: Methods 7: 595–611. doi:10.4319/lom.2009.7.595 Sunda, W., D. J. Kieber, R. P. Kiene, and S. Huntsman. 2002. An antioxidant function for DMSP and DMS in marine algae. Nature418: 317–320. doi:10.1038/nature00851 Thomas, M. A., P. Suntharalingam, L. Pozzoli, S. Rast, A. Devasthale, S. Kloster, J. Feichter, and T. M. Lenton. 2010. Quantification of DMS aerosol-cloud-climate interactions usingtheECHAM5-HAMMOZmodelinacurrentclimatesce- nario. Atmos. Chem. Phys.10: 7425–7438. doi:10.5194/acp- 10-7425-2010 Toole, D. A., and D. A. Siegel. 2004. Light-driven cycling of dimethylsulfide (DMS) in the Sargasso Sea: Closing the loop. Geophys.Res.Lett.31:1–4.doi:10.1029/2004GL019581 Tripp, H. J., J. B. Kitner, M. S. Schwalbach, J. W. H. Dacey, L. J. Wilhelm, and S. J. Giovannoni. 2008. SAR11 marine bacte- ria require exogenous reduced sulphur for growth. Nature 452: 741–744. doi:10.1038/nature06776 Turner, S. M., G. Malin, P. S. Liss, D. S. Harbour, and P. M. Holligan. 1988. The seasonal variation of dimethylsulfide and dimethylsulfoniopropionate concentrations in near- shore waters. Limnol. Oceanogr. 33: 364–375. doi:10.4319/ lo.1988.33.3.0364 vanRijssel,M.,andW.W.C.Gieskes.2002.Temperature,light, andthedimethylsulfoniopropionate(DMSP)contentof Emi- liania huxleyi (Prymnesiophyceae). J. Sea Res.48: 17–27. doi: 10.1016/S1385-1101(02)00134-X Vila-Costa, M., R. Simó, H. Harada, J. M. Gasol, D. Slezak, and R. P. Kiene. 2006. Dimethylsulfoniopropionate uptake by marine phytoplankton. Science 314: 652–654. doi: 10.1126/science.1131043 Vogt, M., S. M. Vallina, E. T. Buitenhuis, L. Bopp, and C. Le Quéré. 2010. Simulating dimethylsulphide seasonality with the Dynamic Green Ocean model PlankTOM5. J. Geophys. Res. Oceans115:1–21. doi:10.1029/2009JC005529 Wang, S., S. Elliott, M. Maltrud, and P. Cameron-Smith. 2015. InfluenceofexplicitPhaeocystisparameterizationsontheglobal distribution of marine dimethyl sulfide. J. Geophys. Res. Bio- geosci.120:2158–2177.doi:10.1002/2015JG003017 Ward,B.A.,S.Dutkiewicz,O.Jahn,andM.J.Follows.2012.Asize- structuredfood-webmodelfortheglobalocean.Limnol.Ocea- nogr.57:1877–1891.doi:10.4319/lo.2012.57.6.1877 Acknowledgments We thank S. Dutkiewicz and O. Jahn for the DARWIN model output (a product of NSF grant OCE-1434007). We acknowledge K. Mackey, J. Archibald, D. Caron, J. Kling, D. Hutchins, and F. Fu for monoculture isolates, E. Kujawinski and K. Soule for TSQ-MS analyses, M. Saito and R/V Falkor crew for in situ DMSPt samples, and finally, the BATS and Palmer LTER scientists and staff for providing freely available data. We also thank M. Lee, E. Webb, and E. Zakem for insightful discussions and manuscript comments. We are grateful for the constructive feedback of two anony- mous reviewers. This work was supported by funding from the Rose Hills Foundation, the University of Southern California, and a National Defense Science and Engineering Graduate Fellowship. Conflict of Interest None declared. Submitted 6 May 2018 Revised 4 October 2018 Accepted 10 October 2018 Associate editor: M. Dileep Kumar McParland and Levine Driving mechanisms of global surface DMSP 17 ! 39! Supplementary Information The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations Erin McParland 1a and Naomi M. Levine 1b 1 Department of Marine and Environmental Biology, University of Southern California, Los Angeles, CA, USA a mcparlan@usc.edu b n.levine@usc.edu Correspondence to: Erin McParland Department of Marine and Environmental Biology University of Southern California 3616 Trousdale Parkway, AHF 107 Los Angeles, CA 90089 Email: mcparlan@usc.edu Phone: 617-201-7038 ! 40! Supplementary Information S1: Meta-analysis of other stressor studies! Previous studies looking at the response of DMSP production to non-nutrient stressors, including temperature, light and global change stressors (increased temperature and/or increased CO 2 ), were also analyzed as part of this study. However, due to the significant bias towards high DMSP producer responses, it was not possible to conduct a comprehensive comparison between low producers (LoDPs) and high producers (HiDPs) for non-nutrient stressors. Below, we analyze the responses of the HiDPs only for these studies. Overall, changes in HiDPs intracellular DMSP in response to these other stressors were similar to those observed for nutrient stress (i.e. relatively little change). Most evidence for the role of temperature in DMSP regulation comes from pure enzyme extracts, where DMSP has been shown to stabilize enzymes at low temperatures suggesting that DMSP might be upregulated at low temperatures as a cryoprotectant (Nishiguchi and Somero 1992, Karsten et al. 1996, Stefels 2000). To our knowledge, only two studies have reported the effect of temperature on intracellular DMSP (normalized to cell volume) in monoculture experiments. van Rijssel and Gieskes (2002) assessed the accumulation of intracellular DMSP at lower temperatures in Emiliania huxleyi (E. huxleyi), whereas McLenon and DiTullio (2012) used high temperature as an oxidative stressor in Symbiodinium microadriaticum. The increase in intracellular DMSP under the stress temperature relative to ‘optimal’ temperature was surprisingly quite similar for both of these very different studies, 1.7 and 1.2 fold change, respectively. Two of the leading hypothesized physiological functions for DMSP, an overflow mechanism and an antioxidant, relate to light stress. Consequentially, there has been a significant number of studies (n=8, (Stefels and Leeuwe 1998; Sunda et al. 2002; van Rijssel and Buma 2002; van Rijssel and Gieskes 2002; Slezak and Herndl 2003; Harada et al. 2009; Archer et al. 2010; Darroch et al. 2015)) that investigated the impact of PAR and UV stress on intracellular DMSP regulation in monocultures. However, these results are heavily biased towards a single HiDP species, E. huxleyi (n=6), such that it was not possible to identify over-arching trends as was done for nutrient stress responses. In general, intracellular DMSP was seen to increase (n=6) in response to UV and/or high PAR, with one E. huxleyi study reporting no change in DMSP content in response to UV doses (van Rijssel and Buma 2002). While cross-comparisons between studies are complicated by different acclimation conditions and light doses, HiDPs under these different light stressors exhibited an average fold change of 1.3 ± 0.2, similar to the nutrient stress response. Only ! 41! one light stressor study has reported the response of a LoDP, where UV stress resulted in a 3-fold decrease (p=0.004) in intracellular DMSP by T. oceanica over 60 hours relative to the PAR only treatment (Harada et al. 2009). Further characterization of LoDPs’ regulation of intracellular DMSP is required to confirm whether there is a consistent trend of a large response in intracellular DMSP to UV stress in LoDPs relative to HiDPs. The impact of global change stressors on intracellular DMSP regulation is the least well described: n=3 studies for HiDPs (Spielmeyer and Pohnert 2012; Arnold et al. 2013; Webb et al. 2016) and n=1 for LoDPs (Spielmeyer and Pohnert 2012). The role of global change variables as ‘stressors’ is complicated and often studies that compare across different species and phytoplankton functional types are inconclusive due to multiple mechanisms at play (Hutchins et al. 2009). For example, Spielmeyer and Pohnert (2012) reported decreased intracellular DMSP under CO 2 and temperature ‘stressors’ for an LoDP. However, increased growth rates were also reported, possibly suggesting that growth rate stressors were alleviated and therefore caused a downregulation of intracellular DMSP. Due to the complex nature of global change ‘stressors’, future studies are required before a comparative analysis can be conducted looking at the response of HiDPs and LoDPs to these variables. While our model predictions of DMSPp only included nutrient stress regulation, we have shown that community composition, not cellular physiological responses, is the primary driver of variability in in situ DMSPp concentrations (main text, Figure 5 and Figure 7). Though we lack a sufficient number of studies for LoDP response to other stressors to provide a prediction for the magnitude of the response, even if strong regulation was observed (similar to the nutrient stress response) it would still not be possible for this group to contribute significantly to in situ DMSP concentrations due to the large contribution of HiDPs. Explicitly including the response of HiDPs to non-nutrient stressors would in general increase our predictions of DMSPp. However, given the low-fold changes by HiDPs in response to temperature and light stressors, we do not believe that including these compounding stressors would significantly change our results. Downregulation under acute UV stress, or abiotic oxidation, (Archer et al. 2010; Galindo et al. 2016), is another mechanism that is not included in our predictions. However, based on the small 5% decrease of cellular DMSP reported by Archer et al. (2010), we do not expect this response to significantly change our predictions of DMSPp. ! 42! Supplementary Information S2: Osmolarity calculations Marine phytoplankton live in an environment of high water potential (~1013 osmol!m -3 , (Boyd and Gradmann 2002) and therefore need to maintain mechanisms that will quickly re- equilibrate intracellular osmolarity and maintain cellular homeostasis when external changes cause diffusion across the cell membrane (Bisson and Kirst 1995; Stefels 2000). Previously measured intracellular ionic and organic osmolyte concentrations vary widely across species (Keller et al. 1999; Boyd and Gradmann 2002; Gebser and Pohnert 2013) and the internal mass balance of intracellular osmolarity is still not fully understood (Raven and Doblin 2014). Intracellular osmolarity accounted for by the primary ions, K + , Na + and Cl - , ranges across species from 497 to 1188 osmol!m -3 , with a median of 895 osmol!m -3 (Boyd and Gradmann 2002). Using this median value of ionic osmolarity and assuming an intracellular osmolarity 20% higher than seawater (Sikes and Wilbur 1982; Lavoie et al. 2015), we found that 74% of total intracellular osmolarity would be accounted for by these primary ions. This would require organic osmolytes to account for an additional ~320 osmol!m -3 (Supplementary Table 6). The intracellular DMSP concentrations of representative high DMSP producers (E. huxleyi, C. leptoporus, P. antarctica) (HiDPs) could account for 55-128% of this organic osmolarity, or 14-34% of total intracellular osmolarity, suggesting that DMSP could serve as an osmolyte in these species. In representative low DMSP producers (Trichodesmium, Thalassiosira, Skeletenoma) (LoDPs), the contribution of DMSP to organic osmolarity was much lower ranging from 0.02-20%, or 0.004-5% of total intracellular osmolarity. Even in T. pseudonana under CO 2 limitation, when intracellular DMSP increased ~70 fold to 64mM, DMSP would only account for 20% of organic osmolarity. These minimal contributions to intracellular osmolarity challenge the hypothesis that the primary function of DMSP in LoDPs is as a major osmolyte. However, if the bulk intracellular DMSP concentrations actually reflect accumulation of DMSP in specific cellular sites including vacuoles, chloroplasts and the cytoplasm (Raina et al. 2017), DMSP could significantly contribute to cellular osmolarity even at these much lower concentrations (Lyon et al. 2016). Complementary to the osmolyte function is the hypothesis that, rather than acting in response to osmotic shock, DMSP is substituted for high nitrogen content osmolytes such as glycine betaine (Keller et al. 2004; Bertrand and Allen 2012). For every DMSP substituted, one mole of nitrogen would be alleviated for the cell. This scenario is most plausible for HiDPs, where, for example, a cellular quota of 5 fmole DMSP·cell -1 in E. huxleyi would alleviate 70 fg N·cell -1 , ! 43! or 7% of its organic nitrogen content (Keller et al. 1999). In comparison, a cellular quota of 0.5 fmole DMSP·cell -1 in the LoDP Thalassiosira pseudonana would only alleviate 7 fg N·cell -1 , or 0.6% of its organic nitrogen content (Keller et al. 1999). However, evidence for a direct relationship demonstrating this hypothesis is conflicting. Keller et al. (1999) found an inverse relationship between DMSP and the nitrogen containing osmolytes homarine and glycine betaine in chemostat cultures of T. pseudonana, but no relationship in E. huxleyi or A. carteraea, which maintained much higher concentrations of DMSP than any other osmolyte. This relationship is even more complicated in mixed communities, where the physiological history (rather than absolute nitrate concentration) appears to regulate changes in nitrogen containing osmolytes and DMSP (Keller et al. 2004). Finally, while nitrate limitation caused the greatest upregulation of DMSP by T. pseudonana, phosphate, silicate and carbon dioxide limitation also caused an increase in DMSP production, suggesting the nitrogen osmolyte substitution hypothesis cannot be the only driver of DMSP production (Bucciarelli et al. 2003). ! 44! Supplementary Figures and Tables Supp Figure 1: Individual fits for each LoDP species in the meta-analysis, where intracellular DMSP is a function of growth limitation due to nutrient stress (γ). Colors indicate different nutrient limitations (purple = iron, blue = nitrate, green = senescence, orange = phosphate, pink = silicate) and shapes indicate different species. If a sigmoidal fit was used and there were ≥ 4 points, the R 2 value is reported. The intracellular DMSP of picoprokaryotes (top, left panel) has only been reported in replete conditions. The intracellular DMSP concentrations for limited conditions were extrapolated based on observed responses in Trichodesmium (see Methods in main text). 10 20 Picoprokaryotes Corn et al. 1996 Prochlorococcus Synechococcus 10 30 50 Skeletonema marinoi Spielmeyer and Pohnert 2012 0 20 30 Thalassiosira oceanica Bucciarelli et al. 2013 5 10 15 0 10 20 30 10 20 1 3 5 7 0 20 40 2 3 4 8 12 γ (Growth Limitation) 1 10 20 30 r 2 = 0.97 0 r 2 = 0.97 r 2 = 0.98 r 2 = 0.99 Thalassiosira oceanica Harada et al. 2009 Thalassiosira pseudonana Franklin et al. 2012 Thalassiosira pseudonana Keller et al. 1999 0 Thalassiosira pseudonana Kettles et al. 2014 Thalassiosira pseudonana Sunda et al. 2002 Thalassiosira pseudonana Bucciarelli and Sunda 2003 1 0 γ (Growth Limitation) γ (Growth Limitation) 3 2 1 0 Trichodesmium Bucciarelli et al. 2013 Intracellular DMSP (mM) Intracellular DMSP (mM) Intracellular DMSP (mM) Intracellular DMSP (mM) Thalassiosira pseudonana Bucciarelli and Sunda 2003 Thalassiosira pseudonana Bucciarelli and Sunda 2003 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Increasing nutrient stress ! 45! Supp Figure 2: Individual fits for each HiDP species in the meta-analysis, where intracellular DMSP is a function of growth limitation due to nutrient stress (γ). Colors indicate different nutrient limitations (purple = iron, blue = nitrate, red = k/50) and shapes indicate different species. All were fit with a linear regression. 260 300 340 380 150 250 350 Phaeocystis antarctica Stefels and van Leeuwe 1998 116 120 124 Amphidinium carterae Keller et al. 1999 50 70 90 110 130 180 190 200 Emiliania huxleyi Sunda et al. 2007 100 300 500 700 Calcidiscus leptoporus Franklin et al. 2010 520 560 600 Gephyrocapsa oceanica Franklin et al. 2010 290 310 330 350 Umbilicosphaera sibogae Franklin et al. 2010 280 290 300 310 320 Coccolithus braarudii Franklin et al. 2010 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Phaeocystis antarctica Kinsey et al. 2016 Emiliania huxleyi Keller et al. 1999 Intracellular DMSP (mM) γ (growth limitation) γ (growth limitation) γ (growth limitation) Intracellular DMSP (mM) Intracellular DMSP (mM) Increasing nutrient stress ! 46! Supp Figure 3: Ratio of particulate organic carbon (μg·L -1 ) to chlorophyll (μg·L -1 ) measured at BATS (dashed line). As higher ratios may be due to heterotrophic biomass, the maximum POC: Chl a ratio was bound by the mean phytoplankton C: Chl a found by Marañón 2005 and used to convert HPLC Chl a to carbon (solid line). The seasonal signal in POC: Chl a at BATS is expected to be driven primarily by phototrophs as heterotrophic biomass has been shown to be constant across seasons (Carlson et al. 1996). J M J A O D C:Chla Ratio 100 300 500 700 900 1100 BATS F A M J S N ! 47! Supp Figure 4: Comparison between predicted DMSPp and in situ measured DMSPp at (a) Palmer and (b) BATS. In situ measurements are shown in black and DARWIN DMSPp is shown in red. DMSPp was predicted as a function of Chl a by performing a linear regression of in situ measured DMSPp and Chl a. The resulting slope was used to predict in situ DMSPp (green). The satellite derived algorithms of Galí et al. (2015) were also used to predict in situ DMSPp (blue). The mixed algorithm (ratio of euphotic depth:mixed layer depth <1) was used for Palmer and the stratified algorithm (ratio >1) for BATS. The mixed algorithm requires knowledge of the euphotic and mixed layer depth climatologies. We estimated the euphotic depth as the mean 0.5% PAR depth reported for each month at Palmer Station B (1% PAR depth is preferable (Galí et al. 2015) but was not available for this dataset). Mixed layer depth data for the times DMSPp was measured were not available and therefore we estimated the depth from ARGO float profiles presented in Dong et al. (2008). While our resulting ratio for January was well within the bounds of ratios measured at this site during the same time (Huang et al. 2012), we note that any apparent errors in DMSPp at Palmer Station predicted with the Galí mixed algorithm could certainly be caused by our estimate of the euphotic:mixed layer depth ratio. DMSP (nM) 0 100 200 300 400 500 600 Palmer In Situ DARWIN Galí Chl a J F M A M J J A S O N D Nov Dec Jan Feb March In Situ DARWIN Galí Chl a (a) (b) 0 5 10 15 20 25 30 DMSP (nM) BATS ! 48! Supp Figure 5: Cumulative HiDP cellular biovolume (μL cell·L -1 ) versus in situ measured DMSPp (nM) at BATS. Cumulative cellular biovolume concentrations were calculated using HiDP biomass from HPLC pigments (μL cell·L -1 ). Each point is color-coded by upper mixed layer UV light dose (W·m -2 ). While a relationship was observed between cumulative HiDP cellular volume and DMSPp (R 2 =0.3, p-value=0.07), no relationship was observed between DMSP and UV light dose (R 2 =0.02, p-value=0.7). UV Dose (W·m -2 ) 0.01 0.03 0.05 0.07 0.09 0.11 10 14 18 22 26 0.01 0.03 0.05 0.07 HiDP Cellular Volume (µL cell·L -1 ) m = 0.002 R 2 = 0.3 p = 0.07 BATS DMSPp (nM) ! 49! Supp Figure 6: Seasonal cycle of DARWIN predicted DMSPp for three Longhurst provinces: NADR (a), SATL (b), and SANT (c). The predicted DMSPp for the region (mean=blue line, standard deviation=light blue bounds) are compared against in situ measurements (boxplots) for the same provinces binned by month black dots=median, edges of blue boxes = 25 th and 75 th percentile, whiskers = min and max values within 1.5*IQR, blue circles= outliers, greater than 50 100 150 200 250 0 50 100 150 1 150 250 365 0 20 40 60 80 100 120 140 160 North Atlantic 43°-55°N 11°-43°W South Atlantic 33°-40°S 30°-50°W Antarctic/Indian 40°-54°S 22°-75°E 50 Day of Year DMSP (nM) 23 17 9 79 10 87 8 13 16 22 17 3 30 5 25 (a) (b) (c) ! 50! 1.5*IQR). If a boxplot is missing, there were no measurements in the province for that month. The numbers above each box represent the number of observations for the box. Supp Figure 7: Percent contribution of cyanobacteria (picoprokaryote (n=2) and diazotroph (n=5) phytoplankton types) to total mean annual biomass in the DARWIN model. High Eukaryote Archaeplastida Green algae Prasinophyceae Micromonas pusilla DW8 JF794057.1 0.2 161.94 Keller et al. 1989 High Eukaryote Archaeplastida Green algae Prasinophyceae Micromonas pusilla IB 4 0.1 287.31 Keller et al. 1989 High Eukaryote Archaeplastida Green algae Prasinophyceae Pycnococcus provasolii CCMP 1203 AF122889.1 2.4E-01 57 Corn et al. 1996 High Eukaryote Archaeplastida Green algae Prasinophyceae Pycnococcus sp. CCMP 1192 1.5E-01 58 Corn et al. 1996 High Eukaryote Archaeplastida Green algae Prasinophyceae unidentified coccoid VB 1 10.3 156.72 Keller et al. 1989 High Eukaryote Archaeplastida Green algae Prasinophyceae unidentified coccoid VH 2 ax 14.3 126.87 Keller et al. 1989 High Eukaryote Archaeplastida Green algae Prasinophyceae unidentified flagellate BE92 0.2 484.33 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Alexandrium minutum VGO 651 200 Berdalet et al. 2011 High Eukaryote Chromalveolata Alveolates Dinophyceae Amphidinium carterae AMPHI EF057407.1 143.8 2201.5 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Amphidinium carterae CCMP1314 123.3 Keller et al. 1996 High Eukaryote Chromalveolata Alveolates Dinophyceae Brandtodinium nutricula RCC3468 200 490ˆ Gutierrez-Rodriguez et al. 2017 High Eukaryote Chromalveolata Alveolates Dinophyceae Cachonina niei CACH EF492499.1 320.4 192.54 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Crypthecodinium cohnii CCOHNII FJ821501.1 340.6 377.61 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Crypthecodinium cohnii CCMP316 36.2 Keller et al. 1996 High Eukaryote Chromalveolata Alveolates Dinophyceae Gymnodinium simplex CCMP419 346.8 Keller et al. 1996 High Eukaryote Chromalveolata Alveolates Dinophyceae Gymnodinium sp. 94GYR 178.8 124.63 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Gymnoxanthella radiolariae RCC3507 AB920349.1 100 462ˆ Gutierrez-Rodriguez et al. 2017 High Eukaryote Chromalveolata Alveolates Dinophyceae Heterocapsa pygmaea GYMNO EF492500.1 145.3 451.49 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Heterocapsa sp. GT23 582.0 190.3 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Pelagodinium beii RCC1491 JX661028.1 900 272ˆ Gutierrez-Rodriguez et al. 2017 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum micans CCMP691 1781.7 Keller et al. 1996 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum minimum EXUV FJ587221.1 159.5 888.06 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum minimum CCMP1329 284.0 Keller et al. 1996 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum minimum 260 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum sp. IIB 2 b 1 122.2 1082.1 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum sp. M12-11 4419.0 190.3 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Protogonyaulax tamarensis GT429 1974.8 139.55 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Scrippsiella trochoidea PERI AJ415515.1 2861.6 350 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Symbiodinium microadriaticum CCMP2467 283 Curson et al. 2018 High Eukaryote Chromalveolata Alveolates Dinophyceae Symbiodinium microadriaticum HIPP M88521.1 180.3 344.78 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae Thoracosphaera heimii L603 LC054944.1 198.2 194.03 Keller et al. 1989 High Eukaryote Chromalveolata Alveolates Dinophyceae unidentified dinoflagellate DDT 682.6 83.58 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Cylindrotheca closterium CCMP342 2.3 Keller et al. 1996 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Melosira nummuloides MEL3 257.1 264.18 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Melosira nummuloides CCMP482 HQ912566.1 51.9 Keller et al. 1996 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Nitzschia laevis CCMP560 3.6 Keller et al. 1996 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Skeletonema costatum CCMP1332 5.3 Keller et al. 1996 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Skeletonema costatum RCC75 2 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Stramenopile Bacillariophyceae Skeletonema marinoi G4 3 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Stramenopile Chrysophyceae Chrysamoeba sp. IG5 EF165102.1 20.0 596.27 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Chrysophyceae Ochromonas sp. IC 1 U42381.1 2.8 200.75 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Chrysophyceae Ohromonas sp. VT 1 4.3 529.1 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Chrysophyceae Rhizochromulina sp. CCMP1150 U14388.1 10.4 Keller et al. 1996 High Eukaryote Chromalveolata Stramenopile Chrysophyceae unidentified coccoid IVR5ax 1.8 422.39 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Cryptophyceae Cryptochloris sp. 4 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Stramenopile Cryptophyceae Cryptomonas sp. ID 2 AJ420696.1 21.3 345.52 Keller et al. 1989 High Eukaryote Chromalveolata Stramenopile Cryptophyceae unidentified cryptophyte CCMP1178 3.3 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Calcidiscus leptoporus RCC1130 AJ544116.1 96.7 412 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Calcidiscus leptoporus (haploid) RCC1154 32.2 187 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Calcidiscus quadriperforatus RCC1159 483.7 596 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Chrysochromulina ericina NEPCC 109A 28.4 251.49 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Chrysochromulina herdlansis NEPCC 186 AJ246278.1 27.0 412.69 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Chrysochromulina sp. CCMP288 9.7 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Coccolithus braarudii RCC1200 713.7 541 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Coccolithus neohelis CONE AJ246262.1 18.9 85.08 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Coccolithus neohelis CCMP298 55.7 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Emiliania huxleyi CCMP 370 and RCC1216 2.1 295 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Emiliania huxleyi BT6 KC404141.1 5.6 166.42 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Emiliania huxleyi CCMP378 5.9 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Emiliania huxleyi CCMP376 8.4 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Gephyrocapsa oceanica RCC1291, RCC1319,RCCC1313 AJ246276.2 10.7 174 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Helicosphaera carteri RCC1333 AM490983.2 359.0 628 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Imantonia rotunda IIE 6 AJ246267.1 1.3 159.7 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Imantonia rotunda 1197NTA 1.9 87.31 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Imantonia sp. PCC 1851 1.48 149 Corn et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Isochrysis galbana 8701 4.3 Li et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Isochrysis galbana ISO AJ246266.1 3.7 56.87 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Isochrysis galbana 2.8 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Prymnesiophyceae Oolithotus fragilis RCC1376 AM491026.2 172.0 386 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova pinguis CCMP609 7.9 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIB 3 + 1.6 53.36 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIG 3 + 3.4 51.19 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIG 3 ax 3.9 59.18 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIG 6 + 4.8 73.66 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIB 3 ax 4.8 156.72 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova sp. IIG 6 ax 5.4 83.58 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Pavlova viridis 5 Li et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis antarctica 275 Kinsey et al. 2016 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis cordata 12 358 Decelle et al. 2012 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. RCC1383 JN381495.1 10 307ˆ Gutierrez-Rodriguez et al. 2017 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. 1209 7.5 113.43 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. 677-3 17.1 260.45 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. CCMP628 30.5 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. (likely antarctica) 125 Stefels and van Leeuwe 1998 High Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis sp. (likely globosa) 71 Stefels and van Boekel 1993 High Eukaryote Chromalveolata Prymnesiophyceae Pleurochrysis carterae COCCOII AJ544120.1 89.4 170.15 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Prymnesium parvum CCAP946 55 Curson et al. 2018 High Eukaryote Chromalveolata Prymnesiophyceae Prymnesium parvum PRYM AJ246269.1 12.7 111.94 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Prymnesium parvum CCMP708 13.6 Keller et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae Prymnesium parvum 17 Spielmeyer et al. 2011 High Eukaryote Chromalveolata Prymnesiophyceae Umbilicosphaera foliosa RCC1471 47.3 79 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Umbilicosphaera hulburtiana RCC1474 23.6 715 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Umbilicosphaera sibogae RCC1469 105.3 291 Franklin et al. 2010 High Eukaryote Chromalveolata Prymnesiophyceae Umbilicosphaera sibogae L1178 AJ544118.1 102.8 195.52 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae Unidentified CCMP 625 8.9E-01 128 Corn et al. 1996 High Eukaryote Chromalveolata Prymnesiophyceae unidentified coccolithophore 8613COCCO 8.2 125.37 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae unidentified flagellate 3D 11.7 179.1 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae unidentified flagellate 8610C3 23.5 358.21 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae unidentified flagellate 8610G6 25.1 139.55 Keller et al. 1989 High Eukaryote Chromalveolata Prymnesiophyceae unidentified flagellate 326-1 41.4 365.67 Keller et al. 1989 High Eukaryote Rhizaria Chlorarchiniophyte Bigelowiella natans CCMP621 FJ973362.1 1.25 This study High Eukaryote unidentified unidentified green flagellate LOBD 20.9 94.78 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Chlorophyceae Chlamydomonas sp. CCMP231 GQ122368.1 2.24E-03 Keller et al. 1996 Low Eukaryote Archaeplastida Green algae Chlorophyceae Chlorella capsulata OPT10 AY044652.1 4.5 25.22 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Chlorophyceae Chlorella capsulata CCMP246 0.0 Keller et al. 1996 Low Eukaryote Archaeplastida Green algae Chlorophyceae Chlorella sp. 0.3 Li et al. 2010 Low Eukaryote Archaeplastida Green algae Chlorophyceae Chlorococcum sp. Chloro-1 AB058336.1 0.6 0.14 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Chlorophyceae Nannochloris sp. CCMP 515 8.9E-04 0.74 Corn et al. 1996 Low Eukaryote Archaeplastida Green algae Chlorophyceae Platymonas sp. 0.2 Li et al. 2010 Low Eukaryote Archaeplastida Green algae Prasinophycea Micromonas pusilla CCMP490 0.4 Keller et al. 1996 Low Eukaryote Archaeplastida Green algae Prasinophyceae Bathycoccus prasinos type strain JF794058.1 3.4E-02 35.6 Corn et al. 1996 Low Eukaryote Archaeplastida Green algae Prasinophyceae Mantoniella squamata PLY189 X73999.1 1.0 29.18 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae Micromonas pusilla CCMP 490 1.6E-02 17.9 Corn et al. 1996 Low Eukaryote Archaeplastida Green algae Prasinophyceae Nephroselmis pyriformis PLY58 AB058391.1 8.0 36.57 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae Pseudoscourfielda marina IVP 11 AJ132619.1 0.1 5.65 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae Pyraminonas sp. 13-10PYR AB052289.1 0.1 0.53 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae Tetraselmis levis PLATY 1 DQ207405.1 12.1 32.76 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae Tetraselmis sp. OPT4 8.1 45.45 Keller et al. 1989 Low Eukaryote Archaeplastida Green algae Prasinophyceae unidentified coccoid Ω48-23 0.7 47.69 Keller et al. 1989 Low Eukaryote Archaeplastida Red algae Rhodophyceae Rhodosorus marinus RHODO AF168625.1 2.3 35.67 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Ceratium longipes 90201 AF022192.1 14.9 0.23 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Dissodinium lunula L823 FJ473378.1 864.4 1.94 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gambierdiscus toxicus GT200A AB764301.1 1192.3 10.08 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gonyaulax polyedra GP60e AJ415511.1 134.1 4.01 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gonyaulax spinifera W1 1080.5 16.49 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gymndinium nelsoni GSBL 1818.3 29.55 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gymnodinium simplex WT8 JF791031.1 238.5 45.5 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gyrodinium aureolum PLY497A AJ415517.1 5.1 0.36 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Gyrodinium aureolum KT3 5.4 0.65 Keller et al. 1989 Low Eukaryote Chromalveolata Alveolates Dinophyceae Pyrocystis noctiluca CCMP4 AF022156.1 45.2 0.01 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Asterionella formosa SAG 8.95 detected Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Biddulphia sp. L1474 EF192990.1 7.5 0.69 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros didymum L162 AY485449.1 0.3 18.21 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros didymus 0.5 Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros simplex CCMP199 0.2 Keller et al. 1996 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Coscinodiscus sp. COSC1 KC309533.1 77.5 0.17 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Cylindrotheca closterium NCLOST KC899347.1 11.1 41.42 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Ditylum brightwellii L154 AY485444.1 73.1 4.61 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Fragilariopsis cylindrus CCMP 1102 6.7 Curson et al. 2018 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Fragilariopsis cylindrus recent isolate EF140624.1 0.6 15.2 This study Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Minidiscus triocultaus GMe41 DQ093363.1 0.7 32.91 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Minidiscus triocultaus CCMP495 0.8 Keller et al. 1996 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Navicula sp. detected Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Nitzschia closterium 0.6 Li et al. 2010 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Nitzschia laevis O7 KF177775.1 3.9 7.34 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Phaeodactylum tricornutum UTEX646 0.8 Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Porosira glacialis 18 HQ912619.1 69.8 2.09 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Pseudo-nitzschia subcurvata recent isolate GU373970.1 0.2 0.6 This study Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Rhizosolenia setigera RHIZ0 AY485461.1 112.5 0.46 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Skeletonema costatum SKEL EF433519.1 3.7 21.87 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Skeletonema menzellii MEN5 1.0 30.3 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira oceanica AF374479.2 2.8 Bucciarelli et al. 2013 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira pseudonana 3H 0.6 16.64 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira pseudonana CCMP1335 0.5 Keller et al. 1996 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira pseudonana CCMP 1335 0.97 Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira rotula MB411 1.9 1.05 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira sp. PP86A 40.8 1.82 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira weissflogii RCC 76 detected Spielmeyer et al. 2011 Low Eukaryote Chromalveolata Stramenopile Cryptophyceae Rhodomonas salina CCMP1319 HM126532.1 0.1 Keller et al. 1996 Low Eukaryote Chromalveolata Stramenopile Eustigmatophyceae Nannochloropsis GSB Sticho KT031997.1 0.1 22.76 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Pelagophyceae Pelagococcus subviridus PelaCl 2 0.7 29.9 Keller et al. 1989 Low Eukaryote Chromalveolata Stramenopile Pelagophyceae Pelagomonas sp. EUM 8 U14389.1 5.2E-02 23.4 Corn et al. 1996 Low Eukaryote Chromalveolata Pavlovophyceae Exanthemachysis sp. recent isolate 0.6 This study Low Eukaryote Chromalveolata Prymnesiophyceae Chrysochromulina sp. PCC307 0.2 Curson et al. 2018 Low Eukaryote Chromalveolata Prymnesiophyceae Chrysochromulina tobin CCMP 291 0.6 Curson et al. 2018 Low Eukaryote Chromalveolata Prymnesiophyceae Pavlova lutheri MONO 0.4 3.28 Keller et al. 1989 Low Eukaryote Chromalveolata Prymnesiophyceae Pavlova pinguis IG 7 AF102370.1 5.3 46.87 Keller et al. 1989 Low Eukaryote Chromalveolata Prymnesiophyceae Syracosphaera elongata SE62 KF422621.1 147.5 35.3 Keller et al. 1989 Low Eukaryote Rhizaria Radiolarian Acantharia Amphilonche elongata JN811152.1 15200 17.1† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Acantharia Star shape morphotype 36400 2.4† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Acantharia Translucid morphotype 20900 0.3† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Collodaria Collozoum sp. AB690554.1 3652000 0.1† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Collodaria Sphaerozoum sp. AB690556.1 3135000 0.1† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Collodaria Thalassicolla sp. AY266297.1 2757000 0.2† Gutierrez-Rodriguez et al. 2017 Low Eukaryote Rhizaria Radiolarian Foraminifera Globigerinella sp. Z83959.1 37800 7.5† Gutierrez-Rodriguez et al. 2017 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae Chlamydomonas sp. EPT3 BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae Chlorella sp. O17 BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae Dunaliella tertiolecta DUN BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae Nannochloris atomus GSBNanno BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae Stichococcus sp. NB3-18 BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Chlorophyceae unidentified coccoid IVP16AX BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Prasinophyceae Pedinomonas minutissima VA3 BD Keller et al. 1989 Undetermined Eukaryote Archaeplastida Green algae Prasinophyceae Unidentified EUM 16B BD Corn et al. 1996 Undetermined Eukaryote Archaeplastida Red algae Rhodophyceae Porphyridium cruentum PORPH BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Alveolates Dinophyceae Karenia brevis CCMP 2281 detected del Valle et al. 2011 Undetermined Eukaryote Chromalveolata Alveolates Dinophyceae Oxyrrhis marina NEPCC534 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Alveolates Dinophyceae Prorocentrum sp. 86183 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Amphora coffaeformis 47M BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Asterionella glacialis A6 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros affinis CCUR BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros decipiens WTCD BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Chaetoceros simplex BBSM BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Navicula pelliculosa O4 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Phaeodactylum tricornutum PHAEO BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira antarctica CCMP 982 detected del Valle et al. 2011 Undetermined Eukaryote Chromalveolata Stramenopile Bacillariophyceae Thalassiosira guillardii 7-15 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Chloromonadophyceae Chattonella harima WTOMO BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Chloromonadophyceae Chattonella luteus OLISTH BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Chrysophyceae Chrysosphaera sp. UW397 BD <dl Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Chrysophyceae unidentified flagellate MC1 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae Chroomonas salina 3C BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae Cryptomonas sp. PHI BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae Cyanophora paradoxa CY2 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae Proteomonas sulcata CCMP 1175 detected del Valle et al. 2011 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae Rhodomonas lens RLENS BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae unidentified cryptomonad IVF 3 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae unidentified cryptomonad IVP 8 BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Cryptophyceae unidentified cryptomonad 10C BD Keller et al. 1989 Undetermined Eukaryote Chromalveolata Stramenopile Pelagophyceae Pelagococcus sp. detected del Valle et al. 2011 Undetermined Eukaryote Chromalveolata Prymnesiophyceae Phaeocystis globosa CCMP 627 detected del Valle et al. 2011 Undetermined Eukaryote Excavata Euglenophyceae Eutreptia sp. EEUI BD Keller et al. 1989 Low Prokaryote Alphaproteobacteria Purple non-sulfur bacteria Rhodovulum euryhalinum DSM 4868 NR_043406.1 0.3‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple non-sulfur bacteria Rhodovulum sulfidophilum DSM 1374 D16430.1 0.06‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Chromatium gracile DSM 203 X93473.2 0.4‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Chromatium minus DSM 178 Y12372.2 0.6‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Chromatium minutissimum DSM 1376 Y12369.3 0.07‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Chromatium violascens DSM 198 AJ224438.1 0.04‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Thiocapsa pfennigii DSM 226 Y12373.3 0.2‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Purple sulfur bacteria Thiocapsa roseopersicina DSM 5653 Y12364.3 0.1‡ Karsten et al. 1996 Low Prokaryote Alphaproteobacteria Amorphus coralli DSM19760 DQ097300.1 8.2 § 1.3 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Labrenzia aggregata IAM12614 3.6 § 0.6 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Labrenzia aggregata LZB033 EU694387.1 8.4 § 1.3 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Oceanicola batsensis HTCC2597 AY424897.1 40.5 § 6.3 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Pelagibaca bermudensis HTCC2601 DQ178660.1 259 § 40.6 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Sagittula stellata E-37 HG315014.1 11.1 § 1.7 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Sediminimonas qiaohouensis DSM21189 EU878003.1 122 § 19.1 Curson et al. 2017 Low Prokaryote Alphaproteobacteria Thalassobaculum salexigens DSM19539 NR_116122.1 4.8 § 0.8 Curson et al. 2017 Low Prokaryote Cyanobacteria Anabaena sp. Bo 70 AJ133156.1 6.28‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Crocosphaera WH0003 NR_115288.1 Detected This study Low Prokaryote Cyanobacteria Crocosphaera WH0005 Detected This study Low Prokaryote Cyanobacteria Lyngbya aestuarii Bo 9 AB075989.1 4.9‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Microcoleus chthonoplastes MPI MEL-1 0.06‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Microcoleus chthonoplastes MPI EBD-1 GQ402024.1 0.1‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Microcoleus chthonoplastes MPI TOW-1 0.1‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Microcoleus chthonoplastes MPI GNL-1 0.319‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Nostoc sp. Bo 84 EU022731.1 1.17-2.39‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus NATL1 8.2E-07 0.02 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus TATL1 1.2E-06 0.03 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus TATL2 1.2E-06 0.03 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus TATL4 1.4E-06 0.03 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus PAC1A 2.2E-06 0.05 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus SS120 X63140.1 5.3E-06 0.12 Corn et al. 1996 Low Prokaryote Cyanobacteria Prochlorococcus MED 4 AF001466.1 1.6 - 3 E-06 0.03 - 0.06 Corn et al. 1996 Low Prokaryote Cyanobacteria Synechococcus DC2 AF311291.1 7.5E-06 0.03 Corn et al. 1996 Low Prokaryote Cyanobacteria Synechococcus MAX42 AY172805.1 1.0E-04 0.7 Corn et al. 1996 Low Prokaryote Cyanobacteria Synechococcus sp WH307 BD + This study Low Prokaryote Cyanobacteria Synechocystis sp. Bo79 0.14-0.75‡ Karsten et al. 1996 Low Prokaryote Cyanobacteria Trichodesmium erythraeum IMS101 AF013030.1 BD + 0.05 Bucciarelli et al. 2013 Low Prokaryote Cyanobacteria Trichodesmium erythraeum 2175 Detected This study Low Prokaryote Gammaproteobacteria Purple non-sulfur bacteria Rhodopseudomonas salexigens DSM 2132 D12700.1 0.6‡ Karsten et al. 1996 Low Prokaryote Gammaproteobacteria Purple non-sulfur bacteria Rhodopseudomonas sulfoviridis DSM 729 D86514.1 2.2-12.5‡ Karsten et al. 1996 Undetermined Prokaryote Cyanobacteria Synechococcus EUM11 BD Corn et al. 1996 Undetermined Prokaryote Cyanobacteria Synechococcus elongatus CCMP1631 AY946243.1 confirmed Fiore et al. 2015 Undetermined Prokaryote Cyanobacteria Synechococcus sp RCC 555 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp RCC 2673 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp RCC 2369 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp RCC 2553A BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp RCC 2553B BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp WH8102 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp WH8020 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp WH 7803 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp WH8109 BD This study Undetermined Prokaryote Cyanobacteria Synechococcus sp WH8113 BD This study Undetermined Prokaryote Cyanobacteria Synechocystis sp PCC6803 AB041937.1 Detected This study Undetermined Prokaryote Cyanobacteria Trichodesmium thiebautii ND This study Undetermined Prokaryote Cyanophyceae Synechococcus bacillaris SYN BD Keller et al. 1989 Undetermined Prokaryote Cyanophyceae Synechococcus sp. DC 2 BD Keller et al. 1989 Undetermined Prokaryote Cyanophyceae Synechococcus sp. L1602 BD Keller et al. 1989 Undetermined Prokaryote Cyanophyceae Synechococcus sp. L1604 BD Keller et al. 1989 Undetermined Prokaryote Cyanophyceae Synechocystis sp. CN0117 BD Keller et al. 1989 Undetermined Prokaryote Cyanophyceae Trichodesmium sp. MACC0993 BD Keller et al. 1989 Species Strain High or low producer Temp ( C) Light (umol photon/m2/s) Light:Dark Growth Treatment Growth rate (d-1) µ (limited growth/ replete growth) Intracellular DMSP (mM) Franklin et al. 2010 Calcidiscus leptoporus RCC1154 High 17 85 14:10 Batch k/5 vs k/50 0.47 1.00 187 Calcidiscus leptoporus 0.23 0.49 409 Coccolithus braarudii RCC1200 High 17 85 14:10 Batch k/5 vs k/50 0.49 1.00 541 Coccolithus braarudii 0.42 0.86 613 Gephyrocapsa oceanica RCC1291 High 17 85 14:10 Batch k/5 vs k/50 0.34 1.00 174 Gephyrocapsa oceanica 0.32 0.94 350 Helicosphaera carteri RCC1333 High 17 85 14:10 Batch k/5 vs k/50 0.19 1.00 628 Helicosphaera carteri 0.23 1.21 523 Umbilicosphaera sibogae RCC1469 High 17 85 14:10 Batch k/5 vs k/50 0.4 1.00 291 Umbilicosphaera sibogae 0.27 0.68 312 Franklin et al. 2012 Emiliania huxleyi CCMP 1516 High 17 100 14:10 Batch Grow to senescence 0.51 1.00 111.49 0.21 0.41 114.66 0.05 0.09 105.80 Keller et al. 1999 Amphidinium carterae CCMP 1314 High 20 250 14:10 Chemostat Nitrate limitation 0.2 0.29 117.6 0.4 0.57 125.3 0.7 1 116.5 Emiliania huxleyi CCMP 378 High 20 250 14:10 Chemostat Nitrate limitation 0.3 0.38 90 0.4 0.50 110.8 0.8 1 59 Kinsey et al. 2016 Phaeocystis antarctica CCMP 3314 High 2 67 - Batch Fe limitation 0.06 0.40 365 0.03 0.20 364 0.04 0.27 332 0.04 0.27 278 0.15 1 266 0.08 1 275 0.02 1 261 0.01 1 268 Stefels and van Leeuwe 1998 Phaeocystis sp. - High 4 25 16:8 Batch Fe limitation, low light 0.36 0.57 128 0.38 0.57 122 0.63 1.00 161 0.67 1.00 154 Phaeocystis sp. - High 4 110 16:8 Batch Fe limitation, high light 0.3 0.32 284 0.26 0.29 313 0.94 1.00 158 0.9 1.00 181 Sunda et al. 2002 Emiliania huxleyi CCMP 374 High 20 500 14:10 Semi-con CO2 lim 0.73 1.00 185 0.73 1.00 167 0.73 1.00 200 0.73 1.00 193 0.23 1.51 232 0.23 0.32 267 0.23 0.32 302 0.09 0.12 365 0.09 0.12 396 0.09 0.12 478 0.09 0.12 461 Sunda et al. 2007 Emiliania huxleyi CCMP 374 High 20 500 12:12 batch NH4 limitation 0.25 0.31 190 NO3 limitation 0.25 0.31 250 0.8 1.00 175 Emiliania huxleyi CCMP 374 High 20 500 12:12 semi-con NO3 limitation 0.78 1.00 196 0.71 1.00 202 0.16 0.21 200 0.18 0.25 180 Bucciarelli and Sunda 2003 Thalassiosira pseudonana CCMP 1335 Low 20 560 12:12 Batch CO2 lim 1.4 1.00 1.5 0.02 0.01 10.2 Bucciarelli and Sunda 2003 Thalassiosira pseudonana CCMP 1335 Low 20 560 12:12 Batch NO3 limi 1.5 1.00 2 0.01 0.01 30 Bucciarelli and Sunda 2003 Thalassiosira pseudonana CCMP 1335 Low 20 560 12:12 Batch PO4 lim 1.3 1.00 1.1 0.01 0.01 2.6 Bucciarelli and Sunda 2003 Thalassiosira pseudonana CCMP 1335 Low 20 560 12:12 Batch SiO4 lim 1.6 1.00 1.5 0.01 0.01 10.3 Bucciarelli et al. 2013 Trichodesmium erythraeum IMS101 Low 27 150 12:12 Semi-con Fe limitation 0.5 0.05 0.1 0.28 2.27 Bucciarelli et al. 2013 Thalassiosira oceanica CCMP 1005 Low 18 75 14:10 Semi-con Fe limitation 0.7 0.80 5.2 0.8 0.87 6 0.5 0.59 16 0.5 0.60 16 0.03 0.03 33.5 0.2 0.20 30 0.8 1.00 3.8 0.9 1.00 4 Franklin et al. 2012 Thalassiosira pseudonana CCMP 1335 Low 17 100 14:10 Batch Grow to senescence 0.37 12.26 0.04 0.12 24.66 0.05 0.14 31.03 Harada et al. 2009 Thalassiosira oceanica CCMP 1005 Low 600 12:12 Batch NO3 lim 1.02 1.05 3.57 0.49 0.51 2.84 0.12 0.12 10.54 0.00 0 14.90 0.97 1 1.32 0.89 1 0.96 0.00 1 1.29 0.00 1 3.08 Keller et al. 1999 Thalassiosira pseudonana CCMP 1335 Low 20 250 14:10 Chemostat Nitrate limitation 0.3 0.38 25.9 0.5 0.63 23.4 0.8 1.00 0.9 Kettles et al. 2014 Thalassiosira pseudonana CCMP 1335 Low 15 115 14:10 Batch NO3 starved 1.3 1.0 4.28 0.228 0.2 17.6 0.036 0.0 24.22 1.31 1 1.08 0.48 1 5.13 0.125 1 9.9 Spielmeyer and Pohnert 2012 Skeletonema marinoi G4 Low 15 70 14:10 Batch NO3 lim 0.93 1.08 25.7 0.58 0.67 37.65 0.37 0.43 41.04 0.04 0.04 40.83 0.86 1.00 11.6 0.68 1.00 37.2 0.33 1.00 60.5 0.04 1.00 66.6 Sunda et al. 2002 Thalassiosira pseudonana CCMP 1335 Low 20 500 12:12 Semi-con CO2 limitation 0.31 0.21 18.9 0.31 0.21 13.5 0.15 0.10 59.6 0.18 0.12 64.2 Thalassiosira pseudonana CCMP 1335 Low 20 500 12:12 Semi-con Fe limitation 0.04 0.03 6.5 0.04 0.03 7.6 0.04 0.03 20.5 1.45 1.00 0.88 1.45 1.00 1.46 1.45 1.00 0.85 PFT # PFT Biovolume Cellular C Quota ksat SiO 2 ksat PO 4 ksat FeT ksat NO 3 ksat NO 2 ksat NH 4 um 3 mmol C cell -1 mmol Si m 3 mmol P m 3 µmol Fe m 3 µmol N m 3 µmol N m 3 µmol N m 3 1 Picoprokaryote 1.1E-01 2.6E-12 0.0E+00 5.2E-04 5.2E-07 8.3E-03 8.3E-03 4.1E-03 2 Picoprokaryote 3.8E-01 7.8E-12 0.0E+00 8.8E-04 8.8E-07 1.4E-02 1.4E-02 7.0E-03 3 Picoeukaryote 1.4E+00 2.4E-11 0.0E+00 1.5E-03 1.5E-06 2.4E-02 2.4E-02 1.2E-02 4 Picoeukaryote 4.2E+00 7.2E-11 0.0E+00 2.5E-03 2.5E-06 4.1E-02 4.1E-02 2.0E-02 5 Coccolithophore 1.4E+01 2.2E-10 0.0E+00 4.4E-03 4.4E-06 7.0E-02 7.0E-02 3.5E-02 6 Coccolithophore 4.8E+01 6.6E-10 0.0E+00 6.2E-03 6.2E-06 9.9E-02 9.9E-02 5.0E-02 7 Coccolithophore 1.5E+02 2.0E-09 0.0E+00 8.7E-03 8.7E-06 1.4E-01 1.4E-01 7.0E-02 8 Coccolithophore 5.2E+02 6.1E-09 0.0E+00 1.2E-02 1.2E-05 2.0E-01 2.0E-01 9.9E-02 9 Coccolithophore 1.8E+03 1.8E-08 0.0E+00 1.7E-02 1.7E-05 2.8E-01 2.8E-01 1.4E-01 10 Diazotroph 1.4E+01 2.2E-10 0.0E+00 1.2E-03 4.8E-06 4.8E-02 4.8E-02 2.4E-02 11 Diazotroph 4.8E+01 6.6E-10 0.0E+00 1.7E-03 6.7E-06 6.7E-02 6.7E-02 3.4E-02 12 Diazotroph 1.5E+02 2.0E-09 0.0E+00 2.4E-03 9.5E-06 9.5E-02 9.5E-02 4.7E-02 13 Diazotroph 5.2E+02 6.1E-09 0.0E+00 3.3E-03 1.3E-05 1.3E-01 1.3E-01 6.7E-02 14 Diazotroph 1.8E+03 1.8E-08 0.0E+00 4.7E-03 1.9E-05 1.9E-01 1.9E-01 9.4E-02 15 Diatom 1.4E+01 2.2E-10 2.0E-01 1.2E-02 1.2E-05 2.0E-01 2.0E-01 9.8E-02 16 Diatom 4.8E+01 6.6E-10 2.8E-01 1.7E-02 1.7E-05 2.8E-01 2.8E-01 1.4E-01 17 Diatom 1.5E+02 2.0E-09 3.9E-01 2.4E-02 2.4E-05 3.9E-01 3.9E-01 1.9E-01 18 Diatom 5.2E+02 6.1E-09 5.5E-01 3.4E-02 3.4E-05 5.5E-01 5.5E-01 2.7E-01 19 Diatom 1.8E+03 1.8E-08 7.7E-01 4.8E-02 4.8E-05 7.7E-01 7.7E-01 3.9E-01 20 Diatom 5.6E+03 5.6E-08 1.1E+00 6.8E-02 6.8E-05 1.1E+00 1.1E+00 5.4E-01 21 Diatom 1.7E+04 1.7E-07 1.5E+00 9.6E-02 9.6E-05 1.5E+00 1.5E+00 7.7E-01 22 Diatom 5.4E+04 5.2E-07 2.2E+00 1.4E-01 1.4E-04 2.2E+00 2.2E+00 1.1E+00 23 Diatom 1.8E+05 1.6E-06 3.0E+00 1.9E-01 1.9E-04 3.0E+00 3.0E+00 1.5E+00 24 Diatom 5.9E+05 4.8E-06 4.3E+00 2.7E-01 2.7E-04 4.3E+00 4.3E+00 2.1E+00 25 Diatom 1.9E+06 1.4E-05 6.0E+00 3.8E-01 3.8E-04 6.0E+00 6.0E+00 3.0E+00 26 Dinoflagellate 1.5E+02 2.0E-09 0.0E+00 1.1E-02 1.1E-05 1.7E-01 1.7E-01 8.5E-02 27 Dinoflagellate 5.2E+02 6.1E-09 0.0E+00 1.5E-02 1.5E-05 2.4E-01 2.4E-01 1.2E-01 28 Dinoflagellate 1.8E+03 1.8E-08 0.0E+00 2.1E-02 2.1E-05 3.4E-01 3.4E-01 1.7E-01 29 Dinoflagellate 5.6E+03 5.6E-08 0.0E+00 3.0E-02 3.0E-05 4.7E-01 4.7E-01 2.4E-01 30 Dinoflagellate 1.7E+04 1.7E-07 0.0E+00 4.2E-02 4.2E-05 6.7E-01 6.7E-01 3.3E-01 31 Dinoflagellate 5.4E+04 5.2E-07 0.0E+00 5.9E-02 5.9E-05 9.4E-01 9.4E-01 4.7E-01 32 Dinoflagellate 1.8E+05 1.6E-06 0.0E+00 8.3E-02 8.3E-05 1.3E+00 1.3E+00 6.6E-01 33 Dinoflagellate 5.9E+05 4.8E-06 0.0E+00 1.2E-01 1.2E-04 1.9E+00 1.9E+00 9.3E-01 34 Dinoflagellate 1.9E+06 1.4E-05 0.0E+00 1.6E-01 1.6E-04 2.6E+00 2.6E+00 1.3E+00 35 Dinoflagellate 6.2E+06 4.4E-05 0.0E+00 2.3E-01 2.3E-04 3.7E+00 3.7E+00 1.9E+00 PFT # Biovolume Cellular C Quota Palmer um 3 mmol C cell -1 Flagellates 26 1.5E+02 2.0E-09 Diatoms 16 4.8E+01 6.6E-10 Prasinophytes 5 1.4E+01 2.2E-10 Cryptophytes 5 1.4E+01 2.2E-10 Phaeocystis 6 4.8E+01 6.6E-10 BATS Prochlorophytes 1 1.5E+02 2.0E-09 Diatoms 17 1.8E+05 1.6E-06 Prymnesiophytes 5 4.8E+01 6.6E-10 Pelagophytes 3 1.8E+03 1.8E-08 Dinoflagellates 26 1.8E+05 1.6E-06 R2 p RMSE MRB MAPE Palmer DARWIN.predicted 0.77 <0.0001 46 (5.3 39.5 Gali.predicted.(mixed) 0.78 <0.0001 65 36 56 Chl.a'predicted 0.62 <0.0001 50 22 46 BATS DARWIN.predicted 0.43 0.03 4.4 (2.7 28 Gali.predicted.(stratified) 0.05 0.52 4.9 0.3 32 Chl.a'predicted 0.04 0.57 13 51 73 (a) Seawater salinity Seawater osmolarity Total Intracellular osmolarity Median % ionic contribution Osmolarity contribution by organic osmolytes ppt osmol·m -3 osmol·m -3 osmol·m -3 34.19 962.35 1154.82 78% 259.54 34.91 982.61 1179.13 76% 283.86 35.63 1002.87 1203.44 74% 308.17 35.99 1013 1215.6 74% 320.32 36.71 1033.26 1239.91 72% 344.64 37.43 1053.52 1264.22 71% 368.95 (b) % Contribution to organic osmolarity %"Contribution" to"total" osmolarity Study Species Treatment DMSP (mM) Bucciarelli"et"al."2013 T."erythraeum Replete 0.05 0.02% 0.004% Nutrient"lim 2.3 0.7% 0.2% Sunda"et"al."2002 T."pseudonana Replete 0.9 0.3% 0.1% Nutrient"lim 64 20% 5% Spielmeyer"and"Pohnert"2012 S."marinoi Replete 17 5% 1% Nutrient"lim 29 9% 2% Sunda"et"al."2007 E."huxleyi Replete 175 55% 14% Nutrient"lim 250 78% 21% Franklin"et"al."2010 C."leptoporus Replete 187 58% 15% Nutrient"lim 409 128% 34% Kinsey"et"al."2016 P."antarctica Replete 266 83% 22% Nutrient"lim 365 114% 30% Intracellular DMSP values tested ! 60! References Archer, S. D., M. Ragni, R. Webster, R. L. Airs, and R. J. Geider. 2010. Dimethylsulfoniopropionate and dimethylsulfide production in response to photoinhibition in Emiliania huxleyi. Limnol. Oceanogr. 55: 1579–1589. doi:10.4319/lo.2010.55.4.1579 Arnold, H. E., P. Kerrison, and M. Steinke. 2013. Interacting effects of ocean acidification and warming on growth and DMS-production in the haptophyte coccolithophore Emiliania huxleyi. Glob. Chang. Biol. 19: 1007–16. doi:10.1111/gcb.12105 Berdalet, E., G. Llaveria, and R. Simo. 2011. Modulation of dimethylsulfoniopropionate (DMSP) concentration in an Alexandrium minutum (Dinophyceae) culture by small-scale turbulence: A link to toxin production? Harmful Algae 11: 88–95. doi:10.1016/j.hal.2011.08.003 Bertrand, E. M., and A. E. Allen. 2012. Influence of vitamin B auxotrophy on nitrogen metabolism in eukaryotic phytoplankton. Front. Microbiol. 3: 1–16. doi:10.3389/fmicb.2012.00375 Bisson, M., and G. Kirst. 1995. Osmotic acclimation and turgor pressure regulation in algae. Naturwissenschaften 82: 461–471. Boyd, C. M., and D. Gradmann. 2002. Impact of osmolytes on buoyancy of marine phytoplankton. Mar. Biol. 141: 605–618. doi:10.1007/s00227-002-0872-z Bucciarelli, E., C. Ridame, W. G. Sunda, C. Dimier-Hugueney, M. Cheize, and S. Belviso. 2013. Increased intracellular concentrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum. Limnol. Oceanogr. 58: 1667–1679. doi:10.4319/lo.2013.58.5.1667 Bucciarelli, E., W. G. Sunda, and P. Nicolas. 2003. Influence of CO2, nitrate, phosphate, and silicate limitation on intracellular DMSP of the coastal diatom Thalassiosira pseudonana in batch cultures. Limnol. Oceanogr. 48: 2256–2265. Carlson, C. A., H. W. Ducklow, and T. D. Sleeter. 1996. Stocks and dynamics of bacterioplankton in the northwestern Sargasso Sea. Deep Sea Res. Part II 43: 491–515. doi:10.1016/0967-0645(95)00101-8 Corn, M., S. Belviso, F. Partensky, N. Simon, and U. Christaki. 1996. Origin and importance of picoplanktonic DMSP, In Biological and environmental chemistry of DMSP and related sulfonium compounds. ! 61! Curson, A. R. J., J. Liu, A. Bermejo Martínez, and others. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nat. Microbiol. 2: 1–9. doi:10.1038/nmicrobiol.2017.9 Curson, A., B. Williams, B. Pinchbeck, and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat. Microbiol. doi:10.1038/s41564-018-0119-5 Darroch, L., M. Lavoie, M. Levasseur, and others. 2015. Effect of short-term light- and UV- stress on DMSP, DMS, and DMSP lyase activity in Emiliania huxleyi. Aquat. Microb. Ecol. 74: 173–185. doi:10.3354/ame01735 Dong, S., J. Sprintall, S. T. Gille, and L. Talley. 2008. Southern ocean mixed-layer depth from Argo float profiles. J. Geophys. Res. Ocean. 113: 1–12. doi:10.1029/2006JC004051 Fiore, C. L., K. Longnecker, M. C. Kido Soule, and E. B. Kujawinski. 2015. Release of Ecologically Relevant Metabolites by the Cyanobacterium, Synechococcus elongatus CCMP 1631. Environ. Microbiol. 17: n/a-n/a. doi:10.1111/1462-2920.12899 Franklin, D. J., M. Steinke, J. Young, I. Probert, and G. Malin. 2010. Dimethylsulphoniopropionate (DMSP), DMSP lyase activity (DLA) and dimethylsulphide (DMS) in 10 species of coccolithophore. Mar. Ecol. Prog. Ser. 410: 13–23. doi:10.3354/meps08596 Galí, M., E. Devred, M. Levasseur, S. J. Royer, and M. Babin. 2015. A remote sensing algorithm for planktonic dimethylsulfoniopropionate (DMSP) and an analysis of global patterns. Remote Sens. Environ. 171: 171–184. doi:10.1016/j.rse.2015.10.012 Galindo, V., M. Levasseur, C. J. Mundy, and others. 2016. Contrasted sensitivity of DMSP production to high light exposure in two Arctic under-ice blooms. J. Exp. Mar. Bio. Ecol. 475: 38–48. doi:10.1016/j.jembe.2015.11.009 Gebser, B., and G. Pohnert. 2013. Synchronized regulation of different zwitterionic metabolites in the osmoadaption of phytoplankton. Mar. Drugs 11: 2168–2182. doi:10.3390/md11062168 Gutierrez-Rodriguez, A., L. Pillet, T. Biard, W. Said-Ahmad, A. Amrani, R. Simó, and F. Not. 2017. Dimethylated sulfur compounds in symbiotic protists: A potentially significant source for marine DMS(P). Limnol. Oceanogr. 62: 1139–1154. doi:10.1002/lno.10491 Harada, H., M. Vila-costa, J. Cebrian, and R. P. Kiene. 2009. Effects of UV radiation and nitrate ! 62! limitation on the production of biogenic sulfur compounds by marine phytoplankton. Aquat. Bot. 90: 37–42. doi:10.1016/j.aquabot.2008.05.004 Huang, K., H. Ducklow, M. Vernet, N. Cassar, and M. L. Bender. 2012. Export production and its regulating factors in the West Antarctica Peninsula region of the Southern Ocean. Global Biogeochem. Cycles 26: 1–13. doi:10.1029/2010GB004028 Hutchins, D., M. Mulholland, and F. Fu. 2009. Nutrient Cycles and Marine Microbes in a CO2 Enriched Ocean. Oceanography 22: 128–145. Karsten, U., K. Kück, C. Vogt, and G. O. Kirst. 1996. Dimethylsulfoniopropionate production in phototrophic organisms and its physiological functions as a cryoprotectant, p. 143–153. In Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds. Keller, M., W. Bellows, and R. Guillard. 1989. Dimethyl sulfide production in marine phytoplankton, p. 167–182. In Biogenic sulfur in the environment. Keller, M. D. 1989. 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 Keller, M. D., R. P. Kiene, P. A. Matrai, and W. K. Bellows. 1999. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. II. N-limited chemostat cultures. Mar. Biol. 135: 249–257. doi:10.1007/s002270050622 Keller, M. D., and W. Korjeff-Bellows. 1996. Physiological Aspects of the Production of Dimeyhtlsulfoniopropionate (DMSP) by Marine Phytoplankton, p. 131–142. In Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds. Keller, M. D., P. A. Matrai, R. P. Kiene, and W. K. Bellows. 2004. Responses of coastal phytoplankton populations to nitrogen additions: dynamics of cell-associated dimethylsulfoniopropionate (DMSP), glycine betaine (GBT), and homarine. Can. J. Fish. Aquat. Sci. 61: 685–699. doi:10.1139/f04-058 Kinsey, J. D., D. J. Kieber, and P. J. Neale. 2016. Effects of iron limitation and UV radiation on Phaeocystis antarctica growth and dimethylsulfoniopropionate, dimethylsulfoxide and acrylate concentrations. Environ. Chem. 13: 195–211. doi:10.1071/EN14275 Lavoie, M., M. Levasseur, and M. Babin. 2015. Testing the potential ballast role for dimethylsulfoniopropionate in marine phytoplankton: A modeling study. J. Plankton Res. 37: 699–711. doi:10.1093/plankt/fbv050 ! 63! Li, C., G. Yang, J. Pan, and H. Zhang. 2010. Experimental studies on dimethylsulfide (DMS) and dimethylsulfoniopropionate (DMSP) production by four marine microalgae. Acta Oceanol. Sin. 29: 78–87. doi:10.1007/s13131-010-0052-9 Lyon, B. R., J. M. Bennett-Mintz, P. A. Lee, M. G. Janech, and G. R. Ditullio. 2016. 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 Marañón, E. 2005. Phytoplankton growth rates in the Atlantic subtropical gyres. Limnol. Oceanogr. 50: 299–310. doi:10.4319/lo.2005.50.1.0299 Raina, J. B., P. Clode, S. Cheong, and others. 2017. Subcellular tracking reveals the location of dimethylsulfoniopropionate in microalgae and visualises its uptake by marine bacteria. Elife 1–17. doi:10.7554/eLife.23008 Raven, J. A., and M. A. Doblin. 2014. Active water transport in unicellular algae: Where, why, and how. J. Exp. Bot. 65: 6279–6292. doi:10.1093/jxb/eru360 van Rijssel, M., and A. Buma. 2002. UV radiation induced stress does not affect DMSP synthesis in the marine prymnesiophyte Emiliania huxleyi. Aquat. Microb. Ecol. 28: 167– 174. doi:10.3354/ame028167 van Rijssel, M., and W. W. C. Gieskes. 2002. Temperature, light, and the dimethylsulfoniopropionate (DMSP) content of Emiliania huxleyi (Prymnesiophyceae). J. Sea Res. 48: 17–27. doi:10.1016/S1385-1101(02)00134-X Sikes, C., and K. Wilbur. 1982. Functions of coccolith formation. Limnol. Oceanogr. 27: 18–26. Slezak, D., and G. Herndl. 2003. Effects of ultraviolet and visible radiation on the cellular concentratins of dimethylsulfoniopropionate (DMSP) in Emiliania hyxleyi (strain L). Mar. Ecol. Prog. Ser. 246: 61–71. doi:10.3354/meps246061 Spielmeyer, A., B. Gebser, and G. Pohnert. 2011. Dimethylsulfide sources from microalgae: Improvement and application of a derivatization-based method for the determination of dimethylsulfoniopropionate and other zwitterionic osmolytes in phytoplankton. Mar. Chem. 124: 48–56. doi:10.1016/j.marchem.2010.12.001 Spielmeyer, A., and G. Pohnert. 2012. Influence of temperature and elevated carbon dioxide on the production of dimethylsulfoniopropionate and glycine betaine by marine phytoplankton. Mar. Environ. Res. 73: 62–69. doi:10.1016/j.marenvres.2011.11.002 ! 64! Stefels, J. 2000. Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43: 183–197. Stefels, J., and W. van Boekel. 1993. Production of DMS from dissolved DMSP in axenic cultures of the marine phytoplankton species Phaeocystis sp. Mar. Ecol. Prog. Ser. 97: 11– 18. Stefels, J., and M. A. Van Leeuwe. 1998. Effects of iron and light stres on the biochemical composition of antarctic Phaeocystis sp. I. Intracellular DMSP concentrations. J. Phycol. 34: 486–495. Sunda, W., D. J. Kieber, R. P. Kiene, and S. Huntsman. 2002. An antioxidant function for DMSP and DMS in marine algae. Nature 418: 317–320. del Valle, A., D. Slezak, C. M. Smith, A. N. Rellinger, D. J. Kieber, and R. P. Kiene. 2011. Effect of acidification on preservation of DMSP in seawater and phytoplankton cultures : Evidence for rapid loss and cleavage of DMSP in samples containing. Mar. Chem. 124: 57– 67. doi:10.1016/j.marchem.2010.12.002 Webb, A. L., G. Malin, F. E. Hopkins, K.-L. Ho, U. Riebesell, K. Schulz, A. Larsen, and P. Liss. 2016. Ocean acidification has different effects on the production of dimethylsulfide and dimethylsulfoniopropionate measured in cultures of Emiliania huxleyi and a mesocosm study: a comparison of laboratory monocultures and community interactions. Environ. Chem. 13: 314–329. doi:http://dx.doi.org/10.1071/EN14268 ! 65 Chapter Two: Evidence for contrasting roles of dimethylsulfoniopropionate (DMSP) production in Emiliania huxleyi and Thalassiosira oceanica Erin L. McParland, Anna Wright, Kristin Art, Meagan He and Naomi M. Levine Department of Marine and Environmental Biology, University of Southern California Los Angeles, CA Abstract DMSP is a globally abundant metabolite produced by most marine eukaryotic phototrophs that constitutes an important carbon source for microbial ecosystems and influences climate regulation. However, the physiological function of DMSP has remained enigmatic for 30+ years. Recent insight hypothesizes different roles for DMSP based on cellular DMSP concentrations in producers. Here, this differential production was tested with multiple physiological experiments that altered growth rate with salinity, temperature and nitrate stress to target different metabolic conditions in Emiliania huxleyi, a high DMSP producer, and Thalassiosira oceanica, a low DMSP producer. Emiliania huxleyi did not significantly change intracellular DMSP in metabolically imbalanced growth. Thalassiosira oceanica intracellular DMSP was correlated to growth rate across all stressors tested and exhibited a plastic response to non-steady-state changes on a timescale of hours. This work shifts the current paradigm of a universal DMSP mechanism for all producers to two distinct ecological roles for DMSP, defined by fundamentally different drivers of production and cellular mechanisms in the two groups of producers, where 1). the primary role of DMSP in high producers is a compatible solute, and 2). where DMSP is likely a signaling molecule in low producers. 66 Introduction Dimethylsulfoniopropionate (DMSP) is a globally abundant organic sulfur and carbon metabolite produced by a diverse array of marine microbes, including most marine eukaryotic phototrophs and some marine prokaryotes (Keller et al. 1989; McParland and Levine 2019). DMSP can comprise up to 11% of cellular carbon in marine phytoplankton, and DMSP production accounts for as much as 5% of total primary production in both coastal and open ocean regimes (Stefels et al. 2007; Galí et al. 2013; Levine et al. 2015). When released into the dissolved pool, DMSP is estimated to supply up to 13% and 100% of the bacterial carbon and sulfur demand, respectively (Levine et al. 2015; Kiene et al. 2000; Tripp et al. 2008). Additionally, a DMSP degradation product, dimethylsulfide (DMS), is thought to be the most significant natural source of sulfur to the atmosphere and consequentially plays an important role in climate regulation as a source of cloud condensation nuclei (Charlson et al. 1987; Lana et al. 2012). Understanding the physiological function of DMSP in producers is critical for quantifying the significant role of DMSP in global carbon cycling, climate and marine microbial ecosystem dynamics. Intracellular DMSP concentrations span a wide range from 0.01 to >1000 mM in different producers. This distribution appears to be bi-modal with two types of producers: high producers (HiDPs) with intracellular DMSP concentrations > 50mM and low producers (LoDPs) with intracellular DMSP < 50mM (Caruana and Malin 2014; McParland and Levine 2019) (Figure 1). In general, prymnesiophytes and dinoflagellates are classified as HiDPs, and other important primary producers, such as diatoms and cyanobacteria, are typically LoDPs (Keller 1989; Keller et al. 1989; McParland and Levine 2019). Current hypotheses for the cellular mechanism of DMSP production include its use as a compatible solute, a cryoprotectant, a ballasting mechanism, a signaling molecule, an overflow mechanism, and an antioxidant (Karsten et al. 1996; Stefels and Leeuwe 1998; Stefels 2000; Sunda et al. 2002; Seymour 2010; Lavoie et al. 2015). Most likely, DMSP plays multiple roles in the cell and/or different roles for different phytoplankton groups (Archer et al. 2010; Bucciarelli et al. 2013). However, despite over thirty years of research, we currently lack an understanding of what these roles are and how they vary across different DMSP producers. Experimental results from studies investigating the DMSP mechanism are often conflicting (e.g. van Rijssel and Buma 2002; Archer et al. 2010) and typically extrapolate the results for a single species to all DMSP producers. Further complications have arisen as measurements of 67 cellular concentration changes, rather than synthesis rates, are highly dependent on normalization factors (e.g. cell number, cell carbon, or cell volume) (Keller 1989; Stefels et al. 2009). Previous studies have also compounded decreased and imbalanced growth, and steady-state and non-steady- state conditions, making it difficult to distinguish potential mechanisms (McParland and Levine 2019). A mechanistic understanding of the primary drivers behind DMSP production is essential for accurate predictions of in situ DMSP cycling, the sensitivity of the DMSP cycle to environmental changes, and the implications of these changes for marine ecosystems and the global climate. Recently, a meta-analysis of all previous monoculture studies that assessed DMSP production proposed that HiDPs and LoDPs differentially regulate DMSP: HiDP intracellular DMSP did not appear to respond to nutrient stress, suggesting it may be a constitutive metabolite in these producers, whereas LoDP intracellular DMSP significantly increased as a predictable function of nutrient stress, suggesting it is a stress mechanism (McParland and Levine 2019). This contrasting behavior of HiDPs and LoDPs was originally hypothesized by Stefels et al. (2007), but has never been tested by directly comparing HiDP and LoDP DMSP production under multiple types of stressed growth. Here, we methodically compared intracellular DMSP changes of a HiDP coccolithophore, Emiliania huxleyi, and a LoDP diatom, Thalassiosira oceanica, in response to four different environmental conditions that uniquely altered cellular metabolism and growth rate: salinity, temperature and nitrogen stress under steady-state conditions, and non-steady-state nitrogen stress. The four conditions isolate the responses of DMSP production by the HiDP and LoDP to different low growth conditions, either metabolically balanced or imbalanced growth, with or without nutrient stress. We also tested the plasticity of the DMSP mechanism with fine-scale temporal measurements in non-steady-state. We confirmed the hypothesis of differential regulation under steady-state nutrient stress proposed by McParland and Levine (2019) with the first direct comparison of a HiDP and LoDP and build on this hypothesis with observations of differential DMSP production across multiple metabolic conditions in steady- and non-steady-state. The results demonstrate that there are (at least) two distinct groups of DMSP producers for which DMSP serves significantly different physiological roles. Moreover, these groups regulate DMSP differently, one as a constitutive function of cellular growth and the other as a function of stressed growth. Critically, we show that the current paradigm of viewing DMSP cycling through a single 68 lens is incorrect; rather, DMSP functions in two independent ecological cycles where it serves different cellular functions and responds to different ecological cues. Methods Culturing Axenic cultures of Thalassiosira (T.) oceanica (CCMP 1005) and Emiliania (E.) huxleyi (CCMP 373) were obtained from the National Center for Marine Algae and Microbiota. Cultures were grown in f/25 media: 0.2 µm filtrate of natural seawater (collected at the San Pedro Ocean Time- series station, 33º 33’N, 118º 24’W, and incubated in the dark for >2 weeks) was microwave sterilized and nutrients were added ([NO3] = 100 µM, [PO4] = 6.25 µM, all other nutrients at f/2 concentrations) (Iglesias-Rodriguez et al. 2008). Cultures were maintained semi-continuously in 1-liter polycarbonate bottles at 23°C under fluorescent light (100 µmol photons m -2 s -1 , 14:10 h light: dark cycle). For all steady-state experiments, in-vivo fluorescence was used to semi-continuously acclimate each species to each experimental condition in triplicate until growth rates remained constant across two transfers (Wood et al. 2005). After the third transfer, the bottles were sampled for cell counts, cellular biovolume, chlorophyll-a, Fv/Fm, and DMSP total in mid-exponential phase and again 24 h later. Maintenance culturing conditions were altered as follows to create different types of steady-state growth stress. Experiment 1: Each species was grown semi-continuously under maintenance conditions in low NO3 f/25 media ([NO3] = 8 µM). Two low NO3 conditions were assessed: one where each species was semi-continuously acclimated to low NO3 in mid- exponential phase and one where each species was semi-continuously acclimated to low NO3 in late-exponential phase. Experiment 2: Each species was grown semi-continuously under maintenance conditions in water baths set to 14°C, 16°C, 20°C, 23°C, 26°C, or 28°C. Experiment 3: Media of different salinities was made by mixing 0.2 µm filter sterile seawater or MilliQ water with a hyper-saline solution (65‰) of ESAW artificial salts (Harrison and Berges 2005). Media with final salinities of 25‰, 35‰, 40‰, and 50‰ for T. oceanica and 25‰, 30‰, 35‰, 40‰, and 45‰ for E. huxleyi were confirmed with a refractometer after microwave sterilization. Nutrients were added at maintenance concentrations (replete). By diluting all salts, these experiments altered not only salinity, but also the availability of sulfate, which has been shown to alter DMSP concentrations in Emiliania huxleyi, but not in T. pseudonana (Bochenek et al. 2013; 69 Kettles et al. 2014). All steady-state results are presented as the mean of biological triplicates, except for E. huxleyi grown at 25‰ and late-exponential phase low NO3 for which only the mean of biological duplicates is presented as the third bottles crashed. Steady-state low NO3 acclimated cultures of both species were used to quantify the non- steady-state response to the alleviation of nutrient stress. Experiment 4: The experiment was started in mid-exponential phase by adding back NO3 to replete concentrations ([NO 3] = 100 µM) in triplicate. Control cultures were maintained in low NO3 conditions in triplicate ([NO3] = 8 µM). Samples for cell counts, cellular biovolume, chlorophyll-a, Fv/Fm, and DMSP total were collected at 0, 1, 3, 6, 9, 12, and 24 h after N add-back. DMSP Culture bottles were gently inverted three times before sampling for DMSP total (DMSPt). Unfiltered culture (10ml) was collected in Falcon tubes and immediately acidified to pH 2 with 50% H 2SO4 (3.3 µl per 1ml of sample). DMSPt samples were stored for a minimum of 24 h to allow for the complete oxidation of DMS (del Valle et al. 2011). An aliquot of sample (1ml for E. huxleyi and 3ml for T. oceanica) was dispensed into an acid-washed, combusted 14ml serum vial with 1ml of 5N NaOH and crimped closed with a gas-tight Teflon lid. Samples were vortexed and incubated for 10 min to cleave DMSP to DMS via alkaline hydrolysis (Kiene and Service 1991). DMSP (derivatized to DMS) was quantified via headspace analysis using a gas-tight Hamilton syringe with a custom Shimadzu 2016 gas chromatograph with a flame photometric detector and a Chromosil 330 packed column (Supelco). Column temperature was 70ºC and retention time was 0.92 min. DMSP concentrations were calculated with a 4-point DMS standard curve (R 2 > 0.95). Experimental triplicates and technical duplicates (n=6 per treatment) were quantified. DMSPt concentrations are presented as nmoles DMSP·liter of culture -1 (nM) and as intracellular DMSP, calculated as cellular DMSP normalized to cell volume in mmoles DMSP·liter of cell volume -1 (mM). Fv/Fm Unfiltered culture (3ml) was collected in duplicate and dark adapted for 20 min. The maximum PSII quantum yield (Fv/Fm) was measured with a WALZ Phyto-PAM. Fo was measured under low 70 frequency (25Hz) pulses and Fm was measured with a saturating pulse of 2600 µmol photons·m - 2 ·s -1 for 200ms. Fv/F m was calculated as: ! " /! $ =& ! $ −! ( ! $ Ancillary measurements 5-10ml of culture were gently vacuum filtered onto a 25mm Whatman GF/F filter for chlorophyll- a (Chla) and stored at -20ºC. Filters were extracted in 2.5 ml of 90% HPLC-grade acetone within 36 h of collection and analyzed with a Turner Trilogy fluorometer (Welschemeyer 1994). Cell counts were quantified using 1.5 ml of culture preserved with 20 µl of 37% formaldehyde (final concentration = 0.5%). Cell counts were then enumerated with light microscopy (Zeiss) using a 0.1mm Neubauer hemacytometer (Hausser Scientific). Cell images were taken with a Zeiss AxioCam MRc 5 at 400x and dimensions were measured in MATLAB (diameter for E. huxleyi and diameter and height for T. oceanica). Number of pixels along a straight line were converted to µm using an image of a stage micrometer for calibration. Cell biovolumes were calculated using the geometric volume formula of a sphere for E. huxleyi and a cylinder for T. oceanica. Cellular biovolumes for each species were used to calculate intracellular DMSP (Supp Table 1). Finally, growth rates were calculated as: ) = *+,-. / 0−+,(. 2 )4 5 ⁄ where Xf is cell concentration at the final time point, Xi is cell concentration at the initial time point, and t is the interval between the measurements in days. Cellular osmolarity calculations The media osmolarity was calculated after Boyd and Gradmann (2002). Total cellular osmolarity was assumed to be maintained 20% higher than the media osmolarity and equal to the sum of ionic osmolarity and organic osmolarity (Sikes and Wilbur 1982; Lavoie et al. 2015). Ionic osmolarity was calculated based on previously measured concentrations of ions and organic osmolarity was calculated as the difference in total cellular osmolarity - total ionic osmolarity. Previously measured ionic and organic osmolarity varies widely across species (Keller et al. 1999a; Boyd and Gradmann 2002; Gebser and Pohnert 2013) and the internal mass balance of intracellular osmolarity is still not fully understood (Raven and Doblin 2014). Therefore, we assumed that ionic 71 osmolarity was equal to that predicted for E. huxleyi and T. pseudonana by Lavoie et al. (2015) and that major ion concentrations linearly scaled with changes in external osmolarity (Dickson and Kirst 1987b; a) (Supp Table 2). The contribution of DMSP to cellular osmolarity is presented as a percent contribution to the calculated total organic osmolarity. Results We quantified DMSP production by a HiDP (E. huxleyi) and a LoDP (T. oceanica) under temperature, nutrient and salinity stress to isolate differential responses by these two species under metabolically balanced and unbalanced growth. As McParland and Levine (2019) observed two distinct, but consistent, responses to nutrient stress in multiple strains of LoDPs and HiDPs, we assumed that the responses of the two model DMSP producers chosen for these experiments reflect those of all HiDPs and LoDPs. Previous work has demonstrated that growth at temperatures below optimal growth temperature (T opt) is metabolically balanced, with similar decreases in photosynthesis and respiration (Thomas et al. 2012; Baker et al. 2016; Barton et al. 2018). In contrast, growth at temperatures above T opt and under nutrient limitation (steady- and non-steady- state) both result in lower growth rates linked to metabolic imbalances, where photosynthesis and respiration are decoupled (Hockin et al. 2012; Baker et al. 2016; Wordenweber et al. 2018). The metabolic conditions of marine phytoplankton resulting from hyper- and hypo-saline stress in steady-state are relatively unknown as most previous studies observed responses of non-steady- state osmotic shocks or euryhaline species (Qasim et al. 1972; Macler 1988; Jahnke and White 2003; Bussard et al. 2017). Extreme salinity stress is however well-known to induce oxidative stress (Acosta-Motos et al. 2017) and Jahnke & White (2003) found that different types of oxidative stress may occur in hyper- versus hypo-saline conditions. Temperature stress (balanced and imbalanced growth): E. huxleyi and T. oceanica were grown in steady-state, nutrient replete conditions across a thermal gradient. Both E. huxleyi and T. oceanica growth rates (µ) exhibited classic, skewed, thermal response curves with an optimal growth temperature (T opt) at 23°C and 26°C, respectively (Supp Figure 1, Table 1). We compared intracellular DMSP concentrations at temperatures ! T opt (balanced growth) and > T opt (imbalanced growth). In both species grown at temperatures < T opt, Fv/Fm was positively correlated with temperature (R 2 >0.7, p!0.05) (Supp Figure 2) and cellular Chla was significantly higher than at 72 T opt (t-test, p!0.05), consistent with the established low temperature response (Geider 1987). For E. huxleyi at temperatures > T opt, Fv/F m continued to increase (Supp Figure 2) and cellular Chla was slightly higher at 28°C than at T opt (t-test, p=0.07). The temperature response curve for T. oceanica was more skewed than for E. huxleyi with µ at 28°C not significantly lower than µ at T opt (t-test, p=0.7) but with zero growth at 30°C (Supp Figure 1). Therefore, the response of T. oceanica to reduced growth above T opt could not be quantified. Intracellular DMSP at T opt was 145 ± 19 mM in E. huxleyi and 2 ± 0.2 mM in T. oceanica (Table 1), which are consistent with previous measurements of the same strains (Steinke et al. 1998; Bucciarelli et al. 2013; Arnold et al. 2013). Under balanced low growth (temperatures ! T opt), intracellular DMSP significantly positively correlated with µ in both E. huxleyi and T. oceanica (R 2 >0.8, p!0.05) (Figures 2a,b). Specifically, over a 3-fold decrease in µ (c. 0.9 d -1 to 0.3 d -1 ), intracellular DMSP increased 3-fold and 6-fold for E. huxleyi and T. oceanica, respectively (Supp Figure 3, Table 1). Under imbalanced low growth (temperatures > T opt), E. huxleyi intracellular DMSP became decoupled from µ as intracellular DMSP was either lower than or unchanged from the T opt concentration (Figure 2a, Supp Figure 3). N stress (imbalanced growth): E. huxleyi and T. oceanica were grown in steady-state under three different nitrate (NO3) conditions: NO3 replete in exponential phase (Nss + ), NO3 stress in exponential (Nss - ) and NO3 stress in late exponential phase (Nss -- ). µ significantly decreased with increasing NO3 stress in both species (t-test, p!0.05) (Table 1). In E. huxleyi, Fv/Fm significantly decreased with increasing NO3 stress (" Fv/Fm = 0.24) (t-test, p!0.05) (Supp Figure 2), while cellular Chla remained unchanged (t-test, p=0.9) (Table 1). This suggests that photosynthetic efficiency decreased under NO3 stress in E. huxleyi, even though optimal cellular Chla was maintained (unchanged relative to Nss + condition). In T. oceanica, both Fv/Fm and cellular Chla significantly decreased as NO3 stress increased (t-test, p!0.05) (Supp Figure 2, Table 1). Though Fv/Fm changes were significant, the magnitude of change was small in T. oceanica ("Fv/Fm = 0.05). We confirmed the negative relationship between Fv/Fm and NO3 stress in T. oceanica with additional measurements (Supp Figure 4), which further suggest that these small changes in Fv/Fm reflect the cellular response to NO3 stress (i.e. imbalanced growth). E. huxleyi intracellular DMSP remained high with no significant changes under NO3 stress (Nss + = 165 ± 23 mM versus Nss -- 146 ± 36) (t-test, p=0.2) (Figure 2b, Supp Figure 3). In contrast, 73 T. oceanica intracellular DMSP linearly increased with increasing NO3 stress from 4 ± 0.3 mM in Nss + to 12 ± 0.5 mM in Nss -- (R 2 =0.9, p!0.05) (Figure 2a). Specifically, relative to Nss + , as µ decreased 3- fold in T. oceanica, intracellular DMSP increased 3-fold in Nss -- . Salinity stress: E. huxleyi and T. oceanica were grown in steady-state under nutrient replete conditions across a salinity gradient. While the metabolic conditions of these two species in steady- state hypo- and hyper-saline stress are unknown, salinity changes significantly decreased µ of E. huxleyi and T. oceanica relative to µ at optimal salinity (35‰) (Table 1). In hyper-saline conditions, µ decreased 3-fold in E. huxleyi at 45‰ and 7-fold in T. oceanica at 50‰ (Table 1). Hypo-saline conditions (25‰) decreased E. huxleyi µ 8-fold (Table 1). No growth was observed for T. oceanica in the hypo-saline condition of 25‰ and thus the hypo-saline response could not be tested. Similar to the response observed under NO3 stress, cellular Chla in E. huxleyi did not change significantly under hyper-saline conditions (t-test, p=0.1) (Table 1), but Fv/Fm exhibited a small, significant decrease relative to optimal salinity ("Fv/Fm = 0.05) (t-test, p!0.05) (Supp Figure 2). In hypo-saline conditions, cellular Chla in E. huxleyi decreased (t-test, p!0.05) but no significant change was observed in Fv/Fm (t-test, p=0.2). In T. oceanica, both cellular Chla and Fv/Fm were significantly lower under hyper-saline conditions ("Fv/Fm = 0.2, t-test, p!0.05) (Table 1, Supp Figure 2). For both species, intracellular DMSP was significantly positively correlated to both salinity and µ in hyper-saline conditions (R 2 >0.9, p!0.05) (Supp Figure 3, Figures 2a,b). The increase in E. huxleyi intracellular DMSP more than compensated for the increased osmotic demand, with the predicted contribution of DMSP to organic osmolarity increasing from 20% to 37% in hyper-saline conditions (Figure 2c). A small increase in the contribution of DMSP to organic osmolarity from 0.2% to 2% was also predicted for T. oceanica in hyper-saline conditions (Figure 2d). All steady-state: Depressed growth rates in T. oceanica resulted in increased intracellular DMSP across all steady-state experiments. Critically, the same linear relationship between µ and intracellular DMSP was observed (R 2 =0.8, p!0.05) (Figure 2b), where the greatest decrease in µ resulted in the greatest intracellular DMSP change (" = 13 ± 1 mM at 50‰) and subsequently the greatest contribution to organic osmolarity (Figures 2b,d). This constant relationship between DMSP and growth suggests that T. oceanica produced DMSP as a function of stressed growth, 74 independent of the stressor type or different metabolic conditions associated with each (Figure 2b). E. huxleyi exhibited a very different, and noisier, relationship between µ and intracellular DMSP (Figure 2a). Changes in intracellular DMSP in E. huxleyi were significantly correlated to µ in hyper-saline stress and temperatures ! T opt (R 2 =0.8, p!0.05), but not under NO3 stress or at temperatures > T opt (R 2 <0.6, p>0.05) (Figure 2a). However, under salinity stress, most of the increase in E. huxleyi intracellular DMSP could be attributed to cellular osmotic adjustments in response to media osmolarity changes (Figure 2c). Therefore, the only significant change in E. huxleyi intracellular DMSP concentrations, not directly linked to salinity changes, occurred under temperatures ! T opt (Figure 2a). Non-steady-state N stress (imbalanced growth): To compare the plastic response (non-steady- state) to the steady-state condition, intracellular DMSP changes in E. huxleyi and T. oceanica were tracked for 24 h after alleviation of NO 3 stress in Nss - cultures (+N). Control cultures were maintained in Nss - (-N). The significant increase in both Chla and cell concentrations in +N conditions for both species after 24 h (t-test, p!0.05) confirms that NO3 limited the starting conditions (Figure 3). The timecourse of Chla and cell counts in response to the NO3 add-back differed between species (Figure 3). +N T. oceanica responded rapidly by increasing both Chla and cell concentrations continuously over the 24 h, with cell concentrations significantly higher than –N beginning at 9 h (t-test, p!0.05). After 24 h, cellular Chla in T. oceanica was statistically higher in +N conditions than in -N (t-test, p!0.05) and was similar to the Nss + cellular Chla (Figure 3f, Table 1). In contrast, only Chla in +N E. huxleyi significantly increased in the first 12 h following the NO3 add-back, while cell concentrations remained constant (Figure 3c). Cell concentrations then significantly increased between 12 and 24 h in +N, indicating that E. huxleyi divided during this time period (Fig 3b). Other than the deviation prior to cell growth (timepoints at 9 and 12 h), E. huxleyi cellular Chla concentrations were not statistically different in +N and -N cultures (t- test, p=0.4) (Figure 3c). Despite the differences in cellular Chla and cell division, changes in Fv/F m over the experiment were consistent for T. oceanica and E. huxleyi (Figure 4). In both species after 24 h, a small but significant increase in Fv/Fm in +N cultures ("Fv/F m = 0.04 and 0.03 for T. oceanica and E. huxleyi, respectively) was significantly higher than Fv/Fm in -N cultures (t-test, p!0.05). Each 75 species exhibited a unique diel feature at the mid-point of the light cycle, but these features were determined not to have significant implications for DMSP production and therefore are only discussed in the Supplement (Supp Note, Supp Figure 5). Fv/F m in both species was significantly correlated to total Chla concentrations (R 2 >0.5, p!0.05) (Figures 3,4), consistent with the response of Fv/Fm to NO3 stress observed in the steady-state experiments. While DMSPt significantly increased in the +N E. huxleyi experiment (t-test, p!0.05) (Supp Figure 6), on a per cell basis, intracellular DMSP was not significantly different from -N after 24 h (t-test, p=0.3) (Figure 5a). The significant diel variability in E. huxleyi intracellular DMSP was correlated with the observed changes in cellular Chla (R 2 =0.4, p!0.05) (Figures 3c,5a), suggesting that DMSP was produced at a similar rate as Chla in response to the NO3 add-back. The statistically similar intracellular DMSP at the beginning and end of the experiment is consistent with the non-significant changes observed in steady-state NO3 stress (Supp Figure 3). The lack of intracellular DMSP response in E. huxleyi to the alleviation of NO3 stress indicates that DMSP was maintained constitutively, independent of stressed growth. In contrast, +N T. oceanica DMSPt was lower than -N DMSPt after 24 h, though not significantly (t-test, p=0.1), and intracellular DMSP rapidly decreased 2-fold over the experiment in +N (Figure 5b, Supp Figure 6). Specifically, the rapid increase in T. oceanica biomass and Chla concentrations in response to the NO3 add-back (Figures 3d,e) was matched by a similar, rapid decrease in intracellular DMSP (R 2 =0.5, p!0.05) (Figure 5b). We attribute this decrease in T. oceanica intracellular DMSP to dilution due to cell division (Supp Figure 7) and downregulation of DMSP production to a basal rate after NO3 stress alleviation. This downregulation of DMSP production was comparable to the observed intracellular DMSP concentrations under steady-state replete conditions (Table 1). The plasticity of DMSP production by T. oceanica in non-steady- state suggests that DMSP production was actively regulated in response to the NO3 add-back. All experiments: While a consistent relationship with µ was lacking under steady-state conditions in E. huxleyi (Figure 2a), intracellular DMSP was positively correlated to cellular Chla across all experiments (R 2 =0.6, p!0.05) (Figure 6a). This relationship was not observed for T. oceanica (R 2 =0.06, p=0.04) (Figure 6b), suggesting a decoupling of these two cellular processes. Instead, responses of T. oceanica Fv/Fm to all experimental conditions were significantly correlated to intracellular DMSP concentrations (R 2 =0.5, p!0.05) (Figure 6d), but not in E. huxleyi (R 2 =0.09, 76 p=0.01) (Figure 6c). This further suggests that intracellular DMSP production in T. oceanica was correlated with cellular responses to different stressors, both under steady and non-steady-state. In contrast, intracellular DMSP production in E. huxleyi was correlated with a constitutive component of cellular growth (Figure 6a). Discussion and Conclusion Changes in intracellular DMSP were quantified for a HiDP, E. huxleyi, and a LoDP, T. oceanica, in a series of monoculture experiments in order to disentangle the different physiological mechanisms of DMSP production in HiDPs and LoDPs. Specifically, to target the two hypothesized types of DMSP regulation (constitutive versus stress-related), DMSP production was contrasted under metabolically balanced growth conditions (temperatures below T opt) and metabolically imbalanced growth conditions (temperatures above T opt, and nutrient limited growth). We also directly tested the role of DMSP as a compatible solute in both species under different salinity conditions. Simultaneous physiology measurements (µ, cellular Chla and Fv/Fm) provided insight into whether DMSP was playing a role in stressor response or whether the cells were constitutively producing DMSP as a function of growth. Under all conditions, including steady- and non-steady-state experiments, E. huxleyi intracellular DMSP and cellular Chla were significantly positively correlated (Figure 6a). E. huxleyi intracellular DMSP did not significantly respond to N stress or temperatures > T opt, despite significant growth limitation, indicating that DMSP production was not altered by these metabolic imbalances (Figures 2a,5a). Salinity shifts did induce changes in E. huxleyi intracellular DMSP (Figure 2a), but these changes were predicted to be primarily accounted for by shifts in DMSP production to maintain internal osmolarity (Figure 2c). Low temperature growth (< T opt) resulted in the only significant change in E. huxleyi intracellular DMSP that was independent of salinity changes (Figure 2c). The general patterns of changes in intracellular DMSP observed here are consistent with previous studies of E. huxleyi (Turner et al. 1988; Keller and Korjeff-Bellows 1996; Keller et al. 1999b; a; Sunda et al. 2002, 2007; van Rijssel and Gieskes 2002). It is also important to note that the large diurnal variability in cellular Chla and intracellular DMSP (2-fold) (Figures 3c,5a) in E. huxleyi highlights the importance of both sampling time and normalization factor for assessing DMSP production in E. huxleyi. 77 The maintenance of high cellular DMSP concentrations by E. huxleyi, independent of two contrasting metabolically imbalanced growth conditions (N stress and > T opt), and the scaling of intracellular DMSP in response to changes in media osmolarity, imply that DMSP is likely a compatible solute in HiDPs. One of the original hypotheses for the DMSP mechanism was that DMSP replaces N-containing osmolytes under N stress (Andreae 1986). However, we found no evidence to support this for E. huxleyi as intracellular DMSP concentrations did not increase under N stress; in fact, they decreased slightly (Supp Figure 3). Intracellular DMSP in E. huxleyi did significantly respond to low temperatures (< T opt) (Figure 2a). It has been hypothesized that DMSP serves to stabilize enzymes and proteins at low temperatures (Nishiguchi and Somero 1992; Karsten et al. 1996), however this work was conducted at much lower temperatures (6ºC). Further work is necessary to determine whether the significant DMSP response under low temperatures reflects a shift in compatible solute preference or a secondary role for DMSP in this HiDP. Maintaining cellular DMSP concentrations in the 100s of mM is believed to require a significant proportion of cellular energy, with a particular demand on methionine synthesis (Stefels 2000). Given the versatility of DMSP as a compatible solute, free radical scavenger and overflow mechanism, this significant energy demand for synthesis may be justified if HiDPs utilize the multi-functionality of DMSP. While intracellular DMSP concentrations did not vary significantly across a wide range of environmental conditions, this does not mean that DMSP is not being actively consumed within the cell (e.g. through reaction with ROS), only that any consumption is being matched by production. This is consistent with the observation that carbon and sulfur appear to be efficiently recycled as part of DMSP production in E. huxleyi (Bochenek et al. 2013). Nonetheless, while E. huxleyi may regulate DMSP production to match internal consumption rates, it clearly does not actively regulate intracellular DMSP concentrations in response to metabolic imbalances. Furthermore, cellular Chla and DMSP in E. huxleyi were always significantly correlated (Figure 6a). Altogether, these experiments suggest that DMSP is regulated as a constitutive metabolite in this HiDP, likely as an essential compatible solute, but also has the potential to serve other cryptic roles simultaneously. Significant changes in T. oceanica intracellular DMSP under both balanced and imbalanced growth conditions suggest that T. oceanica actively regulated DMSP production in response to stressed growth. In all steady-state conditions, intracellular DMSP concentrations were significantly correlated with µ (Figure 2b). This is consistent with the upregulation of intracellular 78 DMSP in response to nutrient stress observed in a previous meta-analysis of LoDP monoculture studies (McParland and Levine 2019). Intracellular DMSP concentrations were still quantifiable under non-stressed growth suggesting that DMSP production is part of the basal metabolism of T. oceanica, similar to E. huxleyi, but at much lower concentrations (Table 1). The plasticity of DMSP production and consistent responses across multiple metabolic conditions suggest that elevated intracellular DMSP concentrations are essential for stressed growth in T. oceanica. A strong negative correlation between intracellular DMSP and Fv/Fm was observed across all experiments only for T. oceanica (Figure 6d). Previous studies have used changes in Fv/Fm to draw conclusions about the relationship between DMSP production and ROS damage as Fv/Fm is typically considered an oxidative stress marker (Bucciarelli et al. 2003; Harada et al. 2009; Archer et al. 2010; Darroch et al. 2015). However, Fv/Fm can also be impacted by the number and configuration of PSII reaction centers (Butler 1978). Changes in cellular Chla concentrations and photosystem proteins in response to nutrient status, but independent of oxidative stress, will result in a re-organization of PSII, and therefore a change in Fv/F m (Butler 1978; Hailemichael et al. 2016). The significant decrease of Fv/Fm in T. oceanica under hyper-saline conditions ("Fv/Fm = 0.2) may reflect oxidative stress (Jahnke and White 2003; Bussard et al. 2017), but all other small, yet significant, changes in Fv/Fm in our experiments ("Fv/Fm = c. 0.05) (Supp Figure 2) suggest that Fv/Fm was likely a signature of PSII re-organization, not of ROS damage. Therefore, we conclude that the strong correlation between intracellular DMSP and Fv/Fm in T. oceanica is not due to a direct linkage (e.g. ROS damage and therefore antioxidant function of DMSP) but because both are responding to the same environmental changed. Thus, our results suggest that DMSP is not being regulated as an antioxidant for T. oceanica as there was no indication of ROS damage co-occurring with elevated intracellular DMSP in temperature and N stressed growth. Unlike E. huxleyi where DMSP was predicted to contribute up to 100% of total organic osmolarity, intracellular DMSP was predicted to contribute a maximim c. 2% of total organic osmolarity in T. oceanica (Figures 2c,d). This small contribution suggests that the regulation of intracellular DMSP by T. oceanica is not driven by a compatible solute role. It is possible that DMSP could contribute significantly to cellular osmolarity if maintained within vacuoles or organelles (Lyon et al. 2016), though significant concentrations of DMSP appear to be stored in the cytoplasm (Raina et al. 2017). Finally, it has been proposed that DMSP may serve as an electron sink during metabolically imbalanced growth (Stefels 2000). However, T. oceanica 79 intracellular DMSP significantly increased at low temperatures, when growth was limited but metabolic balance is expected to be maintained (Barton et al. 2018). Of the current proposed DMSP mechanisms, our findings of DMSP upregulation across different metabolic conditions of stressed growth are most consistent with DMSP serving as a signaling molecule in T. oceanica (Seymour et al. 2010; Johnson et al. 2016). Microbial interactions are mediated by infochemicals and are critical for diatoms adapting to different environmental stressors (Amin et al. 2012, 2015; Arandia-Gorostidi et al. 2017). DMSP is a strong chemoattractant (Seymour et al. 2010) and therefore, if DMSP has evolved to serve as a signaling molecule in LoDPs under stress, intracellular DMSP production would be expected to increase across all conditions of stressed growth, as observed here. Future work incorporating recently discovered DMSP synthesis genes (Curson et al. 2018; Kageyama et al. 2018) and microfluidics will be critical for better understanding the role of low cellular DMSP concentrations in LoDPs (Lambert et al. 2017; Seymour et al. 2017). While this work only used two model species of DMSP producers, these findings support the hypothesis by McParland and Levine (2019) of differential regulation by HiDPs and LoDPs, which was based on a large number of previous monoculture experiments with many different strains (n = 15). This provides confidence that the differential production of DMSP by E. huxleyi and T. oceanica observed in this study is representative of other HiDPs and LoDPs. The contrasting strategies of DMSP regulation in HiDPs and LoDPs mirrors the recent discovery of two DMSP synthesis genes (DYSB and TpMT) that share little homology (Curson et al. 2018; Kageyama et al. 2018). The two genes appear to be differentially present based on HiDP and LoDP taxonomy (McParland and Levine 2019), suggesting that the HiDP and LoDP phenotypes evolved separately (McParland 2019). More advanced biochemistry methods are needed to investigate this potential evolutionary history and the two cellular mechanisms of DMSP. The first direct monoculture comparisons of HiDPs and LoDPs across multiple metabolic conditions presented here lays the foundation for future investigations to use omics-based approaches to further define the different cellular functions of DMSP in HiDPs and LoDPs. To provide a unifying framework for which to interpret 30 years of conflicting experiments on the cellular function of DMSP, we characterized changes in intracellular DMSP concentrations under metabolically balanced and imbalanced growth and under steady-state and non-steady state conditions in parallel experiments for a HiDP and LoDP. We found a consistent response of 80 intracellular DMSP to a wide range of environmental stressors for the LoDP (T. oceanica) and a different but also consistent response for the HiDP (E. huxleyi). This work suggests that DMSP serves a fundamentally different physiological role for HiDPs and LoDPs and thus production is also regulated differently. Specifically, our findings are most consistent with the primary role of DMSP being an essential compatible solute in HiDPs, independent of environmental stressors, and being a signaling molecule produced in response to stressed growth in LoDPs. Previous work has tried to understand variations in in situ DMSP production by assuming that DMSP serves a similar physiological function in all DMSP producers. Breaking DMSP cycling into two different ecological cycles with different underlying genes, different regulation, and different environmental drivers dramatically shifts understanding of in situ DMSP cycling. The constitutive regulation of high intracellular DMSP concentrations by HiDPs, independent of environmental stressors, explains why HiDPs always dominate in situ DMSP production, even in the most nutrient limited regions of the ocean where HiDPs are a sub-dominant community (McParland and Levine 2019). This dominant production of DMSP as a constitutive component of cellular growth likely means HiDPs contribute most significantly to atmospheric release of DMS and climate control at a global-scale. In contrast, LoDPs must use limited resources to maintain the fine-tuned regulation of DMSP production in response to environmental stressors. This suggests that DMSP is essential for stressed growth in LoDPs and that DMSP produced by LoDPs is likely more important for microbial interactions at a microscale. The paradigm of a universal mechanism for DMSP should be reconsidered in the context of this study. Future work should consider the importance of differential regulation across DMSP producer taxonomy presented here when quantifying the impact of DMSP on carbon cycling, climate and the marine microbial ecosystem. Acknowledgements This work was supported by funding from the Rose Hills Foundation, the University of Southern California, and a National Defense Science and Engineering Graduate Fellowship. We thank Kate Mackey, Eric Webb, Dave Hutchins and Seth John for insightful discussions about experimental design and the manuscript. We acknowledge Alexandra Koops and Emily Vainstein for assistance with sample collection. Finally, we would like to thank Ron Kiene for his inspiration, support and encouragement. 81 Figures and Tables Figure 1: Histogram of intracellular DMSP concentrations in all previously measured DMSP producers in replete conditions. Adapted from Supp. Table 1 in McParland and Levine (2019). 82 Figure 2: Change in intracellular DMSP and the predicted percent contribution of intracellular DMSP to organic osmolarity versus the relative fold change in µ (relative to treatment control) for E. huxleyi (a, c) and T. oceanica (b, d) under steady-state salinity, temperature and NO3 stress. Black dotted line at 0 represents no change in intracellular DMSP (a, b). Solid blue and magenta lines are significant linear regressions for E. huxleyi under temperatures ! T opt and hyper-saline conditions, respectively (a, c). Solid black lines are significant linear regressions for T. oceanica across all steady-state experiments (b, d). 83 Figure 3: Cell counts, Chla, and Chla per cell after NO3 add-back for E. huxleyi (a,b,c) and T. oceanica (d,e,f). Solid lines indicate the NO3 add-back treatment (+N). The dashed lines represent the control treatment (-N). Grey shading indicates dark period (14:10 light:dark cycle). Errorbars are error propagation of SD in biological triplicates. 84 Figure 4: Fv/Fm after N add-back for E. huxleyi (a) and T. oceanica (b). Solid lines indicate the NO3 add-back treatment (+N). The dashed lines represent the control treatment (-N). Grey shading indicates dark period (14:10 light:dark cycle). Errorbars are error propagation of SD in biological triplicates. 85 Figure 5: Intracellular DMSP after N add-back for E. huxleyi (a) and T. oceanica (b). Solid lines indicate the NO3 add-back treatment (+N). The dashed lines represent the control treatment (-N). Grey shading indicates dark period (14:10 light:dark cycle). Errorbars are error propagation of SD in biological triplicates. 86 Figure 6: Intracellular DMSP versus cellular Chla and Fv/Fm for E. huxleyi (a,c) and T. oceanica (b,d) across all experiments (both steady- and non-steady-state). Solid circles represent steady- state experiments, open circles represent non-steady-state experiments. Solid black lines are significant linear regressions. 87 Table 1: µ, Fv/Fm, intracellular DMSP and cellular Chla concentrations for T. oceanica and E. huxleyi in all steady-state experiments. ± represents SD for µ and error propagation of SD for Fv/Fm, intracellular DMSP and cellular Chla concentrations. 88 Supplemental Figures and Tables Supp Figure 1: Thermal response curve of µ and µin-vivo (µin-vivo = measured with in-vivo fluorescence) for E. huxleyi and T. oceanica. Errorbars are error propagation of SD in biological triplicates. 89 Supp Figure 2: Fv/Fm measured in E. huxleyi and T. oceanica relative to steady-state temperature (a,b), NO3 stress (c,d) and salinity treatments (e,f). Errorbars are error propagation of SD in biological triplicates. 90 Supp Figure 3: Intracellular DMSP measured in E. huxleyi and T. oceanica relative to steady-state temperature (a,b), NO3 stress (c,d) and salinity treatments (e,f). Errorbars are error propagation of SD in biological triplicates. 91 Supp Figure 4: Fv/F m versus NO3 concentration measured in steady-state exponential N stress. 92 Supp Figure 5: Fv/Fm diel cycle for E. huxleyi (a) and T. oceanica (b) in Nss + and Nss - growth conditions and NO3 add-back. 93 Supp Figure 6: Total DMSP (DMSPt) after NO3 add-back for E. huxleyi (a) and T. oceanica (b). Solid lines indicate the NO3 add-back treatment (+N). The dashed lines represent the control treatment (-N). Grey shading indicates dark period (14:10 light:dark cycle). Errorbars are error propagation of SD in biological triplicates. 94 Supp Figure 7: Dilution of intracellular DMSP in T. oceanica after NO3 add-back due to cell division. The blue line represents the observed values (main text, Figure 4). The black line represents intracellular DMSP calculated holding the initial DMSPt concentration before NO3 addback constant, divided by the observed changes in cellular concentrations and biovolume. 95 Supp Table 1: The number of measurements of cell diameter and height made to calculate cellular biovolumes of T. oceanica and E. huxleyi measured across all experiments. Cellular biovolumes for each species did not significantly change across treatments in any of the experiments (t-test, p>0.1) and therefore for a more direct comparison, the mean cellular biovolume was used to calculate intracellular DMSP. 96 Supp Table 2: The following values were used to predict the percent contribution of intracellular DMSP to total organic osmolarity: calculated osmolarity of media in each steady-state treatment (osmol · m -3 ), the predicted concentrations of major ions (mM), ionic osmolarity (sum of major ions, mM), the percent contribution of inorganic ions to total cellular osmolarity, and the predicted organic osmolarity needed to balance the total cellular osmolarity (osmol · m -3 ), and the measured intracellular DMSP concentration (mM). 97 Supplemental Note During the non-steady-state NO3 add-back experiment, both +N E. huxleyi and +N T. oceanica and -N T. oceanica exhibited a midday Fv/Fm minimum that was significantly lower than Fv/Fm at the beginning of the add-back (t-test, p!0.05) (Main text, Figure 3). This diel variation in Fv/Fm was also present in steady-state N replete and N stress T. oceanica (Supp Figure 5), suggesting that the mid-day minimum in T. oceanica Fv/Fm was an inherent component of the species’ PSII reaction centers, independent of nutrient status and similar to that observed previously in natural communities of diatoms and picoplankton (Villareal, 2004; Mackey et al., 2008). In contrast, the diel variability in Fv/Fm was absent in steady-state N replete and N stress E. huxleyi (Supp Figure 5), and the midday minimum in +N E. huxleyi corresponds to the significant increase in Chla concentrations at 6 h (Main text, Figure 2b). Therefore, the midday minimum in Fv/Fm likely reflects a response to the NO3 add-back in E. huxleyi, but not T. oceanica. 98 References Acosta-Motos, J., M. Ortuño, A. Bernal-Vicente, P. Diaz-Vivancos, M. Sanchez-Blanco, and J. Hernandez. 2017. Plant Responses to Salt Stress: Adaptive Mechanisms. Agronomy 7: 18. doi:10.3390/agronomy7010018 Amin, S. A., L. R. Hmelo, H. M. van Tol, and others. 2015. Interaction and signalling between a cosmopolitan phytoplankton and associated bacteria. Nature 522: 98–101. doi:10.1038/nature14488 Amin, S., M. Parker, and E. Armbrust. 2012. Interactions between diatoms and bacteria. Microbiol. Mol. Biol. Rev. 76: 667–684. doi:10.1128/MMBR.00007-12 Arandia-Gorostidi, N., P. K. Weber, L. Alonso-Sáez, X. A. G. Morán, and X. Mayali. 2017. Elevated temperature increases carbon and nitrogen fluxes between phytoplankton and heterotrophic bacteria through physical attachment. ISME J. 11: 641–650. doi:10.1038/ismej.2016.156 Archer, S. D., M. Ragni, R. Webster, R. L. Airs, and R. J. Geider. 2010. Dimethylsulfoniopropionate and dimethylsulfide production in response to photoinhibition in Emiliania huxleyi. Limnol. Oceanogr. 55: 1579–1589. doi:10.4319/lo.2010.55.4.1579 Arnold, H. E., P. Kerrison, and M. Steinke. 2013. Interacting effects of ocean acidification and warming on growth and DMS-production in the haptophyte coccolithophore Emiliania huxleyi. Glob. Chang. Biol. 19: 1007–16. doi:10.1111/gcb.12105 Baker, K. G., C. M. Robinson, D. T. Radford, A. S. Mcinnes, C. Evenhuis, and M. A. Doblin. 2016. Thermal Performance Curves of Functional Traits Aid Understanding of Thermally Induced Changes in Diatom-Mediated Biogeochemical Fluxes. Front. Mar. Sci. 3: 1–14. doi:10.3389/fmars.2016.00044 Barton, S., J. Jenkins, A. Buckling, C. Schaum, and N. Smirnoff. 2018. Universal metabolic constraints on the thermal tolerance of marine phytoplankton. bioRxiv doi:10.1101/358002. Bochenek, M., G. J. Etherington, A. Koprivova, S. T. Mugford, T. G. Bell, G. Malin, and S. Kopriva. 2013. Transcriptome analysis of the sulfate deficiency response in the marine microalga Emiliania huxleyi. New Phytol. 199: 650–662. doi:10.1111/nph.12303 Boyd, C. M., and D. Gradmann. 2002. Impact of osmolytes on buoyancy of marine phytoplankton. Mar. Biol. 141: 605–618. doi:10.1007/s00227-002-0872-z Bucciarelli, E., C. Ridame, W. G. Sunda, C. Dimier-Hugueney, M. Cheize, and S. Belviso. 2013. 99 Increased intracellular concentrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum. Limnol. Oceanogr. 58: 1667–1679. doi:10.4319/lo.2013.58.5.1667 Bucciarelli, E., W. G. Sunda, and P. Nicolas. 2003. Influence of CO2, nitrate, phosphate, and silicate limitation on intracellular DMSP of the coastal diatom Thalassiosira pseudonana in batch cultures. Limnol. Oceanogr. 48: 2256–2265. Bussard, A., E. Corre, C. Hubas, and others. 2017. Physiological adjustments and transcriptome reprogramming are involved in the acclimation to salinity gradients in diatoms. Environ. Microbiol. 19: 909–925. doi:10.1111/1462-2920.13398 Butler, W. L. 1978. Energy distribution in the photochemical apparatus of photosynthesis. Annu. Rev. Plant Physiol. 29: 345–378. Caruana, A. M. N., and G. Malin. 2014. 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 Charlson, R., J. Lovelock, M. Andrae, and S. Warren. 1987. Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Nature 326: 655–661. Curson, A., B. Williams, B. Pinchbeck, and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat. Microbiol. 4: 430– 439. doi:10.1038/s41564-018-0119-5 Darroch, L., M. Lavoie, M. Levasseur, and others. 2015. Effect of short-term light- and UV- stress on DMSP, DMS, and DMSP lyase activity in Emiliania huxleyi. Aquat. Microb. Ecol. 74: 173–185. doi:10.3354/ame01735 Dickson, D., and G. Kirst. 1987a. Osmotic adjustment in marine eukaryotic algae: The role of inorganic ions, quaternary ammonium, tertiary sulphonium and carbohydrate solutes. I. Diatoms and a rhodophyte. New Phytol. 106: 645–655. Dickson, D. M. J., and G. O. Kirst. 1987b. Osmotic adjustment in marine eukaryotic algae: The role of inorganic ions, quaternary ammonium, tertiary sulphonium and carbohydrate solutes. I. Diatoms and a rhodophyte. New Phytol. 106: 645–655. Galí, M., R. Simó, G. L. Pérez, C. Ruiz-González, H. Sarmento, S. J. Royer, A. Fuentes-Lema, and J. M. Gasol. 2013. Differential response of planktonic primary, bacterial, and dimethylsulfide production rates to static vs. dynamic light exposure in upper mixed-layer 100 summer sea waters. Biogeosciences 10: 7983–7998. doi:10.5194/bg-10-7983-2013 Gebser, B., and G. Pohnert. 2013. Synchronized regulation of different zwitterionic metabolites in the osmoadaption of phytoplankton. Mar. Drugs 11: 2168–2182. doi:10.3390/md11062168 Geider, R. J. 1987. Light and Temperature Dependence of the Carbon to Chlorophyll a Ratio in Microalgae and Cyanobacteria: Implications for Physiology and Growth of Phytoplankton. New Phytol. 106: 1–34. Hailemichael, G., A. Catalina, M. Gonzalez, and P. Martin. 2016. Relationships between water status, leaf chlorophyll content and photosynthetic performance in Tempranillo Vineyards. South African J. Enol. Vitic. 37: 149–156. Harada, H., M. Vila-costa, J. Cebrian, and R. P. Kiene. 2009. 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 Harrison, J., and A. Berges. 2005. Marine culture medium, In R. Andersen [ed.], Algal culturing techniques. Hockin, N. L., T. Mock, F. Mulholland, S. Kopriva, and G. Malin. 2012. 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 Iglesias-rodriguez, M. D., P. R. Halloran, R. E. M. Rickaby, and others. 2008. Phytoplankton calcification in a high CO2 world. Science (80-. ). 320: 336–340. doi:10.1126/science.1154122 Jahnke, L. S., and A. L. White. 2003. Long-term hyposaline and hypersaline stresses produce distinct antioxidant responses in the marine alga Dunaliella tertiolecta. J. Plant Physiol. 160: 1193–1202. doi:10.1078/0176-1617-01068 Johnson, W. M., M. C. K. Soule, and E. B. Kujawinski. 2016. Evidence for quorum sensing and differential metabolite production by the marine heterotroph, Ruegeria pomeroyi, in response to DMSP. ISME J. 10: 2304–2316. doi:10.1038/ismej.2016.6 Kageyama, H., Y. Tanaka, A. Shibata, R. Waditee-Sirisattha, and T. Takabe. 2018. Dimethylsulfoniopropionate biosynthesis in a diatom Thalassiosira pseudonana: Identification of a gene encoding MTHB-methyltransferase. Arch. Biochem. Biophys. 645: 100–106. doi:10.1016/j.abb.2018.03.019 101 Karsten, U., K. Kück, C. Vogt, and G. O. Kirst. 1996. Dimethylsulfoniopropionate production in phototrophic organisms and its physiological functions as a cryoprotectant, p. 143–153. In Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds. Keller, M., W. Bellows, and R. Guillard. 1989. Dimethyl sulfide production in marine phytoplankton, p. 167–182. In Biogenic sulfur in the environment. Keller, M. D. 1989. 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 Keller, M. D., R. P. Kiene, P. A. Matrai, and W. K. Bellows. 1999a. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. II. N-limited chemostat cultures. Mar. Biol. 135: 249–257. doi:10.1007/s002270050622 Keller, M. D., R. P. Kiene, P. A. Matrai, and W. K. Bellows. 1999b. Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. I. Batch cultures. Mar. Biol. 135: 237–248. doi:10.1007/s002270050621 Keller, M. D., and W. Korjeff-Bellows. 1996. Physiological Aspects of the Production of Dimeyhtlsulfoniopropionate (DMSP) by Marine Phytoplankton, p. 131–142. In Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds. Kettles, N. L., S. Kopriva, and G. Malin. 2014. Insights into the regulation of DMSP synthesis in the diatom Thalassiosira pseudonana through APR activity, proteomics and gene expression analyses on cells acclimating to changes in salinity, light and nitrogen. PLoS One 9: e94795. doi:10.1371/journal.pone.0094795 Kiene, R. P., L. J. Linn, and J. A. Bruton. 2000. New and important roles for DMSP in marine microbial communities. J. Sea Res. 43: 209–224. Kiene, R. P., and S. K. Service. 1991. 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 Lambert, B. S., J. B. Raina, V. I. Fernandez, and others. 2017. A microfluidics-based in situ chemotaxis assay to study the behaviour of aquatic microbial communities. Nat. Microbiol. 2: 1344–1349. doi:10.1038/s41564-017-0010-9 Lana, A., R. Simó, S. M. Vallina, and J. Dachs. 2012. Re-examination of global emerging patterns of ocean DMS concentration. Biogeochemistry 110: 173–182. doi:10.1007/s10533- 102 011-9677-9 Lavoie, M., M. Levasseur, and M. Babin. 2015. Testing the potential ballast role for dimethylsulfoniopropionate in marine phytoplankton: A modeling study. J. Plankton Res. 37: 699–711. doi:10.1093/plankt/fbv050 Levine, N. M., D. A. Toole, A. Neeley, N. R. Bates, S. C. Doney, and J. W. H. Dacey. 2015. Revising upper-ocean sulfur dynamics near Bermuda: New lessons from 3 years of concentration and rate measurements. Environ. Chem. 13: 302–313. doi:10.1071/EN15045 Lin, Q., J. Liang, Q. Huang, and others. 2017. Differential cellular responses associated with oxidative stress and cell fate decision under nitrate and phosphate limitations in Thalassiosira pseudonana#: Comparative proteomics. PLoS One 12: e0184849. Lyon, B. R., J. M. Bennett-Mintz, P. A. Lee, M. G. Janech, and G. R. Ditullio. 2016. 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 Macler, B. A. 1988. Salinity Effects on Photosynthesis, Carbon Allocation, and Nitrogen Assimilation in the Red Alga, Gelidium coulteri. Plant Physiol. 88: 690–694. doi:10.1104/pp.88.3.690 McParland, E. L., and N. M. Levine. 2019. The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations. Limnol. Oceanogr. 64: 757–773. doi:10.1002/lno.11076 Nishiguchi, M. K., and G. N. Somero. 1992. Temperature- and concentration-dependence of compatibility of the organic osmolyte β-dimethylsulfoniopropionate. Cryobiology 29: 118– 124. doi:10.1016/0011-2240(92)90011-P Qasim, S. Z., P. M. A. Bhattathiri, and V. P. Devassy. 1972. The influence of salinity on the rate of photosynthesis and abundance of some tropical phytoplankton. Mar. Biol. 12: 200–206. doi:10.1007/BF00346767 Raina, J. B., P. Clode, S. Cheong, and others. 2017. Subcellular tracking reveals the location of dimethylsulfoniopropionate in microalgae and visualises its uptake by marine bacteria. Elife 6: e23008. doi:10.7554/eLife.23008 Raven, J. A., and M. A. Doblin. 2014. Active water transport in unicellular algae: Where, why, and how. J. Exp. Bot. 65: 6279–6292. doi:10.1093/jxb/eru360 103 van Rijssel, M., and A. G. J. Buma. 2002. UV radiation induced stress does not affect DMSP synthesis in the marine prymnesiophyte Emiliania huxleyi. Aquat. Microb. Ecol. 28: 167– 174. doi:10.3354/ame028167 van Rijssel, M., and W. W. C. Gieskes. 2002. Temperature, light, and the dimethylsulfoniopropionate (DMSP) content of Emiliania huxleyi (Prymnesiophyceae). J. Sea Res. 48: 17–27. doi:10.1016/S1385-1101(02)00134-X Seymour, J. R., S. A. Amin, J. B. Raina, and R. Stocker. 2017. Zooming in on the phycosphere: The ecological interface for phytoplankton-bacteria relationships. Nat. Microbiol. 2: 1–12. doi:10.1038/nmicrobiol.2017.65 Seymour, J. R., R. Simó, T. Ahmed, and R. Stocker. 2010. Chemoattraction to Dimethylsulfoniopropionate throughout the Marine Microbial Food Web. Science (80-. ). 329: 342–345. doi:10.1126/science.1188418 Sikes, C., and K. Wilbur. 1982. Functions of coccolith formation. Limnol. Oceanogr. 27: 18–26. Stefels, J. 2000. Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43: 183–197. Stefels, J., J. W. H. Dacey, and J. T. M. Elzenga. 2009. In vivo DMSP-biosynthesis measurements using stable-isotope incorporation and proton-transfer-reaction mass spectrometry (PTR-MS). Limnol. Oceanogr. Methods 7: 595–611. doi:10.4319/lom.2009.7.595 Stefels, J., M. Steinke, S. Turner, G. Malin, and S. Belviso. 2007. 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 Steinke, M., G. V Wolfe, and G. Kirst. 1998. Partial characterisation of dimethylsulfonio- propionate (DMSP) lyase isozymes in 6 strains of Emiliania huxleyi. Mar. Ecol. Prog. Ser. 175: 215–225. Sunda, W. G., R. Hardison, R. P. Kiene, E. Bucciarelli, and H. Harada. 2007. 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 Sunda, W., D. J. Kieber, R. P. Kiene, and S. Huntsman. 2002. An antioxidant function for DMSP and DMS in marine algae. Nature 418: 317–320. 104 Thomas, M. K., C. T. Kremer, C. a. Klausmeier, and E. Litchman. 2012. A Global Pattern of Thermal Adaptation in Marine Phytoplankton. Science (80-. ). 338: 1085–1089. doi:10.1126/science.1224836 Tripp, H. J., J. B. Kitner, M. S. Schwalbach, J. W. H. Dacey, L. J. Wilhelm, and S. J. Giovannoni. 2008. SAR11 marine bacteria require exogenous reduced sulphur for growth. Nature 452: 741–744. doi:10.1038/nature06776 Turner, S. M., G. Malin, P. S. Liss, D. S. Harbour, and P. M. Holligan. 1988. The seasonal variation of dimethylsulfide and dimethylsulfoniopropionate concentrations in nearshore waters. Limnol. Oceanogr. 33: 364–375. doi:10.4319/lo.1988.33.3.0364 del Valle, A., D. Slezak, C. M. Smith, A. N. Rellinger, D. J. Kieber, and R. P. Kiene. 2011. Effect of acidification on preservation of DMSP in seawater and phytoplankton cultures#: Evidence for rapid loss and cleavage of DMSP in samples containing. Mar. Chem. 124: 57– 67. doi:10.1016/j.marchem.2010.12.002 Welschemeyer, N. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll b and pheopigments. Limnol. Oceanogr. 39: 1985–1992. doi:10.1111/j.1540- 5931.2009.00571.x Wood, A. M., R. C. Everroad, and L. M. Wingard. 2005. Measuring growth rates in microalgal cultures, p. 269–286. In R.A. Andersen [ed.], Algal culturing techniques. Elsevier Academic Press. Wordenweber, R., S. D. Rokitta, E. Heidenreich, and others. 2018. Phosphorus and nitrogen starvation reveal life-cycle specific responses in the metabolome of Emiliania huxleyi (Haptophyta). Limnol. Oceanogr. 63: 203–226. doi:10.1002/lno.10624 105 Chapter Three: Using DMSP synthesis genes as marker genes for high and low DMSP producers to predict in situ DMSP production Erin L. McParland 1 , Bonnie Laverock 2 , Sarah Hu 3 , Michael D. Lee 4 , Harriet Alexander 3 , Justin Seymour 2 , and Naomi M. Levine 1 1 Department of Marine and Environmental Biology, University of Southern California, Los Angeles, California 2 Climate Change Cluster, University of Technology Sydney, Australia 3 Woods Hole Oceanographic Institution, Marine Chemistry and Geochemistry, Woods Hole, MA, USA 4 NASA Ames Research Center, California, USA Abstract Dimethylsulfoniopropionate (DMSP) is a globally abundant metabolite produced primarily by marine primary producers that has the potential to influence the cycling of marine organic carbon and Earth’s climate. In order to understand how DMSP production will change in future oceans, a better mechanistic understanding of the environmental drivers of DMSP production is needed. Recently it was proposed that the abundance of high DMSP producers (HiDPs, contain >50mM intracellular DMSP) is responsible for variability of in situ DMSP concentration variability. However, HiDPs are typically a sub-dominant part of the marine microbial community and accurately capturing their abundance is difficult. Here, we confirm this hypothesis by conducting natural community incubations in four contrasting oceanographic regimes where NO3 stress was alleviated and changes in the DMSP producer community and DMSP production were tracked over the course of the grow-out. We demonstrate that two recently discovered eukaryotic DMSP synthesis genes are differentially present in transcriptomes of HiDPs and low DMSP producers (LoDPs, contain <50mM DMSP), and for the first time, use these synthesis genes to predict the abundances of the HiDP and LoDP communities. We confirm that HiDP abundance drives changes in DMSP production after NO3 add-back and demonstrate the potential for using a HiDP marker gene to quantify the abundance of HiDPs. Future work to better characterize the 106 HiDP marker gene will result in accurate prediction of in situ DMSP concentrations and a better method for predicting how HiDP abundances will change in future oceans. Introduction Dimethylsulfoniopropionate (DMSP) is a globally abundant labile sulfur and carbon metabolite produced in the surface ocean primarily by marine phytoplankton (McParland and Levine 2019). DMSP can account for up to 11% of phytoplankton cellular carbon (Stefels et al. 2007) and, in both coastal and open ocean regimes, DMSP production may account for up to 5% of total primary production (Galí et al. 2013; Levine et al. 2015). Once released into the water column, dissolved DMSP plays a critical role in the marine microbial loop, supplying up to 13% and 100% of the bacterial carbon and sulfur demand, (Levine et al. 2015; Kiene et al. 2000; Tripp et al. 2008). One of the degradation products of DMSP, dimethylsulfide (DMS), is a volatile trace gas estimated to be the most significant natural source of sulfur to the atmosphere and thought to play an important role in climate regulation as a source of cloud condensation nuclei (Charlson et al. 1987; Lana et al. 2012). While the strength of the negative feedback loop due to atmospheric DMS oxidation is debated (Quinn and Bates 2011), there are multiple lines of evidence that changes in DMS(P) production in future oceans due to global climate change will result in potential feedback mechanisms (Six et al. 2013). Significant decreases in oceanic DMS emissions by future changes in community composition or ocean chemistry are predicted to result in a 1.6ºC increase in global temperature (Gunson et al. 2006). In addition, shifts in DMSP production have the potential to significantly alter microbial community composition and rates of carbon cycling. DMSP is utilized as a growth substrate by a diverse array of marine heterotrophs (Reisch et al. 2011). Specifically, the SAR11 lineage, which are sulfur auxotrophs, significantly contributes to global carbon cycling as the most abundant bacterioplankton and DMSP is a significant source of its sulfur nutrition (Morris et al. 2002; Tripp et al. 2008). A better understanding of the environmental controls on DMSP production by marine primary producers is critical for understanding how the supply of DMSP will change in future oceans and predicting any subsequent impacts on the climate. Previous in situ mesocosm experiments investigating how DMSP production will change in the future ocean are conflicting (Lee et al. 2011; Archer et al. 2013; Hopkins and Archer 2014; Webb et al. 2016), and responses to climate change stressors vary significantly by species 107 (Spielmeyer and Pohnert 2012; Arnold et al. 2013; Webb et al. 2016). Reconciling these studies requires a mechanistic understanding of the environmental drivers of DMSP production. Using a synthesis of previous studies and the global DMSP database (Kettle et al. 1999; Lana et al. 2011), McParland and Levine (2019) conclude that high DMSP producers (HiDPs) dominate in situ DMSP production due to their high cellular concentrations. As a result, they show that HiDP biomass alone is sufficient for predicting in situ bulk DMSP concentrations and that physiological changes in DMSP production due to environmental stress do not play a primary role in determining in situ DMSP concentrations. Accurate quantification of the HiDP community, not the physiological responses to environmental drivers, is necessary for predicting in situ DMSP in future oceans. Dinoflagellates and haptophytes are the most well-known HiDPs with intracellular DMSP concentrations >>50 mM. However, in much of the global oceans, HiDPs are the sub-dominant community and so accurate estimates of the biomass for these groups is hard to determine (Carradec et al. 2018). On the other hand, more abundant primary producers (picocyanobacteria and diatoms) are typically non-DMSP producers or low DMSP producers (LoDPs) with <50mM intracellular DMSP concentrations. Even though LoDPs significantly upregulate cellular DMSP in response to environmental stress, they are not predicted to contribute significantly to bulk DMSP concentrations (McParland and Levine 2019, McParland et al. in prep). Changes in global climate will significantly impact primary producer community structure (Dutkiewicz et al. 2013, 2019) and therefore quantifying how HiDP abundance will change will be critical for predicting how in situ DMSP will change in the future. Despite the biosynthesis pathway for DMSP production in algae being characterized biochemically 20 years ago (Gage et al. 1997), it was only recently that genes for eukaryotic phytoplankton DMSP synthesis were discovered (Curson et al. 2018; Kageyama et al. 2018). The genes (DSYB, TpMT1 and TpMT2) are SAM dependent methyltransferases that perform the same critical third step of DMSP synthesis of adding the second methyl group to methylthiohydroxybutryate (MTHB). All three genes encode for the same pathway, suggesting that they are isoforms (Curson et al. 2018). DSYB appears to be primarily found in dinoflagellates and haptophytes (HiDPs), while TpMT1 and TpMT2 have only been reported in diatoms (Curson et al. 2018; Kageyama et al. 2018). Therefore, we hypothesize that these synthesis genes may differentiate between HiDP and LoDP species, and that the differential regulation of DMSP 108 production in these groups (McParland and Levine 2019, McParland et al. in prep) may be due to different regulation of these genes. Quantification of DMSP producers with a marker gene would significantly improve the ability to predict variability of in situ DMSP production. However, to date it has not been possible to target these genes in situ as universal primers have not been developed due to species specificity of sequences. Previous studies observe conflicting responses of mixed natural communities to different future climate change stressors. As well, HiDP abundance is hypothesized to drive variability of in situ particulate DMSP concentrations. However, as HiDPs are typically the sub-dominant community, quantifying their abundance is difficult. Previous studies have correlated HiDP pigment markers to DMSP concentrations (Riseman and DiTullio 2004), but this method cannot capture the entire HiDP community and the correlation does not always exist, particularly in the oligotrophic oceans (Bell et al. 2010). Here, we demonstrate that DSYB and TpMT1/TpMT2 can differentiate between HiDPs and LoDPs by quantifying the prevalence of these synthesis genes in marine eukaryotic transcriptomes, and show that DSYB can be used as a marker gene for quantifying HiDP abundance by conducting the first analysis of DMSP synthesis genes in situ and show that the marker gene successfully tracks HiDPs in mixed natural communities. We track the response of DMSP production in natural community grow-out experiments in four contrasting oceanographic regimes while simultaneously measuring the DMSP producer community by assigning putative DMSP producer functionality to OTUs of 18S ssu rRNA sequences. By quantifying the DMSP producer community, we confirm the hypothesis that HiDPs dominate in situ DMSP production. Methods In situ NO3 add-back experiment: Grow-out experiments were conducted in four contrasting oceanographic sites: a coastal, oligotrophic regime on the western side of the Great Barrier Reef, Australia (herein referred to as “Oligo 1”: 18.585°S, 146.981°E); a coastal regime north of Vancouver Island, Canada (“Coast 1”: 50.973°N, 128.224°W); an open ocean, oligotrophic regime off the coast of Washington, USA (“Oligo 2”: 45.315°N, 126.618°W); and a coastal, nutrient replete regime at Catalina Island, California, USA (“Coast 2”: 33.445°N, 118.484°W). At all sites, 60L of surface water was collected from Niskin bottles and pre-filtered through acid-washed Nitex mesh directly into an acid-washed carboy. Pre-filtered water was then dispensed into 6 acid- 109 washed bottles, and 3 bottles were spiked with NaNO3. Final NaNO3 concentration at Oligo 1 (1µM) was lower than the other sites (~20µM) to avoid potential poisoning of the extremely low biomass system. Bottles were immediately placed in a deckboard incubator connected to underway seawater flowthrough to maintain in situ temperatures and covered with black mesh that filtered 50% of ambient sunlight (resulting in 200-500 mmol quanta·m -2 ·s -1 ). All incubations were tracked in real-time by subsampling 3mL of incubation water from each bottle to measure in-vivo fluorescence. Incubations were started in the early morning (at or before sunrise) and incubated for 36 hours, except Coast 2 which was incubated for 42 hours as there were no significant differences in in-vivo fluorescence after 36 hours. All incubations were sampled for starting conditions from the original carboy water collected to set-up experiments, and all were sampled for a final timepoint at 36 hours (except Coast 2 which was sampled at 42 hours). Incubation bottles were gently rotated three times before sampling for nutrients, chlorophyll-a, DMSP total and community analyses. Sampling timepoints over the course of the experiment differed slightly at each site and are described in Supp Table 1 along with sampling date, depth, and filter size. Nutrients: At each timepoint, 8mL of sample was filtered (Supp Table 1) and stored at -20°C. NO3+NO2, PO4 and SiOH4 were measured onboard the R/V Investigator for Oligo 1. For all other sites, NO3+NO2 and PO4 were measured by the UCSB MSI Analytical Lab. Both facilities used a Lachat Instrument QuikChem Flow Injection Analysis System to simultaneously measure nutrients following the protocols of Knpele and Bogren 2002 and Smith and Bogren 2001. Chlorophyll-a: At each timepoint, a minimum 200mL of sample was gently vacuum filtered onto a Whatman 25mm GF/F and stored at -20°C. Filters were extracted in 2.5 mL of 90% HPLC-grade acetone and analyzed with a Turner Trilogy fluorometer. DMSP: At each timepoint, 50mL of sample was collected in Falcon tubes and immediately acidified to pH 2 with 50% H 2SO4 (3.3 µL per 1mL of sample). Total DMSP samples were stored for a minimum of 24 hours to allow for the complete oxidation of DMS. Before analysis, an aliquot of sample was dispensed into an acid-washed, combusted, 14mL serum vial with 1mL of 5N NaOH and crimped closed with a gas-tight Teflon lid. Samples were vortexed and then incubated for 10 minutes, which has been determined to be a sufficient reaction time for DMSP to be cleaved via 110 alkaline hydrolysis to DMS (Kiene and Service 1991). DMSP (derivitized to DMS) was then quantified via purge and trap analysis (Kiene and Service 1991). The sample was directly connected to a sparging needle and sparged with ultrapure helium. Sparged DMS was passed through a Nafion dryer to remove water vapor and cryotrapped on a Teflon loop immersed in liquid nitrogen. The DMS was volatilized by rapidly switching the cryotrap loop into hot water, followed by a switching of a six-way-port valve to inject DMS onto a custom Shimadzu 2016 gas chromatograph equipped with a flame photometric detector and a Chromosil 330 packed column (Supelco) for quantification. Column temperature was 70ºC and retention time was 0.92 minutes. DMSP concentrations were calculated with a 4-point DMS standard curve (R 2 > 0.92). Experimental replication and technical triplicates were quantified. DNA sampling, extraction and sequencing of 18S rRNA genes: At each timepoint, a minimum 1000mL was gently vacuum filtered from each bottle (see Supp Table 1 for filter type). Filters were placed in sterile cryovials dry, flash-frozen, and stored at -80ºC. DNA was extracted with a DNeasy PowerWater Kit (Qiagen) following the manufacturer’s protocol with additional lysis steps added to insure complete lysis of heterogenous eukaryotic cells (Hu et al. 2018b). After placing dry filters into Qiagen bead tubes, tubes with filter and lysis buffer were heated at 65ºC for 5 min and vortexed for 5 min, 3-times. MilliQ water (1000mL) was vacuum filtered onto the three different filter types and extracted as blanks. DNA concentrations after extraction were quantified with the Qubit HS dsDNA Assay. DNA was PCR amplified using forward (5!- CCAGCA[GC]C[CT]GCGGTAATTCC-3!) and reverse (5!-ACTTTCGTTCTTGAT[CT][AG]A- 3!) primers that targeted the V4 hypervariable region (Stoeck et al. 2010). The PCR reactions were performed in duplicate in 50uL volumes consisting of 1x Q5 High Fidelity Master Mix (New England Biolabs, #M0492S, Ipswich, MA), 0.5µM forward primer, 0.5µM reverse primer and 1ng of genetic material (final concentrations). The PCR thermal profile from Hu et al. 2018 was used: initial activation step (Q5 activation temperature) of 98ºC for 2 min, followed with 10 cycles of 98ºC for 10 s, 53ºC for 30 s, 72ºC for 30 s, and 15 cycles of 98ºC for 10 s, 48ºC for 30 s, and 72ºC for 30 s, and a final extension of 72ºC for 2 min. Successful PCR amplification of the V4 region was confirmed with gel chromatography using 10% TAE buffer and a 1kb DNA ladder (Promega, #G7541). PCR products were sent to a commercial vendor (Laragen, Culver City, CA, USA) for PCR clean-up, library preparation and Illumina MiSeq paired-end 2x250 bp sequencing. 111 Sequence processing and OTU analyses: The sequence processing pipeline was primarily conducted in Qiime1 (v 1.9.1) (Caporaso et al. 2010), unless otherwise noted, and is identical to the protocol used by Hu et al. (2018). Raw sequences were de-multiplexed by the sequencing center. Reads of paired end sequences were merged requiring a minimum 20bp overlap with fastqjoin (v 1.3.1), and merged sequences were filtered using a Q score of 30 (Caporaso et al. 2010). Primers were trimmed using cutadapt (v 1.18) (Martin 2011) allowing for a 30% error. Any resulting sequences shorter than 150bp or longer than 500bp were removed. Pooled chimera checks were performed de novo and reference-based on quality controlled sequences with vsearch (v 2.8.0) (Rognes et al. 2016) and the Protist Ribosomal (PR2) database (release 4.10) (Guillou et al. 2013). Finally, sequences were clustered into operational taxonomic units (OTUs) at 97% sequence similarity. OTUs were generated with a subsampling open-reference OTU clustering algorithm (Rideout et al. 2014) that combined both de novo and closed-reference (with the PR2 database) OTU clustering. Taxonomy based on the PR2 database was assigned to the OTUs with uclust (v 1.2.22) (Edgar 2010) at 90% similarity. The remaining analyses were performed in RStudio (v 3.3.3; RStudio Team 2016). Global tripletons (OTUs with only 3 sequences or less across all samples) were removed. Potential cross-contamination was accounted for by removing OTUs with an average total sequence count 10-fold higher in the blanks than the average total count across all samples (Lee et al. 2015). Samples were rarified with the vegan package (Oksanen et al. 2019) by randomly subsampling each so that total sequence number in all samples was equal to the lowest total sequence value. Putative DMSP production: A custom DMSP producer database of full length 18S sequences was built using monocultures of eukaryotic protists with confirmed DMSP production (Supp Table 1 in McParland and Levine 2019). A total of 119 species were included, of which, 61 were classified as high producers and 58 as low producers. DMSP production was assigned to the add-back experiment OTUs by BLASTing each against the DMSP database with BLASTn (Altschul et al. 1990). Putative high or low DMSP production was assigned to an OTU if the query sequence was >= 90% identical across at least 90% of the reference sequence in the DMSP producer database. 112 DMSP synthesis genes in MMETSP: A custom nucleotide BLAST database was built with the recently re-assembled MMETSP transcriptomes (n=678) (Johnson et al. 2018). MMETSP transcriptomes contain n=393 unique strains, with an additional 285 transcriptomes from the same strains cultured in different conditions. Corresponding genomes are not available for MMETSP. The entire database was used to identify DMSP synthesis genes in order to account for potential differences in transcription of the genes under different conditions. It is possible that a strain is identified as a non-DMSP producer because transcription did not occur under the tested conditions. However, given that constitutive production is observed in HiDPs, this is most likely only an issue for LoDPs. tBLASTn was used to search the database for the amino acid sequences of DSYB (Scrippsiella trochoidea CCMP3099-1 from Curson et al. 2018), TpMT1 (Accession XP_002296978) and TpMT2 (Accession XP_002291473). Significant hits for the genes (E value ≤ 1e -30 ) are presented as presence or absence of the total unique strains, though some strains had multiple copies of a single gene. It should be noted that MMETSP is not axenic (Keeling et al. 2014) and the eukaryotic DSYB used has high homology with the bacterial dsyB DMSP synthesis gene (Curson et al. 2017). Future analyses will use a recursive BLAST of each MMETSP sequence with a significant hit to a DMSP synthesis gene against the NCBI RefSeq database to further confirm them as likely of eukaryotic origin, or to be removed if found to be bacterial. Putative DMSP synthesis gene presence: A reference alignment of full-length 18S sequences from unique MMETSP strains was built with Infernal (v 1.1.2) (Nawrocki and Eddy 2013) using the RFam (RF01960) (Kalvari et al. 2018)for eukaryotic ssu rRNA. The alignment was manually curated in JalView (v 2.10.5) (Waterhouse et al. 2009)to remove large gaps due to single sequences. The reference tree was built in RAxML (v 8.2.12) (Stamatakis 2014). Phylogenetic placement of add-back experiment OTUs onto the 18S MMETSP phylogeny was performed with pplacer (v 1.1) (Matsen et al. 2010) keeping the top 10 most likely placements. The highest-weight placement for each query was extracted and placements with posterior probability >90% were kept. An OTU with significant placement on an edge of an MMETSP strain containing a DMSP synthesis gene (based on the transcriptome BLAST searches described above) was putatively assigned to also contain the respective synthesis gene. 113 Calculations: We conducted a set of calculations to confirm the hypothesis that in situ DMSP in the add-back experiments could be predicted based on HiDP abundance, and that LoDPs did not contribute significantly to in situ DMSP. We also used the same set of calculations to show that similar results could be obtained using the ratio of DSYB:TpMT2 as HiDP:LoDP. Specifically, the ratio of total DMSP: total chlorophyll-a at timepoint t (Rt) was calculated as: ! " = $ % " ∗ℎ∗(! " + ℎ∗*! " eq. 1 where h is the fractional contribution of HiDPs to total Chla at timepoint t, $ % " is the ratio of LoDPs to HiDPs at timepoint t, LRt is the cellular DMSP:chla ratios of LoDPs at timepoint t, and HRt is the cellular DMSP:chla ratios of LoDPs at timepoint t. Rt and $ % " were measured. Based on a review of previous monoculture measurements (McParland and Levine 2019), HRt was assumed to be 300nmole/µg. LRt was assumed to be 10nmole/µg and assumed to decrease 10-fold to 1 nmole/µg in bottles that were spiked with NO3. Equation 1 was solved for h, and then h was used to solve for the Chla produced by HiDPs, LoDPs and ‘other’ groups, where ‘other’ is non-DMSP producers (or insignificant contributors). Finally, we solve for the concentration of DMSP produced by HiDPs and LoDPs during each experiment in the starting and final communities. +ℎ,- %./0 =ℎ∗1213+ℎ,- eq. 2 +ℎ,- %./0 = $ % ∗ℎ∗1213+ℎ,- eq. 3 ′51ℎ67 8 =1213+ℎ,-−*:;<−(2;< eq. 4 ;=>< %./0 =+ℎ,- %./0 ∗3*! eq. 5 ;=>< $?/0 =+ℎ,- $?/0 ∗3(! eq. 6 TARA eukaryotic metagenome and metatranscriptome search: The same sequences for DSYB and TpMT2 used to search MMETSP transcriptomes were used to search the Marine Atlas of TARA Ocean Unigenes (MATOU) genome and transcriptome using the Ocean Gene Atlas portal (Villar et al. 2018) with BLASTp using a cut-off e-value of 1e -10 . Abundances of DMSP synthesis genes were defined as RPKM (reads per kilobase covered per million of mapped reads). Based on MATOU taxonomy, the percent contribution of gene abundance by four major eukaryotic groups (dinophyceae, haptophyceae, bacillariophyta, and pelagophyceae/chlorophyta) to total DMSP synthesis gene abundance was calculated. Maps of relative contributions of the major eukaryotic 114 groups were made in RStudio with an adaptation of the world map script found here: https://github.com/merenlab/world-map-r. Results We first assess the prevalence of DMSP synthesis genes across major taxonomic group of eukaryotic marine protists and demonstrate that DSYB is most dominant in known HiDP groups, whereas TpMT1 and TpMT2 are found in LoDP groups. We then use the DMSP producer database and synthesis genes to putatively assign DMSP production to OTUs of mixed natural communities. Finally, we use the relative abundance ratios of HiDP:LoDP and DSYB:TpMT2 to predict the abundance of HiDPs and LoDPs as a contribution to total observed Chla to test: 1. if synthesis genes accurately capture the DMSP producing community, and 2. if HiDPs dominate in situ DMSP as hypothesized by McParland and Levine (2019). Prevalence of DMSP synthesis genes: The DMSP producer database (Figure 1a) reflects the current state of knowledge of marine eukaryotes tested for DMSP production in monoculture (n=119). The majority of the database was measured in a seminal paper by Keller et al. (1989), but also includes more recent measurements reviewed or measured by McParland and Levine (2019). The DMSP producer database is composed primarily of Haptophyceae (35%), Dinophyceae (24%), Bacillariophyta (16%), Chlorophyta (15%), Rhizaria (6%) and others (4%, includes Cryptophyceae (1), Eustigmatophyceae (1), Pelagophyceae (2) and Rhodophyta (1)). The majority of Haptophyceae (88%) and dinophycaea (64%) are classified as HiDPs (>50mM DMSP measured). All Rhizaria species, and all but one Bacillariophyta species were LoDPs (<50mM DMSP measured). Chlorophyta species have almost equal numbers of HiDP and LoDP representatives (61% HiDP and 39% LoDP) indicating species specificity for this group’s DMSP phenotype. 10 species of dinophyceae were classified as LoDPs (Figure 1a), with intracellular concentrations between 0.01 – 16 mM (McParland and Levine 2019). While cellular DMSP content is most accurate when normalized to cellular volume, cellular volume is notoriously difficult to measure accurately with cell microscopy, particularly for the unique shape of dinoflagellates (Harrison et al. 2015). These dinophyceae species have very high cellular concentrations of DMSP (45 - 1192 fmole/cell) (Keller et al. 1989), but the reported cellular volumes for these species (4.8x10 5 - 3.4x10 7 µm 3 ) are as much as two orders of magnitude higher 115 than other reported cell volume measurements for the same genus (Menden-Deuer and Lessard 2000; Harrison et al. 2015). We maintain the 50mM cut-off for HiDPs and LoDPs in the DMSP producer database for consistency, but this highlights the importance of accurately quantifying the normalization factor when assigning HiDP and LoDP phenotype. We tested the hypothesis that DSYB is a marker gene for HiDPs by comparing the abundance of the DMSP synthesis genes (DSYB, TpMT1 and TpMT2) in the transcriptomes of the Marine Microbial Eukaryotic Transcriptomes Project (MMETSP) to the designation of HiDP versus LoDP of the same genus in the DMSP producer database. MMETSP is currently the largest reference database of cultured marine protist transcriptomes. 64% of the unique strains (n=393) in the database are representatives of Bacillariophyta, Dinophyceae, Chlorophyta and Haptophyceae, which are also taxa that are well represented in the DMSP producer database (Figure 1). In the MMETSP database, 36% of the total unique strains overlapped with genus in the DMSP producer database (Figure 1b). Of this fraction, 100% of the bacillariophyta species were classified as LoDPs in the DMSP producer database and 88% of the dinophyceae species were classified as HiDPs. Chlorophyta and Haptophyceae contained more intraspecies variability. Of the Chlorophyta and Haptophyceae MMETSP species with representatives in the DMSP producer database, 19% and 67% were HiDPs, respectively. Half of the unique strains in MMETSP (n=393) contained at least one of the DMSP synthesis genes (Figure 1c). The fraction of strains with DMSP synthesis genes varied between the major groups, with 93% of Haptophyceae, 81% of Dinophyceae, 71% of Bacillariophyta and 36% of Chlorophyta transcriptomes containing at least one DMSP synthesis gene. DSYB and TpMT1 were most abundant representing 30% and 32% of total DMSP synthesis gene-containing transcriptomes, respectively, while TpMT2 only accounted for 7%. Many of the DMSP synthesis gene-containing transcriptomes appeared to have more than one synthesis gene (either DSYB and one of the TpMT genes (22%) or both TpMT genes (9%)). The dinophyceae transcriptomes, which are typically considered HiDPs, were dominated by DSYB (70%) or DSYB+TpMT1 or TpMT2 (11%). Bacillariophyta transcriptomes, which are typically considered LoDPs, were dominated by TpMT1 and/or TpMT2 (71%). The presence of different TpMT genes in bacillariophyta transcriptomes was variable: 12% contained TpMT2, 23% contained TpMT1, 12% contained both TpMT2 and TpMT1, and 23% contained a TpMT gene and DSYB (Figure 1c). While there is some overlap of the potential HiDP and LoDP marker 116 genes in these two groups, the overlap was only observed when multiple genes were present in the transcriptomes. For example, in dinophyceae, 13% of transcriptomes have TpMT1 or TpMT2 but all of these strains also contain DSYB. In bacillariophyta, 33% of transcriptomes have DSYB, but also contain either TpMT1 or TpMT2. In these two groups the opposing marker gene never occurred in a transcriptome alone. Haptophyceae are classically considered HiDPs, and the majority of this group’s transcriptomes contained either DSYB or DSYB and a TpMT gene (89%). Three Haptophyceae transcriptomes contained TpMT1 only: Pavlova lutheri RCC 1537 (MMETSP id 1463), Exanthemachrysis gayraliae RCC 1523 (MMETSP id 1464) and Phaeocystis cordata RCC 1383 (MMETSP id 1465). Pavlova and Exanthemachrysis are pavlophytes that were confirmed to have low intracellular DMSP concentrations when measured in monocultures (Keller et al. 1989; McParland and Levine 2019), which is consistent with the potential for TpMT to serve as a LoDP marker gene. Phaeocystis cordata appears to be the exception as it has been shown to produce 358mM intracellular DMSP (Decelle et al. 2012). However, the potential TpMT1 homologues in all three species have low percent identity (<40%) in the transcriptomes and so we hypothesize that they (and particularly Phaeocystis cordata) may contain a third, as of yet to be identified, DMSP production gene (see Discussion). Chlorophyta contain both HiDP and LoDP representatives in the DMSP producer database. In the MMETSP transcriptomes, a majority of the chlorophyta transcriptomes did not contain DMSP synthesis genes (64%), and the ones that did contained either one or both of the TpMT genes, but no DSYB. The remaining synthesis gene-containing transcriptomes were unidentified eukaryotes (5%), or part of the following groups (16%): Cryptophyta, Pelagophyceae, Dictyochophyceae, Chrysophyceae, Euglenida, Opisthokonta, Rhodophyta, Alveolata, Amoebozoa, Rhizaria, Stramenopiles, and Ciliophora. To our knowledge, DMSP production has been confirmed in at least one representative of all these groups, except for Dictyochophyceae, Euglenida, Opisthokonta, Amoebozoa and Ciliophora (McParland and Levine 2019). All of the transcriptomes with only a DSYB representative sequence are known or assumed HiDPs based on the DMSP culture database. However, not all confirmed HiDPs contain a DSYB, for example Phaeocystis cordata (Figure 1). Specifically, DSYB is missing in some potentially important HiDPs: 24% of dinophyceae transcriptomes and 6% of haptophyceae transcriptomes do not contain any synthesis genes (Figure 1b) despite the widespread ability to produce DMSP in these groups (Figure 1a). Some of these species may truly be non-DMSP producers, such as the 117 heterotrophic dinoflagellate Oxyrrhis marina, which has been identified to have no measurable DMSP production (Keller et al. 1989) and did not contain any synthesis genes (Curson et al. 2018). However, the autotrophic dinoflagellate Prorocentrum micans has very high cellular DMSP concentrations (Keller et al. 1989) but did not contain any synthesis genes. This again suggests that there may be an additional HiDP DMSP synthesis gene or genes that have yet to be identified. TpMT1 and TpMT2 were most prevalent in bacillariophyta and chlorophyta suggesting that they could be LoDP marker genes (Figure 1c). The majority of bacillariophyta contained at least one copy of TpMT1 or TpMT2 (71%). TpMT1 or TpMT2 is missing in 9 MMETSP transcriptomes of confirmed LoDPs: 1 species of Pseudo-nitzschia australis, 3 species of Chaetoceros, and 5 strains of Ditylum brightwelli, again suggesting that TpMT is a general LoDP marker gene but others likely exist (Figure 1c). Not enough is known about chlorophyta DMSP production to determine if the 64% of chlorophyta transcriptomes with no synthesis gene sequences are truly non-DMSP producers, are actually HiDPs with a different synthesis gene (not DSYB), or are LoDPs with a different synthesis gene (not TpMT1 or TpMT2). While not perfect marker genes, the DMSP synthesis genes clearly capture the general taxonomic differences of HiDPs and LoDPs (Figure 1). Although TpMT1 appears to be more prevalent, purified recombinant protein TpMT1 does not exhibit methyltransferase activity (the third step of DMSP synthesis) and therefore cannot be assumed to be an active synthesis gene (Kageyama et al. 2018). Purified recombinant protein of TpMT2 does catalyze methyltransferase of MTHB and will therefore be considered the only LoDP marker gene in our remaining analyses. DMSP dynamics in add-back experiments: The change in DMSP concentrations produced by natural communities in response to NO3 add-back was assessed in four different oceanographic regimes. +N refers to bottles that were spiked with NO3, and -N refers to bottles that did not receive NO3. Oligo 1 was most oligotrophic based on Chla concentrations, followed by Oligo 2 << Coast 1 < Coast 2 (Figure 2). Oligo 1 had the lowest starting N:P ratio (N:P=0.5), followed by Oligo 2 and Coast 2 which had similar N:P ratios of 4 and 5, and Coast 1 which had the highest ratio of 9. The addition of NO3 alleviated potential NO3 limitation in the +N bottles at all sites (N:P >> 25 after add-back). Chla in the +N treatments significantly increased over the course of the experiment at all sites (Figure 2, t-test, p< 0.01), but not necessarily due to NO3 drawdown. Changes in Chla and 118 NO3 concentrations suggest that only Oligo 1 and Coast 1 responded to the N addition, this was most significant at Oligo 1 where Chla increased significantly after only 12 hours (t-test, p<0.05). By 36 hours, 84 ± 0.02% of NO3 was drawn-down at Oligo 1 and 14 ± 0.02% of NO3 was drawn- down at Coast 1. Starting concentrations of SiOH4 were also drawn-down 80% at Oligo 1 indicating potential diatom growth (data not shown). At Oligo 2 and Coast 2, NO3 was likely not the primary limiting nutrient as final NO3 concentrations were not significantly drawn-down at either site (t-test, p>0.1). At Coast 2, Chla increased in both +N and -N, where +N was indistinguishable from -N over the first 12 hours indicating that the increase was due to a bottle effect (potentially iron contamination relieving iron limitation). +N continued to grow in the Coast 2 experiment between 24 and 48 hours while -N crashed to starting concentrations. We attribute the step-change in Chla at Oligo 2 between 6 and 9 hours to a sharp increase in the temperature of the deckboard incubators due to a front crossing during this time. We believe, this site was iron limited (Ribalet et al. 2010) and so had no response to the NO3 addition. Total DMSP was correlated with the NO 3 concentrations in starting water for each site (linear regression, R 2 =0.9, p=0.06), with 10 ± 0.01 nM DMSP at Oligo 1, followed by Oligo 2 (17 ± 0.3 nM) < Coast 1 (80 ± 1 nM) < Coast 2 (90 ± 0.5 nM). Total DMSP increased in +N relative to both starting concentrations and final -N concentrations at Coast 1, Oligo 2, and Coast 2 (t-test for Oligo 2 and Coast 2: p < 0.01, t-test for Coast 1: p=0.05) (Figure 3). In contrast, at Oligo 1, total DMSP decreased in +N after just 12 hours (t-test, p=0.06) and remained low throughout the remainder of the experiment. DMSP: Chla at Oligo 1 and Coast 1 significantly decreased after N addition (t-test, p<0.05) (Figure 4) due to changes in both total DMSP and Chla. At Oligo 2 and Coast 2, DMSP: Chla did not significantly change after N addition (t-test, p>0.06) as both total DMSP and Chla increased simultaneously. The significant increase in DMSP: Chla in -N at Oligo 2 and Coast 2 was primarily driven by the decrease in Chla, though total DMSP in -N did significantly decrease as well at Coast 2 (t-test, p<0.05). Community analysis: The V4 region of 18S rRNA genes was amplified and sequenced in the initial and final community of each experiment. After QC (see Methods), each sample had an average 74,200 ± 30,830 sequences (min = 41,437, max = 162,472). While there was little overlap of the 4083 OTUs across sites as might be expected, each sample had a consistent number of 1050 ± 209 OTUs (min = 750, max = 1624). For the samples with biological duplicates (Oligo 1), an average 119 49 ± 11% of OTUs were found in both bottles. For the samples with biological triplicates (Coast 1, Oligo 2, Coast 2), an average 40 ± 2% of OTUs were found in all three bottles and an average 22 ± 1% of OTUs were found in two of the three bottles. DMSP production assignment: At Oligo 1 and Coast 1, alveolata (primarily Dinophyceae and Syndiniales) dominated more than more than 50% of the total community, and at Oligo 2 and Coast 2, alveolatea and chlorophyta (primarily Micromonas) combined comprised more than 50% of the total community (Figure 5). Bacilliarophyta were most abundant at Oligo 1 (19% of starting community). The eukaryotic community assessed with 18S will include not only DMSP producers, but also non-DMSP producers, such as heterotrophic ciliates and MAST that have not been tested for DMSP production. Therefore, we assigned we used the sequences of known DMSP producers (both from monoculture and synthesis gene presence) to quantify the relative abundance of HiDPs and LoDPs in each experiment. Each OTU from the add-back experiments was assigned a putative HiDP or LoDP phenotype using two different methods: 1) by BLASTing against 18S sequences of known DMSP producers from the DMSP producer database and 2) by phylogenetic similarity with 18S sequences of species that potentially contain one or more of the DMSP synthesis genes. Potential DMSP producers identified in this way comprised a maximum 54% of the eukaryotic community at all sites (Figure 5). This is a conservative estimate as the DMSP producer database is limited to only 119 species and, as discussed above, we have reason to believe that there is a yet-to-be identified DMSP production gene based on our analysis of n=393 eukaryotic transcriptomes. However, this analysis provides the first attempt to link DMSP phenotype (HiDP and LoDP) and DMSP synthesis genes to observed changes in DMSP concentrations. Using the DMSP producer database assignment, 1554 of 4083 OTUs were assigned as either a LoDP (26%) or HiDP (74%). HiDPs belonged to the Alveolata (82%), Prymnesiophyceae (8%), Chlorophyta (7%), and Others (Chrysophyceae, Haptophyta, Pavlovophyceae and Unassigned) (3%). LoDPs were Stramenopiles (primarily Bacillariophyta) (56%), Chlorophyta (25%), Alveolata (12%), Rhizaria (4%) and Others (Haptophyta, Pavlovophyceae and Prymnesiophyceae) (3%). Using the gene-based assignment, 900 of 4083 OTUs were assigned to putatively contain DMSP synthesis genes, 96% of which were assigned to be DSYB-carriers. This is consistent with the observation that DSYB is more prevalent than TpMT2 (Figure 1c). Of the 120 864 OTUs predicted to contain DSYB, the majority were Alveolata (42%) as expected, but also included Rhizaria (21%), Chlorophyta (13%) and the remainder were Hacrobia, Stramenopiles (primarily Chrysophyceae and Ciliophora) and Opisthokonta (24%). The 74 OTUs predicted to contain TpMT2 were Bacillariophyta (38%), Ciliophora (38%), Cryptophyta (12%), and Hacrobia (12%). The Chlorophyta OTUs, particularly Micromonas pusilla, are important DMSP producers at Oligo 2 and Coast 2. The majority of MMETSP Chlorophyta transcriptomes contained no synthesis gene or TpMT1 or TpMT2, but 55% of Chlorophyta OTUs in the add-back experiments (n=206) contained DSYB. This is consistent with the DMSP producer database, where Micromonas pusilla contains both high and low DMSP producing representatives (i.e. intraspecies variation) and indicates that pplacer found these OTUs to be more similarly related to DSYB- carriers than the Micromonas TpMT1 or TpMT2-carriers in the MMETSP phylogeny (Figure 1c). Overall the percentage of OTUs assigned to HiDP versus LoDP and DSYB versus TpMT2 were consistent at the group level. However, there was very little overlap between the OTU’s placed on the MMETSP phylogeny and those related to the DMSP producer database. Specifically, only 15% of the OTUs identified as a DMSP producer (either HiDP or LoDP) using the DMSP producer database were also phylogenetically similar enough to a synthesis gene-carrying strain in the MMETSP phylogeny to be assigned a DMSP synthesis gene (Figure 5). We attribute this to 3 factors: 1) taxonomic differences of species included within the two databases (Figure 1b vs. Figure 1c), 2) unidentified DMSP synthesis genes, and 3) the stronger statistical power required by pplacer to assign DMSP synthesis genes. Similarly, of the 900 OTUs identified as containing either DSYB or TpMT2, only 41% also had cultured representatives in the DMSP producer database (Figure 1a). Despite these discrepancies, both methods captured the same major HiDP groups and both predicted that HiDPs dominate the DMSP producer community relative to LoDPs at all sites. Across all sites the relative abundance of HiDPs to LoDPs and DSYB to TpMT2 was significantly linearly related (linear regression, R 2 =0.6, p<0.01) (Figure 6a). This suggests that, while TpMT2 underestimates LoDPs, both methods captured the same relative changes in the DMSP producer community. Additionally, the total number of OTUs assigned to be HiDPs or DSYB-carriers were also significantly linearly related across three of the four sites (linear regression, R 2 =0.4, p<0.01) (Figure 6b). While Oligo 1 showed the same relationship between HiDP:LoDP and DSYB:TpMT2 as the other sites, only about half as many OTUs were assigned 121 to be DSYB-carriers as were assigned to be HiDP. (Figure 6b). Oligo 1 had a significantly higher OTU richness than the other sites (t-test, p<0.01), suggesting that the MMETSP phylogeny did not provide enough coverage for assigning DSYB to the diverse community. Predicting DMSP with measured community composition: Previous work suggested that in situ DMSP concentrations are determined by the abundance of HiDPs (McParland and Levine 2019). Using the estimated relative abundances of high versus low producers (HiDP:LoDP) (Supp Table 2) assigned using the DMSP producer database, we estimated the absolute abundance of HiDPs over the course of the add-back experiments. We assumed that the rest of the community was made up of LoDPs and ‘other’, which represents non-DMSP producers and non-eukaryotic phytoplankton (i.e. cyanobacteria). We then repeated the analysis using the relative abundance of DSYB versus TpMT2 carriers from the synthesis gene assignments as a proxy for the high:low producer ratio (Supp Table 2). The ratio of HiDP:LoDP and DSYB:TpMT2 were significantly related (Figure 6), indicating that the gene-based assignment method should produce similar results as the DMSP producer database assignment. As this is a compositional dataset, any changes in non-DMSP producers will change total relative abundance of HiDPs and LoDPs to maintain a sum of 1 (Weiss et al. 2017). Therefore, to reduce this bias, we used a ratio of relative abundances to most accurately capture changes in the DMSP producer community. Both the predicted abundance of HiDPs (and to a lesser extent LoDPs) and the predicted change over the course of the experiment were similar using both methodologies for determining the DMSP producer community composition (Figure 4, Table 1). In all starting communities, the ‘other’ group was predicted to contribute more than 50% of total Chla, with the greatest contribution at the most oligotrophic site, Oligo 1 (Figure 4). This was expected as 18S taxonomy does not capture important cyanobacteria primary producers which are known to dominate in-situ Chla (e.g. Flombaum et al. 2013). At Oligo 1 and Coast 1, DMSP: Chla in -N did not significantly change and the community was predicted to remain relatively constant. In +N, a significant decrease in the relative abundance of HiDPs over the course of the experiment was necessary to produce the observed decrease in DMSP: Chla (Figure 4a and 4b). However, total HiDP Chla was not predicted to decrease, but rather total 'other' Chla increased significantly making up the majority of the increase in total Chla. In contrast, at Oligo 2 and Coast 2, the relative abundance of HiDPs in -N was predicted to significantly increase as total Chla 122 declined in order to maintain high DMSP concentrations (Figure 4c and 4d). Total HiDP Chla was predicted to decrease slightly, but the ‘other’ Chla was predicted to significantly crash resulting in the shift in relative abundance towards HiDPs. Community composition in +N at Oligo 2 and Coast 2, was not predicted to significantly change. At all of these sites, HiDPs were predicted to produce " 96% of total DMSP (Table 1). If we assumed that the ‘other’ group was in fact made up of LoDPs, this only changes slightly to " 92% of total DMSP attributed to HiDPs. These calculations assumed that LoDPs significantly decreased (10-fold change) cellular DMSP in response to N add-back (McParland and Levine 2019; McParland et al in prep). Even with significant upregulation under the nitrogen stressed condition (-N), LoDPs were predicted to only contribute 0.1 to 0.6 nM DMSP (1 to 6 %), which is close to measurement errors that have been reported for in situ DMSP studies (Kiene et al. 2007; Franklin et al. 2009; Bell et al. 2010). The one exception could be at Oligo 1 where total DMSP is low and LoDPs comprised 50 ± 5% of DMSP producers based on the DMSP producer database. If LoDPs had significantly higher cellular DMSP concentrations (5-fold higher) at Oligo 1, they could contribute 14% - 32% of total DMSP (1.2 ± 0.1 nM or 2.9 ± 0.3 nM) (Supp Figure 1c and 1d, Supp Table 3). The estimated abundance of HiDPs is extremely sensitive to the assumed HiDP intracellular concentration of DMSP (cellular DMSP: Chla) – consistent with our hypothesis that HiDPs are the primary determinants of in situ DMSP. Lowering HiDP intracellular DMSP concentrations in the calculations results in a corresponding increase in HiDP Chla and decrease in ‘other’ Chla (Supp Figure 1 & Supp Table 3). A significant decrease in HiDP intracellular DMSP concentrations though still requires a bloom of the ‘others’ to produce the decreased DMSP: Chla ratio measured in +N. While individual HiDP species are not expected to shift cellular DMSP in response to N add-back (McParland and Levine 2019; McParland et. al. in prep), a shift in the HiDP community could result in an apparent significant change in HiDP intracellular DMSP concentrations. For example, if dinophyceae species which on average have a higher intracellular DMSP concentration (mean 469 ± 501 mM based on the DMSP producer database) were replaced with haptophyceae species (223 ± 178 mM) this would result in an apparent decrease in HiDP intracellular DMSP concentrations in response to the add-back even though there was not a physiological response of the HiDPs to the NO3 addition. For the add-back experiments, this would change the total predicted 123 HiDP Chla over the course of the experiment but does not change the conclusions. For example, at Oligo 1, even with a change in HiDP intracellular DMSP (either increase or decrease), most of the increase in Chla was still predicted to be in the ‘other’ group as this was necessary to produce the observed decrease in DMSP (Supp Figure 2a and 2b). Similarly, at Coast 2, a significant increase or decrease in HiDP intracellular DMSP would still result in a significant crash of the ‘other’ group while the HiDPs dominate total-chla (Supp Figure 2c and 2d). In fact, such a low HiDP cellular DMSP at Coast 2 (Supp Figure 2d) would require more HiDP Chla than was measured for total Chla to produce the observed total DMSP, suggesting that HiDP cellular DMSP in the initial community at Coast 2 must have been greater than 100 nmole/µg. Global distribution of DMSP synthesis genes: Both DSYB and TpMT2 were globally abundant in the eukaryotic metagenomes and metatranscriptomes of the Marine Atlas of Tara Ocean Unigenes (Figure 7). Dinoflagellates dominated the majority of synthesis gene expression, except in the Southern Ocean where haptophytes dominated. The differentiation of synthesis genes by group supports our hypothesis that DSYB is a marker gene for HiDPs, and vice versa for TpMT2. Dinoflagellata and haptophyte only contained DSYB. Bacillariophyta on average contain >95% TpMT2, but did also contain DSYB which was expected based on MMETSP transcriptomes (Figure 1c). At certain sites, DSYB was more highly expressed than TpMT2 in the Bacillariophyta. Pelagophyceae and chlorophyta genomes are dominated by TpMT2, where chlorophyta only contained TpMT2 and a small number of pelagophyceae contained DSYB. Both groups almost exclusively express TpMT2. Discussion and Conclusion In situ DMSP production was previously hypothesized to be dominated by HiDPs, with minimal contributions by LoDPs despite a significant upregulation of cellular DMSP under nutrient stress (McParland and Levine 2019). Here, we confirmed this hypothesis by quantifying the DMSP producer community with marker genes of HiDP and LoDP phenotypes for the first time in situ while simultaneously measuring the community’s response to nitrogen add-back. Even when LoDPs comprised a significant portion of the DMSP community (Oligo 1) and upregulated cellular DMSP in response to N stress, they contributed <1nM DMSP. 124 We confirm that community composition determines in situ DMSP concentrations as a function of both HiDP intracellular DMSP concentrations and HiDP biomass. At Oligo 1, where DMSP: Chla significantly decreased, HiDP biomass remained low, while the ‘other’ group dominated the observed increase in Chla. Intracellular DMSP of individual HiDP species do not respond significantly to N stress (McParland and Levine 2019, McParland et al. in prep). However, due to the large variability in cellular DMSP concentrations among HiDP species and groups (Keller 1989; Keller et al. 1989; Caruana and Malin 2014; McParland and Levine 2019), a shift in the community composition of HiDPs would change the average HiDP intracellular DMSP concentration. Therefore, in situ DMSP changes will be correlated to environmental conditions at both large (e.g. different oceanographic regions) and small (e.g. during the add-back experiment) scales not due to physiological responses of DMSP producers to the environment but due to environmental controls on community composition. Finally, we demonstrate for the first time that DSYB and TpMT2 synthesis genes can be used to differentiate HiDPs and LoDPs in both monocultures and in situ (Figure 1 and 7). DSYB dominated the major HiDP groups of Dinophyceae and Haptophyceae, though some HiDP species were missing a synthesis gene (Figure 1). The DSYB phylogeny (Curson et al. 2018) was originally reported by searching only 119 MMETSP eukaryotic transcriptomes. Here, we included all MMETSP transcriptomes (n=678) which provides a more complete quantification of the diversity of DSYB among major taxon groups. While TpMT2 was much less abundant than DSYB, this gene was shown to be a LoDP marker as it was found in the specific species reported by Curson et al. (2018) to be missing DSYB, including Phaeodactylum tricornutum, Thalassiosira pseudonana, Thalassiosira oceanica and many members of the Chlorophyta group. Significant TpMT2 homology was only reported in three species previously (Kageyama et al. 2018), but here, we find TpMT2 present in at least 13 MMETSP transcriptomes and to be globally abundant. While there was overlap of the two genes in HiDPs and LoDPs, in both the monoculture transcriptomes and in situ, if an opposing marker gene was present, the respective marker gene was also present. As DSYB and TpMT2 are isoforms that both perform the third step of the DMSP synthesis pathway, known DMSP producing species missing a synthesis gene (Figure 1) likely contain another, as of yet to be identified, isoform (Curson et al. 2018). Differences in the assignment of OTUs in the add-back experiments using the DMSP producer database and the gene-based database were largely driven by the differences in taxonomy represented by each database (Figure 125 1b and 1c). Though the culture-based assignment appears to capture more HiDPs, this is biased towards knowledge of cultured phototrophic eukaryotes. DSYB was present in a number of unknown potential HiDPs (Figure 1b) that have not been described in culture as DMSP producers. In particular, Ciliophora are unexpected HiDPs as DMSP synthesis is typically attributed to phototrophs. However, some heterotrophs have previously been reported to produce DMSP (Raina et al. 2013; Curson et al. 2018), and also, marine ciliophora are known to host potential HiDP symbionts (Dziallas et al. 2012; Mordret et al. 2016). While DSYB is not a perfect marker gene, future work towards developing a universal primer for DSYB (and potentially other important HiDP synthesis gene(s)) would result in better prediction of in situ DMSP. A gene-based approach would best capture the entire HiDP community which inevitably includes uncultured eukaryotes that would significantly contribute to in situ DMSP as HiDPs. qPCR methods would allow for quantitative measurements of the HiDP community, rather than just relative abundances. Additionally, transcription rates of DSYB appear to be correlated with intracellular DMSP concentrations (Curson et al. 2018). More work is needed to validate this correlation across HiDPs, but this relationship would eliminate the need to know cellular DMSP concentrations of HiDPs in situ when predicting total DMSP concentrations. Despite underestimation of LoDPs with TpMT2, in situ DMSP production dynamics were accurately predicted when using either the DMSP producer database or gene-based assignments of LoDPs. This is because the constraint on in situ DMSP concentrations was the biomass and intracellular concentrations of HiDPs, not those of LoDPs. This suggests that correctly identifying LoDP synthesis gene(s) is not required for predicting in situ DMSP. A ratio of the relative abundances of HiDPs versus Bacillariophyta would likely have predicted the observed DMSP production dynamics just as accurately. However, LoDP synthesis genes are still important for understanding the cellular mechanism of DMSP in these producers. LoDPs were only recently recognized for their potentially significant DMSP production and monoculture studies have only investigated the physiological role of DMSP in these producers with diatom species and one cyanobacteria (Bucciarelli et al. 2013; McParland and Levine 2019). Here, we show that other important primary producers identified as LoDPs contain TpMT2, particularly the Chlorophyta (Keller 1989; McParland and Levine 2019). The LoDP mechanism likely plays an important role at the cellular level as a signaling molecule and may be involved in microbial interactions 126 (Seymour et al. 2010; Johnson et al. 2016). Further development of the LoDP synthesis genes will allow for probing of the potential signaling role of DMSP at a cellular-level. The recently discovered DMSP synthesis genes, DSYB and TpMT, are found in a diverse phylogeny of primarily marine phototrophs and are significantly represented throughout the global ocean. DMSP synthesis genes in both natural community metagenomes and metatranscriptomes and monoculture transcriptomes are differentiated across known HiDP and LoDP functional groups confirming our hypothesis that the synthesis genes can serve as marker genes for HiDPs and LoDPs. The differential changes in DMSP production by HiDPs and LoDPs in monoculture in response to environmental stressors (McParland et al. in prep) suggest DMSP serves different roles in HiDPs and LoDPs and these different mechanisms are likely regulated by the two genes analyzed here. HiDP biomass controls in situ DMSP production and therefore prediction of DMSP is dependent on knowledge of HiDP species abundance and intracellular DMSP concentrations. Assigning DMSP production through monoculture screening can be significantly influenced by normalization values. For example, the dinoflagellates classified as LoDPs in the DMSP producer database, but contain DSYB, have extremely high cellular concentrations of DMSP but, based on cell volume measurements, have <50 mM DMSP. Additionally, while high intracellular DMSP concentrations are typically associated with DSYB, it may be that the constitutive regulation of DMSP by HiDPs is actually what defines this group. Chrysochromulina sp., a haptophyte, contains DSYB but would be considered a LoDP based on cellular concentrations. However, Chrysochromulina does not alter transcription in response to nitrogen limitation (Curson et al. 2018), suggesting that it exhibits a constitutive regulation of cellular DMSP despite low cellular concentrations. Finally, a number of uncultured eukaryotes contained DSYB (Figure 1). Small changes in HiDP abundance (i.e. less than 1 cell/µL) will significantly alter in situ DMSP concentrations and therefore quantification of all HiDPs is critical for prediction of in situ DMSP (McParland and Levine 2019). A universal assignment of the HiDP phenotype with a marker gene(s) would result in the most accurate classification of all HiDPs and thus prediction of in situ DMSP. 127 Acknowledgements We thank Chief Scientists Ron Kiene, Phil Tortelle, Zoran Ristovski, and Graham Jones. We thank Elaina Graham for discussions about methodology. Finally we thank the crew of the R/V Oceanus, R/V Investigator and the Miss Christy. This project was supported by funding from the Rose Hills Foundation, the Australian Research Council, a Gerald Bakus Graduate Research Fellowship and a NDSEG Research Fellowship. 128 Figures and Tables Figure 1: (a) The DMSP producer database of DMSP production previously confirmed in monocultures. (b) Known DMSP producers confirmed previously with monoculture production in the MMETSP database. (c) Presence and absence of DMSP synthesis genes, DSYB, TpMT1, and TpMT2 in the MMETSP database. Many transcriptomes had multiple gene copies (see Supp Table 1 for more details). n is the total sum of the circle. 129 Figure 2: Chla concentrations (left) and NO3 concentrations (right) during N add-back experiments at Oligo 1 (a), Coast 1 (b), Oligo 2 (c) and Coast 2 (d). +N is solid line, - N is dashed line. The double y axis on NO3 plots represents +N (left) and -N (right). 130 Figure 3: Total DMSP concentrations during N add-back experiments at Oligo 1 (a), Coast 1 (b), Oligo 2 (c) and Coast 2 (d). +N is solid line, - N is dashed line. 131 Figure 4: DMSP: Chla during N add-back experiments at Oligo 1 (a), Coast 1 (b), Oligo 2 (c) and Coast 2 (d). +N is solid line, - N is dashed line. Pies represent predicted contribution of HiDPs, LoDPs and ‘other’s to total Chla. Top pie was calculated using HiDP:LoDP ratio from the DMSP producer database assignment, and bottom pie was calculated using HiDP:LoDP ratio from the gene-based assignment. 132 Figure 5: The relative abundance of major eukaryotic groups in the starting community (left) at Oligo 1 (a), Coast 1 (b), Oligo 2 (c) and Coast 2 (d). The relative abundance of DMSP producers 133 in the starting community (right) as determined by the DMSP producer database and gene-based assignment. Figure 6: (a) The ratio of HiDP:LoDP as assigned with the DMSP producer database versus the ratio of DSYB:TpMT2 as assigned with the synthesis genes in OTUs across all sites. (b) The total number of OTUs assigned to be HiDPs versus DSYB-carriers normalized to the total number of OTUs at each site. The linear regression is significant for both relationships (see main text). 134 Figure 7: Global distribution of DMSP synthesis genes in TARA eukaryotic metagenomes (left) and metatranscriptomes (right) in major eukaryotic classes (Dinophyceae, Haptophyceae, 135 Bacillariophyta, and Pelagophyceae+Chlorophyta). Size of circle at each station represents the contribution of each group to the total abundance of DMSP synthesis genes. The color of each circle represents the % of DSYB or TpMT present. 136 Table 1: The total Chla produced by HiDPs and LoDPs, and the sum of total Chla produced by DMSP producers at Oligo 1, Coast 1, Oligo 2 and Coast 2, in the starting community and final communities (-N and +N) as predicted with the HiDP:LoDP ratio (top) and DSYB:TpMT2 (bottom). 137 Supplemental Figures and Tables Supp Figure 1: A sensitivity test demonstrating the importance of cellular DMSP concentrations in determining contribution of HiDPs to total Chla (and total DMSP). HiDP and LoDP cellular DMSP concentrations were held constant at the given value across the experiment at Oligo 1 and ‘others’ were solved for using the observed changes in DMSP, Chla and ratio of highs:lows using the culture-based assignment. 138 Supp Figure 2: A sensitivity test demonstrating the impact of a community shift in the HiDPs between starting and final communities on resulting HiDP contribution to total Chla. HiDP cellular concentrations either decreased (left panels) or increased (right panels) and the resulting impacts are shown for Oligo 1 in +N (top panels) and for Coast 2 in -N (bottom panels). 139 Supp Table 1: Differences in experimental design across different sites. 140 Supp Table 2: The ratios of LoDP to HiDPs using the DMSP producer database and synthesis gene assignment. 141 Supp Table 3: Calculated concentrations of Chla and DMSP produced by HiDPs and LoDPs in the sensitivity tests (Supp Figure 1 and 2). !! 142 References Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215: 403–410. doi:https://doi.org/10.1016/S0022-2836(05)80360- 2 Archer, S. D., S. a. Kimmance, J. a. Stephens, F. E. Hopkins, R. G. J. Bellerby, K. G. Schulz, J. Piontek, and a. Engel. 2013. Contrasting responses of DMS and DMSP to ocean acidification in Arctic waters. Biogeosciences 10: 1893–1908. doi:10.5194/bg-10-1893- 2013 Arnold, H. E., P. Kerrison, and M. Steinke. 2013. Interacting effects of ocean acidification and warming on growth and DMS-production in the haptophyte coccolithophore Emiliania huxleyi. Glob. Chang. Biol. 19: 1007–16. doi:10.1111/gcb.12105 Bell, T. G., A. J. Poulton, and G. Malin. 2010. Strong linkages between dimethylsulphoniopropionate (DMSP) and phytoplankton community physiology in a large subtropical and tropical Atlantic Ocean data set. Global Biogeochem. Cycles 24: n/a-n/a. doi:10.1029/2009GB003617 Bucciarelli, E., C. Ridame, W. G. Sunda, C. Dimier-Hugueney, M. Cheize, and S. Belviso. 2013. Increased intracellular concentrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum. Limnol. Oceanogr. 58: 1667–1679. doi:10.4319/lo.2013.58.5.1667 Caporaso, J. G., J. Kuczynski, J. Stombaugh, and others. 2010. QIIME allows analysis of high- throughput community sequencing data. Nat. Methods 7: 335. Carradec, Q., E. Pelletier, C. Da Silva, and others. 2018. A global ocean atlas of eukaryotic genes. Nat. Commun. 9: 373. doi:10.1038/s41467-017-02342-1 Caruana, A. M. N., and G. Malin. 2014. 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 Charlson, R., J. Lovelock, M. Andrae, and S. Warren. 1987. Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Nature 326: 655–661. Curson, A. R. J., J. Liu, A. Bermejo Martínez, and others. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nat. Microbiol. 2: 1–9. doi:10.1038/nmicrobiol.2017.9 143 Curson, A., B. Williams, B. Pinchbeck, and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat. Microbiol. 4: 430– 439. doi:10.1038/s41564-018-0119-5 Decelle, J., I. Probert, L. Bittner, and others. 2012. An original mode of symbiosis in open ocean plankton. Proc Natl Acad Sci 109. doi:10.1073/pnas.1212303109/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1212303109 Dutkiewicz, S., A. E. Hickman, O. Jahn, S. Henson, C. Beaulieu, and E. Monier. 2019. Ocean colour signature of climate change. Nat. Commun. 10. doi:10.1038/s41467-019-08457-x Dutkiewicz, S., J. R. Scott, and M. J. Follows. 2013. Winners and losers: Ecological and biogeochemical changes in a warming ocean. Global Biogeochem. Cycles 27: 463–477. doi:10.1002/gbc.20042 Dziallas, C., M. Allgaier, M. T. Monaghan, and H. P. Grossart. 2012. Act together-implications of symbioses in aquatic ciliates. Front. Microbiol. 3: 1–17. doi:10.3389/fmicb.2012.00288 Edgar, R. C. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. doi:10.1093/bioinformatics/btq461 Flombaum, P., J. L. Gallegos, R. a Gordillo, and others. 2013. Present and future global distributions of the marine Cyanobacteria Prochlrococcus and Synechococcus. Pnas 110: 9824–9829. doi:10.1073/pnas.1307701110/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1307701110 Franklin, D. J., A. J. Poulton, M. Steinke, J. Young, I. Peeken, and G. Malin. 2009. 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 Gage, D., D. Rhodes, K. Nolte, W. A. Hicks, T. Leustek, A. J. L. Cooper, and A. Hanson. 1997. A new route for synthesis of in marine algae. Nat. Lett. 387: 891–894. Galí, M., R. Simó, G. L. Pérez, C. Ruiz-González, H. Sarmento, S. J. Royer, A. Fuentes-Lema, and J. M. Gasol. 2013. Differential response of planktonic primary, bacterial, and dimethylsulfide production rates to static vs. dynamic light exposure in upper mixed-layer summer sea waters. Biogeosciences 10: 7983–7998. doi:10.5194/bg-10-7983-2013 Guillou, L., D. Bachar, S. Audic, and others. 2013. The Protist Ribosomal Reference database (PR 2 ): A catalog of unicellular eukaryote Small Sub-Unit rRNA sequences with curated 144 taxonomy. Nucleic Acids Res. 41: 597–604. doi:10.1093/nar/gks1160 Gunson, J. R., S. A. Spall, T. R. Anderson, A. Jones, I. J. Totterdell, and M. J. Woodage. 2006. Climate sensitivity to ocean dimethylsulphide emissions. Geophys. Res. Lett. 33: 2–5. doi:10.1029/2005GL024982 Harrison, P. J., A. Zingone, M. J. Mickelson, and others. 2015. Cell volumes of marine phytoplankton from globally distributed coastal data sets. Estuar. Coast. Shelf Sci. 162: 130–142. doi:10.1016/j.ecss.2015.05.026 Hopkins, F. E., and S. D. Archer. 2014. Consistent increase in dimethyl sulfide (DMS) in response to high CO2 in five shipboard bioassays from contrasting NW European waters. Biogeosciences 11: 4925–4940. doi:10.5194/bg-11-4925-2014 Hu, S. K., P. E. Connell, L. Y. Mesrop, and D. A. Caron. 2018a. A Hard Day’s Night: Diel Shifts in Microbial Eukaryotic Activity in the North Pacific Subtropical Gyre. Front. Mar. Sci. 5: 1–17. doi:10.3389/fmars.2018.00351 Hu, S. K., Z. Liu, H. Alexander, V. Campbell, P. E. Connell, S. T. Dyhrman, K. B. Heidelberg, and D. A. Caron. 2018b. Shifting metabolic priorities among key protistan taxa within and below the euphotic zone. Environ. Microbiol. 20: 2865–2879. doi:10.1111/1462- 2920.14259 Johnson, L. K., H. Alexander, and C. T. Brown. 2018. Re-assembly, quality evaluation, and annotation of 678 microbial eukaryotic reference transcriptomes. Gigascience 1–12. doi:10.1093/gigascience/giy158 Johnson, W. M., M. C. K. Soule, and E. B. Kujawinski. 2016. Evidence for quorum sensing and differential metabolite production by the marine heterotroph, Ruegeria pomeroyi, in response to DMSP. ISME J. 10: 2304–2316. doi:10.1038/ismej.2016.6 Kageyama, H., Y. Tanaka, A. Shibata, R. Waditee-Sirisattha, and T. Takabe. 2018. Dimethylsulfoniopropionate biosynthesis in a diatom Thalassiosira pseudonana: Identification of a gene encoding MTHB-methyltransferase. Arch. Biochem. Biophys. 645: 100–106. doi:10.1016/j.abb.2018.03.019 Kalvari, I., A. Bateman, J. Argasinska, and others. 2018. Rfam 13.0: shifting to a genome-centric resource for non-coding RNA families. Nucleic Acids Res. 46: D335–D342. doi:10.1093/nar/gkx1038 Keeling, P. J., F. Burki, H. M. Wilcox, and others. 2014. The Marine Microbial Eukaryote 145 Transcriptome Sequencing Project (MMETSP): Illuminating the Functional Diversity of Eukaryotic Life in the Oceans through Transcriptome Sequencing. PLoS Biol. 12: e1001889. doi:10.1371/journal.pbio.1001889 Keller, M., W. Bellows, and R. Guillard. 1989. Dimethyl sulfide production in marine phytoplankton, p. 167–182. In Biogenic sulfur in the environment. Keller, M. D. 1989. 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 Kettle, A. J., M. O. Andreae, D. Amouroux, and others. 1999. 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. Cycles 13: 399–444. Kiene, R. P., D. J. Kieber, D. Slezak, D. a. Toole, D. a. Valle, J. Bisgrove, J. Brinkley, and A. Rellinger. 2007. Distribution and cycling of dimethylsulfide, dimethylsulfoniopropionate, and dimethylsulfoxide during spring and early summer in the Southern Ocean south of New Zealand. Aquat. Sci. 69: 305–319. doi:10.1007/s00027-007-0892-3 Kiene, R. P., L. J. Linn, and J. A. Bruton. 2000. New and important roles for DMSP in marine microbial communities. J. Sea Res. 43: 209–224. Kiene, R. P., and S. K. Service. 1991. 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 Lana, a., T. G. Bell, R. Simó, and others. 2011. An updated climatology of surface dimethlysulfide concentrations and emission fluxes in the global ocean. Global Biogeochem. Cycles 25: n/a-n/a. doi:10.1029/2010GB003850 Lana, A., R. Simó, S. M. Vallina, and J. Dachs. 2012. Re-examination of global emerging patterns of ocean DMS concentration. Biogeochemistry 110: 173–182. doi:10.1007/s10533- 011-9677-9 Lee, M. D., N. G. Walworth, J. B. Sylvan, K. J. Edwards, and B. N. Orcutt. 2015. Microbial communities on seafloor basalts at Dorado Outcrop reflect level of alteration and highlight global lithic clades. Front. Microbiol. 6: 1–20. doi:10.3389/fmicb.2015.01470 Lee, P. a., S. F. Riseman, C. E. Hare, D. a. Hutchins, K. Leblanc, and G. R. DiTullio. 2011. Potential impact of increased temperature and CO2 on particulate 146 dimethylsulfoniopropionate in the Southeastern Bering Sea. Adv. Oceanogr. Limnol. 2: 33– 47. doi:10.1080/19475721.2011.574433 Levine, N. M., D. A. Toole, A. Neeley, N. R. Bates, S. C. Doney, and J. W. H. Dacey. 2015. Revising upper-ocean sulfur dynamics near Bermuda: New lessons from 3 years of concentration and rate measurements. Environ. Chem. 13: 302–313. doi:10.1071/EN15045 Matsen, F., R. Kodner, and E. Armbrust. 2010. pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree. BMC Bioinformatics 11: 538. McParland, E. L., and N. M. Levine. 2019. The role of differential DMSP production and community composition in predicting variability of global surface DMSP concentrations. Limnol. Oceanogr. 64: 757–773. doi:10.1002/lno.11076 Menden-Deuer, S., and E. J. Lessard. 2000. Carbon to volume relationships for dinoflagellates, diatoms and other protist plankton. Limnol. Oceanogr. 45: 569–579. Mordret, S., S. Romac, N. Henry, and others. 2016. The symbiotic life of Symbiodinium in the open ocean within a new species of calcifying ciliate (Tiarina sp.). ISME J. 10: 1424–1436. doi:10.1038/ismej.2015.211 Morris, R. M., M. S. Rappe, K. L. Vergin, W. A. Siebold, C. A. Carlson, and S. J. Giovannoni. 2002. SAR11 clade dominates ocean surface bacterioplankton communities. Nature 420: 806–810. doi:10.1038/nature01281.1. Nawrocki, E. P., and S. R. Eddy. 2013. Infernal 1.1: 100-fold faster RNA homology searches. Bioinformatics 29: 2933–2935. doi:10.1093/bioinformatics/btt509 Quinn, P. K., and T. S. Bates. 2011. The case against climate regulation via oceanic phytoplankton sulphur emissions. Nature 480: 51–56. doi:10.1038/nature10580 Raina, J.-B., D. M. Tapiolas, S. Forêt, and others. 2013. DMSP biosynthesis by an animal and its role in coral thermal stress response. Nature. doi:10.1038/nature12677 Reisch, C. R., M. A. Moran, and W. B. Whitman. 2011. Bacterial catabolism of dimethylsulfoniopropionate (DMSP). Front. Microbiol. 2: 1–12. doi:10.3389/fmicb.2011.00172 Ribalet, F., A. Marchetti, K. A. Hubbard, and others. 2010. Unveiling a phytoplankton hotspot at a narrow boundary between coastal and offshore waters. Proc. Natl. Acad. Sci. 107: 16571– 16576. doi:10.1073/pnas.1005638107 147 Rideout, J. R., Y. He, J. A. Navas-Molina, and others. 2014. Subsampled open-reference clustering creates consistent, comprehensive OTU definitions and scales to billions of sequences. PeerJ 2: e545. doi:10.7717/peerj.545 Riseman, S. F., and G. R. DiTullio. 2004. 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. doi:10.1139/f04-052 Rognes, T., T. Flouri, B. Nichols, C. Quince, and F. Mahé. 2016. VSEARCH: a versatile open source tool for metagenomics. PeerJ 4: e2584–e2584. doi:10.7717/peerj.2584 Seymour, J. R., R. Simó, T. Ahmed, and R. Stocker. 2010. Chemoattraction to Dimethylsulfoniopropionate throughout the Marine Microbial Food Web. Science (80-. ). 329: 342–345. doi:10.1126/science.1188418 Six, K. D., S. Kloster, T. Ilyina, S. D. Archer, K. Zhang, and E. Maier-Reimer. 2013. Global warming amplified by reduced sulphur fluxes as a result of ocean acidification. Nat. Clim. Chang. 3: 1–4. doi:10.1038/nclimate1981 Spielmeyer, A., and G. Pohnert. 2012. Influence of temperature and elevated carbon dioxide on the production of dimethylsulfoniopropionate and glycine betaine by marine phytoplankton. Mar. Environ. Res. 73: 62–69. doi:10.1016/j.marenvres.2011.11.002 Stamatakis, A. 2014. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30: 1312–1313. doi:10.1093/bioinformatics/btu033 Stefels, J., M. Steinke, S. Turner, G. Malin, and S. Belviso. 2007. 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 Stoeck, T., D. Bass, M. Nebel, R. Christen, and D. Meredith. 2010. Multiple marker parallel tag environmental DNA sequencing reveals a highly complex eukaryotic community in marine anoxic water. 19: 21–31. doi:10.1111/j.1365-294X.2009.04480.x Tripp, H. J., J. B. Kitner, M. S. Schwalbach, J. W. H. Dacey, L. J. Wilhelm, and S. J. Giovannoni. 2008. SAR11 marine bacteria require exogenous reduced sulphur for growth. Nature 452: 741–744. doi:10.1038/nature06776 Villar, E., T. Vannier, C. Vernette, and others. 2018. The Ocean Gene Atlas: Exploring the biogeography of plankton genes online. Nucleic Acids Res. 46: W289–W295. 148 doi:10.1093/nar/gky376 Waterhouse, A. M., D. M. A. Martin, G. J. Barton, J. B. Procter, and M. Clamp. 2009. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 25: 1189–1191. doi:10.1093/bioinformatics/btp033 Webb, A. L., G. Malin, F. E. Hopkins, K.-L. Ho, U. Riebesell, K. Schulz, A. Larsen, and P. Liss. 2016. Ocean acidification has different effects on the production of dimethylsulfide and dimethylsulfoniopropionate measured in cultures of Emiliania huxleyi and a mesocosm study: a comparison of laboratory monocultures and community interactions. Environ. Chem. 13: 314–329. doi:http://dx.doi.org/10.1071/EN14268 Weiss, S., Z. Z. Xu, S. Peddada, and others. 2017. Normalization and microbial differential abundance strategies depend upon data characteristics. Microbiome 5: 27. doi:10.1186/s40168-017-0237-y 149$ Dissertation Conclusion Since the Industrial Revolution, humans have released an estimated 535 Gt of carbon into the atmosphere, of which ~30% has been absorbed by the oceans (IPCC 2013, 2018). Changes in ocean pH, temperature, light, oxygen, and nutrient availability are all expected to change with climate and significantly impact the marine ecosystem (Boyd et al. 2010; Gruber 2011; Dutkiewicz et al. 2013, 2015; Oschlies et al. 2018). The biogeochemical cycle of DMS(P) will inevitably be altered by changes in the future ocean but a better mechanistic understanding of the cycle, particularly for the drivers of DMSP production, is needed to predict these changes. However, we can expect that any changes in DMSP production and subsequent release into the dissolved pool will have the potential to alter the marine organic carbon cycle and Earth’s climate. Before DMS enters the atmosphere it is cycled through a complex microbial loop, where >95% of DMSP is consumed (Bates 1994; Simó et al. 2002). It is estimated that approximately 60% of marine bacteria have the potential to consume DMSP. For some marine bacterium, such as the SAR11 lineage which are sulfur auxotrophs, DMSP is a critical source of sulfur (Tripp et al. 2008; Moran et al. 2012). As marine bacteria are responsible for recycling ~50% of the dissolved organic matter pool (Hansell and Carlson 2001), changes in the supply of DMSP to these microbes could alter the organic carbon cycle in the surface ocean. In addition, changes in DMSP production have the potential to alter atmospheric DMS emissions, and hence global climate. In particular, a significant decrease in the emission of DMS to the atmosphere (greater than a 50% reduction) has been shown to increase global temperatures by 1.6°C. The response of climate to changes in DMS is nonlinear in time and space, where more significant changes to the Earth’s albedo are observed when DMS emission changes occur in areas of low aerosol concentration or in areas of high wind speeds (Gunson et al. 2006; Woodhouse et al. 2013). Considering that such small quantities of DMSP reach the atmosphere as DMS (1-5%), and the spatial dependency of the impacts of DMS on Earth’s albedo, any small, spatially heterogeneous changes in the production of DMSP have the potential to significantly impact the climate. Production of particulate DMSP by marine phytoplankton is the first step of the DMS(P) sulfur cycle, representing the total pool of DMSP available to be released as dissolved DMSP and subsequent degradation and/or consumption. This thesis focuses on this critical first step of the DMS(P) cycle and demonstrates that, despite significant effects of environmental stressors on the physiology of cellular DMSP production by low DMSP producers (LoDPs), community 150$ composition, specifically that of high DMSP producers (HiDPs), controls variability of in situ particulate DMSP concentrations. As the climate warms, significant changes in community composition are predicted to shift towards a more dominant picoplankton community, making HiDPs less abundant (Dutkiewicz et al. 2013, 2015). This thesis quantified the contribution of HiDPs to DMSP production and will ultimately improve predictions of in situ DMSP, which is critical for understanding how the supply of DMS(P) to the microbial loop and atmosphere will change in future oceans. Chapter 1 combines a meta-analysis of previously published monoculture studies, new measurements of DMSP production in, previously, poorly characterized DMSP producing groups, and a global ecosystem model to predict in situ particulate DMSP concentrations. I demonstrated the prevalence of DMSP production across the tree of life and successfully predicted in situ DMSP concentrations by modeling the mechanistic regulation of DMSP by HiDPs and LoDPs observed in monocultures. While the potential for differential regulation of HiDPs and LoDPs was previously hypothesized (Stefels et al. 2007), this was the first quantification of this hypothesis. This chapter establishes that HiDP abundance, not physiological stressors, drive in situ DMSP variability. Chapter 2 used targeted monoculture experiments to alter the metabolisms of a HiDP, Emiliania huxleyi and a LoDP, Thalassiosira oceanica. As hypothesized in Chapter 1, E. huxleyi did not change intracellular DMSP concentrations in metabolically imbalanced growth, suggesting that DMSP is produced constitutively. In contrast, T. oceanica consistently regulated intracellular DMSP concentrations in response to all stressed growth conditions tested, both metabolically balanced and metabolically imbalanced growth, suggesting that DMSP served the same stress mechanism across all conditions. The previously proposed stress mechanisms of antioxidant and overflow mechanism were not able to explain all of the experimental results. Based on this study, I hypothesize that the role of DMSP as a signaling molecule in LoDPs is the most likely and should be assessed further. Chapter 2 highlights the need for direct rate measurements of DMSP production. As DMSP synthesis genes were not available when I started my dissertation, I quantified DMSP production rates by adapting the method of Stefels et al. (2009). This rate measurement uses 13C labeled sodium bicarbonate NaH 13 CO3 to track the rate of DMSP synthesis, but required the use of a proton transfer reaction mass spectrometer which is an instrument not readily available in the 151$ oceanography field. I successfully developed a similar method using HPLC to isolate and fraction collect DMSP from cell lysates, and subsequently analyze the 13C/12C ratios of CO 2 from combusted DMSP on a Picarro isotope and gas analyzer. Across a 3 day growth curve, the ratio of 13C/12C in DMSP became more heavy. While this method was not included in the formal chapters of this thesis, it has promising implications for making the DMSP production rate measurement more accessible to the field in the future. Chapter 3 used natural community grow-out experiments to quantify the response of a mixed community to alleviation of nutrient stress while simultaneously measuring the DMSP producer community. Previous studies have quantified the DMSP degradation community, but, to our knowledge, this is the first study to use tag sequencing methods to quantify the DMSP producer community. These experiments confirmed that HiDPs dominate in situ DMSP production. Measuring the eukaryotic community 18S tag sequencing provided us the opportunity to apply the DMSP synthesis genes to in situ measurements. We demonstrate that not only do the synthesis genes differentiate between the types of DMSP producers, but also the HiDP synthesis gene has the potential to be a successful marker gene for quantifying HiDP abundance. Chapter 2 and 3 strongly support a role of a signaling molecule in LoDPs as T. oceanica upregulated DMSP across multiple different metabolic stressors, and LoDPs contribute <4% to bulk in situ DMSP. During my PhD, I made the first steps towards quantifying the impact of phytoplankton-heterotrophic interactions on DMSP production in a LoDP. The model HiDP and LoDP species used in Chapter 2 were grown axenic (confirmed with plating tests and DAPI stains) and then natural heterotrophic communities (media made with 1.2µm filtrate, rather than sterile seawater) were added-back and the production of DMSP was tracked over time. However, given that these interactions happen at such a small scale, bulk observations made it impossible to observe statistical differences. If DMSP synthesis genes had been available during my PhD, measurements of transcription during these experiments likely would have been sensitive enough to assess the signaling role of LoDPs. In addition to my own work, during my PhD, significant advances in knowledge of the DMS(P) cycle have been made. The most significant of these is certainly the discovery of the DMSP synthesis genes for heterotrophic bacteria and eukaryotic marine protists (Curson et al. 2017, 2018; Kageyama et al. 2018). Though from my analyses in Chapter 3, I believe that more synthesis gene likely exist. Additionally, Rhizaria, a critical group of marine protists that 152$ significantly contribute to organic carbon export (Lampitt et al. 2009), were identified as potential hotspots of DMSP production. Rhizaria host symbiotic partners that are typically HiDPs which, compared to free-living partners, appear to produce up to 100-fold higher intracellular concentrations (>700mM DMSP) when living with the host (Gutierrez-Rodriguez et al. 2017). Finally, a new component of the DMS(P) cycle was characterized in monocultures of HiDPs and LoDPs, dimethylsulfoxonium-propionate (DMSOP). DMSOP constitutes concentrations equal to ~0.1 -1% of cellular DMSP, and the authors propose this new metabolite to be a potential sink for reactive oxygen species in support of the antioxidant mechanism. Efforts towards identifying DMSP producer types in situ, increasing knowledge of the cellular mechanism of DMSP, and quantifying HiDP abundances should continue to be prioritized for improving predictions of in situ DMSP and ultimately understanding how DMSP supply will change in future oceans. For three decades the DMSP community has attempted to define a universal mechanism to explain production of DMSP in all marine phytoplankton. This has produced seven proposed hypotheses and countless laboratory and field studies often with conflicting results, resulting in insignificant progress towards defining the cellular mechanism and environmental drivers of DMSP production. My thesis has made a considerable contribution to the field by using multiple lines of evidence (models, laboratory, and field) to demonstrate that community composition and scale explain previously conflicting results and must be accounted for when quantifying the drivers of DMSP production and its ecosystem implications. Specifically, I showed that there are two different types of DMSP producers that regulate DMSP differently and for which DMSP appears to serve two different cellular functions. High DMSP producers (HiDPs) constitutively produce DMSP, likely as a compatible solute, and Low DMSP producers (LoDPs) actively regulate DMSP production in response to environmental conditions, possibly as a signaling molecule. I also provide evidence that DMSP production in these two groups evolved independently as two different genes encode for DMSP synthesis in these two groups. The cellular mechanisms defined in this thesis for HiDPs and LoDPs (constitutive and stress) allow for the first cohesive explanation of DMSP production across all marine phytoplankton, where previous studies often invoked shifting functions and environmental drivers to explain apparently conflicting experimental results. This thesis lays the foundation for future studies which must consider differential regulation of DMSP when quantifying DMSP production at different scales. To understand how DMSP production will change on a global scale in future oceans, the abundance and identity of 153$ species making up the HiDP community, not the physiology of the DMSP mechanism, must be targeted and accurately quantified. Global ecosystem models and satellite algorithms have significantly improved prediction power of the microbial community, but future improvements of predicting the sub-dominant community are critical for DMSP prediction. In this thesis I show that HiDP marker genes have the potential to accurately quantify the HiDP fraction of the microbial community. To apply this in the field, a universal HiDP primer(s) is needed as is the identification of a second, as-of-yet unknown HiDP gene. To understand the role of DMSP production at the cellular level, particularly as it pertains to microbial interactions that have the potential to influence primary production, the physiology of DMSP production must be considered. However, monoculture experiments are not sensitive enough to capture these interactions and LoDP DMSP synthesis genes are needed to best understand the role of DMSP at the micrometer scale. My thesis defines the regulating mechanisms of DMSP production at two different scales that were previously only considered as a whole and will ultimately improve predictions of how DMSP production will change in future oceans. Future work While the identification of two DMSP synthesis genes was a huge leap forward for the field, I have shown that we are still missing at least one important DMSP synthesis gene as several dominant HiDPs do not contain either of the current genes. Similarly, the majority of known LoDPs are missing both TpMT2 and DYSB. Future work should prioritize characterizing the DMSP synthesis genes and developing universal primers for the HiDP and LoDP genes. DSYB and TpMT2 discriminate between HiDPs and LoDPs, but more effort is needed to fully capture the entire DMSP producing community. One starting place could be TpMT1. TpMT1 is annotated as a methyltransferase enzyme and actually captures more of the LoDP community than TpMT2. Future work could include TpMT1 to understand why recombinant protein did not exhibit methyltransferase activity. Additionally, better quantifying the correlation of DSYB transcription and cellular DMSP concentrations would ultimately result in the best predictive power for particulate DMSP as this would eliminate the need to know cellular concentrations of producers in situ. Lastly, more work on the role of picoeukaryotes as dominant HiDPs in oligotrophic oceans should be prioritized. Picoeukaryotes were originally not included in our predictions of in situ DMSP in Chapter 1 as the DMSP phenotype of these species is poorly constrained. However, not 154$ including this sub-dominant population resulted in significant underestimation of in situ DMSP. Curson et al. (2018) found DSYB, the hypothesized HiDP marker gene, to be abundant in eukaryotes within the picoplankton size fraction of TARA. Picoeukaryotes likely dominate DMSP production in the oligotrophic oceans, and characterizing their importance will greatly benefit from omics based methods. Efforts towards characterizing the DMSP mechanism of LoDPs has clearly shown that these producers significantly upregulate DMSP in stressed growth conditions (here and by others, reviewed in Chapter 1). However, future efforts to assess the potential signaling mechanism of DMSP in LoDPs and role of DMSP in microbial interactions must include the DMSP synthesis gene and ideally microfluidic experiments as bulk monoculture experiments do not provide the sensitivity needed to quantify this mechanism. 155$ Conclusion References Bates, T. S. 1994. The cycling of sulfur in surface seawater of the northeast Pacific. J. Geophys. Res. 99: 7835–7843. doi:10.1029/93JC02782 Boyd, P. W., R. Strzepek, F. Fu, and D. a. Hutchins. 2010. Environmental control of open-ocean phytoplankton groups: Now and in the future. Limnol. Oceanogr. 55: 1353–1376. doi:10.4319/lo.2010.55.3.1353 Curson, A. R. J., J. Liu, A. Bermejo Martínez, and others. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nat. Microbiol. 2: 1–9. doi:10.1038/nmicrobiol.2017.9 Curson, A., B. Williams, B. Pinchbeck, and others. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat. Microbiol. 4: 430– 439. doi:10.1038/s41564-018-0119-5 Dutkiewicz, S., J. J. Morris, M. J. Follows, J. Scott, O. Levitan, S. T. Dyhrman, and I. Berman- Frank. 2015. Impact of ocean acidification on the structure of future phytoplankton communities. Nat. Clim. Chang. 5: 1002–1006. doi:10.1038/nclimate2722 Dutkiewicz, S., J. R. Scott, and M. J. Follows. 2013. Winners and losers: Ecological and biogeochemical changes in a warming ocean. Global Biogeochem. Cycles 27: 463–477. doi:10.1002/gbc.20042 Gruber, N. 2011. Warming up, turning sour, losing breath: ocean biogeochemistry under global change. Philos. Trans. A. Math. Phys. Eng. Sci. 369: 1980–96. doi:10.1098/rsta.2011.0003 Gunson, J. R., S. A. Spall, T. R. Anderson, A. Jones, I. J. Totterdell, and M. J. Woodage. 2006. Climate sensitivity to ocean dimethylsulphide emissions. Geophys. Res. Lett. 33: 2–5. doi:10.1029/2005GL024982 Gutierrez-Rodriguez, A., L. Pillet, T. Biard, W. Said-Ahmad, A. Amrani, R. Simó, and F. Not. 2017. Dimethylated sulfur compounds in symbiotic protists: A potentially significant source for marine DMS(P). Limnol. Oceanogr. 62: 1139–1154. doi:10.1002/lno.10491 Hansell, D. a, and C. a Carlson. 2001. Marine Dissolved Organic Matter and the Carbon Cycle. Society 14: 41–49. doi:10.5670/oceanog.2001.05 IPCC. 2013. Climate Change 2013: The physical science basis. Contribution of working group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, T.F. Stocker, D. Qin, G.-K. Plattner, et al. [eds.]. Cambridge University Press. 156$ IPCC. 2018. Summary for policymakers, In Masson-Delmotte, V.P. Zhai, H.-O. Pörtner, et al. [eds.], Global Warming of 1.5C. An IPCC Special Report on the impacts of global warming of 1.5C above pre-industrial levels and related global greenhouse gas emission pathways, in the cntext of strengthening the global response to the threat of climate change, su. World Meteorological Organization. Kageyama, H., Y. Tanaka, A. Shibata, R. Waditee-Sirisattha, and T. Takabe. 2018. Dimethylsulfoniopropionate biosynthesis in a diatom Thalassiosira pseudonana: Identification of a gene encoding MTHB-methyltransferase. Arch. Biochem. Biophys. 645: 100–106. doi:10.1016/j.abb.2018.03.019 Lampitt, R. S., I. Salter, and D. Johns. 2009. Radiolaria: Major exporters of organic carbon to the deep ocean. Global Biogeochem. Cycles 23: 1–9. doi:10.1029/2008GB003221 Moran, M. A., C. R. Reisch, R. P. Kiene, and W. B. Whitman. 2012. Genomic Insights into Bacterial DMSP Transformations. Ann. Rev. Mar. Sci. 4: 523–542. doi:10.1146/annurev- marine-120710-100827 Oschlies, A., P. Brandt, L. Stramma, and S. Schmidtko. 2018. Drivers and mechanisms of ocean deoxygenation. Nat. Geosci. 11: 467–473. doi:10.1038/s41561-018-0152-2 Simó, R., S. D. Archer, C. Pedrós-Alió, L. Gilpin, and C. E. Stelfox-Widdicombe. 2002. Coupled dynamics of dimethylsulfoniopropionate and dimethylsulfide cycling and the microbial food web in surface waters of the North Atlantic. Limnol. Oceanogr. 47: 53–61. doi:10.4319/lo.2002.47.1.0053 Stefels, J., J. W. H. Dacey, and J. T. M. Elzenga. 2009. In vivo DMSP-biosynthesis measurements using stable-isotope incorporation and proton-transfer-reaction mass spectrometry (PTR-MS). Limnol. Oceanogr. Methods 7: 595–611. doi:10.4319/lom.2009.7.595 Stefels, J., M. Steinke, S. Turner, G. Malin, and S. Belviso. 2007. 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 Tripp, H. J., J. B. Kitner, M. S. Schwalbach, J. W. H. Dacey, L. J. Wilhelm, and S. J. Giovannoni. 2008. SAR11 marine bacteria require exogenous reduced sulphur for growth. Nature 452: 741–744. doi:10.1038/nature06776 157$ Woodhouse, M. T., G. W. Mann, K. S. Carslaw, and O. Boucher. 2013. Sensitivity of cloud condensation nuclei to regional changes in dimethyl-sulphide emissions. Atmos. Chem. Phys. 13: 2723–2733. doi:10.5194/acp-13-2723-2013
Abstract (if available)
Abstract
Dimethylsulfoniopropionate (DMSP) is a labile sulfur and carbon metabolite that significantly contributes to both the cycling of marine dissolved organic carbon and the balance of Earth’s albedo. DMSP is produced by the majority of eukaryotic marine phytoplankton and by many prokaryotes, but despite decades of research, the cellular mechanism and environmental drivers of DMSP production remain unknown. My thesis confirms that the cellular mechanism of DMSP is differentiated by the cellular concentrations of DMSP in different producers, where high DMSP producers (e.g. dinoflagellates and haptophytes) constitutively produce DMSP and low DMSP producers (e.g. cyanobacteria and diatoms) actively regulate DMSP production in response to environmental stress. However, with natural community experiments and global model predictions, my thesis demonstrates that variability of in situ DMSP production is driven by the biomass of high producers. My thesis highlights the potential for predicting in situ DMSP concentrations with a high DMSP producer marker gene and demonstrates the importance of accurately capturing the sub-dominant community for prediction of DMSP, or other similar metabolites produced by a small fraction of the marine microbial community. Insight into the differential regulation of DMSP by HiDPs and LoDPs presented here should dramatically shift understanding of in situ DMSP cycling. Previous work assumes a universal mechanism of DMSP in all producers, but this thesis clearly demonstrates the importance of differential regulation across DMSP producer taxonomy which should be considered when resolving the significance of DMSP in carbon cycling and climate regulation.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Thermal diversity within marine phytoplankton communities
PDF
Iron-dependent response mechanisms of the nitrogen-fixing cyanobacterium Crocosphaera to climate change
PDF
The impact of Lagrangian environmental variability on the growth of phytoplankton
PDF
Identifying functional metabolic guilds: a computational approach to classifying heterotrophic diversity in the marine system
PDF
Characterizing protistan diversity and quantifying protistan grazing in the North Pacific Subtropical Gyre
PDF
The impact of mesoscale and submesoscale physical processes on phytoplankton biomass, community composition, and carbon dynamics in the oligotrophic ocean
PDF
The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes
PDF
Thermal acclimation and adaptation of key phytoplankton groups and interactions with other global change variables
PDF
Spatial and temporal dynamics of marine microbial communities and their diazotrophs in the Southern California Bight
PDF
Examining potential triggers of algal blooms and harmful algae in the Southern California bight region
PDF
Future impacts of warming and other global change variables on phytoplankton communities of coastal Antarctica and California
PDF
Spatial and temporal investigations of protistan grazing impact on microbial communities in marine ecosystems
PDF
Genetic characterization of microbial eukaryotic diversity and metabolic potential
PDF
Enhancing recovery of understudied and uncultured lineages from metagenomes
PDF
Unexplored microbial communities in marine sediment porewater
PDF
Carbonate dissolution at the seafloor: fluxes and drivers from a novel in situ porewater sampler
PDF
Disentangling the ecology of bacterial communities in cnidarian holobionts
PDF
Temporal variability of marine archaea across the water column at SPOT
PDF
The distributions and geochemistry of iodine and copper in the Pacific Ocean
PDF
The distribution of B-vitamins in two contrasting aquatic systems, and implications for their ecological and biogeochemical roles
Asset Metadata
Creator
McParland, Erin Lynn
(author)
Core Title
The dynamic regulation of DMSP production in marine phytoplankton
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Biology (Marine Biology and Biological Oceanography)
Publication Date
07/24/2019
Defense Date
05/08/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
biogeochemical cycling,community composition,compatible solute,dimethylsulfoniopropionate (DMSP),OAI-PMH Harvest,organic sulfur,phytoplankton,stress molecule
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Levine, Naomi M. (
committee chair
), Hutchins, David (
committee member
), John, Seth (
committee member
), Webb, Eric (
committee member
)
Creator Email
erinmcp426@gmail.com,mcparlan@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-189261
Unique identifier
UC11663111
Identifier
etd-McParlandE-7606.pdf (filename),usctheses-c89-189261 (legacy record id)
Legacy Identifier
etd-McParlandE-7606.pdf
Dmrecord
189261
Document Type
Dissertation
Format
application/pdf (imt)
Rights
McParland, Erin Lynn
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
biogeochemical cycling
community composition
compatible solute
dimethylsulfoniopropionate (DMSP)
organic sulfur
phytoplankton
stress molecule