Significance of N2 fixingplanktonic symbioses for openocean ecosystems Marcus Stenegren

Marcus Stenegren    Sign

ificance of N

2 fixing plan

ktonic sym

bioses for open ocean

ecosystems

Doctoral Thesis in Marine Ecology at Stockholm University, Sweden 2020

Department of Ecology, Environment andPlant Sciences

ISBN 978-91-7797-933-3

Marcus Stenegren

   I. Stenegren M, Caputo A, Berg C, Bonnet S and Foster RA. (2018).Distribution and drivers of symbiotic and free-living diazotrophiccyanobacteria in the western tropical South Pacific. Biogeosciences, 15,1559–1578.      II. Stenegren M, Berg C, Padilla CC, David S-S, Montoya JP, YagerPL and Foster RA. (2017). Piecewise Structural Equation Model (SEM)disentangles the environmental conditions favoring Diatom DiazotrophAssociations (DDAs) in the Western Tropical North Pacific (WTNA).Front. Microbiol. 8:810.      III. Caputo A, Stenegren M, Pernice MC and Foster RA. (2018). Ashort comparison of two marine planktonic diazotrophic symbioseshighlights an unquantified disparity. Front. Mar. Sci. 5:2.      IV. Stenegren M, Weber SC, Brodin D, Montoya JP, SubramaniamA, Voss M and Foster RA. (Manuscript). Insights on symbiotic diatomdiazotroph associations in the South China Sea by targeted microarrayanalysis of three symbiont strains.

Significance of N2 fixing planktonic symbioses foropen ocean ecosystemsMarcus Stenegren

Academic dissertation for the Degree of Doctor of Philosophy in Marine Ecology at StockholmUniversity to be publicly defended on Friday 10 January 2020 at 10.00 in Vivi Täckholmssalen,NPQ-huset, Svante Arrhenius väg 20.

AbstractDi-nitrogen (N2) fixers, also called diazotrophs, are able to reduce atmospheric N2 into bioavailable nitrogen, giving theman advantage in open ocean regions with low dissolved inorganic nitrogen concentrations. The focus of this thesis arethree lineages of symbiotic heterocystous filamentous types (het-1, het-2 and het-3), that associate with several genera ofmicroalgae called diatoms (collectively referred to as Diatom Diazotroph Associations, DDAs). Other major cyanobacterialdiazotrophs in the ocean are the filamentous Trichodesmium spp., and the unicellular UCYN-A, UCYN-B and UCYN-C. Although widespread in the tropics and subtropics, and first described in the early 20th century, the DDAs are anunderstudied group of diazotrophs. Hence, our knowledge of their distribution, abundance, activity, and how these areconstrained by the environment is limited.

Initially we investigated the abundances and distributions of eight cyanobacterial diazotrophs, and two proposed micro-algal hosts of UCYN-A1 and A2, in the western tropical south Pacific (WTSP), using quantitative polymerase chainreaction (qPCR). Trichodesmium spp. was the most abundant diazotroph and het-1 was the most abundant DDA symbiont.Using correlation analysis a distinct vertical separation was observed between UCYN-A and the other diazotrophs(Trichodesmium spp., UCYN-B and DDA symbionts). The most influential environmental parameter on the diazotrophabundances in the WTSP was temperature, and in order to investigate this further we compiled qPCR data from 11 publiclyavailable datasets from four ocean basins. Using a weighted meta-analysis we found that temperature was a robust factorgoverning the diazotroph abundances (except for UCYN-A) across ocean basins.

Attempting to identify differences in environmental impacts on two of the DDA symbiont strains (het-1 and het-2),we applied a new statistical tool called piecewise structural equation model, on qPCR abundance data from the westerntropical North Atlantic. We saw that the two strains had a direct positive interaction between each other, but two parameters(salinity and dissolved inorganic phosphorous) differed. Based on a direct positive effect of salinity on het-1, and an indirectnegative effect of salinity on het-2, we concluded that het-2 prefers intermediate salinities (30-35 PSU), which is consistentwith where observations of het-2 blooms have been made.

Although DDA and UCYN-A symbionts both are major contributors of new N, and are symbiotic, they have severalunique differences. The host partners differ in phylogeny (diatom vs prymnesiophyte), size (80-250 vs 7-10 µm) and thesymbiotic life history (colonial vs solitary). The larger, colonial nature of DDAs make them difficult to collect, and hencethey are often under-sampled and undetected. In fact, after reviewing 46 qPCR studies we found that < 30% of the studies(13 out of 46) quantified all three DDA symbionts, compared to UCYN-A (96%, 44 out of 46).

In order to study the DDA symbiont gene expressions we developed a highly specific DDA symbiont microarray (748probes), which was applied on samples collected in the South China Sea. Although the gene expression levels were highlyvariable, we observed an upregulation of the nifH gene (for N2 fixation) in the night. Investigating environmental impacton overall gene expression levels, we found that fluorescence, temperature and salinity was most important. Temperatureand salinity also constrained abundances, but fluorescence could be seen as a proxy for either other phytoplankton or lightavailability, suggesting that daylight and host influence DDA symbiont gene expression levels.

The results of this thesis broaden our understanding of the DDAs and how their ambient environment influences them. Ithas also opened up new possibilities for in depth analysis of these complex environmental impacts. Lastly, it has providednew analysis tools for further development on how the symbionts and hosts potentially impact each other’s activities.

Keywords: nitrogen fixation, diazotrophs, symbiosis, DDA, cyanobacteria, qPCR, piecewise SEM, microarray, tropics,subtropics, marine, open ocean, Richelia, Calothrix.

Stockholm 2020http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-176275

ISBN 978-91-7797-933-3ISBN 978-91-7797-934-0

Department of Ecology, Environment and Plant Sciences

Stockholm University, 106 91 Stockholm

SIGNIFICANCE OF N2 FIXING PLANKTONIC SYMBIOSES FOROPEN OCEAN ECOSYSTEMS 

Marcus Stenegren

Significance of N2 fixingplanktonic symbioses for openocean ecosystems 

Marcus Stenegren

©Marcus Stenegren, Stockholm University 2020 ISBN print 978-91-7797-933-3ISBN PDF 978-91-7797-934-0 Cover image: The cyanobacteria Richelia intracellularis in symbiosis with the diatom Rhizosolenia cleveii,observed with light microscopy from waters outside station Aloha, Hawaii, North Pacific Ocean. The image is acourtesy of Cherry Gao. Paper I cover image: Aerial view of R/V L'Atalante in the Melanesian Archipelago. The image is a courtesy of thethe OUTPACE 2015 research team.Paper II cover image: Amazon River influenced seawater near the coast of French Guinea.Paper III cover image: The Hemiaulus-Richelia symbiosis as seen in an epifluorescence microscope. The image is acourtesy of Rachel A. Foster.Paper IV cover image: Boundary between river plume influenced seawater and oceanic water near the MekongRiver in the South China Sea. The image is a courtesy of the FK160603 research team. Printed in Sweden by Universitetsservice US-AB, Stockholm 2019

Mikaela, Haley, Levi,Tindra, Max, Flora

6

Sammanfattning

Kvävefixerare, även kallade diazotrofer, kan reducera atmosfärisk kvävgas till biotillgängligt kväve, vilket ger dem en fördel i regioner i öppna hav med låga koncentrationer av löst oorganiskt kväve. Fokus för denna avhandling är tre stammar av heterocysta, filamentösa diazotrofer (het-1, het-2 och het-3) som bildar symbioser med ett antal släkten av kiselalger (tillsammans kallas de ”Diatom Diazotroph Associations”, DDA). Andra viktiga cyanobacteriella diazotrofer i öppna hav är de filamentösa Trichodesmium spp., och de encelliga UCYN-A, UCYN-B och UCYN-C. Trots att de är utbredda i tropikerna och sub-tropikerna och beskrevs redan i början av 1900-talet, är DDA symbionterna en understuderad grupp av diazotrofer. Därför är vår kunskap om deras distribution, abundans, aktivitet och fördelaktiga miljöbetingelser begränsad.

Inledningsvis undersökte vi, med hjälp av ”quantitative polymerase chain reaction” (qPCR) -analyser abundans och distribution för åtta cyanobakteriella diazotrofer och två föreslagna symbiotiska värdceller av UCYN-A1 och A2 i sydvästra tropiska Stilla Havet. Trichodesmium spp. var den vanligast förekommande diazotrofen och het-1 var den vanligast förekommande DDA-symbionten. Med hjälp av korrelationsanalys observerade vi en distinkt vertikal separation mellan UCYN-A och de andra diazotroferna (Trichodesmium spp., UCYN-B och DDA-symbionterna). Den viktigaste miljöfaktorn för diazotrofernas abundans i sydvästra Stilla Havet var temperatur, och för att undersöka detta ytterligare sammanställde vi 11 publika qPCR-dataset från fyra olika hav. Med hjälp av en ”viktad metaanalys” visade det sig att temperatur återigen var en viktig miljöfaktor (med undantag för UCYN-A).

I ett försök att identifiera skillnader i miljöpåverkan på två av DDA-symbiont-stammarnas (het-1 och het-2) abundanser i nordvästra tropiska Atlanten, använde vi oss av ett nytt statistiskt verktyg som kallas ”piecewise structural equation model”. Även om de två stammarna hade en direkt positiv interaktion mellan varandra, verkade salthalt och löst oorganiskt fosfor ha olika effekter på dem. Baserat på en direkt positiv effekt av salinitet på het-1 och en indirekt negativ effekt av salinitet på het-2, drog vi slutsatsen att het-2 föredrar salinitet som är något lägre (30-35 PSU), vilket stämmer med observationer av blommande het-2.

Trots att DDA och UCYN-A båda är viktiga kvävefixerare och bildar symbioser, har de flertalet unika skillnader. Deras värdpartners skiljer sig i fylogeni (prymnesiofyt/kiselalg), storlek (7-10 µm/80-250 µm) och livshistoria (enskild/kolonial). DDA:ernas större och koloniala celler gör dem svåra att samla in och därför är de ofta understuderade och inte detekterade. Faktum är att efter att ha granskat 46 qPCR-studier fann vi att <30% av studierna (13 av 46) kvantifierade alla tre DDA-symbionterna, jämfört med UCYN-A (96%, 44 av 46).

7

För att kunna studera DDA-symbionternas genuttryck utvecklade vi ett specifikt DDA-symbiont-microarray (748 prober), som applicerades på prover insamlade i Sydkinesiska havet. Även om genuttrycken varierade mycket, observerade vi en uppreglering av nifH-genen (för kvävefixering) på natten. Vid undersökningen av påverkan av olika miljöparametrar på genuttryck var fluorescens, temperatur och salinitet viktigast. Vi observerade tidigare att temperatur och salinitet även påverkar abundansen, men fluorescens kan ses som en proxy för antingen andra fytoplankton eller ljustillgänglighet, vilket antyder att dagsljus och värdcellen påverkar DDA-symbionternas genuttryck.

Resultaten av denna avhandling breddar vår förståelse för DDA:erna och hur de påverkas av deras omgivande miljö. Dessutom öppnar det upp förr ytterligare möjligheter att undersöka den komplexa påverkan som miljön har. Avslutningsvis bidrar den med nya analysverktyg för att studera hur symbionten och värdcellen potentiellt påverkar varandras aktivitet.

8

List of papers

Papers included in this thesis:

I. Stenegren M, Caputo A, Berg C, Bonnet S and Foster RA.

(2018). Distribution and drivers of symbiotic and free-living

diazotrophic cyanobacteria in the wester tropical South Pacific.

Biogeosciences, 15, 1559–1578.

II. Stenegren M, Berg C, Padilla CC, David S-S, Montoya JP,

Yager PL and Foster RA. (2017). Piecewise Structural Equation

Model (SEM) disentangles the environmental conditions

favoring Diatom Diazotroph Associations (DDAs) in the

Western Tropical North Pacific (WTNA). Front. Microbiol.

8:810.

III. Caputo A, Stenegren M, Pernice MC and Foster RA. (2018). A

short comparison of two marine planktonic diazotrophic

symbioses highlights an unquantified disparity. Front. Mar. Sci.

5:2.

IV. Stenegren M, Weber SC, Brodin D, Montoya JP, Subramaniam

A, Voss M and Foster RA. (Manuscript). Insights on symbiotic

diatom diazotroph associations in the South China Sea by

targeted microarray analysis of three symbiont strains.

9

My contribution to the papers:

I. Sample collection, DNA extraction and qPCR “at sea” and in

lab, microscopy analysis, synthesis of meta-dataset, data

analysis and statistics, co-lead writing and revisions.

II. Data analysis, Piecewise SEM optimization, application and

interpretation, co-lead writing and revisions.

III. Contributed to compilation of data for summary table. Compiled

genomic data and summary of qPCR public datasets, and

generation of main figure. Contributed to writing.

IV. Experimental design, RNA extraction and “in house” quality

control, microarray design and specificity control, Piecewise

SEM optimization, application and interpretation, post-data

analysis and statistics, wrote the first draft.

10

Co-authors’ contributions to the papers:

I. AC collected samples, did microscopy analysis and provided the

manuscript text related to microscopy. CB helped with the

statistical analysis and wrote the R script. SB provided the

opportunity to participate in the OUTPACE2015 cruise. RAF

assisted in planning, execution of the project, microscopy counts,

and co-wrote the paper.

II. CB helped with the R script and design of the statistical model

framework. S-SD and CP performed DNA extraction and qPCR.

JM collected, measured and provided the nutrient data. RAF

designed the project, collected the samples, did microscopy

observations and co-wrote the paper.

III. RAF led the writing with contribution from all co-authors.AC

led the organizing and microscopy content of Table 1.

IV. DB managed the raw data, provided the R scripts and initial heat

maps for hierarchical cluster analysis. JPM, MV and AS

provided the ship time (FK160303) and collected the

environmental data. SCW collected, analyzed and provided the

nutrient data. RAF extracted DNA, conducted the qPCR,

supervised the project and assisted in writing.

11

Additional papers:

Foster RA, Stenegren M, Franzke D, Littman S, Whitehouse M, Kuypers

MMM and White AE. (Manuscript). Pairing up in the plankton: diatoms and

cyanobacteria work together in a nutrient deplete ocean.

Spungin D, Belkin N, Foster RA, Stenegren M, Caputo A, Pujo-Pay M,

Leblond N, Dupouy C, Bonnet S and Berman-Frank I. (2018). Programmed

cell death in diazotrophs and the fate of organic matter in the western

tropical South Pacific Ocean during the OUTPACE cruise. Biogeosciences,

15, 3893–3908.

Bonnet S, Caffin M, Berthelot H, Grosso O, Benavides M, Helias-Nunige S,

Guieu C, Stenegren M and Foster RA. (2018). In-depth characterization of

diazotrophs activity across the western tropical South Pacific hotspot of N2

fixation (OUTPACE cruise). Biogeosciences, 15, 4215–4232.

12

Public data:

Stenegren M, Caputo A, Berg C, Bonnet S and Foster RA. (2018).

Abundance of symbiotic and free-living diazotrophic cyanobacteria in the

western tropical South Pacific during L’Atalante cruise OUTPACE in 2015.

https://doi.org/10.1594/PANGAEA.887730.

Stenegren M, Brodin D and Foster RA. (2019). Targeted microarray

analysis of two symbiotic cyanobacterial diazotroph strains in the western

tropical North Atlantic.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE139589 NCBI

accession number GSE139589.

Stenegren M, Brodin D and Foster RA. (2019). Targeted microarray

analysis of three symbiotic cyanobacterial diazotroph strains in the South

China Sea.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE139588. NCBI

accession number GSE139588.

13

Table of Contents

1 Introduction .............................................................................................. 18

1.1 Marine N2 fixation and its global importance ................................ 19

1.2 Estimating the marine N budget ...................................................... 21

1.3 Measuring pelagic N2 fixation rates ................................................ 24

1.4 Open ocean cyanobacterial diazotrophs ......................................... 25

1.5 Open ocean non-cyanobacterial diazotrophs ................................. 32

1.6 Differing N2 fixation strategies ......................................................... 34

1.7 A symbiotic life history of DDA and UCYN-A ............................... 36

1.8 Environmental constraints on marine diazotrophs ....................... 39

1.8.1 Nutrients ..................................................................................... 39

1.8.2 Hydrographic parameters ......................................................... 44

2 Thesis aims ................................................................................................ 45

3 Comments on methodology ..................................................................... 46

3.1 Sampling for diazotrophs ................................................................. 46

3.2 Quantifying diazotrophs and their gene expression levels ............ 50

3.3 Challenges and caveats in detection of diazotrophs ....................... 54

3.4 Statistics as powerful tools of investigation .................................... 56

4 Results and discussion ............................................................................. 58

4.1 Spatial and vertical detections of diazotrophs ................................ 59

4.2 Applying piecewise SEM on two DDA symbiont strains ............... 63

4.3 Detection discrepancies of the UCYN-A symbioses ....................... 65

4.4 Comparing DDA and UCYN-A symbioses ..................................... 67

4.5 Gene expression levels of the DDA symbionts ................................ 69

5 Conclusions ............................................................................................... 71

6 Future perspectives .................................................................................. 72

6.1 Diazotrophs in future ocean conditions .......................................... 72

6.2 Field observations and modeling environmental constraints ........ 74

6.3 More reliable detections of diazotrophs .......................................... 77

14

6.4 Hypotheses generated by pSEM ...................................................... 78

6.5 Symbiont-host interactions and host control .................................. 80

7 Acknowledgments .................................................................................... 81

8 References ................................................................................................. 82

15

Abbreviations

ANAMMOX Anaerobic Ammonium Oxidation

APA Alkaline Phosphatase Activity

AR Amazon River

ARA Acetylene Reduction Assay

ATP Adenosine Triphosphate

bp Base pair

C Carbon

CalSC01 Draft genome of Calothrix rhizosoleniae (associated with Chaetoceros compressus)

CARD-FISH Catalyzed Reporter Deposition Fluorescence In

Situ Hybridization

CO2 Carbon dioxide

CTD Conductivity, Temperature, Depth

DDA Diatom Diazotroph Association

DIN Dissolved Inorganic Nitrogen: includes nitrate, nitrite and ammonium

DIP Dissolved Inorganic Phosphorous

DNA Deoxyribonucleic Acid

DOC Dissolved Organic Carbon

DOM Dissolved Organic Matter

DOP Dissolved Organic Phosphorous

ESP Environmental sample processor

Fe Iron

FISH Fluorescence In Situ Hybridization

Het-1 Clade designation in nifH phylogeny for Richelia intracellularis in symbiosis with Rhizosolenia sp.

16

Het-2a Clade designation in nifH phylogeny for Richelia intracellularis in symbiosis with Hemiaulus hauckii

Het-2b Clade designation in nifH phylogeny for Richelia intracellularis in symbiosis with Hemiaulus membranaceus

Het-3 Clade designation in nifH phylogeny for Calothrix rhizosoleniae in symbiosis with Chaetoceros compressus

ITS Internal Transcribed Spacer

Mo Molybdenum

N2 Di-nitrogen gas

N Nitrogen

NCD Non-Cyanobacterial Diazotroph

NEP North East Pacific Ocean

NH3 Ammonia

NO3- Nitrate

nifH Gene encoding the highly conserved nitrogenase Fe subunit

nMDS Non-metric Multidimensional Scaling

nt Nucleotide

O2 Oxygen

ODZ Oxygen Deficient Zone

OMZ Oxygen Minimum Zone

P Phosphorous

PAR Photosynthetically Active Radiation

PCR Polymerase Chain Reaction

POM Particulate Organic Matter

PSI Photosystem I

17

PSII Photosystem II

pSEM Piecewise Structural Equation Model

qPCR Quantitative Polymerase Chain Reaction

RDA Redundancy Analysis

RIN RNA Integrity Number

RintRC01 Draft genome of Richelia intracellularis (associated with Rhizosolenia sp.)

RintHH01 Draft genome of Richelia intracellularis (associated with Hemiaulus hauckii)

RintHM01 Draft genome of Richelia intracellularis (associated with Hemiaulus membranaceus)

RNA Ribonucleic Acid

RuBisCo Ribulose- 1,5-bisphosphate carboxylase-oxygenase

SC01 Strain name of the isolated Calothrix

rhizosoleniae

SCS South China Sea

Si Silica

SIMS Secondary Ion Mass Spectrometry

TCA Tricarboxylic Acid

UCYN Unicellular Cyanobacteria (groups A, B, or C)

Vn Vanadium

WTNA Western tropical North Atlantic

WTSP Western tropical South Pacific

18

1 Introduction Nitrogen (N) is an essential nutrient and is part of all living organisms, including the code of life, or deoxyribonucleic acid (DNA). However, despite di-nitrogen (N2) gas making up approximately 78 % of the Earth's atmosphere, it is unavailable as a nutrient source for most organisms since they lack the enzyme required to reduce N2. Instead, most living organisms acquire N in other more bioavailable forms, either as organic (e.g. amino acids, urea) or inorganic N (e.g. ammonium, nitrate and nitrite), from the environment or via food. However, some N2-fixing bacteria and archaea, known as diazotrophs, are able to reduce N2 into bioavailable ammonia (NH3), a process called biological N2 fixation (hereafter N2 fixation).

In nature, N2 fixation occurs in a range of ecosystems, including terrestrial and aquatic habitats, e.g. boreal forests (DeLuca et al., 2002), arctic lowlands (Chapin et al., 1991), marine environments (Capone and Carpenter, 1982), freshwater and brackish estuaries (Howarth et al., 1988; Lehtimaki et al., 1997), deep seas (Dekas et al., 2009) and even the guts of invertebrates and whale falls (Carpenter and Culliney, 1975; Dekas et al., 2018; French et al., 1976). In the environment, where nutrients are scarce, N2 fixation adds a significant amount of bioavailable N to the N budget and therefore fertilizes the surrounding microbial and plant communities (Mulholland et al., 2002). This process of N2 fixation by microbes has been known and implemented in anthropogenic agricultural practices for centuries (Peoples and Craswell, 1992). Due to their cosmopolitan distribution, diazotrophs play vital roles in many ecosystems. However, little is still known about the distribution (vertical and horizontal) of marine diazotrophs and the cellular regulation of their N2 fixation, and how it is affected by ambient environmental conditions. One group of largely understudied N2 fixers is the heterocystous cyanobacterial lineages that form associations, or symbioses, with diatoms. This thesis investigates what biotic and abiotic factors constrain the globally important process of N2 fixation, with an emphasis on the diatom diazotroph associations (DDAs).

19

1.1 Marine N2 fixation and its global importance

The process of N2 fixation is catalyzed by an enzyme complex called nitrogenase, which is encoded by a suite of nif genes (Young, 2005; Zehr et al., 2008). Nitrogenase is highly sensitive to oxygen (O2) (Eady and Leigh, 1994). The following chemical reaction represents N2 reduction by the Fe-Mo nitrogenase enzyme (Howard and Rees, 1996):

N2 + 8H+ + 16ATP + 8e– → 2NH3 + H2 + 16ADP + 16Pi

The most common form of marine nitrogenase is composed of two proteins: di-nitrogen reductase (the iron (Fe) protein) and di-nitrogenase (the Fe-molybdenum (Mo) protein) (Eady and Leigh, 1994). There are also less common alternative nitrogenases (McRose et al., 2017), which instead of the Fe-Mo in the di-nitrogenase protein are composed of Fe-Fe or Fe-vanadium (Vn) (Eady and Leigh, 1994). However, the alternative nitrogenases are more energetically expensive for reducing N2 (Eady and Leigh, 1994) due to a lower flux of electron transport across the Vn nitrogenase (Miller and Eady, 1988). The distribution of alternative nitrogenases in the environment is thought to be coupled with trace metal concentrations (McRose et al., 2017), meaning that there is prevalence for e.g. Vn nitrogenase when Mo is scarce. Moreover, the Fe-Vn nitrogenase also seems to be more efficient at lower temperatures (Miller and Eady, 1988). However, what all nitrogenases have in common is the high demand for Fe and adenosine triphosphate (ATP), and the production of two molecules of bioavailable NH3. The actual reduction of N2 is performed by a sequential transfer of electrons between the di-nitrogenase reductase component and a specialized electron donor component (Howard and Rees, 1996; Wenke, 2019).

It is well known that marine N2 fixation is an important counterweight against ecosystem N loss processes caused by e.g. denitrification and anaerobic ammonium oxidation (ANAMMOX). N2 fixation also stimulates primary production and the subsequent drawdown of carbon dioxide (CO2) from the atmosphere; a process termed the biological pump (Karl et al., 2012; Volk and Hoffert, 1985). Much of the reduced carbon (C) acts as the foundation of marine food webs through grazing by zooplankton and the subsequent consumption by higher trophic levels. Some C also degrades and recycles in the sunlit layer due to dying plankton, and

20

fuels the microbial loop (Sellner, 1997). Lastly, ca 3-14 % of biogenic C and N sinks, as senescent and dying cells, out of the euphotic (sunlit) zone (Lohrenz et al., 1992). It is only a small fraction of this C (on average 5 % (Karl et al., 2012)) that actually sequesters to the ocean floor (Fig. 1). The rest will be re-mineralized and re-introduced to the euphotic zone. Thus, N2 fixation constitutes a major source of new N, and also drives a net C sequestration given that upwelling or advection of re-mineralized N is always accompanied by C (Eppley and Peterson, 1979).

Figure 1. Simple schematic of the biological pump which depicts the drawdown of CO2 and

N2, the incorporation of these elements into biomass, the sinking of particulate organic C

and N (POC and PON) to the sub-euphotic (below the sunlit) zone, the small fraction

sequestered to the sea floor, and lastly the re-introduction of CO2 and re-mineralized N in

the form of nitrate (NO3-) to the euphotic zone. Note that the figure is not to scale. Modified

and reproduced with permission from Sohm et al. (2011).

Some fixed N remains in the euphotic zone, and as mentioned, can be utilized by grazers and other pelagic microbes. Bioavailable N will both ‘leak’ from the living diazotrophs, as well as be released upon cell lysis (Berman-Frank et al., 2004; Mulholland et al., 2006). Thus, diazotrophic growth, particularly regional blooms, may prime the pelagic community for subsequent successions of phytoplankton, e.g. dinoflagellates (Mulholland et al., 2006). Since grazing on marine diazotrophic cyanobacteria usually is limited (Conroy et al., 2017) due to e.g. toxins (O’Neil, 1998), a diazotroph-induced succession of phytoplankton could instead indirectly benefit grazers (Motwani et al., 2018) and subsequently also higher trophic levels.

21

1.2 Estimating the marine N budget

For decades it was assumed that a majority of marine N2 fixation occurred in areas of low to depleted concentrations of dissolved inorganic N (DIN) in the surface waters of tropical and subtropical regions (Fig. 2) (Carpenter et al., 1999; Wang et al., 2019). However, several studies during the last 15 years are challenging this canonical paradigm. For example, measurable rates of N2 fixation in deeper waters, including oxygen minimum zones (OMZ) have been detected (Bonnet et al., 2013; Fernandez et al., 2011; Halm et al., 2009). Moreover, both presence and activity of diazotrophs have been observed in high latitude locations such as recently reported in Arctic marine waters (Díez et al., 2012; Harding et al., 2018; Martínez-Pérez et al., 2016; Shiozaki et al., 2017, 2018), temperate coastal regions where DIN is not always depleted (Bentzon-Tilia et al., 2015; Mulholland et al., 2012, 2019; Short and Zehr, 2007; White et al., 2013), estuaries (Bentzon-Tilia et al., 2015; Messer et al., 2015) and river influenced oceanic regions (Bombar et al., 2011; Carpenter, 1983; Carpenter et al., 1999; Foster et al., 2007, 2009; Goes et al., 2014). The new insights on a broader distribution of active marine N2 fixers, provides cumulative evidence that the proposed imbalanced global marine N budget is due to an underestimation of N2 fixation (Gruber and Sarmiento, 1997; Karl et al., 1997).

Figure 2. Modelled N2 fixation rates in the global oceans. Reproduced with permission by

Wang et al. (2019).

22

Global estimates of pelagic N2 fixation have historically been difficult to model, and there are still uncertainties regarding environmental constraints, and the rates and distributions of marine N2 fixation. Moreover, there is accumulating evidence of an increasingly broader diversity of diazotrophs (Henke et al., 2018; Turk-Kubo et al., 2017). Thus, in light of these recent findings on the distribution of marine N2 fixation, results from new models of global marine N2 fixation have shown increased inputs of fixed N (Landolfi et al., 2018; Wang et al., 2019). Previously, Luo et al. (2012) summarized both the global and regional pelagic N2 fixation rates from ca 45 years’ worth of field data, and concluded that it varied from 0.6-56 Tg N yr-1 across ocean basins, with an estimated 137 Tg N yr-1 globally in the ocean (excluding the Indian Ocean due to lack of data). Another, more recent modeling study, calculated that the global N2 fixation rates equated an average input of 153 Tg N yr-1 (Landolfi et al., 2018). However, Wang et al. (2019) proposed a balanced marine N budget with a global fixed N2 input of 163 Tg N yr-1 (Table 1).

23

Table 1. Estimated N2 fixation rates for each major ocean (Wang et al., 2019), as well as the

total estimates for marine and terrestrial environments. The area of the global oceans

corresponds to the area with hypothesized N2 fixation as reported in Luo et al. (2012). The

table is synthesized from Wang et al. (2019), Luo et al. (2012) and Vitousek et al. (Vitousek et

al., 2013)†.

Region Area (x 106 km2) N2 fixation (Tg N yr-1)

Atlantic Ocean 63 34.0

Pacific Ocean 161 100.6

Indian Ocean 70 26.6

Mediterranean Sea 2.5 0.2

Arctic oceans - 1.8

Global oceans 297 163.2

Terrestrial 149 58†

Despite recent modeling efforts to predict the inputs and outputs of the marine N budget, Landolfi et al. (2018) argued that there are inherent difficulties and uncertainties in making global estimates by model-based approaches. For example, data coverage is low in some areas of the ocean (e.g. Indian Ocean) and a poor understanding of the ecology of marine diazotrophs leads to generalizing input values and/or making many assumptions based on no or low numbers of measurements. One such assumption is that all diazotrophs are equally affected by environmental parameters (Deutsch et al., 2007; Landolfi et al., 2018), while evidence on the contrary is accumulating (Bombar et al., 2011; Church et al., 2005; Langlois et al., 2012; Messer et al., 2016; Moisander et al., 2010; Mulholland et al., 2019; Stenegren et al., 2017, 2018). The ocean is vast and blooms are often short-lived and regional, and therefore routine sampling and measurements required for model efforts are difficult. Thus, N2 fixation is an important process on both global and regional scales, and although models of the marine N budget are seemingly closing in on a balance between N gains and losses, there is still much research needed.

24

1.3 Measuring pelagic N2 fixation rates

N2 fixation rates can be measured either using the 15N2-tracer method, or indirectly by the acetylene reduction assay (ARA). The 15N2-tracer method traditionally added the isotope as a gas bubble to an incubation bottle, but this method has been recently criticized as possibly underestimating N2 fixation. Therefore, Mohr et al. (2010), Großkopf et al. (2012) and Wilson et al. (2012) reported on method comparisons between amending bottles with 15N2 as a gas bubble versus 15N2 enriched seawater. The results showed that the traditional bubble method had underestimated N2 fixation by ca 63 % (Großkopf et al., 2012), and dissolution increased N2 fixation rates 2-6 fold (Wilson et al., 2012). The study by Großkopf et al. (2012) was conducted on a Trichodesmium spp. bloom, hence, underestimation could have greater consequences on smaller (e.g. unicellular) diazotrophs that did not occupy the surface and the immediate vicinity of the 15N2 bubble. Moreover, White et al. (2012) argued that the inconsistencies in sampling procedures when measuring N2 fixation rates add to the uncertainties, since the rates are highly dependent on the diazotroph and time of day the incubation was performed. The reason being that the time of day that the incubation starts influences N2 fixation measurements, since dissolution will vary and diazotrophs are active at different times (day vs. night) (Dron et al., 2012; Taniuchi et al., 2012; Thompson et al., 2014; Zehr et al., 2008). If extrapolated, the resulting global pelagic N2 fixation estimate by Großkopf et al. (2012) was increased to 177 ± 8 Tg N yr-1 compared to their earlier estimate of 103 ± 8 Tg N yr-1. Thus, the method of measuring N2 fixation could have a great impact on modelling and estimates of the marine N budget.

25

1.4 Open ocean cyanobacterial diazotrophs

Box 1. The main cyanobacterial diazotrophic lineages of the open ocean euphotic zone.

Trichodesmium spp. UCYN-A UCYN-B UCYN-C Richelia intracellularis

Calothrix rhizosoleniae

Marine diazotrophy has only been identified in a relatively small, but diverse, group of bacteria and archaea: i.e. heterotrophic proteobacteria (α, β, δ and γ), firmicutes, cyanobacteria and euryarcheaeota (Capone, 2001; Zehr et al., 1995, 2003). Cyanobacteria are one of the most widespread and ancient group of microorganisms and some of them are diazotrophs. Cyanobacteria collectively played a key role in Earth’s great oxygenation event and its aftermath (Bekker, 2014; Holland, 2002, 2006; Luo et al., 2018), and are also the origin of the primary chloroplast, which paved the way for modern day photosynthetic plant life (Martin et al., 2002; Raven and Allen, 2003).

Figure 3. Epifluorescent images of the main cyanobacterial diazotrophic lineages of the open

ocean euphotic zone. A. Trichodesmium spp.; B. colonial (left) and single cells (right) of

UCYN-B (Crocosphaera watsonii); C. FISH image of UCYNA-1 symbiosis; D. apical end of a

Rhizosolenia-Richelia symbiosis; E. chain of Hemiaulus-Richelia symbioses; F. chain of

Chaetoceros-Calothrix symbioses. All images are courtesy of R.A. Foster except image C

which is courtesy of N. Musat.

26

Some of the open ocean diazotrophs also form symbioses with diverse eukaryotic hosts. Cyanobacterial diazotrophs in the tropical and subtropical oceanic surface waters (Box 1) (Fig. 3) are highly diverse in several characters, e.g. size, life history and phylogeny (Fig. 4). Previously, diazotrophs, including the cyanobacterial groups, were categorized based on cell diameter (> 10 µm or < 10µm), but with new information on complex interactions and life histories (colonial, symbioses), a simple size based classification is complicated. For example, the unicellular diazotroph C.

watsonii occurs as single cells (2-6 µm in cell diameter), in colonies (20-1000’s µm in colony diameter), and as a symbiont in a chain-forming centric diatom (1000’s µm in host cell diameter) (Caputo et al., 2019; Carpenter and Janson, 2000; Foster et al., 2011). Thus, their size category may change depending on their life history.

Figure 4. The cyanobacterial diazotrophs of the open ocean, and their different life histories.

Note that the dashed lines signify uncertainty regarding the location of the symbiont within

the host cell (UCYN-A, Crocosphaera and Richelia). The figure is reproduced from Thompson

and Zehr (2013) under the CC BY-NC-ND 4.0 license

(https://creativecommons.org/licenses/by-nc-nd/4.0/).

27

There are two genera of filamentous, heterocystous diazotrophs: R.

intracellularis (hereafter Richelia) and C. rhizosoleniae (hereafter Calothrix), which form symbioses with diatom hosts (mainly Rhizosolenia, Hemiaulus and Chaetoceros). They are collectively referred to as DDAs. It should be noted that the aforementioned symbioses between C. watsonii and a diatom (Climacodium frauenfeldianum) are also a part of the DDAs, but were not a focus of this dissertation. The symbiotic host diatoms are unicellular, photosynthetic microalgae, which can occur as chains (Hemiaulus, Chaetoceros), but also as single cells (e.g. Rhizosolenia clevei), and vary greatly in cell diameter (ca 7-250 µm) (Caputo et al., 2019 and references therein). Diatoms are renown for synthesizing intricate and ornate ‘glass houses’ made out of silica (Si) (Armbrust, 2009). In general diatoms are widely distributed in all aquatic habitats, and contribute to nearly half (45%) of global primary production (Field et al., 1998). The DDAs are amongst the few diatoms that thrive in the open ocean, since most live coastally or in the benthos where dissolved nutrients are high.

Three heterocystous symbiotic lineages have been delineated (based on their nifH gene phylogeny) that form symbioses with three different diatom host genera (Foster and Zehr, 2006) (Fig. 5). They are abbreviated het-1 and het-2 (a-b), which refers to Richelia associated with Rhizosolenia spp. and Hemiaulus spp. diatoms respectively, and lastly het-3, which refers to Calothrix rhizosoleniae associated with Chaetoceros compressus diatoms (Church et al., 2005; Foster and Zehr, 2006). Later, the symbiont genomes were sequenced, and correspond to the following draft genomes names: RintRC01 (het-1), RintHH01 (het-2a), RintHM01 (het-2b) and CalSC01 (het-3). Hereafter the genome names will be used when discussing gene expression (Paper IV), while the het nomenclature is reserved to results of nifH quantitative PCR assays (Paper I-IV).

Depending on the host there is a range of DDA symbiont cellular integration. For example, Calothrix attaches externally on the spines of the host diatom C. compressus (Norris, 1961), while Richelia is partially integrated and situated between the frustule and plasmalemma (inner plasma membrane) of host diatom R. clevei (Villareal, 1990), or appear to have penetrated the cytosol of the host H. hauckii (Caputo et al., 2019). In fact, confocal microscopy observations reported Richelia in H. hauckii as being partially and occasionally fully surrounded by chloroplast auto-fluorescence, and located next to the host mitochondria (Caputo et al., 2019). However, most observations were on chemically fixed samples and require further

28

verification. Of the various DDAs, only the Calothrix symbiont of C.

compressus has been brought into culture (strain SC01) and is readily available, albeit without its host (Foster et al., 2010). Both Hemiaulus and Rhizosolenia have been enriched in a symbiotic state, but only short term (Villareal, 1989, 1990).

DDAs are widespread in the tropics and subtropics and often form large blooms at intermediate salinities (30-35 PSU) i.e. riverine influenced regions (Bombar et al., 2011; Foster et al., 2007, 2009) and the open ocean (Church et al., 2005; Harke et al., 2019). The conditions that constrain the blooms are poorly understood, and it is currently unknown why one DDA might bloom while the others are absent or co-occurring at lower densities. Interestingly, Richelia and Calothrix are among the few heterocystous cyanobacteria known to colonize the open ocean, except for rare observations of Anabaena gerdii in the South Pacific (Carpenter and Janson, 2001). In contrast, brackish waters in temperate regions like the Baltic Sea are dominated seasonally by the heterocystous cyanobacteria Nodularia sp., Aphanizomenon sp. and Dolichospermum sp. (Eigemann et al., 2018; Finni et al., 2001; Sivonen et al., 1989).

29

Figure 5. Neighbor joining phylogeny of partial nifH genes (359 bp) derived from the open

ocean diazotrophs. The phylogenetic difference in the nifH gene is also the basis for the

design of diazotroph specific oligonucleotides for the detection of the organisms by

quantitative polymerase chain reaction (qPCR) assays. The sequences within the subclades

have been collapsed. The scale bar represents 0.1 substitutions per site. A jukes-cantor

correction was used, and three Trichodesmium erythraeum IMS101 sequences were used as

the outgroup. Bootstrap values greater than 50% (from 10000 bootstrap replicates) are

shown at the nodes.

30

Another filamentous diazotroph is the colonial non-heterocystous Trichodesmium spp., which is the most abundant and widespread of the N2-fixers in the tropics and subtropics. It can also form surface ‘slicks’ of aggregating colonies easily visible to the naked eye and by satellites (Dupouy et al., 2011), and has some of the highest observed N2 fixation rates among the marine diazotrophs (Capone et al., 1997; Luo et al., 2012 and references therein). Moreover, Trichodesmium spp. is also considered a holobiont, as it is associated with a community of heterotrophic bacteria (Frischkorn et al., 2017; Hewson et al., 2009; Paerly et al., 1989). A recent study also identified a strain of Calothrix associated with Trichodesmium spp., which was phylogenetically different (nifH gene) from the one (Calothrix SC01; het-3) associated with diatoms (Momper et al., 2015).

Lastly the unicellular group (UCYN) of diazotrophs includes the Candidatus Atelocyanobacterium Thalassa (hereafter UCYN-A), UCYN-B and UCYN-C. The UCYN-A currently has six known sub-lineages (Turk-Kubo et al., 2017), and possibly two more (Henke et al., 2018). Three of these sub-lineages (UCYNA1-A3) are known to be symbiotic with a small microalgae closely related to the prymnesiophyte Braarudosphaera

bigelowii (Cornejo‐Castillo et al., 2019; Hagino et al., 2013; Thompson et al., 2012) which is a small mixotrophic unicellular eukaryote. UCYN-A symbionts have been observed internally, loosely attached or embedded within a symbiosome-like compartment with several UCYN-A cells enclosed by a common envelope (Cornejo-Castillo et al., 2016; Hagino et al., 2013; Thompson et al., 2012). UCYN-A has been observed in a wide geographical range. For example, in addition to the tropics and subtropics, UCYN-A has been detected in temperate oceans (Bentzon-Tilia et al., 2015; Moisander et al., 2010; Needoba et al., 2007), estuaries (Messer et al., 2015), and most recently in arctic marine waters (Harding et al., 2018; Shiozaki et al., 2018). Thus, identification of UCYN-A in these other locations changed the canonical paradigm that marine N2 fixation is largely restricted to the euphotic zone of the subtropical and tropical open oceans.

Based on several genetic markers, e.g. nifH for nitrogenase, 16S rRNA, and the internal transcribed spacer (ITS), the region between the small subunit ribosomal RNA (rRNA) and large subunit rRNA, UCYN-B is most phylogenetically similar to C. watsonii (Webb et al., 2009; Zehr et al., 2007). It is often co-occurring with Trichodesmium spp., where it has been found to be abundant in the euphotic zone (Foster et al., 2007; Moisander et al., 2010; Robidart et al., 2014; Wilson et al., 2017). Six strains of C.

31

watsonii have been isolated, and genome sequenced. These strains demonstrate tremendous phenotypic diversity: e.g. cell sizes, temperature optima, growth rates, and ability to form colonies (Webb et al., 2009). Genomic comparisons of the strains identified an unusual low genetic diversity despite large differences in phenotype (Bench et al., 2013; Zehr et al., 2007). Instead genetic diversity is maintained through genetic rearrangements by mobile genetic elements (transposases) (Bench et al., 2011, 2013). Bench et al. (2016) designed qPCR oligonucleotides to target two different phenotypes (small and large cell diameter sizes) in the field. The small cell size phenotype dominated in the NEP, while the small and large cell size phenotypes were nearly equal in the western tropical south Pacific (WTSP).

UCYN-C is similar (90%) in the nifH sequence (359 bp) to the unicellular diazotrophic cyanobacteria Cyanothece ATCC51142 (Foster et al., 2009), and has been less studied than the other diazotrophs. It appears to have a broad geographic range, and has previously been reported in e.g. riverine influenced regions (Bombar et al., 2011; Foster et al., 2007), temperate waters (Needoba et al., 2007), and recently in a mesocosm study in a coastal region of the southwest Pacific (Turk-Kubo et al., 2015).

32

1.5 Open ocean non-cyanobacterial diazotrophs

A group of marine diazotrophs that have recently gained increased attention are the non-cyanobacterial diazotrophs (NCDs). NCDs are composed of N2 fixing heterotrophic bacterial populations, and to a lesser extent archaeal diazotrophs. The heterotrophic diazotrophs are mainly comprised of proteobacteria (α, β, δ and γ) (Farnelid et al., 2011). N2 fixation rates by NCDs are low compared to the cyanobacterial diazotrophs (Moisander et al., 2017). However, given their potential for occupying vast regions and deeper depths where cyanobacterial diazotrophs are less abundant or absent (e.g. below the euphotic zone, at higher latitudes, in waters with high concentrations of DIN and inside nutrient depleted oceanic gyres) (Farnelid et al., 2011; Riemann et al., 2010; Shiozaki et al., 2014; Wu et al., 2019), depth integrated estimates of their N2 fixation rates suggest NCDs are important contributors to the marine N cycle (Moisander et al., 2017). Regionally, e.g. the South Pacific gyre (Bonnet et al., 2008; Halm et al., 2012) and the Indian Ocean (Shiozaki et al., 2014; Wu et al., 2019), rates of N2 fixation by NCDs were considered significant. However, they are also broadly distributed in the euphotic zone of the tropics and subtropics together with the cyanobacterial diazotrophs (Bonnet et al., 2015; Farnelid et al., 2011; Langlois et al., 2015). Moreover, in oxygen deficient zones (ODZ), spanning both the sub-euphotic and euphotic zones, much of the fixed N can be attributed to heterotrophic diazotrophs (Bonnet et al., 2013; Fernandez et al., 2011; Moisander et al., 2017). It is tempting to assume that N2 fixation in ODZ’s is an effective strategy for minimizing the negative influence of O2 on the nitrogenase complex, however, studies on heterotrophic diazotrophic activity are variable in terms of ambient O2 concentrations and show no clear pattern (Moisander et al., 2017). Moreover, ODZ’s are generally regions of denitrification, meaning that concentrations of DIN are high, making N2 fixation superfluous (Gaye et al., 2013). More research is needed to understand the possible correlation between heterotrophic diazotrophs and O2 concentration, as well as the influences of other biotic and abiotic factors on heterotrophic diazotroph ecology.

N2 fixation by archaea has only been identified in some species of the methanogenic lineage of the euryarchaeota (Leigh, 2000). However, despite observations of significant fractions of euryarchaeota in the euphotic zone (Karner et al., 2001; Massana et al., 1997), the diazotrophic activity and contribution to regional diazotrophy by methanogens in the euphotic zone has rarely been reported (Bombar et al., 2016; Man‐Aharonovich et al., 2007

33

and references therein). Similar to NCDs, much is still to be learned about N2 fixing archaea populations.

34

1.6 Differing N2 fixation strategies

The nitrogenase complex is highly sensitive to O2, due to the irreversible oxidation of its Fe-S clusters (Eady and Leigh, 1994). Exposure to O2 will rapidly decrease N2 fixation efficiency (Stewart, 1969 and references therein), and paradoxically, many of the open ocean diazotrophs residing in the euphotic zone are cyanobacteria, and therefore photoautotrophs, which evolve O2 as part of photosynthesis. The one known exception is UCYN-A, which have lost genes for photosystem II (PSII), the system where O2 evolves, and subsequently the ability to photosynthesize (Zehr et al., 2008) The lineages of UCYN-A that are symbiotic have hosts that perform oxygenic photosynthesis, and therefore evolve O2. However, it is not known how symbiotic UCYN-A combats the O2. To counteract O2 and its negative impact on the nitrogenase enzyme, other diazotrophs have evolved different strategies. For example, either by temporally or spatially separating N2 fixation from photosynthesis (Berman-Frank et al., 2001; Haselkorn, 1978; Mitsui et al., 1986). Moreover, genes for hopanoid lipids, known for barriers for O2 diffusion, have recently been found in all major non-heterocystous cyanobacterial diazotrophs of the open ocean, including UCYN-A, presumably functioning to decrease O2 penetration to the nitrogenase enzyme (Cornejo-Castillo and Zehr, 2019).

Heterocystous, filamentous cyanobacteria apply a spatial separation of nitrogenase by localizing N2 fixation to a thick-walled heterocyst with decreased O2 permeability (Haselkorn, 1978), and the O2 evolving photosynthesis step to the vegetative cells. In contrast, some unicellular cyanobacteria (e.g. Crocosphaera strains) temporally separate N2 fixation by fixing N2 at night and photosynthesizing during the day (Dron et al., 2012; Taniuchi et al., 2012). However, UCYN-A appears to fix N2 during daytime (Thompson et al., 2014; Zehr et al., 2008). This temporal shift in UCYN-A is proposed to be possible due to the symbiosis. Furthermore, the prymnesiophyte host supplies carbohydrates, and UCYN-A applies a cyclic photophosphorylation with photosystem I (PSI) (Muñoz-Marín et al., 2019).

The filamentous colonial cyanobacteria, Trichodesmium spp., have evolved a unique N2 fixation strategy by using both temporal and spatial separation. The latter is made possible by a proposed use of localized regions of the filaments composed of several cells, which were named diazocytes (Fredriksson and Bergman, 1997). Since diazocytes, in contrast to heterocysts, contain both active PSI and PSII (Fredriksson and Bergman, 1997), Trichodesmium spp. combine a strict diurnal cycle of N and C

35

fixation, coupled with a Mehler reaction to minimize O2 (Berman-Frank et al., 2001). Little is known about the N2 fixation strategies of the heterotrophic diazotrophs, but it has been suggested that they could use O2 deficient microhabitats by associating with organic matter where dissolved organic matter (DOM) could fuel their N2 fixation (Moisander et al., 2017; Thiele et al., 2015). Moreover, they could also reside in areas of low O2, such as the euphotic and sub-euphotic ODZ’s.

36

1.7 A symbiotic life history of DDA and UCYN-A

Symbiosis is a relatively old concept, introduced by deBary (deBary, 1879), where he defined symbiosis as two (or more) differently named organisms living together. Today the concept of symbiosis involves interspecies benefits and includes mutualism, commensalism, and parasitism. Among the diazotroph-eukaryote symbioses (where the diazotroph is the symbiont and the eukaryote is the host), many are believed to be mutualistic, where a transfer of nutrients (e.g. N and/or C) makes up the basis of the partnerships (Foster et al., 2011; Foster and Zehr, 2019; Thompson et al., 2012). In the open ocean, symbioses between diazotrophs and eukaryotic phytoplankton are highly specific (Foster and Zehr, 2006; Janson et al., 1999; Thompson et al., 2014) and in some cases potentially obligatory for the symbiont (Hilton et al., 2013; Thompson et al., 2012; Thompson and Zehr, 2013). These partnerships have given the eukaryotic host distinct advantages in colonizing the nutrient scarce open ocean, however, the potential benefit(s) to the cyanobacterial symbiont remain uncertain.

Genome reduction, gene loss and decreased GC content are common characters in symbionts (Moran et al., 2008), and in several planktonic cyanobacteria-eukaryote symbioses of the open ocean, genome reduction and streamlining has been observed (Hilton et al., 2013; Nakayama et al., 2014; Tripp et al., 2010; Zehr et al., 2008). The relatively small sizes for some of the symbiotic diazotroph genomes, together with the streamlining of metabolic pathways (e.g. N assimilation and acquisition in Richelia and PSII, tricarboxylic acid cycle, and several amino acids and purines in UCYN-A) (Table 2), give an idea of the progression of the symbiosis in an evolutionary perspective. The genome size and content for the symbionts of DDAs is directly linked to the symbiont cellular location (Table 2) (Caputo et al., 2018; Hilton et al., 2013). For example, the external symbiont Calothrix has the largest genome (CalSC01, 5.97 Mbp), followed by the partially integrated Richelia (RintRC01, 5.4 Mbp), and the most internal symbiont, Richelia, has the smallest genome (RintHM01 and RintHH01, 2.21 and 3.24, Mbp respectively). Moreover, all of the DDA symbiont genomes are smaller than e.g. the free-living heterocystous diazotroph Anabaena variabilis ATCC 29413 (7.1 Mbp) (Thiel et al., 2014). The UCYN-A genome has evolved more, and is more reduced (1.4-1.8 Mbp) than the DDA symbionts (Bombar et al., 2014; Caputo et al., 2019; Cornejo-Castillo et al., 2016; Tripp et al., 2010, p.2010; Zehr et al., 2008). Furthermore, cellular entities, referred to as spheroid bodies, that have N2 fixation activity and resemble

37

cyanobacteria in genome content and sequence similarity, have been observed in freshwater diatoms of the genera Rhopalodia and Epithemia (Kneip et al., 2007; Nakayama et al., 2014; Trapp et al., 2012). These spheroid bodies are similar to UCYN-A in genome size (2.6-2.79 Mbp) and genome content that has been lost (Kneip et al., 2007; Nakayama et al., 2014).

Table 2. Summary of genomic and life history information of the DDAs compared to the

older and more streamlined UCYN-A1 and A2 symbionts. Reproduced from Caputo et al.

(2018) (Paper III) under the CC BY 4.0 license

(https://creativecommons.org/licenses/by/4.0/).

38

Compared to their free-living counterparts, there are likely trade-offs to a symbiotic life-history. The genome reduction and loss of genomic content observed in Richelia (RintHH01, RintHM01 and RintRC01) and UCYN-A symbionts (Table 2) impacts their dependency on the host. For example, UCYN-A lost the ability to photosynthesize, while the Richelia strains (RintHH01, RintHM01) lost several N acquisition pathways resulting in N2 fixation as the primary means to acquire bioavailable N (Hilton et al., 2013). Moreover, the GC content seen in the DDA symbiont genomes is decreased (34-39 %) (Hilton et al., 2013) compared to e.g. the free-living heterocystous Anabaena variabilis ATCC 29413 (41 %) (Thiel et al., 2014). Furthermore, the symbiont UCYN-A lacks e.g. ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCo), tricarboxylic acid (TCA) cycle, several amino acid synthesis pathways and PS II, rendering them incapable of fixing their own C (Cabello et al., 2016; Cornejo-Castillo et al., 2016; Farnelid et al., 2016; Zehr et al., 2008), and resulting in a host dependency for carbohydrates.

39

1.8 Environmental constraints on marine diazotrophs

1.8.1 Nutrients

As previously mentioned N2 fixation fertilizes the environment by reducing N2 gas into bioavailable ammonia, and has been estimated to contribute more than 50 % of the N demand for the surrounding phytoplankton community (Capone et al., 2005; Karl et al., 1997). However, N2 fixation is largely constrained by the ambient environment, both biotic (e.g. other plankton) and abiotic (e.g. nutrients and temperature) factors. Some oceanic regions are identified as conducive to N2 fixation based on their dissolved inorganic nutrient concentrations (Bonnet et al., 2017; Messer et al., 2016; Moutin et al., 2017), which would include low to immeasurable concentration of DIN and elevated dissolved inorganic phosphorous (DIP) and high Fe deposition.

In general, dissolved nutrients (e.g. N, phosphorous (P), Fe and Si) are added to the euphotic zone through 1) hydrographic processes (advection, upwelling and eddy diffusion) that transport re-mineralized nutrients from the deep ocean, 2) terrestrial inputs (rivers, land run-off and sewage) (Eppley and Peterson, 1979), and 3) atmospheric deposition. In most cases a mix of the dissolved nutrients are transported simultaneously to the surface ocean, close to the Redfield equilibrium (C:N:P = 106:16:1) (Eppley and Peterson, 1979). Regionally, atmospheric dust deposition also plays an important role in supplying e.g. Fe and P to the surface ocean. For example, approximately 0.11 % of Saharan dust is P and ca 3.5 % Fe, making dust deposition an important stimulant for surface N2 fixation (Ridame and Guieu, 2002). One prime example is the seasonal input of Saharan dust to the tropical and subtropical North Atlantic Ocean (Chen and Siefert, 2004; Langlois et al., 2012; Pabortsava et al., 2017).

The nitrogenase enzyme has, due to its composition, a high demand for Fe which is therefore considered the major limiting nutrient for diazotrophy (Chappell and Webb, 2010; Kustka et al., 2003; Raven, 1988). In addition, photosynthesis requires Fe for both PSI and PSII complexes, as well as in Fe-S and cytochrome proteins (Raven, 1990; Sunda and Huntsman, 2015), making photoautotrophic diazotroph Fe requirements even higher. Heterotrophic diazotrophs do not photosynthesize and would therefore have a lower Fe requirement than their photoautotrophic counterparts. The lower Fe requirement is a proposed reason for the dominance of heterotrophic diazotrophs in ultra-oligotrophic ocean regions like the South Pacific gyre, where Fe input is low and dissolved organic carbon (DOC) concentration is high (Halm et al., 2012). Moreover, the lack of atmospheric dust deposition

40

is one of the proposed reasons why the eastern South Pacific Ocean (Bonnet et al., 2008) and southern Atlantic Ocean (Langlois et al., 2012) is lacking in terms of marine N2 fixation. However, there are more inconspicuous sources of Fe in marine environments, like hydrothermal activity (Tagliabue et al., 2010) and animal feces (Lavery et al., 2010) that can be important on local and regional scales.

The strategies for dealing with Fe limitation, although not well known, differ between phytoplankton, including diazotrophs (Morrissey and Bowler, 2012). For example, Fe depleted conditions induce colony formation of Trichodesmium spp. (Tzubari et al., 2018), whose colonies have previously been shown to actively and directly mobilize and utilize dust particles (Rubin et al., 2011), and Shi et al. (2007) have suggested that Trichodesmium spp. prioritize photosynthesis over N2 fixation due to a more rapid down-regulation of N2 fixation genes in response to Fe deficiency. Among the unicellular diazotrophic cyanobacteria both Crocosphaera and Cyanothece have similar patterns of increased cellular Fe, coupled with an up-regulation of Fe uptake genes in the evening, linking Fe uptake with N2 fixation, which occurs at night (Saito et al., 2011; Shi et al., 2010; Stöckel et al., 2008). Furthermore, the bacterioferritin protein, which is a bacterial Fe storage protein (Rivera, 2017), peak at times of maximum Fe utilization by N2 fixation and photosynthesis (Morrissey and Bowler, 2012). Combined, it suggests that Crocosphaera minimizes its Fe quota by transferring Fe between the N2 fixation and photosynthesis systems (Morrissey and Bowler, 2012). Little is known about the Fe limitation strategies in DDAs, however, being in symbiosis with diatoms, which in general have a rapid response to Fe availability (Marchetti et al., 2012; Morrissey and Bowler, 2012) may have its advantages. Moreover, both Richelia and Calothrix possess the Fur gene for regulation of Fe uptake, while Calothrix also possess tonB transport and a complete siderophore pathway for biosynthesis and active transport of Fe3+-binding siderophores (Nieves-Morión, unpubl.). In light of the varying DDA symbiont cellular integration with their hosts, mediation, competition, and possible coordination of Fe acquisition between the partners is an interesting and still open question.

Another important nutrient for diazotrophs is P, which has been proposed to be a co-limiting nutrient of N2 fixation together with Fe in marine environments (Mills et al., 2004), and is the primary limiting factor of diazotrophy in Fe replete conditions (Sañudo-Wilhelmy et al., 2001). Moreover, it has been suggested to drive Fe limitation on diazotrophy (e.g.

41

in the SCS), and subsequent N limitation on the phytoplankton community (Wu et al., 2003). It has been shown in laboratory cultures of Trichodesmium and Crocosphaera that the co-limitation of P and Fe increases both growth and N2 fixation rates compared to either replete or single nutrient limited conditions (Garcia et al., 2015). A later culture study on Trichodesmium showed that some of the characteristic genes up-regulated during Fe-P co-limitation included e.g. Fe stress (isiB), P transport (sphX) and cell morphology (ezrA). The ezrA gene has previously been linked to cell size (Jorge et al., 2011), suggesting that regulation of cell size (smaller cells, larger cellular surface to volume ratio), and gene expression levels in Trichodesmium is an adaptation to chronic nutrient co-limited conditions of the oligotrophic open oceans (Walworth et al., 2016).

One of the main arguments for the importance of P in marine ecosystems is the widespread prevalence of P uptake genes, even among the smallest and most streamlined prokaryotic genomes (Dyhrman et al., 2007). However, a majority of all marine P is in the form of dissolved organic P (DOP) (Clark et al., 1999), which may only be utilized by phytoplankton that possess alkaline phosphatase, phosphoesterase or metallophosphoesterase (Dyhrman and Haley, 2006). These three enzymes catalyze the hydrolysis of P monoesters, which constitutes ca 80 % of all marine DOP (Young and Ingall, 2010). A wide range of phytoplankton has the ability to utilize DOP via P hydrolyzing enzymes (Dyhrman and Haley, 2006), including three of the DDA symbionts (Fig. 6), which could give them an advantage in DIP limited conditions.

42

Figure 6. Presence or absence of P scavenging and acquisition genes for the four DDA

symbiont strains Richelia-Hemiaulus hauckii (R-HH), Richelia-Hemiaulus membranaceus (R-

HM), Richelia-Rhizosolenia clevei (R-RC) and Calothrix-Chaetoceros (C-C), and the unicellular

diazotroph Crocosphaera watsonii (CW). A colored gene signifies its presence in the

corresponding genome.

R-HH R-HM R-RC C-C CW Gene Function

ppa Inorganic pyrophosphatasepitA Inorganic phosphate transporterphoU Phosphate transport system regulatorphoR Phosphate regulon sensor proteinphoB Phosphate regulon transcriptional regulatory proteinphoH Phosphate starvation-inducible proteinpstB Phosphate transport ATP-binding proteinpstA Phosphate transport system permease proteinpstC Phosphate transport system permease proteinpstS Phosphate ABC transporter, periplasmic phosphate-binding proteinppk Polyphosphate kinaseppx ExopolyphosphatasephnE Phosphonate ABC transporter permeasephoD Phosphodiesterase/alkaline phosphatase DphoA Alkaline phosphataseypbG Phosphoesterasempe Metallophosphoesterase

DOP hydrolysis and transport

43

Molybdenum (Mo) is a trace metal highly relevant for N2 fixation since it is a central part of the Fe-Mo nitrogenase enzyme. Although needed in much less quantities than Fe, it has been proposed to be a possible limiting factor of marine diazotrophy due to lower availability in some regions (Howarth and Cole, 1985). Molybdenum in marine environments can reach concentrations ca 100 times higher than in freshwater systems. The decreased availability in some habitats could be related to an active competition in uptake between the ions molybdate and sulfate, which has a lower concentration in freshwater. However, marine diazotrophs are unaffected (Paulsen et al., 1991), and their activity is not directly linked to environmental Mo concentrations. One plausible explanation is that marine diazotrophs possess’ genes for highly specific uptake of Mo via a homolog of the modEABC Mo transport system first identified in the heterotrophic soil diazotroph Azenobacter vinelandii (Tuit et al., 2004), which has been found in both Trichodesmium spp. and the DDA symbionts.

Si is an important nutrient for the diatom hosts of the DDAs, since they incorporate Si by building intricate cell walls, called frustules. Much of the production of biogenic Si in the oceans is by diatoms (Tréguer and De La Rocha, 2013), and therefore, the DDAs are at a disadvantage in regions of low Si inputs and concentrations. However, there are plenty of sources of Si to the marine environment, the most prominent being river discharges (Conley, 1997; Tréguer and De La Rocha, 2013), which is likely one of the reasons why the highest abundances of DDAs have been observed in regions heavily influenced by river discharges, e.g. the Amazon, Orinoco and Mekong river plumes (Bombar et al., 2011; Carpenter et al., 1999; Foster et al., 2007).

44

1.8.2 Hydrographic parameters

Temperature is often proposed to be one of the most important parameters governing the distribution and activity of marine diazotrophs (Chen et al., 2019; Messer et al., 2016; Moisander et al., 2010; Stenegren et al., 2018). However, it is not the only relevant hydrographic parameter. Depending on spatial and temporal variability on both small and large scales photosynthetically active radiation (PAR), salinity and depth can also play important roles in shaping diazotroph distributions in pelagic ecosystems.

Variation in gene expression levels for certain metabolic pathways based on daylight has recently been demonstrated for one DDA strain (Harke et al., 2019), and it is well known that some unicellular photoautotrophic diazotrophs (i.e. UCYN-B and UCYN-C) regulate their N2 fixation and photosynthesis based on daylight, or the absence thereof (Dron et al., 2012; Taniuchi et al., 2012). Although many studies have observed decreasing abundances of diazotrophs as a function of decreasing PAR with depth (Church et al., 2005; Foster et al., 2007; Goebel et al., 2010; Kong et al., 2011; Langlois et al., 2008; Moisander et al., 2008, 2010; Stenegren et al., 2018), less is known about the regulation of e.g. nitrogenase transcription with decreasing PAR, due to increasing depth or turbidity. In terms of phytoplankton abundance, decreasing PAR due to increasing turbidity has been observed to have a negative impact on abundance of the greater phytoplankton community in the Amazon River plume (Goes et al., 2014). Moreover, transcription of the nifH gene (nitrogenase Fe protein) in particular has been observed to decrease with depth for one DDA symbiont strain in the open ocean of the NEP (Church et al., 2005).

Salinity is a less obvious predictor of diazotroph abundance and activity since most of the open ocean euphotic zone has a uniform salinity regime (> 35 PSU). The exception is in regions near freshwater discharge from large rivers (e.g. Amazon, Niger and Mekong) where decreased salinity can be measured far from the river mouth (Borstad, 1982; Muller-Karger et al., 1988). In these regions salinity shapes the mosaic and spatial succession of e.g. phytoplankton where coastal species tend to dominate inside the river plume, and open ocean species next to or underneath the river plume (Goes et al., 2014). Moreover, Foster et al. (2007) observed a succession of the phytoplankton community from non-symbiotic phytoplankton, to symbiotic diazotrophs, to free-living diazotrophs along a salinity (and nutrient) gradient (ca. 24-36 PSU) of the Amazon River plume.

45

2 Thesis aims It is well known that diazotrophs are a small but diverse group of microbes, with a wide range in life histories, N2 fixation strategies, distributions and N2 fixation rates. However, little is known about the environmental conditions that influences N2 fixers and their activity. Therefore, an overarching aim of this thesis was to investigate cyanobacterial diazotroph distributions, abundances and gene transcriptions in relation to the different environmental conditions. Here, gene expression is used as an indirect proxy for activity and the thesis largely focused on the DDAs. Increased knowledge of what environmental parameters govern diazotroph abundance and activity, as well as how the diazotrophs are affected, will help us better understand ecosystem dynamics and the role of the diazotrophs in marine biogeochemical cycles. Moreover, it will help in modeling of marine N and C budgets over both spatial and temporal scales.

Specifically, the aims of this thesis were to determine the distribution and abundance patterns of cyanobacterial diazotrophs in various habitats (Paper I, II, IV), including a comparison of regional patterns with other ocean basins (Paper I), and a statistical model for predicting parameter influence (Paper II, IV). Determine which ambient environmental parameters and conditions potentially select for the occurrence of certain diazotrophs, and how the environment impacts abundances of diazotrophic groups (Paper I, II, IV). The thesis further highlights a disparity in research attention in the field of N2 fixation for both abundance and activity measurements of DDAs, and provided a brief comparison of the two main N2 fixing planktonic symbioses: DDAs and prymnesiophyte-UCYN-A (Paper III). Lastly, a DDA symbiont specific microarray was developed and applied to determine strain similarities and differences in gene expression levels, and how environmental parameters influence the transcription of select functional genes (Paper IV).

46

3 Comments on methodology

3.1 Sampling for diazotrophs

Due to the relatively remote occurrence of the DDAs, sampling for them was conducted during three 4-6 week-long scientific cruises in the open ocean (Fig. 7). All three regions of sampling were located within the subtropical-tropical belt (30°N-30°S): WTSP (Paper I), WTNA (Paper II) and SCS (Paper IV). The sampling regions of the WTNA and SCS were located near large river discharges (Amazon and Orinoco, and Mekong, respectively) and are therefore dynamic oceanic regions heavily influenced by seasonal freshwater discharges. The sampling region of the WTSP included parts of the Melanesian archipelago and subtropical gyre, which are regions characterized as distinctly oligotrophic to ultra-oligotrophic with little variation in sea surface nutrient concentrations, temperature, and salinity.

47

48

Figure 7. Integrated maps of sampling regions on the inset images, showing the regional

surface area of chlorophyll a concentrations (mg m-3

) during the cruises as two-month

average composites. The red and green circles are all the sampled stations, while RNA for

microarray analyses were from stations marked with green circles. The oceanic regions are

the Western Tropical North Atlantic (WTNA; sampled May-June, 2010), the South China Sea

(SCS; sampled June, 2016) and the Western Tropical South Pacific (WTSP; sampled February-

March, 2015).

At sea, three different sampling strategies were used and dependent on the intended lab method. The most frequently used method of sampling was with niskin bottles (12 L each) mounted on a conductivity, temperature, depth (CTD) rosette (Fig. 8). The CTD is capable of measuring a number of hydrographic parameters during deployment, allowing for the in situ overview of the conditions in the water column. At a given depth, individual niskin bottles can be closed, returned to the surface, and the water from that depth retrieved for further processing on board the research vessel.

Figure 8. CTD rosette with 24 mounted 12L niskin bottles being deployed at the starboard

side of the R/V L’Atalante during the OUTPACE 2015 cruise.

49

Other, simpler ways of sampling used, included a bucket lowered to the surface with a rope, or on some research vessels, water from a sub-surface underway system. The bucket is easy to deploy, and may retrieve relatively large volumes of water in a short time. It also samples from the very surface in which the CTD rosette mechanically cannot sample. On the other hand it is more prone to contaminations and degradation due to its close proximity to the hull of the research vessel and the more direct handling of the sample water itself. Furthermore, the underway pump is, like the bucket, a strategy with virtually unlimited amount of sample water, however, the underway system is usually fitted with a pre-screen (> 200 µm mesh) to exclude larger plankton and debris. Screens are also often subject to biofouling. The pre-screen is also an issue in the case of e.g. the DDAs, Trichodesmium spp. or particle associated diazotrophs, since larger chains, colonies or aggregates of cells and particles might be excluded, or compromised.

After retrieving the sample water it was filtered through different diameter and pore sizes (from 0.2 µm to 10 µm), depending on the downstream application, using a peristaltic pump. Mostly, we used 25 mm and 47 mm diameter filters for collecting material for microscopy and nucleic acids. The filter diameter and pore size depended on the plankton density and volume filtered. When filtering DNA/RNA samples the pressure of the peristaltic pump had to be low enough not to compromise cell integrity, but high enough to collect samples within 40 min, hopefully minimizing nucleic acid degradation and potential changes in transcription.

50

3.2 Quantifying diazotrophs and their gene expression levels

The most straightforward method of quantifying the relative abundance of diazotrophs in the environment is qPCR. It was also the most applied method in this thesis. The basis of the TaqMan qPCR used in this thesis is a pre-designed set of short (16-32 bp each) oligonucleotides (reverse/forward primers and a marker probe) that were applied to the nucleic acids from the samples and an 8-point standard curve of the target at known concentrations. The standards were made from plasmids containing the target or commercially synthesized in G-blocks. A fitted line to the amplification signal of standards allows us to estimate the abundance in the samples. In our case the oligonucleotides are designed to target (and enumerate) a small (49-83) base pair (bp) fragment of a conserved region of the nifH gene (359 bp), (Paper I and II, IV). In addition, two sets of oligonucleotides, including a newly designed set, were applied to target the18S rRNA gene of the UCYN-A1 and UCYN-A2 hosts (Paper I; Table 3). The samples and assays were run on a StepOnePlus (Applied Biosystems, Life Technologies; Stockholm, Sweden) qPCR instrument, which has the possibility of a ‘fast mode’ with a run time of ca 40 min (compared to the normal ca 2 h), allowing for rapid generation of relative abundance data.

Table 3. Summary of the TaqMan oligonucleotides used in this thesis.

51

52

In this thesis, activity was quantified by measuring gene transcription (the first step in assembling proteins and enzymes for metabolic activities) using a microarray approach. The recent cost effectiveness and improvements in high throughput sequencing have led to an increasing number of partial and fully sequenced microbial genomes. This new information opens up new opportunities for targeted microarray analyses since they are based on known sequences (Shilova et al., 2016). Subsequently, the microarray technique can be designed to quantify transcription of whole or specific pathways in genomes. For example, a few recent studies on diazotrophs (Hewson et al., 2007; Muñoz-Marín et al., 2019; Shi et al., 2010; Paper IV), the nifH gene of diazotrophic communities (Moisander et al., 2006) or a wide range of functional genes in the pelagic microbial community (Robidart et al., 2018; Shilova et al., 2014, 2016), have used mRNA based microarrays. Given the nature and quality of our samples, microarrays are more advantageous than total RNA sequencing (e.g. RNA seq) since it allows for detection of low abundance transcripts of species that are relatively less abundant in the ambient and mixed microbial community. Moreover, by designing individual probes for each gene it is possible to distinguish between closely related strains and target a subset of functional genes simultaneously. Before the total RNA extracts could be processed they were quality controlled for degradation using the RNA Integrity Number (RIN). RIN is an objective algorithm based on the entire electrophoresis output, including rRNA ratios (28S/18S for eukaryotes and 23S/16S for prokaryotes). The RIN value is graded on a scale of 1-10, where 10 is the best quality (most intact RNA) (Mueller et al., 2004). However, due the lower relative abundance of the DDAs in our samples, there were difficulties in collecting enough RNA of adequate quality for the microarray.

Much of the preparatory work in designing an organism specific microarray was controlling for potential cross-hybridization to closely related strains (e.g. between the DDA symbionts) and to the co-occurring environmental populations. We were highly conservative in our approach of design and employed several steps: 1) the manufacturers default algorithm when designing the probes, 2) BLASTn comparisons in the National center for Biotechnology Information (NCBI), including the publically available omics datasets and the three DDA symbiont draft reference genomes, 3) pilot microarray that applied extracted total RNA from the asymbiotic Calothrix strain (CalSC01). All probes with nucleotide (nt) similarities of 95 % or higher over the entire probe region were discarded. Moreover, any probes assigned to the RintRC01, RintHH01 or RintHM01 that hybridized in

53

the pilot experiment with RNA derived from CalSC01 culture were also discarded. The final microarray design had a 30 % decrease in the number of probes (mainly for the RintHH01 and RintHM01), but still contained a wide range of the functional genes initially selected for all strains. Due to the high cross-hybridization between RintHH01 and RintHM01, the latter was not included in any further analysis (Fig. 9).

Figure 9. A summary comparison of the number of probes for the main metabolic pathways

on the final micro-array shown according to symbiotic strain: RintRC01, RintHH01, CalSC01.

Values shown as the number of genes normalized to the draft genome size (Mbp) of the

respective symbiont strains.

54

3.3 Challenges and caveats in detection of diazotrophs

The oceans are vast, and dilute, hence collection of materials is difficult, and like many bacteria, not all cyanobacterial diazotrophs (e.g. Richelia, UCYN-A) have been maintained in long-term culture. Some marine diazotrophs, e.g. Trichodesmium, Crocosphaera, Cyanothece and some proteobacterial strains have been isolated and are available for lab-based studies (Boström et al., 2007; Farnelid et al., 2014; Prufert-Bebout et al., 1993; Reddy et al., 1993; Webb et al., 2009). The DDAs have historically been difficult to culture, and currently the Calothrix symbiont (Calothrix SC01) is the only symbiont successfully isolated and subsequently maintained in culture, although without its diatom host (Foster et al., 2010). However, short-term enrichments have been successful for the Rhizosolenia-Richelia symbiosis (Villareal, 1989, 1990).

Experiments and collection of cells for the uncultivated diazotrophs has to be largely done in the field, where detection is the first key step. To date, the most common method of detecting and estimating abundances of marine diazotrophs has been qPCR, and to some extent also epi-fluorescence microscopy (Luo et al., 2012). However, both approaches have their challenges in application. For example oligonucleotide specificity, potential cross-hybridization, DNA extraction efficiency, multiple gene copy number and possible polyploidy are all factors influencing qPCR. The currently used nifH oligonucleotides were designed over a decade ago with a comparatively smaller database of sequences. This directly impacts our detection of the diversity.

Microscopy is limited to cells with a distinct morphological identity, and with epi-fluorescence applications a known excitation wavelength. For example, it is not possible to identify the heterotrophic diazotrophs by microscopy since they look like any other bacteria. A second example is UCYN-A, which do not possess the typical auto-fluorescence of cyanobacterial phycobiliproteins that is usually used to distinguish them with an epi-fluorescence microscope. Hence, UCYN-A require qPCR or 16S rRNA fluorescent oligonucleotides and Fluorescence In situ Hybridization (FISH) assays for identification, which is not easily implemented in the field.

Moreover, detecting any planktonic population by use of either qPCR or microscopy is challenging due to the natural heterogeneity of the water column. Plankton is notoriously patchy and may vary on scales small

55

enough to affect assumed homogeneity within sample collection (niskin) bottles. It is also possible that settling of material within the niskin happens during the sampling procedure. Lastly, diazotrophs living in symbioses, attached, or colonially are also challenging to collect without compromising the cell integrity. Hence, qPCR has been extensively applied (Luo et al., 2012) and proven to be a great tool for rapid detection of marine diazotrophs, and can even be applied at sea (Robidart et al., 2014; Paper I).

Microarrays are a great tool for measuring gene transcription, which could be an indirect proxy for cellular activity. However, gene expression is not related to metabolic activity. Furthermore, some cellular processes like photosynthesis, cell division and N2 fixation are delayed compared to peak expression levels of involved genes (Chen et al., 1998; Shi et al., 2010). Moreover, metabolic pathways require a suite of genes to be expressed, often in sequence, before the protein or process is active. Thus, despite technological improvements in sampling and detection of open ocean diazotrophs, their abundances and activity are still difficult to accurately quantify.

56

3.4 Statistics as powerful tools of investigation

To further our understanding of diazotrophs in a larger context (e.g. global and regional scales), newly refined multivariate statistical tools were applied on the quantitative data (abundances and gene expression levels). For example, a weighted meta-analysis (Koricheva et al., 2013) allowed us to compare distribution patterns both within and between ocean basins. Moreover, a piecewise Structural Equation Model (pSEM) (Lefcheck, 2016) was fitted to the qPCR datasets and measurements of environmental parameters to predict the influence, including causality, of up to 25 biotic and abiotic parameters simultaneously. The concepts and ideas behind multivariate statistics are not new, but as the statistical methodology developed throughout the 20th century, calculations and plotting was limited by computational power (or the lack thereof). Thus, as our potential explanatory output and interpretation of data grows increasingly sophisticated, the philosophy behind many statistical tests (e.g. correlations) is still up for debate. For example, ’correlation does not imply causation’ is a frequently used statement related to many statistical analyses, but recent multivariate models challenge this paradigm.

One statistical tool used in this thesis (Paper II, IV) is, as mentioned, the pSEM. Was developed by Lefcheck (2016) based on previous SEMs by e.g. Grace (2006). Its application within the field of marine ecology has just begun to be explored (Austin et al., 2017; Duffy et al., 2015; Lefcheck, 2016; Lefcheck and Duffy, 2015), and the potential for modeling complex interactions within dynamic marine ecosystems is great. The idea behind pSEM is the possibility to model several different models (e.g. linear and generalized linear) in a piecewise manner as a single model. Thus, the pSEM can be modified based on the data to be fitted. Lefcheck (2016) argues, as have others, that correlation does imply causation, but the directionality of that correlation is unresolved. One can argue that this may never be true unless we are dealing with simple and closed systems with much background information. However, pSEM is not only a predictive tool, but also useful for generating new testable hypotheses. By constructing a framework (based on previous knowledge) of possible and probable interactions between parameters in the system we can analyze the parameter effects and explore the possibilities of missing interactions (based on statistical significance). With this statistical information, coupled with previous knowledge, we can start to form a new understanding of the system and generate testable hypotheses. For example, unexplained direct positive

57

effects between the DDA symbionts were observed (Paper II and IV), and this interaction could be further explored (to investigate causes).

58

4 Results and discussion Field observations and detections of diazotrophs based on direct measurements are essential for understanding diazotroph ecology and to improve model predictions for their contributions to marine biogeochemical cycles. Many studies in the past decade have expanded our knowledge of the distribution of marine diazotrophs (e.g. Bombar et al., 2011; Bonnet et al., 2015, p.215; Goebel et al., 2010; Harding et al., 2018; Langlois et al., 2012; Moisander et al., 2010; Mulholland et al., 2019; Paper I, II). Furthermore, others have highlighted that this kind of distribution data is still lacking for our greater understanding of marine biogeochemical cycles and the diazotrophs’ role in them (Landolfi et al., 2018; Luo et al., 2012). However, simply knowing distribution and abundance of marine diazotrophs does not inform whether they are active and significantly contributing to regional and global N budgets.

Recent efforts on quantifying the expression levels of functional genes related to diazotrophy (Harke et al., 2019; Robidart et al., 2018; Shilova et al., 2014, 2016); Paper IV) has increased our understanding of the complex cellular response to the ambient environment required for N2 fixation. This thesis has contributed to describing how regional environmental conditions govern the diazotroph abundances, distribution and activity (estimated by gene expression).

59

4.1 Spatial and vertical detections of diazotrophs

One of the major contributions of this thesis was sampling and reporting of diazotrophs in regions not commonly sampled. The WTSP is one such region (Bonnet et al., 2009; Moisander et al., 2010; Paper I). Previous observations in the region have attributed most of the diazotrophy to Trichodesmium spp. (Dupouy et al., 2011), UCYN-A and to a lesser extent UCYN-B (Moisander et al., 2010). The DDA symbionts have been less reported in the WTSP. In fact, the het-3 had never been quantified prior to our work. In our study in the WTSP, eight diazotrophs (het-1, het-2, het-3, UCYN-A1, UCYN-A2, UCYN-B, UCYN-C and Trichodesmium spp.) and the two UCYN-A hosts (UCYN-A1 and A2 hosts) (Paper I) were quantified by qPCR. Despite obvious difficulties in applying qPCR in the field, like a moving vessel and the high risk for contaminations, we successfully extracted DNA for application of a shipboard qPCR on the four unicellular diazotroph strains (UCYN-A1, UCYN-A2, UCYN-B, and UCYN-C). The abundances were estimated within three hours after sample collection and allowed informed decisions for sampling, such as which depths to sample for incubation experiments.

The distribution and abundance patterns were similar to earlier work showing Trichodesmium spp. (Dupouy et al., 2011) as the overall dominant diazotroph, followed by UCYN-B (Moisander et al., 2010). However, unlike earlier studies that reported UCYN-A as one of the most dominant diazotroph (Farnelid et al., 2016; Moisander et al., 2010), we found both strains to be largely undetected (63 and 79 of 120 samples, respectively). The DDA symbionts often co-occurred and were all reported for the first time in the WTSP. The het-1 was the most abundant followed by het-2 and het-3. Congruent with other studies on marine pelagic diazotrophs (Chen et al., 2019; Messer et al., 2015; Moisander et al., 2010), correlation and multivariate redundancy analyses (RDA), identified temperature as the most important environmental parameter driving diazotroph abundance and distribution. In fact, our statistical analyses showed that both temperature and depth separated the diazotrophs into two groups: 1) a deep-dwelling group (UCYN-A with their hosts: > 50m) and 2) a surface-dwelling group (Trichodesmium spp, het-groups and UCYN-B: < 50 m) (Fig. 9). The clear vertical separation of the UCYN-A suggests that conditions favoring UCYN-A (and its partner) are different from the other diazotrophs. It was interesting and unexpected given the diverse life histories (e.g. symbiotic, free-living and colonial) within the surface-dwelling group that these would form a

60

robust group. Thus, many of the other measured environmental parameters (e.g. nutrients) did not seem to influence the abundance of the diazotrophs differently in the open ocean of the WTSP.

The results in the WTSP were combined with 11 additional qPCR studies (four ocean basins) to further investigate environmental influence on diazotroph abundances in the world’s oceans. Here, a weighted meta-analysis was applied (Paper I). Again, temperature was a major environmental parameter governing diazotroph abundance. The one exception was UCYN-A. This result makes sense in light of recent and previous detections of UCYN-A in cooler waters (Harding et al., 2018; Mulholland et al., 2019; Needoba et al., 2007) and the prevalence of UCYN-A in cooler waters at depth in the tropics and subtropics (Moisander et al., 2010; Paper I). Our work contributes evidence of UCYN-A as a cosmopolitan and important marine diazotroph, and also expands the marine waters where N2 fixation can occur. The analyses in the larger dataset also highlighted the global dominance of Trichodesmium spp. with only a couple of regional exceptions, e.g. het-2 in the WTNA (Foster et al., 2007) and the UCYN-A in the North Atlantic (Goebel et al., 2010).

61

Figure 10. Hierarchical cluster analysis based on spearman's rank correlations identified two

separate groups (brackets) with different environmental parameters. The colors signify the

strength of the correlation, either positive (green) or negative (blue). The asterisks mark

significant correlations (p < 0.05). The figure is reproduced from Stenegren et al. (2018)

(Paper I) under the CC BY 3.0 license (https://creativecommons.org/licenses/by/3.0/).

62

In light of potential cross-hybridization in qPCR assays we also conducted experiments with the assays that target the most closely related strains (DDA symbiont strains and UCYN-A1 and A2) using different cycle modes (fast vs. slow) on the qPCR instrument. Similar to earlier works (Foster et al., 2007; Thompson et al., 2014) we found cross-hybridization between the assays. We encountered cross-hybridization between het-1 and het-2 (both modes) if the two strains (based on standard concentrations) should co-occur at abundances > 105 nifH copies L-1, which have rarely been observed in the environment. The UCYN-A1 and A2 strains, despite following suggestions to increase annealing temperature (Thompson et al., 2014), cross-hybridizations were encountered at all standard concentrations when run in fast mode. Moreover, only increased cycle time (slow mode) coupled with an increased annealing temperature negated cross-hybridization. Hence, there is a possibility that studies using the UCYN-A1 and A2 oligonucleotides that do not increase annealing temperature and cycle time, will overestimate the abundance of UCYN-A1 and A2 strains. Moreover, we did not perform cross-hybridization tests against newly discovered strains, e.g. the UCYN-A (3-6), however, they have been proposed to occupy different ecological niches (Turk-Kubo et al., 2017). Abundance is often seen as indicative of significance, and therefore caution is advised when interpreting qPCR data.

63

4.2 Applying piecewise SEM on two DDA symbiont strains

Simple correlations cannot determine the possible environmental cause for the subtle differences in distributions between populations, e.g. the DDA symbiont strains. Applying the multivariate pSEM to qPCR datasets of het-1 and het-2 generated in the WTNA (Paper II) we could start to understand how the environment interacts with the biology (Fig. 11). Blooms of DDAs, e.g. Hemiaulus-Richelia (het-2), are consistently and seasonally observed in the WTNA (Foster et al., 2007; Subramaniam et al., 2008), so we hypothesized that the Amazon River (AR) plume would have a significant impact in shaping DDA occurrences and abundances. We found impacts of the river plume on the het-2, which was mediated directly and indirectly by turbidity. For example, fluorescence had an indirect negative effect on het-2 mediated through a positive effect on turbidity (and turbidity was negatively and directly affecting het-2) (Fig. 11). This estimate effect is visualized by the following example:

Fluorescence Turbidity Het-2

The pSEM also identified that both salinity and DIP had direct positive influence on het-1 abundance, while salinity was indirectly negative on het-2. These results are also consistent with observations of het-2 at intermediate salinities (Foster et al., 2007, 2009; Paper II) and the dominance of het-1 in oceanic salinities (> 36 PSU) like the WTSP (Paper I) and North Pacific gyre (Church et al., 2005; Mague et al., 1974; Venrick, 1974). In paper III, we also found that the draft genome of het-1 (RintRC01) had more genes related to P uptake and DOP hydrolysis than the draft genome of het-2 (RintHH01) (Fig. 6), which could also drive a niche separation between the two closely related symbiont strains. A similar scenario has been reported for Trichodesmium spp. and UCYN-B, although these are not phylogenetically similar (Dyhrman et al., 2006).

Building upon the pSEM framework from the WTNA we subsequently applied an expanded model on qPCR data from the SCS and the Mekong river plume (Paper IV). Since the freshwater discharge of the Mekong River was at a record low at the time of sampling (Nguyen, 2017), parameters directly related to the river plume (e.g. nutrients and turbidity) were expected to have minimal impact on the diazotrophs. This is in contrast

64

to the conditions in the WTNA and AR plume, and as expected we found contrasting environmental impacts in our pSEM output. None of the environmental parameters associated with the river plume (e.g. DIP, DIN, beam transmission) had an effect on the DDA symbionts, while depth and salinity both had a negative impact. This is congruent with the observations in the open ocean of the WTSP (Paper I). Similar to the pSEM results from the WTNA (Paper II), interactions between the DDA symbionts were positive and direct. Thus, there is a re-occurring strong positive interaction amongst the DDA symbionts, which could shift based on abundance.

Figure 11. Part of the pSEM showing the environmental parameter effects on the DDA

symbiont strains during the 2010 cruise in the WTNA. Note that only the closest indirect

effects on the DDA symbionts are shown. The arrows represent the direction of effect, either

positive (green) or negative (red), and the values assigned to each arrow is the strength of

the effect. The values in parentheses are the indirect effect on either DDA symbiont strain.

The figure is reproduced from Stenegren et al. (2017) (Paper II) under the CC BY 4.0 license

(https://creativecommons.org/licenses/by/4.0/).

65

4.3 Detection discrepancies of the UCYN-A symbioses

In the WTSP, one of our results was particularly interesting in light of the high host fidelity reported for UCYN-A and its prymnesiophyte host (Cabello et al., 2016; Cornejo-Castillo et al., 2016; Farnelid et al., 2016). There are currently three strains of UCYN-A known to be in symbiosis with a eukaryote host, and two of these have specifically designed oligonucleotides for both the symbiont and host (Thompson et al., 2014; Paper I). Thus, the UCYN-A strains and their respective hosts can be quantified in parallel. Previous studies have suggested a high host dependency and selectivity (Cabello et al., 2016; Cornejo-Castillo et al., 2016; Farnelid et al., 2016), due to the loss of PSII, TCA cycle, and several amino acids and purines (Tripp et al., 2010) rendering it dependent and/or an obligate symbiont. Interestingly, our qPCR data revealed several samples with detections of the UCYN-A symbiont without its proposed host, while the UCYN-A hosts were always detected in samples with the highest abundances of the UCYN-A (Fig. 12). This suggests a higher obligateness for the prymnesiophyte host. However, since qPCR is based on a specific set of oligonucleotides, we could be missing a greater diversity of UCYN-A hosts.

Despite the discrepancy in detections of the UCYN-A strains (A1 and A2) and their hosts, they still significantly correlated with each other, and showed the same patterns of correlations with environmental parameters (e.g. negatively with temperature and PAR). However, using RDA the O2 concentration and fluorescence in the ambient water separated the symbionts and the hosts. Interestingly, the abundance of UCYN-A strains was more positively affected by both O2 and fluorescence than their proposed hosts.

66

Figure 12. Detections of the UCYN-A1 strain (LOG nifH genes L-1

) and its proposed host (LOG

18S rRNA genes L-1

) throughout the OUTPACE cruise. Each dot in the depth plot is one

sampling point. The black boxes highlight the distribution disparities, and the red circles are

the samples with highest detections of UCYN-A1. The figure is modified from Stenegren et al.

(2018) (Paper I).

67

4.4 Comparing DDA and UCYN-A symbioses

Despite the DDAs being first described more than a century ago (Lemmermann, 1905), comparably less is known about these important symbioses. Moreover, in Paper III we highlighted a disparity in research attention where the DDAs are less assayed by qPCR than the other open ocean cyanobacterial diazotrophs. UCYN-A on the other hand was more recently discovered (Zehr et al., 1998), and has received more attention. When reviewing 46 qPCR nifH studies, we found that all three DDAs were simultaneously quantified in less than 30 % of the studies, and at most, het-1 was quantified in ca 50 % of the studies. UCYN-A was quantified in 96 % of the studies. Moreover, in five studies the DDAs were only quantified in a subset of the samples collected. Estimating the regional and global significance of a population is often coupled with its abundances. For example, biogeochemical models rely on observations and reports of abundances. Thus, the importance of these widespread DDAs has been largely overlooked, especially considering that regional N2 fixation rates by the DDAs (0.19-0.62 T mol N yr-1) (Foster et al., 2011) are similar to Trichodesmium spp. (0.36-0.71 T mol N yr-1), which is largely considered as the main N2 fixer (Capone et al., 2008). In summary, the cumulative disparity in attention has resulted in an undersampling and poorer understanding of DDAs, their distributions, and their impact on biogeochemical cycles.

Although DDAs and the UCYN-A symbioses have a similar function and N2 fixation strategy, there are still significant differences in the genome sizes and contents between the UCYN-A and DDA symbionts (Fig. 13). Furthermore, there are also differences within the strains (Bombar et al., 2014) that for the DDA symbionts is directly related to their cellular location (Hilton et al., 2013). For example, we identified that both RintHH01 (het-2) and the UCYN-A, normalized to their smaller genome sizes, retain more genes related to N2 fixation than do the partially integrated symbiont (RintRC01, het-1).

68

Figure 13. Comparisons of genome content for selected pathways between the DDA

symbionts and the UCYN-A symbionts (A1 and A2). The number of genes are normalized to

the respective genome size. Reproduced from Caputo et al. (2018) (Paper III) under the CC

BY 4.0 license (https://creativecommons.org/licenses/by/4.0/).

69

4.5 Gene expression levels of the DDA symbionts

Building on the genome comparisons of the DDA symbionts we designed a DDA symbiont specific microarray to investigate in situ gene expression levels (Paper IV). Given the differences in genome content, associated hosts, and symbiont cellular location (external, internal), we expected to see differences in gene expression levels. Our microarray study was specifically aimed at investigating gene expression levels as a function of 1) time (dark/light), and 2) in a range of in situ environmental conditions (e.g. temperature, depth, and dissolved nutrients) of the SCS. The co-occurrence of all three DDA symbionts in this region enabled us to additionally measure their co-occurring gene expression patterns.

The level and pattern of gene expression differed between the DDA symbionts. For example, RintRC01 (het-1) had the highest detection (60 %), likely related to its higher abundances, and the PS I gene (psaA) was the most highly expressed gene. The RintHH01 (het-2) devoted a higher relative transcriptional capacity to N2 fixation, while genes related to P and K transport were highly expressed by the least abundant CalSC01 (het-3).

Non-metric multidimensional scaling (nMDS) and subsequent Adonis analyses found temperature and fluorescence to be the main drivers of gene expression levels for the DDA symbionts. The fluorescence sensor on the CTD measures chlorophyll fluorescence and is an indirect measure of phytoplankton. Temperature is a well-known driver of diazotroph abundance and distribution, but fluorescence is more surprising, since it was either negatively or non-significantly correlated with DDA symbiont abundance in the WTSP (Paper I). Fluorescence also emerged as a strong predictor of gene expression levels of Richelia in samples from the WTNA and AR plume (Foster et al. in prep). Furthermore, when investigating the nifH gene expression for all symbiont strains, expression levels were temporally separated (except for one early morning sample) (Fig. 14), where the highest expression levels of nifH were during the dark period. The higher nifH expression level during the dark period is congruent with results reported on RintRC01 (het-1) nifH expression by Harke et al. (2019) in the NEP. When comparing the environmental impacts on individual strains gene expression patterns, RintHH01 deviated from the two other strains by being significantly (p < 0.05) influenced by salinity and O2 concentration. Salinity has been shown to distribute diazotrophs along a gradient in e.g. the AR influenced WTNA (Foster et al., 2007; Paper II), but its impact on gene expression is unclear. The effect of O2 on RintHH01 could be related to the

70

internal location within its host, or a subsequent transcriptional response following a larger relative investment into the O2 sensitive nitrogenase (Eady and Leigh, 1994).

Figure 14. Time separated nifH gene expression levels by nMDS. The circles represent the

samples in a multi-dimensional space, where yellow circles are day time samples (09:00-

12:20) and black circles are night time samples (00:20-02:30). The circle colored brown was

sampled later than all other samples (03:20). The number next to each circle signifies the

station where the sample was collected.

71

5 Conclusions The DDAs are an important part of the microbial and diazotrophic community in the open ocean, and contribute significantly to biogeochemical cycles like C and N. Despite their importance, comparably little research has been conducted on the DDAs since they were first described. The disparity in attention was highlighted in this thesis. This thesis also attempted to increase our knowledge on their distributions, abundances and activity (by gene expression levels) in relation to ambient environmental conditions.

Comparisons between free-living and symbiotic N2 fixing cyanobacteria abundance were also included, and revealed several complex networks of interactions amongst cyanobacterial diazotrophs. Overall, we found that the distributions of cyanobacterial diazotrophs in the oligotrophic open ocean (WTSP) are largely constrained by temperature while in regions influenced by rivers (SCS, WTNA) parameters related to freshwater river discharge (turbidity, salinity and nutrients) drive niche separations. Our results were surprising in light of how closely related some strains are and how different the populations are in life history and genomic content.

The DDA symbiont specific microarray identified in situ expression patterns and highlighted how the strains responded to their ambient environment. Overall patterns of gene expression were highly variable and symbiont strain specific. These results highlight that the symbionts are under a varying influence of the respective host partners dictated by their cellular location. Collectively, the results of this thesis further our understanding of how the environment shapes the abundances and activities of open ocean cyanobacterial diazotrophs, and highlights the need for further research on diazotroph ecology.

72

6 Future perspectives The last few decades have seen great improvements in the methodology (e.g. single cell, next generation sequencing) and in situ tools (e.g. robotics, remotely operated vehicles) applied to the study of marine microbes. Results of recent studies using these new tools have dramatically shifted canonical views of marine diazotrophs and the marine N cycle. Still, much effort remains to map and understand the ecology, activity and distribution of marine pelagic diazotrophs within the complex N cycle. Furthermore, with the environment and climate currently changing at an unprecedented rate (Solomon et al., 2007), it is imperative that we improve our understanding of how the environment impacts the key microorganisms mediating the biogeochemical cycles, e.g. diazotrophs. Thus, we would be better equipped to model and predict the future climate impacts on the open ocean diazotrophic community.

6.1 Diazotrophs in future ocean conditions

In the face of climate change, the global ocean is predicted to increase in surface temperature (Hansen et al., 2006), and decrease in pH (Caldeira and Wickett, 2003; Wolf‐Gladrow et al., 1999), as the ocean absorbs more heat and CO2 from the atmosphere. Since many of the microbes, diazotrophs included, have been shown to have relatively narrow ranges of temperature for optimal growth, and an upper temperature limit (Breitbarth et al., 2007; Falcón et al., 2005; Langlois et al., 2005; Mazard et al., 2004), an increase in temperature in the tropical and subtropical belt could quickly reshape the diazotrophic community and their activity (Boyd and Doney, 2002; Breitbarth et al., 2007). Moreover, previous studies have showed variable responses to an increase in dissolved CO2 concentration among the diazotrophs (Böttjer et al., 2014; Hutchins et al., 2013). Therefore, some species and strains might be favored over others as dissolved CO2 concentration increases, leading to a shift in the diazotrophic community. Interestingly, the earliest appearance of DDAs was in the mid-cretaceous period (120-80 Myr ago), a time when atmospheric CO2 concentrations were higher than today (Bauersachs et al., 2010; Caputo et al., 2019; Kashiyama et al., 2008; Sachs and Repeta, 1999). Hence, one might expect these populations to be favored in future climate scenarios. However, how modern day DDAs would respond to increased temperature and CO2 concentrations remains an open question.

73

In the meta-analysis of Paper I, the DDAs (except for het-3) most often had a strong significant and positive correlation with Trichodesmium spp. Thus, it is likely that the DDAs in general will be similarly impacted as Trichodesmium spp., when ocean surface temperature increases. However, Crocosphaera have a higher optimal temperature for growth than e.g. Trichodesmium (Fu et al., 2014; Moisander et al., 2010), suggesting that Crocosphaera could be favored by increased temperature (Boatman et al., 2017; Fu et al., 2014). Although, many marine pelagic diazotrophs are found in environments outside their optimal temperature for growth, it is not well known how their N2 fixation rates are affected at elevated temperatures (Breitbarth et al., 2007; Fu et al., 2014). In this regard, no studies have been made on the DDAs. Given the DDAs’ importance for fueling the biological pump and the drawdown of CO2 from the atmosphere (Karl et al., 2012), it is crucial that we better understand how future open ocean conditions will affect them.

74

6.2 Field observations and modeling environmental

constraints

Most attempts at modelling marine N2 fixation and the marine N budget assume that all diazotrophs are equally influenced by the same environmental parameters (Deutsch et al., 2007; Landolfi et al., 2015). Hence, the call for more observations and a better understanding of the ecology of the marine pelagic diazotrophs by Landolfi et al. (2018) is a valid request. Many cruises have historically been regionally and seasonally limited. However, the recent shift in the canonical paradigm that pelagic diazotrophs are not restricted to the tropical and subtropical belt (Bentzon-Tilia et al., 2015; Harding et al., 2018; Mulholland et al., 2019; Shiozaki et al., 2018) highlights the need for more measurements in other oceanic regions and in different seasons.

Recent technological advances have allowed oceanographers to apply seagoing robotics, such as environmental sample processors (ESPs), fitted with highly specific gene based assays for detecting microbial populations, including diazotrophs (Ottesen et al., 2013; Preston et al., 2011; Robidart et al., 2014). Using these autonomous sampling approaches, it is possible to acquire a more comprehensive picture, with a higher spatiotemporal resolution. Thus, the development of advanced autonomous robotics fitted with custom seagoing tools could greatly increase data collection of marine microbes, both in terms distribution/abundance and transcription/activity.

75

76

Figure 15. Global map of the Tara Oceans cruise track (2009-2013). Each label represents

one station and the blue areas signify OMZs. Image reproduced from Pesant et al. (2015)

under the CC BY 4.0 license (https://creativecommons.org/licenses/by/4.0/).

Lastly, an increasing amount of new sequence and image data generated from pan global surveys (GOS, Tara, Malaspina) (Fig. 15) has become available in online databases. This new information greatly opens up the possibilities for more specific and sophisticated modeling. Studies using these resources have begun to unravel the greater distribution and diversity of marine microbes, diazotrophs included (Caputo et al., 2019; Cornejo-Castillo, 2017). Collectively this information and knowledge could lead to a better understanding of diazotroph ecology as a whole, and subsequently make it easier to target and study these microbes in the field, regardless of their abundances. However, the unquantified disparity reported in Paper III, should be considered in future studies since it highlights an underrepresentation of all populations, the DDAs in particular. Together, the generation of data in undersampled regions of the world’s oceans coupled with autonomous robotics and the continuous input of new data to open source online databases holds promise for providing the necessary input data for better predictive modeling. Improvements in model predictions of marine N2 fixation will also help field campaigns to select appropriate study sites.

77

6.3 More reliable detections of diazotrophs

As demonstrated in Paper I, there are a number of caveats to the qPCR method that are difficult to address. The most obvious one concerns the matter of cross-hybridization of the nifH oligonucleotides between the closely related diazotroph strains. Other genes have been successfully applied in newly designed qPCR assays for the DDA symbiont strains with no cross-hybridizations (e.g. exbB, pstS, rnpB) (Ley, 2019). However, some genes are more/less common and broadly distributed, and perhaps not well inventoried in the databases. Thus, it is difficult to properly test for cross-hybridization against the wider microbial community. Therefore, the preferred genes should be specific, and in terms of distinguishing between closely related strains, less conserved. For the DDA symbionts, the hetR gene could be one option, since it had a greater nucleotide sequence (83.8%) divergence amongst the strains than nifH (91.1%) (Foster and Zehr, 2006). Moreover, based on the rigorous cross-hybridization tests when designing the microarray, both nifK and nifE also appear as candidate genes for distinguishing the DDA symbiont strains. Another challenge to more reliable estimates of a particular population (e.g. by qPCR) is related to the natural environmental heterogeneity in samples. The heterogeneity within e.g. a niskin bottle could be decreased by filtering the entire content or by mixing the sample water prior to sampling. However, this is logistically difficult in the field. Alternatively, sequential sampling from a niskin bottle (sample 1-4) then subsequent assaying could determine the extent of heterogeneity. Quantitative PCR assays do not distinguish environmental DNA or DNA from dead cells, and also assumes extraction efficiencies are sufficient. A spike-in control could be used to evaluate DNA extraction efficiencies (O’Connell et al., 2017). Currently, the method of quantifying diazotrophs risks both over and underestimation, which influences our current understanding of their distributions and importance. Method development will be an important aspect for understanding the role of diazotrophs in the global ocean cycle.

78

6.4 Hypotheses generated by pSEM

Increasing knowledge of a system, together with larger datasets including more biotic and abiotic environmental parameters, would allow us to model complex cascading effects within an ecosystem. Subsequently, it would open up the possibility to identify novel and testable interactions within a given system. The environmental parameter input to the pSEM is largely restricted to the traditional oceanographic parameters regularly sampled or measured e.g. by the CTD. Thus, it is possible that a wider range of environmental parameters included in the model could help explain some of the observed patterns and interactions of the diazotrophs. An extended list of biotic environmental parameters, including pigments (e.g. diatoxanthin and fucoxanthin), was investigated with pSEM in the SCS qPCR nifH datasets, and revealed direct influences of pigment concentrations on some of the diazotrophs (Fig. 16). These interactions could be interpreted as proxies for other co-occurring phytoplankton, but at the moment no greater interpretation of these results has been made. The pSEM generated many new testable hypotheses. Assuming the eventual isolation of the DDAs, several parameters highlighted in the pSEM as significantly influencing abundance or gene expression (salinity, P, etc.), could be tested experimentally. Additionally, these pSEM analyses also demonstrated the importance of the other diazotrophs and the direct positive effects they have on each other (Fig. 15). One such example is the UCYN-C and the het-3, which could be grown together in a perturbation experiment with differing inputs (e.g. nutrients), while abundances, cell integrity and gene expression levels could be measured.

79

Figure 16. The complete pSEM output of qPCR and environmental data from the Mekong

river plume and the SCS. The network is structured with the strongest predictor parameters

on the top and cascading down in strength. Green arrows are positive effects and red arrows

are negative effects, and the arrow thickness signifies the strength of the effect. The ellipses

are color coded based on input group (green=pigments, orange=diazotrophs,

purple=nutrients and blue=abiotic parameters). The following abbreviations apply: perid:

peridinin, fuco: fucoxanthin, zea: zeaxanthin, diato: diatoxanthin, diadino: diadinoxanthin,

Chl a: chlorophyll a, Chl b: chlorophyll b, Chl c: chlorophyll c, beam.trans: beam transmission,

PAR: photosynthetically active radiation, cond: conductivity, sal: salinity, temp: temperature,

O2 sat: O2 saturation, tricho: Trichodesmium.

80

6.5 Symbiont-host interactions and host control

As demonstrated in Paper I there were discrepancies in the detections of UCYN-A and their proposed hosts. All possible qPCR caveats aside, this raised the question on UCYN-A host specificity. Similarly, oligonucleotides could be designed for the host of the DDAs to investigate the coupling of the DDA symbiont and their respective hosts as we showed with the UCYN-A symbioses (Paper I). Furthermore, several of the highly expressed genes identified in the microarray (nifH, psaA, groEL, psbO) could be used for designing gene-fluorescence in situ hybridization (gene-FISH) in order to identify (in a single cell) how the symbiont gene expression responds in situ. Double gene-FISH assays that target genes of both partners would be an invaluable approach to studying the partner interactions.

A very open and obvious question remains as to how much the host controls the symbiont growth and activity. Earlier work has suggested that the symbionts are more highly active in the symbiotic state than living freely (Foster et al., 2011). In addition, eukaryotic inhibitor studies on wild DDA populations measured reduced N2 and C fixation when inhibiting the host (Foster et al. in prep). Using the microarray of Paper IV we also identified strong proxies for the host impact on DDA symbiont gene expression levels. Ultimately two steps in properly addressing the host-symbiont interactions of DDAs involve closing the draft genomes of the DDA symbionts and fully sequencing the host diatom genomes. Having both partner genomes also allows for a deeper investigation into which genes, if any, have been transferred between partners. Endosymbiont gene transfer (EGT) is common in other symbioses (Nowack et al., 2008, 2016; Yurchenko et al., 2018).

After genome sequencing of both partners, subsequent investigations could include application of environmental transcriptomic studies (e.g. RNA seq) or the development of a host-symbiont specific microarray. Both can be applied to whole water and include experiments with eukaryotic inhibitors. Thus, both partner gene expression patterns can be studied, and additionally the results from the latter experiment could address the amount of control that the host asserts over its symbiont gene expression. A recent RNA seq study was successful for studying the Rhizosolenia-Richelia symbioses in the North Pacific gyre, however it required filtering of ca 10 L per sample for successful application (Harke et al., 2019).

81

7 Acknowledgments I would like to thank my supervisor, Rachel Foster, for this great opportunity and memorable experience (financed by Knut and Alice Wallenberg Foundation), and all the help and support provided, both practical and theoretical, along the way. None of this would have been possible without you. I would also like to thank the rest of our lab group over the years (Andrea Caputo, Sofie Vonlanthen, Massimo Pernice, Mercedes Nieves Morión, Lotta Berntzon, Charlotta Ekblom, Elina Viinamäki, Andrew Prevett and Theo Vigil Stenman) for all the help, feedback and laughs. A special thanks to Carlo Berg, Anders Bignert, Daniel Lundin and David Brodin for their invaluable insights on statistical analysis and bioinformatics. Thanks to Thierry Moutin, Sophie Bonnet and the OUTPACE2015 scientists and ship crew of the R/V L’Atalante for providing me the opportunity to sample for the first study of this thesis, and thanks to Joseph Montoya, Ajit Subramaniam and the EN614 scientists and ship crew of the R/V Endeavor, especially First Mate Chris Armanetti (for making me feel right at home) and Ship Engineer Kurt Rethorn (for keeping me company during my long nights of sampling). I would also like to thank my office mates Alfred Burian, Konrad Karlsson, Per Hedberg and Andreas Novotny for all our inspiring discussions, not only concerning science, and Peter Bruce and Tom Staveley for always having your doors open. Not to forget the good beers we shared over the years. Also a big thanks to all my other friends I’ve made and met during my PhD, both in Sweden and abroad, who made my days brighter and experiences better. Lastly, a very special thank you to my family, wife and all you crazy kids who have shown great patience, support and love when my PhD has required me to be elsewhere, both physically and mentally.

82

8 References

Armbrust, E. V. (2009). The life of diatoms in the world’s oceans. Nature, 459(7244), 185–192. https://doi.org/10.1038/nature08057

Austin, Å. N., Hansen, J. P., Donadi, S., & Eklöf, J. S. (2017). Relationships between aquatic vegetation and water turbidity: A field survey across seasons and spatial scales. PLOS ONE, 12(8), https://doi.org/10.1371/journal.pone.0181419

Bauersachs, T., Speelman, E. N., Hopmans, E. C., Reichart, G-J., Schouten, S & Damsté J. S. S. (2010). Fossilized glycolipids reveal past oceanic N2 fixation by heterocystous cyanobacteria. Proceedings of the National Academy of

Sciences, 107(45), 19190–19194. Bekker, A. (2014). Great oxygenation event. Encyclopedia of Astrobiology, 1–9. Bench, S. R., Frank, I., Robidart, J., & Zehr, J. P. (2016). Two subpopulations of

Crocosphaera watsonii have distinct distributions in the North and South Pacific. Environmental Microbiology, 18(2), 514–524. https://doi.org/10.1111/1462-2920.13180

Bench, S. R., Heller, P., Frank, I., Arciniega, M., Shilova, I. N., & Zehr, J. P. (2013). Whole genome comparison of six Crocosphaera watsonii strains with differing phenotypes. Journal of Phycology, 49(4), 786–801. https://doi.org/10.1111/jpy.12090

Bench, S. R., Ilikchyan, I. N., Tripp, H. J., & Zehr, J. P. (2011). Two strains of Crocosphaera watsonii with highly conserved genomes are distinguished by strain-specific features. Frontiers in Microbiology, 2. https://doi.org/10.3389/fmicb.2011.00261

Bentzon-Tilia, M., Severin, I., Hansen, L. H., & Riemann, L. (2015). Genomics and ecophysiology of heterotrophic nitrogen-fixing bacteria isolated from estuarine surface water. The ISME Journal, 6(4). https://doi.org/10.1128/mBio.00929-15

Berman-Frank, I., Bidle, K. D., Haramaty, L., & Falkowski, P. G. (2004). The demise of the marine cyanobacterium, Trichodesmium spp., via an autocatalyzed cell death pathway. Limnology and Oceanography, 49(4), 997–1005. https://doi.org/10.4319/lo.2004.49.4.0997

Berman-Frank, I., Lundgren, P., Chen, Y.-B., Küpper, H., Kolber, Z., Bergman, B., & Falkowski, P. (2001). Segregation of nitrogen fixation and oxygenic photosynthesis in the marine cyanobacterium Trichodesmium. Science, 294(5546), 1534–1537. https://doi.org/10.1126/science.1064082

Boatman, T. G., Lawson, T., & Geider, R. J. (2017). A key marine diazotroph in a changing ocean: The interacting effects of temperature, CO2 and light on the

83

growth of Trichodesmium erythraeum IMS101. PLOS ONE, 12(1). https://doi.org/10.1371/journal.pone.0168796

Bombar, D., Moisander, P. H., Dippner, J. W., Foster, R. A., Voss, M., Karfeld, B., & Zehr, J. P. (2011). Distribution of diazotrophic microorganisms and nifH gene expression in the Mekong River plume during intermonsoon. Marine

Ecology Progress Series, 424, 39–52. https://doi.org/10.3354/meps08976 Bombar, D., Heller, P., Sanchez-Baracaldo, P., Carter, B. J., & Zehr, J. P. (2014).

Comparative genomics reveals surprising divergence of two closely related strains of uncultivated UCYN-A cyanobacteria. The ISME Journal, 8(12), 2530–2542. https://doi.org/10.1038/ismej.2014.167

Bombar, D., Paerl, R. W., & Riemann, L. (2016). Marine non-cyanobacterial diazotrophs: moving beyond molecular detection. Trends in Microbiology, 24(11), 916–927. https://doi.org/10.1016/j.tim.2016.07.002

Bonnet, S., Biegala, I. C., Dutrieux, P., Slemons, L. O., & Capone, D. G. (2009). Nitrogen fixation in the western equatorial Pacific: Rates, diazotrophic cyanobacterial size class distribution, and biogeochemical significance. Global Biogeochemical Cycles, 23(3). https://doi.org/10.1029/2008GB003439

Bonnet, S., Caffin, M., Berthelot, H., & Moutin, T. (2017). Hot spot of N2 fixation in the western tropical South Pacific pleads for a spatial decoupling between N2 fixation and denitrification. Proceedings of the National Academy of

Sciences, 114(14). https://doi.org/10.1073/pnas.1619514114 Bonnet, S., Dekaezemacker, J., Turk-Kubo, K. A., Moutin, T., Hamersley, R. M.,

Grosso, O., et al. (2013). Aphotic N2 fixation in the eastern tropical South Pacific Ocean. PLOS ONE, 8(12). https://doi.org/10.1371/journal.pone.0081265

Bonnet, S., Guieu, C., Bruyant, F., Prášil, O., Wambeke, F. V., Raimbault, P., et al. (2008). Nutrient limitation of primary productivity in the Southeast Pacific (BIOSOPE cruise). Biogeosciences, 5(1), 215–225. https://doi.org/https://doi.org/10.5194/bg-5-215-2008

Bonnet, S., Rodier, M., Turk-Kubo, K. A., Germineaud, C., Menkes, C., Ganachaud, A., et al. (2015). Contrasted geographical distribution of N2 fixation rates and nifH phylotypes in the Coral and Solomon Seas (southwestern Pacific) during austral winter conditions. Global

Biogeochemical Cycles, 29(11). https://doi.org/10.1002/2015GB005117 Borstad, G. A. (1982). The influence of the meandering Guiana Current and

Amazon River discharge on surface salinity near Barbados. Journal of

Marine Research, 40(2), 421–433. Boström, K. H., Riemann, L., Kühl, M., & Hagström, Å. (2007). Isolation and

gene quantification of heterotrophic N2-fixing bacterioplankton in the Baltic

84

Sea. Environmental Microbiology, 9(1), 152–164. https://doi.org/10.1111/j.1462-2920.2006.01124.x

Böttjer, D., Karl, D. M., Letelier, R. M., Viviani, D. A., & Church, M. J. (2014). Experimental assessment of diazotroph responses to elevated seawater pCO2 in the North Pacific subtropical gyre. Global Biogeochemical Cycles, 28(6), 601–616. https://doi.org/10.1002/2013GB004690

Boyd, P. W., & Doney, S. C. (2002). Modelling regional responses by marine pelagic ecosystems to global climate change. Geophysical Research Letters, 29(16), 53-1-53–54. https://doi.org/10.1029/2001GL014130

Breitbarth, E., Oschlies, A., & Laroche, J. (2007). Physiological constraints on the global distribution of Trichodesmium: effect of temperature on diazotrophy. Biogeosciences, 4(1), 53–61.

Cabello, A. M., Cornejo-Castillo, F. M., Raho, N., Blasco, D., Vidal, M., Audic, S., et al. (2016). Global distribution and vertical patterns of a prymnesiophyte–cyanobacteria obligate symbiosis. The ISME Journal, 10(3), 693–706. https://doi.org/10.1038/ismej.2015.147

Caldeira, K., & Wickett, M. E. (2003). Anthropogenic carbon and ocean pH. Nature, 425(6956), 365–365. https://doi.org/10.1038/425365a

Capone, D. G. (2001). Marine nitrogen fixation: what’s the fuss? Current Opinion

in Microbiology, 4(3), 341–348. https://doi.org/10.1016/S1369-5274(00)00215-0

Capone, D. G., Bronk, D. A., Mulholland, M. R., & Carpenter, E. J. (2008). Nitrogen in the marine environment. Elsevier.

Capone, D. G., Burns, J. A., Montoya, J. P., Subramaniam, A., Mahaffey, C., Gunderson, T., et al. (2005). Nitrogen fixation by Trichodesmium spp.: An important source of new nitrogen to the tropical and subtropical North Atlantic Ocean. Global Biogeochemical Cycles, 19(2). https://doi.org/10.1029/2004GB002331

Capone, D. G., & Carpenter, E. J. (1982). Nitrogen fixation in the marine environment. Science, 217(4565), 1140–1142. https://doi.org/10.1126/science.217.4565.1140

Capone, D. G., Zehr, J. P., Paerl, H. W., Bergman, B., & Carpenter, E. J. (1997). Trichodesmium, a globally significant marine cyanobacterium. Science, 276(5316), 1221–1229. https://doi.org/10.1126/science.276.5316.1221

Caputo, A., Nylander, J. a. A., & Foster, R. A. (2019). The genetic diversity and evolution of diatom-diazotroph associations highlights traits favoring symbiont integration. FEMS Microbiology Letters, 366(2). https://doi.org/10.1093/femsle/fny297

85

Caputo, A., Stenegren, M., Pernice, M. C., & Foster, R. A. (2018). A short comparison of two marine planktonic diazotrophic symbioses highlights an un-quantified disparity. Frontiers in Marine Science, 5. https://doi.org/10.3389/fmars.2018.00002

Carpenter, E. J. (1983). Nitrogen in the marine environment. In E. J. Carpenter & D. G. Capone (Eds.), Nitrogen fixation by marine Oscillatoria. Elsevier.

Carpenter, E. J., & Culliney, J. L. (1975). Nitrogen fixation in marine shipworms. Science, 187(4176), 551–552. https://doi.org/10.1126/science.187.4176.551

Carpenter, E. J., & Janson, S. (2000). Intracellular cyanobacterial symbionts in the marine diatom Climacodium frauenfeldianum (bacillariophyceae). Journal of

Phycology, 36(3), 540–544. https://doi.org/10.1046/j.1529-8817.2000.99163.x

Carpenter, E. J., & Janson, S. (2001). Anabaena gerdii sp. nov., a new planktonic filamentous cyanobacterium from the South Pacific Ocean and Arabian Sea. Phycologia; Lawrence, 40(2), 105–110.

Carpenter, E. J., Montoya, J. P., Burns, J., Mulholland, M. R., Subramaniam, A., & Capone, D. G. (1999). Extensive bloom of a N₂-fixing diatom/cyanobacterial association in the tropical Atlantic Ocean. Marine

Ecology Progress Series, 185, 273-283. https://doi.org/10.3354/meps185273 Chapin, D. M., Bliss, L. C., & Bledsoe, L. J. (1991). Environmental regulation of

nitrogen fixation in a high arctic lowland ecosystem. Canadian Journal of

Botany, 69(12), 2744–2755. https://doi.org/10.1139/b91-345 Chappell, P. D., & Webb, E. A. (2010). A molecular assessment of the iron stress

response in the two phylogenetic clades of Trichodesmium. Environmental

Microbiology, 12(1), 13–27. https://doi.org/10.1111/j.1462-2920.2009.02026.x

Chen, M., Lu, Y., Jiao, N., Tian, J., Kao, S.-J., & Zhang, Y. (2019). Biogeographic drivers of diazotrophs in the western Pacific Ocean. Limnology and Oceanography, 64(3), 1403–1421. https://doi.org/10.1002/lno.11123

Chen, Y., & Siefert, R. L. (2004). Seasonal and spatial distributions and dry deposition fluxes of atmospheric total and labile iron over the tropical and subtropical North Atlantic Ocean. Journal of Geophysical Research:

Atmospheres, 109(D9). https://doi.org/10.1029/2003JD003958 Chen, Y.-B., Dominic, B., Mellon, M. T. & Zehr, J. P. (1998) Circadian rhythm of

nitrogenase gene expression in the diazotrophic filamentous nonheterocystous cyanobacterium Trichodesmium sp. strain IMS 101, J. Bacteriology, 180(14), 3598–3605.

86

Church, M. J., Jenkins, B. D., Karl, D. M., & Zehr, J. P. (2005). Vertical distributions of nitrogen-fixing phylotypes at Stn ALOHA in the oligotrophic North Pacific Ocean. Aquatic Microbial Ecology, 38(1), 3–14. https://doi.org/10.3354/ame038003

Clark, L. L., Ingall, E. D., & Benner, R. (1999). Marine organic phosphorus cycling: Novel insights from nuclear magnetic resonance. American Journal

of Science, 724–737. Conley, D. J. (1997). Riverine contribution of biogenic silica to the oceanic silica

budget. Limnology and Oceanography, 42(4), 774–777. https://doi.org/10.4319/lo.1997.42.4.0774

Conroy, B. J., Steinberg, D. K., Song, B., Kalmbach, A., Carpenter, E. J., & Foster, R. A. (2017). Mesozooplankton graze on cyanobacteria in the Amazon River Plume and western tropical North Atlantic. Frontiers in Microbiology, 8. https://doi.org/10.3389/fmicb.2017.01436

Cornejo-Castillo, F. M. (2017). Diversity, ecology and evolution of marine diazotrophic microorganisms. Retrieved from https://digital.csic.es/handle/10261/164042

Cornejo-Castillo, F. M., Cabello, A. M., Salazar, G., Sánchez-Baracaldo, P., Lima-Mendez, G., Hingamp, P., et al. (2016). Cyanobacterial symbionts diverged in the late Cretaceous towards lineage-specific nitrogen fixation factories in single-celled phytoplankton. Nature Communications, 7(1), 1–9. https://doi.org/10.1038/ncomms11071

Cornejo‐Castillo, F. M., Muñoz‐Marín, M. del C., Turk‐Kubo, K. A., Royo‐Llonch, M., Farnelid, H., Acinas, S. G., & Zehr, J. P. (2019). UCYN-A3, a newly characterized open ocean sublineage of the symbiotic N2-fixing cyanobacterium Candidatus Atelocyanobacterium thalassa. Environmental

Microbiology, 21(1), 111–124. https://doi.org/10.1111/1462-2920.14429 Cornejo-Castillo, F. M., & Zehr, J. P. (2019). Hopanoid lipids may facilitate

aerobic nitrogen fixation in the ocean. Proceedings of the National Academy

of Sciences, 116(37), 18269–18271. https://doi.org/10.1073/pnas.1908165116

deBary, A. (1879). Die Erscheinung der Symbiose. Vortrag Auf Der Versammlung

Der Naturforscher Und Artze Zu Cassel, 1–30. Dekas, A. E., Poretsky, R. S., & Orphan, V. J. (2009). Deep-sea archaea fix and

share nitrogen in methane-consuming microbial consortia. Science, 326(5951), 422–426. https://doi.org/10.1126/science.1178223

Dekas, A. E., Fike, D. A., Chadwick, G. L., Green-Saxena, A., Fortney, J., Connon, S.A., et al. (2018). Widespread nitrogen fixation in sediments from

87

diverse deep-sea sites of elevated carbon loading. Environmental

Microbiology 20(12), 4281-4296. DeLuca, T. H., Zackrisson, O., Nilsson, M.-C., & Sellstedt, A. (2002).

Quantifying nitrogen-fixation in feather moss carpets of boreal forests. Nature, 419(6910), 917–920. https://doi.org/10.1038/nature01051

Deutsch, C., Sarmiento, J. L., Sigman, D. M., Gruber, N., & Dunne, J. P. (2007). Spatial coupling of nitrogen inputs and losses in the ocean. Nature, 445(7124), 163–167. https://doi.org/10.1038/nature05392

Díez, B., Bergman, B., Pedrós‐Alió, C., Antó, M., & Snoeijs, P. (2012). High cyanobacterial nifH gene diversity in Arctic seawater and sea ice brine. Environmental Microbiology Reports, 4(3), 360–366. https://doi.org/10.1111/j.1758-2229.2012.00343.x

Dron, A., Rabouille, S., Claquin, P., Le Roy, B., Talec, A., & Sciandra, A. (2012). Light–dark (12:12) cycle of carbon and nitrogen metabolism in Crocosphaera watsonii WH8501: Relation to the cell cycle. Environmental

Microbiology, 14(4), 967–981. https://doi.org/10.1111/j.1462-2920.2011.02675.x

Duffy, J. E., Reynolds, P. L., Boström, C., Coyer, J. A., Cusson, M., Donadi, S., et al. (2015). Biodiversity mediates top–down control in eelgrass ecosystems: A global comparative-experimental approach. Ecology Letters, 18(7), 696–705. https://doi.org/10.1111/ele.12448

Dupouy, C., Benielli-Gary, D., Neveux, J., Dandonneau, Y., & Westberry, T. K. (2011). An algorithm for detecting Trichodesmium surface blooms in the south western tropical Pacific. Biogeosciences, 8, 3631–3647. https://doi.org/10.5194/bg-8-3631-2011

Dyhrman, S. T., Ammerman, J. W., & Van Mooy, B. (2007). Microbes and the marine phosphorus cycle. Oceanography, 20(2), 110–116.

Dyhrman, S. T., Chappell, P. D., Haley, S. T., Moffett, J. W., Orchard, E. D., Waterbury, J. B., et al. (2006). Phosphonate utilization by the globally important marine diazotroph Trichodesmium, Nature, 439(7072), 68–71, doi:10.1038/nature04203.

Dyhrman, S. T., & Haley, S. T. (2006). Phosphorus scavenging in the unicellular marine diazotroph Crocosphaera watsonii. Applied and Environmental

Microbiology, 72(2), 1452–1458. https://doi.org/10.1128/AEM.72.2.1452-1458.2006

Eady, R. R., & Leigh, J. G. (1994). Metals in the nitrogenases. Journal of the

Chemical Society, Dalton Transactions, 0(19), 2739–2747. https://doi.org/10.1039/DT9940002739

88

Eigemann, F., Schwartke, M., & Schulz-Vogt, H. (2018). Niche separation of Baltic Sea cyanobacteria during bloom events by species interactions and autecological preferences. Harmful Algae, 72, 65–73. https://doi.org/10.1016/j.hal.2018.01.001

Eppley, R. W., & Peterson, B. J. (1979). Particulate organic matter flux and planktonic new production in the deep ocean. Nature, 282, 677–680. https://doi.org/10.1038/282677a0

Falcón, L. I., Pluvinage, S., & Carpenter, E. J. (2005). Growth kinetics of marine unicellular N2-fixing cyanobacterial isolates in continuous culture in relation to phosphorus and temperature. Marine Ecology Progress Series, 285, 3–9. https://doi.org/10.3354/meps285003

Farnelid, H., Andersson, A. F., Bertilsson, S., Al-Soud, W. A., Hansen, L. H., Sørensen, S., et al. (2011). Nitrogenase gene amplicons from global marine surface waters are dominated by genes of non-cyanobacteria. PLOS ONE, 6(4), e19223. https://doi.org/10.1371/journal.pone.0019223

Farnelid, H., Harder, J., Bentzon‐Tilia, M., & Riemann, L. (2014). Isolation of heterotrophic diazotrophic bacteria from estuarine surface waters. Environmental Microbiology, 16(10), 3072–3082. https://doi.org/10.1111/1462-2920.12335

Farnelid, H., Turk-Kubo, K., Muñoz-Marín, M. del C., & Zehr, J. P. (2016). New insights into the ecology of the globally significant uncultured nitrogen-fixing symbiont UCYN-A. Aquatic Microbial Ecology, 77(3), 125–138. https://doi.org/10.3354/ame01794

Fernandez, C., Farías, L., & Ulloa, O. (2011). Nitrogen fixation in denitrified marine waters. PLOS ONE, 6(6). https://doi.org/10.1371/journal.pone.0020539

Field, C. B., Behrenfeld, M. J., Randerson, J. T., & Falkowski, P. (1998). Primary production of the biosphere: Integrating terrestrial and oceanic components. Science, 281(5374), 237–240.

Finni, T., Kononen, K., Olsonen, R., & Wallström, K. (2001). The history of cyanobacterial blooms in the Baltic Sea. AMBIO: A Journal of the Human

Environment, 30(4), 172–178. https://doi.org/10.1579/0044-7447-30.4.172 Foster, R. A., Goebel, N. L., & Zehr, J. P. (2010). Isolation of Calothrix

rhizosoleniae (cyanobacteria) Strain SC01 from Chaetoceros (bacillariophyta) spp. Diatoms of the Subtropical North Pacific Ocean. Journal of Phycology, 46(5), 1028–1037. https://doi.org/10.1111/j.1529-8817.2010.00885.x

Foster, R. A., Kuypers, M. M. M., Vagner, T., Paerl, R. W., Musat, N., & Zehr, J. P. (2011). Nitrogen fixation and transfer in open ocean diatom–

89

cyanobacterial symbioses. The ISME Journal, 5(9), 1484–1493. https://doi.org/10.1038/ismej.2011.26

Foster, R. A., Subramaniam, A., Mahaffey, C., Carpenter, E. J., Capone, D. G., & Zehr, J. P. (2007). Influence of the Amazon River plume on distributions of free-living and symbiotic cyanobacteria in the western tropical North Atlantic Ocean. Limnology and Oceanography, 52(2), 517–532. https://doi.org/10.4319/lo.2007.52.2.0517

Foster, R. A., Subramaniam, A., & Zehr, J. P. (2009). Distribution and activity of diazotrophs in the Eastern Equatorial Atlantic. Environmental Microbiology, 11(4), 741–750. https://doi.org/10.1111/j.1462-2920.2008.01796.x

Foster, R. A., & Zehr, J. P. (2006). Characterization of diatom–cyanobacteria symbioses on the basis of nifH, hetR and 16S rRNA sequences. Environmental Microbiology, 8(11), 1913–1925. https://doi.org/10.1111/j.1462-2920.2006.01068.x

Foster, R. A., & Zehr, J. P. (2019). Diversity, genomics, and distribution of phytoplankton-cyanobacterium single-cell symbiotic associations. Annual

Review of Microbiology, 73(1), 435–456. https://doi.org/10.1146/annurev-micro-090817-062650

Fredriksson, C., & Bergman, B. (1997). Ultrastructural characterization of cells specialized for nitrogen fixation in a non-heterocystous cyanobacterium, Trichodesmium spp. Protoplasma, 197(1), 76–85. https://doi.org/10.1007/BF01279886

French, J. R. J., Turner, G. L., & Bradbury, J. F. (1976). Nitrogen fixation by bacteria from the hindgut of termites. Microbiology, 95(2), 202–206. https://doi.org/10.1099/00221287-95-2-202

Frischkorn, K. R., Rouco, M., Mooy, B. A. S. V., & Dyhrman, S. T. (2017). Epibionts dominate metabolic functional potential of Trichodesmium colonies from the oligotrophic ocean. The ISME Journal, 11(9), 2090–2101. https://doi.org/10.1038/ismej.2017.74

Fu, F.-X., Yu, E., Garcia, N. S., Gale, J., Luo, Y., Webb, E. A., & Hutchins, D. A. (2014). Differing responses of marine N2 fixers to warming and consequences for future diazotroph community structure. Aquatic Microbial

Ecology, 72(1), 33–46. https://doi.org/10.3354/ame01683 Garcia, N. S., Fu, F., Sedwick, P. N., & Hutchins, D. A. (2015). Iron deficiency

increases growth and nitrogen-fixation rates of phosphorus-deficient marine cyanobacteria. The ISME Journal, 9(1), 238–245. https://doi.org/10.1038/ismej.2014.104

Gaye, B., Nagel, B., Dähnke, K., Rixen, T., & Emeis, K.-C. (2013). Evidence of parallel denitrification and nitrite oxidation in the ODZ of the Arabian Sea

90

from paired stable isotopes of nitrate and nitrite. Global Biogeochemical

Cycles, 27(4), 1059–1071. https://doi.org/10.1002/2011GB004115 Goebel, N. L., Turk, K. A., Achilles, K. M., Paerl, R., Hewson, I., Morrison, A. E.,

et al. (2010). Abundance and distribution of major groups of diazotrophic cyanobacteria and their potential contribution to N2 fixation in the tropical Atlantic Ocean. Environmental Microbiology, 12(12), 3272–3289. https://doi.org/10.1111/j.1462-2920.2010.02303.x

Goes, J. I., Gomes, H. do R., Chekalyuk, A. M., Carpenter, E. J., Montoya, J. P., Coles, V. J., et al. (2014). Influence of the Amazon River discharge on the biogeography of phytoplankton communities in the western tropical north Atlantic. Progress in Oceanography, 120, 29–40. https://doi.org/10.1016/j.pocean.2013.07.010

Grace, J. B. (2006). Structural equation modeling and natural systems. Cambridge

University Press. Großkopf, T., Mohr, W., Baustian, T., Schunck, H., Gill, D., Kuypers, M. M. M.,

et al. (2012). Doubling of marine dinitrogen-fixation rates based on direct measurements. Nature, 488(7411), 361–364. https://doi.org/10.1038/nature11338

Gruber, N., & Sarmiento, J. L. (1997). Global patterns of marine nitrogen fixation and denitrification. Global Biogeochemical Cycles, 11(2), 235–266. https://doi.org/10.1029/97GB00077

Hagino, K., Onuma, R., Kawachi, M., & Horiguchi, T. (2013). Discovery of an endosymbiotic nitrogen-fixing cyanobacterium UCYN-A in Braarudosphaera bigelowii (Prymnesiophyceae). PLOS ONE, 8(12), e81749. https://doi.org/10.1371/journal.pone.0081749

Halm, H., Lam, P., Ferdelman, T. G., Lavik, G., Dittmar, T., LaRoche, J., et al. (2012). Heterotrophic organisms dominate nitrogen fixation in the South Pacific Gyre. The ISME Journal, 6(6), 1238–1249. https://doi.org/10.1038/ismej.2011.182

Halm, H., Musat, N., Lam, P., Langlois, R., Musat, F., Peduzzi, S., et al. (2009). Co-occurrence of denitrification and nitrogen fixation in a meromictic lake, Lake Cadagno (Switzerland). Environmental Microbiology, 11(8), 1945–1958. https://doi.org/10.1111/j.1462-2920.2009.01917.x

Hansen, J., Sato, M., Ruedy, R., Lo, K., Lea, D. W., & Medina-Elizade, M. (2006). Global temperature change. Proceedings of the National Academy of

Sciences, 103(39), 14288–14293. https://doi.org/10.1073/pnas.0606291103 Harding, K., Turk-Kubo, K. A., Sipler, R. E., Mills, M. M., Bronk, D. A., & Zehr,

J. P. (2018). Symbiotic unicellular cyanobacteria fix nitrogen in the Arctic

91

Ocean. Proceedings of the National Academy of Sciences, 115(52), 13371–13375. https://doi.org/10.1073/pnas.1813658115

Harke, M. J., Frischkorn, K. R., Haley, S. T., Aylward, F. O., Zehr, J. P., & Dyhrman, S. T. (2019). Periodic and coordinated gene expression between a diazotroph and its diatom host. The ISME Journal, 13(1), 118–131. https://doi.org/10.1038/s41396-018-0262-2

Haselkorn, R. (1978). Heterocysts. Annual Review of Plant Physiology, 29(1), 319–344. https://doi.org/10.1146/annurev.pp.29.060178.001535

Henke, B. A., Turk-Kubo, K. A., Bonnet, S., & Zehr, J. P. (2018). Distributions and abundances of sublineages of the N2-fixing cyanobacterium Candidatus Atelocyanobacterium thalassa (UCYN-A) in the New Caledonian coral lagoon. Frontiers in Microbiology, 9. https://doi.org/10.3389/fmicb.2018.00554

Hewson, I., Moisander, P. H., Morrison, A. E., & Zehr, J. P. (2007). Diazotrophic bacterioplankton in a coral reef lagoon: phylogeny, diel nitrogenase expression and response to phosphate enrichment. The ISME Journal, 1(1), 78–91. https://doi.org/10.1038/ismej.2007.5

Hewson, I., Poretsky, R. S., Dyhrman, S. T., Zielinski, B., White, A. E., Tripp, H. J., et al. (2009). Microbial community gene expression within colonies of the diazotroph, Trichodesmium, from the Southwest Pacific Ocean. The ISME

Journal, 3(11), 1286–1300. https://doi.org/10.1038/ismej.2009.75 Hilton, J. A., Foster, R. A., James Tripp, H., Carter, B. J., Zehr, J. P., & Villareal,

T. A. (2013). Genomic deletions disrupt nitrogen metabolism pathways of a cyanobacterial diatom symbiont. Nature Communications, 4, 1767. https://doi.org/10.1038/ncomms2748

Holland, H. D. (2002). Volcanic gases, black smokers, and the great oxidation event. Geochimica et Cosmochimica Acta, 66(21), 3811–3826. https://doi.org/10.1016/S0016-7037(02)00950-X

Holland, H. D. (2006). The oxygenation of the atmosphere and oceans. Philosophical Transactions of the Royal Society B: Biological Sciences, 361(1470), 903–915. https://doi.org/10.1098/rstb.2006.1838

Howard, J. B., & Rees, D. C. (1996). Structural basis of biological nitrogen fixation. Chemical Reviews, 96(7), 2965–2982. https://doi.org/10.1021/cr9500545

Howarth, R. W., & Cole, J. J. (1985). Molybdenum availability, nitrogen limitation, and phytoplankton growth in natural waters. Science, 229(4714), 653–655. https://doi.org/10.1126/science.229.4714.653

Howarth, R. W., Marino, R., Lane, J., & Cole, J. J. (1988). Nitrogen fixation in freshwater, estuarine, and marine ecosystems: Rates and importance.

92

Limnology and Oceanography, 33(4 part 2), 669–687. https://doi.org/10.4319/lo.1988.33.4part2.0669

Hutchins, D. A., Fu, F.-X., Webb, E. A., Walworth, N., & Tagliabue, A. (2013). Taxon-specific response of marine nitrogen fixers to elevated carbon dioxide concentrations. Nature Geoscience, 6(9), 790–795. https://doi.org/10.1038/ngeo1858

Janson, S., Wouters, J., Bergman, B., & Carpenter, E. J. (1999). Host specificity in the Richelia-diatom symbiosis revealed by hetR gene sequence analysis. Environmental Microbiology, 1(5), 431–438.

Jorge, A. M., Hoiczyk, E., Gomes, J. P., & Pinho, M. G. (2011). EzrA contributes to the regulation of cell size in Staphylococcus aureus. PLOS ONE, 6(11), e27542. https://doi.org/10.1371/journal.pone.0027542

Karl, D., Letelier, R., Tupas, L., Dore, J., Christian, J., & Hebel, D. (1997). The role of nitrogen fixation in biogeochemical cycling in the subtropical North Pacific Ocean. Nature, 388(6642), 533–538. https://doi.org/10.1038/41474

Karl, D. M., Church, M. J., Dore, J. E., Letelier, R. M., & Mahaffey, C. (2012). Predictable and efficient carbon sequestration in the North Pacific Ocean supported by symbiotic nitrogen fixation. Proceedings of the National

Academy of Sciences, 109(6), 1842–1849. https://doi.org/10.1073/pnas.1120312109

Karner, M. B., DeLong, E. F., & Karl, D. M. (2001). Archaeal dominance in the mesopelagic zone of the Pacific Ocean. Nature, 409(6819), 507–510. https://doi.org/10.1038/35054051

Kashiyama, Y., Ogawa, N. O., Kuroda, J., Shiro, M., Nomoto, S., Tada, R. et al. (2008). Diazotrophic cyanobacteria as the major photoautotrophs during mid-Cretaceous oceanic anoxic events: Nitrogen and carbon isotopic evidence from sedimentary porphyrin. Organic Geochemistry, 39(5), 532–549. https://doi.org/10.1016/j.orggeochem.2007.11.010

Kneip, C., Lockhart, P., Voß, C., & Maier, U.-G. (2007). Nitrogen fixation in eukaryotes – New models for symbiosis. BMC Evolutionary Biology, 7(1), 55. https://doi.org/10.1186/1471-2148-7-55

Kong, L., Jing, H., Kataoka, T., Sun, J., & Liu, H. (2011). Phylogenetic diversity and spatio-temporal distribution of nitrogenase genes (nifH) in the northern South China Sea. Aquatic Microbial Ecology, 65(1), 15–27. https://doi.org/10.3354/ame01531

Koricheva, J., Gurevitch, J., & Mengersen, K. (2013). Handbook of meta-analysis in ecology and evolution. Princeton University Press.

Kustka, A. B., Sañudo-Wilhelmy, S. A., Carpenter, E. J., Capone, D., Burns, J., & Sunda, W. G. (2003). Iron requirements for dinitrogen- and ammonium-

93

supported growth in cultures of Trichodesmium (IMS 101): Comparison with nitrogen fixation rates and iron: carbon ratios of field populations. Limnology and Oceanography, 48(5), 1869–1884. https://doi.org/10.4319/lo.2003.48.5.1869

Landolfi, A., Kähler, P., Koeve, W., & Oschlies, A. (2018). Global marine N2 fixation estimates: From observations to models. Frontiers in Microbiology, 9. https://doi.org/10.3389/fmicb.2018.02112

Landolfi, A., Koeve, W., Dietze, H., Kähler, P., & Oschlies, A. (2015). A new perspective on environmental controls of marine nitrogen fixation. Geophysical Research Letters, 42(11), 2015GL063756. https://doi.org/10.1002/2015GL063756

Langlois, R., Großkopf, T., Mills, M., Takeda, S., & LaRoche, J. (2015). Widespread distribution and expression of Gamma A (UMB), an uncultured, diazotrophic, γ-proteobacterial nifH phylotype. PLOS ONE, 10(6), e0128912. https://doi.org/10.1371/journal.pone.0128912

Langlois, R. J., Hümmer, D., & LaRoche, J. (2008). Abundances and distributions of the dominant nifH phylotypes in the northern Atlantic Ocean. Applied and

Environmental Microbiology, 74(6), 1922–1931. https://doi.org/10.1128/AEM.01720-07

Langlois, R. J., LaRoche, J., & Raab, P. A. (2005). Diazotrophic diversity and distribution in the tropical and subtropical Atlantic Ocean. Applied and

Environmental Microbiology, 71(12), 7910–7919. https://doi.org/10.1128/AEM.71.12.7910-7919.2005

Langlois, R. J., Mills, M. M., Ridame, C., Croot, P., & LaRoche, J. (2012). Diazotrophic bacteria respond to Saharan dust additions. Marine Ecology

Progress Series, 470, 1–14. https://doi.org/10.3354/meps10109 Lavery, T. J., Roudnew, B., Gill, P., Seymour, J., Seuront, L., Johnson, G., et al.

(2010). Iron defecation by sperm whales stimulates carbon export in the Southern Ocean. Proceedings of the Royal Society B: Biological Sciences, 277(1699), 3527–3531. https://doi.org/10.1098/rspb.2010.0863

Lefcheck, J. S. (2016). piecewiseSEM: Piecewise structural equation modelling in R for ecology, evolution, and systematics. Methods in Ecology and

Evolution, 7(5), 573–579. https://doi.org/10.1111/2041-210X.12512 Lefcheck, J. S., & Duffy, J. E. (2015). Multitrophic functional diversity predicts

ecosystem functioning in experimental assemblages of estuarine consumers. Ecology, 96(11), 2973–2983. https://doi.org/10.1890/14-1977.1

Lehtimaki, J., Moisander, P., Sivonen, K., & Kononen, K. (1997). Growth, nitrogen fixation, and nodularin production by two Baltic Sea cyanobacteria. Applied and Environmental Microbiology, 63(5), 1647–1656.

94

Leigh, J. A. (2000). Nitrogen fixation in methanogens - the archaeal perspective. Prokaryotic Nitrogen Fixation: A Model System for the Analysis of a

Biological Processes, 657–669. Lemmermann, E. (1905). Die algenflora der sandwich-inseln. ergebnisse einer

reise nach dem Pacific. H. Schauinsland 1896/97. Bot. Jahrbuch Syst.

Pflanzengesch. Pflanzengeogr. (34), 607–663. Ley, P. (2019). Impact of nutrient deficiency on diatom-diazotroph associations

(DDAs) abundance, gene expression, and activity (Master thesis). Stockholm

University, Stockholm. Lohrenz, S.E., Knauer, G.A., Asper, V.L., Tuel, M., Michaels, A.F., & Knap, A.

H. (1992) Seasonal variability in primary production and particle flux in the northwestern Sargasso Sea: U.S. JGOFS Bermuda Atlantic time-series study. Deep Sea Research Part A. Oceanographic Research Papers, 39(7), 1373-1391.

Luo, G., Junium, C. K., Izon, G., Ono, S., Beukes, N. J., Algeo, T. J., et al. (2018). Nitrogen fixation sustained productivity in the wake of the Palaeoproterozoic great oxygenation event. Nature Communications, 9(1), 1–9. https://doi.org/10.1038/s41467-018-03361-2

Luo, Y.-W., Doney, S. C., Anderson, L. A., Benavides, M., Berman-Frank, I., Bode, A., et al. (2012). Database of diazotrophs in global ocean: abundance, biomass and nitrogen fixation rates. Earth System Science Data, 4(1), 47–73. http://dx.doi.org/10.5194/essd-4-47-2012

Mague, T. H., Weare, N. M., & Holm-Hansen, O. (1974). Nitrogen fixation in the North Pacific Ocean. Marine Biology, 24(2), 109–119. https://doi.org/10.1007/BF00389344

Man‐Aharonovich, D., Kress, N., Zeev, E. B., Berman‐Frank, I., & Béjà, O. (2007). Molecular ecology of nifH genes and transcripts in the eastern Mediterranean Sea. Environmental Microbiology, 9(9), 2354–2363. https://doi.org/10.1111/j.1462-2920.2007.01353.x

Marchetti, A., Schruth, D. M., Durkin, C. A., Parker, M. S., Kodner, R. B., Berthiaume, C. T., et al. (2012). Comparative metatranscriptomics identifies molecular bases for the physiological responses of phytoplankton to varying iron availability. Proceedings of the National Academy of Sciences, 109(6), E317–E325. https://doi.org/10.1073/pnas.1118408109

Martin, W., Rujan, T., Richly, E., Hansen, A., Cornelsen, S., Lins, T., et al. (2002). Evolutionary analysis of Arabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phylogeny and thousands of cyanobacterial genes in the nucleus. Proceedings of the National Academy of Sciences of the United

States of America, 99(19), 12246–12251. https://doi.org/10.1073/pnas.182432999

95

Martínez-Pérez, C., Mohr, W., Löscher, C. R., Dekaezemacker, J., Littmann, S., Yilmaz, P., et al. (2016). The small unicellular diazotrophic symbiont, UCYN-A, is a key player in the marine nitrogen cycle. Nature Microbiology, 1(11), 16163. https://doi.org/10.1038/nmicrobiol.2016.163

Massana, R., Murray, A. E., Preston, C. M., & DeLong, E. F. (1997). Vertical distribution and phylogenetic characterization of marine planktonic Archaea in the Santa Barbara Channel. Applied and Environmental Microbiology, 63(1), 50–56.

Mazard, S. L., Fuller, N. J., Orcutt, K. M., Bridle, O., & Scanlan, D. J. (2004). PCR analysis of the distribution of unicellular cyanobacterial diazotrophs in the Arabian Sea. Applied and Environmental Microbiology, 70(12), 7355–7364. https://doi.org/10.1128/AEM.70.12.7355-7364.2004

McRose, D. L., Zhang, X., Kraepiel, A. M. L., & Morel, F. M. M. (2017). Diversity and activity of alternative nitrogenases in sequenced genomes and coastal environments. Frontiers in Microbiology, 8. https://doi.org/10.3389/fmicb.2017.00267

Messer, L. F., Doubell, M., Jeffries, T. C., Brown, M. V., & Seymour, J. R. (2015). Prokaryotic and diazotrophic population dynamics within a large oligotrophic inverse estuary. Aquatic Microbial Ecology, 74(1), 1–15. https://doi.org/10.3354/ame01726

Messer, L. F., Mahaffey, C., Robinson, C. M., Jeffries, T. C., Baker, K. G., Bibiloni Isaksson, J., et al. (2016). High levels of heterogeneity in diazotroph diversity and activity within a putative hotspot for marine nitrogen fixation. The ISME Journal, 10(6), 1499–1513. https://doi.org/10.1038/ismej.2015.205

Miller, R. W., & Eady, R. R. (1988). Molybdenum and vanadium nitrogenases of Azotobacter chroococcum. Low temperature favours N2 reduction by vanadium nitrogenase. Biochemical Journal, 256(2), 429–432. https://doi.org/10.1042/bj2560429

Mills, M. M., Ridame, C., Davey, M., La Roche, J., & Geider, R. J. (2004). Iron and phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature, 429(6989), 292–294. https://doi.org/10.1038/nature02550

Mitsui, A., Kumazawa, S., Takahashi, A., Ikemoto, H., Cao, S., & Arai, T. (1986). Strategy by which nitrogen-fixing unicellular cyanobacteria grow photoautotrophically. Nature, 323, 720–722. https://doi.org/10.1038/323720a0

Mohr, W., Großkopf, T., Wallace, D. W. R., & LaRoche, J. (2010). Methodological underestimation of oceanic nitrogen fixation rates. PLOS

ONE, 5(9), e12583. https://doi.org/10.1371/journal.pone.0012583

96

Moisander, P. H., Beinart, R. A., Hewson, I., White, A. E., Johnson, K. S., Carlson, C. A., et al. (2010). Unicellular cyanobacterial distributions broaden the oceanic N2 fixation domain. Science, 327(5972), 1512–1514. https://doi.org/10.1126/science.1185468

Moisander, P. H., Beinart, R. A., Voss, M., & Zehr, J. P. (2008). Diversity and abundance of diazotrophic microorganisms in the South China Sea during intermonsoon. The ISME Journal, 2(9), 954–967. https://doi.org/10.1038/ismej.2008.51

Moisander, P. H., Benavides, M., Bonnet, S., Berman-Frank, I., White, A. E., & Riemann, L. (2017). Chasing after non-cyanobacterial nitrogen fixation in marine pelagic environments. Frontiers in Microbiology, 8. https://doi.org/10.3389/fmicb.2017.01736

Moisander, P. H., Shiue, L., Steward, G. F., Jenkins, B. D., Bebout, B. M., & Zehr, J. P. (2006). Application of a nifH oligonucleotide microarray for profiling diversity of N2-fixing microorganisms in marine microbial mats. Environmental Microbiology, 8(10), 1721–1735. https://doi.org/10.1111/j.1462-2920.2006.01108.x

Momper, L. M., Reese, B. K., Carvalho, G., Lee, P., & Webb, E. A. (2015). A novel cohabitation between two diazotrophic cyanobacteria in the oligotrophic ocean. The ISME Journal, 9(4), 882–893. https://doi.org/10.1038/ismej.2014.186

Moran, N. A., McCutcheon, J. P., & Nakabachi, A. (2008). Genomics and evolution of heritable bacterial symbionts. Annual Review of Genetics, 42(1), 165–190. https://doi.org/10.1146/annurev.genet.41.110306.130119

Morrissey, J., & Bowler, C. (2012). Iron utilization in marine cyanobacteria and eukaryotic algae. Frontiers in Microbiology, 3. https://doi.org/10.3389/fmicb.2012.00043

Motwani, N. H., Duberg, J., Svedén, J. B., & Gorokhova, E. (2018). Grazing on cyanobacteria and transfer of diazotrophic nitrogen to zooplankton in the Baltic Sea. Limnology and Oceanography, 63(2), 672–686. https://doi.org/10.1002/lno.10659

Moutin, T., Doglioli, A. M., Verneil, A. de, & Bonnet, S. (2017). Preface: The Oligotrophy to the UlTra-oligotrophy PACific Experiment (OUTPACE cruise, 18 February to 3 April 2015). Biogeosciences, 14, 3207–3220. https://doi.org/10.5194/bg-14-3207-2017

Mueller, O., Lightfoot, S., & Schroeder, A. (2004). RNA integrity number (RIN)–standardization of RNA quality control. Agilent Application Note, publication 1, 1–8.

97

Mulholland, M. R., Bernhardt, P. W., Blanco-Garcia, J. L., Mannino, A., Hyde, K., Mondragon, E., et al. (2012). Rates of dinitrogen fixation and the abundance of diazotrophs in North American coastal waters between Cape Hatteras and Georges Bank. Limnology and Oceanography, 57(4), 1067–1083. https://doi.org/10.4319/lo.2012.57.4.1067

Mulholland, M. R., Bernhardt, P. W., Heil, C. A., Bronk, D. A., & O’Neil, J. M. (2006). Nitrogen fixation and release of fixed nitrogen by Trichodesmium spp. in the Gulf of Mexico. Limnology and Oceanography, 51(4), 1762–1776. https://doi.org/10.4319/lo.2006.51.4.1762

Mulholland, M. R., Bernhardt, P. W., Widner, B. N., Selden, C. R., Chappell, P. D., Clayton, S., et al. (2019). High rates of N2 fixation in temperate, western North Atlantic coastal waters expand the realm of marine diazotrophy. Global Biogeochemical Cycles, 33(7), 826–840. https://doi.org/10.1029/2018GB006130

Mulholland, M. R., Heil, C. A., Bronk, D. A., O’Neil, J. M., & Bernhardt, P. (2002). Does nitrogen regeneration from the N2 fixing cyanobacteria Trichodesmium spp. fuel Karenia brevis blooms in the Gulf of Mexico? Harmful Algae, 47–49.

Muller-Karger, F. E., McClain, C. R., & Richardson, P. L. (1988). The dispersal of the Amazon’s water. Nature, 333(6168), 56–59. https://doi.org/10.1038/333056a0

Muñoz-Marín, M. del C., Shilova, I. N., Shi, T., Farnelid, H., Cabello, A. M., & Zehr, J. P. (2019). The transcriptional cycle is suited to daytime N2 fixation in the unicellular cyanobacterium “Candidatus Atelocyanobacterium

thalassa” (UCYN-A). mBio, 10(1). https://doi.org/10.1128/mBio.02495-18 Nakayama, T., Kamikawa, R., Tanifuji, G., Kashiyama, Y., Ohkouchi, N.,

Archibald, J. M., & Inagaki, Y. (2014). Complete genome of a nonphotosynthetic cyanobacterium in a diatom reveals recent adaptations to an intracellular lifestyle. Proceedings of the National Academy of Sciences, 111(31), 11407–11412. https://doi.org/10.1073/pnas.1405222111

Needoba, J. A., Foster, R. A., Sakamoto, C., Zehr, J. P., & Johnson, K. S. (2007). Nitrogen fixation by unicellular diazotrophic cyanobacteria in the temperate oligotrophic North Pacific Ocean. Limnology and Oceanography, 52(4), 1317–1327. https://doi.org/10.4319/lo.2007.52.4.1317

Nguyen, N. A. (2017). Historic drought and salinity intrusion in the Mekong delta in 2016: Lessons learned and response solutions. Vietnam Journal of Science,

Technology and Engineering, 59(1), 93–96. Norris, R. E. (1961). Observations on phytoplankton organisms collected on the

NZOI Pacific cruise. New Zeeland Journal of Science, 1961(4), 162–188.

98

Nowack, E. C. M., Melkonian, M., & Glöckner, G. (2008). Chromatophore genome sequence of Paulinella sheds light on acquisition of photosynthesis by eukaryotes. Current Biology, 18(6), 410–418. https://doi.org/10.1016/j.cub.2008.02.051

Nowack, E. C. M., Price, D. C., Bhattacharya, D., Singer, A., Melkonian, M. & Grossman, A. R. (2016). Gene transfers from diverse bacteria compensate for reductive genome evolution in the chromatophore of Paulinella chromatophora, Proceedings of the National Academy of Sciences, 113(43), 12214–12219, doi:10.1073/pnas.1608016113.

O’Connell, G. C., Chantler, P. D., & Barr, T. L. (2017). High interspecimen variability in nucleic acid extraction efficiency necessitates the use of spike-in control for accurate qPCR-based measurement of plasma cell-free DNA levels. Laboratory Medicine, 48(4), 332–338. https://doi.org/10.1093/labmed/lmx043

O’Neil, J. M. (1998). The colonial cyanobacterium Trichodesmium as a physical and nutritional substrate for the harpacticoid copepod Macrosetella gracilis. Journal of Plankton Research, 20(1), 43–59. https://doi.org/10.1093/plankt/20.1.43

Ottesen, E. A., Young, C. R., Eppley, J. M., Ryan, J. P., Chavez, F. P., Scholin, C. A., & DeLong, E. F. (2013). Pattern and synchrony of gene expression among sympatric marine microbial populations. Proceedings of the National

Academy of Sciences, 110(6), E488–E497. https://doi.org/10.1073/pnas.1222099110

Pabortsava, K., Lampitt, R. S., Benson, J., Crowe, C., McLachlan, R., Le Moigne, F. A. C., et al. (2017). Carbon sequestration in the deep Atlantic enhanced by Saharan dust. Nature Geoscience, 10(3), 189–194. https://doi.org/10.1038/ngeo2899

Paerly, H. W., Bebout, B. M., & Prufert, L. E. (1989). Bacterial associations with marine Oscillatoria sp. (Trichodesmium sp.) populations: Ecophysiological implications1. Journal of Phycology, 25(4), 773–784. https://doi.org/10.1111/j.0022-3646.1989.00773.x

Paulsen, D. M., Paerl, H. W., & Bishop, P. E. (1991). Evidence that molybdenum-dependent nitrogen fixation is not limited by high sulfate concentrations in marine environments. Limnology and Oceanography, 36(7), 1325–1334. https://doi.org/10.4319/lo.1991.36.7.1325

Peoples, M. B., & Craswell, E. T. (1992). Biological nitrogen fixation: Investments, expectations and actual contributions to agriculture. Plant and

Soil, 141(1), 13–39. https://doi.org/10.1007/BF00011308

99

Pesant, S., Not, F., Picheral, M., Kandels-Lewis, S., Le Bescot, N., Gorsky, G., et al. (2015). Open science resources for the discovery and analysis of Tara Oceans data, Scientific Data, 2(1), 1–16, doi:10.1038/sdata.2015.23

Preston, C. M., Harris, A., Ryan, J. P., Roman, B., Iii, R. M., Jensen, S., et al. (2011). Underwater application of quantitative PCR on an ocean mooring. PLOS ONE, 6(8). https://doi.org/10.1371/journal.pone.0022522

Prufert-Bebout, L., Paerl, H. W., & Lassen, C. (1993). Growth, nitrogen fixation, and spectral attenuation in cultivated Trichodesmium species. Applied and

Environmental Microbiology, 59(5), 1367–1375. Raven, J. A. (1988). The iron and molybdenum use efficiencies of plant growth

with different energy, carbon and nitrogen sources. New Phytologist, 109(3), 279–287. https://doi.org/10.1111/j.1469-8137.1988.tb04196.x

Raven, J. A. (1990). Predictions of Mn and Fe use efficiencies of phototrophic growth as a function of light availability for growth and of C assimilation pathway. New Phytologist, 116(1), 1–18. https://doi.org/10.1111/j.1469-8137.1990.tb00505.x

Raven, J. A., & Allen, J. F. (2003). Genomics and chloroplast evolution: what did cyanobacteria do for plants? Genome Biology, 4(3), 209. https://doi.org/10.1186/gb-2003-4-3-209

Reddy, K. J., Haskell, J. B., Sherman, D. M., & Sherman, L. A. (1993). Unicellular, aerobic nitrogen-fixing cyanobacteria of the genus Cyanothece. Journal of Bacteriology, 175(5), 1284–1292. https://doi.org/10.1128/jb.175.5.1284-1292.1993

Riemann, L., Farnelid, H., & Steward, G. F. (2010). Nitrogenase genes in non-cyanobacterial plankton: Prevalence, diversity and regulation in marine waters. Aquatic Microbial Ecology, 61(3), 235–247. https://doi.org/10.3354/ame01431

Rivera, M. (2017). Bacterioferritin: Structure, dynamics, and protein-protein interactions at play in iron storage and mobilization. Accounts of Chemical

Research, 50(2), 331–340. https://doi.org/10.1021/acs.accounts.6b00514 Robidart, J. C., Church, M. J., Ryan, J. P., Ascani, F., Wilson, S. T., Bombar, D.,

et al. (2014). Ecogenomic sensor reveals controls on N2-fixing microorganisms in the North Pacific Ocean. The ISME Journal, 8(6), 1175–1185. https://doi.org/10.1038/ismej.2013.244

Robidart, J. C., Magasin, J. D., Shilova, I. N., Turk-Kubo, K. A., Wilson, S. T., Karl, D. M., et al. (2019). Effects of nutrient enrichment on surface microbial community gene expression in the oligotrophic North Pacific Subtropical Gyre. The ISME Journal, 13(2), 374–387. https://doi.org/10.1038/s41396-018-0280-0

100

Rubin, M., Berman-Frank, I., & Shaked, Y. (2011). Dust- and mineral-iron utilization by the marine dinitrogen-fixer Trichodesmium. Nature

Geoscience, 4(8), 529–534. https://doi.org/10.1038/ngeo1181 Sachs, J. P., & Repeta, D. J. (1999). Oligotrophy and nitrogen fixation during

eastern Mediterranean sapropel events. Science, 286(5449), 2485–2488. https://doi.org/10.1126/science.286.5449.2485

Saito, M. A., Bertrand, E. M., Dutkiewicz, S., Bulygin, V. V., Moran, D. M., Monteiro, F. M., et al. (2011). Iron conservation by reduction of metalloenzyme inventories in the marine diazotroph Crocosphaera watsonii. Proceedings of the National Academy of Sciences, 108(6), 2184–2189. https://doi.org/10.1073/pnas.1006943108

Sañudo-Wilhelmy, S. A., Kustka, A. B., Gobler, C. J., Hutchins, D. A., Yang, M., Lwiza, K., et al. (2001). Phosphorus limitation of nitrogen fixation by Trichodesmium in the central Atlantic Ocean. Nature, 411(6833), 66–69. https://doi.org/10.1038/35075041

Sellner, K. G. (1997). Physiology, ecology, and toxic properties of marine cyanobacteria blooms. Limnology and Oceanography, 42(5), 1089–1104.

Shi, T., Ilikchyan, I., Rabouille, S., & Zehr, J. P. (2010). Genome-wide analysis of diel gene expression in the unicellular N2-fixing cyanobacterium Crocosphaera watsonii WH 8501. The ISME Journal, 4(5), 621–632. https://doi.org/10.1038/ismej.2009.148

Shi, T., Sun, Y., & Falkowski, P. G. (2007). Effects of iron limitation on the expression of metabolic genes in the marine cyanobacterium Trichodesmium erythraeum IMS101. Environmental Microbiology, 9(12), 2945–2956. https://doi.org/10.1111/j.1462-2920.2007.01406.x

Shilova, I. N., Robidart, J. C., DeLong, E. F., & Zehr, J. P. (2016). Genetic diversity affects the daily transcriptional oscillations of marine microbial populations. PLOS ONE, 11(1). https://doi.org/10.1371/journal.pone.0146706

Shilova, I. N., Robidart, J. C., James Tripp, H., Turk-Kubo, K., Wawrik, B., Post, A. F., et al. (2014). A microarray for assessing transcription from pelagic marine microbial taxa. The ISME Journal, 8(7), 1476–1491. https://doi.org/10.1038/ismej.2014.1

Shiozaki, T., Bombar, D., Riemann, L., Hashihama, F., Takeda, S., Yamaguchi, T., et al. (2017). Basin scale variability of active diazotrophs and nitrogen fixation in the North Pacific, from the tropics to the subarctic Bering Sea. Global Biogeochemical Cycles, 31(6), 996–1009. https://doi.org/10.1002/2017GB005681

101

Shiozaki, T., Fujiwara, A., Ijichi, M., Harada, N., Nishino, S., Nishi, S., et al. (2018). Diazotroph community structure and the role of nitrogen fixation in the nitrogen cycle in the Chukchi Sea (western Arctic Ocean). Limnology

and Oceanography, 63(5), 2191–2205. https://doi.org/10.1002/lno.10933 Shiozaki, T., Ijichi, M., Kodama, T., Takeda, S., & Furuya, K. (2014).

Heterotrophic bacteria as major nitrogen fixers in the euphotic zone of the Indian Ocean. Global Biogeochemical Cycles, 28(10), 1096–1110. https://doi.org/10.1002/2014GB004886

Short, S. M., & Zehr, J. P. (2007). Nitrogenase gene expression in the Chesapeake Bay estuary. Environmental Microbiology, 9(6), 1591–1596. https://doi.org/10.1111/j.1462-2920.2007.01258.x

Sivonen, K., Kononen, K., Carmichael, W. W., Dahlem, A. M., Rinehart, K. L., Kiviranta, J., & Niemela, S. I. (1989). Occurrence of the hepatotoxic cyanobacterium Nodularia spumigena in the Baltic Sea and structure of the toxin. Applied and Environmental Microbiology, 55(8), 1990–1995.

Sohm, J. A., Webb, E. A., & Capone, D. G. (2011). Emerging patterns of marine nitrogen fixation. Nature Reviews Microbiology, 9(7), 499–508. https://doi.org/10.1038/nrmicro2594

Solomon, S., Qin, D., Manning, M., Averyt, K., Marquis, M., & Tignor, M. M. (2007). Climate change 2007 - The physical science basis: Working group I contribution to the fourth assessment report of the IPCC. Cambridge

University Press. Stenegren, M., Berg, C., Padilla, C. C., David, S.-S., Montoya, J. P., Yager, P. L.,

& Foster, R. A. (2017). Piecewise structural equation model (SEM) disentangles the environmental conditions favoring diatom diazotroph associations (DDAs) in the western tropical North Atlantic (WTNA). Frontiers in Microbiology, 8. https://doi.org/10.3389/fmicb.2017.00810

Stenegren, M., Caputo, A., Berg, C., Bonnet, S., & Foster, R. A. (2018). Distribution and drivers of symbiotic and free-living diazotrophic cyanobacteria in the western tropical South Pacific. Biogeosciences, 15(5), 1559–1578. https://doi.org/https://doi.org/10.5194/bg-15-1559-2018

Stewart, W. D. P. (1969). Biological and ecological aspects of nitrogen fixation by free-living micro-organisms. Proceedings of the Royal Society of London.

Series B, Biological Sciences, 172(1029), 367–388. Stöckel, J., Welsh, E. A., Liberton, M., Kunnvakkam, R., Aurora, R., & Pakrasi, H.

B. (2008). Global transcriptomic analysis of Cyanothece 51142 reveals robust diurnal oscillation of central metabolic processes. Proceedings of the

National Academy of Sciences, 105(16), 6156–6161. https://doi.org/10.1073/pnas.0711068105

102

Subramaniam, A., Yager, P. L., Carpenter, E. J., Mahaffey, C., Björkman, K., Cooley, S., et al. (2008). Amazon River enhances diazotrophy and carbon sequestration in the tropical North Atlantic Ocean. Proceedings of the

National Academy of Sciences, 105(30), 10460–10465. https://doi.org/10.1073/pnas.0710279105

Sunda, W. G., & Huntsman, S. A. (2015). High iron requirement for growth, photosynthesis, and low-light acclimation in the coastal cyanobacterium Synechococcus bacillaris. Frontiers in Microbiology, 6. https://doi.org/10.3389/fmicb.2015.00561

Tagliabue, A., Bopp, L., Dutay, J.-C., Bowie, A. R., Chever, F., Jean-Baptiste, P., et al. (2010). Hydrothermal contribution to the oceanic dissolved iron inventory. Nature Geoscience, 3(4), 252–256. https://doi.org/10.1038/ngeo818

Taniuchi, Y., Chen, Y. L., Chen, H.-Y., Tsai, M.-L., & Ohki, K. (2012). Isolation and characterization of the unicellular diazotrophic cyanobacterium Group C TW3 from the tropical western Pacific Ocean. Environmental Microbiology, 14(3), 641–654. https://doi.org/10.1111/j.1462-2920.2011.02606.x

Thiel, T., Pratte, B. S., Zhong, J., Goodwin, L., Copeland, A., Lucas, S., et al. (2014). Complete genome sequence of Anabaena variabilis ATCC 29413. Standards in Genomic Sciences, 9(3), 562–573. https://doi.org/10.4056/sigs.3899418

Thiele, S., Fuchs, B. M., Amann, R., & Iversen, M. H. (2015). Colonization in the photic zone and subsequent changes during sinking determine bacterial community composition in marine snow. Applied and Environmental

Microbiology, 81(4), 1463–1471. https://doi.org/10.1128/AEM.02570-14 Thompson, A., Carter, B. J., Turk-Kubo, K., Malfatti, F., Azam, F., & Zehr, J. P.

(2014). Genetic diversity of the unicellular nitrogen-fixing cyanobacteria UCYN-A and its prymnesiophyte host. Environmental Microbiology, 16(10), 3238–3249. https://doi.org/10.1111/1462-2920.12490

Thompson, A. W., Foster, R. A., Krupke, A., Carter, B. J., Musat, N., Vaulot, D., et al. (2012). Unicellular cyanobacterium symbiotic with a single-celled eukaryotic alga. Science, 337(6101), 1546–1550. https://doi.org/10.1126/science.1222700

Thompson, A. W., & Zehr, J. P. (2013). Cellular interactions: lessons from the nitrogen-fixing cyanobacteria. Journal of Phycology, 49(6), 1024–1035. https://doi.org/10.1111/jpy.12117

Trapp, E. M., Adler, S., Zau-ner, S., & Maier, U.-G. (2012). Rhopalodia gibba and its endosymbionts as a model for early steps in a cyanobacterial primary endosymbiosis. Endocytobiosis & Cell Research, 23, 21–24.

103

Tréguer, P. J., & De La Rocha, C. L. (2013). The world ocean silica cycle. Annual

Review of Marine Science, 5(1), 477–501. https://doi.org/10.1146/annurev-marine-121211-172346

Tripp, H. J., Bench, S. R., Turk, K. A., Foster, R. A., Desany, B. A., Niazi, F., et al. (2010). Metabolic streamlining in an open-ocean nitrogen-fixing cyanobacterium. Nature, 464(7285), 90–94. https://doi.org/10.1038/nature08786

Tuit, C., Waterbury, J., & Ravizza, G. (2004). Diel variation of molybdenum and iron in marine diazotrophic cyanobacteria. Limnology and Oceanography, 49(4), 978–990. https://doi.org/10.4319/lo.2004.49.4.0978

Turk‐Kubo, K. A., Farnelid, H. M., Shilova, I. N., Henke, B., & Zehr, J. P. (2017). Distinct ecological niches of marine symbiotic N2-fixing cyanobacterium Candidatus Atelocyanobacterium thalassa sublineages. Journal of Phycology, 53(2), 451–461. https://doi.org/10.1111/jpy.12505

Turk-Kubo, K. A., Frank, I. E., Hogan, M. E., Desnues, A., Bonnet, S., & Zehr, J. P. (2015). Diazotroph community succession during the VAHINE mesocosm experiment (New Caledonia lagoon). Biogeosciences, 12(24), 7435–7452. https://doi.org/https://doi.org/10.5194/bg-12-7435-2015

Tzubari, Y., Magnezi, L., Be’er, A., & Berman-Frank, I. (2018). Iron and phosphorus deprivation induce sociality in the marine bloom-forming cyanobacterium Trichodesmium. The ISME Journal, 12(7), 1682–1693. https://doi.org/10.1038/s41396-018-0073-5

Venrick, E. L. (1974). The distribution and significance of Richelia intracellularis Schmidt in the North Pacific central gyre. Limnology and Oceanography, 19(3), 437–445. https://doi.org/10.4319/lo.1974.19.3.0437

Villareal, T. A. (1989). Division cycles in the nitrogen-fixing Rhizosolenia (Bacillariophyceae)-Richelia (Nostocaceae) symbiosis. British Phycological

Journal, 24(4), 357–365. https://doi.org/10.1080/00071618900650371 Villareal, T. A. (1990). Laboratory culture and preliminary characterization of the

nitrogen-fixing Rhizosolenia-Richelia symbiosis. Marine Ecology, 11(2), 117–132. https://doi.org/10.1111/j.1439-0485.1990.tb00233.x

Vitousek, P. M., Menge, D. N. L., Reed, S. C., & Cleveland, C. C. (2013). Biological nitrogen fixation: Rates, patterns and ecological controls in terrestrial ecosystems. Philosophical Transactions of the Royal Society B:

Biological Sciences, 368(1621). https://doi.org/10.1098/rstb.2013.0119 Volk, T., & Hoffert, M. I. (1985). Ocean carbon pumps: Analysis of relative

strengths and efficiencies in ocean-driven atmospheric CO2 changes. In E. T. Sundquist & W. S. Broecker (Eds.), The Carbon Cycle and Atmospheric

104

CO2: Natural Variations Archean to Present (pp. 99–110). Retrieved from http://onlinelibrary.wiley.com/doi/10.1029/GM032p0099/summary

Walworth, N. G., Fu, F.-X., Webb, E. A., Saito, M. A., Moran, D., Mcllvin, M. R., et al. (2016). Mechanisms of increased Trichodesmium fitness under iron and phosphorus co-limitation in the present and future ocean. Nature

Communications, 7(1), 1–11. https://doi.org/10.1038/ncomms12081 Wang, W.-L., Moore, J. K., Martiny, A. C., & Primeau, F. W. (2019). Convergent

estimates of marine nitrogen fixation. Nature, 566(7743), 205–211. https://doi.org/10.1038/s41586-019-0911-2

Webb, E. A., Ehrenreich, I. M., Brown, S. L., Valois, F. W., & Waterbury, J. B. (2009). Phenotypic and genotypic characterization of multiple strains of the diazotrophic cyanobacterium, Crocosphaera watsonii, isolated from the open ocean. Environmental Microbiology, 11(2), 338–348. https://doi.org/10.1111/j.1462-2920.2008.01771.x

Wenke, B. B. (2019). The many roles of the nitrogenase iron protein (PhD, California Institute of Technology). Retrieved from http://resolver.caltech.edu/CaltechTHESIS:04222019-175445561

White, A. E. (2012). The trouble with the bubble. Nature, 488(7411), 290–291. https://doi.org/10.1038/nature11481

White, A. E., Foster, R. A., Benitez-Nelson, C. R., Masqué, P., Verdeny, E., Popp, B. N., et al. (2013). Nitrogen fixation in the Gulf of California and the eastern tropical North Pacific. Progress in Oceanography, 109, 1–17. https://doi.org/10.1016/j.pocean.2012.09.002

Wilson, S. T., Aylward, F. O., Ribalet, F., Barone, B., Casey, J. R., Connell, P. E., et al. (2017). Coordinated regulation of growth, activity and transcription in natural populations of the unicellular nitrogen-fixing cyanobacterium Crocosphaera. Nature Microbiology, 2(9), 17118. https://doi.org/10.1038/nmicrobiol.2017.118

Wilson, S. T., Böttjer, D., Church, M. J., & Karl, D. M. (2012). Comparative assessment of nitrogen fixation methodologies, conducted in the oligotrophic North Pacific Ocean. Applied and Environmental Microbiology, 78(18), 6516–6523. https://doi.org/10.1128/AEM.01146-12

Wolf‐Gladrow, D. A., Riebesell, U., Burkhardt, S., & Bijma, J. (1999). Direct effects of CO2 concentration on growth and isotopic composition of marine plankton. Tellus B, 51(2), 461–476. https://doi.org/10.1034/j.1600-0889.1999.00023.x

Wu, C., Kan, J., Liu, H., Pujari, L., Guo, C., Wang, X., & Sun, J. (2019). Heterotrophic bacteria dominate the diazotrophic community in the eastern

105

Indian Ocean (EIO) during pre-southwest monsoon. Microbial Ecology. https://doi.org/10.1007/s00248-019-01355-1

Wu, J., Chung, S.-W., Wen, L.-S., Liu, K.-K., Chen, Y. L., Chen, H.-Y., & Karl, D. M. (2003). Dissolved inorganic phosphorus, dissolved iron, and Trichodesmium in the oligotrophic South China Sea. Global Biogeochemical

Cycles, 17(1), 8-1-8–10. https://doi.org/10.1029/2002GB001924 Young, C. L., & Ingall, E. D. (2010). Marine dissolved organic phosphorus

composition: Insights from samples recovered using combined electrodialysis/reverse osmosis. Aquatic Geochemistry, 16(4), 563–574. https://doi.org/10.1007/s10498-009-9087-y

Young, J. P. W. (2005). The phylogeny and evolution of nitrogenases. In R. Palacios & W. E. Newton (Eds.), Genomes and Genomics of Nitrogen-fixing

Organisms (pp. 221–241). Retrieved from http://link.springer.com/chapter/10.1007/1-4020-3054-1_14

Yurchenko, T., Ševčíková, T., Přibyl, P., Karkouri, K. E., Klimeš, V., Amaral, R., et al. (2018). A gene transfer event suggests a long-term partnership between eustigmatophyte algae and a novel lineage of endosymbiotic bacteria. The

ISME Journal, 12(9), 2163–2175.https://doi.org/10.1038/s41396-018-0177-y Zehr, J. P., Bench, S. R., Carter, B. J., Hewson, I., Niazi, F., Shi, T., et al. (2008).

Globally distributed uncultivated oceanic N2-fixing cyanobacteria lack oxygenic photosystem II. Science, 322(5904), 1110–1112. https://doi.org/10.1126/science.1165340

Zehr, J. P., Bench, S. R., Mondragon, E. A., McCarren, J., & DeLong, E. F. (2007). Low genomic diversity in tropical oceanic N2-fixing cyanobacteria. Proceedings of the National Academy of Science, 104, 17807–17812. https://doi.org/10.1073/pnas.0701017104

Zehr, J. P., Jenkins, B. D., Short, S. M., & Steward, G. F. (2003). Nitrogenase gene diversity and microbial community structure: a cross-system comparison. Environmental Microbiology, 5(7), 539–554. https://doi.org/10.1046/j.1462-2920.2003.00451.x

Zehr, J. P., Mellon, M., Braun, S., Litaker, W., Steppe, T., & Paerl, H. W. (1995). Diversity of heterotrophic nitrogen fixation genes in a marine cyanobacterial mat. Applied and Environmental Microbiology, 61(7), 2527–2532.

Zehr, J. P., Mellon, M. T., & Zani, S. (1998). New nitrogen-fixing microorganisms detected in oligotrophic oceans by amplification of nitrogenase (nifH) genes. Applied and Environmental Microbiology, 64(9), 3444–3450.