University of Southern California Dissertations and Theses

Genetic characterization of microbial eukaryotic diversity and metabolic potential

(USC Thesis Other)

Genetic characterization of microbial eukaryotic diversity and metabolic potential

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Genetic characterization of microbial eukaryotic
diversity and metabolic potential

Sarah K. Hu
A Dissertation Presented to the Faculty of the
USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirement for the Degree
DOCTOR OF PHILOSOPHY
BIOLOGICAL SCIENCES
MAY 2018
1

Approved by Advisory Committee:
David A. Caron (Chair)
Eric A. Webb
Karla Heidelberg
William Berelson

2
Table of Contents
Table of Contents ............................................................................................................................ 2
Acknowledgements ......................................................................................................................... 4
Dissertation Abstract ....................................................................................................................... 7
Dissertation Introduction ................................................................................................................ 8
Introduction References ................................................................................................................ 13
Chapter One: Estimating Protistan Diversity using High-throughput Sequencing ...................... 15
Abstract ............................................................................................................................................. 16
1.1 Introduction ................................................................................................................................. 17
1.2 Materials and Methods ................................................................................................................ 19
1.3 Results and Discussion ............................................................................................................... 22
1.4 Chapter One Figures and Tables ................................................................................................ 26
1.5 Chapter One Literature Cited ..................................................................................................... 27
Chapter Two: Protistan diversity and activity inferred from RNA and DNA at a coastal ocean
site in the eastern North Pacific .................................................................................................... 28
Abstract ............................................................................................................................................. 29
2.1 Introduction ................................................................................................................................. 30
2.2 Materials and Methods ............................................................................................................... 32
2.3 Results ......................................................................................................................................... 37
2.4 Discussion ................................................................................................................................... 43
2.5 Chapter Two Figures and Tables ............................................................................................... 49
2.6 Chapter Two Literature Cited ..................................................................................................... 57
Chapter Three: A Hard Day’s Night: Shifts in microbial eukaryotic activity in the North Pacific
Subtropical Gyre ........................................................................................................................... 64
Abstract ............................................................................................................................................. 65
3.1 Introduction ................................................................................................................................. 66
3.2 Material and Methods ................................................................................................................. 68
3.3 Results ......................................................................................................................................... 72
3.4 Discussion ................................................................................................................................... 77
3.5 Chapter Three Figures and Tables ............................................................................................. 88
3.6 Chapter Three Literature Cited .................................................................................................. 97
Chapter Four: Shifting metabolic priorities among key protistan taxonomic groups within and
below the euphotic zone .............................................................................................................. 104
Abstract ............................................................................................................................................ 105
4.1 Introduction ................................................................................................................................ 106
4.2 Results ........................................................................................................................................ 109
4.3 Discussion .................................................................................................................................. 113
4.4 Experimental Procedures .......................................................................................................... 123
4.5 Chapter Four Figures and Tables .............................................................................................. 128
4.6 Chapter Four Literature Cited ................................................................................................... 134
Supplementary Materials ............................................................................................................ 141
Supplementary Material for Chapter One ....................................................................................... 142
Supplementary Material for Chapter Two ....................................................................................... 144
Supplementary Material for Chapter Three .................................................................................... 155
Supplementary Material for Chapter Four ...................................................................................... 160
3

4
Acknowledgements
The presented dissertation is a product of years of work that would not have been
possible without the support of many people in my life. I would first and foremost like to
acknowledge the incredible mentorship from my advisor Dave Caron. With the encouragement
and advisement from Dave, I’ve improved my skills as a scientist, travelled all over the world to
present my research, and built a foundation for a career. Dave’s guidance has had a large impact
on how I define myself as a scientist, and I greatly appreciate the time and effort he puts into
ensuring the members of the Caron Lab feel supported.
I would also like to thank the USC community, especially the Marine Environmental
Biology (MEB) faculty, post docs, and graduate students. Advice and insight from everyone in
the MEB Department has been an integral component to my growth at USC. I would especially
like to thank the current and former members of my qualifying and dissertation committees,
Karla Heidelberg, Eric Webb, William Berelson, John Heidelberg, and Jed Fuhrman. Your
insight into the research process provided me extensive personal and professional guidance. I
also extend my gratitude to staff in the MEB program, USC Women In Science & Engineering
program, Wrigley Institute of Environmental Studies, and the USC Graduate School for support
and advisement over the years.
The immense amount of time and work put forth for this dissertation would not have
been possible without the love and encouragement from my family and friends at every turn,
especially Markie Hu, Phil Hu, Pam Hu, Phil Hu Jr., Paige Connell, Sonika Ung, and Barlee.
You are all a consistent source of kindness, fun, and love. Parents (Phil and Pam), thank you for
inspiring my love for science and always supporting my career choices. Markie, I cannot express
how much your patience, undying optimism, silliness, homemade pretzels, and commitment to
our relationship has contributed to this work and made me a better person and scientist.
5
I would especially like to acknowledge the amazing Caron Lab group (past and present):
Paige Connell, Jay Liu, Jayme Smith, Alle Lie, Victoria Campbell, Ramon Terrado, Lisa
Mesrop, Avery Tatters, Diane Kim, Gerid Ollison, Julie Hopper, and Emily Aquirre. I am lucky
to have had such a positive experience in graduate school and this is largely due to this fantastic
group. I extend my gratitude to my peers and friends whom have had an impact on me during my
time as a graduate student, Megan Hall, Johanna Holm, Ben Tully, Rohan Sachdeva, Erin
MacParland, Erin Fichot, Yubin Raut, Christopher Suffridge, Jacob Cram, David Needham, Kyle
Frischkorn, Emma Timmins-Schiffman, Sean Luis, and Ross Whippo.
This dissertation work would not have been possible without the fieldwork support from
the Captain and Crew of the R/V Yellow Fin and the Sundiver Express, Troy Gunderson, Captain
and Crew of the R/V Kilo Moana and the SCOPE ops field team (University of Hawaii),
especially Sam Wilson and Tara Clemente. In the process of analyzing data and writing each
manuscript, I learned a great deal from each of my co-authors, including: Sonya Dyhrman,
Harriet Alexander, Peter Countway, Adriane Jones, Rebecca Gast, Craig Cary, Evelyn Sherr, and
Barry Sherr.
I would also like to acknowledge the numerous funding resources: Chapter One work
was supported by grants from the National Science Foundation (OCE-0550829, MCB-0703159,
MCB-0084231, OCE-1136818) and the Gordon and Betty Moore Foundation. Sequencing was
conducted by the U.S. Department of Energy Joint Genome Institute and supported by the Office
of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.
Sequence assembly and processing at JGI was performed by Ed Kirton, Jim Bristow, and
Sussannah Tringe. William C. Nelson assisted with early bioinformatic analysis of the full-length
sequence data. Chapter Two was supported by the National Science Foundation [grant number
6
OCE-1136818]. Chapter Three was supported by the Simons Foundation (SCOPE: Simon’s
Collaboration on Ocean Processes and Ecology) #P49802. Funding for Chapter Four was
provided by the Gordon and Betty Moore Foundation #3299 and the National Science
Foundation #1136818.

7
Dissertation Abstract
Single-celled microbial eukaryotes (protists) mediate critical elemental transformations that
support ecosystem function. An overarching goal in microbial ecology is to link the metabolic
roles of individual microbes to food web structure, when will enable a better understanding of
the biogeochemical processes driving marine ecosystems. The central goal of my dissertation
work was to characterize protistan community structure and diversity in two regions in the
Pacific Ocean: off the coast of California in the San Pedro Channel and northeast of Hawai’i in
the North Pacific Subtropical Gyre. I used molecular techniques and bioinformatics to improve
how we describe the diversity and activity of in situ protistan communities. By pairing traditional
DNA-based tag sequencing with ribosomal RNA sequences, changes in protistan community
structure and activity were found to correspond to depth, proximity to coastline, season, or time
of day. Further, a metatranscriptome survey provided one of the first accounts of protistan
physiological ecology across a depth gradient. Comparing transcript abundances (mRNA) among
euphotic and sub-euphotic zone depths revealed the metabolic flexibility among key protistan
lineages to rely on alternate metabolic modes/nutritional strategies (e.g. phototrophy at the
surface, heterotrophy below the euphotic zone). This dissertation work addressed long-standing
ecological questions regarding the metabolic roles of protists that we otherwise have no other
way of observing in situ by using RNA-derived sequence information to characterize protistan
diversity and metabolic potential.

8
Dissertation Introduction

Single-celled microbial eukaryotes (protists) display a wide range of nutrient acquisition
strategies, which enable them to fulfill roles as both primary producers and consumers at the base
of aquatic food webs (Caron 2001; Caron et al. 2012; Sherr et al. 2010; Worden et al. 2015b).
Describing the composition and ecology of natural protistan assemblages is therefore critical for
understanding the major elemental transformations that drive globally important biogeochemical
cycles (C, N, P, Fe, or Si) in marine ecosystems (de Vargas et al. 2015). However, the vast
diversity (morphological and physiological) of protistan species poses challenges for
characterizing their naturally occurring populations and subsequent ecological impact (Behnke et
al. 2011; Caron et al. 2017a). This dissertation addresses this challenge by using genetic
techniques to characterize protistan community composition and diversity at different temporal
and spatial scales and investigate group-specific nutritional strategies.
Advances in DNA sequencing capabilities over the past 20 years have enhanced studies
of microbial ecology by providing an additional metric for testing phylogenetic hypotheses,
supplementing classic microscopy-based approaches for species identification, and uncovering
vast amounts of previously unknown diversity in virtually all environments on the planet (López-
García et al. 2001; Massana 2015a; Massana et al. 2013; Moon-van der Staay et al. 2001). A
major challenge in using DNA sequence information to catalog protistan species composition is
the process of accurately assigning taxonomic identities and drawing ecological interpretations
from sequence results. The most common approach for taxonomy characterization of a diverse
community is to perform high-throughput tag-sequencing of a selected hypervariable region,
which falls within the highly conserved 18S rRNA gene (Hugerth et al. 2014b; Mahé et al.
2014b; Medlin et al. 1988a; Stoeck et al. 2010b). However, limitations in high-throughput
9
sequence read lengths [currently, maximum lengths generated from Illumina paired end
sequencing are ~400-500 bps] do not provide the same taxonomic resolution as longer sequences
(e.g. the entire 18S rRNA gene). Chapter One demonstrated how downstream ecological
interpretations from various hypervariable regions (sequence lengths) may differ when compared
to results derived from full length 18S rRNA gene sequences (Chapter One; Hu et al. 2015).
An additional challenge in characterizing microbial eukaryote diversity using high-
throughput tag-sequencing data is that most studies rely on rDNA from the environment, which
does not discriminate the genetic material from live, active microorganisms from non-viable,
non-living genetic material (Not et al. 2009; Stoeck et al. 2007). Ribosomal RNA (rRNA) is
more susceptible to degradation than rDNA (Karl and Bailiff 1989b; Lorenz and Wackernagel
1987), thus sequence information derived from rRNA provides an estimate of ribosomal content
and thus a proxy for protein synthesis in a cell (Egge et al. 2015; Massana 2015b; Poulsen et al.
1993). In Chapters Two and Three, traditional rDNA tag-sequencing was paired with sequences
derived from rRNA (cDNA reverse transcribed from extracted total RNA); this is some of the
first work to use differences in RNA:DNA sequence read ratios to infer changes in relative
metabolic activity among individual protistan taxonomic groups on both temporal and spatial
scales.
Results from Chapter Two demonstrated how the community composition and metabolic
activity of dominant protistan taxa fluctuated with respect to season and location, inferred from a
combined rRNA and rDNA tag-sequencing approach. Samples from the San Pedro Channel, off
the coast of California, were collected seasonally at three environmentally distinct sites: the San
Pedro Ocean Time-series (SPOT) station from four depths (surface, deep chlorophyll maximum,
150 m, and 890 m), surface waters offshore from Santa Catalina Island, and from the Port of Los
10
Angeles (Chapter Two; Hu et al. 2016). rRNA estimates of community diversity reflected
seasonal changes in community structure, while rDNA-based estimates did not. Temporal
changes in the RNA:DNA ratios of regionally important phytoplankton corresponded to known
bloom activity, thus validating the utility of combined rRNA and rDNA sequencing methods in
efforts to monitor phytoplankton blooms. Previous rDNA-based work at SPOT detected ciliates
and radiolaria below the euphotic zone (150 m and 890 m), but was unable to confirm if
sequences originated from live cells or dead sinking material. Our results supported the presence
of metabolically active ciliate and radiolarian groups at 150 m and 890 m, demonstrating how
comparisons of RNA:DNA ratios were useful for gathering ecological information from species
that are otherwise difficult to characterize.
Microbially-mediated biological processes are known to synchronize with regular light-
dark (diel) cycles, but short-term temporal dynamics influencing protistan community structure
are largely undocumented. A Lagrangian survey in the oligotrophic North Pacific Subtropical
Gyre (NPSG) was conducted every 4 hours over a period of 3 days to investigate diel
rhythmicity in the relative abundances and activities of euphotic zone protists (Chapter Three;
Hu et al. In prep). The relative metabolic activity (inferred from RNA:DNA ratios) of several
phytoplankton groups was found to coordinate with the light cycle; the exact timing of daily
peaks in RNA:DNA ratios varied with respect to group, which reflected the varied phytoplankton
metabolic strategies (e.g. biomass-specific nutrient uptake rates or timing of photosynthetic
activity). Further, the relative metabolic activities of phagotrophic protists were more similar to
one another consistently peaking at dusk and throughout the dark cycle, suggesting that grazing
activity increased in response to availability of prey. These results implicated temporal niche
partitioning within the protistan community, influenced by light-dark cycling. Time-dependent,
11
significantly co-occurring OTUs (as a proxy for species) throughout the diel cycle were
hypothesized to represent important predator-prey, parasitic, and mutualistic relationships. More
specifically, negative or positive temporally time-delayed interactions with the parasitoid
Syndiniales included a high number of dinoflagellates and ciliates, suggesting Syndiniales
parasitism may contribute more to mortality in the NPSG than previously thought. Further, the
frequency of mutualistic relationships between endosymbiotic algae and heterotrophic rhizarian
hosts (e.g. acantharia) were inferred from positively co-correlated OTUs. These findings are
unique, as traditional molecular surveys do not typically include these symbiotic trophic
interactions.
To gain a better understanding of the ecological activities of marine protists, a
metatranscriptome survey (the collective mRNA of all eukaryotes) was conducted from a vertical
depth profile at the SPOT station (Chapter Four; Hu et al. In revision.) to elucidate both the
taxonomic composition and metabolic potential of in situ protistan communities. The availability
of improved genetic databases for microbial eukaryotes in recent years has enabled researchers
to apply transcriptomic approaches to natural samples; however this technique is still
infrequently applied in the field due to the limited number of reference transcriptomes (and
subsequent annotations) in sequence databases. Comparative metatranscriptomic surveys
indicated shifts in metabolic potential (changes in transcript abundance) with respect to
taxonomic group and environmental factors, providing a transcript-level metric of the microbial
functional diversity underscoring important biological processes. These efforts are transforming
the way biologists and ecologists describe microbial populations, as incorporation of functional
traits into predictive ecosystem models (e.g. phytoplankton functional type, PFT) enables clearer
representation of ocean ecosystems (Ward and Follows 2016). To date, metatranscriptome
12
surveys that target protists mainly focus on euphotic zone processes; this study (Chapter Four) is
one of the first metatranscriptome surveys of protists to compare in situ communities at both
euphotic and sub-euphotic depths.
Comparisons of relative transcript abundances revealed depth-related shifts in the
nutritional modes of key protistan taxonomic groups, which demonstrated the variable roles
protists contribute to elemental cycling throughout the water column at SPOT (Chapter Four; Hu
et al. In revision.). Depth-related differences in transcript abundance enabled nutritional
strategies, such as phototrophy, heterotrophy, or mixotrophy, to be linked to specific taxonomic
groups. Functional annotations from sunlit samples were typical of phytoplankton communities
(i.e. glycolysis, Calvin cycle, and fatty acid synthesis), while taxa capable of phagotrophy below
the euphotic zone were characterized by an upregulation of transcripts associated with the
breakdown of fatty acids and the glyoxylate cycle. Additionally, we found evidence for
anaerobic metabolism among ciliates, and adaptations to prolonged darkness in diatoms and
chlorophytes. We show that separating group-specific nutritional modes is a novel approach for
examining linkages between protistan community composition and ecosystem function (i.e.
microbially mediated biological processes). A broader implication of this work is to use
transcriptional signatures as proxies for microbial nutritional strategies in predictive ecosystem
models (Stec et al. 2017; Ward and Follows 2016).
13
Introduction References
Behnke, A., M. Engel, R. Christen, M. Nebel, R. R. Klein, and T. Stoeck. 2011. Depicting more
accurate pictures of protistan community complexity using pyrosequencing of
hypervariable SSU rRNA gene regions. Environ. Microbiol. 13: 340-349.
Caron, D. A. 2001. Protistan herbivory and bacterivory. Methods Microbiol 30: 289-315.
Caron, D. A. and others 2016. Probing the evolution, ecology and physiology of marine protists
using transcriptomics. Nat Rev Microbiol: 1-15.
Caron, D. A., P. D. Countway, A. C. Jones, D. Y. Kim, and A. Schnetzer. 2012. Marine protistan
diversity. Annu. Rev. Mar. Sci. 4: 467-493.
de Vargas, C. and others 2015. Eukaryotic plankton diversity in the sunlit ocean. Science 348:
DOI: 10.1126/science.1261605.
Egge, E. S. and others 2015. Seasonal diversity and dynamics of haptophytes in the Skagerrak,
Norway, explored by high-throughput sequencing. Mol. Ecol. 24: 3026-3042.
Hu, S. K. and others 2016. Protistan diversity and activity inferred from RNA and DNA at a
coastal ocean site in the eastern North Pacific. FEMS Microbiol. Ecol. 92: 1-13.
Hu, S. K., P. E. Connell, L. Y. Mesrop, and D. A. Caron. In prep. A Hard Day’s Night: Diel
shifts in microbial eukaryotic activity in the North Pacific Subtropical Gyre.
Hu, S. K. and others Submitted. Shifting metabolic priorities among key protistan taxonomic
groups within and below the euphotic zone.
Hu, S. K. and others 2015. Estimating Protistan Diversity Using High-Throughput Sequencing. J.
Eukaryot. Microbiol. 62: 688-693.
Hugerth, L. W. and others 2014. Systematic Design of 18S rRNA Gene Primers for Determining
Eukaryotic Diversity in Microbial Consortia. PloS One 9: e95567.
Karl, D. M., and M. D. Bailiff. 1989. The measurement and distribution of dissolved nucleic
acids in aquatic environments . Limnol. Oceanogr. 34: 543-558.
López-García, P., F. Rodriguez-Valera, and C. Pedros-Alio. 2001. Unexpected diversity of small
eukaryotes in deep-sea Antarctic plankton. Nature 409: 603-607.
Lorenz, M. G., and W. Wackernagel. 1987. Adsorption of DNA to sand and variable degradation
rates of adsorbed DNA. Appl. Environ. Microb. 53: 2948-2952.
Mahé, F. and others 2014. Comparing High-throughput Platforms for Sequencing the V4 Region
of SSU-rDNA in Environmental Microbial Eukaryotic Diversity Surveys. J. Eukaryot.
Microbiol.: n/a-n/a.
Massana, R. 2015a. Getting specific: making taxonomic and ecological sense of large
sequencing data sets Mol. Ecol.: 1-3.
---. 2015b. Protistan diversity in environmental molecular surveys, p. 3-21. In S. Ohtsuka,
Suzaki, T., Horiguchi, T., Suzuki, N., Not, F. [ed.], Marine Protists: Diversity and
Dynamics. Springer Japan.
Massana, R., J. del Campo, M. E. Sieracki, S. e. p. Audic, and R. Logares. 2013. Exploring the
uncultured microeukaryote majority in the oceans: reevaluation of ribogroups within
stramenopiles. 8: 854-866.
Medlin, L., H. J. Elwood, S. Stickel, and M. L. Sogin. 1988. The characterization of
enzymatically amplified eukaryotic 16S-like rRNA-coding regions Gene: 1-9.
Moon-van der Staay, S. Y., R. De Wachter, and D. Vaulot. 2001. Oceanic 18S rDNA sequences
from picoplankton reveal unsuspected eukaryotic diversity. Letters to Nature 409: 607-
610.
14
Not, F., J. del Campo, V. Balagué, C. de Vargas, and R. Massana. 2009. New insights into the
diversity of marine picoeukaryotes. PloS One 4: e7143. DOI:
7110.1371/journal.pone.0007143.
Poulsen, L. K., G. Ballard, and D. A. Stahl. 1993. Use of rRNA fluorescence in situ
hybridization for measuring the activity of single cells in young and established biofilms.
Appl. Environ. Microb. 59: 1354-1360.
Sherr, B. F., E. B. Sherr, D. A. Caron, D. Vaulot, and A. Z. Worden. 2010. A sea of microbes:
Oceanic protists. Oceanography 20: 130-134.
Stec, K. F. and others 2017. Modelling plankton ecosystems in the meta-omics era. Are we
ready? Marine Genomics: 1-17.
Stoeck, T. and others 2010. Multiple marker parallel tag environmental DNA sequencing reveals
a highly complex eukaryotic community in marine anoxic water. Mol. Ecol. 19: 21-31.
Stoeck, T., A. Zuendorf, H.-W. Breiner, and A. Behnke. 2007. A molecular approach to identify
active microbes in environmental eukaryote clone libraries. Microb. Ecol. 53: 328-339.
Ward, B. A., and M. J. Follows. 2016. Marine mixotrophy increases trophic transfer efficiency,
mean organism size, and vertical carbon flux. P. Natl. Acad. Sci. USA: 201517118-
201517116.
Worden, A. Z., M. J. Follows, S. J. Giovannoni, S. Wilken, A. E. Zimmerman, and P. J. Keeling.
2015. Rethinking the marine carbon cycle: Factoring in the multifarious lifestyles of
microbes. Science 347: 1257594-1257594.

15
Chapter One: Estimating Protistan Diversity using High-throughput
Sequencing
Sarah K. Hu
a
, Zhenfeng Liu
a
, Alle A. Y. Lie
a
, Peter D. Countway
b
, Diane Y. Kim
a
,
Adriane C. Jones
c
, Rebecca J. Gast
d
, S. Craig Cary
e,f
, Evelyn B. Sherr
g
, Barry F. Sherr
g
and
David A. Caron
a

a Department of Biological Sciences, University of Southern California, Los Angeles, California
90089
b Bigelow Laboratory for Ocean Sciences, East Boothbay, Maine 04544
c Mount St. Mary’s College, Los Angeles, California 90049
d Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543
e Environmental Research Institute, School of Science, University of Waikato, Hamilton, New
Zealand 3240
f College of Earth and Ocean Science, University of Delaware, Newark, Delaware 19716
g College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, Oregon
97331

Citation: Hu, S. K. and others 2015. Estimating Protistan Diversity Using High-Throughput
Sequencing. J. Eukaryot. Microbiol. 62: 688-693. doi:10.1111/jeu.12217.

16
Abstract
Sequencing hypervariable regions from the 18S rRNA gene is commonly employed to
characterize protistan biodiversity, yet there are concerns that short reads do not provide the
same taxonomic resolution as full-length sequences. A total of 7,432 full-length sequences were
used to perform an in silico analysis of how sequences of various lengths and target regions
impact downstream ecological interpretations. Sequences that were longer than 400 nucleotides
and included the V4 hypervariable region generated results similar to those derived from full-
length 18S rRNA gene sequences. Present high-throughput sequencing capabilities are
approaching protistan diversity estimation comparable to whole gene sequences.

17
1.1 Introduction
High-throughput sequencing (HTS) of hypervariable regions within the small subunit 18S
ribosomal RNA gene provides researchers the means to delve deeply into the species richness of
natural protistan communities (Grattepanche et al. 2014; Mahé et al. 2014a). Nevertheless, read
lengths enabled by these methods are still relatively short (ca. 100-400 nt) and the taxonomic and
phylogenetic resolution afforded by them can be limited. As a consequence, we presently have a
relatively poor understanding of how 18S rRNA gene sequences of various lengths, and specific
regions targeted for sequencing, impact downstream ecological interpretations (Hadziavdic et al.
2014; Hugerth et al. 2014a; Stoeck et al. 2010a).
Previous work comparing sequencing results from full-length 18S rRNA genes with
hypervariable regions of the 18S gene has demonstrated inconsistencies in the conclusions
gleaned from these datasets regarding species richness and diversity. For example, full-length
18S rRNA sequences of natural protistan communities and short sequences of the V4 or V9
regions led to different conclusions on the composition of abundant taxa (Wolf et al. 2014) and
the level of taxonomic detail (Stoeck et al. 2010a). Discrepancies between hypervariable region
results are related, in part, to methodological problems of sequencing errors, PCR bias, and
differences among applications used to call operational taxonomic units (OTUs). Inconsistencies
in results may also be attributed to differing rates of evolution among hypervariable regions.
These issues affect the number of OTUs measured and therefore the species richness estimated in
natural samples (Decelle et al. 2014; Dunthorn et al. 2012; Hugerth et al. 2014a; Kim et al. 2011;
Wolf et al. 2014; Youssef et al. 2009).
One approach to eliminate PCR and sequencing artifacts from sequence diversity
comparisons is to bioinformatically extract the shorter fragments of interest from longer
18
sequences. Analysis of in silico extractions of short sequence reads from full-length or nearly
full-length 18S rRNA sequences have recently been reported in the literature (Hadziavdic et al.
2014; Hugerth et al. 2014a). These studies concluded that hypervariable regions within the 18S
rRNA gene did not yield similar estimates of protistan biodiversity to each other or to full-length
sequences. The choice of the hypervariable region played a greater role in influencing
interpretations of microbial biodiversity than the choice of sequencing technology and sequence
length (Hadziavdic et al. 2014; Hugerth et al. 2014a).
In this study we compared diversity analyses derived from 7,432 Sanger-sequenced, full-
length 18S rRNA genes obtained from natural microbial eukaryotic communities dominated by
protists, with several bioinformatically-extracted regions from the same full-length sequences.
Full-length and nearly full-length sequences served to groundtruth the ecological interpretations
attained from in silico extracted sequence fragments between 107 and 1,200 nucleotides (nt).
Short sequence fragments included the V7, V4, V1-V3, V1-V4, V4-V7, and V1-V7 regions
(Table S1) determined by 18S rRNA gene specific primers previously used in protistan studies
(Hadziavdic et al. 2014; Medlin et al. 1988b; Stoeck et al. 2010a; Weekers et al. 1994b). OTUs
clustered at 97%, 98%, and 99% sequence similarity from full-length and short fragment
sequence datasets revealed that short reads (< 400 nt) predicted fewer OTUs and yielded lower
values for two commonly employed diversity indices than the full-length sequences.
Nevertheless, sequences ≥ 400 nt yielded ecological interpretations similar to full-length
sequences, especially when the extracted sequence included a taxonomically-informative region
such as the V4 hypervariable region.

19
1.2 Materials and Methods
1.2.1 Sample collection and processing
Samples from a global survey of natural microbial eukaryotic communities (Lie et al. 2014) were
collected for DNA extraction and sequencing (Countway et al 2010). A total of 12 seawater
samples were obtained from depths ranging from 5 m to 2,500 m at 5 locations (Table S2):
Arctic Ocean (AO), San Pedro Ocean time series station in the eastern North Pacific (ENP), East
Pacific Rise (EPR) in the eastern Pacific, Gulf Stream (GS) in the western North Atlantic, and
Ross Sea, Antarctica (RS). Water was collected using Niskin bottles mounted on a rosette, and 2-
20 liters of water was prefiltered through 200 µm Nitex mesh in order to exclude most
multicellular organisms (samples from ENP were prescreened with both 200-µm and 80-µm
Nitex mesh). Microbial biomass was collected onto GF/F filters (Whatman), rolled and placed
into vials containing 2-mL of 2X lysis buffer (100-mM Tris pH 8, 40-mM EDTA pH 8, 100-mM
NaCl, 1% SDS), and stored frozen (-20 °C) aboard the ship or flash frozen in liquid nitrogen for
later DNA extraction.
Genomic DNA was extracted by thawing filters on a 70 °C heating block followed by
three rounds of bead-beating on a vortex mixer (3 min) and heating at 70 °C (3 min) before
extraction using phenol-chloroform as outlined by Countway et al (2007a). Extracted DNA was
precipitated, dried, and then resuspended in sterile water. Genomic DNA was PCR-amplified to
enrich 18S rRNA genes using universal eukaryotic primers Euk-A and Euk-B (Table S1, Medlin
et al. 1988). PCR amplifications were performed as described by Countway et al (2010). The
PCR thermal protocol consisted of a single cycle at 95 ˚C for 2 minutes, 35 cycles of 95 ˚C for
30 seconds, 50 ˚C for 30 seconds, and 72 ˚C for 2.5 minutes, and a final elongation step at 72 ˚C
for 7 minutes.
20
Cloning procedures were performed as described in Countway et al (2010). 18S rDNA
products were run on a 1.2% SeaKem agarose gel, bands were excised and PCR amplicons were
cloned with the TOPO-TA kit (Invitrogen) using the TOP-10 electrocompetent cells (Countway
et al. 2007a; Countway et al. 2010). Initial plating was done to ensure successful transformation
and to estimate cloning efficiency. 10% glycerol stocks were prepared for shipment to the Joint
Genome Institute (JGI, http://www.jgi.doe.gov/) for sequencing.
1.2.2 Sequencing and data analysis
Clones from glycerol stocks were plated and discrete colonies were robotically-picked at JGI
after overnight growth. Sanger sequencing was conducted by the JGI on an ABI 3730 capillary
DNA sequencer (Applied Biosystems). A total of 1,152 clones were sequenced for each sample
using T3 and V7 vector primers (Sambrook et al. 1989) and the 570-F internal primer (Weekers
et al. 1994b). Assembly was performed by JGI using the program Phrap v0.990319 (Ed Kirton).
The resulting full-length and nearly full-length 18S rRNA gene sequences had an average of
1,647 nt in length.
Sequences were screened for chimeras using a local implementation of the Pintail
algorithm (Ashelford et al. 2005). Each sequence was searched with BLASTN v2.2.25+
(Altschul et al. 1990) against a database consisting of the eukaryotic SSU sequences from
SILVA (Pruesse et al. 2007) to calculate a deviation of expected (DE) value (Ashelford et al.
2005). Sequences with a DE value greater than 5 were identified as possible chimeras and
removed (6% of the sequences). Sequences were submitted to GenBank with accession numbers:
KJ757035-KJ764638 (Lie et al. 2014).
A global alignment using 7,432 of the sequences was performed using Geneious v6.1.6
(http://www.geneious.com, Kearse et al. 2012), followed by in silico isolation of the short
21
sequence fragments. A virtual ‘bioinformatic PCR’ was performed in Geneious using forward
and reverse primers for V7, V4, V1-V3, V1-V4, V4-V7, and V1-V7 regions with an allowance
of 3 mismatches per primer (indels not considered, Table S1). Targeted regions were chosen
based on 18S primers previously employed to evaluate microbial eukaryotic biodiversity
(Hadziavdic et al. 2014; Medlin et al. 1988b; Stoeck et al. 2010a; Weekers et al. 1994b). Each
region was then isolated in silico from all of the 7,432 sequences based on the location of the
primer hit to a position on the aligned full-length 18S rRNA gene sequences. Sequence
fragments used in downstream analysis included the forward and reverse primers. These
fragments and the original full-length sequences were then processed in mothur v1.33.1 (Schloss
et al. 2009) to call OTUs and calculate diversity indices. Distances between sequences were
calculated using the furthest neighbor algorithm at 97%, 98%, and 99% sequence similarity.
Representative OTUs from each sequence dataset were blasted against the SILVA database to
assign taxonomic identities based on an e-value threshold of 1e
-
10. In the case of more than one
equal blast hit, the first one provided in the BLAST output was chosen for taxonomic
assignment.
1.2.3 Choice of primers
Primers chosen for this study originated from previously published work that targeted natural
protistan communities (Table S1). The effectiveness of these 18S rRNA gene primers has been
evaluated for a number of protistan lineages (Dunthorn et al. 2012; Ki 2012), although there are
taxa for which these primers are ineffective due to variances in the length of the 18S, such as
foraminifera (Pawlowski and Lecroq 2010).

22
V1-V3 and V4 primers reflect commonly used primers, including Euk-A from Medlin et
al. (1988b), 570-R from Weekers et al. (1994b) (which is approximately the end of the V3
region), and V4 primers from Stoeck et al. (2010a). The V1-V3 and V4 regions examined in this
study overlapped by 94 nt (including primers, Table S1).
1.2.4 V9 hypervariable region
The V9 region was not included in our analysis because complete Euk-B reverse primer
sequences were not obtained for many of the full-length sequences. Only approximately 2,000
sequences had suitable V9 regions (intact Euk-B reverse primer), significantly less than the
original 7,432 full-length and nearly full-length 18S rRNA gene sequences.
1.2.5 Diversity and community similarity of protistan assemblages
Inverse Simpson and Shannon indices were calculated to provide a measurement of diversity
based on the abundance and evenness of OTUs generated from extracted fragments and full-
length sequences (Schloss et al. 2009). Community composition was dominated by protists in
each sample. Community composition was compared using the Bray-Curtis similarity parameter.
OTU abundances for each dataset were square-root transformed and clustered by non-metric
Multidimensional Scaling plots in Primer-E v6 (Clark and Warwick 2001).

1.3 Results and Discussion
Assessing sequence diversity of rRNA genes is commonly used to investigate the structure and
composition of natural protistan communities, but our understanding of the ecological
information contained in sequence data is still evolving. We show here that the ecological
inference of protistan species richness and diversity obtained from 18S rRNA gene sequences
varied as a factor of read length and the inclusion of the V4 hypervariable region (Figure 1.1).
23
Previous work, specifically in regards to ciliates, has shown that in comparison to the shorter V9
hypervariable region, the V4 region was better for resolution of taxonomies (Dunthorn et al.
2012) and phylogenetic placement (Dunthorn et al. 2014). Use of the V4 region for protistan
biodiversity studies has been shown to be superior to other 18S hypervariable regions, as the
entire length (ca. 400 nt) is easily obtainable with current paired end HTS methods, namely
Illumina MiSeq and HiSeq (Mahé et al. 2014a).
The number of OTUs in this study generated from in silico extracted sequences was
directly related to sequence length (Figure 1.1A), while both read length and inclusion of the V4
region appeared to influence diversity indices (Figure 1.1B,C). In particular, read lengths
between 400 and 600 bases recovered between 300 and 900 fewer OTUs compared to full-length
or nearly full-length 18S gene sequences (>1,200 nt). Analysis of the shortest region (V7)
yielded only a small fraction of the species richness obtained using longer reads. Results
generated using different levels of sequence similarity to form OTUs yielded different absolute
numbers of OTUs, as expected, but relative changes in OTUs between the various fragment
lengths were consistent across the dataset (different shading in Figure 1.1A).
Values for inverse Simpson and Shannon diversity indices reflected the trends in total
number of OTUs generated. Diversity indices for the V7 region were several-fold lower than
values obtained using longer reads, and read lengths greater than 400 nt were only moderately
lower than values obtained using full-length 18S rRNA gene sequences (Figure 1.1B,C). The
V4-V7 and V1-V7 reads had diversity indices that were virtually indistinguishable from those
based on full-length reads (Figure 1.1B,C), in accordance with findings that the V4 region is
taxonomically-informative (Dunthorn et al. 2012; Hugerth et al. 2014a; Mahé et al. 2014a;
Nickrent and Sargent 1991).
24
A non-metric Multi-Dimensional Scaling (nmMDS) plot of Bray-Curtis dissimilarity
values from the short V7 region was unable to resolve more than a few differences in protistan
community structure among the twelve globally distributed natural samples analyzed in this
study (Figure S1; Table S2). In contrast, data sets derived from DNA fragments with lengths >
400 nt all yielded patterns that were relatively similar. There were only small inconsistencies
among patterns obtained using longer fragments of the rRNA genes and the full-length
sequences, a finding that paralleled results from the diversity indices (Figure 1.1B,C; Figure S1).
Pernice et al (2013) recognized the usefulness of establishing a reference database of
well-curated, relatively long rRNA gene sequences for vetting the many shorter sequences
presently produced by HTS. Using a similar approach, we used in silico extraction of shorter
reads from the same dataset of full-length sequences and examined the consistency of the
ecological interpretations derived from DNA fragments of various lengths and target regions.
Our study systematically confirms that longer (>400 nt) sequences enabled by HTS technologies
provide consistent ecological information on protistan richness and diversity, relative to full-
length sequences of 18S rRNA genes. This result is comforting in that it supports the application
of present-day HTS in protistan ecology (Mahé et al. 2014a), although it does not necessarily
capture the total eukaryotic diversity present in natural ecosystems. This study did not address
the more fundamental limitations of using a single gene for investigating the enormous breadth
of microbial eukaryote diversity, such as inconsistencies between the morphospecies and genetic
species concepts, or the potential variable rates of mutations of rRNA genes among protistan
lineages (Caron 2013a). Given this inherent limitation, the present study used full-length 18S
rRNA environmental sequences to groundtruth the use of smaller gene fragments. DNA
sequence information will continue to augment our understanding of microbial diversity, and
25
future efforts to incorporate sequences from multiple genes into these approaches will improve
our ability to capture true protistan diversity in the environment (Caron 2013a; Stoeck et al.
2008).

26
1.4 Chapter One Figures and Tables

Figure 1.1 Total number of observed OTUs and diversity estimates for full-length and in
silico extracted regions of the small subunit 18S ribosomal RNA gene. (A) Numbers of
operational taxonomic units (OTUs) obtained using full-length sequences or gene fragments of
different lengths (Table S1). Sequences were clustered at 97%, 98% or 99% sequence similarity
in mothur v. 1.33.1 (Schloss et al. 2009). Full-length, V1-V7, and V4-V7 (>1,200 nt) sequence
datasets yielded the largest number of OTUs (average of 1,680 OTUs at 97% sequence
similarity) compared to the V4, V1-V4, and V1-V3 (400-700 nt) sequence results (average of
1,430 OTUs at 97% sequence similarity) and the short V7 (107 nt) sequence results (300 OTUs
at 97% sequence similarity). OTUs from various sequence fragments were proportionally the
same for 97%, 98%, or 99% sequence similarities. (B) Inverse Simpson diversity index and (C)
Shannon diversity index calculated in mothur v. 1.33.1 (Schloss et al. 2009) using the OTU
information presented in (A) as a function of sequence length (nt). Sequences longer than 1,200
nt (V4-V7, V1-V7) yielded inverse Simpson and Shannon values that were comparable to values
for the full-length sequences. Values for the V4, V1-V3 and V1-V4 fragments were marginally
lower than the three longer sequences for both indices, while values for the V7 region were
markedly lower. Results from the various gene fragments in (B) and (C) are in the same order as
(A), depicted left to right in order of average read length (Table S1): V7 (107 nt), V4 (418 nt),
V1-V3 (533 nt), V1-V4 (690 nt), V4-V7 (1,200 nt), V1-V7 (1,260 nt), and full length (1,647 nt).
A
B C
Average Read Length (nt)
Shannon Inverse Simpson
V7
V4
V1-V3
V1-V4
V4-V7
V1-V7
FL
V7
V4
V1-V3
V1-V4
V4-V7
V1-V7
FL
27
1.5 Chapter One Literature Cited
Decelle, J., S. Romac, E. Sasaki, F. Not, and F. Mahé. 2014. Intracellular diversity of the V4 and
V9 regions of the 18S rRNA in marine protists (radiolarians) assessed by high-throughput
sequencing. PLoS ONE 9: e104297.
Dunthorn, M., J. Klier, J. Bunge, and T. Stoeck. 2012. Comparing the hyper-variable V4 and V9
regions of the small subunit rDNA for assessment of ciliate environmental diversity. J.
Eukaryot. Microbiol. 59: 185-187.
Grattepanche, J.-D., L. F. Santoferrara, G. B. McManus, and L. A. Katz. 2014. Diversity of
diversity: conceptual and methodological differences in biodiversity estimates of
eukaryotic microbes as compared to bacteria. Trends Microbiol. 22: 432-437.
Hadziavdic, K., K. Lekang, A. Lanzen, I. Jonassen, E. M. Thompson, and C. Troedsson. 2014.
Characterization of the 18S rRNA gene for designing universal eukaryote specific
primers. PLoS ONE 9: e87624.
Hugerth, L. W. and others 2014. Systematic design of 18S rRNA gene primers for determining
eukaryotic diversity in microbial consortia. PLoS ONE 9: e95567.
Kim, M., M. Morrison, and Z. Yu. 2011. Evaluation of different partial 16S rRNA gene
sequence regions for phylogenetic analysis of microbiomes. J. Microbiol. Methods 84:
81-87.
Mahé, F. and others 2014. Comparing high-throughput platforms for sequencing the V4 region
of SSU- rDNA in environmental microbial eukaryotic diversity surveys. J. Eukaryot.
Microbiol. e-pub ahead of print 19 Sept 2014, doi:10.1111/jeu.12187-4559.
Medlin, L., H. J. Elwood, S. Stickel, and M. L. Sogin. 1988. The characterization of
enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71: 491-499.
Stoeck, T. and others 2010. Multiple marker parallel tag environmental DNA sequencing reveals
a highly complex eukaryotic community in marine anoxic water. Mol. Ecol. 19 Suppl 1:
21-31.
Weekers, P. H. H., R. J. Gast, P. A. Fuerst, and T. J. Byers. 1994. Sequence variations in small-
subunit ribosomal RNAs of Hartmannella vermiformis and their phylogenetic
implications. Mol. Biol. Evol. 11: 684-690.
Wolf, C., E. Silva Kilias, and K. Metfies. 2014. Evaluating the potential of 18S rDNA clone
libraries to complement pyrosequencing data of marine protists with near full-length
sequence information. Mar. Biol. Res. 10: 771-780.
Youssef, N., C. S. Sheik, L. R. Krumholz, F. Z. Najar, B. A. Roe, and M. S. Elshahed. 2009.
Comparison of species richness estimates obtained using nearly complete fragments and
simulated pyrosequencing-generated fragments in 16S rRNA gene-based environmental
surveys. Appl. Environ. Microbiol. 75: 5227-5236.

28
Chapter Two: Protistan diversity and activity inferred from RNA and DNA at
a coastal ocean site in the eastern North Pacific

Sarah K. Hu
a
*, Victoria Campbell
a,b
, Paige Connell
a
, Alyssa G. Gellene
a
, Zhenfeng Liu
a
, Ramon
Terrado
a
, David A. Caron
a

a
Department of Biological Sciences, University of Southern California, 3616 Trousdale Parkway,
Los Angeles, California 90089-0371
b
UW Medicine, Division Allergy and Infectious Diseases, 750 Republican St, Seattle, WA 98109

Citation: Hu, S. K. et al. Protistan diversity and activity inferred from RNA and DNA at a
coastal ocean site in the eastern North Pacific. FEMS Microbiol Ecol fiw050–13 (2016).
doi:10.1093/femsec/fiw050

29
Abstract
Microbial eukaryotes fulfill key ecological positions in marine food webs. Molecular approaches
that connect protistan diversity and biogeography to their diverse metabolisms will greatly
improve our understanding of marine ecosystem function. The majority of molecular-based
studies to date use 18S rRNA gene sequencing to characterize natural microbial assemblages, but
this approach does not necessarily discriminate between active and non-active cells. We
incorporated RNA sequencing into standard 18S rRNA gene sequence surveys with the purpose
of assessing those members of the protistan community contributing to biogeochemical cycling
(active organisms), using the ratio of cDNA (reverse transcribed from total RNA) to 18S rRNA
gene sequences within major protistan taxonomic groups. Trophically important phytoplankton,
such as diatoms and chlorophytes exhibited seasonal trends in relative activity. Additionally,
both radiolaria and ciliates displayed previously unreported high relative activities below the
euphotic zone. This study sheds new light onto on the relative metabolic activity of specific
protistan groups and how microbial communities respond to changing environmental conditions.

30
2.1 Introduction
Protists are fundamental for maintaining the functional stability of marine ecosystems (Caron et
al. 2012; Sherr and Sherr 1994). Microbial eukaryotes serve as important links to higher trophic
levels as primary producers and consumers (Sherr et al. 2010; Sherr and Sherr 2002). The vast
morphological and genetic variability, and consequent physiological diversity among protistan
species enables them to fulfill these diverse roles, yet complicates our efforts to study protistan
community ecology (Behnke et al. 2011; Caron 2013b).
Sequencing genes from environmental samples to supplement classical approaches of
microscopy and culture for studying the diversity of microorganisms has revolutionized the field
of microbial ecology. DNA sequencing has allowed researchers the means to rapidly characterize
entire natural assemblages of protists and resolve some taxonomic and phylogenetic relationships
among these species (Caron 2013b; López-García et al. 2001; Massana 2015b; Moon-van der
Staay et al. 2001). High-throughput sequencing of the small subunit 18S rRNA gene is currently
a cost-effective method for probing the diversity of these communities. Such studies have
revealed immense genetic diversity in virtually all environments sampled to date (Armbrust and
Palumbi 2015; Caron et al. 2012; de Vargas et al. 2015; Moreira and López-García 2001; Stoeck
et al. 2010b; Worden et al. 2015a).
A number of recent molecular studies have promoted sequencing cDNA (reverse
transcribed from extracted total RNA, hereafter referred to as RNA) as a supplement to 18S
rRNA gene sequencing (hereafter, referred to as DNA) (Charvet et al. 2014a; Egge et al. 2015;
Logares et al. 2014; Massana et al. 2015; Not et al. 2009; Stoeck et al. 2007; Terrado et al.
2011). DNA-based sequencing methods provide genetic identifications of Operational
Taxonomic Units (OTUs) of the protistan species present in a sample, but they do not necessarily
31
discriminate active organisms from moribund, encysted, metabolically inactive, or even non-
living genetic material (Not et al. 2009; Stoeck et al. 2007). DNA is generally more resistant to
degradation than RNA (Karl and Bailiff 1989a; Lorenz and Wackernagel 1987), and therefore
sequence information derived from intact total RNA implies ribosomal activity and the potential
for protein synthesis (Blazewicz et al. 2013; Corinaldesi et al. 2011; Egge et al. 2015;
Lejzerowicz et al. 2013; Massana et al. 2015; Poulsen et al. 1993). Consequently, the ratio of
RNA to DNA sequences (RNA:DNA) has been used as an index of metabolic activity (Charvet
et al. 2014a; Logares et al. 2014; Massana et al. 2015). Results derived from RNA sequencing
can provide significant insight into the potential for protein synthesis in the microbial
community, as long as the limitations and assumptions of the method are made clear (Blazewicz
et al. 2013).
Comparisons of RNA and DNA sequence abundances have revealed environmental
selection of active species in natural and experimental systems (Charvet et al. 2014a; Massana et
al. 2015; Not et al. 2009; Poulsen et al. 1993; Stoeck et al. 2007). In an Arctic freshwater lake
study, changes in the RNA-based sequence community correlated with experimental conditions,
while DNA-based sequence diversity remained relatively unchanged among treatments (Charvet
et al. 2014a). Thus, community diversity derived from RNA appears more responsive to
environmental conditions than DNA (Terrado et al. 2011). RNA-DNA comparisons may be
useful for characterizing the dynamics of the protistan rare biosphere (Caron and Countway
2009), where metabolic activity has been detected (Debroas et al. 2015; Logares et al. 2014).
Our objectives were to examine differences in community structure based on RNA or
DNA sequence results and use RNA:DNA ratios to provide new insight into the dynamics of the
functionally active component of the protistan community. Changes in the diversity of the total
32
protistan community (inferred from DNA) and the subset of the community presumed
metabolically active (inferred from RNA) were examined at three stations with varying levels of
anthropogenic influence in the eastern North Pacific off southern California. Samples were
collected seasonally at the San Pedro Ocean Time-series (SPOT) station at four depths (surface,
subsurface chlorophyll maximum, 150 m, and 890 m), and also from surface waters in the Port
of Los Angeles (Port of LA), and near Santa Catalina Island (Catalina). RNA and DNA
sequences of the V4 hypervariable region were compared in order to infer relative activity of
protistan communities temporally (seasonally) and spatially (horizontally and vertically). We
also address the limitations of evaluating diverse protistan communities using an RNA:DNA
ratio approach by separating observations according to individual taxonomic groups. The
majority of sequence-based studies to date do not address in situ metabolic activity of microbial
communities with respect to season or location. This study contributes to our knowledge of the
biologically important roles that protists play in marine ecosystems, by evaluating protistan
community structure and relative activity.

2.2 Materials and Methods
Sample collection
Seawater samples were collected in April, July, October, and January (2013-14) from the San
Pedro Ocean Time-series station (SPOT; 33°33’N, 118°24’W), the Port of Los Angeles (Port of
LA; 33°42.75 N, 118°15.55 W), and approximately 1 km offshore from Santa Catalina Island
(Catalina; 33°27.17 N, 118°28.51 W) in the eastern North Pacific (Figure 2.1, Table S1). The
SPOT station was sampled from 5 m, the subsurface chlorophyll maximum (SCM), 150 m, and
890 m using 10 L Niskin bottles mounted on a CTD rosette, during regularly scheduled cruises
33
(https://dornsife.usc.edu/spot/). The depth of the SCM was determined using real-time
fluorescence data from the CTD downcast, with seawater collected during the CTD upcast.
Surface water from the Port of LA and Catalina was collected within 1 week of the SPOT cruise
sampling dates (Table S1), by submerging acid-cleaned and pre-rinsed 20 L carboys below the
seawater surface and capping the containers while underwater to minimize bubble formation and
contamination with the surface microlayer.
Seawater from all samples was sequentially prefiltered through 200 µm and 80 µm Nitex
mesh to reduce abundances of multicellular eukaryotes (metazoa). Near-surface and SCM
seawater (2 L) and 150 m and 890 m seawater (4 L) was filtered onto GF/F filters (nominal pore
size 0.7 µm; Whatman, International Ltd., Florham Park, NJ, USA) and immediately flash frozen
in liquid nitrogen for later DNA and RNA extraction. Oxygen, temperature, salinity, and
fluorescence data were obtained from either CTD sensors or a profiling natural fluorometer
(PNF). Samples for chlorophyll a and nutrients were collected at all three surface sites; details
can be found in Supporting Information.

Nucleic acid extraction and sequencing
Total DNA and RNA were extracted simultaneously from each sample using the All Prep
DNA/RNA Mini kit (Qiagen, Valencia, CA, USA, #80204). Genomic DNA was removed during
the RNA extraction with RNase-Free DNase reagents (Qiagen, #79254). Total extracted RNA
was checked for residual genomic DNA by performing a polymerase chain reaction (PCR) using
DNA specific primers to ensure that no amplified products appeared when run on an agarose gel.
RNA was reverse transcribed into cDNA using iScript Reverse Transcription Supermix with
random hexamers (Bio-Rad Laboratories, Hercules, CA, USA, #170-8840).
34

The resulting cDNA and DNA from each sample were PCR amplified using V4 forward
(5’-CCAGCA[GC]C[CT]GCGGTAATTCC-3’) and reverse (5’-
ACTTTCGTTCTTGAT[CT][AG]A-3’) primers (Stoeck et al. 2010b). Duplicate PCR reactions
were performed in 50 µl volumes of: 1X Phusion High-Fidelity DNA polymerase (New England
Biolabs, Ipswich, MA, USA, #M0530S), 200 µM of dNTPs, 0.5 µM of each V4 forward and
reverse primer, 3% DMSO, 50 mM of MgCl, and 5 ng of either DNA or cDNA template per
reaction. The PCR thermal cycler program consisted of a 98°C denaturation step for 30 seconds
(s), followed by ten cycles of 10 s at 98°C, 30 s at 53°C, and 30 s at 72°C, and then fifteen cycles
of 10 s at 98°C, 30 s at 48°C, and 30 s at 72°C, and a final elongation step at 72°C for ten
minutes, as described in Rodríguez-Martínez et al. (2012). PCR products were purified (Qiagen,
#28104) and duplicate samples were pooled. The approximately 400 base pair (bp) cDNA and
DNA PCR products were quality checked on an Agilent Bioanalyzer 2100 (Agilent
Technologies, Santa Clara, CA).

Sequence analysis
250 X 250 bp paired-end sequencing was performed on an Illumina MiSeq (V2 chemistry).
Nucleotide bases with a Q score lower than 20 for the last 30 bp of each sequence were trimmed.
Paired-end sequences were merged using FLASh (Magoc and Salzberg 2011) with a minimum
of 10 bp and maximum of 150 bp overlap between each sequence pair. Sequences shorter than
350 bp, longer than 460 bp, or which had an average quality score lower than 25 were discarded
using QIIME v1.8 (Caporaso et al. 2010). Chimeric sequences were identified and removed, by
either de novo or reference-based chimera checking (identify_chimeric_seqs.py in QIIME,
35
intersection method). Sequence data have been deposited in the NCBI BioProject database under
accession number PRJNA311248.
OTUs were generated using QIIME’s subsampled open-reference OTU clustering
protocol, which combined closed-reference (using the SILVA v.111 database (Quast et al. 2012))
and de novo (without any reference database) OTU clustering to decrease run time and ensure
that the maximum number of sequences were placed into OTUs at 97% sequence similarity (as
described in Rideout et al. (2014)). OTUs with only one (singleton) or two (doubleton)
sequences were removed from the dataset for all downstream analyses (Behnke et al. 2011).
For each site and month, OTUs with only RNA (RNA-only OTUs) or DNA (DNA-only
OTUs) were not included in the analysis. RNA- or DNA-only OTUs may have arisen due to
sequencing error, differences in detection level with regard to RNA or DNA copy number, or
reverse transcription error (in the case of RNA-only OTUs (Egge et al. 2013)). Defining DNA-
only OTUs as dormant was deemed inappropriate, as dormant cells often have detectable levels
of RNA (Blazewicz et al. 2013). We therefore focused on OTUs with detectable levels of both
RNA and DNA.

Protistan community diversity and composition
Whole protistan community composition was evaluated by first compiling all RNA and DNA-
based OTUs into phylum or class level taxonomic designations (SILVA). Each OTU was
manually assigned to a major taxonomic group, consisting of: dinoflagellates, ciliates, other
alveolates (the latter mainly comprised of Syndiniales as defined in Ohtsuka et al. (2015)),
chlorophytes, cryptophytes, haptophytes, MArine STramenopiles (MASTs, as defined in
Massana et al. (2004)), diatoms, other stramenopiles (the latter mainly comprised of
36
Chrysophyceae and Dictyochophyceae), cercozoa, radiolaria (also known as retaria, Adl et al.
(2012)), metazoa, and others.
Bray-Curtis dissimilarity matrices were constructed based on the RNA or DNA sequence
results for the three surface stations and four depths at SPOT separately, using the function
‘simprof’ in the ‘clustsig’ R package (Clarke et al. 2008; Whitaker and Christman 2014). The
data were normalized across samples by calculating the relative abundance of each OTU. Then
Bray-Curtis distance matrices were visualized by creating dendrograms based on average
hierarchical clustering (‘hclust’ function in R (R Core Team 2014)). Weighted UniFrac analyses,
which incorporated phylogenetic distances among OTUs, were also computed; details can be
found in Supporting Information.

Relative activity of protistan groups
RNA:DNA ratios were calculated for each OTU, then average ratios for each major taxonomic
group were used as a proxy for relative activity. The relative activity was only compared across
sites (spatial) or months (temporal) within an individual taxonomic group because gene copy
number per cell can vary widely among different protistan lineages (groups) (Godhe et al. 2008;
Massana and Pedros-Alio 2008; Vaulot et al. 2002; Zhu et al. 2005). Differences in RNA:DNA
ratios by month and site were evaluated using analysis of variance (ANOVA) in conjunction
with Tukey’s honest significant difference test using R (R Core Team 2014)(P<0.05, conf. 0.95).

37
2.3 Results
Environmental parameters at the study sites
The SPOT station is located approximately midway between the Port of LA and Santa Catalina
Island (Figure 2.1). Sea surface temperatures at SPOT ranged seasonally from 16°C in April to
20°C in July during the study (Figure 2.2A, Table S2). The depth and intensity of the subsurface
chlorophyll maximum (SCM) at SPOT varied with season (Figure 2.2B). Chlorophyll a at the
SCM was highest during the spring and summer (2-4 µg L
-1
) and lowest during the fall and
winter (1.0-1.7 µg L
-1
, Figure 2.2B). The SCM was sampled at 32 m in April, 33 m in July, 34 m
in October, and 42 m in January. The April SCM sample was obtained from the lower edge of
the chlorophyll a maximum at 32 m (2 µg L
-1
) instead of ca. 25 m (Figure 2.2B).
SPOT is situated over a deep coastal basin (the San Pedro Basin) with a maximum depth
of ca. 900 m. Underwater sills at ca. 740 m to the north and south, and Catalina Island to the west
restrict water circulation in the lower half of the water column. Temperatures did not fluctuate
seasonally below the euphotic zone and remained ca. 9.6°C at 150 m and 5.2°C at 890 m (Figure
2.2A). Dissolved oxygen concentrations at 150 m during the study ranged between 90 and 130
µmol L
-1
(Figure 2.2C). There was a persistent oxycline extending from below the euphotic zone
to approximately 350 m, and dissolved oxygen levels were <45 µmol L
-1
. Below 350 m, the
water column remained hypoxic throughout the course of the study (Figure 2.2C).
Surface salinity across all three stations and sampling dates remained between 31 and 34
PSU and sea surface temperatures ranged from 14 to 20°C (Table S2). The Port of LA had
higher concentrations of chlorophyll a, phosphate, silicate, nitrite, nitrate, and ammonium
compared to Catalina and SPOT, indicating the more eutrophic nature of the nearshore station
38
(Table S2). Chlorophyll a concentrations were highest during April at all three surface stations
(Table S2).

Eukaryotic RNA and DNA sequences
High-throughput sequencing initially recovered ca. 8 million RNA and DNA sequences from 24
samples. Approximately 3.1 million of these sequences were discarded following quality
checking, paired-end merging, and chimera detection (1.6 million chimeric sequences), resulting
in 4.88 million high-quality RNA (2.3 million) and DNA (2.5 million) V4 sequences. Sequences
were clustered into OTUs at 97% sequence similarity, which generated 30 781 OTUs at
approximately species level designations. Following removal of singleton and doubleton OTUs
there were 4.57 million sequences and 18 393 OTUs (See Table S3 for sequence and OTU
counts). For all downstream analyses, only OTUs with both RNA and DNA sequences for each
sample were used in the analysis, leaving 4.27 million sequences (6.6% decrease in the total
number of sequences) and 6 931 OTUs (Table S3). Final RNA and DNA sequence libraries
consisted of 2.1 million and 2.2 million sequences, respectively (Table 2.1). The total number of
sequences found in each OTU and taxonomic designation for each OTU can be found in Table
S7. Analyses for community structure and sequence abundances were repeated with all OTUs
and showed the same trends as the final culled dataset, see Supporting Information (Table S8 and
Figure S1).

Protistan community diversity and composition
The same major protistan taxonomic groups (manually designated groups, see Materials and
Methods) were detected across all months and sites in both RNA and DNA sequence libraries,
39
although not always in the same relative proportions (Table 2.1). Alveolates (including:
dinoflagellates, ciliates, and other alveolates) made the largest single contribution to both RNA
and DNA sequence libraries (over 40% each, Table 2.1). After alveolates, chlorophytes,
cryptophytes, haptophytes, and stramenopiles (the latter includes diatoms, MAST, and other
stramenopiles) made up a combined 43% of the RNA sequence library, while the same groups
contributed only 17% of the DNA sequences (Table 2.1). Within the rhizaria, less than 2% of the
RNA and DNA library was identified as cercozoa (Table 2.1). Radiolaria comprised less than 2%
of the total number of RNA sequences, but contributed 11% of the total number of DNA
sequences. Metazoan sequences made up ca. 21% of the DNA sequence library, while fewer than
2% of the RNA sequences were identified as metazoa (Table 2.1).
Community dissimilarity based on DNA samples grouped the surface stations
approximately by site, Port of LA samples formed a cluster separate from SPOT and Catalina,
with the exception of the Port of LA in April and Catalina in January (Figure 2.3A). DNA-based
diversity in the Port of LA in April was made up of 72% dinoflagellate sequences, while
dinoflagellates made up <50% of the total number of sequences in other samples (Table S4). In
January, the DNA sample at Catalina revealed comparatively larger abundances of
dinoflagellates (44%) and ciliates (20%) relative to other SPOT and Catalina samples (Table S4).
Bray-Curtis dissimilarity results from the RNA sequence library formed more distinct
clusters with respect to both site and season relative to DNA-based estimates (compare Figures
2.3A and 2.3B). Samples from the Port of LA formed a cluster separate from SPOT and Catalina,
with the exception of the Port of LA in April (Figure 2.3B). Diatoms generally made up <10% of
the total number of RNA sequences in surface samples, but in April, diatoms contributed 45% of
the sequences in the Port of LA (Table S4).
40
RNA-based communities were not significantly different between SPOT and Catalina
within each month sampled (P<0.05, Figure 2.3B), although dominant taxonomic groups varied
by month (Table S4). In January, both SPOT and Catalina were dominated by dinoflagellates
(SPOT: 21%, Catalina: 15%), ciliates (SPOT: 35%, Catalina: 31%), and haptophytes (SPOT:
18%, Catalina: 21%). The same taxonomic groups were found in April at SPOT and Catalina,
but at different relative abundances compared to January: dinoflagellates (SPOT: 11%, Catalina:
16%), ciliates (SPOT: 14%, Catalina: 17%), and haptophytes (SPOT: 17%, Catalina: 22%)
(Table S4). Compared to January and April, dinoflagellates in July and October contributed
fewer sequences (between 2-5%), ciliates were still dominant in July (SPOT: 13%, Catalina:
18%), but less abundant in October (SPOT: 3%, Catalina: 9%), and haptophytes were found to
be at lower abundances in July (SPOT: 10%, Catalina: 9%), and high abundances in October
(SPOT: 34%, Catalina 26%) (Table S4). Additionally, the total number of chlorophyte RNA
sequences was higher in July (SPOT: 13%, Catalina: 23%) and October (SPOT: 10%, Catalina
23%), relative to January and April (Table S4).
Analysis of protistan community dissimilarity from four depths at the SPOT station
revealed greater agreement between the clustering patterns of the RNA and DNA libraries
(Figures 2.3C and 2.3D) relative to the dissimilarity among the three surface stations (Figures
2.3A and 2.3B). The clearest trend among these samples was that both RNA and DNA based
communities from 150 m and 890 m grouped apart from shallower depths (5 m and SCM)
(Figures 2.3C and 2.3D). Diversity from shallower depths was mainly comprised of
dinoflagellates, ciliates, chlorophytes, haptophytes, stramenopiles, and (specific only to DNA)
metazoa (Table S4). Communities at 150 m and 890 m were not significantly different from one
another (P<0.05, dashed lines in Figures 2.3C and 2.3D, except for January 890 m DNA sample),
41
and were mainly comprised of dinoflagellates, ciliates, radiolaria, and (specific only to DNA)
metazoa (Table S4). The one exception at 890 m where the DNA sample in January did not
group with other 890 m samples (Figure 2.3C) coincided to substantially higher abundances of
chlorophyte (3%), haptophyte (18%), and other stramenopile (11%) sequences, relative to other
890 m depths (Table S4).
DNA-based communities at 5 m from all months at the SPOT station were not
significantly different (P<0.05, Figure 2.3C), while season affected the RNA-derived community
structure at 5 m (Figure 2.3D). RNA samples from April and July were not significantly different
from one another (P<0.05, Figure 2.3D). There were higher abundances of MAST sequences in
April (23%) and July (21%) relative to January (3%) and October (6%) (Table S4). SCM
samples from January and October were not significantly different from one another in either
DNA or RNA-derived Bray-Curtis dissimilarity results (P<0.05, Figures 2.3C and 2.3D). The
April sample at SCM did not cluster with other euphotic zone samples (Figures 2.3C and 2.3D),
coinciding with high relative abundances of diatom sequences (RNA: 33%, DNA: 18%, Table
S4).

Relative activity of protistan groups
Interpretations of relative activity (RNA:DNA ratios) were conducted separately for each major
taxonomic group (Table S5) and graphically visualized for select groups (Figure 2.4). Average
RNA:DNA ratios for each major group varied widely (Figure 2.5). Statistically significant
differences in RNA:DNA ratios revealed seasonal and depth-related trends in relative activity
(P<0.05, Figures 2.4, 2.6, and 2.7, Tables S5 and S6).
42
A common seasonal trend in the euphotic zone (surface and SCM at the SPOT station,
Port of LA, or Catalina) was that the majority of protistan groups had significantly higher
RNA:DNA ratios in April, relative to other months (P<0.05, Figure 2.4, Table S5).
Dinoflagellates had significantly higher RNA:DNA ratios in all euphotic zone samples in April
(P<0.05, Figure 2.4A, Tables S5 and S6). In April, diatom (Figures 2.4B and 2.6) and cercozoa
RNA:DNA ratios were significantly higher at the Port of LA, compared to all other months and
depths (P<0.05, Tables S5 and S6). Chlorophytes (Figure 2.4C), haptophytes and MASTs had
significantly higher relative activity in April at the surface and SCM at the SPOT station and at
the Port of LA compared to other months (P<0.05, Tables S5 and S6). Ciliates had higher levels
of relative activity at the SCM (SPOT) and the Port of LA in April (P<0.05, Figure 2.4D, Tables
S5 and S6). Higher RNA:DNA ratios at the SCM at SPOT were also observed for cryptophytes
(P<0.05, Tables S5 and S6).
There were fewer seasonal trends in RNA:DNA ratios in deep samples (150 m and 890
m) than in the euphotic zone. Dinoflagellates had significantly higher RNA:DNA ratios at 890 m
in January (P<0.05, Figure 2.4A, Tables S5 and S6). Relative activity of diatoms (Figure 2.4B)
and ciliates (significant P<0.05, Figure 2.4D) at 150 m was highest in October relative to other
months (Tables S5 and S6).
Average RNA:DNA ratios revealed vertical trends and a few horizontal spatial trends.
Overall, ciliate ratios at the SPOT station were significantly higher at 150 m relative to other
depths (P<0.05, Figure 2.7A, Tables S5 and S6). Haptophytes and cercozoa also had higher
relative activities at 150 m at the SPOT station, while dinoflagellates, cryptophytes and MASTs
had significantly higher ratios at 150 m and 890 m relative to the euphotic zone (P<0.05, Tables
S5 and S6). At the SPOT station in April, radiolarian RNA:DNA ratios were significantly higher
43
at the SCM compared to 150 m and 890 m, although species richness was greater at deeper
depths than at the SCM (P<0.05, Figure 2.7B, Tables S5 and S6).

2.4 Discussion
RNA-DNA comparisons (diversity based on RNA or DNA, RNA:DNA ratios) in this study
revealed insights into natural protistan assemblages at different stations and depths in the eastern
North Pacific. DNA-based molecular surveys have contributed considerable knowledge of
protistan biogeography and diversity [e.g. (Caron et al. 2012; de Vargas et al. 2015; López-
García et al. 2001; Massana et al. 2015)], but there is still relatively little information regarding
members of the community that are active and how protistan activity is impacted by
environmental conditions (Jing et al. 2015; Parris et al. 2014; Stecher et al. 2015).
One major difference between RNA and DNA-derived community composition in this
study was that metazoa were more abundant in DNA samples, while fewer sequences in RNA
samples were identified as metazoa (Tables 2.1, S4, and S6). High abundances of metazoan
sequences in microbial molecular surveys can mask the presence of rare protistan sequences,
thus their contribution is often minimized through pre-filtration during seawater collection, or
removed during downstream bioinformatic analyses (Countway et al. 2010; Schnetzer et al.
2011), see Materials and Methods. Differences between the RNA and DNA libraries indicated
either the presence of extracellular metazoan DNA (e.g. broken cellular material from the pre-
filtration process) or that metazoa have fewer copies of RNA relative to DNA. Regardless, the
small numbers of metazoan sequences found in the RNA libraries were considered advantageous
for future protistan-focused studies that strive to reduce the presence of metazoan sequences.

44
Observed seasonality across surface stations
A major finding of this study was that RNA-based Bray-Curtis dissimilarity analysis more
clearly clustered surface samples according to general environmental and seasonal conditions
(chlorophyll and nutrient concentrations, Table S2) relative to DNA (Figures 2.3A and 2.3B).
Community similarity analysis broadly grouped samples relative to site, and then more finely
according to season (Figure 2.3B). The ability of the RNA-based Bray-Curtis dissimilarity
estimates to differentiate seasons suggests that RNA provided a better snapshot of the active
component of the protistan community relative to DNA (Countway et al. 2010; Schnetzer et al.
2011).
We acknowledge limitations in the use of RNA:DNA ratios to evaluate microbial activity
by making comparisons of relative activity within individual taxonomic groups. Variations in
gene copy number (Godhe et al. 2008; Gong et al. 2013; Prokopowich et al. 2003), cell size
(Godhe et al. 2008; Zhu et al. 2005), and dormant (including cyst-forming) cells make the
interpretation of diversity from DNA sequence counts difficult. Similarly, rRNA copy numbers
vary as a factor of metabolic state or cell size (Blazewicz et al. 2013). Therefore, using a
RNA:DNA ratio to infer metabolic activity of a diverse (both metabolically and taxonomically)
protistan community is complicated. For example, in the present study, the maximum average
dinoflagellate RNA:DNA ratios were approximately 6, while maximum average RNA:DNA
ratios for MASTs were approximately 62 (Figure 2.5, Table S5). A direct comparison of
dinoflagellate and MAST RNA:DNA ratios would imply that dinoflagellates were consistently
less active than MASTs. It is prudent to compare RNA:DNA ratios among related species that
have comparable cell sizes or metabolisms. For this reason, we used average RNA:DNA ratios
45
within manually designated taxonomic groups as an indication of relative activity (see Materials
and Methods).
RNA:DNA ratios provided insights into compositional and seasonal changes in the
activity of phytoplankton taxa. Most protistan groups had high RNA:DNA ratios in April.
Specifically, relative changes in RNA:DNA ratios of diatoms highlighted an intensification of
their activity in April (Figure 2.6, Tables S5 and S6). Springtime increases in diatom absolute
and relative abundance have previously been recorded in the Port of LA and the San Pedro
Channel (Kim et al. 2009; Schnetzer et al. 2013; Seubert et al. 2013; Venrick 2002). In this
study, high relative activity of diatoms in the Port of LA in April co-occurred with the highest
chlorophyll a measurements observed in this study (Figures 2.4B and 2.6, Tables S5 and S2) and
a high biomass of diatoms from microscopy counts (Connell et al. In prep). Diatom OTUs at that
time were dominated by a few dominant taxa (Table S7, Figure 2.6).
Chlorophytes in April, comprised mainly of picoeukaryotes Osterococcus, Micromonas,
and Bathycoccus, had high relative activities at the SCM at SPOT. This finding was expected as
previous studies have noted episodic dominance of minute chlorophytes in the San Pedro
Channel (Countway and Caron 2006; Kim et al. 2014; Kim et al. 2012; Lie et al. 2013; Schnetzer
et al. 2011). Countway and Caron (2006) noted a trend in which high Ostreococcus abundance
often followed diatom blooms. While both chlorophyte and diatom RNA:DNA ratios were high
in the Port of LA and the SCM at the SPOT station in April (Figures 2.4B and 2.4C, Table S5),
sampling frequency in the current study was not sufficient to resolve temporal coupling of
diatom and Ostreococcus abundances and activities. Nevertheless, RNA:DNA ratios could
provide an additional tool for examining temporal relationships among biologically important
phytoplankton groups (Worden 2006).
46

Seasonal and spatial trends along the vertical profile at SPOT
RNA-DNA comparisons along the vertical profile at SPOT revealed depth-specific trends in
community composition and relative activity that were not addressed in DNA-based analyses,
emphasizing the value of combining RNA and DNA-based sequencing efforts. Findings were
consistent with previous studies at the SPOT station, where seasonal patterns in community
structure were generally restricted to shallower depths (surface and SCM), and shallow and deep
(150 and 890 m) protistan communities were distinct (Countway et al. 2010; Kim et al. 2014;
Schnetzer et al. 2011) (Figures 2.3C and 2.3D, Table S4). However, RNA-based community
dissimilarity analysis provided additional insight into how environmental conditions may
influence microbial activity. DNA-derived diversity at 890 m in January grouped separate from
other deep samples due to a large abundance of chlorophytes, haptophytes, and other
stramenopiles, which was not seen in the RNA library (Table S6). This difference appears to
indicate that the DNA library was influenced by sinking inactive or non-viable genetic material.
Functional activities of ciliates revealed seasonal trends in the euphotic zone and an
uptick in activity at 150 m at the SPOT station (oxycline). RNA:DNA ratios for ciliates found in
the euphotic zone were consistently higher in April, relative to other months (Figures 2.4D and
2.7A), which may be coupled to seasonal availability of prey, as ciliates are known to be
significant grazers of bacteria and other protists (Sherr and Sherr 1994; Sherr and Sherr 2002).
Previous studies at SPOT have found evidence of decreases in the relative abundance of ciliate
DNA sequences in deep samples (150 m and 890 m) relative to shallower samples (surface to
SCM) (Kim et al. 2014; Schnetzer et al. 2011). The same trend was observed in DNA samples
from the present study, but relative abundances of ciliate RNA sequences were higher at depth
47
compared to 5 m and SCM (Figure 2.7A, Table S4). Moreover, relative activities of ciliates and
species richness was highest at 150 m (Figure 2.7A, Tables S5 and S4), suggesting that ciliates
play an important role at the oxycline not previously documented at SPOT. Increased relative
abundances and activities have also been described along major transition zones, where a
stratified gradient (e.g. sharp oxygen decrease) may promote high prey abundances for ciliate
grazers [e.g. (Anderson et al. 2012; Edgcomb et al. 2011a; Edgcomb and Pachiadaki 2014; Lin et
al. 2008; Parris et al. 2014; Stock et al. 2009; Stoeck et al. 2007; Wylezich and Jürgens 2011)].
Relative activities of dinoflagellates (Figure 2.4A), haptophytes, and MASTs were also higher at
150 m (Table S5), in concordance with observations by Schnetzer et al. (2011). The authors
speculated that the oxycline supported diverse microbial physiologies.
Molecular approaches allow us to study the diversity of natural protistan communities
using culture-independent methods, and here our RNA and DNA results shed new light on
radiolaria – an enigmatic group of which current knowledge is mainly drawn from surface
species and which are difficult to rear in the laboratory (Burki and Keeling 2014; Caron and
Swanberg 1990; de Vargas et al. 2015; Decelle et al. 2012a; Decelle et al. 2013; Gilg et al.
2010). Previous DNA-based protistan diversity studies have found that radiolaria make up a
substantial component of the community in deep water [e.g. (Countway et al. 2007b; Edgcomb et
al. 2002; Not et al. 2007)]. Radiolarian sequences have also been consistently found at the SPOT
station below the euphotic zone (Countway et al. 2010; Kim et al. 2014; Schnetzer et al. 2011).
However, since the ecology of deep-dwelling radiolaria remains largely undocumented and these
studies were DNA-based, it has remained unresolved if radiolarian sequences detected in deep
water originated from metabolically active radiolaria or non-viable cellular material sinking from
the euphotic zone (Gilg et al. 2010; Not et al. 2007). Our results support the former, suggesting
48
the presence of a significant assemblage of active radiolaria at 150 m and 890 m, indicated by
high RNA:DNA ratios (Figure 2.7B, Table S5). Radiolaria species richness was also highest at
150 m (Figure 2.7B), which was mostly attributed to a larger abundance of polycystine OTUs
compared to other depths (Table S7).

Conclusions
This study is the first to use differences in RNA:DNA ratios to infer relative activity in
individual protistan taxonomic groups on both temporal and spatial scales. We characterized how
the protistan community responded to environmental change, with respect to season and spatially
distinct locations (i.e. Port of LA versus SPOT and Catalina) by documenting how community
composition and potential activity of dominant taxa fluctuated. Frequent sampling for
community RNA:DNA ratios is a method to enhance DNA-based approaches for monitoring
phytoplankton bloom initiation and demise [e.g. Countway and Caron (2006)]. The combined
RNA and DNA approach provided evidence that ciliates play an important role along the
oxycline and demonstrated that RNA:DNA ratios were useful for gathering ecological
information from species that are difficult to characterize, such as radiolaria. DNA-based
molecular methods characterize the diversity and biogeography of protistan communities,
incorporating RNA sequencing enabled us to connect protistan diversity to ecosystem function.
Future studies using combined RNA and DNA sequencing to evaluate metabolically active
protists must address the limitations of the approach and link relative changes in RNA:DNA
ratios to measurements of activities. Characterizing protistan community activity using RNA is
one step towards fully understanding ecosystem function and how microbes respond to
environmental change.
49
2.5 Chapter Two Figures and Tables

Figure 2.1. Map of San Pedro Channel, off the coast of southern California. Black circles
indicate the three locations sampled, the San Pedro Ocean Time-series (SPOT) station (33°33 N,
118°24 W), the Port of Los Angeles (33°42.75 N, 118°15.55 W), and offshore from Santa
Catalina Island (33°27.17 N, 118°28.51 W). Surface (5 m), subsurface chlorophyll maximum,
150 m, and 890 m were sampled at the SPOT station, while surface water was sampled at the
Port of Los Angeles and offshore from Santa Catalina Island. Seawater and environmental
parameters were collected at each location quarterly in April, July October (2013), and January
(2014) (Table S1).

50

Figure 2.2. Environmental parameters throughout the water column at the San Pedro
Ocean Time-series (SPOT) station during January, April, July, and October (Table S1).
Values for (A) temperature (°C), (B) chlorophyll a (µg L
-1
) (note the change in y-axis in B), (C)
oxygen (µmol L
-1
), and (D) salinity (PSU) were obtained from CTD sensor data (Seabird
Electronics) during water collection.
51
Figure 2.3. Cluster dendrograms based on average hierarchical clustering. Dendrograms
depict level of dissimilarity among (A) DNA samples from SPOT, the Port of LA, and Catalina
(surface stations), (B) RNA samples from surface stations, (C) DNA samples from four depths at
the SPOT station (5 m, subsurface chlorophyll maximum (SCM), 150 m, and 890 m), and (D)
RNA samples from the four depths at the SPOT station. Data were normalized by calculating the
relative abundance of each OTU, and then Bray-Curtis dissimilarity matrices were constructed
for surface stations (A and B) and for each depth sampled at SPOT (C and D), using either DNA
(A and C) or RNA (B and D) sequence libraries. The percent dissimilarity among samples is
depicted by horizontal axes. Dashed lines represent samples that were not significantly different
from one another (P<0.05).
52

Figure 2.4. Average RNA:DNA ratios from select taxonomic groups. Ratios represent
relative activity at (from top to bottom) the Port of LA, Catalina, SPOT station surface/5 m,
SCM, 150 m, and 890 m, in each month sampled for (A) dinoflagellates, (B) diatoms, (C)
chlorophytes, (D) ciliates, (E), and radiolaria. Average RNA:DNA ratios for all taxonomic
groups reported in Table S5. Asterisks (*) denote significantly higher (P<0.05) average
RNA:DNA ratios relative to other months (pairwise comparisons, Table S6).

53

Figure 2.5. Range of average RNA:DNA ratios for each taxonomic group. Boxplots
represent variation among the average RNA:DNA ratios in each sample. Whiskers denote
minimum and maximum values. Average RNA:DNA ratios among individual taxonomic groups
were highly variable, making direct comparisons between different groups difficult. See Table
S5 for all ratios.

54

Figure 2.6. Relative abundance of RNA (y-axis) and DNA (x-axis) sequences for each
diatom OTU detected in the euphotic zone. From top to bottom: Port of LA (surface), Catalina
(surface), SPOT (surface/5 m), and SPOT subsurface chlorophyll maximum (SCM). Months are
depicted from left to right. Each data point depicts an OTU (n=total number of OTUs).
RNA:DNA ratios for diatoms were significantly higher at the Port of LA in April, compared to
all other months and depths (P<0.05, Tables S5 and S6).

55

Figure 2.7. Relative abundance of RNA (y-axis) and DNA (x-axis) sequences for ciliate and
radiolarian OTUs. From top to bottom, 5 m, subsurface chlorophyll maximum (SCM), 150 m,
and 890 m at the SPOT station in April. Each data point depicts (A) ciliate or (B) radiolarian
OTUs (n=total number of OTUs). (A) Ciliate RNA:DNA ratios were significantly higher at the
SCM and 150 m at the SPOT station in April (P<0.05, Tables S5 and S6). Species richness of
ciliates was also high at 150 m (n=118). (B) RNA:DNA ratios for radiolaria were significantly
higher at the SCM compared to 150 m and 890 m at SPOT (P<0.05, Tables S5 and S6).
Radiolarian species richness was highest at 150 m (n=101), which was attributed to a higher
number of polycystine OTUs (Table S7). Generally, radiolarian RNA:DNA ratios were not
significantly different among depths (Tables S5 and S6).
56
Table 2.1. Total number of RNA and DNA sequences, and percentage of the total number of
sequences by major taxonomic groups. Values for RNA are shaded.

RNA DNA

Sequences Percent Sequences Percent
Alveolates
Dinoflagellates 373,698 18.2 825,700 37.2
Ciliates 501,779 24.4 161,796 7.29
Other 8,999 0.44 8,792 0.40

Chlorophytes 115,167 5.61 81,233 3.66

Cryptophytes 73,146 3.56 32,586 1.47

Haptophytes 227,453 11.1 99,109 4.47
Stramenopiles
Diatoms 139,817 6.80 88,447 3.99
MAST 136,428 6.64 29,227 1.32
Other 193,555 9.42 38,625 1.74
Rhizaria
Cercozoa 39,954 1.94 30,425 1.37
Radiolaria 39,145 1.91 244,545 11.0

Metazoa 26,236 1.28 467,102 21.1

Other 80,378 3.91 53,632 2.42

Unassigned 98,919 4.81 57,716 2.60

Total: 2,054,674

2,218,935

57
2.6 Chapter Two Literature Cited
Adl, S. M. and others 2012. The revised classification of eukaryotes. J. Eukaryot. Microbiol. 59:
429-493.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic Local
Alignment Search Tool. J. Mol. Biol. 215: 403-410.
Anderson, R., C. Winter, and K. Jürgens. 2012. Protist grazing and viral lysis as prokaryotic
mortality factors at Baltic Sea oxic−anoxic interfaces. Mar. Ecol. Prog. Ser. 467: 1-14.
Armbrust, E. V., and S. R. Palumbi. 2015. Uncovering hidden worlds of ocean biodiversity.
Science 348: 865-867.
Ashelford, K. E., N. A. Chuzhanova, J. C. Fry, A. J. Jones, and A. J. Weightman. 2005. At least
1 in 20 16S rRNA sequence records currently held in public repositories is estimated to
contain substantial anomalies. Appl. Environ. Microbiol. 71: 7724-7736.
Behnke, A., M. Engel, R. Christen, M. Nebel, R. R. Klein, and T. Stoeck. 2011. Depicting more
accurate pictures of protistan community complexity using pyrosequencing of
hypervariable SSU rRNA gene regions. Environ. Microbiol. 13: 340-349.
Blazewicz, S. J., R. L. Barnard, R. A. Daly, and M. K. Firestone. 2013. Evaluating rRNA as an
indicator of microbial activity in environmental communities: limitations and uses. ISME
J. 7: 2061-2068.
Burki, F., and P. J. Keeling. 2014. Rhizaria. Curr Biol 24: R103-R107.
Caporaso, J. G. and others 2010. QIIME allows analysis of high-throughput community
sequencing data. Nat. Methods 7: 335-336.
Caron, D. A. 2001. Protistan herbivory and bacterivory. Methods Microbiol 30: 289-315.
Caron, D. A. 2013a. Towards a molecular taxonomy for protists: Benefits, risks and applications
in plankton ecology. J. Eukaryot. Microbiol. 60: 407-413.
Caron, D. A. 2013b. Towards a molecular taxonomy for protists: Benefits, risks, and applications
in plankton ecology. J. Eukaryot. Microbiol. 60: 407-413.
Caron, D. A., and P. D. Countway. 2009. Hypotheses on the role of the protistan rare biosphere
in a changing world. Aquat. Microb. Ecol. 57: 227-238.
Caron, D. A., P. D. Countway, A. C. Jones, D. Y. Kim, and A. Schnetzer. 2012. Marine protistan
diversity. Annu. Rev. Mar. Sci. 4: 467-493.
Caron, D. A., and N. R. Swanberg. 1990. The ecology of planktonic sarcodines. Rev Aquat Sci
3: 147-180.
Charvet, S., W. F. Vincent, and C. Lovejoy. 2014. Effects of light and prey availability on Arctic
freshwater protist communities examined by high-throughput DNA and RNA
sequencing. FEMS Microbiol. Ecol. 88: 550-564.
Clark, K. R., and R. M. Warwick. 2001. Change in marine communities: An approach to
statistical analysis and interpretation. Plymouth, UK: Primer-E.
Clarke, K. R., P. J. Somerfield, and R. N. Gorley. 2008. Testing of null hypotheses in
exploratory community analyses: similarity profiles and biota-environment linkage. J Exp
Mar Biol Ecol 366: 56-69.
Connell, P., V. Campbell, A. G. Gellene, S. K. Hu, and D. A. Caron. Seasonal and spatial trends
in microbial growth and mortality at three coastal ocean sites. (In prep).
Corinaldesi, C., M. Barucca, G. M. Luna, and A. Dell'Anno. 2010. Preservation, origin and
genetic imprint of extracellular DNA in permanently anoxic deep-sea sediments. Mol.
Ecol. 20: 642-654.
58
Countway, P. D., and D. A. Caron. 2006. Abundance and distribution of Ostreococcus sp. in the
San Pedro Channel, California, as revealed by quantitative PCR. Appl. Environ. Microb.
72: 2496-2506.
Countway, P. D., R. J. Gast, M. R. Dennett, P. Savai, J. M. Rose, and D. A. Caron. 2007a.
Distinct protistan assemblages characterize the euphotic zone and deep sea (2500 m) of
the western North Atlantic (Sargasso Sea and Gulf Stream). Environ. Microbiol. 9: 1219-
1232.
Countway, P. D., R. J. Gast, M. R. Dennett, P. Savai, J. M. Rose, and D. A. Caron. 2007b.
Distinct protistan assemblages characterize the euphotic zone and deep sea (2500 m) of
the western North Atlantic (Sargasso Sea and Gulf Stream). Environ. Microbiol. 9: 1219-
1232.
Countway, P. D., P. D. Vigil, A. Schnetzer, S. D. Moorthi, and D. A. Caron. 2010. Seasonal
analysis of protistan community structure and diversity at the USC Microbial
Observatory (San Pedro Channel, North Pacific Ocean). Limnol. Oceanogr. 55: 2381-
2396.
de Vargas, C. and others 2015. Eukaryotic plankton diversity in the sunlit ocean. Science 348:
DOI: 10.1126/science.1261605.
Debroas, D., M. Hugoni, and I. Domaizon. 2015. Evidence for an active rare biosphere within
freshwater protists community. Mol. Ecol. 24: 1236-1247.
Decelle, J., L. Bittner, C. de Vargas, and F. Not. 2012. An original mode of symbiosis in open
ocean plankton. P. Natl. Acad. Sci. USA 109: 18000-18005.
Decelle, J. and others 2013. Diversity, ecology and biogeochemistry of cyst-forming acantharia
(radiolaria) in the oceans. PloS One 8: e53598. DOI: 53510.51371/journal.pone.0053598.
Decelle, J., S. Romac, E. Sasaki, F. Not, and F. Mahé. 2014. Intracellular diversity of the V4 and
V9 regions of the 18S rRNA in marine protists (radiolarians) assessed by high-throughput
sequencing. PLoS ONE 9: e104297.
Dunthorn, M., J. Klier, J. Bunge, and T. Stoeck. 2012. Comparing the hyper-variable V4 and V9
regions of the small subunit rDNA for assessment of ciliate environmental diversity. J.
Eukaryot. Microbiol. 59: 185-187.
Dunthorn, M. and others 2014. Placing environmental next-generation sequencing amplicons
from microbial eukaryotes into a phylogenetic context. Mol. Biol. Evol. 31: 993-1009.
Edgcomb, V. and others 2011. Protistan microbial observatory in the Cariaco Basin, Caribbean.
I. Pyrosequencing vs Sanger insights into species richness. ISME J. 5: 1344–1356.
Edgcomb, V. P., and M. Pachiadaki. 2014. Ciliates along oxyclines of permanently stratified
marine water columns. J. Eukaryot. Microbiol. 61: 434-445.
Edgcomb, V. P., A. Teske, and M. L. Sogin. 2002. Benthic eukaryotic diversity in the Guaymas
Basin hydrothermal vent environment. P. Natl. Acad. Sci. USA 99: 7658-7662.
Egge, E. S., L. Bittner, T. Anderson, S. Audic, C. de Vargas, and B. Edvardsen. 2013. 454
pyrosequencing to describe microbial eukaryotic community composition, diversity and
relative abundance: a test for marine haptophytes. PloS One 8: e74371. DOI:
74310.71371/journal.pone.0074371.
Egge, E. S. and others 2015. Seasonal diversity and dynamics of haptophytes in the Skagerrak,
Norway, explored by high-throughput sequencing. Mol. Ecol. 24: 3026-3042.
Gilg, I. C., L. Amaral-Zettler, P. D. Countway, S. Moorthi, A. Schnetzer, and D. A. Caron. 2010.
Phylogenetic affiliations of mesopelagic acantharia and acantharian-like environmental
18S rRNA genes off the southern California coast. Protist 161: 197-211.
59
Godhe, A., M. E. Asplund, K. Harnstrom, V. Saravanan, A. Tyagi, and I. Karunasagar. 2008.
Quantification of diatom and dinoflagellate biomasses in coastal marine seawater samples
by real-time PCR. Appl. Environ. Microb. 74: 7174-7182.
Gong, J., J. Dong, X. Liu, and R. Massana. 2013. Extremely high copy numbers and
polymorphisms of the rDNA operon estimated from single cell analysis of oligotrich and
peritrich ciliates. Protist 164: 369-379.
Grattepanche, J.-D., L. F. Santoferrara, G. B. McManus, and L. A. Katz. 2014. Diversity of
diversity: conceptual and methodological differences in biodiversity estimates of
eukaryotic microbes as compared to bacteria. Trends Microbiol. 22: 432-437.
Hadziavdic, K., K. Lekang, A. Lanzen, I. Jonassen, E. M. Thompson, and C. Troedsson. 2014.
Characterization of the 18S rRNA gene for designing universal eukaryote specific
primers. PLoS ONE 9: e87624.
Hugerth, L. W. and others 2014. Systematic design of 18S rRNA gene primers for determining
eukaryotic diversity in microbial consortia. PLoS ONE 9: e95567.
Jing, H., E. Rocke, L. Kong, X. Xia, H. Liu, and M. R. Landry. 2015. Protist communities in a
marine oxygen minimum zone off Costa Rica by 454 pyrosequencing. Biogeosci Discuss
12: 13483-13509.
Karl, D. M., and M. D. Bailiff. 1989. The measurement and distribution of dissolved nucleic
acids in aquatic environments. Limnol. Oceanogr. 34: 543-558.
Kearse, M. and others 2012. Geneious Basic: An integrated and extendable desktop software
platform for the organization and analysis of sequence data Bioinformatics 28: 1647-
1649.
Ki, J.-S. 2012. Hypervariable regions (V1–V9) of the dinoflagellate 18S rRNA using a large
dataset for marker considerations. J Appl Phycol 24: 1035-1043.
Kiefer, D., W. Chamberlin, and C. Booth. 1989. Natural fluorescence of chlorophyll a:
Relationship to photosynthesis and chlorophyll concentration in the western South Pacific
gyre. Limnol. Oceanogr. 34: 868-881.
Kim, D. Y. and others 2014. Monthly to interannual variability of microbial eukaryote
assemblages at four depths in the eastern North Pacific. ISME J. 8: 515-530.
Kim, D. Y., P. D. Countway, W. Yamashita, and D. A. Caron. 2012. A combined sequence-
based and fragment-based characterization of microbial eukaryote assemblages provides
taxonomic context for the Terminal Restriction Fragment Length Polymorphism (T-
RFLP) method. J. Microbiol. Methods 91: 527-536.
Kim, H.-J., A. J. Miller, J. McGowan, and M. L. Carter. 2009. Coastal phytoplankton blooms in
the Southern California Bight. Prog Oceanogr 82: 137-147.
Kim, M., M. Morrison, and Z. Yu. 2011. Evaluation of different partial 16S rRNA gene
sequence regions for phylogenetic analysis of microbiomes. J. Microbiol. Methods 84:
81-87.
Lejzerowicz, F., I. Voltsky, and J. Pawlowski. 2013. Identifying active foraminifera in the Sea of
Japan using metatranscriptomic approach. Deep-Sea Res PT II 86-87: 214-220.
Lie, A. A. Y., D. Y. Kim, A. Schnetzer, and D. A. Caron. 2013. Small-scale temporal and spatial
variations in protistan community composition at the San Pedro Ocean Time-series
station off the coast of southern California. Aquat. Microb. Ecol. 70: 93-110.
Lie, A. A. Y. and others 2014. Investigating microbial eukaryotic diversity from a global census:
Insights from a comparison of pyrotag and full-length sequences of 18S rRNA gene
sequences. Appl. Environ. Microbiol. 80: 4363-4373.
60
Lin, X., M. I. Scranton, A. Y. Chistoserdov, R. Varela, and G. T. Taylor. 2008. Spatiotemporal
dynamics of bacterial populations in the anoxic Cariaco Basin. Limnol. Oceanogr. 53:
37-51.
Logares, R. and others 2014. Patterns of rare and abundant marine microbial eukaryotes. Curr
Biol 24: 813-821.
López-García, P., F. Rodriguez-Valera, and C. Pedros-Alio. 2001. Unexpected diversity of small
eukaryotes in deep-sea Antarctic plankton. Nature 409: 603-607.
Lorenz, M. G., and W. Wackernagel. 1987. Adsorption of DNA to sand and variable degradation
rates of adsorbed DNA. Appl. Environ. Microb. 53: 2948-2952.
Lozupone, C., and R. Knight. 2005. UniFrac: a new phylogenetic method for comparing
microbial communities. Appl. Environ. Microb. 71: 8228-8235.
Magoc, T., and S. L. Salzberg. 2011. FLASH: fast length adjustment of short reads to improve
genome assemblies. Bioinformatics 27: 2957-2963.
Mahé, F. and others 2014. Comparing high-throughput platforms for sequencing the V4 region
of SSU- rDNA in environmental microbial eukaryotic diversity surveys. J. Eukaryot.
Microbiol. e-pub ahead of print 19 Sept 2014, doi:10.1111/jeu.12187-4559.
Massana, R. 2015. Protistan diversity in environmental molecular surveys, p. 3-21. In S.
Ohtsuka, Suzaki, T., Horiguchi, T., Suzuki, N., Not, F. [ed.], Marine Protists: Diversity
and Dynamics. Springer Japan.
Massana, R. and others 2004. Phylogenetic and ecological analysis of novel marine
stramenopiles. Appl. Environ. Microb. 70: 3528-3534.
Massana, R. and others 2015. Marine protist diversity in European coastal waters and sediments
as revealed by high-throughput sequencing. Environ. Microbiol. 17: 4035-4049.
Massana, R., and C. Pedros-Alio. 2008. Unveiling new microbial eukaryotes in the surface
ocean. Curr. Opin. Microbiol. 11: 213-218.
Medlin, L., H. J. Elwood, S. Stickel, and M. L. Sogin. 1988. The characterization of
enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71: 491-499.
Moon-van der Staay, S. Y., R. De Wachter, and D. Vaulot. 2001. Oceanic 18S rDNA sequences
from picoplankton reveal unsuspected eukaryotic diversity. Letters to Nature 409: 607-
610.
Moreira, D., and P. López-García. 2001. The molecular ecology of microbial eukaryotes unveils
a hidden world. Trends Microbiol. 10: 31-38.
Nickrent, D. L., and M. L. Sargent. 1991. An overview of the secondary structure of the V4
region of eukaryotic small-subunit ribosomal RNA. Nucleic Acids Res. 19: 227-235.
Not, F., J. del Campo, V. Balagué, C. de Vargas, and R. Massana. 2009. New insights into the
diversity of marine picoeukaryotes. PloS One 4: e7143. DOI:
7110.1371/journal.pone.0007143.
Not, F., R. Gausling, F. Azam, J. F. Heidelberg, and A. Z. Worden. 2007. Vertical distribution of
picoeukaryotic diversity in the Sargasso Sea. Environ. Microbiol. 9: 1233-1252.
Ohtsuka, S., T. Suzaki, T. Horiguchi, N. Suzuki, and F. Not. 2015. Marine Protists: Diversity and
Dynamics. Springer: Japan.
Parris, D. J., S. Ganesh, V. P. Edgcomb, E. F. DeLong, and F. J. Stewart. 2014. Microbial
eukaryote diversity in the marine oxygen minimum zone off northern Chile. Front
Microbiol 5: 543 DOI: 510.3389/fmicb.2014.00543.
Pawlowski, J., and B. Lecroq. 2010. Short rDNA barcodes for species identification in
foraminifera. J. Eukaryot. Microbiol. 57: 197-205.
61
Pernice, M. C., R. Logares, L. Guillou, and R. Massana. 2013. General patterns of diversity in
major marine microeukaryote lineages. PLoS ONE 8: e57170.
Poulsen, L. K., G. Ballard, and D. A. Stahl. 1993. Use of rRNA fluorescence in situ
hybridization for measuring the activity of single cells in young and established biofilms.
Appl. Environ. Microb. 59: 1354-1360.
Price, M. N., P. S. Dehal, and A. P. Arkin. 2009. FastTree: Computing large minimum evolution
trees with profiles instead of a distance matrix. Mol. Biol. Evol. 26: 1641-1650.
Prokopowich, C. D., T. R. Gregory, and T. J. Crease. 2003. The correlation between rDNA copy
number and genome size in eukaryotes Genome 46: 48-50.
Pruesse, E. and others 2007. SILVA: a comprehensive online resource for quality checked and
aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res. 35:
7188-7196.
Quast, C. and others 2012. The SILVA ribosomal RNA gene database project: improved data
processing and web-based tools. Nucleic Acids Res. 41: D590-D596.
R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for
Statistical Computing.
Rideout, J. R. and others 2014. Subsampled open-reference clustering creates consistent,
comprehensive OTU definitions and scales to billions of sequences. PeerJ 2: e545. DOI:
510.7717/peerj.7545.
Rodríguez-Martínez, R., G. Rocap, R. Logares, S. Romac, and R. Massana. 2012. Low
evolutionary diversification in a widespread and abundant uncultured protist (MAST-4).
Mol. Biol. Evol. 29: 1393-1406.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual
(2nd Ed.), 2nd
ed.
Schloss, P. D. and others 2009. Introducing mothur: Open-source, platform-independent,
community-supported software for describing and comparing microbial communities.
Appl. Environ. Microbiol. 75: 7537-7541.
Schnetzer, A. and others 2013. Coastal upwelling linked to toxic Pseudo-nitzschia australis
blooms in Los Angeles coastal waters, 2005-2007. J. Plankton Res. 35: 1080-1092.
Schnetzer, A., S. D. Moorthi, P. D. Countway, R. J. Gast, I. C. Gilg, and D. A. Caron. 2011.
Depth matters: Microbial eukaryote diversity and community structure in the eastern
North Pacific revealed through environmental gene libraries. Deep-Sea Res. PT. I 58: 16-
26.
Seubert, E. L. and others 2013. Seasonal and annual dynamics of harmful algae and algal toxins
revealed through weekly monitoring at two coastal ocean sites off southern California,
USA. Environ Sci Pollut Res 20: 6878-6895.
Sherr, B. F., E. B. Sherr, D. A. Caron, D. Vaulot, and A. Z. Worden. 2010. A sea of microbes:
Oceanic protists. Oceanography 20: 130-134.
Sherr, E. B., and B. F. Sherr. 1994. Bacterivory and herbivory: key roles of phagotrophic protists
in pelagic food webs. Microb. Ecol. 28: 223-235.
---. 2002. Significance of predation by protists in aquatic microbial food webs. Antonie Van
Leeuwenhoek 81: 293-308.
Stecher, A. and others 2015. rRNA and rDNA based assessment of sea ice protist biodiversity
from the central Arctic Ocean. Eur. J. Phycol.:
DOI:10.1080/09670262.09672015.01077395.
62
Stock, A., K. Jürgens, J. Bunge, and T. Stoeck. 2009. Protistan diversity in suboxic and anoxic
waters of the Gotland Deep (Baltic Sea) as revealed by 18S rRNA clone libraries. Aquat.
Microb. Ecol. 55: 267-284.
Stoeck, T. and others 2010a. Multiple marker parallel tag environmental DNA sequencing
reveals a highly complex eukaryotic community in marine anoxic water. Mol. Ecol. 19
Suppl 1: 21-31.
Stoeck, T. and others 2010b. Multiple marker parallel tag environmental DNA sequencing
reveals a highly complex eukaryotic community in marine anoxic water. Mol. Ecol. 19:
21-31.
Stoeck, T., S. Jost, and J. Boenigk. 2008. Multigene phylogenies of clonal Spumella-like strains,
a cryptic heterotrophic nanoflagellate, isolated from different geographical regions. Int. J.
Syst. Evol. Microbiol. 58: 716-724.
Stoeck, T., A. Zuendorf, H.-W. Breiner, and A. Behnke. 2007. A molecular approach to identify
active microbes in environmental eukaryote clone libraries. Microb. Ecol. 53: 328-339.
Terrado, R., E. Medrinal, C. Dasilva, M. Thaler, W. F. Vincent, and C. Lovejoy. 2011. Protist
community composition during spring in an Arctic flaw lead polynya. Polar Biol. 34:
1901-1914.
Vaulot, D., K. Romari, and F. Not. 2002. Are autotrophs less diverse than heterotrophs in marine
picoplankton? Trends Microbiol. 10: 266-267.
Venrick, E. L. 2002. Floral patterns in the California current system off southern California:
1990-1996. J Mar Res 60: 171-189.
Weekers, P. H., R. J. Gast, P. A. Fuerst, and T. J. Byers. 1994a. Sequence variations in small-
subunit ribosomal RNAs of hartmannella vermiformis and their phylogenetic
implications. Mol. Biol. Evol. 11: 684-690.
Weekers, P. H. H., R. J. Gast, P. A. Fuerst, and T. J. Byers. 1994b. Sequence variations in small-
subunit ribosomal RNAs of Hartmannella vermiformis and their phylogenetic
implications. Mol. Biol. Evol. 11: 684-690.
Welschmeyer, N. A. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll
b and pheopigments. Limnol. Oceanogr. 39: 1985-1992.
Whitaker, D., and M. Christman. 2014. Package 'clustsig'. R package version 1.1
https://cran.hafro.is/web/packages/clustsig/clustsig.pdf.
Wolf, C., E. Silva Kilias, and K. Metfies. 2014. Evaluating the potential of 18S rDNA clone
libraries to complement pyrosequencing data of marine protists with near full-length
sequence information. Mar. Biol. Res. 10: 771-780.
Worden, A. Z. 2006. Picoeukaryote diversity in coastal waters of the Pacific Ocean. Aquat.
Microb. Ecol. 43: 165-175.
Worden, A. Z., M. J. Follows, S. J. Giovannoni, and S. Wilken. 2015. Rethinking the marine
carbon cycle: Factoring in the multifarious lifestyles of microbes. Science 347: 735. DOI:
710.1126/science.1257594.
Wylezich, C., and K. Jürgens. 2011. Protist diversity in suboxic and sulfidic waters of the Black
Sea. Environ. Microbiol. 13: 2939-2956.
Youssef, N., C. S. Sheik, L. R. Krumholz, F. Z. Najar, B. A. Roe, and M. S. Elshahed. 2009.
Comparison of species richness estimates obtained using nearly complete fragments and
simulated pyrosequencing-generated fragments in 16S rRNA gene-based environmental
surveys. Appl. Environ. Microbiol. 75: 5227-5236.
63
Zhu, F., R. Massana, F. Not, D. Marie, and D. Vaulot. 2005. Mapping of picoeucaryotes in
marine ecosystems with quantitative PCR of the 18S rRNA gene. FEMS Microbiol. Ecol.
52: 79-92.

64
Chapter Three: A Hard Day’s Night: Shifts in microbial eukaryotic activity in
the North Pacific Subtropical Gyre

Sarah K. Hu, Paige E. Connell, Lisa Y. Mesrop, & David A. Caron

Department of Biological Sciences, University of Southern California, 3616 Trousdale Parkway,
Los Angeles, California 90089-0371

65
Abstract
Molecular analysis revealed diel rhythmicity in the metabolic activity of single-celled microbial
eukaryotes (protists) at station ALOHA in the North Pacific Subtropical Gyre. Diel trends among
different protistan taxonomic groups reflected distinct nutritional capabilities and temporal niche
partitioning. Changes in relative metabolic activities among phototrophs corresponded to the
light cycle, generally peaking in the mid- to late-afternoon. Phagotrophic protist metabolic
activities were higher at night, relative to daytime, potentially in response to increased
availability of prey. Tightly correlated Operational Taxonomic Units throughout the diel cycle
indicated the existence of parasitic and mutualistic relationships within microbial eukaryotic
communities, underscoring the need to define and include these symbiotic interactions in marine
food webs. This study provided a new high-resolution view into the ecologically important
interactions among primary producers and consumers that mediate the efficient transfer of
carbon through marine food webs. Characterizations of the temporal dynamics of protistan
activities contribute knowledge for predicting how these microorganisms respond to
environmental forcing factors.

66
3.1 Introduction
Single-celled microbial eukaryotes (protists) fulfill vital ecological roles as primary producers
and consumers at the base of marine food webs. Environmental factors (e.g. light-dark cycling)
influence protistan trophic interactions, thus studies to characterize how protistan biological
activity varies or responds to such changes are critical for understanding ecosystem functioning.
Microbial cell cycles are known to be synchronized with regular light-dark cycling (diel),
especially among photoautotrophs. However, there is still a need for a comprehensive
understanding of how taxon-specific functional roles and trophic interactions are shaped by
short-term temporal dynamics (i.e. diel periodicity).
Picophytoplankton behavior, including cell division and growth, are key underlying
biological mechanisms affecting ecosystem function, and thus diel synchronization among these
species has implications for primary productivity and the transfer of carbon to higher trophic
levels. Flow cytometric measurements have demonstrated that phytoplankton cell size typically
increases throughout the course of the light period, followed by a decrease in mean cell size and
increase in cell abundances corresponding to division events (Vaulot and Marie 1999). Cell
division rates among picophytoplankton, Synechococucs, Prochlorococcus, and picoeukaryotes
have been found to peak in the late afternoon and dusk, although the exact timing appears to be
somewhat taxon-specific (Binder and DuRand 2002; Tsai et al. 2009; Vaulot and Marie 1999).
Molecular analyses have provided further support for diel periodicity in cell cycle
regulation among dominant phytoplankton. In situ metatranscriptomic studies found transcripts
associated with energy acquisition pathways to be more abundant during the day, relative to
nighttime (Aylward et al. 2015; Poretsky et al. 2009). Further, whole microbial community
temporal dynamics were likely driven by the metabolic activity of dominant phytoplankton,
67
evidenced by high daily abundances in transcripts responsible for carbon fixation and
photosynthesis among photoautotrophs followed by peaks in transcripts associated with
translation and oxidative phosphorylation in heterotrophic species (Aylward et al. 2015). While
both Aylward and Poretsky detected Ostreococcus (green alga, chlorophyte) as one of the
dominant photoautotrophic contributors to daily peaks in photosynthesis and carbon fixation
pathways, molecular studies devoted to the short-term temporal dynamics (e.g. daily) of in situ
protistan populations are rare.
Our ability to characterize the diversity of in situ assemblages of microbial eukaryotes
has become more feasible with high-throughput sequencing (de Vargas et al. 2015; Le Bescot et
al. 2016; Massana et al. 2015). Recent tag-sequencing efforts which target cellular ribosomal
RNA (rRNA) have provided an approach for determining the metabolically active members of
the community rather than using rRNA gene (rDNA)-derived tag sequences, which are more
likely to include dead, inactive cellular material (Blazewicz et al. 2013). Ribosomal RNA can
serve as a proxy for presumed transcriptional activity and the precursor for protein synthesis
(Corinaldesi et al. 2011; Egge et al. 2015; Lejzerowicz et al. 2013; Massana 2015b; Poulsen et al.
1993).
We investigated diel shifts in community composition and relative activity of naturally
occurring protistan assemblages (18S V4 tag sequencing of rDNA and rRNA, respectively) in
the North Pacific Subtropical Gyre (NPSG) at 4 hour intervals over a period of 3 days
(Lagrangian sampling). While DNA-based estimates of species richness were relatively
unchanging, RNA-derived results demonstrated regular, daily fluctuations in species richness
and relative abundance that typically peaked mid-day. We show how many key photoautotrophic
species exhibited a diel periodicity synchronized to the light period, but were temporally offset
68
from one another. The relative activity of many phagotrophic-capable taxa also had diel
periodicity, which may be linked to the availability of prey (e.g. phytoplankton). Finally, we
inferred intimate ecological relationships, specifically symbiosis, from significantly co-occurring
Operational Taxonomic Units (OTUs), specifically revealing the frequency of parasitism
between Syndiniales and other alveolates and mutualism among rhizarian hosts and
endosymbiotic algae.

3.2 Material and Methods
Sample collection
Seawater samples were collected from a depth of 15 m every 4 hours for 3 days following a
Lagrangian sampling schematic in an anticyclonic eddy in the North Pacific Subtropical Gyre
(Figure 3.1A), as a part of the Simons Collaboration on Ocean Processes and Ecology (SCOPE;
http://scope.soest.hawaii.edu/) cruise efforts in July 2015 (Table S1). Samples were collected
using 10 L Niskin bottles mounted on a CTD rosette at approximately 6AM, 10AM, 2PM, 6PM,
10PM, and 2AM. Corresponding temperature, salinity, dissolved oxygen and chlorophyll a data
were derived from the same CTD casts. For each molecular sample, 3.5 L of seawater was
prefiltered through 80 µm Nitex mesh, to minimize the presence of multicellular eukaryotes, and
vacuum filtered onto sterile GF/F filters (nominal pore size 0.7 µm, Whatman, International Ltd.
Florham Park, NJ, USA). Filters were stored immediately in 1.5 mL of RLT+ buffer (with β-
Mercaptoethanol, Qiagen, Valencia, CA, USA), and flash frozen in liquid nitrogen (protocols.io:
dx.doi.org/10.17504/protocols.io.hisb4ee). Samples for flow cytometry were taken at
approximately the same time points (16 samples) and processed as described in protocols.io
69
(Enumeration of bacteria and cyanobacteria by flow cytometry, protocols.io:
dx.doi.org/10.17504/protocols.io.hisb4ee).

PCR and sequence library preparation
Frozen filters were simultaneously thawed and bead-beaten by adding RNase-free silica beads
and vortexing for 5 minutes. Total DNA and RNA was extracted using the DNA/RNA AllPrep
kit (Qiagen, Valencia, CA, USA, #80204) with an in-line genomic DNA removal step (RNase-
free DNase reagents, Qiagen #79254). Total RNA was reverse transcribed into cDNA using
BioRad (iScript Select cDNA Synthesis Kit #1708896, Hercules, CA). The extraction protocol
can be found online at protocols.io: dx.doi.org/10.17504/protocols.io.hk3b4yn.
The V4 hypervariable region of the 18S rRNA gene was targeted using (Stoeck et al.
2010b) primers (Forward-CCAGCASCYGCGGTAATTCC, Reverse-
TYRATCAAGAACGAAAGT). The V4 hypervariable region was chosen because the length
(~400 bps) provides more phylogenetic resolution and better estimates of diversity compared to
shorter hypervariable regions (Hu et al. 2015). The PCR reactions consisted of a final
concentration of 1X Q5 High Fidelity Master Mix (NEB #M0492S, Ipswich, MA), 0.5 µM each
of forward and reverse primers, and 10 ng of genetic material. PCR thermal profile was derived
from (Rodríguez-Martínez et al. 2012), with an initial activation step (Q5 specific) of 98°C for 2
minutes, followed with 10 cycles of 98°C for 10 seconds, 53°C for 30 seconds, 72°C for 30
seconds, and 15 cycles of 98°C for 10 seconds, 48°C for 30 seconds, and 72°C for 30 seconds,
and a final extension of 72°C for 2 minutes. PCR products were purified using an AMPure bead
clean up (Beckman Coulter #A63881, Brea, CA). Purified PCR product (approximately 400 bp
region in length) concentrations were normalized and indexed using Illumina-specific P5 and P7
70
indices. Indexed samples were pooled at equimolar concentrations and sequenced using MiSeq
250 bp PE sequences (Laragen, Culver City, CA). Molecular protocols can be found at
dx.doi.org/10.17504/protocols.io.hdmb246.

Sequence processing and data analysis
Low quality sequence ends and primers were trimmed using Trimmomatic (Bolger et al. 2014).
Paired end sequences were merged with a 20 bp minimum overlap using fastqjoin (Aronesty
2011) and quality checked (requiring Q score > 30) using QIIME (Caporaso et al. 2010).
Chimeras were removed using vsearch (Rognes T 2016) by searching against the Protist
Ribosomal database (PR2, Guillou et al. 2013). Finally, sequences were clustered into
approximately species designations Operational Taxonomic Units (OTUs) defined as groups of
sequences with at least 97% sequence similarity to one another. OTUs were generated using the
open-reference OTU clustering algorithm (Rideout et al. 2014) in QIIME, which combines both
de novo and reference based (PR2 database) OTU clustering (Guillou et al. 2013). Taxonomy
was assigned based on the PR2 database taxonomic identities using uclust (Edgar 2010) at 90%
similarity. To normalize between sequence libraries (DNA and RNA), all samples were
randomly subsampled to ensure an equal number of sequences per sample. Scripts for sequence
analysis can be found at GitHub (https://github.com/shu251/18Sdiversity_diel) and raw sequence
data are available under SRA BioProject PRJNA393172.

Data synthesis and statistical analyses
Final sequence counts and OTU taxonomic information were synthesized in R (R Core Team
2014); all scripts are available at GitHub (https://github.com/shu251/18Sdiversity_diel) and the
71
OTU table is available as supplementary material. Global singletons (OTUs that exist only once
in the whole dataset and have only one sequence) were removed from the final results.
Sequences in each sample (including both RNA and DNA derived samples) were randomly
subsampled so that all samples had the same number of sequences (equal to the sample with the
fewest sequences). Manual taxonomic group names were assigned at approximately class or
phylum levels in order to visualize the complex microbial community (Hu et al. 2016). Major
taxonomic groups discussed in the downstream analysis comprised > 0.1% of the either the
rRNA or rDNA sequence libraries.
The RNA:DNA ratio was calculated for each OTU (after normalization) in order to infer
relative activity, but OTUs were removed if they did not have either RNA or DNA sequences
(Fu and Gong 2015; Hu et al. 2016). All comparisons of relative abundance (OTU or sequence
abundance) or activity (RNA:DNA ratio) were performed within a single taxonomic group to
avoid problems related to varying gene copy number among the various protistan groups. Means
and standard mean errors of RNA:DNA ratios from each time point were used to examine
relative changes throughout the course of a day.
Two types of statistical tests were employed to investigate periodicity and possible co-
occurrence among taxonomic groups and OTUs. First, the Rhythmicity Analysis Incorporating
Nonparametric method (RAIN) was applied (Thaben and Westermark 2014) to look for
significant rhythmicity (p<0.05) in RNA or DNA sequence abundance for each OTU. RAIN uses
non-parametric methods to detect oscillations in time series data.
Extended Local Similarity Analysis (eLSA) was computed for OTUs based on RNA
sequence abundances, in order to identify temporally significant interactions between OTUs and
environmental conditions (i.e. temperature or time of day); a time lag of 12 hours was
72
incorporated into the analysis to detect temporally shifted co-occurrences (Xia et al. 2013; Xia et
al. 2011). The RNA-sequence library was chosen for eLSA to investigate the metabolically
active component of the community. OTUs based on rRNA were first filtered so that each OTU
had more than one sequence and appeared in all 19 time points. OTUs with no taxonomy
assignment were removed and only highly significant correlations were selected based on the p-
value (p<0.05), q-value (q<0.05), and Spearman rank correlation coefficient (> 0.5 or < -0.5). To
estimate the frequency of potential parasitism, we assumed negatively correlated or positive, but
time-delayed OTUs were indicative of the parasitoid lifestyle, while positively correlated OTUs
reflected the mutualism between rhizarian hosts and putative endosymbiots (Chow et al. 2014).

3.3 Results
Environmental characterization of an anticyclonic eddy
Lagrangian sampling (total of 19 samples) was conducted in an anticyclonic eddy northeast of
station ALOHA (Figure 3.1A). This oligotrophic region of the Pacific Ocean is persistently
nutrient deplete in the upper euphotic zone and experiences only moderate seasonal changes
(Karl and Church 2014). Changes in the majority of environmental variables were minor
throughout the sampling period (Figures 3.1B-C), with the exception of chlorophyll a
fluorescence, which decreased during the day (minimum: 0.19 µg/L) and increased at night
(maximum: 0.27 µg/L; dot-dashed line in Figure 3.1C). Cell counts obtained using flow
cytometry showed that populations of heterotrophic bacteria, Prochlorococcus, Synechococcus,
and picoeukaryote abundances did not change dramatically during the sampling period.
Picoeukaryote cell abundances were highest at night (0.01x10
5
cells/mL) and lowest during the
transition from day to night (0.004x10
5
cells/mL) (Figure S1).
73

Sequence results
The 18S rRNA gene (rDNA) and ribosomal RNA (rRNA) sequence libraries resulted in an
average of 142,995 and 106,787 sequences per sample, which clustered into an average of 1,459
and 986 OTUs per sample, respectively (Figure S2). To normalize across the entire dataset,
samples from both rDNA and rRNA sequence libraries were randomly subsampled to 42,719
sequences per sample, which left a total of 3,353 OTUs from the combined rDNA and rRNA
sequence library.

Temporal shifts in total community richness and diversity
The total number of rRNA OTUs varied more than 2-fold across the 19 samples collected
throughout the study (min: 458 OTUs, max: 1,159 OTUs; red line in Figure 3.2), while the total
number of OTUs based on rDNA remained similar (1,000-1,200 OTUs, black line in Figure 3.2).
Numbers of rRNA OTUs were highest in the mid- to late-afternoon, which typically
corresponded to the total number of rDNA OTUs (compare red and black lines, Figure 3.2).
Figures 3.3 and S3 showed that the community composition derived from rDNA did not change
over the sampling period (Figures 3.3A and 3.3C), while fluctuations in the total number of
rRNA OTUs were driven by changes in community composition (Figures 3.3B and 3.3D).
Specifically, relative increases in the total number of rRNA OTUs (red line in Figure 3.2)
corresponded to increases in the total number of dinoflagellate (Figure 3.3B), haptophyte,
chlorophyte, and diatom rRNA OTUs (Figure 3.3D).

74
Similar to richness, relative abundances of rRNA sequences fluctuated throughout the
diel sampling period while rDNA sequence abundances remained static for most taxonomic
groups (Figure 3.4). The rDNA sequence library was comprised of almost 60% dinoflagellate
reads, including Karlodinium, Gyrodinium, Prorocentrum, and Dinophyceae, followed by ca.
20% metazoan reads, and ca. 10% Syndiniales Group-I (Table S2, Figure 3.4A). Approximately
1/3
rd
of the rRNA sequence library were dinoflagellates (ca. 33%), including Karlodinium,
Gyrodinium, and Dinophyceae, this was followed by ciliates, which were further identified as
Oligohymenophorea (ca. 15%), and then stramenopiles (ca. 11%), mainly diatoms and members
from the MArine STramenopile group (MAST)-3 (Table S2, Figure 3.4B). Canonical
Correspondence Analysis (CCA) with respect to time of day revealed variability among samples
based on rRNA to be greater than samples based on rDNA (CCA1=18.5 vs. CCA1=8.75%,
respectively, Figure S4).
Significant diel periodicities among dinoflagellate, ciliate, Syndinales, haptophyte, and
stramenopile OTUs were identified in the rRNA dataset (but not rDNA) based on the non-
parametric RAIN analysis (Table 3.1, p<0.05). There were only 11 rDNA OTUs compared to 87
rRNA OTUs with significant diel rhythmicity (Table 3.1, p<0.05). OTUs found to have
significant diel periodicity accounted for 44.5% of the total rDNA sequence library and 46.4% of
the rRNA sequence library. Only 2 OTUs were found to have significant diel rhythmicity
inferred from rRNA and rDNA sequence results, the ciliate Strombidium and a rhizarian OTU,
acantharia (OTUs denoted with asteriks in Table S3).

75
Relative metabolic activity varied as a factor of taxonomic group and time of day
Patterns of relative metabolic activity throughout the course of a day were investigated by
calculating the mean and standard mean error for the RNA:DNA ratios of OTUs at each time
point based on assigned taxonomic group (Figures 3.5 and 3.6). Excluding OTUs with missing
rDNA or rRNA sequences (also see Hu et al. 2016; Laroche et al. 2017) resulted in 1,423 total
OTUs (89% of the total sequences, Table 3.1). Taxonomic groups were broadly classified as
either phototrophic/mixotrophic (chloroplast-containing) or non-phototrophic to examine trends
in RNA:DNA ratios over the diel period (Figures 3.5 and 3.6, respectively). Taxon-specific gene
copy number, cell size, and metabolic capabilities contribute variation in rRNA tag-sequencing
results (Blazewicz et al. 2013). We acknowledge this variability by comparing ratios only within
major taxonomic groups (also see Hu et al. 2016).
RNA:DNA ratios of dinoflagellates and diatoms peaked at dusk (6PM; Figure 3.5).
Individual dinoflagellate OTUs with significant diel rhythmicity included (n=34): Gyrodinium,
Karlodinium, Pentapharsodinium, Suessiales, Prorocentrum, and several non-classified
Dinophyceae OTUs (Table 3.1, p<0.05). Diatom RNA:DNA ratios at night (6PM-2AM) were
generally higher relative to daytime (6AM-6PM, Figure 3.5) and two OTUs (Raphid-pennate,
based on PR2 assignment) exhibited significant periodicity (Table 3.1 and S3, p<0.05).
Relative metabolic activity among pelagophytes, haptophytes, and chlorophytes peaked
mid-day (10AM or 2PM) and again at night (10PM or 2AM, Figure 3.5). No chlorophyte or
pelagophyte OTUs were found to have significant rhythmicity, while 3 haptophyte OTUs
belonging to the classes Chrysochromulineaceae and Phaeocystis were found to have significant
periodicity (Table 3.1 and S3, p<0.05).
76
RNA:DNA ratios for ciliates (which are primarily non-phototrophic) and MAST peaked
at dusk and again towards the end of the dark cycle (6PM and 2AM, Figure 3.6). 15 of 245
ciliate OTUs exhibited significant rhythmicity, which included species belonging to
Litostomatea, Spirotrichea, Prostomatea, Oligohymenophorea, and Phyllopharyngea (Table 3.1
and S3, p<0.05). Relative activities among Syndiniales had a clear diel trend, with a sharp
increase in relative metabolic activity from daytime to dusk (2PM-6PM) followed by a decrease
throughout the dark cycle (Figure 3.6). Syndiniales OTUs with significant rhythmicity belonged
to Groups I, II, and III (Table 3.1, p<0.05). Rhizarian groups, including acantharia and radiolaria
(RAD groups A, B, and C), generally had higher relative metabolic activities during the night
compared to daytime, but only one OTU (acantharia) was found to have significant rhythmicity
(Figure 3.6, Table 3.1; p<0.05).

Potential mutualistic relationships supported by significantly co-correlating OTUs
eLSA analysis based on rRNA OTUs (n=201) and associated metadata resulted in over
22,366 total correlations. The analysis yielded 4,242 significant correlations after filtering for
highly significant OTU pairs (see Materials and Methods). The top 20 correlations between
major taxonomic groups (based on frequency of significantly co-occurring OTUs) are
summarized in Table 3.2, a full list is reported in Table S4. Dinoflagellate and ciliate OTUs
made up the majority of the significant interactions, followed by stramenopiles (Table 3.2 and
S4).
Results from eLSA were examined for possible symbiosis, specifically Syndiniales
parasitism of other protists or mutualism between rhizarian hosts and endosymbionts (Table 3.3).
Syndiniales OTUs with either negative or positive with time-delayed relationships with other
77
OTUs made up 55% of the significant correlations (209/380 OTU pairs; Table 3.3A and S4). The
majority of correlated OTUs with Syndinilaes were identified as other alveolates, ciliates or
dinoflagellates, or stramenopiles (Table 3.3A). 47.8% of the significantly co-occurring OTUs
with rhizaria were positive (100/209; Table 3.3B and S4). All rhizarian OTUs were identified as
acantharia (Table S4), and were mainly found to significantly correlate with ciliate and
dinoflagellate OTUs (Table 3.3B).

3.4 Discussion
Studies with high-resolution sampling on shorter time scales, such as this one, contribute
important knowledge of the underlying mechanisms that govern microbial community structure
and function. Metabolic coordination to the diel cycle was characterized in euphotic zone
protists, revealing how taxa exhibiting different trophic strategies fulfill temporally separate
ecological niches. Temporally-dependent trends in relative metabolic activity allowed us to infer
ecologically relevant trophic linkages, including predator-prey interactions and symbiotic
relationships. Regular, long-term ecological studies in the North Pacific Subtropical Gyre
(NPSG) have provided valuable information on the complex microbial interactions that drive
ecosystem-level biogeochemical cycling (Church et al. 2013). Here, we expand on current
knowledge of protistan roles in marine food webs by documenting the short-term temporal
dynamics of protistan behavior – which can ultimately be used to predict community responses
to environmental changes.

78
Total species richness and metabolic reassembly based on rDNA and rRNA
Daily shifts in community diversity were more clearly represented in the rRNA sequence results
compared to rDNA results (Figures 3.2, 3.3, 3.4, S3, and S4). rRNA-derived estimates of
community diversity have previously been shown to be more sensitive to environmental and
seasonal conditions (Charvet et al. 2014b; Hu et al. 2016). Species richness of the protistan
community in surface waters of the NPSG based on rDNA remained relatively unchanged during
this study, while rRNA-derived estimates of species richness yielded a temporally varying subset
of the protistan community (Figures 3.2 and 3.3). These changes represented differences in
protistan activity at different times of day and corresponded to changes in community
composition. We hypothesize that diel periodicity observed in the total rRNA OTUs represented
changes in transcriptional activity within the protistan population, where decreased cellular
activity causes a given OTU to fall below the level of detection via high-throughput sequencing.
The taxonomic composition of total OTUs provides further support for this hypothesis, where
samples with relatively higher total numbers of rRNA OTUs corresponded to samples with
separate taxonomic composition (compare red line in Figure 3.2 to Figures 3.3B and 3.3D). This
was most apparent in comparisons between day and night samples; relative to samples taken
during the dark cycle, the taxonomic composition of daytime samples were enriched with
dinoflagellate and haptophyte OTUs, and to some extent, chlorophyte, pelagophyte, and diatom
OTUs.

Changes in the relative abundance of rRNA reads for each major taxonomic group were
also repeated daily throughout the sampling period (Figure 3.4) and a higher number of rRNA
OTUs were found to have significant diel rhythmicity based on RAIN analysis compared to
79
rDNA OTUs (Table 3.1). These results provided evidence that rRNA-derived results offer a way
to follow microbial community responses to changing conditions. Previous work found that
minimal spatial and temporal variability during sample collection did not negatively impact long-
term (e.g. monthly or seasonal) ecological studies (Lie et al. 2013). We expand this observation:
studies based on DNA sequencing provide temporally stable estimates of total community
diversity suitable to detect monthly or seasonal shifts, while those derived from RNA sequencing
can reflect short-term community dynamics. Therefore, sample collection time of day must be
taken into account in order to characterize daily or weekly microbial responses to environmental
changes.
Similar timing in peak relative sequence abundances and metabolic activities (RNA:DNA
ratios) for several taxonomic groups across the diel cycle suggested that common ecological
dynamics may influence different protistan lineages on very short time scales (Figures 3.4-3.6).
However, the exact timing of these peaks were often dependent on taxonomic group, presumably
reflecting the varying importance in availability of solar energy, prey, or other dissolved
nutrients to different functional groups of protists.

Phytoplankton metabolic activity synced to the light cycle
Diurnal trends in the biological activities of marine phytoplankton have been reported in both
culture- and field-based studies, including periodicity in photosynthetic capacity (Harding et al.
1982; Lorenzen 1963), chlorophyll a accumulation (Owens et al. 1980), the uptake and
assimilation of nutrients (Bender et al. 2012; Eppley et al. 1971a; Eppley et al. 1971b; Goering et
al. 1964), cell division (Nelson and Brand 1979; Smayda 1975), and transcriptional regulation of
cellular processes (Askworth et al. 2013; Bender et al. 2012; Smith et al. 2016). While the
80
magnitude and phase of these trends is typically taxonomic group-specific (Nelson and Brand
1979), light-dark cycling is the shared environmental variable. Overall increased biological
activity, especially photosynthesis among photoautotrophs, will promote cell growth, increase
carbon content per cell, and result in larger average cell sizes among photoautotrophs (Binder
and DuRand 2002). Increases in relative metabolic activity (higher RNA:DNA ratio), in this
study, may therefore be attributed to higher rRNA content per cell, assuming the rDNA remains
constant (Zhu et al. 2005). RNA:DNA ratios generally increased from morning to evening
among primarily phototrophic taxa (Figure 3.5); likely reflecting increased photosynthesis and
carbon fixation with the availability of sunlight.
The relative metabolic activity of chlorophytes (inferred from RNA:DNA ratios)
correlated with the light cycle: increasing throughout the day, with a peak mid-day/afternoon,
and then decreasing at night (Figure 3.5). This was consistent with two molecular-based studies,
Aylward et al. and Poretsky et al., that found Ostreococcus (belonging to the chlorophyte group)
energy acquisition transcripts to peak during the day. Further, Aylward et al. found a diel
progression, where transcripts associated with carbon fixation peaked in the morning and were
followed by photosynthesis-related transcripts mid-day (Aylward et al. 2015). Other taxonomic
groups demonstrated increased relative metabolic activity throughout the day, but RNA:DNA
ratio maxima were often temporally offset (e.g. 6PM for diatoms, 10AM for pelagophytes, or
6PM for chlorophytes and haptophytes, Figure 3.5) or had secondary peaks during the dark cycle
(e.g. 2AM for diatoms, pelagophytes, and chlorophytes, Figure 3.5).
Peaks in relative metabolic activity during the dark cycle among primarily phototrophic
taxa may be a result of continuing to synthesize intracellular compounds, perhaps relying on
carbon stores for energy. Phytoplankton continue synthesizing protein throughout the nighttime
81
at rates that may equal or exceed synthesis during the day (Cuhel et al. 1984), which may explain
secondary increases in RNA:DNA ratios observed in diatoms, pelagophytes, and chlorophytes in
Figure 3.5. In support of this observation, both field- and culture-based work have found
continued transcription in phytoplankton during dark periods (Askworth et al. 2013; Aylward et
al. 2015; Poretsky et al. 2009). For example, diatoms in light-dark cycle culture conditions were
found to switch from the expression of genes associated with the synthesis of chloroplast-related
proteins during the light cycle to mitochondria-related proteins during the dark cycle (Bender et
al. 2012), and in Smith et al. (2016), diatoms in the dark cycle were found to increase
transcription of genes associated with cellular processes such as the catabolism of stored carbon.
Another factor contributing to temporally distinct trends in phytoplankton metabolic
activity are species-specific differences in nutrient uptake kinetics. The rate at which
microorganisms take up nutrients is size dependent (Finkel et al. 2009; Hein et al. 1995), which
subsequently impacts group-specific photosynthetic activity and growth rate. This may explain
the temporally offset peaks in RNA:DNA ratios observed in larger diatoms compared to smaller
chlorophytes in Figure 3.5. Interestingly, similar to the diel trend in diatom activity (Figure 3.5),
diel measurements of C:Chl a ratios in a natural phytoplankton population dominated by diatoms
was found to increase during the transition from light to dark and decrease at the start of the next
light cycle (Owens et al. 1980). Additionally, increases in relative metabolic activity during the
transition from dark to light may also be attributed to anticipatory increases in transcripts related
to photosynthetic machinery (Askworth et al. 2013). Understanding temporal fluctuations in the
biological activity of ecologically photoautotrophs, such as chlorophytes and diatoms, is key for
characterizing environmental parameters that may impact primary production.
82
One other explanation for the secondary peaks in RNA:DNA ratios found in Figure 3.5 is
the potential for species to switch or supplement phototrophy with phagotrophic ingestion of
prey (i.e. mixotrophy). Increases in relative metabolic activity in the absence of light may reflect
grazing activity among taxonomic groups that include species known to be mixotrophic,
including, pelagophytes, haptophytes, and dinoflagellates. Microscopical analysis from the same
cruise found mixotrophic nanoflagellates to not display significant diel periodicity (Connell et al.
In prep). Together, this suggests that while the ingestion of prey by mixotrophic nanoflagellates
does not have a clear diel trend, some mixotrophic species may exhibit diel trends in overall
metabolic activity (reflected in RNA:DNA ratios). Nonetheless, additional work is required to
directly link mixotrophic ingestion of prey to transcriptional activity in situ, especially as
mixotrophic contributions to biogeochemical transformations are ecologically significant (Ward
and Follows 2016).

Diel trends in heterotrophic metabolic activity implied predator-prey interactions
While the synchronization of light-driven phototrophic protists with available solar energy is
more obvious, diel periodicity observed in most phagotrophic protists was unexpected. For most
phagotrophic taxa there was higher relative metabolic activities at night compared to daytime
(Figure 3.6) and several groups were found to have significant diel periodicities, including
dinoflagellates, haptophytes, ciliates, and MAST (Table 3.1 and S2).
Temporal studies at station ALOHA have found cell division rates among
picophytoplankton to consistently peak at dusk, which presumably increases the number of prey
available to grazers (Binder and DuRand 2002; Tsai et al. 2009; Vaulot and Marie 1999; Wilson
et al. 2017). Additional work from the same cruise as this study found that despite cell division
83
events consistently occurring at dusk, picophytoplankton abundances remained stable,
suggesting that mortality via protistan grazing activity and viral lysis contribute to consistent
picophytoplankton abundances (Connell et al. In prep; Ribalet et al. 2015). Higher RNA:DNA
ratios at night compared to daytime among phagotrophic protists demonstrated a putative
predator-prey interaction, where increased prey abundance following peak picophytoplankton
division events, may promote increased grazing activity (Figure 3.6). Supporting this
observation, Connell et al. found grazing rates among heterotrophic nanoflagellates to have
significant diel periodicity, with the highest grazing rates recorded at night.
Since a large percentage of primary production in the euphotic zone is consumed by
protistan grazers (Calbet and Landry 2004; Schmoker et al. 2012; Strom et al. 1997), it is
important to characterize the composition and metabolic activity of the grazer community (Strom
2008). There are few studies devoted to identifying the composition of the protistan grazer
community using molecular approaches; however, two previous studies which tracked the
consumption of isotopically-labeled prey, cyanobacteria Prochlorococcous and Synechococcus
(Frias-Lopez et al. 2009) and Micromonas (Orsi et al. 2018), confirmed the dominant members
of the grazer assemblage to consist of haptophytes, dinoflagellates, MAST, and ciliates.
Heterotrophic rhizaria have also been observed to prey on dinoflagellates, ciliates, haptophytes
(Anderson et al. 1984). In this study, we were also able to infer similar predator-prey interactions
from significantly co-occurring OTUs, which included dinoflagellates, ciliates, MAST, and
rhizaria (Massana et al. 2009; Sherr and Sherr 1994; Sherr and Sherr 2007), with putative prey
(Table 3.2 and S4).

84
Significantly co-correlated OTUs inferred protistan parasitism and mutualism
Symbiotic associations between protistan species are well documented, but typically not included
in current food web models, due to the complex and diverse nature of these symbioses (Worden
et al. 2015b). Here, we examined symbioses among protists by targeting possible partnerships
using eLSA to identify significant time-dependent associations among OTUs (Xia et al. 2011).
While we cannot definitively identify the nature of these interactions, we characterized
significantly co-occurring OTUs based on the ecology of the putative symbioses: parasitism with
Syndiniales or mutualism with rhizaria.
Syndiniales are an exclusively parasitic order known to obligately kill their hosts (Guillou
et al. 2008), therefore parasitism in marine environments may be a significant source of mortality
in aquatic food webs (Lima-Mendez et al. 2015). Based on the parasitoid ecology, the strongest
putative parasite-host relationships were hypothesized to be either negative or positive but time-
delayed interactions (see also: Chow et al. 2014). 55% of the significant interactions (n=209)
with Syndinilaes were either negative or positive with a time-delay and the majority of these
interactions were with their most common host, dinoflagellates (Table 3.3A). Network analyses
from a recent global scale ocean survey (de Vargas et al. 2015; Lima-Mendez et al. 2015)
detected a significant number of parasitic Syndinales, but these interactions were derived from
rDNA sequences, therefore variable rDNA copies may have inflated the number of OTU
correlations (Krabberød et al. 2017). The use of rRNA sequences in this study supported putative
host-parasite relationships between Syndinales and other alveolates (Table 3.2, 3A, and S4) and
demonstrated a potentially underestimated source of protistan mortality and carbon export near
station ALOHA.
85
Recent global surveys of protistan diversity have documented the unrealized abundance
and contribution to carbon export by rhizaria (Biard et al. 2016; Burki and Keeling 2014; Caron
2016; Guidi et al. 2016). While difficult to study, due to the delicate nature of their skeletal
structures and lack of cultured representatives (Ohtsuka et al. 2015), molecular surveys have
been more successful in characterizing the in situ ecological roles of rhizaria (Biard et al. 2016;
Guidi et al. 2016; Hu et al. 2016). Rhizaria occupying the euphotic zone commonly form
symbiotic partnerships with phototrophs (Decelle et al. 2015; Decelle et al. 2012b; Ohtsuka et al.
2015). Especially in oligotrophic environments, this mutualistic relationship contributes to host
nutrition and creates a hub of primary productivity (Michaels 1988). Rates of primary production
can be much higher in association with the holobiont (rhizarian host plus endosymbiont) than in
the surrounding environment (Caron et al. 1995b; Caron and Swanberg 1990); therefore,
understanding the nature of this holobiont has implications for improving our ability to measure
the amount of carbon available in marine food webs. We found 100 OTUs to be significantly
correlated with rhizaria, based on eLSA results (Table 3.3B). Specifically, positive interactions
with acantharian hosts included likely endosymbionts, such as dinoflagellates and haptophytes
(Table 3.3B). Consistent with our findings, many dinoflagellate and haptophyte
(Chrysochromulina) endosymbionts have been found with acantharian hosts, sometimes with
multiple microalgal symbionts (Decelle et al. 2015; Decelle et al. 2012b) (Table 3.3B). Rhizaria
in the NPSG (in either solitary or colony form) may exceed the size of our prefiltration (see
Materials and Methods), thus the frequency of acantharia with endosymbionts may be
underestimated. Nevertheless, we demonstrated how this ecologically significant interaction can
be estimated in situ.

86
Summary
This study contributed new insight into the protistan group-specific metabolic activities
occurring over a diel cycle and the ecology of key trophic interactions in the oligotrophic NPSG.
Depending on the frequency and nature of these trophic interactions, the amount of carbon yield
via primary production or subsequent transfer via grazing in marine food webs may vary.
Therefore, understanding the short-term temporal dynamics of protistan ecological roles
enhances our ability to predict how protistan trophic interactions may change under varying
conditions. Species richness estimates revealed diel periodicity in transcriptionally active
component of the community, demonstrating how we can infer community responses to
environmental change from molecular surveys. The light cycle was found to correspond to
progressively increasing relative metabolic activity (i.e. rRNA:rDNA ratios) in primarily
phototrophic taxa. Protistan grazers, both heterotrophic and mixotrophic had significant diel
trends in relative metabolic activity, emphasizing the important role grazers play in structuring
microbial communities. Additionally, time-dependent partnerships from the diel cycle revealed
trophic interactions with parasitic Syndinales and mutualistic symbiont-bearing heterotrophs,
highlighting the ecologically significant, but understudied, contribution of symbiosis to marine
food web dynamics. Results from this study characterized the temporal dynamics governing key
protistan trophic interactions, including the availability of light or prey; these observations can be
applied in future efforts to predict how environmental changes may influence protistan
community dynamics.

87
Acknowledgments: This work was funded by the Simons Foundation as part of the Simon’s
Collaboration on Ocean Processes and Ecology (SCOPE), grant # P49802 awarded to DAC.
Authors would like to thank the captain and crew of the R/V Kilo Moana and the SCOPE
operations team. In particular, we would like to acknowledge Sam Wilson as the Chief Scientist
on the cruise and Benedetto Barone for generation of Sea Level Anomalies map. For helpful
discussion on the statistics used in this manuscript authors would like to thank Jacob Cram,
members of the Weitz lab group, especially Stephen Beckett.

88
3.5 Chapter Three Figures and Tables

Figure 3.1. (A) Location of Lagrangian sampling within an anticyclonic eddy in the North
Pacific Subtropical Gyre, colors represent Sea Level Anomalies (SLA) in cm adapted from:
Wilson et al. (2017). Additional information regarding sample collection can be found in Table
S1. Environmental parameters were obtained from CTD casts for each time point at the depth
sampled (15 m). Variables included (from left to right axes): (B) temperature in Celsius (solid
line with circles), salinity (dashed line), (C) chlorophyll a fluorescence concentration in µg/L
(dot-dash line) and dissolved oxygen in µmol/L (solid line with diamonds). Shaded regions in B-
C indicate dark cycle (night period).

Figure 1.
A"
89

Figure 3.2. Total number of Operational Taxonomic Units (OTUs) based on rDNA and rRNA
results (black and red lines, respectively) throughout the diel cycle (x-axis), which is
representative of species richness. The number of rRNA OTUs changed more than two-fold over
the diel period and was typically higher in the mid- to late- afternoon (min: 458, max: 1,159
OTUs). rDNA estimates of species richness (total OTUs) remained approximately the same
throughout the diel sampling period (1,000-2,000 OTUs). Shaded regions indicate dark cycle
(night period).

Total OTUs
90

Figure 3.3. Taxonomic composition of total OTUs at each sampling point based on either the
rDNA (A and C) or rRNA (B and D) sequence library. Since alveolate sequences dominated
each sample (Table S2), alveolates (A and B) and non-alveolates (C and D) are visualized
separately. A full breakdown of taxonomic groups by total OTUs can be found in Figure S3.

A
B
C
D
91

Figure 3.4. Area plots represented the relative sequence abundance of rDNA (top) or rRNA
(bottom) reads for each major taxonomic group. Black horizontal lines indicate periods of
darkness (6PM-6AM). Only major taxonomic groups that made up > 0.1% of the total rDNA or
rRNA reads are represented here; a full list of sequence abundances can be found in Table S2.
Dinoflagellates dominated both sequence libraries (rDNA and rRNA), but the rRNA-based
library had fewer total dinoflagellate sequences. The second most abundant sequences in the
rDNA sequence library belonged to Syndiniales, while ciliates, MArine STramenopiles (MAST),
and diatoms were next in abundance in the rRNA library.

92

Figure 3.5. Diel trends in relative metabolic activity inferred from changes in mean RNA:DNA
ratios among phototrophic/mixotrophic taxonomic groups: dinoflagellates, diatoms,
pelagophytes, haptophytes, chlorophytes, and chrysophytes. For each time point, the average
RNA:DNA ratio for OTUs within the same taxonomic group was calculated to better illustrate
relative changes throughout the course of a day (circles). Shading surrounding each line
represents the standard mean error of the RNA:DNA ratios for OTUs assigned to each taxonomic
group. Total number of OTUs included for analysis and a summary of the OTUs found to have
significantly diel periodicity are reported in Table 1. Grey shaded regions indicate dark cycle
(night period). To better represent the cyclical nature of the study, 6AM was plotted twice.

93

Figure 3.6. Diel trends in relative metabolic activity inferred from changes in mean RNA:DNA
ratios among non-phototrophic taxonomic groups: ciliates, MArine STramenopiles (MAST),
Syndiniales, and rhizarian groups (including acantharia, cercozoa, polycystines, and radiolaria).
For each time point, the average RNA:DNA ratio for OTUs within the same taxonomic group
was calculated to better illustrate relative changes throughout the course of a day (circles).
Shading surrounding each line represents the standard mean error of the RNA:DNA ratios for
OTUs assigned to each taxonomic group. Total number of OTUs included for analysis and a
summary of the OTUs found to have significantly diel periodicity are reported in Table 1. Grey
shaded regions indicate dark cycle (night period). To better represent the cyclical nature of the
study, 6AM was plotted twice.

94
Table 3.1. Total number of OTUs, total number of OTUs found to have significant rhythmicity
based on RAIN analysis (p<0.05), and total number of OTUs with both RNA and DNA
sequences (Figures 3.5-3.6). rRNA or rDNA OTU identity lists additional taxonomic information
for OTUs found to have significantly diel rhythmicity (RAIN analysis, see Materials and
Methods); a full list is available in Table S3. Taxonomic identities are derived from the PR2
database (Guillou et al. 2013), which may fall short of full species-level characterizations for
uncultured representatives (i.e. Other-unclassified).

Table 1.
Taxonomic group
Total
OTUs
Total
rRNA
OTUs rRNA OTU identity
Total
rDNA
OTUs rDNA OTU identity
Total
OTUs
with
RNA:DN
A ratio
Alveolates-Ciliates 245 15
Litostomatea, Spirotrichea, Prostomatea,
Oligohymenophorea, Phyllopharyngea
1 Spirotrichea 133
Alveolates-Dinoflagellates 1344 34
Gyrodinium, Dinophyceae, Karlodinium,
Pentapharsodinium, Suessiales, Prorocentrum,
and Other-unclassified
3 Prorocentrum, Dinophyceae 509
Alveolates-Other 2 2
Alveolates-Syndiniales 736 12 Dino-Group-II, Dino-Group-III, Dino-Group-I 2 Dino-Group-II 342
Archaeplastids-Chlorophytes 44 19
Archaeplastids-Other 9 0
Cryptophytes 2 1
Excavates 1 0
Haptophytes 132 3 Chrysochromulinaceae, Phaeocystaceae 63
Opisthokont-Fungi 19 2 Chytridiomycota 8
Opisthokont-Metazoa 234 1 Crustacea 4 Urochordata, Crustacea 50
Opisthokonts-Other 29 6 Choanoflagellatea 16
Other/unknown 100 1 Hacrobia 44
Rhizaria-Acantharia 51 1 Acantharea 1 Acantharea 37
Rhizaria-Cercozoa 12 8
Rhizaria-Polycystines 51 14
Rhizaria-RAD (A,B,C) 27 16
Stramenopiles-Chrysophytes 14 6
Stramenopiles-Diatoms 43 2 Raphid-pennate 28
Stramenopiles-MAST 116 3 MAST-1, MAST-3, MAST-7 45
Stramenopiles-Other 70 5
Rhizochromulinales, Thraustochytriaceae,
MOCH-2, and Other-unclassified
45
Stramenopiles-Pelagophytes 12 3
Unassigned 60 2 Other-unclassified 34
Totals: 3353 87 11 1423
OTUs with significant diel rhythmicity (RAIN analysis)
95
Table 3.2. Frequency and percentage of significantly co-occurring OTUs based on Local
Similarity Analysis results (see Materials and Methods for significance thresholds). Table reports
the top 20 OTU interactions based on assigned taxonomic group. A full list of significantly co-
correlated OTUs is in Table S4.

Table 2.
Interactions
Number of
Pairs
Percentage
of total
Alveolates-Dinoflagellates & Alveolates-Dinoflagellates 631 14.9
Alveolates-Ciliates & Alveolates-Dinoflagellates 629 14.8
Alveolates-Dinoflagellates & Stramenopiles-MAST 256 6.0
Alveolates-Dinoflagellates & Stramenopiles-Other 204 4.8
Alveolates-Ciliates & Alveolates-Ciliates 194 4.6
Alveolates-Ciliates & Stramenopiles-MAST 139 3.3
Alveolates-Dinoflagellates & Alveolates-Syndiniales 127 3.0
Alveolates-Ciliates & Stramenopiles-Other 124 2.9
Alveolates-Dinoflagellates & Stramenopiles-Diatoms 106 2.5
Alveolates-Dinoflagellates & Haptophytes 90 2.1
Alveolates-Dinoflagellates & Stramenopiles-Chrysophytes 81 1.9
Alveolates-Ciliates & Alveolates-Syndiniales 68 1.6
Alveolates-Dinoflagellates & Rhizaria-Acantharia 66 1.6
Alveolates-Ciliates & Stramenopiles-Diatoms 61 1.4
Alveolates-Dinoflagellates & Opisthokont-Fungi 48 1.1
Stramenopiles-MAST & Stramenopiles-Other 48 1.1
Alveolates-Ciliates & Stramenopiles-Chrysophytes 47 1.1
Alveolates-Ciliates & Haptophytes 42 1.0
Alveolates-Ciliates & Rhizaria-Acantharia 39 0.9
Alveolates-Syndiniales & Stramenopiles-MAST 36 0.8
96
Table 3.3. Summary of significantly co-correlated OTUs associated with either (A) Syndiniales
or (B) rhizaria. These two taxonomic groups were chosen to examine the frequency of putative
parasitic (Syndiniales) or mutualistic (rhizarian) relationships within the protistan community.
For Syndiniales, putative parasitic relationships were hypothesized to be from either negative or
positive with a time-delay interactions (A). Positive interactions were hypothesized to represent
the mutualistic relationship between heterotrophic rhizarian hosts and endosymbiotic algae.
Additional taxonomic information for OTUs correlated to wither Syndinilaes or rhizaria are
listed; taxonomic identities are derived from the PR2 database (Guillou et al. 2013), which may
fall short of full species-level characterizations for uncultured representatives (i.e. Other-
unclassified). A full list of significantly co-correlated OTUs is in Table S4.

Table 3.
A. Significantly correlated
with Syndiniales Taxonomic detail
Total
significant
OTU
interactions
Alveolates-Ciliates
Oligohymenophorea, Spirotrichea, Colpodea,
Phyllopharyngea, & Litostomatea
47
Alveolates-Dinoflagellates
Protoperidinium, Cochlodinium, Karlodinium, Dinophyceae,
Gyrodinium, Prorocentrum, Suessiales, Heterocapsa, &
Pentapharsodinium
43
Stramenopiles-MAST MAST-1, MAST-2, MAST-3, MAST-4, MAST-7, & MAST-9 28
Stramenopiles-Other Bolidomonas, MOCH-2, Pedinellales, & Other-unclassified 20
Other/unknown Other-unclassified 13
Stramenopiles-Diatoms
Guinardia, Ditylum, Cylindrotheca, Haslea, & Other-
unclassified
11
Stramenopiles-Chrysophytes Clade-G, Clade-H, & Clade-I 9
Unassigned Other-unclassified 9
Rhizaria-Acantharia Acantharea, Lychnaspis, & Other-unclassified 9
Alveolates-Syndiniales Dino-Group-II, Dino-Group-III 6
Opisthokont-Fungi Other-unclassified 5
Environmental parameter Dissolved O2, light, dark, & salinity 4
Cryptophytes Other-unclassified 2
Stramenopiles-Pelagophytes Pelagomonas 2
Haptophytes Chrysochromulinaceae
1
Total: 209
B. Significantly correlated
with Rhizaria Taxonomic detail
Total positive
interactions
Alveolates-Ciliates
Spirotrichea,Oligohymenophorea, Colpodea,
Phyllopharyngea, & Litostomatea
19
Alveolates-Dinoflagellates
Protoperidinium, Karlodinium, Cochlodinium, Prorocentrum,
Gyrodinium, Dinophyceae, & Pentapharsodinium
29
Alveolates-Syndiniales Dino-Group-II, Dino-Group-III 7
Archaeplastids-Chlorophytes Pyramimonadales 1
Haptophytes Chrysochromulinaceae 3
Opisthokont-Fungi Other-unclassified 2
Other/unknown Other-unclassified 2
Stramenopiles-Chrysophytes Clade-G, Clade-H, & Clade-I 5
Stramenopiles-Diatoms Ditylum, Haslea, Guinardia 5
Stramenopiles-MAST MAST-1, MAST-3, & MAST-9 8
Stramenopiles-Other Bolidomonas, MOCH-2, & Other-unclassified 12
Stramenopiles-Pelagophytes Pelagomonas 2
Unassigned Other-unclassified 3
Environmental parameter Chlorophyll a & light 2
Total: 100
97
3.6 Chapter Three Literature Cited
Anderson OR, Swanberg NR, Bennett P. 1984. An estimate of predation rate and relative
preference for algal versus crustacean prey by a spongiose skeletal radiolarian. Mar Biol.
78:205-207.
Aronesty E. 2011. ea-utils : Command-line tools for processing biological sequencing data.
Askworth J, Coesel S, Lee A, Armbrust EV, Orellana MV, Baliga NS. 2013. Genome-wide diel
growth state transitions in the diatom Thalassiosira pseudonana. P Natl Acad Sci USA.
110:7518-7523.
Aylward FO, Eppley JM, Smith JM, Chavez FP, Scholin CA, DeLong EF. 2015. Microbial
community transcriptional networks are conserved in three domains at ocean basin scales.
P Natl Acad Sci USA. 112:5443-5448.

Bender SJ, Parker MS, Armbrust EV. 2012. Coupled Effects of Light and Nitrogen Source on
the Urea Cycle and Nitrogen Metabolism over a Diel Cycle in the Marine Diatom
Thalassiosira pseudonana. Protist. 163:232-251.

Biard T, Stemmann L, Picheral M, Mayot N, Vandromme P, Hauss H, Gorsky G, Guidi L, Kiko
R, Not F. 2016. In situ imaging reveals the biomass of giant protists in the global ocean.
Nature. 532:1-16.

Binder BJ, DuRand MD. 2002. Diel cycles in surface waters of the equatorial Pacific. Deep-Sea
Res PT II.2601-2617.

Blazewicz SJ, Barnard RL, Daly RA, Firestone MK. 2013. Evaluating rRNA as an indicator of
microbial activity in environmental communities: limitations and uses. ISME J. 7:2061-
2068.

Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence
data. Bioinformatics. 30:2114-2120.

Burki F, Keeling PJ. 2014. Rhizaria. Curr Biol. 24:R103-R107.

Calbet A, Landry MR. 2004. Phytoplankton growth, microzooplankton grazing, and carbon
cycling in marine systems. Limnol Oceanogr. 49:51-57.

Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Peña
AG, Goodrich JK, Gordon JI, et al. 2010. QIIME allows analysis of high-throughput
community sequencing data. Nat Methods. 7:335-336.

Caron DA. 2016. Ocean science: The rise of Rhizaria. Nature. 532:444-445.

Caron DA, Michaels AF, Swanberg NR, Howse FA. 1995. Primary productivity by symbiont-
bearing planktonic sarcodines (Acantharia, Radiolaria, Foraminifera) in surface waters
near Bermuda J Plankton Res. 17:103-129.
Caron DA, Swanberg NR. 1990. The ecology of planktonic sarcodines. Aquat Sci. 3:147-180.
98
Charvet S, Vincent WF, Lovejoy C. 2014. Effects of light and prey availability on Arctic
freshwater protist communities examined by high ‐throughput DNA and RNA
sequencing. FEMS Microbiol Ecol. 88:550-564.
Chow C-ET, Kim DY, Sachdeva R, Caron DA, Fuhrman JA. 2014. Top-down controls on
bacterial community structure: microbial network analysis of bacteria, T4-like viruses
and protists. The ISME Journal. 8:816-829.

Church MJ, Bjorkman KM, Karl DM. 2013. HOT turns 25: A quarter century of Hawaii Ocean
Time-series OCB News. 6:1-35.

Connell P, Ribalet F, Armbrust EV, Caron DA. In prep. Nutritional requirements and prey cell
cycles dictate temporal feeding strategies of heterotrophic and mixotrophic nanoplankton.

Corinaldesi C, Barucca M, Luna GM, Dell'Anno A. 2011. Preservation, origin and genetic
imprint of extracellular DNA in permanently anoxic deep-sea sediments. Mol Ecol.
20:642-654.

Cuhel RL, Ortner PB, Lean DRS. 1984. Night synthesis of protein by algae. Limnol Oceanogr.
29:731-744.

de Vargas C, Audic S, Henry N, Decelle J, Mahé F, Logares R, Lara E, Berney C, Le Bescot N,
Probert I, et al. 2015. Eukaryotic plankton diversity in the sunlit ocean. Science.
348:DOI: 10.1126/science.1261605.

Decelle J, Colin S, Foster RA. 2015. Photosymbiosis in Marine Planktonic Protists Tokyo:
Springer Japan. p. 465-500.

Decelle J, Siano R, Probert I, Poirier C, Not F. 2012. Multiple microalgal partners in symbiosis
with the acantharian Acanthochiasma sp. (Radiolaria). Symbiosis. 58:233-244.

Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics.
26:2460-2461.

Egge ES, Johannessen TV, Andersen T, Eikrem W, Bittner L, Larsen A, Sandaa R-A, Edvardsen
B. 2015. Seasonal diversity and dynamics of haptophytes in the Skagerrak, Norway,
explored by high-throughput sequencing. Mol Ecol. 24:3026-3042.

Eppley RW, Carlucci AF, Holm-Hansen O, Kiefer D, McCarthy JJ, Venrick EL, Williams PM.
1971. Phytoplankton growth and composition in shipboard cultures supplied with nitrate,
ammonium, or urea as the nitrogen source. Limnol Oceanogr. 16:741-751.

Eppley RW, Rogers JN, McCarthy JJ, Sournia A. 1971. Light/dark periodicity in nitrogen
assimilation of the marine phytoplankters Skeletonema costatum and Coccolithus huxleyi
in N-limited chemostat culture. J Phycol. 7:150-154.

99
Finkel ZV, Beardall J, Flynn KJ, Quigg A, Rees TAV, Raven JA. 2009. Phytoplankton in a
changing world: cell size and elemental stoichiometry. J Plankton Res. 32:119-137.

Frias-Lopez J, Thompson A, Waldbauer J, Chisholm SW. 2009. Use of stable isotope-labelled
cells to identify active grazers of picocyanobacteria in ocean surface waters. Environ
Microbiol. 11:512-525.

Fu R, Gong J. 2015. Single Cell Analysis Linking Ribosomal (r)DNA and rRNA Copy Numbers
to Cell Size and Growth Rate Provides Insights into Molecular Protistan Ecology. J
Eukaryot Microbiol. 38:42-49.

Goering JJ, Dugdale RC, Menzel DW. 1964. Cyclic diurnal variations in the uptake of ammonia
and nitrate by photosynthetic organisms in the Sargasso Sea. Limnol Oceanogr. 9:48-451.

Guidi L, Chaffron S, Bittner L, Eveillard D, Larhlimi A, Roux S, Darzi Y, Audic S, Berline L,
Brum J, et al. 2016. Plankton networks driving carbon export in the oligotrophic ocean.
Nature. 532:465-470.

Guillou L, Bachar D, Audic S, Bass D, Berney C, Bittner L, Boutte C, Burgaud G, de Vargas C,
Decelle J, et al. 2013. The Protist Ribosomal Reference database (PR2): a catalog of
unicellular eukaryote Small Sub-Unit rRNA sequences with curated taxonomy. Nucleic
Acids Res. 41:D597-D604.

Guillou L, Viprey M, Chambouvet A, Welsh RM, Kirkham AR, Massana R, Scanlan DJ,
Worden AZ. 2008. Widespread occurrence and genetic diversity of marine parasitoids
belonging to Syndiniales (Alveolata). Environ Microbiol. 10:3349-3365.

Harding LW, Prézelin BB, Sweeney BM, COX JL. 1982. Diel Oscillations of the
Photosynthesis-Irradiance (P-I) Relationship in Natural Assemblages of Phytoplankton.
Mar Biol. 67:167-178.

Hein M, Pedersen MF, Sand-Jensen K. 1995. Size-Dependent Nitrogen Uptake in Micro- and
Macroalgae. Mar Ecol Prog Ser. 118:247-253.

Hu SK, Campbell V, Connell P, Gellene AG, Liu Z, Terrado R, Caron DA. 2016. Protistan
diversity and activity inferred from RNA and DNA at a coastal ocean site in the eastern
North Pacific. FEMS Microbiol Ecol. 92:1-13.

Hu SK, Liu Z, Lie AAY, Countway PD, Kim DY, Jones AC, Gast RJ, Cary SC, Sherr EB, Sherr
BF, et al. 2015. Estimating Protistan Diversity Using High-Throughput Sequencing. J
Eukaryot Microbiol. 62:688-693.

Karl DM, Church MJ. 2014. Microbial oceanography and the Hawaii Ocean Time-series
programme. Nat Rev Microbiol. 12:699-713.

100
Krabberød AK, Bjorbækmo M, Shalchian-Tabrizi K, Logares R. 2017. Exploring the oceanic
microeukaryotic interactome with metaomics approaches. Aquat Microb Ecol. 79:1-12.

Laroche O, Wood SA, Tremblay LA, Lear G, Ellis JI, Pochon X. 2017. Metabarcoding
monitoring analysis: the pros and cons of using co-extracted environmental DNA and
RNA data to assess offshore oil production impacts on benthic communities. PeerJ.
5:e3347-3324.

Le Bescot N, Mahé F, Audic S, Dimier C, Garet M-J, Poulain J, Wincker P, de Vargas C, Siano
R. 2016. Global patterns of pelagic dinoflagellate diversity across protist size classes
unveiled by metabarcoding. Environ Microbiol. 18:609-629.

Lejzerowicz F, Voltsky I, Pawlowski J. 2013. Identifying active foraminifera in the Sea of Japan
using metatranscriptomic approach. Deep-Sea Res PT II. 86-87:214-220.

Lie AAY, Kim DY, Schnetzer A, Caron DA. 2013. Small-scale temporal and spatial variations
in protistan community composition at the San Pedro Ocean Time-series station off the
coast of southern California. Aquat Microb Ecol. 70:93-110.

Lima-Mendez G, Faust K, Henry N, Decelle J, Colin Sb, Carcillo F, Chaffron S, Ignacio-
Espinosa JC, Roux S, Vincent F, et al. 2015. Determinants of community structure in the
global plankton interactome. Science. 348:1-10.

Lorenzen CJ. 1963. Diurnal variation in photosynthetic activity of natural phytoplankton
populations. Limnol Oceanogr. 8:56-62.

Massana R. 2015. Protistan diversity in environmental molecular surveys. In: Marine Protists:
Diversity and Dynamics. Tokyo: Springer Japan. p. 3-21.

Massana R, Gobet A, Audic S, Bass D, Bittner L, Boutte C, Chambouvet A, Christen R, Claverie
J-M, Decelle J, et al. 2015. Marine protist diversity in European coastal waters and
sediments as revealed by high-throughput sequencing. Environ Microbiol. 17:4035-4049.

Massana R, Unrein F, Rodríguez-Martínez R, Forn I, Lefort T, Pinhassi J, Not F. 2009. Grazing
rates and functional diversity of uncultured heterotrophic flagellates. ISME J. 3:588-596.

Michaels AF. 1988. Vertical-Distribution and Abundance of Acantharia and Their Symbionts.
Mar Biol. 97:559-569.

Nelson DM, Brand LE. 1979. Cell division periodicity in 13 species of marine phytoplankton on
a light:dark cycle. J Phycol. 15:67-75.

Ohtsuka S, Suzaki T, Horiguchi T, Suzuki N, Not F. 2015. Marine Protists: Diversity and
Dynamics: Springer: Japan.

101
Orsi WD, Wilken S, del Campo J, Heger T, James E, Richards TA, Keeling PJ, Worden AZ,
Santoro AE. 2018. Identifying protist consumers of photosynthetic picoeukaryotes in the
surface ocean using stable isotope probing. Environ Microbiol. 41:65-13.

Owens TG, Falkowski PG, Whitledge TE. 1980. Diel Periodicity in Cellular Chlorophyll
Content in Marine Diatoms. Mar Biol. 59:71-77.

Poretsky RS, Hewson I, Sun S, Allen AE, Zehr JP, Moran MA. 2009. Comparative day/night
metatranscriptomic analysis of microbial communities in the North Pacific subtropical
gyre. Environ Microbiol. 11:1358-1375.

Poulsen LK, Ballard G, Stahl DA. 1993. Use of rRNA fluorescence in situ hybridization for
measuring the activity of single cells in young and established biofilms. Appl Environ
Microb. 59:1354-1360.

R: A language and environment for statistical computing [2014. Version http://www.R-
project.org/. Vienna, Austria.

Ribalet F, Swalwell J, Clayton S, Jiménez V, Sudek S, Lin Y, Johnson ZI, Worden AZ,
Armbrust EV. 2015. Light-driven synchrony of Prochlorococcus growth and mortality in
the subtropical Pacific gyre. P Natl Acad Sci USA. 112:8008-8012.

Rideout JR, He Y, Navas-Molina JA, Walters WA, Ursell LK, Gibbons SM, Chase J, McDonald
D, Gonzalez A, Robbins-Pianka A, et al. 2014. Subsampled open-reference clustering
creates consistent, comprehensive OTU definitions and scales to billions of sequences.
PeerJ. 2:e545. DOI: 510.7717/peerj.7545.

Rodríguez-Martínez R, Rocap G, Logares R, Romac S, Massana R. 2012. Low evolutionary
diversification in a widespread and abundant uncultured protist (MAST-4). Mol Biol
Evol. 29:1393-1406.

Rognes T FT, Nichols B, Quince C, Mahé F. 2016. VSEARCH: a versatile open source tool for
metagenomics. PeerJ.4:e2584.

Schmoker C, Hernández-León S, Calbet A. 2012. Microzooplankton grazing in the oceans:
impacts, data variability, gaps of knowledge, and future directions J Plankton Res.
35:691-706.

Sherr EB, Sherr BF. 1994. Bacterivory and herbivory: key roles of phagotrophic protists in
pelagic food webs. Microb Ecol. 28:223-235.

Sherr EB, Sherr BF. 2007. Heterotrophic dinoflagellates: a significant component of
microzooplankton biomass and major grazers of diatoms in the sea. Mar Ecol Prog Ser.
352:187-197.

102
Smayda TJ. 1975. Phased cell division in natural populations of the marine diatom Ditylum
brightwelli and the potential significance of diel phytoplankton behavior in the sea. Deep-
Sea Res. 22:151-165.

Smith SR, Gillard JTF, Kustka AB, McCrow JP, Badger JH, Zheng H, New AM, Dupont CL,
Obata T, Fernie AR, et al. 2016. Transcriptional Orchestration of the Global Cellular
Response of a Model Pennate Diatom to Diel Light Cycling under Iron Limitation. PLoS
Genet. 12:e1006490-1006439.

Stoeck T, Bass D, Nebel M, Christen R, Jones MDM, Breiner H-W, Richards TA. 2010.
Multiple marker parallel tag environmental DNA sequencing reveals a highly complex
eukaryotic community in marine anoxic water. Mol Ecol. 19:21-31.

Strom SL. 2008. Microbial ecology of ocean biogeochemistry: a community perspective.
Science. 320:1043-1045.

Strom SL, Benner R, Ziegler S. 1997. Planktonic grazers are a potentially important source of
marine dissolved organic carbon. Limnol Oceanogr. 42:1364-1374.

Thaben P, Westermark P. 2014. Detecting Rhythms in Time Series with RAIN. J Biol Rhythms.
29:391-400.

Tsai A-Y, Wen-Mei C, Kuo-Ping C. 2009. Diel patterns of grazing by pigmented nanoflagellates
on Synechococcus spp. in the coastal ecosystem of subtropical western Pacific.
Hydrobiologia. 636:249-256.

Vaulot D, Marie D. 1999. Diel variability of photosynthetic picoplankton in the equatorial
Pacific. J Geophys Res Oceans. 104:3297-3310.

Ward BA, Follows MJ. 2016. Marine mixotrophy increases trophic transfer efficiency, mean
organism size, and vertical carbon flux. P Natl Acad Sci USA. 113:2958-2963.

Wilson ST, Aylward FO, Ribalet F, Barone B, Casey JR, Connell PE, Eppley JM, Ferrón S,
Fitzsimmons JN, Hayes CT, et al. 2017. Coordinated regulation of growth, activity and
transcription in natural populations of the unicellular nitrogen-fixing cyanobacterium
Crocosphaera. Nat Microbiol. 2:1-9.

Worden AZ, Follows MJ, Giovannoni SJ, Wilken S, Zimmerman AE, Keeling PJ. 2015.
Rethinking the marine carbon cycle: Factoring in the multifarious lifestyles of microbes.
Science. 347:1257594.

Xia LC, Ai D, Cram J, Fuhrman JA, Sun F. 2013. Efficient statistical significance approximation
for local similarity analysis of high-throughput time series data. Bioinformatics. 29:230-
237.

103
Xia LC, Steele JA, Cram JA. 2011. Extended local similarity analysis (eLSA) of microbial
community and other time series data with replicates. BMC Syst Biol. 5:S15.

Zhu F, Massana R, Not F, Marie D, Vaulot D. 2005. Mapping of picoeucaryotes in marine
ecosystems with quantitative PCR of the 18S rRNA gene. FEMS Microbiol Ecol. 52:79-
92.

104
Chapter Four: Shifting metabolic priorities among key protistan taxonomic
groups within and below the euphotic zone

Sarah K. Hu
1*
, Zhenfeng Liu
1
, Harriet Alexander
2
, Victoria Campbell
3
, Paige E. Connell
1
, Sonya
T. Dyhrman
4
, Karla B. Heidelberg
1
, & David A. Caron
1

1
Department of Biological Sciences, University of Southern California, Los Angeles, CA,
2
Department of Population Health and Reproduction, University of California Davis, Davis, CA,
3
UW Medicine, Division Allergy and Infectious Diseases, Seattle, WA,
4
Department of Earth
and Environmental Sciences, Lamont-Doherty Earth Observatory, Columbia University,
Palisades, NY

105
Abstract
A metatranscriptome study targeting the protistan community was conducted off the coast of
Southern California, at the San Pedro Ocean Time-series station at the surface, 150 m (oxycline),
and 890 m to link putative metabolic patterns to distinct protistan lineages. Comparison of
relative transcript abundances revealed depth-related shifts in the nutritional modes of key
taxonomic groups. Eukaryotic gene expression in the sunlit surface environment was dominated
by phototrophs, such as diatoms and chlorophytes, and high abundances of transcripts associated
with synthesis pathways (e.g. photosynthesis, carbon fixation, fatty acid synthesis). Sub-euphotic
depths (150 m and 890 m) exhibited strong contributions from dinoflagellates and ciliates, and
were characterized by transcripts relating to digestion or intracellular nutrient recycling (e.g.
breakdown of fatty acids and V-type ATPases). These transcriptional patterns underlie the
distinct nutritional modes of ecologically important protistan lineages that drive marine food
webs, and provide a framework to investigate trophic dynamics across diverse protistan
communities.

106
4.1 Introduction
Natural assemblages of microbial eukaryotes are dominated by a huge diversity of protistan
species. Protists fulfill multiple roles in marine food webs due to their diverse morphologies,
behaviors, and metabolic capabilities (Adl et al. 2012; Caron et al. 2012; de Vargas et al. 2015;
Hess et al. 2016; Ohtsuka et al. 2015; Worden et al. 2015b). Understanding their nutritional
strategies, including phototrophy, heterotrophy, and mixotrophy, is essential for characterizing
their ecological interactions, responses to environmental conditions, and for modeling the
emergent properties of these communities. Recent global surveys of protistan diversity based on
molecular approaches have uncovered higher species richness at all depths of the water column
(de Vargas et al. 2015; Pernice et al. 2015; Pernice et al. 2016). Understanding the in situ
functional roles of these species is necessary to assess their roles in marine biogeochemical
cycles.
Transcriptomic analyses have augmented traditional physiological studies of protists by
identifying the core genes and pathways that serve as a way to determine if an organism is reliant
on a primarily autotrophic or heterotrophic mode of nutrition (Beisser et al. 2017; Koid et al.
2014; Liu et al. 2016). Primarily photoautotrophic modes of nutrition in phytoplankton are
generally characterized by higher expression of genes related to photosynthetic machinery and
the downstream fixation and intracellular partitioning of carbon in the cell (Smith et al. 2012).
Further, transcript-based studies have revealed information on the metabolic adaptations of
phytoplankton to specific environmental conditions, such as light or nutrient availability (e.g.
Bender et al. 2014; Beszteri et al. 2012; Dyhrman et al. 2012; Frischkorn et al. 2014; Harke et al.
2017; Liu et al. 2015b; Meyer et al. 2015; Moustafa et al. 2010; Wurch et al. 2011).
107
Compared to phototrophy, the metabolic pathways for phagotrophic ingestion of prey are
not as well documented, but recent work on mixed nutrition in protists (combined phototrophic
and heterotrophy; mixotroph) has identified genes enriched in organisms dependent on prey
ingestion. These studies have included investigations of organelle sequestration (Johnson 2015;
Johnson et al. 2007; Lasek-Nesselquist et al. 2015; Santoferrara et al. 2014), endosymbioses
(Balzano et al. 2015), and phytoflagellate mixotrophy (Lie et al. 2017; Lie et al. 2018; Liu et al.
2016; Liu et al. 2015a). For instance, increased expression of genes associated with the catabolic
breakdown of compounds or vacuolar activity in mixotrophic species provided putative target
genes for a heterotrophic physiology (Liu et al. 2016). Beisser et al. (2017) found heterotrophic
strains of chrysophyte species to have reduced expression of genes integral for carbon fixation
and photosysthetic machinery and increased expression of genes involved in the breakdown and
adsorption of ingested organic material.
Metatranscriptomic surveys of natural communities of microbial eukaryotes have
recently become feasible with advances in sequencing capability and improved reference
databases, and are rapidly becoming a tool for probing the “black box” of environmental
microbial ecology. This approach has enabled the identification of shifts in metabolic potential of
multiple taxonomic groups simultaneously with respect to environmental forcing factors (e.g.
coastal vs. oligotrophic conditions, light or salinity gradients; Alexander et al. 2015a; Allen et al.
2012; Aylward et al. 2015; Dupont et al. 2015; Grossmann et al. 2016; Pearson et al. 2015;
Zielinski et al. 2016).
To date, in situ protistan-specific metatranscriptomic surveys have largely focused on
communities within the euphotic zone (Alexander et al. 2015a; Alexander et al. 2015b; Carradec
et al. 2018; Dupont et al. 2015; Marchetti et al. 2012). Efforts to characterize subsurface
108
protistan diversity have demonstrated the important contribution that these species make in sub-
euphotic, low-oxygen microbial food webs (Edgcomb 2016; Pernice et al. 2015; Pernice et al.
2016; Stoeck et al. 2014). Previous work has found evidence for transcriptionally active
unicellular eukaryotes (mainly fungi) in deep hypersaline anoxic basins based on the presence of
actin and tubulin transcripts (Edgcomb et al. 2016; Pachiadaki et al. 2014b) and other subsurface
environments (Orsi et al. 2013). However, details of the metabolic potential of protistan species
at water column depths below the euphotic zone still remain undocumented, warranting further
exploration.
Gene expression profiles in and below the euphotic zone at the San Pedro Ocean Time-
series (SPOT) station were used to examine the distribution of ecologically significant nutritional
modes in five taxonomic groups: dinoflagellates, ciliates, haptophytes, diatoms, and
chlorophytes. The surface environment was characterized by synthesis-related pathways such as
photosynthesis and downstream fixation of carbon, while below the euphotic zone, at 150 m and
890 m, there were higher abundances of transcripts associated with lysosomes, and vacuolar type
ATPases, and the breakdown of fatty acids. Depth-related differences in transcript abundance
(based on several replicates) allowed us to form molecular-level descriptions of in situ nutritional
strategies among dominant taxonomic groups, which included evidence for anaerobic
metabolism among ciliates at 150 m and 890 m, and adaptation to prolonged darkness in diatoms
and chlorophytes. Exploring the physiological traits of dominant protists throughout the water
column allowed us to infer depth-specific roles of protistan-mediated production and turnover of
carbon.

109
4.2 Results
Environmental parameters and cell counts
Water temperatures at the SPOT station in May were 17.1°C at the surface, 9.6°C at 150 m and
5.2°C at 890 m (Figure 4.1B). Chlorophyll a fluorescence was 0.03 µg/L at the surface,
decreasing to un-detectable (<0.01 µg/L) at ≈80 m (Figure 4.1B). Oxygen levels at the surface
were 4.95 ml/L, relatively hypoxic at 150 m (1.98 ml/L) and nearly anoxic at 890 m (0.15 ml/L;
Figure 4.1B, Table S4). All nutrient concentrations (µM) were low at the surface, consistent with
the oligotrophic nature of the site, relative to values at 150 m and 890 m (Table S4).
Microscopical cell counts of all microbial groups decreased from the surface to 150 m
and 890 m (Table S4). Microplankton (20-200 µm), pigmented (phototrophic and mixotrophic)
and heterotrophic nanoplankton (2.0-20 µm), and picophytoplankton (0.2-2.0 µm), including
picoeukaryotes, Synechoccocus, Prochlorococcus, were highest at the surface (Table S4), and
representative of spring-time abundances for these assemblages (Caron et al. 2017b).

Annotation
Assemblies from each depth yielded a total of 7.8 million contigs, generating 5.4 million putative
protein sequences (Table S5). Contigs were clustered into ≈3.9 million unique ortholog groups
(4.2 million total, Table S5). 40.5% of the contigs were assigned taxonomic identities (excluding
“NA from database” and “Multiple hits”) and 25% of the contigs were assigned KEGG
identities.

110
Taxonomic composition
Taxonomic assignments of transcripts revealed the top groups that contributed to metabolic
potential (Figure 4.2, Table S6). The total number of transcripts successfully assigned taxonomy
was highest from the surface community (62%), nearly double the percentage at 150 m or 890 m
(36 and 35%, respectively; Figure 4.2, Table S6). Major taxonomic groups that made up >2% of
transcript counts per million (CPM) across all depths were: dinoflagellates (17%), ciliates
(4.5%), haptophytes (3.7%), diatoms (2.6%) and chlorophytes (5.3%) (Figure 4.2, Table S6).
The surface was mainly comprised of dinoflagellate, haptophyte, and chlorophyte transcripts
(>9% each, Figure 4.2, Table S6). Below the euphotic zone (150 m and 890 m), the majority of
protistan transcripts were dinoflagellates (16-24%; Figure 4.2, Table S6). The relative abundance
of ciliate transcripts at 150 m (6.9%) was more than twice the percentage found at the surface or
890 m (3.1% and 2.5%, respectively; Figure 4.2, Table S6). Taxonomic groups that made up less
than 2% of the community were collapsed into the “Other eukaryote” category but are
summarized in Table S6.
The majority of rRNA reads (5.8% total sequence reads, Table S5) were found to be
protistan (28-50%, Figure S1A). Metagenomic Illumina tag (miTag, as defined by Logares et al.
(2013)) results at the class-level and with more than 500 count were reported in Figure S2 (also
see Table S7), and the distribution of miTags at each depth was shown in Figure S3. However,
we refrain from over interpreting the results, as sequence lengths shorter than 400 bps provide
less accurate assessments of protistan diversity (Hu et al. 2015).

111
Ortholog groups
More than 1 million ortholog groups were unique to each sampling depth (Figure 4.3). Of the
contigs making up these ortholog groups, 13-19% had KEGG identities (Figure 4.3B). Clustering
revealed a higher number of shared ortholog groups between 150 m and 890 m relative to other
comparisons (Figure 4.3B; >120,000 ortholog groups). More than half of the ortholog groups
shared among all depths were given functional assignments (Figure 4.3B, right-most bar:
62.9%), which mainly belonged to the KEGG Class 2 category, “Genetic information
processing” (Figure S4).

Metabolic potential of the whole community with respect to depth
Community-wide gene expression of targeted pathways at the surface was distinct from that
observed from the two depths below the euphotic zone (Figure 4.4). Photosynthesis, carbon
fixation (Calvin cycle, glycolysis, and gluconeogenesis), pyruvate dehydrogenase (PDH) and
fatty acid biosynthesis related transcripts were significantly more abundant at the surface relative
to deeper depths (Figure 4.4, upper portion of ternary plot; Table S8; p<0.05). Transcripts
associated with fatty acid breakdown, TCA and glyoxylate cycles, and vacuolar-type (V-type)
ATPases had similar abundances at all three depths (center of plot; Figure 4.4). Nitrogen uptake
(nitrate/nitrite transporter, NRT) and assimilatory nitrogen reduction transcripts were more
abundant at the surface and 890 m relative to 150 m (Figure 4.4, left portion of ternary plot).
Pyruvate-ferredoxin oxidoreductase (por), lysosome-associated genes, and chitinase genes were
significantly more abundant at 150 m and/or 890 m relative to the surface (Figure 4.4, lower
portion of ternary plot; Table S8; p<0.05).

112
Taxon-specific trends in metabolic potential
Ordination analysis of transcripts with KEGG annotations shared among taxonomic groups
revealed that samples clustered by taxonomic group and depth. Dinoflagellates and ciliates
clustered regardless of depth, while haptophytes, diatoms, and chlorophytes separated by depth
(Figure 4.5). Haptophytes, diatoms, and chlorophytes at the surface clustered away from all 150
m and 890 m samples (Figure 4.5). 150 m and 890 m samples from haptophytes, diatoms, and
chlorophytes more closely clustered with dinoflagellates (Figure 4.5).
In order to examine depth-specific shifts in expressed genes by taxon, transcript counts
were separated by taxonomic group and re-normalized in edgeR (see Experimental Procedures).
Functions with similar transcript abundances at each depth, indicative of unchanging
metabolism, were positioned centrally in each ternary plot (Figure 4.6). The majority of
dinoflagellate transcripts were similar in abundance at all depths (Figure 4.6A, center of plot),
with exceptions including nitrate/nitrite uptake (NRT), which was significantly higher among
dinoflagellates at the surface and 890 m relative to 150 m, and photosynthesis-related genes
expressed at the surface (Figure 4.6A; Table S9; p<0.05). Ciliate relative transcript abundance
was highest at 150 m (Figures 4.2A and 4.6B). Below the euphotic zone, ciliate transcripts that
were significantly more abundant relative to the surface included fatty acid metabolism,
chitinase, TCA and glyoxylate cycles, por, lysosome, and V-type ATPases (Figure 4.6B; Table
S9, p<0.05).
Distinct patterns of gene expression were observed between the surface and sub-euphotic
zone samples for haptophytes. Haptophyte species at the surface had significantly higher
abundances of photosynthesis, NRT, Calvin cycle, glycolysis, gluconeogenesis, and PDH
transcripts (Figure 4.6C; Table S9, p<0.05). Below the euphotic zone, dominant haptophyte
113
transcripts consisted of higher abundances of urea cycle (p<0.05), lysosome (p<0.05), chitinase
(p<0.05), glyoxylate cycle, and glutamate synthase/glutamine oxoglutarate aminotransferase
pathway (GS/GOGAT) associated transcripts (Figure 4.6C; Table S9).
Diatom and chlorophyte species at the surface had significantly more abundant
transcripts associated with photosynthesis, Calvin cycle, glycolysis, and gluconeogenesis
(Figures 4.6D-E; Table S9; p<0.05). Numbers of diatom and chlorophyte reads were low at 150
m and 890 m (Figure 4.2, and pie charts in Figures 4.6D-E), where metabolisms were
characterized by V-type ATPases, chitinase, lysosome, TCA cycle, and glyoxylate cycle related
genes (Figures 4.6D-E).

4.3 Discussion
A community-wide view of gene expression provided insight into the dominant biological
processes governing this coastal ecosystem, comparing the sunlit surface versus sub-euphotic
zone depths, and the taxonomic groups contributing to those processes. High relative abundances
of transcripts associated with core carbon fixation pathways and fatty acid synthesis indicated an
overwhelming contribution of phytoplankton at the surface, with strong contributions by
dinoflagellates, haptophytes, diatoms, and chlorophytes (Figures 4.2 and 4.4). Transcripts
associated with the breakdown of organic carbon compounds and intracellular nutrient cycling
dominated at 150 m and 890 m, particularly among dinoflagellates and ciliates. Lysosome-
associated enzymes are involved in many cellular processes, including the internal breakdown
and recycling of biomolecules (Eskelinen and Saftig 2009; Settembre et al. 2013), while V-type
ATPases play a role in digestive processes by lowering the pH in phagosomes (Finbow and
114
Harrison 1997). Distinct transcriptional profiles with respect to depth demonstrated the variable
roles protists contribute to elemental cycling throughout the water column at SPOT.

Depth-specific differences among taxonomic groups
The ordination analysis (Figure 4.5) and ternary plot (Figure 4.6) illustrated taxon-specific
transcriptional patterns indicative of differences in species composition within major taxonomic
groups, physiological acclimatization among the same species, or a result of evolutionary
adaptation within a lineage at each depth. In addition to verifying high reproducibility among
replicates, the ordination analysis divided some groups by presumably different nutritional
strategies exhibited at different depths in the water column. That is, primarily phototrophic taxa
(chlorophytes, haptophytes, diatoms) at the surface clustered away from samples collected at
sub-euphotic depths, and from phagotrophic-capable taxonomic groups (Figure 4.5, positions on
x-axis). This discrimination was similar to that observed in culture-based transcriptome studies,
where gene expression profiles of species clustered by nutritional mode (chrysophytes; Beisser et
al. 2017; MMETSP transcriptomes; Koid et al. 2014). The distribution of miTags with respect to
depth among most taxonomic groups provided some evidence that closely related species were
found throughout the water column (Figure S2), as there were many miTags shared among all
depths.
In contrast to well-known pathways associated with photosynthesis, there is little known
about the pathways that govern phagotrophy in protists. Yet, heterotrophic protists are important
consumers of microbial prey, and contribute significantly to food web function and nutrient
remineralization (Sherr and Sherr 1994; Sherr and Sherr 2002; Strom et al. 1997). Additionally,
phagotrophy among phytoplankton (mixotrophy) is a widespread and ecologically important
115
strategy for those species, and may increase the transport of carbon to higher trophic levels of
plankton communities (Mitra et al. 2016; Ward and Follows 2016; Worden et al. 2015b).
Dinoflagellate genes made up a significant portion of the transcripts observed in our
dataset (Figure 4.2), and much of their expression (Figure 4.6A) and class-level taxonomic
composition (miTag results, Figure S2A) was similar at each depth. In part this may reflect the
large transcript pools or post-transcriptional gene regulation found among dinoflagellates
(Hackett et al. 2004; Morey et al. 2011; Moustafa et al. 2010). However, those pathways that
were significantly differentially expressed indicated a primarily phototrophic assemblage of
dinoflagellates in the photic zone (photosynthesis-related transcripts in Figure 4.6A), while
pathways that may be associated with heterotrophic nutrition were observed at 150 m and 890 m
(fatty acid breakdown associated transcripts; Figure 4.6A). These results are consistent with a
wide range of nutritional modes exhibited by dinoflagellates (Hackett et al. 2004; Jeong et al.
2010).
Interestingly, dinoflagellate assemblages at different depths appeared to employ different
metabolic strategies with respect to nitrogen transport or transformation (NRT, AMT, and nitrate
reduction). Nitrogen-related transcripts and pathways were significantly increased at the surface
and 890 m at SPOT, where the availability of inorganic nitrogen was lowest and highest,
respectively (Figure 4.6A, Table S4). Phototrophic dinoflagellate species at the surface were
likely scavenging inorganic nitrogen (upregulation of NRT in Figure 4.6A), as was observed in
Karenia brevis under N-starved conditions (Morey et al. 2011), whereas, heterotrophic
dinoflagellate species found at 890 m were unlikely to be actively taking up inorganic nitrogen.
We speculate that high abundances of nitrogen metabolism transcripts at 890 m reflected
repartitioning of intracellular nitrogen (Dagenais-Bellefeuille and Morse 2013) or the release of
116
nitrogen acquired from ingested prey. This is notable because it implies flexibility in the nitrogen
metabolism strategies in sub-euphotic zone dinoflagellates. However, additional work to
characterize these taxa are required as dinoflagellate transcripts detected at sub-euphotic zone
depths may have included species (or life stages) for which we do not have appropriate reference
databases (Figure S2A, “Uncertain” and “uncultured”), including parasitic species (e.g.
syndiniales; Guillou et al. 2008) or encysted life stages (Anderson et al. 1987; McMinn and
Martin 2013).
Sources of labile organic matter that support higher protistan phagotrophy at oxyclines
(and other transition zones), might partially explain high ciliate diversity and activity observed at
subsurface depths in this and previous studies (Pachiadaki et al. 2014a; Stock et al. 2009). Ciliate
gene expression provided evidence of elevated activity at the oxycline and also illustrated that
core ciliate metabolisms did not vary greatly with depth (clustering of red symbols in Figure
4.5). Ciliate reads were highest at 150 m (Figure 4.2), in agreement with previous work at the
SPOT station that indicated higher relative activity at the oxycline (Hu et al. 2016). Transcripts
associated with fatty acid breakdown were significantly increased at 150 m relative to other
depths (Figure 4.6B, Table S9, p<0.05). Class-level taxonomic distribution based on miTag
results showed little variation by depth; ciliates below the euphotic zone were mainly dominated
by Choreotrichia and Oligohymenophora (Figure S2B), which include species previously found
at oxic/anoxic boundaries and anoxic environments (Behnke et al. 2006; Edgcomb et al. 2011a;
Fenchel and Finlay 1987; Fenchel and Finlay 1995; Orsi et al. 2012b; Wylezich and Jürgens
2011). Further, compared to other taxonomic groups, there was a higher number of unique ciliate
miTags detected at 150 m and 890 m (Figure S3C); the presence of surprisingly diverse and
active populations of ciliates below the euphotic zone warrants further study.
117
Ciliates in low oxygen environments may adopt anaerobic metabolisms using a modified
mitochondrial organelle (hydrogenosome), in which pyruvate (from glycolysis) is fermented into
acetyl coenzyme-A (acetyl-CoA), acetate, and hydrogen to generate energy (Fenchel and Finlay
1995; Müller 1988). Pyruvate:ferredoxin oxidoreductase (por) transcripts, responsible for the
initial breakdown of pyruvate into acetyl-CoA and CO
2
(Fenchel and Finlay 1995), were higher
at 150 m and 890 m among ciliates relative to the surface in the present study (Figure 4.6B and
Table S9; p<0.05 at 890 m), indicating that some species may have adopted an anaerobic
metabolism at those depths. In agreement with the frequency of microaerophilic and anaerobic
ciliates in the nearby oxygen-deplete Santa Barbara Basin (Beinart et al. 2017; Bernhard et al.
2000) and other low-oxygen environments (Orsi et al. 2012a), some ciliates at 150 m and 890 m
may possess ectosymbiotic (Fenchel et al. 1977) or endosymbiotic bacteria and archaea to
support an anaerobic metabolism (Edgcomb et al. 2011b; Nowack and Grossman 2012).
Specifically, methanogens or sulfate reducing or oxidizing bacteria associated with ciliates may
serve to deplete the end products from the fermentation of pyruvate; however molecular
characterizations of the interactions between protistan hosts and symbionts at anoxic or low-
oxygen depths in the water column remain largely uncharacterized (Edgcomb 2016; Stoeck et al.
2014).
Haptophytes at the surface had higher transcript abundances associated with nitrogen
uptake, consistent with a nitrogen-limited environment (Figure 4.6C, Tables S4 and S9, p<0.05).
Prymnesium parvum, Isochrysis galbana, and Emiliania huxleyi, have been shown to
significantly upregulate nitrate transporters under N-deplete conditions (Dyhrman et al. 2006;
Kang et al. 2007; Liu et al. 2015b). Conversely, decreased expression of nitrogen uptake genes
by P. parvum was observed when prey were abundant (Liu et al. 2015a), and a natural
118
population of haptophytes exhibited decreased expression of nitrogen transporters when nitrogen
deficiency was alleviated by supplementation with nutrient-rich water (Alexander et al. 2015b).
Combined, these results are consistent with the use of NRT as a putative biomarker for N-
stressed haptophyte populations, where decreased abundances of NRT may reflect an alternate N
acquisition strategy among haptophytes, such as phagotrophy.
Haptophyte transcriptional patterns were similar to other phototrophic groups at the
surface, but clustered with dinoflagellates in the ordination analysis for samples collected below
the euphotic zone (Figure 4.5), perhaps reflecting the mixotrophic capabilities of many of these
species (Cuvelier et al. 2010; Liu et al. 2009; Unrein et al. 2013). Haptophytes from deep
samples in this study had high relative transcript abundances in pathways related to fatty acid
breakdown and the glyoxylate cycle relative to the surface, similar to the response of P. parvum
in the presence of prey (Figure 6C; Liu et al. 2015a). Taken together, these data provide evidence
of the cellular breakdown of fatty acids (likely from consumed prey) to produce acetyl-CoA,
which enters the glyoxylate cycle. Transcripts related to the urea cycle also increased in
abundance among haptophytes at 150 m and 890 m, suggesting that the urea cycle may play an
important role in how haptophytes process nitrogen below the euphotic zone, where phagotrophy
is potentially occurring. While there is molecular evidence for urea metabolism in haptophytes
(culture-based studies), integrating in situ biogeochemical measurements of urea uptake and
utilization are required to fully understand intracellular regulation (or re-distribution) of different
nitrogen pools among haptophytes, and their role in the urea cycle (Solomon et al. 2010;
Solomon and Glibert 2008). Our results imply that phylogenetically related haptophytes (mainly
Prymnesiales, Figure S2C and S3D) occupy different depths in the water column. Additionally,
119
while the data suggest a predominantly phagotrophic metabolism, sinking haptophytes may
reallocate intracellular carbon compounds as a stress response to sub-euphotic zone conditions.

Diatoms and chlorophytes
High abundances of diatoms and chlorophytes in the euphotic zone (Figure 4.2) and high
expression of phototrophy and carbon fixation-related genes at the surface (Figures 4.4, 4.6D-E,
Table S9; p<0.05) was anticipated, and our study confirmed a primarily phototrophic lifestyle for
these taxa in the euphotic zone. Diatoms and picoeukaryotic prasinophytes such as Ostreococcus
are common inhabitants of coastal waters of the eastern Pacific Ocean where they contribute
significantly to primary production (Aylward et al. 2015; Countway and Caron 2006; Worden
2006; Worden et al. 2004). MiTag results confirmed the presence of important coastal diatoms
such as Pseudo-nitzschia, and the chlorophytes Micromonas and Ostreococcus (Figure S2D-E).
Their presence in dark ocean environments is typically overlooked, and generally attributed to
deep vertical mixing, active sinking of cells, or association with sinking particles (Agusti et al.
2015; Jochem 1999). Although found at very low abundances, the presence of transcripts and
unique miTags (Figure S3E-F) from these species below the euphotic zone indicated that some
diatoms and chlorophytes may have the capability to alter their physiology to survive at 150 m
and 890 m, at least for some limited time.
Phytoplankton metabolism below the euphotic zone is not well understood, but several
studies have examined the ability of diatoms to survive periods of prolonged darkness (e.g. up to
10 months; Nymark et al. 2013; Peters and Thomas 1996; Sickogoad et al. 1989; Smayda and
Mitchell-Innes 1974). Subsurface diatoms and chlorophytes transcribed photosynthesis-related
genes, but at lower levels than at the surface (Figure 4.6D, photosynthesis transcripts; Figure
120
4.6E, photosynthesis, Calvin cycle, glycolysis-associated transcripts); similar to Phaeodactylum
tricornutum exposed to darkness for 48 hours (Nymark et al. 2013). Continued transcription of
photosynthesis-related genes may be a mechanism for these phytoplankton to retain the ability to
perform photosynthesis immediately when light becomes available (Nymark et al. 2013; Peters
and Thomas 1996). This conjecture is supported by evidence that even though transcription of
chloroplast-encoded genes may continue in phytoplankton, the transcripts may not be translated
into protein without sunlight, thus maintaining responsiveness while minimizing energy
expenditure (Lee and Herrin 2002).
Diatoms and chlorophytes below the euphotic zone demonstrated a possible reliance on
intracellular lipids for energy acquisition (high abundances of fatty acid, glyoxyclate cycle, V-
type ATPases, and lysosome transcripts, Figures 4.6D-E). Previous studies have reported
increased lipid accumulation in phytoplankton under sub-optimal conditions, including polar
darkness (McKnight et al. 2000; McMinn and Martin 2013), silicon-deficiency (Roessler 1988),
low temperature and N-limitation (Mock and Kroon 2002), or in anticipation of a dark cycle
(Poliner et al. 2015), consistent with the transcriptional patterns reported here.
The abundance of transcripts associated with chitinase production was higher below the
euphotic zone in both diatoms and chlorophytes (Figure 4.6D and E). Increased chitinase in
diatoms has been linked to altering cellular structure in order to reduce sinking (by increasing
drag), fungal infection, or entering a resting state (Armbrust et al. 2004; Mulisch 1993; Round et
al. 1990). Additionally, transcriptional signatures of diatoms and chlorophytes at 150 m and 890
m may be indicative of resting cysts, as many taxa are known to enter a dormant cell cycle (cyst)
in response to unfavorable environmental conditions (McMinn and Martin 2013). However, the
121
transcriptional signatures associated with encysted cells or resting spores among diatom and
chlorophyte species in dark ocean environments are not well characterized.

Improving future metatranscriptome analyses
Community composition at each depth in this study, a determined from the metatranscriptomes,
was in agreement with previous tag-sequencing surveys of protistan diversity at the SPOT
station, confirming and expanding our current knowledge of in situ protistan ecological roles at
this coastal site. However, a high percentage of transcripts and miTags still could not be
annotated or sufficiently assigned taxonomy, especially for samples collected at sub-euphotic
depths (Figure 4.2). While there was some evidence of taxa such as excavates, rhizaria, and the
stramenopile group known as MArine STramenopiles (MAST) in the results, transcripts from
these taxa were not well-represented in our dataset (compare Figures 4.2 and S1, Tables S6 and
S7), presumably due to the lack of reference data for those species. One explanation for this is
the use of a prefilter (see Experimental Procedures), which would have excluded rhizaria larger
than 80 µm; however, using the same collection procedures, an earlier study found rhizaria to be
a potentially significant grazer below the euphotic zone at the SPOT station (Hu et al. 2016).
Further, rhizaria are thought to be an important group throughout the world’s oceans (Biard et al.
2016; Burki and Keeling 2014; Caron 2016; de Vargas et al. 2015; Guidi et al. 2016). We
therefore anticipated a significant contribution of these species to our metatranscriptome survey,
but <2% of the community metatranscriptome was identified as cercozoan, radiolarian, or
foraminiferan (Table S6). These findings indicate that available rhizarian reference
transcriptomes (28/844 in our custom database, Table S2) were phylogenetically distant from
those found in the water column at the SPOT station (del Campo et al. 2014; Keeling et al.
122
2014). Increasing the number of available reference transcriptomes and genomes for poorly
represented groups (e.g. rhizaria or MAST) and increasing sequence read lengths should
dramatically improve the amount of information that can be derived from metatranscriptomic
studies of microbial eukaryotes. Efforts to enhance genetic databases for uncultured eukaryotes
are well underway (Carradec et al. 2018; Mangot et al. 2017; Seeleuthner et al. 2018)
Further, we chose to keep taxonomic assignments uniform for all protistan lineages (e.g.
same taxonomic level for all major groups); therefore we refrained from describing the protistan
community to the class or genus level based on transcript or miTag results. Some protistan
lineages are currently underrepresented in transcriptome databases (see previous section),
therefore the alignment of short reads (125 bps in this study) to assembled contigs does not offer
the phylogenetic resolution required to accurately assign all protistan lineages to the species
level. This attests to the need for improved protistan taxonomic assignment criteria and strategies
for metatranscriptome data analyses. We also acknowledge that sample handling procedures may
have an impact on gene expression (Pachiadaki et al. 2014a). To combat these two factors, we
included a high number of replicates, carried out quick and careful sampling procedures, and
focused our study on abundant taxa (>2%, see Experiment Procedures).

Summary
This study contributes a rich amount of gene expression data underpinning the ecological
activities of marine protists throughout the water column at a coastal study site.
Metatranscriptomic approaches are a way forward to uncover the metabolic activities of
ecologically significant microorganisms which we otherwise have no other way of accessing.
Recent global surveys of marine diversity have revealed the presence and ecological impact of
protists to be more significant than previously thought (de Vargas et al. 2015), highlighting a
123
pressing need to develop tools and techniques to investigate in situ trophic strategies in diverse
microbial communities. Identifying the mechanistic drivers of these trophic modes will help to
better parameterize food web models in the coastal zone and other planktonic ecosystems.

4.4 Experimental Procedures
Sample collection & handling
Seawater was collected from the San Pedro Ocean Time-series (SPOT) station off the coast of
Southern California near the surface (5 m), 150 m and 890 m, in late May 2015 (Figure 4.1A;
Table S1). Briefly, seawater was pre-filtered (80 µm) into 20 L carboys to minimize the presence
of multicellular eukaryotes. Replicate samples (ranging in volume from 1.5-3.5 L, Table S1)
from each depth were filtered onto sterile GF/F filters (nominal pore size 0.7 µm, Whatman,
International Ltd. Florham Park, NJ, USA). While we cannot avoid some impact that sample
handling (i.e. bringing samples to the surface) may have had on our results, filters were
immediately placed in 1.5 mL of lysis buffer and flash frozen in liquid nitrogen in <40 minutes
and away from light to minimize RNA degradation (see Table S1). Detailed protocols for field
collection and preservation of samples can be found at protocols.io
(dx.doi.org/10.17504/protocols.io.hisb4ee) and in Supporting Information. Temperature,
chlorophyll a fluorescence, oxygen, and salinity at the SPOT station were obtained during each
CTD cast (Sea-bird Electronics, Inc., Bellevue, WA, USA). Samples for inorganic nutrients and
cell counts were also taken (Supporting Information).

Molecular sample processing and quality control
Total RNA was extracted from each filter using a DNA/RNA AllPrep kit (Qiagen, Valencia, CA,
USA, #80204) with an in-line genomic DNA removal step (RNase-free DNase reagents, Qiagen
124
#79254) (dx.doi.org/10.17504/protocols.io.hk3b4yn). Extracted RNA was quality checked, low
biomass samples were pooled. Six replicates were processed and sequenced from the surface,
while pairs of filters were pooled for either 150 m or 890 m, yielding triplicates for each of the
latter two depths (Table S1). RNA concentrations were normalized before library preparation
(Supporting Information). ERCC spike-in was added before sequence library preparation with
Kapa’s Stranded mRNA library preparation kit using poly-A tail selection beads to select for
eukaryotic mRNA (Kapa Biosystems, Inc., Wilmington, MA, USA, #KK8420). HiSeq High
Output 125 bp PE sequencing was performed at UPC Genome Core at University of Southern
California, Los Angeles, CA (BioProject PRJNA391503).
Sequence adapters, low quality (phred score <10, from 5’ and 3’ ends, and within a 25 bp
sliding window) or short sequences (<50 bps), and sequences containing more than 50
consecutive As or Ts were removed using Trimmomatic v. 0.32 (Bolger et al. 2014). All quality
trimmed sequences were aligned to ERCC sequences using “align_and_estimate_abundance.pl”
in the Trinity v. 2.1.1 (Grabherr et al. 2011) package. Reads that were aligned to ERCC
sequences were removed using a custom PERL script (available:
https://github.com/shu251/SPOT_metatranscriptome).

miTag analysis
rRNA and mRNA reads were separated using SortMeRNA v. 2.0 (Kopylova et al. 2012).
Forward rRNA reads were further quality checked (Q>16) in QIIME (Caporaso et al. 2010). and
taxonomic identities were assigned to each read using uclust (Edgar 2010); each read had to have
at least 97% similarity to the SILVA database v.128 (Quast et al. 2012). Metagenomic Illumina
125
tags, or miTags, are defined as rRNA reads as an alternative to amplicon sequences for diversity
analyses (Logares et al. 2013).

Assembly and protein prediction
mRNA sequences were assembled using MEGAHIT v. 1.0.3 (Li et al. 2015) with default
parameters. Redundancy among assembled contigs was reduced using CD-HIT at 98% similarity
(Li et al. 2001). Estimated transcript abundances based on paired end reads were determined
using Salmon v. 0.8.2 with quasi mapping strategy (kmer size=31; Patro et al. 2017). Coding
sequences for each contig were predicted using GeneMarkS-T v. 5.1, which specifically
predicting coding regions in eukaryotic transcripts (Tang et al. 2015). GeneMarkS-T should
predict one coding sequence per contig, but in rare cases (<0.001%) more than one coding
sequence was predicted. In those cases, the longest coding sequence was used. Predicted protein
sequences (>300 bps) were clustered to produce orthologous groups with 75% identity cutoff
using uclust v. 1.2.22 (Edgar 2010). Contigs without predicted coding sequences were not
analyzed.

Annotation
Contigs were assigned taxonomic identities using BLASTX (e-value cutoff: 1e-5) and then
MegaBLAST (e-value cutoff: 1e-5) to customized protein and cDNA reference databases,
respectively, which included data from the Marine Microbial Eukaryote Transcriptome
Sequencing Project and other publically available protistan genomes and transcriptomes (Table
S2 and Supporting Information). We elected to use nucleotide sequences for taxonomic
assignments in order to evade missing taxonomy information from shorter or missing predicted
126
protein sequences. First, contigs were aligned to the custom protein database using a faster
BLASTX search provided by DIAMOND (Buchfink et al. 2014) with the sensitive alignment
mode, an e-value cutoff of 1e-5, and an identity cutoff of 40%. The best hit and all hits with bit
scores within 90% of the best score were considered. A taxonomic identity was assigned to a
contig if all extracted hits were from the same taxonomic group (Table S2). If no consensus
could be reached among all top hits from the BLASTX results, the contig was assigned to
“Multiple hits”. Second, to increase the number of contigs with taxonomic assignments, any
contigs that fell into the “Multiple hits” category or did not hit any reference in the protein
database (“NA from database”) were searched against the custom cDNA database using
MegaBLAST (Camacho et al. 2009) with the same cutoffs. We found that taxonomic assignment
using this approach was reasonably accurate (>90%) at the phylum or class level, but unreliable
at genus or species level for every protistan lineage. While assignment to the class or genus level
may be accurate among well-represented taxa in the database, we chose to assign taxonomy
uniformly across all lineages, thus taxonomic assignment was restricted to phylum/class levels
(See Table S2 for taxonomic grouping).
KEGG annotation of putative protein sequences (GeneMark S-T) was performed using
GhostKOALA (Kanehisa et al. 2016). Protein annotations for each contig included KEGG
module information when available. Pathways and genes were chosen to represent the core
biogeochemical functions relevant to known metabolisms (Table S3). Genes that appeared in
multiple KEGG modules, and subsequent pathways, were counted multiple times after
normalization steps (Table S3).

127
Data synthesis and statistical analyses
Transcript counts were normalized across libraries (replicates) using the trimmed mean
normalization method in ‘edgeR’ (Lund et al. 2012) to generate transcript counts per million
(CPM) for all downstream analyses. To evaluate differences at the pathway level, analysis of
variance statistical tests (ANOVA) were performed on normalized data using replicates from
each depth. ANOVAs were followed by Tukey’s Honestly Significantly Different test to obtain
p-values for pairwise comparisons among depths. Library normalization with ‘edgeR’ and
ANOVA statistical tests were performed twice, once with the whole community (all transcripts)
and a second time for each individual taxonomic group (i.e. dinoflagellates, ciliates, haptophytes,
diatoms, and chlorophytes). The mean CPM of the replicates for a given depth was used to
visualize depth-specific differences in gene expression, unless specified in the figure legend.
Transcripts with KEGG identities in all five taxonomic groups were used for ordination analysis.
All figures were generated in R using the vegan, UpSetR, ggplot, and ggtern packages (Conway
et al. 2017; Hamilton 2016; R Core Team 2014; Wickham 2009).

Access to data
All sequences are publicly available with accession numbers SAMN07269826-SAMN07269838,
in the Short Read Archive (Table S1). Processed data is available at zenodo (DOI:
10.5281/zenodo.846379) and all source code is available
https://github.com/shu251/SPOT_metatranscriptome.

128

4.5 Chapter Four Figures and Tables

Figure 4.1. (A) Map of San Pedro Ocean Time-series (SPOT) station, located ~20 km from Los
Angeles in the San Pedro Channel (33° 33’ N, 118° 24’ W). Samples were collected from the
surface (5 m), 150 m, and 890 m in late May. (B) Vertical profile based on CTD cast data,
including temperature (°C), chlorophyll a fluorescence (ug/L), oxygen (ml/L), and salinity (ppt).
Dotted horizontal lines in mark sampling depths.

km
Pacific Ocean
California
Los Angeles
San Pedro Channel
SPOT
10 20 0
0
250
500
750
0.00 0.25 0.50 0.75 1.00 1.25
Chlorophyll a (ug/L)
33 34 35 36
Salinity (ppt)
33 34 35 36
10 20 30
0246
Oxygen (ml/L)
Temperature (C)
A
B
129

Figure 4.2. Average (across replicates) relative abundance of transcript counts per million
(CPM). Contigs were assigned a taxonomic identity against a custom database (See
Experimental Procedures). Contigs that hit more than one reference with no consensus among
the references were labeled “Multiple hits”. When no reference was hit, the contig was labeled
“NA from database”. There were more total hits (excluding NA from database and Multiple hits)
to the reference database in the surface sample (62%), relative to 150 m (36%) and 890 m (35%).
“Other eukaryotes” are comprised of taxonomic groups that made up less than 2% of the total
composition of contigs; a full summary of “Other eukaryotes” can be found in Table S6.

130

Figure 4.3. Distribution of ortholog groups among depths sampled. (A) There were a total of 3.9
million ortholog groups found at the surface, 150 m, and 890 m (top to bottom). (B) Bar plots
show the total number of shared or unique ortholog groups across the three depths. Depths
included in a set are represented by filled dots in the matrix below each bar. Most ortholog
clusters were unique to a single depth (three left-most bars), although 36,656 orthologous groups
were common across depths (right most bar). The number of ortholog groups that contain at least
one contig with an annotated KEGG identity are represented by light grey (also given as
percentages). Plot generated using ‘UpSetR’ R package Conway et al. 2017.

131

Figure 4.4. Ternary plot of relative transcript abundance in counts per million (CPM) with
respect to depth for key metabolic pathways (colors). Circle placement is representative of
relative transcript abundance among the three depths (one depth per axis, clockwise with
increasing abundance) and circle size is proportional to the total CPM for the three depths. A full
list of genes and pathways is in Table S3. Related statistical analyses presented in Table S7. Plot
generated using ggtern R package, Hamilton 2016. Abbreviations: AMT, ammonium transporter
(K03320); NRT, nitrate/nitrite transporter (K02575); GS/GOGAT, glutamate synthase/glutamine
oxoglutarate aminotransferase pathway; PDH, pyruvate dehydrogenase subunits alpha (K00161)
and beta (K00162); por, pyruvate-ferredoxin/flavodoxin oxidoreductase (K03737).

Total CPM Pathway/Gene
132

Figure 4.5. Ordination analysis (constrained correspondence analysis – CCA) plot shows the
placement of major taxonomic groups at each depth based on normalized transcript counts per
million (CPM). Contigs included in this analysis had common KEGG identities among all five
taxonomic groups (n = 1 603). Transcript CPM was normalized by sample to the total number of
transcripts belonging to each taxonomic group (colors). Shapes depict depth at the SPOT station:
squares - surface, circles – 150m, and triangles – 890m. Replicate samples are shown as the
same symbol and color; there were 6 replicates from the surface, 3 from 150 m, and 4 from 890
m (Table S1). The x- and y-axes demonstrate the percent variance explained by the data.

133

Figure 4.6. Ternary plots (similar to Figure 4.4) of the relative transcript abundance in counts
per million (CPM) with respect to depth for key metabolic pathways (colors, listed in Table S3)
found in each taxonomic group: (A) dinoflagellates, (B) ciliates, (C) haptophytes, (D) diatoms,
and (E) chlorophytes. Each side of the triangle corresponds to a depth, relative CPM was
normalized to taxonomic group and depth, white and black pie charts represent the relative
abundance of taxa at each depth (Figure 4.1). Circle placement is representative of transcript
abundance relative to the three depths (one depth per axis, clockwise with increasing abundance)
and circle size is proportional to the total CPM. A full list of genes and pathways is in Table S3.
Related statistical analyses presented in Table S8. Abbreviations: AMT, ammonium transporter
(K03320); NRT, nitrate/nitrite transporter (K02575); GS/GOGAT, glutamate synthase/glutamine
oxoglutarate aminotransferase pathway; PDH, pyruvate dehydrogenase subunits alpha (K00161)
and beta (K00162); por, pyruvate-ferredoxin/flavodoxin oxidoreductase (K03737).
Pathway/Gene
Total CPM
134
4.6 Chapter Four Literature Cited
Adl, S. M. and others 2012. The revised classification of eukaryotes. J. Eukaryot. Microbiol. 59:
429-493.
Agusti, S. and others 2015. Ubiquitous healthy diatoms in the deep sea confirm deep carbon
injection by the biological pump. Nat. Commun. 6: 7608.
Alexander, H., B. D. Jenkins, T. A. Rynearson, and S. T. Dyhrman. 2015a. Metatranscriptome
analyses indicate resource partitioning between diatoms in the field. P. Natl. Acad. Sci.
USA 112: E2182-E2190.
Alexander, H., M. Rouco, S. T. Haley, S. T. Wilson, D. M. Karl, and S. T. Dyhrman. 2015b.
Functional group-specific traits drive phytoplankton dynamics in the oligotrophic ocean.
P. Natl. Acad. Sci. USA 112: E5972–E5979.
Allen, L. Z. and others 2012. Influence of nutrients and currents on the genomic composition of
microbes across an upwelling mosaic. ISME J. 6: 1403-1414.
Anderson, D. M., C. D. Taylor, and E. V. Armbrust. 1987. The effects of darkness and
anaerobiosis on dinoflagellate cyst germination. Limnol. Oceanogr. 32: 340-351.
Armbrust, E. V. and others 2004. The genome of the diatom Thalassiosira pseudonana: Ecology,
evolution, and metabolism. Science 306: 79-86.
Aylward, F. O., J. M. Eppley, J. M. Smith, F. P. Chavez, C. A. Scholin, and E. F. DeLong. 2015.
Microbial community transcriptional networks are conserved in three domains at ocean
basin scales. P. Natl. Acad. Sci. USA 112: 5443-5448.
Balzano, S. and others 2015. Transcriptome analyses to investigate symbiotic relationships
between marine protists. Front Microbiol 6: 1-14.
Behnke, A., J. Bunge, K. Barger, H.-W. Breiner, V. Alla, and T. Stoeck. 2006. Microeukaryote
Community Patterns along an O2/H2S Gradient in a Supersulfidic Anoxic Fjord
(Framvaren, Norway) Appl. Environ. Microb. 72: 3626-3636.
Beinart, R. A., D. J. Beaudoin, J. M. Bernhard, and V. P. Edgcomb. 2018. Insights into the
metabolic functioning of a multipartner ciliate symbiosis from oxygen-depleted
sediments. Mol. Ecol. 00: 1-14.
Beisser, D. and others 2017. Comprehensive transcriptome analysis provides new insights into
nutritional strategies and phylogenetic relationships of chrysophytes. PeerJ 5: e2832.
Bender, S. J., C. A. Durkin, C. T. Berthiaume, and R. L. Morales. 2014. Transcriptional
responses of three model diatoms to nitrate limitation of growth. Front. Mar. Sci. 1.
Bernhard, J. M., K. R. Buck, M. A. Farmer, and S. S. Bowser. 2000. The Santa Barbara Basin is
a symbiosis oasis. Nature 403: 77-80.
Beszteri, S. and others 2012. Transcriptomic response of the toxic prymnesiophyte Prymnesium
parvum (N. Carter) to phosphorus and nitrogen starvation. Harmful Algae 18: 1-15.
Biard, T. and others 2016. In situ imaging reveals the biomass of giant protists in the global
ocean. Nature 532: 1-16.
Bolger, A. M., M. Lohse, and B. Usadel. 2014. Trimmomatic: a flexible trimmer for Illumina
sequence data. Bioinformatics 30: 2114-2120.
Buchfink, B., C. Xie, and D. H. Huson. 2014. Fast and sensitive protein alignment using
DIAMOND. Nat. Methods 12: 59-60.
Burki, F., and P. J. Keeling. 2014. Rhizaria. Curr Biol 24: R103-R107.
Camacho, C. and others 2009. BLAST+: architecture and applications. BMC Bioinformatics 10:
421-429.
135
Caporaso, J. G. and others 2010. QIIME allows analysis of high-throughput community
sequencing data. Nat. Methods 7: 335-336.
Caron, D. A. 2016. Ocean science: The rise of Rhizaria. Nature 532: 444-445.
Caron, D. A. and others 2017. Planktonic food web structure at a coastal time-series site: I.
Partitioning of microbial abundances and carbon biomass. Deep-Sea Res PT I 121: 14-
29.
Caron, D. A., P. D. Countway, A. C. Jones, D. Y. Kim, and A. Schnetzer. 2012. Marine protistan
diversity. Annu. Rev. Mar. Sci. 4: 467-493.
Carradec, Q. and others 2018. A global ocean atlas of eukaryotic genes. Nat. Commun.: 1-13.
Conway, J. R., A. Lex, and N. Gehlenborg. 2017. UpSetR: An R Package for the visualization of
intersecting sets and their properties. Bioinformatics: 1-3. DOI:
10.1093/bioinformatics/btx1364.
Countway, P. D., and D. A. Caron. 2006. Abundance and distribution of Ostreococcus sp. in the
San Pedro Channel, California, as revealed by quantitative PCR. Appl. Environ. Microb.
72: 2496-2506.
Cuvelier, M. L., A. E. Allen, A. Monier, C. L. Dupont, and A. Z. Worden. 2010. Targeted
metagenomics and ecology of globally important uncultured eukaryotic phytoplankton. P.
Natl. Acad. Sci. USA 107: 14679–14684.
Dagenais-Bellefeuille, S., and D. Morse. 2013. Putting the N in dinoflagellates. Front Microbiol
4.
de Vargas, C. and others 2015. Eukaryotic plankton diversity in the sunlit ocean. Science 348:
DOI: 10.1126/science.1261605.
del Campo, J., M. E. Sieracki, R. Molestina, P. Keeling, R. Massana, and I. Ruiz-Trillo. 2014.
The others: our biased perspective of eukaryotic genomes. Trends Ecol. Evol.: 1-8.
Dupont, C. L. and others 2014. Genomes and gene expression across light and productivity
gradients in eastern subtropical Pacific microbial communities. ISME J.: 1-17.
Dyhrman, S. T., S. T. Haley, S. R. Birkeland, L. L. Wurch, M. J. Cipriano, and A. G. McArthur.
2006. Long Serial Analysis of Gene Expression for Gene Discovery and Transcriptome
Profiling in the Widespread Marine Coccolithophore Emiliania huxleyi. Appl. Environ.
Microb. 72: 252-260.
Dyhrman, S. T. and others 2012. The Transcriptome and Proteome of the Diatom Thalassiosira
pseudonana Reveal a Diverse Phosphorus Stress Response. PloS One 7: e33768-33710.
Edgar, R. C. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics
26: 2460-2461.
Edgcomb, V. and others 2011a. Protistan microbial observatory in the Cariaco Basin, Caribbean.
I. Pyrosequencing vs Sanger insights into species richness. ISME J. 5: 1344–1356.
Edgcomb, V. P. 2016. Marine protist associations and environmental impacts across trophic
levels in the twilight zone and below. Curr. Opin. Microbiol. 31: 169-175.
Edgcomb, V. P., E. R. Leadbetter, W. Bourland, D. Beaudoin, and J. M. Bernhard. 2011b.
Structured multiple endosymbiosis of bacteria and archaea in a ciliate from marine
sulfidic sediments: a survival mechanism in low oxygen, sulfidic sediments? Front
Microbiol 2.
Edgcomb, V. P., M. G. Pachiadaki, P. Mara, K. A. Kormas, E. R. Leadbetter, and J. M.
Bernhard. 2016. Gene expression profiling of microbial activities and interactions in
sediments under haloclines of E. Mediterranean deep hypersaline anoxic basins. ISME J.:
1-15.
136
Eskelinen, E.-L., and P. Saftig. 2009. Autophagy: A lysosomal degradation pathway with a
central role in health and disease. BBA-Mol Cell Res 1793: 664-673.
Fenchel, T., and B. J. Finlay. 1987. The biology of free-living anaerobic ciliates. Eur. J. Phycol.
26: 201-215.
Fenchel, T., and B. J. Finlay. 1995. Ecology and Evolution in Anoxic Worlds. Oxford University
Press Inc.
Fenchel, T., and T. Perry. 1977. Anaerobiosis and Symbiosis with Bacteria in Free-living
Ciliates. J. Protozool. 24: 154-163.
Finbow, M. E., and M. A. Harrison. 1997. The vacuolar H+-ATPase: a universal proton pump of
eukaryotes. Biochem J 324: 697-712.
Frischkorn, K. R., M. J. Harke, C. J. Gobler, and S. T. Dyhrman. 2014. De novo assembly of
Aureococcus anophagefferens transcriptomes reveals diverse responses to the low
nutrient and low light conditions present during blooms. Front Microbiol 5.
Grabherr, M. G. and others 2011. Full-length transcriptome assembly from RNA-Seq data
without a reference genome. Nat. Biotechnol. 29: 644-652.
Grossmann, L. and others 2016. Trade-off between taxon diversity and functional diversity in
European lake ecosystems. Mol. Ecol.: 1-27.
Guidi, L., S. Chaffron, L. Bittner, D. Eveillard, and A. Larhlimi. 2016. Plankton networks
driving carbon export in the oligotrophic ocean. Nature 532: 465-470.
Guillou, L. and others 2008. Widespread occurrence and genetic diversity of marine parasitoids
belonging to Syndiniales (Alveolata). Environ. Microbiol. 10: 3349-3365.
Hackett, J. D., D. M. Anderson, and D. L. Erdner. 2004. Dinoflagellates: a remarkable
evolutionary experiment. Am J Bot 91: 1523-1534.
Hamilton, N. 2016. ggtern: An Extension to 'ggplot2', for the Creation of Ternary Diagrams. R
package version 2.1.5. https://CRAN.R-project.org/package=ggtern.
Harke, M. J., A. R. Juhl, S. T. Haley, H. Alexander, and S. T. Dyhrman. 2017. Conserved
Transcriptional Responses to Nutrient Stress in Bloom-Forming Algae. Front Microbiol
8: E2182-2118.
Hess, W. R., L. Garczarek, U. Pfreundt, and F. Partensky. 2016. Phototrophic Microorganisms:
The Basis of the Marine Food Web, p. 57-97. Springer International Publishing.
Hu, S. K. and others 2016. Protistan diversity and activity inferred from RNA and DNA at a
coastal ocean site in the eastern North Pacific. FEMS Microbiol. Ecol. 92: 1-13.
Hu, S. K. and others 2015. Estimating Protistan Diversity Using High-Throughput Sequencing. J.
Eukaryot. Microbiol. 62: 688-693.
Jeong, H.-J., Y.-D. Yoo, J. S. Kim, K. A. Seong, N. S. Kang, and T. H. Kim. 2010. Growth,
feeding and ecological roles of the mixotrophic and heterotrophic dinoflagellates in
marine planktonic food webs. Ocean Sci J 45: 65-91.
Jochem, F. J. 1999. Dark survival strategies in marine phytoplankton assessed by cytometric
measurement of metabolic activity with fluorescein diacetate. Mar. Biol. 135: 721-728.
Johnson, M. D. 2015. Inducible Mixotrophy in the Dinoflagellate Prorocentrum minimum. J.
Eukaryot. Microbiol.: 1-13.
Johnson, M. D., D. Oldach, C. F. Delwiche, and D. K. Stoecker. 2007. Retention of
transcriptionally active cryptophyte nuclei by the ciliate Myrionecta rubra. Nature 445:
426-428.
137
Kanehisa, M., Y. Sato, and K. Morishima. 2016. BlastKOALA and GhostKOALA: KEGG Tools
for Functional Characterization of Genome and Metagenome Sequences. J. Mol. Biol.
428: 726-731.
Kang, L.-K., S.-P. L. Hwang, G.-C. Gong, H.-J. Lini, P.-C. Chen, and J. Chang. 2007. Influences
of nitrogen deficiency on the transcript levels of ammonium transporter, nitrate
transporter and glutamine synthetase genes in Isochrysis galbana (Isochrysidales,
Haptophyta). Phycologia 46: 521-533.
Keeling, P. J. and others 2014. The Marine Microbial Eukaryote Transcriptome Sequencing
Project (MMETSP): Illuminating the functional diversity of eukaryotic life in the oceans
through transcriptome sequencing. PLoS Biology 12: e1001889.
Koid, A. E., Z. Liu, R. Terrado, A. C. Jones, D. A. Caron, and K. B. Heidelberg. 2014.
Comparative Transcriptome Analysis of Four Prymnesiophyte Algae. PloS One 9:
e97801.
Kopylova, E., L. Noe, and H. Touzet. 2012. SortMeRNA: fast and accurate filtering of ribosomal
RNAs in metatranscriptomic data. Bioinformatics 28: 3211-3217.
Lasek-Nesselquist, E., J. H. Wisecaver, J. D. Hackett, and M. D. Johnson. 2015. Insights into
transcriptional changes that accompany organelle sequestration from the stolen nucleus of
Mesodinium rubrum. BMC Genomics: 1-14.
Lee, J., and D. L. Herrin. 2002. Assessing the relative importance of light and the circadian clock
in controlling chloroplast translation in Chlamydomonas reinhardtii. Photosyn. Res. 72:
295-306.
Li, D., C.-M. Liu, R. Luo, K. Sadakane, and T.-W. Lam. 2015. MEGAHIT: an ultra-fast single-
node solution for large and complex metagenomics assembly via succinct de Bruijn
graph. Bioinformatics 31: 1674-1676.
Li, W., L. Jaroszewski, and A. Godzik. 2001. Clustering of highly homologous sequences to
reduce the size of large protein databases. Bioinformatics 17: 282-283.
Lie, A. A. Y., Z. Liu, R. Terrado, A. O. Tatters, K. B. Heidelberg, and D. A. Caron. 2017. Effect
of light and prey availability on gene expression of the mixotrophic chrysophyte,
Ochromonas sp. BMC Genomics 18: 1-17.
---. 2018. A tale of two mixotrophic chrysophytes: Insights into the metabolisms of two
Ochromonas species (Chrysophyceae) through a comparison of gene expression. PloS
One 13: e0192439-0192420.
Liu, H. and others 2009. Extreme diversity in noncalcifying haptophytes explains a major
pigment paradox in open oceans. P. Natl. Acad. Sci. USA 106: 12803-12808.
Liu, Z., V. Campbell, K. B. Heidelberg, and D. A. Caron. 2016. Gene expression characterizes
different nutritional strategies among three mixotrophic protists. FEMS Microbiol. Ecol.
92: fiw106-111.
Liu, Z., A. C. Jones, V. Campbell, K. D. Hambright, K. B. Heidelberg, and D. A. Caron. 2015a.
Gene expression in the mixotrophic prymnesiophyte, Prymnesium parvum, responds to
prey availability. Front Microbiol: 1-44.
Liu, Z., A. E. Koid, R. Terrado, V. Campbell, D. A. Caron, and K. B. Heidelberg. 2015b.
Changes in gene expression of Prymnesium parvum induced by nitrogen and phosphorus
limitation. Front Microbiol 6: 1-13.
Logares, R. and others 2013. Metagenomic 16S rDNA Illumina tags are a powerful alternative to
amplicon sequencing to explore diversity and structure of microbial communities.
Environ. Microbiol. 16: 2659-2671.
138
Lund, S. P., D. Nettleton, D. J. McCarthy, and G. K. Smyth. 2012. Detecting Differential
Expression in RNA-sequence Data Using Quasi-likelihood with Shrunken Dispersion
Estimates. Stat Appl Genet Molec Biol 11: 1-44.
Mangot, J.-F. and others 2017. Accessing the genomic information of unculturable oceanic
picoeukaryotes by combining multiple single cells. Scientific Reports 7: 1-12.
Marchetti, A., C. A. Durkin, M. S. Parker, A. E. Allen, and E. V. Armbrust. 2012. Comparative
metatranscriptomics identifies molecular bases for the physiological responses of
phytoplanktonto varying iron availability. P. Natl. Acad. Sci. USA 109: E317-E325.
McKnight, D. M., B. L. Howes, C. D. Taylor, and D. D. Goehringer. 2000. Phytoplankton
dynamics in a stably stratified Antarctic lake during winter darkness. J. Phycol. 36: 852-
861.
McMinn, A., and A. Martin. 2013. Dark survival in a warming world. P. Roy. Soc. B-Biol. Sci.
280: 20122909.
Meyer, J. M. and others 2015. Transcriptomic characterisation and genomic glimps into the
toxigenic dinoflagellate Azadinium spinosum, with emphasis on polykeitde synthase
genes. BMC Genomics 16: 79.
Mitra, A. and others 2016. Defining Planktonic Protist Functional Groups on Mechanisms for
Energy and Nutrient Acquisition: Incorporation of Diverse Mixotrophic Strategies.
Protist 167: 106-120.
Mock, T., and B. Kroon. 2002. Photosynthetic energy conversion under extreme conditions - I:
important role of lipids as structural modulators and energy sink under N-limited growth
in Antarctic sea ice diatoms. Phytochemistry 61: 41-51.
Morey, J. S., E. A. Monroe, and A. L. Kinney. 2011. Transcriptomic response of the red tide
dinoflagellate, Karenia brevis, to nitrogen and phosphorus depletion and addition. BMC
Genomics 12: 549.
Moustafa, A. and others 2010. Transcriptome Profiling of a Toxic Dinoflagellate Reveals a
Gene-Rich Protist and a Potential Impact on Gene Expression Due to Bacterial Presence.
PloS One 5: e9688.
Mulisch, M. 1993. Chitin in protistan organisms: distribution, synthesis and deposition. Eur. J.
Protistol. 29: 1-18.
Müller, M. 1988. Energy metabolism of protozoa without mitochondria. Annu. Rev. Microbiol.
42: 465-488.
Nowack, E., and A. R. Grossman. 2012. Trafficking of protein into the recently established
photosynthetic organelles of Paulinella chromatophora. P. Natl. Acad. Sci. USA 109:
5340-5345.
Nymark, M. and others 2013. Molecular and Photosynthetic Responses to Prolonged Darkness
and Subsequent Acclimation to Re-Illumination in the Diatom Phaeodactylum
tricornutum. PloS One 8: e58722.
Ohtsuka, S., T. Suzaki, T. Horiguchi, N. Suzuki, and F. Not. 2015. Marine Protists: Diversity and
Dynamics. Springer: Japan.
Orsi, W., S. Charvet, P. Vd’acˇný, J. M. Bernhard, and V. P. Edgcomb. 2012a. Prevalence of
partnerships between bacteria and ciliates in oxygen-depleted marine water columns.
Front Microbiol 3: 1-8.
Orsi, W. and others 2012b. Class Cariacotrichea, a novel ciliate taxon from the anoxic Cariaco
Basin, Venezuela. Int. J. Syst. Evol. Micr. 62: 1425-1433.
139
Orsi, W. D., V. P. Edgcomb, G. D. Christman, and J. F. Biddle. 2013. Gene expression in the
deep biosphere. Nature 499: 205-208.
Pachiadaki, M. G., C. Taylor, A. Oikonomou, M. M. Yakimov, T. Stoeck, and V. Edgcomb.
2014a. In situ grazing experiments apply new technology to gain insights into deep-sea
microbial food webs. Deep-Sea Res PT II: 1-32.
Pachiadaki, M. G., M. M. Yakimov, V. LaCono, E. Leadbetter, and V. Edgcomb. 2014b.
Unveiling microbial activities along the halocline of Thetis, a deep-sea hypersaline
anoxic basin. ISME J. 8: 2478-2489.
Patro, R., G. Duggal, M. I. Love, R. A. Irizarry, and C. Kingsford. 2017. Salmon provides fast
and bias-aware quantification of transcript expression. Nat. Methods 14: 1-10.
Pearson, G. A. and others 2015. Metatranscriptomes reveal functional variation in diatom
communities from the Antarctic Peninsula. ISME J. 9: 2275-2289.
Pernice, M. C. and others 2014. Global abundance of planktonic heterotrophic protists in the
deep ocean. ISME J. 9: 782-792.
Pernice, M. C. and others 2015. Large variability of bathypelagic microbial eukaryotic
communities across the world's oceans. ISME J. 10: 945–958.
Peters, E., and D. N. Thomas. 1996. Prolonged darkness and diatom mortality I: Marine
Antarctic species. J Exp Mar Biol Ecol 207: 25-41.
Poliner, E. and others 2015. Transcriptional coordination of physiological responses in
Nannochloropsis oceanica CCMP1779 under light/dark cycles. Plant J. 83: 1097-1113.
Quast, C. and others 2012. The SILVA ribosomal RNA gene database project: improved data
processing and web-based tools. Nucleic Acids Res. 41: D590-D596.
R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for
Statistical Computing.
Roessler, P. G. 1988. Changes in the activities of various lipid and carbohydrate biosynthetic
enzymes in the diatom Cyclotella cryptica in response to silicon deficiency. Arch
Biochem Biophys 267: 521-528.
Round, F. E., R. M. Crawford, and D. G. Mann. 1990. The Diatoms: Biology and Morphology of
the Genera. Cambridge University Press.
Santoferrara, L. F., S. Guida, H. Zhang, and G. B. McManus. 2014. De Novo Transcriptomes of
a Mixotrophic and a Heterotrophic Ciliate from Marine Plankton. PloS One 9: e101418.
Seeleuthner, Y. and others 2018. Single-cell genomics of multiple uncultured stramenopiles
reveals underestimated functional diversity across oceans. Nat. Commun.: 1-10.
Settembre, C., A. Fraldi, D. L. Medina, and A. Ballabio. 2013. Signals from the lysosome: a
control centre for cellular clearance and energy metabolism. Nat Rev Mol Cell Biol 14:
283-296.
Sherr, E. B., and B. F. Sherr. 1994. Bacterivory and herbivory: key roles of phagotrophic protists
in pelagic food webs. Microb. Ecol. 28: 223-235.
---. 2002. Significance of predation by protists in aquatic microbial food webs. Antonie Van
Leeuwenhoek 81: 293-308.
Sickogoad, L., E. F. Stoermer, and J. P. Kociolek. 1989. Diatom Resting Cell Rejuvenation and
Formation - Time Course, Species Records and Distribution. J. Plankton Res. 11: 375-
389.
Smayda, T. J., and B. Mitchell-Innes. 1974. Dark survival of autotrophic, planktonic marine
diatoms. Mar. Biol. 25: 195-202.
140
Smith, S. R., R. M. Abbriano, and M. Hildebrand. 2012. Comparative analysis of diatom
genomes reveals substantial differences in the organization of carbon partitioning
pathways. ALGAL 1: 2-16.
Solomon, C. M., J. L. Collier, G. M. Berg, and P. M. Glibert. 2010. Role of urea in microbial
metabolism in aquatic systems: a biochemical and molecular review. Aquat. Microb.
Ecol. 59: 67-88.
Solomon, C. M., and P. M. Glibert. 2008. Urease activity in five phytoplankton species. Aquat.
Microb. Ecol. 52: 149-157.
Stock, A., K. Jürgens, J. Bunge, and T. Stoeck. 2009. Protistan diversity in suboxic and anoxic
waters of the Gotland Deep (Baltic Sea) as revealed by 18S rRNA clone libraries. Aquat.
Microb. Ecol. 55: 267-284.
Stoeck, T. and others 2014. Living at the Limits: Evidence for Microbial Eukaryotes Thriving
under Pressure in Deep Anoxic, Hypersaline Habitats. Adv. Ecol. Res 2014: 1-9.
Strom, S. L., R. Benner, and S. Ziegler. 1997. Planktonic grazers are a potentially important
source of marine dissolved organic carbon. Limnol. Oceanogr. 42: 1364-1374.
Tang, S., A. Lomsadze, and M. Borodovsky. 2015. Identification of protein coding regions in
RNA transcripts. Nucleic Acids Res. 43: e78-e78.
Unrein, F., J. M. Gasol, F. Not, I. Forn, and R. Massana. 2013. Mixotrophic haptophytes are key
bacterial grazers in oligotrophic coastal waters. ISME J. 8: 164-176.
Ward, B. A., and M. J. Follows. 2016. Marine mixotrophy increases trophic transfer efficiency,
mean organism size, and vertical carbon flux. P. Natl. Acad. Sci. USA: 201517118-
201517116.
Wickham, H. 2009. ggplot2: Elegant Graphics for Data Analysis. Springer Publishing Company,
Incorporated.
Worden, A. Z. 2006. Picoeukaryote diversity in coastal waters of the Pacific Ocean. Aquat.
Microb. Ecol. 43: 165-175.
Worden, A. Z., M. J. Follows, S. J. Giovannoni, S. Wilken, A. E. Zimmerman, and P. J. Keeling.
2015. Rethinking the marine carbon cycle: Factoring in the multifarious lifestyles of
microbes. Science 347: 1257594-1257594.
Worden, A. Z., J. K. Nolan, and B. Palenik. 2004. Assessing the dynamics and ecology of marine
picophytoplankton: The importance of the eukaryotic component Limnol. Oceanogr. 49:
168-179.
Wurch, L. L., S. T. Haley, E. D. Orchard, C. J. Gobler, and S. T. Dyhrman. 2011. Nutrient-
regulated transcriptional responses in the brown tide-forming alga Aureococcus
anophagefferens. Environ. Microbiol. 13: 468-481.
Wylezich, C., and K. Jürgens. 2011. Protist diversity in suboxic and sulfidic waters of the Black
Sea. Environ. Microbiol. 13: 2939-2956.
Zielinski, B. L. and others 2016. Patterns of Transcript Abundance of Eukaryotic
Biogeochemically-Relevant Genes in the Amazon River Plume. PloS One 11: e0160929-
0160920.

141
Supplementary Materials
Supporting materials for all chapters can be found either below or at (zenodo) DOI:
10.5281/zenodo.1204376. Noted below when supporting figures, tables, and documents are only
available at this DOI.
142
Supplementary Material for Chapter One

Shallow
Deep
10
20
40
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.08
G) Full-length
Similarity (%)
Depth
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.08
E) V4-V7
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.09
F) V1-V7
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.08
C) V1-V3
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.1
A) V7
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.09
B) V4
ENP5
ENP500
RS20
RS600
GS15
GS105
GS2500
AO35
AO500
EPR20
EPR1500
EPR2500
2D Stress: 0.08
D) V1-V4
143
Figure S1. Comparison of nmMDS plots of full-length and in silico extracted regions.
Nonmetric Multidimensional Scaling (nmMDS) plots of the Bray-Curtis dissimilarity parameter
based on OTUs called at 97% sequence similarity. nmMDS plots and Bray-Curtis dissimilarity
values for full-length sequences and extracted regions were generated in Primer Ev. 6 (Clark and
Warwick 2001). Similarity overlays represent 10% (solid black lines), 20% (dashed black lines),
and 40% community similarity (solid grey lines). Each sample is labeled by a location and depth
identifier (e.g. AO500: Arctic Ocean, 500 m), Arctic Ocean (AO), the San Pedro Ocean time
series station in the eastern North Pacific (ENP), East Pacific Rise (EPR) in the eastern Pacific,
Gulf Stream (GS) in the western North Atlantic, and Ross Sea, Antarctica (RS), see Table S2.
The pattern of the nmMDS for the V7 region (A) was not consistent (lacked the resolving power)
with nmMDS plots for the V4 (B), V1-V3 (C), V1-V4 (D), V4-V7 (E), V1-V7 (F), and full-
length sequences (G). In contrast, nmMDS plots of the datasets consisting of sequences ≥400 nt
yielded patterns that were generally consistent with one another (B-F). Only minor differences
(e.g. the grouping of the Ross Sea samples, and some North Pacific and North Atlantic samples)
were apparent among the patterns generated by longer sequences (B-F).

SUPPLEMENTAL TABLES
Table S1. Summary of mean nucleotide (nt) length of each sequence fragment used, along with
forward and reverse primer and reference position on the 18S rRNA gene used for in silico
extraction. Position reference refers to the nucleotide position in the aligned full-length 18S
rRNA gene sequence database. Fragment lengths reported include the primer sequences. The V1-
V3 and V4 regions overlap along the 18S rRNA gene. Primers chosen represent commonly used
primers for sequencing environmental microbial eukaryotic communities and therefore serve to
provide realistic examples of short fragments that would be used in HTS.
Region ID
Mean
length
(nt)
Primer sequence (5’-3’):
Forward
Reverse
Position
Reference Primer reference
V7 107 AAT TTG ACT CAA CAC GGG 3,324 Hadziavdic et al. (2014)
ACT AAG AAC GGC CAT GCA CC 3,518
V4 418 CCA GCA [GC]C [CT]G CGG TAA TTC C 1,141 Stoeck et al. (2010a)
ACT TTC GTT CTT GAT [CT][GA]A 1,593
V1-V3 533 AAC CTG GTT GAT CCT GCC AGT 0 Medlin et al. (1988b)
GTA ATT CCA GCT CCA ATA GC 1,213 Weekers et al. (1994a)
V1-V4 690 AAC CTG GTT GAT CCT GCC AGT 0 Medlin et al. (1988b)
CCA GCA [GC]C [CT]G CGG TAA TTC C 1,593 Stoeck et al. (2010a)
V4-V7 1,200 CCA GCA [GC]C [CT]G CGG TAA TTC C 1,141 Stoeck et al. (2010a)
ACT AAG AAC GGC CAT GCA CC 3,518 Hadziavdic et al. (2014)
V1-V7 1,260 AAC CTG GTT GAT CCT GCC AGT 0 Medlin et al. (1988b)
ACT AAG AAC GGC CAT GCA CC 3,518 Hadziavdic et al. (2014)
Full-length 1,647 AAC CTG GTT GAT CCT GCC AGT Medlin et al. (1988b)
TGA TCC TTC TGC AGG TTC ACC TAC

144

Table S2. Summary of sampling dates, coordinates, depth, and number of quality checked
sequences recovered from each sample.
Samples

Abbrev.
Date of
collection
(mm/dd/yy) Latitude Longitude
Depth
(meters)
Sequences
per sample
Ross Sea
RS
11/14/05 76.04 S 170.30 E 20 654

600 694
Arctic Ocean
AO
8/12/02 73.42 N 157.40 W 35 729

8/11/02 500 857
Eastern North
ENP
10/29/01 33.55 N 118.40 W 5 592
Pacific

500 339
East Pacific
EPR
12/8/03 9.84 N 104.35 W 20 441
Rise

1,500 848

2,500 675
Gulf Stream
GS
8/24/00 34.73 N 73.95 W 15 471

105 500

2,500 632

Supplementary Material for Chapter Two
Supplementary Materials and Methods

Vertical profiles of temperature, salinity, oxygen, and chlorophyll fluorescence at the SPOT
station were recorded during each CTD cast (Sea-bird Electronics, Inc., Bellevue, WA, USA).
Temperature and salinity for the Port of LA and Catalina were recorded using a profiling natural
fluorometer (Kiefer et al. 1989)(PNF, Biospherical Instruments Inc., San Diego, CA).
Chlorophyll a was measured in 100-400 mL samples filtered onto GF/F filters, using the non-
acidification method. Chlorophyll a was extracted with 4 mL 100% acetone at -20°C overnight
in the dark, and subsequently measured using a Trilogy Lab fluorometer (Turner Designs,
Sunnyvale, CA, USA) (Caron 2001; Welschmeyer 1994). Samples for inorganic nutrients were
collected at each surface station by freezing 0.2 µm seawater filtrate at -20°C in acid-cleaned
plastic scintillation vials. Frozen samples were analyzed at the Marine Science Institute
145
Analytical Lab (University of California Santa Barbara, CA, USA) for ammonium, phosphate,
nitrite (0.1 µM limit of detection), nitrate plus nitrite (0.2 µM limit of detection), and silicate (1.0
µM limit of detection).

Table S1. Collection dates for surface station samples and SPOT vertical profile. SPOT samples
were collected in conjunction with regularly scheduled monthly SPOT cruises, while seawater
from the Port of LA and nearshore to Santa Catalina Island was conducted on separate cruises
within 1 week of SPOT cruises.

Year Port of L.A. and Catalina
(surface)
SPOT (surface/5 m, SCM,
150 m, and 890 m)
2013 24 April 24 April
15 July 18 July
16 October 15 October
2014 7 January 15 January

Table S2. Surface temperature, salinity, chlorophyll a, and inorganic nutrients at Santa Catalina
Island (Catalina), the San Pedro Ocean Time series station (SPOT), and the Port of LA. Salinity
measurements concurrent with surface sampling in April (Table S1) were not measured.
*Indicates below detection; limit of detection for ammonium, phosphate, and nitrite was 0.1 µM,
nitrate plus nitrite was 0.2 µM, and silicate was 1.0 µM. Where applicable, nitrate was calculated
by subtracting nitrite from nitrate plus nitrite values.

Table S2.
Temp
(°C)
Salinity
(PSU)
Chl a
(µg L
-1
)
µM
Station Month Phosphate Silicate Nitrite Nitrate Ammonium
SPOT
January 15.5 32.3 0.22 0.17 1.73 0.12 0.08 0.30
April 16.0 33.5 0.58 0.22 1.80 <0.10*

<0.10*
July 19.6 31.0 0.34 0.10 2.75 <0.10*

<0.10*
October 18.6 33.6 0.22 0.11 <1.00* <0.10* <0.10*
Port of LA
January 15.4 32.1 2.20 0.23 6.02 0.29 2.62 2.70
April 14.5 - 12.7 0.20 1.68 <0.10*

0.78
July 18.0 32.0 5.00 0.25 8.49 0.23 0.44 0.74
October 17.2 33.6 4.53 0.16 7.39 0.15 0.65 0.53
Santa January 17.5 32.3 0.64 0.14 3.33 <0.10*

0.54
Catalina April 15.9 - 1.02 0.12 <1.00* 0.11 0.15 0.22
Island July 20.0 31.5 0.38 0.12 2.32 <0.10*

0.56

October 19.4 33.7 0.68 0.11 <1.00* <0.10* 0.34

146
Table S3. The total number of sequences and OTUs (97% sequence similarity) recovered from
each sample (see Materials and Methods for more information). Table summarizes the total
number of sequences and OTUs discarded following removal of doubleton OTUs and RNA-only
or DNA-only OTUs from each sample.

147
Table S4. Total number of RNA and DNA sequences for each major taxonomic group
separated by site and month. Major protistan taxonomic groups as defined in main text.
Port%of%LA%(surface) Catalina%(surface)
rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA %
Dinoflagellates
10,071 18.8 38,550 38.8 6,904 14.3 68,003 72.3 8,269 19.6 55,898 50.4 12,473 13.5 42,320 50.5 7,954 14.6 30,646 44.1 10,214 15.9 33,819 43.0 2,639 3.2 16,625 15.8 2,547 2.5 26,126 23.9
Ciliates
6,839 12.8 6,130 6.2 6,352 13.1 846 0.9 7,538 17.8 10,299 9.3 8,176 8.8 2,816 3.4 16,930 31.0 13,774 19.8 11,168 17.4 4,621 5.9 14,989 18.3 14,622 13.9 9,375 9.2 8,305 7.6
Other
59 0.1 129 0.1 1 0.0 1 0.0 609 1.4 5,262 4.7 103 0.1 245 0.3 101 0.2 213 0.3 110 0.2 157 0.2 7 0.0 53 0.1 12 0.0 88 0.1
Chlorophytes
3,412 6.4 3,702 3.7 994 2.1 532 0.6 4,313 10.2 6,847 6.2 11,438 12.4 7,082 8.5 2,985 5.5 1,345 1.9 2,030 3.2 2,356 3.0 18,912 23.1 2,798 2.7 23,253 22.8 22,720 20.8
Cryptophytes
11,411 21.3 4,707 4.7 1,139 2.4 105 0.1 4,721 11.2 3,787 3.4 15,662 16.9 4,009 4.8 1,402 2.6 506 0.7 1,366 2.1 312 0.4 5,862 7.1 1,927 1.8 1,973 1.9 2,688 2.5
Haptophytes
3,896 7.3 4,737 4.8 1,531 3.2 214 0.2 4,036 9.5 3,049 2.7 15,373 16.6 4,250 5.1 11,406 20.9 3,234 4.6 14,286 22.3 3,878 4.9 7,430 9.1 976 0.9 26,786 26.3 12,493 11.4
Diatoms
3,007 5.6 11,615 11.7 21,655 44.7 2,923 3.1 3,480 8.2 7,805 7.0 4,425 4.8 3,698 4.4 2,342 4.3 4,734 6.8 1,602 2.5 5,760 7.3 778 0.9 1,442 1.4 3,172 3.1 9,241 8.5
MAST
2,211 4.1 1,846 1.9 2,119 4.4 425 0.5 973 2.3 651 0.6 2,975 3.2 596 0.7 1,712 3.1 782 1.1 2,463 3.8 947 1.2 8,400 10.2 1,883 1.8 3,153 3.1 2,265 2.1
Other
6,213 11.6 4,882 4.9 2,728 5.6 228 0.2 3,451 8.2 2,254 2.0 11,187 12.1 1,965 2.3 5,831 10.7 1,070 1.5 8,534 13.3 3,341 4.3 8,754 10.7 1,426 1.4 18,389 18.0 3,556 3.3
Cercozoa
2,028 3.8 3,198 3.2 3,414 7.1 863 0.9 1,900 4.5 1,432 1.3 3,920 4.2 1,376 1.6 96 0.2 244 0.4 2,619 4.1 4,200 5.3 1,767 2.2 1,600 1.5 5,791 5.7 3,558 3.3
Radiolaria
51 0.1 236 0.2 72 0.1 52 0.1 194 0.5 257 0.2 206 0.2 701 0.8 171 0.3 1,336 1.9 686 1.1 845 1.1 914 1.1 1,312 1.2 286 0.3 1,723 1.6
Metazoa
47 0.1 13,210 13.3 159 0.3 19,163 20.4 81 0.2 8,397 7.6 394 0.4 11,664 13.9 81 0.1 7,359 10.6 340 0.5 15,141 19.3 144 0.2 47,481 45.0 59 0.1 5,964 5.5
Other
2,234 4.2 2,529 2.5 444 0.9 37 0.0 1,374 3.3 1,470 1.3 2,618 2.8 1,271 1.5 2,484 4.5 2,710 3.9 6,984 10.9 2,049 2.6 5,424 6.6 1,977 1.9 4,394 4.3 5,800 5.3
Unassigned
2,140 4.0 3,943 4.0 910 1.9 607 0.6 1,326 3.1 3,493 3.1 3,520 3.8 1,746 2.1 1,125 2.1 1,596 2.3 1,669 2.6 1,183 1.5 6,009 7.3 11,382 10.8 2,823 2.8 4,671 4.3
53,619
SPOT%(surface%/%5m) SPOT%SCM
rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA %
Dinoflagellates
12,106 21.3 28,632 29.5 13,654 11.4 19,173 26.8 5,120 4.9 34,467 17.8 3,043 3.1 24,927 22.7 14,629 17.2 50,257 51.8 27,308 19.4 21,427 45.7 7,540 13.4 52,714 36.2 6,820 6.5 20,347 37.3
Ciliates
19,658 34.6 5,308 5.5 17,318 14.4 3,169 4.4 14,165 13.5 24,362 12.6 3,272 3.3 2,223 2.0 22,555 26.5 7,718 8.0 25,302 17.9 1,650 3.5 15,593 27.7 17,304 11.9 10,996 10.6 4,513 8.3
Other
178 0.3 157 0.2 90 0.1 103 0.1 6 0.0 28 0.0 53 0.1 110 0.1 270 0.3 468 0.5 437 0.3 83 0.2 13 0.0 61 0.0 152 0.1 261 0.5
Chlorophytes
729 1.3 677 0.7 1,576 1.3 458 0.6 13,625 12.9 5,207 2.7 10,065 10.3 13,326 12.1 3,039 3.6 2,712 2.8 623 0.4 49 0.1 3,072 5.5 5,710 3.9 14,459 13.9 4,035 7.4
Cryptophytes
307 0.5 202 0.2 1,860 1.5 338 0.5 19,129 18.2 6,599 3.4 159 0.2 378 0.3 2,698 3.2 1,134 1.2 1,434 1.0 84 0.2 2,741 4.9 5,353 3.7 698 0.7 419 0.8
Haptophytes
10,335 18.2 2,968 3.1 20,674 17.2 1,903 2.7 10,630 10.1 3,095 1.6 33,179 33.9 16,727 15.2 11,731 13.8 5,989 6.2 4,576 3.2 297 0.6 15,771 28.0 23,567 16.2 21,601 20.7 4,610 8.4
Diatoms
128 0.2 144 0.1 7,105 5.9 3,435 4.8 4,766 4.5 8,886 4.6 4,279 4.4 4,092 3.7 628 0.7 4,413 4.5 46,632 33.1 8,303 17.7 220 0.4 935 0.6 636 0.6 916 1.7
MAST
1,688 3.0 432 0.4 28,173 23.4 1,094 1.5 21,671 20.6 8,816 4.6 5,961 6.1 2,682 2.4 1,799 2.1 789 0.8 17,954 12.7 445 0.9 1,128 2.0 1,245 0.9 1,898 1.8 585 1.1
Other
6,284 11.0 644 0.7 7,005 5.8 505 0.7 4,565 4.3 2,498 1.3 21,824 22.3 4,614 4.2 21,818 25.7 1,931 2.0 964 0.7 112 0.2 4,061 7.2 1,569 1.1 41,254 39.6 1,624 3.0
Cercozoa
236 0.4 243 0.3 3,071 2.6 1,216 1.7 1,166 1.1 3,104 1.6 3,231 3.3 2,974 2.7 248 0.3 353 0.4 2,677 1.9 1,393 3.0 151 0.3 616 0.4 126 0.1 261 0.5
Radiolaria
640 1.1 1,945 2.0 5,536 4.6 3,202 4.5 5,470 5.2 19,062 9.8 242 0.2 2,429 2.2 232 0.3 1,355 1.4 2,532 1.8 1,004 2.1 352 0.6 4,062 2.8 204 0.2 5,188 9.5
Metazoa
427 0.8 53,833 55.5 3,655 3.0 33,528 46.8 889 0.8 67,839 35.0 470 0.5 22,230 20.2 152 0.2 15,283 15.8 1,909 1.4 9,768 20.8 1,207 2.1 22,710 15.6 111 0.1 8,791 16.1
Other
2,828 5.0 1,313 1.4 7,497 6.2 2,217 3.1 2,545 2.4 4,295 2.2 7,340 7.5 8,296 7.6 4,160 4.9 3,083 3.2 4,930 3.5 553 1.2 3,595 6.4 8,494 5.8 3,475 3.3 1,882 3.4
Unassigned
1,342 2.4 540 0.6 3,083 2.6 1,312 1.8 1,559 1.5 5,422 2.8 4,853 5.0 4,779 4.4 1,063 1.3 1,534 1.6 3,711 2.6 1,726 3.7 870 1.5 1,131 0.8 1,788 1.7 1,145 2.1
SPOT%150%m SPOT%890%m
rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA % rRNA % rDNA %
Dinoflagellates
25,650 30.2 48,168 36.6 19,004 28.9 17,705 30.0 41,391 28.5 29,072 29.2 31,802 32.9 13,140 16.0 22,793 22.9 13,532 41.4 16,473 33.1 24,540 59.0 18,913 22.6 44,235 39.4 46,381 37.8 71,377 76.4
Ciliates
27,132 32.0 6,139 4.7 19,510 29.7 1,532 2.6 55,032 37.8 3,798 3.8 23,569 24.4 506 0.6 59,319 59.6 4,319 13.2 18,501 37.2 2,458 5.9 40,193 48.0 12,139 10.8 48,297 39.4 3,245 3.5
Other
1,564 1.8 522 0.4 445 0.7 58 0.1 1,804 1.2 157 0.2 1,476 1.5 76 0.1 267 0.3 87 0.3 358 0.7 105 0.3 331 0.4 196 0.2 553 0.5 172 0.2
Chlorophytes
68 0.1 179 0.1 95 0.1 76 0.1 61 0.0 66 0.1 112 0.1 34 0.0 60 0.1 1,034 3.2 43 0.1 16 0.0 68 0.1 51 0.0 135 0.1 221 0.2
Cryptophytes
18 0.0 10 0.0 29 0.0 11 0.0 63 0.0 3 0.0 54 0.1 2 0.0 0 0.0 0 0.0 99 0.2 5 0.0 2 0.0 1 0.0 319 0.3 6 0.0
Haptophytes
3,515 4.1 434 0.3 1,317 2.0 82 0.1 2,753 1.9 156 0.2 3,413 3.5 110 0.1 1,162 1.2 6,039 18.5 444 0.9 32 0.1 1,247 1.5 129 0.1 361 0.3 140 0.1
Diatoms
587 0.7 4,108 3.1 10,039 15.3 2,836 4.8 3,003 2.1 1,150 1.2 18,886 19.5 714 0.9 9 0.0 418 1.3 1,683 3.4 335 0.8 531 0.6 338 0.3 224 0.2 206 0.2
MAST
4,070 4.8 772 0.6 2,908 4.4 125 0.2 10,084 6.9 301 0.3 4,388 4.5 105 0.1 3,239 3.3 1,866 5.7 2,049 4.1 104 0.2 4,466 5.3 363 0.3 946 0.8 108 0.1
Other
3,454 4.1 1,052 0.8 2,358 3.6 258 0.4 4,117 2.8 258 0.3 1,711 1.8 138 0.2 1,020 1.0 3,501 10.7 2,410 4.8 154 0.4 4,858 5.8 659 0.6 765 0.6 386 0.4
Cercozoa
5,030 5.9 2,617 2.0 773 1.2 161 0.3 627 0.4 207 0.2 421 0.4 61 0.1 5 0.0 60 0.2 254 0.5 120 0.3 195 0.2 91 0.1 208 0.2 477 0.5
Radiolaria
5,425 6.4 45,657 34.7 3,279 5.0 33,442 56.7 5,577 3.8 43,800 44.0 4,444 4.6 63,339 77.3 371 0.4 320 1.0 1,281 2.6 2,399 5.8 685 0.8 3,680 3.3 295 0.2 7,199 7.7
Metazoa
927 1.1 16,450 12.5 582 0.9 1,502 2.5 4,143 2.8 16,576 16.6 1,700 1.8 2,887 3.5 3,098 3.1 448 1.4 2,158 4.3 10,477 25.2 2,563 3.1 47,821 42.6 940 0.8 8,580 9.2
Other
3,950 4.7 2,340 1.8 2,390 3.6 313 0.5 4,855 3.3 634 0.6 2,755 2.9 247 0.3 1,328 1.3 698 2.1 817 1.6 269 0.6 1,349 1.6 814 0.7 608 0.5 341 0.4
Unassigned
3,410 4.0 2,997 2.3 2,949 4.5 888 1.5 11,924 8.2 3,427 3.4 1,920 2.0 608 0.7 6,922 7.0 367 1.1 3,209 6.4 591 1.4 8,255 9.9 1,719 1.5 22,539 18.4 909 1.0
Table S4. Total number of RNA and DNA sequences for
each major taxonomic group separated by site and month.
Major protistan taxonomic groups as defined in main text.
Rhizaria
January April July
Alveolates Alveolates Stramenopiles Rhizaria
January April July
July
October January April
October
Alveolates Stramenopiles
October
Stramenopiles Rhizaria
January April July
July October
April October January April July October January
148
Table S5. Average RNA:DNA ratios for each taxonomic group. RNA:DNA ratios were
calculated for each OTU, and then the average RNA:DNA ratio for each major taxonomic group
(as defined in the main text) in a given month and site was calculated. See Figure 4 for graphical
representation of select taxonomic groups.
149

Table S6. Differences in RNA:DNA ratios by month and site were evaluated using analysis of
variance (ANOVA) tests and Tukey's test for honestly significantly differences (P<0.05,
conf.=0.95). Pairwise comparisons are separated either by month (i.e. January-April, October-
April, etc) or spatially by site (i.e. SPOT 5m-150m, Catalina-150m, etc. The column labeled "All
months", considers OTUs from all months sampled. Bolded values are significant.
DOI: 10.5281/zenodo.1204376 – “CH2_SupplementaryData_Hu_et_al-2016-
FEMS_Microbiol_Ecol.zip”

Table S7. OTU table summarizing the total number of sequences found in each OTU and
sample. Major taxonomic group assigned and SILVA designated taxonomy also included.
DOI: 10.5281/zenodo.1204376 – “CH2_SupplementaryData_Hu_et_al-2016-
FEMS_Microbiol_Ecol.zip”

150
Table S8. Sequence counts with all OTUs (includes RNA and DNA-only OTUs) show similar
trends to data without RNA and DNA-only OTUs in each sample. See main text for further
details. Compare to Table S4.

rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA
Dinoflagellates 10,221 41,409 6,981 70,596 8,354 58,641 12,709 43,999 8,127 33,739 10,426 37,993 2,699 17,512 2,601 29,806
Ciliates 7,089 6,186 6,677 858 7,674 10,420 8,531 2,832 17,658 13,879 11,542 4,647 15,152 14,727 9,715 8,430
Other 74 165 5 42 615 5,336 114 296 116 238 118 240 11 62 12 104
Chlorophytes 3,499 3,891 1,134 569 4,409 7,007 11,640 7,157 3,109 1,424 2,222 2,438 19,234 2,895 23,371 22,923
Cryptophytes 11,672 4,759 1,174 108 4,792 3,814 15,934 4,013 1,442 523 1,407 325 5,997 1,951 2,052 2,705
Haptophytes 4,195 4,795 1,638 223 4,397 3,162 16,853 4,336 12,963 3,294 15,958 3,971 8,299 1,010 28,565 12,626
Diatoms 3,096 11,809 22,140 2,961 3,526 7,911 4,582 3,739 2,489 4,900 1,694 6,213 877 1,548 3,544 9,373
MAST 2,371 1,896 2,417 429 1,043 681 3,281 605 1,877 858 2,679 978 8,804 1,941 3,263 2,308
Other 6,492 4,928 3,433 245 3,655 2,310 11,663 1,986 6,167 1,146 8,895 3,373 9,291 1,466 18,694 3,638
Cercozoa 2,110 3,311 3,955 902 2,016 1,484 4,141 1,410 137 306 2,693 4,314 1,839 1,655 5,820 3,636
Radiolaria 63 312 75 62 198 267 241 718 194 1,375 706 852 975 1,316 288 1,779
Metazoa 49 14,443 176 19,284 81 8,571 398 12,011 83 7,454 375 15,223 151 47,828 63 6,176
Other 2,345 2,619 663 51 1,482 1,523 2,814 1,330 2,684 2,774 7,235 2,105 5,727 2,042 4,503 5,932
Unassigned 2,302 4,507 1,221 717 1,421 4,217 3,963 2,145 1,399 1,946 2,007 1,381 6,284 11,565 3,050 4,984
rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA
Dinoflagellates 12,355 30,829 13,894 20,459 5,221 37,240 3,107 27,526 14,949 55,270 27,892 22,207 7,572 54,328 6,949 22,367
Ciliates 20,798 5,339 17,718 3,214 14,332 24,456 3,401 2,253 23,220 7,763 26,573 1,664 15,830 17,385 11,606 4,554
Other 196 195 96 129 12 66 61 118 306 479 470 92 15 220 165 280
Chlorophytes 786 757 1,759 518 13,934 5,375 10,191 13,431 3,132 2,863 751 52 3,093 5,837 14,511 4,057
Cryptophytes 335 229 1,891 344 19,444 6,604 163 470 2,746 1,162 1,498 84 2,788 5,377 714 434
Haptophytes 11,480 3,007 22,436 1,954 11,576 3,150 34,966 16,958 12,826 6,018 4,933 300 16,636 23,754 23,545 4,688
Diatoms 146 234 7,546 3,497 5,406 9,019 4,375 4,152 652 4,944 47,147 8,367 249 1,034 670 975
MAST 1,881 489 29,163 1,114 22,383 8,856 6,145 2,734 2,085 805 18,685 451 1,185 1,294 2,037 606
Other 6,739 666 7,725 521 4,839 2,567 22,302 4,734 22,264 1,989 1,614 122 4,122 1,618 41,760 1,683
Cercozoa 291 290 3,227 1,246 1,235 3,289 3,271 3,026 286 399 2,892 1,402 166 686 158 324
Radiolaria 690 1,989 5,695 3,205 5,496 19,089 244 2,498 246 1,809 2,766 1,030 352 4,150 211 5,459
Metazoa 432 53,929 3,742 33,868 904 71,000 477 22,374 156 15,394 1,944 9,802 1,224 23,289 115 8,946
Other 3,098 1,337 7,783 2,244 2,645 4,387 7,572 8,401 4,413 3,135 5,298 565 3,707 8,598 3,716 1,916
Unassigned 1,670 757 3,373 1,516 1,721 5,804 5,231 5,295 1,425 2,683 4,551 1,889 916 1,505 2,068 1,344
rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA rRNA rDNA
Dinoflagellates 26,844 52,582 20,293 18,678 43,405 30,365 33,842 14,198 24,191 18,937 17,455 25,348 19,575 45,579 47,158 73,985
Ciliates 29,997 6,199 21,791 1,542 60,385 3,858 29,544 511 62,120 11,046 19,597 2,470 41,820 12,228 49,227 3,286
Other 1,710 546 501 62 1,940 181 1,646 80 303 117 409 113 353 241 601 210
Chlorophytes 121 219 165 82 141 81 157 45 87 5,505 88 53 137 182 198 274
Cryptophytes 314 10 279 14 629 5 157 3 458 25,210 176 5 377 1 349 9
Haptophytes 4,014 437 1,897 85 3,347 158 4,126 112 1,202 9,404 644 33 1,415 135 452 141
Diatoms 656 4,656 10,469 2,863 3,402 1,183 19,374 724 15 4,373 1,736 386 563 381 257 222
MAST 4,987 845 4,034 151 11,495 304 6,174 115 3,573 2,917 3,000 110 4,990 383 1,094 116
Other 4,223 1,118 3,363 261 5,304 274 2,845 143 2,446 10,594 3,189 180 5,592 719 1,028 422
Cercozoa 5,244 2,710 1,084 168 866 214 583 65 35 2,606 425 134 278 109 238 507
Radiolaria 5,677 48,297 3,622 34,362 6,515 45,303 4,684 64,715 509 404 1,370 4,199 735 5,139 350 10,428
Metazoa 953 16,795 624 1,626 4,254 16,764 1,755 2,974 3,163 534 2,264 10,817 2,615 48,070 999 8,746
Other 4,565 2,411 3,545 318 5,739 675 3,489 267 1,813 3,240 1,153 283 1,669 854 775 418
Unassigned 5,113 3,971 4,876 1,088 16,048 4,036 4,139 862 9,387 3,369 4,071 714 9,066 2,055 23,709 1,107
Alveolates
Stramenopiles
Rhizaria
January April October July October
Alveolates
Stramenopiles
Rhizaria
Alveolates
Stramenopiles
Rhizaria
January April July October
SPOT71507m
January April July
SPOT78907m
January April July October
SPOT7SCM SPOT7(surface7/75m)
January April July October
Table S7. Sequence counts with all OTUs (includes RNA and DNA-only OTUs)
show similar trends to data without RNA and DNA-only OTUs in each sample.
See main text for further details. Compare to Table S4. See excel sheet.
January April July October
Port7of7LA7(surface) Catalina7(surface)
151

Figure S1. Dendrogram depict level of dissimilarity among samples (i.e. with RNA and DNA-
only OTUs in each sample). Patterns in sample dissimilarity were equivalent to dendrograms in
Figure 3 based on OTU data without RNA and DNA-only OTUs. See main text for further
details. Cluster dendrograms based on average hierarchical clustering depict dissimilarity among
samples from (A) SPOT, the Port of LA, and Catalina (surface stations) DNA sequence library,
(B) surface station RNA sequence library, (C) DNA sequence library from four depths at the
A B
C
D
152
SPOT station (5 m, SCM, 150 m, and 890 m), and (D) RNA from the four depths at the SPOT
station. Data were normalized by calculating the relative abundance of each OTU, and then
Bray-Curtis dissimilarity matrices were constructed for surface stations (A and B) and for each
depth sampled at SPOT (C and D), using either DNA (A and C) or RNA (B and D) sequence
libraries. The percent dissimilarity among samples is depicted by horizontal axes. Dashed lines
depict samples that were not significantly different from one another. The only difference
between Figure 3 (based on OTUs without RNA and DNA-only OTUs per sample) is that in (C),
October samples at both 150 m and 890 m were not statistically similar to other samples at 150
m and 890 m as depicted in Figure 3C.

153

Figure S2. Weighted UniFrac cluster analyses from (A) SPOT, the Port of LA, and Catalina
(surface stations) DNA sequence library, (B) surface station RNA sequence library, (C) DNA
sequence library from four depths at the SPOT station (5 m, SCM, 150 m, and 890 m), and (D)
RNA from the four depths at the SPOT station. Weighted UniFrac distances were calculated
using a phylogenetic tree based on OTUs clustered at 97% sequence similarity (see Materials
and Methods) constructed using FastTree in QIIME (Caporaso et al. 2010; Lozupone and Knight
2005; Price et al. 2009). UniFrac results were visualized using average hierarchical clustering in
R (R Core Team 2014). Overall UniFrac results were comparable to cluster dendrograms based
on Bray-Curtis dissimilarity among samples (Figures 3 and S1). RNA-derived samples clustered
with respect to site (Port of LA versus Catalina and SPOT) and season, while these patterns were
not observed in DNA-derived samples (compare A and B). Both DNA- and RNA-derived
154
samples along the vertical profile at SPOT (C and D) support the findings that there are distinct
communities in shallow (5 m and SCM) and deep (150 m and 890 m) samples.

Supplemental Chapter Two Literature Cited

Caporaso, J. G. and others 2010. QIIME allows analysis of high-throughput community
sequencing data. Nat. Methods 7: 335-336.
Caron, D. A. 2001. Protistan herbivory and bacterivory. Methods Microbiol 30: 289-315.
Kiefer, D., W. Chamberlin, and C. Booth. 1989. Natural fluorescence of chlorophyll a:
Relationship to photosynthesis and chlorophyll concentration in the western South Pacific
gyre. Limnol. Oceanogr. 34: 868-881.
Lozupone, C., and R. Knight. 2005. UniFrac: a new phylogenetic method for comparing
microbial communities. Appl. Environ. Microb. 71: 8228-8235.
Price, M. N., P. S. Dehal, and A. P. Arkin. 2009. FastTree: Computing large minimum evolution
trees with profiles instead of a distance matrix. Mol. Biol. Evol. 26: 1641-1650.
R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for
Statistical Computing.
Welschmeyer, N. A. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll
b and pheopigments. Limnol. Oceanogr. 39: 1985-1992.

155
Supplementary Material for Chapter Three

Figure S1. Cell counts (10
5
cells /mL) for heterotrophic bacteria (orange), Prochlorococcus
(green), Synechococcus (pink), and picoeukaryotes (blue) based on flow cytometry. See
Materials and Methods for sample processing information; data courtesy of SCOPE ops team at
the University of Hawaii – protocols.io: dx.doi.org/10.17504/protocols.io.j2wcqfe.

10
5
cells/mL
10
5
cells/mL
Heterotrophic bacteria
Prochlorococcus
Picoeukaryotes
Synechococcus
156

Figure S2. Statistics on total sequences (left) and OTU distribution (right) for rDNA and rRNA
sequence libraries (top and bottom, respectively). Global singletons were excluded in this
analysis, sequences were later subsampled to 42,719 sequences (see Materials and Methods).
OTU statistics are separated by singletons (OTUs with one sequence), doubletons (OTUs with 2
sequences), and OTUs with 3 or more sequences (colors in bar plots on the right).

157

Figure S3. Taxonomic composition of total OTUs at each sampling point based on either the
rDNA (top) or rRNA (bottom) sequence library. Black horizontal lines indicate periods of
darkness (6PM-6AM). To better represent the cyclical nature of the study, 6AM was plotted
twice.

158

Figure S4. Clustering by samples based on either (left) RNA-derived results or (right) DNA-
derived results. Results from correspondence correlation analysis demonstrate the relationship
between each sample based on sequence abundance and diversity. Clustering samples based on
RNA derived results explained more variability among samples (CCA1=18.5%) relative to the
DNA-derived results (CCA1=8.75%).

159
All Supplementary Tables for Chapter Three can be found at (zenodo):
DOI: 10.5281/zenodo.1204376 - “CH3_SupplementaryTables_Hu-et-al_Diel18S.zip”

Table S1. Collection information for each sample. Samples were collected every 4 hours for 4
days, following a Lagrangian sampling schematic. Reported here, are the dates, time of
sampling, location, depth (all 15 m), and amount filtered per sample. All collected seawater was
pre-filtered through an 80 µm Nitex mesh screen to remove multicellular eukaryotes. See
protocols.io dx.doi.org/10.17504/protocols.io.hisb4ee and Materials and Methods section for
more information.

Table S2. Total sequence count results for both rDNA and rRNA sequence libraries. OTUs were
assigned taxonomic identities based on their alignment to the PR2 database (Guillou et al. 2013),
also see Materials and Methods. Manual taxonomic group names were assigned at approximately
class or phylum levels (see columns “Taxa” and “Taxa2”) in order to visualize the complex
microbial community.

Table S3. Full list of OTUs found to have significant diel rhythmicity based on RAIN analysis
(see Material and Methods). Two OTUs were determined to have diel rhythmicity based on both
DNA and RNA OTUs, they are noted by asterisks (* and **). Taxonomic identities (Level 1-8)
are derived from the PR2 database (Guillou et al. 2013), which may fall short of full species-
level characterizations for uncultured representatives (i.e. “XX” and Other-unclassified).

Table S4. List of highly significantly co-correlated OTUs based on Local Similarity Analysis
(Xia et al. 2013; Xia et al. 2011). OTUs were derived from the RNA sequence library and were
first filtered so that each OTU had at least 10 sequences and appeared in all 19 time points.
OTUs without a taxonomy assignment (Unknown or unclassified) were discarded. OTU
interactions were considered significant based on p-value (q<0.05), q-value (q<0.05), and
Spearman rank correlation coefficients (> 0.5 or < -0.5). Table lists type of significantly
correlated pair (top 20 are summarized in main text Table 2), OTU identity, manually assigned
taxonomic name, and full taxonomic identity from the PR2 database.

160
Supplementary Material for Chapter Four
Sample collection

Seawater filtration
Seawater samples were collected in 12 L Niskin bottles mounted on a CTD rosette. Tubing, in-
line filters, and carboys were acid washed (5% HCl), autoclaved, and rinsed with 0.1%
diethylpyrocarbonate (DEPC)-treated water. DEPC-treated water was prepared by incubating
milli-Q water with 0.1% DEPC and then autoclaving the solution to remove DEPC. Seawater
was gravity pre-filtered into 20 L carboys through 200 µm and 80 µm Nitex mesh to minimize
the presence of multicellular eukaryotes. Carboys were kept in the dark and processed
immediately to minimize mRNA degradation (<40 minutes total). Volumes of 1.5-3.5 L of
seawater were filtered onto six sterile GF/F filters (nominal pore size 0.7 µm, Whatman,
International Ltd. Florham Park, NJ, USA) with a downstream peristaltic pump (Cole-Parmer
Master flex peristaltic pump, Vernon Hills, IL, USA), immediately placed in 1.5 mL of RLT+
buffer (with β-Mercaptoethanol, Qiagen, Valencia, CA, USA), and flash frozen in liquid
nitrogen. Further details on sample replication and volume filtered per sample are provided in
Table S1.

Chlorophyll a and inorganic nutrient samples
Discrete samples for chlorophyll a were taken by filtering 100-500 mL of seawater onto GF/F
filters and extracting the filters with 4 mL of 100% acetone overnight at -20°C in the dark.
Chlorophyll a was measured using a Trilogy Lab Fluorometer (Turner Designs, Sunnyvale, CA,
USA) by the non-acidification method (Welschmeyer 1994). Inorganic nutrients were collected
by freezing 0.2-µm filtered seawater at -20°C in acid washed scintillation vials. Nutrients were
analyzed at the Marine Science Institute Analytical Lab (University of California Santa Barbara,
CA, USA) for ammonium, phosphate, nitrite (0.1 µM limit of detection), nitrate plus nitrite (0.2
µM limit of detection), and silicate (1.0 µM limit of detection).

Flow cytometry
Triplicate 4 mL samples for flow cytometry were collected at each depth by preserving 80 µm
filtered seawater with formalin (1% final concentration). Cellular abundances were counted on a
FACSCalibur flow cytometer (Becton Dickinson) with the autofluoresence of photosynthetic
pigments and forward scatter used to distinguish three separate groups of picophytoplankton
(Prochlorococcus, Synechococcus, and picoeukaryotes; 0.2-2.0 µm). Average cell abundances
were converted into biomass using a carbon conversion factor of 200 fg C cell
-1
for
Synechococcus (Caron et al. 1995a), 90 fg C cell
-1
for Prochlorococcus (Casey et al. 2013;
Martiny et al. 2015), and 183 fg C µm
-3
for photosynthetic picoeukaryotes (assuming an average
diameter of 1 µm) (Caron et al. 1995a). Total bacterial and archaeal abundance was determined
using flow cytometry and standard staining with SYTO13 (del Giorgio et al. 1996) and then
converted to biomass using a conversion factor of 15 fg C ml
-1
(Caron et al. 1995a). Reported
flow cytometry values are an average of triplicate measurements.

Microscopy
Microscopical counts for nanoplankton (2-20 µm) and microplankton (20-80 µm) were obtained
from 80 µm pre-filtered samples that were formalin preserved (1% final conc.). Nanoplankton
161
samples were vacuum-filtered down to 1 mL onto 25 mm 0.2 µm blacked polycarbonate filters
and stained with 50 µL of 1 mg mL
-1
of 4’,6’-diamidion-2-phenylindole (DAPI, Sigma-Aldrich,
St. Louis, MO, USA, D9542) for 5 minutes in the dark. The remainder of the water was filtered,
and the polycarbonate filters were quickly transferred onto glass slides. Filters were covered with
immersion oil and a glass cover slip, sealed with clear nail polish, and stored in the dark at -
20°C. Nanoplankton slides were counted using the fluorescence interaction of DAPI with UV
light excitation at 100X magnification on an epifluoresence microscope; phototrophic and
heterotrophic/mixotrophic nanoplankton were determined by the presence or absence of
photosynthetic pigments under blue light excitation. Microplankton counts were settled in
Utermöhl chambers for 24 to 48 hours and counted on an inverted light microscope at 400x
magnification (Utermohl 1958). Cell abundances of diatoms, dinoflagellates, and ciliates
(separated by loricate and aloricate) were converted to biomass using a carbon conversion of
0.138 ng C cell
-1
.

Metatranscriptome library preparation

Filters frozen in RLT+ buffer were simultaneously thawed and bead-beaten by adding RNase-
free silica beads and vortexing for 5 minutes (protocols.io;
dx.doi.org/10.17504/protocols.io.hk3b4yn). Total extracted RNA was quality checked using the
Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA) to ensure integrity of RNA, and the
total RNA concentration was measured using a Qubit fluorometer (ThermoFisher Scientific,
Waltham, MA, USA). Total RNA from some 150 m and 890 m replicates was pooled prior to
library preparation, due to low biomass (see Table S1 for details). Before library preparation, the
total concentration of RNA was normalized by dilution and ERCC spike-in was added (1 µl at
1:100 for surface samples and 1:1000 for 150 m and 890 m, ThermoFisher Scientific,
#4456740). Sequence libraries were prepared with Kapa’s Stranded mRNA library preparation
kit using poly-A tail selection beads to select for eukaryotic mRNA (Kapa Biosystems, Inc.,
Wilmington, MA, USA, #KK8420). 2 x 125 bp PE High Output HiSeq 2500 (Illumina, San
Diego, CA, USA) sequencing was performed at the UPC Genome Core (University of Southern
California, Los Angeles, CA, USA). Replicates from each depth were run on a single lane and
are publicly available under accession numbers SAMN07269826-SAMN07269838 at the Short
Read Archive (Table S1).

Generation of custom database

The custom database included data from the Marine Microbial Eukaryote Transcriptome
Sequencing Project (MMETSP; Table S2). Combined assemblies from MMETSP were used for
species with multiple transcriptomes and single assemblies were used for those without multiple
transcriptomes. Publically available genomes and transcriptomes of most aquatic, non-parasitic
protists, and a selection of fungi and metazoan sequences were included in our custom database
to enhance the number of represented eukaryotic lineages (Table S2). Both cDNA and protein
sequences were downloaded to create separate databases. CD-HIT v. 4.5.7 (Li et al. 2001) was
used to reduce the redundancy in the database at 95% identity for sequences of each genus. Final
database is available at zenodo DOI: 10.5281/zenodo.846379.

162
Bioinformatic pipelines and databases

ERCC spike-in sequences were quantified using Trinity v. 2.1.1 to ensure that similar
percentages (~1% per sample) of spike-in sequence was recovered in each sample and the
number of reads for each ERCC RNA were consistent with their designed concentration
(Ambion LifeTech/Thermo Fisher Scientific, Waltham, MA, USA).

Detailed descriptions of bioinformatic pipelines and protocols can be found at
https://github.com/shu251/SPOT_metatranscriptome. Links to custom database files and
assemblies can be found at zenodo DOI: 10.5281/zenodo.1202041.

Supplemental Chapter Four Literature Cited
Caron, D. A., H. G. Dam, P. Kremer, and E. J. Lessard. 1995. The contribution of
microorganisms to particulate carbon and nitrogen in surface waters of the Sargasso Sea
near Bermuda. Deep-Sea Res PT I 42: 943-972.
Casey, J. R., J. P. Aucan, S. R. Goldberg, and M. W. Lomas. 2013. Changes in partitioning of
carbon amongst photosynthetic pico- and nano-plankton groups in the Sargasso Sea in
response to changes in the North Atlantic Oscillation. Deep-Sea Res PT II 93: 58-70.
del Giorgio, P. A., D. F. Bird, Y. Prairie, and D. Planas. 1996. Flow cytometric determination of
bacterial abundance in lake plankton with the green nucleic acid stain SYTO 13. Limnol.
Oceanogr. 41: 783-789.
Li, W., L. Jaroszewski, and A. Godzik. 2001. Clustering of highly homologous sequences to
reduce the size of large protein databases. Bioinformatics 17: 282-283.
Martiny, A. C. and others 2015. Biogeochemical interactions control a temporal succession in
the elemental composition of marine communities. Limnol. Oceanogr. 61: 531-542.
Utermohl, H. 1958. Zur gewassertypenfrage tropischer seen. Verh Int Ver Limnol 13: 236-251.
Welschmeyer, N. A. 1994. Fluorometric analysis of chlorophyll a in the presence of chlorophyll
b and pheopigments. Limnol. Oceanogr. 39: 1985-1992.

163
Supplemental Figures Chapter Four

Figure S1. Bar plots representing metagenomic Illumina tags (miTags) relative abundance at
each depth at the (A) domain level and (B) major protistan taxonomic group. Following
taxonomy assignment (uclust at 97% identity against the PR2 database, see Experimental
Procedures for more details), miTags were subsampled to select the five main taxonomic groups
that were the focus of this study. “Other eukaryotes” are comprised of taxonomic groups that
made up less than 10% of the total composition of miTags. Relative miTag abundances for
dinoflagellates, ciliates, haptophytes, diatoms, and chlorophytes were generally similar to
mRNA-derived diversity (Figure 4.2, Tables S6 and S7), and to that previously observed at the
SPOT station (Hu et al. 2016). MArine STramenopiles (MAST), syndiniales, excavates, and
rhizaria were represented in the miTag results (>10% miTags, Table S7), but these same groups
made up fewer than 2% of the mRNA-derived results (compare to Figure 4.2, Tables S6 and S7).

164

Figure S2. Bar plots representing metagenomic Illumina tags (miTags) relative abundance at the
class level for (A) dinoflagellates, (B) ciliates, (C) haptophytes, (D) diatoms, and (E)
chlorophytes. Following taxonomy assignment (uclust at 97% identity against the PR2 database,
see Experimental Procedures for more details), miTags were subsampled to select the five main
taxonomic groups that were the focus of this study. miTags with more than 500 hits were
selected and summed based on the approximately class level. Class level designations labeled as
“Uncertain” represent miTags assigned to only the phylum level (e.g. Haptophyte, diatom, etc.),
meaning no additional taxonomic information was available. “uncultured eukaryote” and
“Ambiguous_taxa” designations also denote assignments where no further details were available
from the PR2 database. In this study, miTag results are mainly represented at the approximately
phylum level, as short read alignments (~125 bps) for taxonomy assignment are considered less
accurate than, for instance, longer hypervariable regions of the 18S rRNA gene (e.g. V4, ~400
bps). Full list of miTag results are reported in Table S7.

165

Figure S3. Distribution of miTags among (A) all hits, (B) dinoflagellates, (C) ciliates, (D)
haptophytes, (E) diatoms, and (F) chlorophytes. Bar plots show the total number of shared or
unique miTags among depths, depths included are represented by filled dots in the matrix below
each barplot. Numbers represent the total number of miTags in each category. Most miTags were
shared among all depths (left-most bars), except for chlorophytes where the highest abundance of
miTags was unique to the surface. Plot generated using ‘UpSetR’ R package Conway et al. 2017.
A
B
C D
F
E
166

Figure S4. KEGG module assignments for ortholog groups found to be present at all depths
(surface, 150m, and 890m, n=36 656). See main text Figure 4.3.

167

Figure S5: Distribution of ortholog groups among the five targeted taxonomic groups. Bar plots
show the total number of shared or unique ortholog groups among depths, depths included are
represented by filled dots in the matrix below each barplot. Numbers represent the total number
of ortholog clusters in each category. Most ortholog clusters were unique to a single taxonomic
group (five left-most barplots). Ortholog groups associated with dinoflagellates overwhelmed
sampling efforts. Very few ortholog groups were shared among all taxonomic group (n=196,
right-most barplot). Plot generated using ‘UpSetR’ R package Conway et al. 2017.

168
Supplemental Table Legends Chapter Four

All supplementary material can be found at zenodo:
DOI: 10.5281/zenodo.1204376 – “CH4_SupplementaryMaterials_Hu-et-al_SPOTmetaT.zip”

Table S1. Collection information for all samples, including: time of collection, location, depth,
number of replicates, volume filtered, sampling time, and sample pooling information for library
preparation.

Table S2. List of contributing transcriptomes and genomes for custom database. Level 1, 2, and
3 correspond to manual taxonomic designation used in this analysis. Level 2 used in main
analysis. See Supporting Information for more information on how the custom database was
generated

Table S3. List of gene names and KEGG IDs (K0_number) for labeled targeted metabolic
pathways and genes in main text. Associated labels are based on KEGG modules.

Table S4. Summary of environmental variables and results from microscopy counts for each
depth at the SPOT station. Detailed explanation of cell types for microplankton counted can be
found in Supporting Information and methods.

Table S5. Summary of reads throughout sample processing, including total number of reads post
trimming and quality checking, assembled contigs, predicted protein sequences, and clustered
ortholog groups.

Table S6. Transcript counts per million (CPM, based on the mean across replicates, n) and
relative abundance (%) of taxonomic groups found at each depth. Percentage of the total is in the
right-most columns. The first column details the taxonomic grouping referred to in the main text.
Taxonomic groups that made up more than 2% of the community are highlighted in yellow and
discussed further in main text.

Table S7. Total counts and percentages of miTag results at each depth (see Experimental
Procedures). Top table summarizes the top taxonomic groups also represented in Figure S1. The
bottom table lists the taxonomic groups from Levels 1 to 3, which originate from the SILVA
database. MiTag results needed to make up a large percentage of the population (>10%) to be
represented in Figure S1.

Table S8. Results from Tukey HSD post hoc test (following ANOVA) to look for significant
differences among whole pathways based on transcript abundances (CPM) among replicates
(Figure 4.4 in main text). Significant values are bolded, p<0.05.

Table S9. Results from Tukey HSD post hoc test (following ANOVA) to look for significant
differences among whole pathways for individual taxonomic groups based on transcript
abundances (CPM) among replicates (Figure 4.6 in main text). Raw transcript counts were subset
for each taxonomic group and normalized before this analysis. Significant values are bolded,
p<0.05.

Abstract (if available)

Abstract Single-celled microbial eukaryotes (protists) mediate critical elemental transformations that support ecosystem function. An overarching goal in microbial ecology is to link the metabolic roles of individual microbes to food web structure, when will enable a better understanding of the biogeochemical processes driving marine ecosystems. The central goal of my dissertation work was to characterize protistan community structure and diversity in two regions in the Pacific Ocean: off the coast of California in the San Pedro Channel and northeast of Hawai’i in the North Pacific Subtropical Gyre. I used molecular techniques and bioinformatics to improve how we describe the diversity and activity of in situ protistan communities. By pairing traditional DNA-based tag sequencing with ribosomal RNA sequences, changes in protistan community structure and activity were found to correspond to depth, proximity to coastline, season, or time of day. Further, a metatranscriptome survey provided one of the first accounts of protistan physiological ecology across a depth gradient. Comparing transcript abundances (mRNA) among euphotic and sub-euphotic zone depths revealed the metabolic flexibility among key protistan lineages to rely on alternate metabolic modes/nutritional strategies (e.g. phototrophy at the surface, heterotrophy below the euphotic zone). This dissertation work addressed long-standing ecological questions regarding the metabolic roles of protists that we otherwise have no other way of observing in situ by using RNA-derived sequence information to characterize protistan diversity and metabolic potential.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Characterizing protistan diversity and quantifying protistan grazing in the North Pacific Subtropical Gyre

PDF

Changes in the community composition of marine microbial eukaryotes across multiple temporal scales of measurement

PDF

Annual pattern and response of the bacterial and microbial eukaryotic communities in an aquatic ecosystem restructured by disturbance

PDF

Marine protistan diversity, spatiotemporal dynamics, and physiological responses to environmental cues

PDF

Patterns of molecular microbial activity across time and biomes

PDF

Unexplored microbial communities in marine sediment porewater

PDF

Ecological patterns of free-living and particle-associated prokaryotes, protists, and viruses at the San Pedro Ocean Time-series between 2005 and 2018

PDF

Microbial ecology in the deep terrestrial biosphere: a geochemical, metagenomic and culture-based approach

PDF

Spatial and temporal investigations of protistan grazing impact on microbial communities in marine ecosystems

PDF

Microbe to microbe: monthly microbial community dynamics and interactions at the San Pedro Ocean Time-series

PDF

Temporal variability of marine archaea across the water column at SPOT

PDF

Identifying functional metabolic guilds: a computational approach to classifying heterotrophic diversity in the marine system

PDF

Using molecular techniques to explore the diversity, ecology and physiology of important protistan species, with an emphasis on the Prymnesiophyceae

PDF

The development of novel measures of landscape diversity in assessing the biotic integrity of lotic communities

PDF

Using sequencing techniques to explore the microbial communities associated with ferromanganese nodules and sediment from the South Pacific gyre

PDF

Dynamics of marine bacterial communities from surface to bottom and the factors controling them

PDF

The impact of the concentration and distribution of dissolved and particulate B-vitamins and their congeners on marine microbial ecology

PDF

Ecophysiology of important understudied bacterioplankton through an integrated research and education approach

PDF

Spatial and temporal dynamics of marine microbial communities and their diazotrophs in the Southern California Bight

PDF

Ecological implications of daily-to-weekly dynamics of marine bacteria, archaea, viruses, and phytoplankton

Asset Metadata

Creator Hu, Sarah Kathryn (author)

Core Title Genetic characterization of microbial eukaryotic diversity and metabolic potential

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Biology (Marine Biology and Biological Oceanography)

Publication Date 04/09/2018

Defense Date 01/19/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag 18S rRNA gene,cDNA sequencing,diversity,metatranscriptomics,microbial ecology,microbial eukaryotes,OAI-PMH Harvest,protistan diversity,protists,V4 tag sequencing

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Caron, David (committee chair), Berelson, William (committee member), Heidelberg, Karla (committee member), Webb, Eric (committee member)

Creator Email sarah.hu@usc.edu,shu251@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-2945

Unique identifier UC11672487

Identifier etd-HuSarahKat-6165.pdf (filename),usctheses-c89-2945 (legacy record id)

Legacy Identifier etd-HuSarahKat-6165.pdf

Dmrecord 2945

Document Type Dissertation

Rights Hu, Sarah Kathryn

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA