University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

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

Paige Elizabeth Connell

A Dissertation Presented to the Faculty of the

USC GRADUATE SCHOOL

UNIVERSITY OF SOUTHERN CALIFORNIA

In Partial Fulfillment of the Requirements for the Degree

DOCTOR OF PHILOSOPHY

(BIOLOGICAL SCIENCES)

December 2017

© 2017 by Paige E. Connell
All Rights Reserved

! ii!

Approved by Advisory Committee:

David A. Caron (Chair)

Jed A. Fuhrman

Naomi M. Levine

Seth G. John

© 2017 by Paige E. Connell
All Rights Reserved

! iii!
TABLE&OF&CONTENTS&
TABLE&OF&CONTENTS.........................................................................................................................................iii&
ACKNOWLEDGEMENTS.......................................................................................................................................v&
&
&
DISSERTATION&ABSTRACT................................................................................................................................1&
&
CHAPTER&ONE.&PLANKTONIC&FOOD&WEB&STRUCTURE&AT&A&COASTAL&TIME:SERIES&SITE:&II.&
SPATIOTEMPORAL&VARIABILITY&OF&MICROBIAL&TROPHIC&ACTIVITIES..........................................3&
ABSTRACT!.....................................................................................................................................................................................!4!
1.1!INTRODUCTION!..................................................................................................................................................................!5!
1.2.!MATERIALS!AND!METHODS!.........................................................................................................................................!7!
1.3.!RESULTS!.............................................................................................................................................................................!16!
1.4.!DISCUSSION!......................................................................................................................................................................!33!
1.5.!CONCLUSION!....................................................................................................................................................................!44!
CHAPTER!ONE!LITERATURE!CITED!...............................................................................................................................!45!
!
CHAPTER&TWO.&PHYTOPLANKTON&AND&BACTERIAL&DYNAMICS&ON&THE&CHUKCHI&SEA&SHELF&
DURING&THE&SPRING:SUMMER&TRANSITION...........................................................................................50&
ABSTRACT!..................................................................................................................................................................................!51!
2.1!INTRODUCTION!...............................................................................................................................................................!52!
2.2!MATERIALS!AND!METHODS!.......................................................................................................................................!55!
2.3!RESULTS!..............................................................................................................................................................................!65!
2.4!DISCUSSION!.......................................................................................................................................................................!78!
2.5!CONCLUSION!.....................................................................................................................................................................!87!
CHAPTER!TWO!LITERATURE!CITED!.............................................................................................................................!89!
!
CHAPTER&THREE.&NUTRITIONAL&REQUIREMENTS&AND&PREY&CELL&CYCLES&DICTATE&
TEMPORAL&FEEDING&STRATEGIES&OF&HETEROTROPHIC&AND&MIXOTROPHIC&
NANOPLANKTON................................................................................................................................................93&
ABSTRACT:!.................................................................................................................................................................................!94!
3.1!INTRODUCTION!...............................................................................................................................................................!95!
3.2!METHODS!............................................................................................................................................................................!99!
3.3!RESULTS!...........................................................................................................................................................................!104!
3.4!DISCUSSION!....................................................................................................................................................................!109!
3.5!CONCLUSION!..................................................................................................................................................................!120!
CHAPTER!THREE!LITERATURE!CITED!......................................................................................................................!122!
! iv!
CHAPTER&FOUR.&MICROBIAL&CARBON&FLUXES&AT&A&COASTAL&OCEAN&SITE:&AN&INVERSE&
ECOSYSTEM&MODELING&ANALYSIS............................................................................................................127&
ABSTRACT!...............................................................................................................................................................................!128!
4.1.!INTRODUCTION!...........................................................................................................................................................!130!
4.2!METHODS!.........................................................................................................................................................................!134!
4.3!RESULTS!...........................................................................................................................................................................!147!
4.4!DISCUSSION!....................................................................................................................................................................!161!
4.5!CONCLUSION!..................................................................................................................................................................!176!
CHAPTER!FOUR!LITERATURE!CITED!.........................................................................................................................!178!
&
APPENDIX&A:&SUPPLEMENTAL&TABLES&AND&FIGURES........................................................................184&
APPENDIX&B:&SUPPLEMENTAL&TEXT.........................................................................................................200&
&
! v!
ACKNOWLEDGEMENTS&

The work put forward in this dissertation is not only a reflection of my time, thought, and
passion for knowledge, but reflects the intellect and encouragement of so many wonderful people
supporting me throughout this journey. First and foremost, I would like to thank my outstanding
advisor, Dr. David Caron. I met Dave on my first day of undergraduate studies at USC, sitting in
his BISC121 classroom and was enamored by the protozoa he studied. He provided me the
opportunity to work as an undergraduate researcher for 2.5 years in his lab and instilled a in me a
passion for scientific inquiry and for academic life. In allowing me to join his laboratory as a
graduate student, he has provided me with all of the intellectual, financial, and emotional support
needed to complete the challenging process of obtaining a Ph.D. Through Dave, I have met the
“best-of-the-best”, I have travelled to new places near and afar, and I have improved my skills as
a scientist. Most importantly, I have personally experienced the impact that a caring mentor can
have on an individual. As I proceed forward in my life and my career, it is my aspiration that I
can be an impactful and supportive to those I mentor as Dave was to me.
Next, I would like to thank my family and friends for being my stronghold throughout
this process. My parents, Susan Gabica and Ed Connell, and my brother, Nicholas Connell, have
always been there when I needed encouragement and have consistently flown to Los Angeles to
be there during my biggest triumphs or in times when I just needed some family fun and relax.
Additionally, I am ever thankful to my large support network of friends, both near and far,
especially Philip Hu, Sarah Hu, Markie Miller, Pamela Hu, Philip Hu Sr., Ruth Barber, Jayme
Smith, Lisa Mesrop, Alle Lie, Christopher Suffridge, Erin McParland, and Erica Seubert. You
mean the world to me!
! vi!
My gratitude also extends to the myriad individuals who have supported me
professionally during my time at USC Graduate School. I would like to thank my dissertation
committee members—Naomi Levine, Jed Fuhrman, and Seth John—for their insight and time, as
well as members of my qualifying exam committee: John Heidelberg, Doug Capone, and Doug
Hammond. I am grateful to Eric Webb and Karla Heidelberg as well for their mentorship and
kindness over the years. I would like to acknowledge the staff members of the MBBO
Administrative Offices, of the Wrigley Institute, and of the USC Graduate School for providing
me with resources or help when needed. Additionally, I would like to thank all of the members
of my laboratory who provided an equal dose of scientific insight and laughs throughout the
years (Sarah Hu, Jayme Smith, Alle Lie, Alyssa Gellene, Lisa Mesrop, Avery Tatters, Jay Liu,
Vikki Campbell, Erica Seubert, and Ramon Terrado).
My dissertation was made possible by extensive access to field operations and
collaborators. I would like to thank the captains and crews of the Harmony, Sundiver, Yellowfin,
USCG Healy, and R.V. Kilo Moana for supporting my field excursions. I would like to
acknowledge co-authors of my published work. I would also like to acknowledge collaborators
who have participated in data collection and analysis. Laura-Gómez Consarnau and the Fuhrman
Lab (USC) for providing bacteria and fieldwork support for Chapter 1. For Chapter 2, I would
like to acknowledge Kevin Arrigo (Stanford) for inviting me to participate in the SUBICE cruise,
Christine Michel and Guillaume Meisterhans for providing the bacterial production data
(Fisheries and Oceans Canada), Robert Pickart and Elizabeth Bonk for providing oxygen data
(WHOI), Dan and Laura Schuller for providing the nutrient data (SCRIPPS), Kate Lowry and
Gert van Dijken for providing the satellite-derived sea ice concentrations and imagery
(Stanford), Hannah Joy-Warren and Caroline Ferguson for providing the FRRf data (Stanford),
! vii!
and Andrea Niemi for helping with planning and mobilization for the expedition (Fisheries and
Oceans Canada). I would like to thank Sam Wilson (U. Hawaii) for serving as Chief Scientist for
the cruisework that supported Chapter 3, Benedetto Barone and Sam Wilson (U. Hawaii) for
providing the eddy altimetry map included in Figure 1 of this chapter, François Ribalet and
Ginger Armbrust (U. Washington) for collecting and providing the Seaflow data, and Frank
Aylward (U. Hawaii) for providing custom R-script for the harmonic regression analysis used in
Chapter 3. Lastly, I would like to thank Mike Stukel and Tom Kelly (FSU) for providing me the
base code utilized in Chapter 4 and supporting my education in linear inverse modeling, as well
as Jed Fuhrman (USC) and William Haskell (UCSB) for providing data in an accessible form for
the SPOT model.
This work was funded the University of Southern California, the National Science
Foundation [grant numbers: OCE-1136818 to DAC and PLR-1304563 to KRA], the Simons
Foundation (SCOPE: Simon’s Collaboration on Ocean Processes and Ecology; grant number
P49802 to DAC), and USC WiSE (Women in Science and Engineering). Additionally, we thank
the Natural Sciences and Engineering Research Council of Canada (NSERC) for providing
funding to C. Michel (Discovery Grant) and G. Meisterhans (Visiting Fellowship in Canadian
Laboratory Program), as well as Fisheries and Oceans Canada (International Governance
Strategy); this funding enabled the generation of the bacterial production data integrated in
Chapter 2.
! 1!
DISSERTATION&ABSTRACT&
&
Marine protists—single-celled microbial eukaryotes—are an extraordinarily diverse
group of organisms that span a myriad of sizes, forms, and functions. These abundant organisms
fulfill a wide array of ecological roles and are critical to the global cycling of key elements (e.g.
C,N,P,S). Complex physical, chemical, and biological interactions structure these communities,
which in turn sequester atmospheric carbon dioxide into biomass that is destined for higher
trophic levels or export to the deep ocean. Understanding the abundances and trophic activities
of microbes is therefore essential to accurately modeling carbon cycling. Significant attention
has focused on the environmental factors controlling the production of phytoplankton and
bacteria; yet, relatively little is known about the diversity and overall impact of their consumers
(primarily protistan grazers), and how environmental change might affect these features of ocean
biology.
This dissertation investigated how the growth and mortality rates of marine microbes
vary on spatial and temporal scales to better characterize protistan grazing impact in the ocean
environment. Spatial variability in microbial rate processes was examined at broad geographic
scales (between oceanic regimes) as well along environmental gradients. Specifically, fieldwork
was conducted in the Southern California Bight, the Chukchi Sea, and the North Pacific
Subtropical Gyre, areas that differ greatly in physicochemical properties and consequently may
respond differently to a changing climate. Similarly, temporal variability was examined at long
time scales (seasonally and annually) and at very rapid time scales (hours). The central
motivation for this research was to generate microbial mortality data that can be integrated into
ocean ecosystem models, which aim to accurately predict current and future conditions of our
changing ocean. The first three chapters of my dissertation examine microbial rate processes in
! 2!
situ, while the final chapter of my dissertation implements a statistical modeling technique to
integrate the data gathered here with complementary field measurements to advance our
understanding of microbial carbon fluxes in the coastal ocean ecosystem.
Large phytoplankton (microbial eukaryotes) were the dominant source of prey carbon in
coastal regions (specifically, the Port of Los Angeles (Ch. 1) and the Chukchi Sea (Ch. 2)), while
picoplankton were the dominant source of prey carbon in oceanic regions (the SPOT station and
Catalina Island (Ch. 1) and the North Pacific Subtropical Gyre (Ch. 3)). Looking at temporal
variability, it was found that protistan grazers were in tight synchrony with the division cycles of
their prey in the open ocean (Ch. 3). Whether this relationship persists in nutrient-replete, coastal
regimes remains unclear and is an interesting topic for future study. Finally, the heterotrophic
bacteria were consistently found to be an important food source for protists in all regions studied,
a trophic interaction that is often not implicitly included in biogeochemical models. An inverse
modeling analysis (Ch. 4) confirmed the importance of characterizing the heterotrophic bacterial
assemblage, and the processes that support or remove it, when attempting to model the fate of
primary production in the marine environment. This work adds to the litany of knowledge on
microbial rate processes in marine environments by conducting some of the first studies to (1)
investigate spatiotemporal variability in protistan grazing pressure, and (2) to conduct
contemporaneous measurements of bacterial and phytoplankton carbon consumption in marine
environments.
! 3!
CHAPTER&ONE.&PLANKTONIC&FOOD&WEB&STRUCTURE&AT&A&COASTAL&TIME:SERIES&SITE:&II.&
SPATIOTEMPORAL&VARIABILITY&OF&MICROBIAL&TROPHIC&ACTIVITIES&&

!
By Paige E. Connell

Coauthors: Victoria Campbell, Alyssa G. Gellene, Sarah K. Hu, and David A. Caron

Published as

Connell PE, Campbell V, Gellene AG, Hu SK, Caron DA (2017) Planktonic food web structure
at a coastal time-series site: II. Spatiotemporal variability of microbial trophic activities. Deep
Sea Res I 121:210-223

! 4!
Abstract
The!grazing!activities!of!phagotrophic!protists!on!various!microbial!assemblages!play!key!
roles!in!determining!the!amount!of!carbon!available!for!higher!trophic!levels!and!for!export!
out!of!the!photic!zone.!However,!comparisons!of!the!proportion!of!carbon!consumed!from!
the!phytoplankton!(cyanobacteria!+!photosynthetic!eukaryotes)!and!heterotrophic!bacteria!
(bacteria!+!archaea,!excluding!cyanobacteria)!are!rare.!In!this!study,!microbial!community!
composition,!phytoplankton!growth!and!mortality!rates!(total!chlorophyll!a,#Synechococcus,!
Prochlorococcus,!and!photosynthetic!picoeukaryotes),!and!bacterial!mortality!rates!were!
measured!seasonally!from!2012\2014!in!the!surface!waters!of!three!environmentally!
distinct!sites!in!the!San!Pedro!Channel,!off!the!coast!of!southern!CA,!USA.!Higher!nutrient!
concentrations!at!the!nearshore!site!supported!community!standing!stocks!that!were!1.3\
4.5x!those!found!offshore,!yet!average!growth!and!grazing!rates!of!the!phytoplankton!and!
bacterial!assemblages!were!generally!similar!between!sites!and!across!seasons.!Thus,!the!
amount!of!carbon!consumed!by!the!grazer!assemblage!was!largely!dictated!by!prey!
standing!stocks.!Heterotrophic!bacteria!constituted!an!important!source!of!carbon!for!
microbial!consumers,!particularly!at!the!two!offshore!sites!where!bacterial!carbon!
consumed!was!roughly!equivalent!to!the!amount!of!phytoplankton!carbon!consumed.!
Carbon!removal!by!grazers!at!the!nearshore!station!was!predominantly!from!the!diatoms,!
which!were!the!primary!component!of!the!photosynthetic!community!at!that!site.!This!
study!highlights!the!significant!contribution!of!protistan\bacterial!trophic!interactions!to!
planktonic!food!webs!and!provides!unique!community!composition!and!turnover!data!to!
inform!biogeochemical!models.!!

! 5!
1.1 Introduction
Marine microbes (bacteria, archaea, and microbial eukaryotes) are a functionally diverse
group of single-celled organisms that, together with viruses, play critical roles in the global
cycling of key elements (C, N, P, and S)(Karner et al. 2001, Suttle 2005, Sherr et al. 2007, Caron
et al. 2012, Fuhrman et al. 2015). Oceanic primary production is overwhelmingly dominated by
planktonic microbes (Field et al. 1998), and much of the organic carbon transferred to higher
trophic levels or exported out of the photic zone by these species is directly mediated by the
activities of microbial consumers (Sherr & Sherr 1994, Sherr & Sherr 2002, Calbet & Landry
2004, Schmoker et al. 2013). Yet, accurate descriptions of the abundances, biomasses, and
trophic activities of microbial assemblages are incomplete and confounded by the extremely high
biodiversity and many nutritional and life strategies of marine microbes (Caron et al. 2012,
Worden et al. 2015, Caron 2016). A better understanding of these assemblages and their
activities are necessary as climate change may restructure the community composition, primary
productivity, trophic interactions, and export efficiency of microbial communities (Doney et al.
2012, McMahon et al. 2015).
The considerable amount of phytoplankton or bacterial carbon consumed by protistan grazers
has profound implications for the biogeochemistry of an ecosystem. Food web efficiency
decreases as the number of trophic interactions increases due to respiratory losses of carbon.
Thus, food webs in which large amounts of primary productivity are routed through the
heterotrophic bacterial assemblage (“the microbial loop”; Azam et al. 1983) tend to remineralize
a substantial portion of the organic matter and are less efficient than classical phytoplankton-
macrozooplankton trophic interactions at moving energy and carbon up into higher trophic
levels. Nonetheless, the microbial loop is an important mechanism by which some non-living
! 6!
particulate and dissolved organic carbon is salvaged (Strom 2000). The proportion of carbon
flowing directly from phytoplankton into herbivorous zooplankton, relative to that passing
through non-living organic matter into the heterotrophic bacteria and their consumers, therefore
affects overall food web efficiency in the plankton. Direct comparisons of these key trophic
connections have rarely been conducted.
We investigated the spatial and temporal variability of microbial standing stocks, community
composition, and trophic interactions at a coastal ocean site in the San Pedro Channel. The
region is the site of a long-term oceanographic study of plankton dynamics (the San Pedro Ocean
Time-series, SPOT; https://dornsife.usc.edu/spot/). The SPOT station is located equidistant
between the Port of Los Angeles to the northeast and Santa Catalina Island to the southwest (~45
km offshore). The Port of Los Angeles receives extensive runoff from metropolitan and
industrial land development from a population of ~18 million residents. Waters adjacent to Santa
Catalina Island are more oceanic and pristine in nature than near the port, lack the substantial
anthropogenic impact of a large metropolitan area, and are the site of a marine protected area.
Episodic wind-driven upwelling and rainfall events and mild seasonality drive changes in
primary productivity in the San Pedro Channel (Nezlin et al. 2012). These changes are
accompanied by distinct seasonal shifts in prokaryotic (Chow et al. 2013, Cram et al. 2014) and
eukaryotic (Countway et al. 2010, Kim et al. 2014, Hu et al. 2016) community composition. The
Channel provides an environmentally diverse landscape with recurrent seasonal patterns of
microbial community productivity and composition in which to study seasonal and spatial
drivers of microbial community structure and function.
In this study, microbial community abundances and rates of phytoplankton growth,
phytoplankton mortality, and bacterial mortality were measured seasonally in the surface waters
! 7!
of the Port of Los Angeles, the SPOT station, and near Catalina Island. Microbial community
composition and standing stocks varied spatially and temporally, however, rates of growth and
grazing mortality (d
-1
) were similar among sites and seasons. Thus, differences in the amount of
carbon ingested from the various prey assemblages at the different sites and seasons were largely
related to differences in standing stock. Mean daily phytoplankton carbon ingested by grazers
ranged from 0.18 to 18.5x the mean bacterial carbon consumed, with high values of carbon flow
through herbivorous microconsumers attributable to increased diatom abundances onshore and
during the spring season. The mean daily carbon consumption of the heterotrophic bacteria
(bacteria + archaea, excluding cyanobacteria) and photosynthetic picoplankton (coccoid
cynanobacteria + photosynthetic picoeukaryotes) by grazers was similar in magnitude at all sites.
This study highlights the importance of bacteria as a food source for higher trophic levels, a
trophic interaction that is often not explicitly included in ecosystem models but can be as
important as the link between phytoplankton and their grazers.

1.2. Materials and Methods
1.2.1#Study#site#characterization#and#sample#collection#
!
Seasonal microbial standing stocks and trophic dynamics were examined across a range
of anthropogenically affected sites spanning the San Pedro Channel: the heavily impacted Port of
Los Angeles (Port of LA, ~3 km offshore within the breakwater, 33°42.75’N, 118°15.55’W), the
coastal ocean San Pedro Ocean Time-series station (SPOT, ~17 km offshore, 33°33’N,
118°24’W), and a marine protected area off Santa Catalina Island (Catalina, ~45 km offshore,
33°27.17’N, 118°28.51’W) (Fig. 1.1). The approximate water column depths at the Port of LA,

! 8!

Figure 1.1. Study location in the San Pedro Basin, off southern California, USA, in the
eastern North Pacific. The three sites were located in the Port of Los Angeles (Port of LA,
33°42.75’N, 118°15.55’W), central San Pedro Channel at the San Pedro Ocean Time-series
station (SPOT, 33°33’N, 118°24’W), and adjacent to Santa Catalina Island (Catalina,
33°27.17 N, 118°28.51’W).

! 9!
SPOT, and Catalina stations were 15 m, 900 m, and 140 m, respectively. A thorough discussion
of the hydrography and vertical structure at the SPOT station can be found in Caron et al. (2016).
Water collection was conducted quarterly to assess microbial standing stocks and
measure seasonal differences in microbial rate processes. Sampling occurred during the winter
(January), spring (March-April), summer (July), and fall (October) of 2012-2014. Surface water
(0 to 1 m depth) was collected by completely submerging 23 L, acid-cleaned (5% HCl),
polycarbonate carboys, inverted with the spigot open, and allowing water to gently enter through
the mouth of the carboy. Once full, the spigot was closed, the lid replaced underwater, and the
carboy was stored in the dark onboard the ship until processing within 4 h of sampling. The
collection method was carried out to reduce bubbling that damages delicate microzooplankton.
Physical and chemical parameters were measured at each site to characterize differences
across the San Pedro Channel. Temperature was measured during each cruise using a YSI 63
handheld meter (Yellow Springs Instruments, Yellow Springs, Ohio) or profiling natural
fluorometer (PNF) (Kiefer et al. 1989). Salinity was measured using a YSI 63 handheld meter or
refractometer. Samples for inorganic nutrient measurements were collected at each site by
adding 0.2 µm-filtered seawater to 20 mL plastic scintillation vials that were pre-washed with
5% HCl and rinsed three times with 0.2 µm-filtered seawater. Nutrient samples were frozen at -
20°C until processed for ammonium, phosphate, nitrite (all 0.1 µM limit of detection), nitrate
plus nitrite (0.2 µM limit of detection), and silicate (1.0 µM limit of detection) at the Marine
Sciences Institute Analytical Lab at UCSB (http://www.msi.ucsb.edu/services/analytical-lab).
#
1.2.2#Community#composition#and#biomass#
The abundances of various components of the picophytoplankton (0.2-2.0 µm in size)—
specifically Synechococcus, Prochlorococcus, and photosynthetic picoeukaryotes—were
! 10!
determined by flow cytometry. Triplicate samples were collected at the beginning of the
incubation for each site, passed through a 20 µm filter, preserved with 1% formalin (final
concentration), and stored at -80
o
C until flow cytometric analysis. Flow cytometry was
performed using a FACSCalibur flow cytometer (Becton Dickinson, San Jose, California). The
autofluoresence of photosynthetic pigments and forward scatter were used to distinguish these
three groups of picoplankton. Average cell abundances of these plankton categories were
converted to biomasses using carbon conversion factors of 200 fg C cell
-1
for Synechococcus
(Caron et al. 1995), 90 fg C cell
-1
for Prochlorococcus (Casey et al. 2013, Martiny et al. 2016),
and 183 fg C µm
-3
for photosynthetic picoeukaryotes (assuming an average picoeukaryote radius
of 1 µm)(Caron et al. 1995). Bacterial abundances (bacteria + archaea) were measured by flow
cytometry using a standard staining procedure (SYTO 13; ThermoFisher Scientific S7575) (del
Giorgio et al. 1996). Bacterial abundance was converted to carbon biomass assuming a carbon
conversion factor of 15 fg C cell
-1
, which is a conservative estimate for bacteria characteristic of
coastal regions (Caron et al. 1995, Fukuda et al. 1998, Kawasaki et al. 2011, Buitenhuis et al.
2012).
Nanoplankton abundances (photosynthetic and heterotrophic protists, 2-20 µm in size)
were determined using epifluorescence microscopy on samples collected from the initial,
unfiltered seawater for each experiment. Aliquots of seawater (150 mL) were preserved in 1%
formalin (final concentration) for each site. Preserved aliquots (15 mL for Port of LA; 30 mL for
SPOT and Catalina) were filtered down to ~1 mL onto 25 mm, 0.2 µm blackened polycarbonate
filters and stained with 50 µL of a 1 mg mL
-1
working solution of 4’,6’-diamidino-2-
phenylindole (DAPI; Sigma D9542) for 5-10 minutes (Sherr et al. 1993). Samples were then
filtered and rinsed, the filters placed onto glass slides, a drop of immersion oil placed on each
! 11!
filter, and a coverslip added and sealed with clear nail polish. Slides were prepared in triplicate
and stored at -20°C until analyzed by epifluorescence microscopy. Phototrophic and
heterotrophic nanoplankton were differentiated by the presence or absence of chlorophyll
autofluorescence when viewed with blue-light excitation. Nanoplankton biomasses were
calculated from cell abundances using a carbon conversion factor of 183 fg C µm
-3
(Caron et al.
1995), with an assumed average nanoplankton radius of 1.5 µm. Only slides prepared for March
2012, July 2012 (Port of LA only), October 2013, and all months in 2014 were included in
subsequent analyses because some slides were unsuitable for counting. Nanoplankton biomass
was excluded from the total phytoplankton standing stock value when calculating carbon flow to
higher trophic levels (Fig. 1.7) in order to facilitate the comparison of all sampling dates.
Microplankton abundances were enumerated from formalin-preserved, 200 mL samples
(final concentration 1%) collected from the initial, unfiltered seawater for each experiment.
Aliquots of preserved samples (25-100 mL) were settled in Utermöhl chambers for 24 to 48 h
and counted using an inverted light microscope at 400x magnification (Utermöhl 1958). Diatom,
dinoflagellate, and ciliate abundances were recorded, and carbon biomasses were calculated
using a carbon conversion factor of 138 pg C cell
-1
. Diatom cells ranging from 1000 to 4000 µm
3

in size were the most common in our region (see section 1.3.3.1), which corresponds to cellular
carbon contents of 78 to 240 pg C cell
-1
according to the equations of Menden-Deuer and Lessard
(2000). The selected value (138 pg C cell
-1
) was determined from the phototrophic
microplankton data presented in Caron et al. (1995) and represents an average carbon content
given the expected range of values in our region. We used a constant carbon conversion factor
for all microplankton groups because diatoms and dinoflagellates < 3000 µm
3
in size

have
comparable carbon content (Menden-Deuer & Lessard 2000), Photosynthetic and heterotrophic
! 12!
dinoflagellates were not distinguished in this study, and therefore total dinoflagellate biomass
was split evenly between these nutritional modes when assessing the composition of the
phytoplankton and grazer communities.
Carbon conversion factors are poorly constrained in the literature, but provide a useful
tool to assess microbial standing stocks when the empirical determination of the carbon contents
of each microbial assemblage is not feasible. A detailed discussion of the carbon biomass
conversion factors selected for our study region can be found in Caron et al. (in revision).
Cellular abundances have also been provided in Table A1 to facilitate an assessment of how the
choice of different conversion factors may impact the results of this study.

1.2.3#Phytoplankton#growth#and#grazing#experiments#
!
1.2.3.1. Experimental set-up
A modified dilution method was used to determine the growth and grazing rates of the
total phytoplankton community as well as three ecologically distinct groups of picoplanktonic
phototrophs: Synechococcus, Prochlorococcus, and photosynthetic picoeukaryotes (Landry &
Hassett 1982, Landry et al. 1995). The dilution method allows simultaneous measurement of
phytoplankton growth (µ) and mortality (m) through sequential dilution of experimental water
with 0.2 µm diluent. Whole, unfiltered seawater (WSW) from two 23 L carboys collected at each
site was joined into a 50 L carboy to ensure homogeneity between treatments. Diluent was
prepared from water collected at each site by filtering WSW through an acid-cleaned and rinsed
Pall 0.2 µm Acropak 1550 Capsule Filter with Supor Membrane. A five-point dilution series
(100%, 80%, 60%, 40% and 20% WSW) was established in acid-washed, polycarbonate, 2.3 L
bottles for each site. All treatments were prepared and incubated in triplicate, with care taken to
! 13!
reduce bubbling that may harm delicate microzooplankton. Bottles in the dilution series were
nutrient-enriched, with 1 mL of nutrient stock (final incubation concentrations of 10 µM NaNO
3
,
1 µM NH
4
Cl, 0.7 µM NaP
2
PO
4
·H
2
O) added to each bottle to account for the possibility of
nutrient-limited phytoplankton growth in the more-dilute treatments (Landry et al. 1995). An
unenriched treatment of 100% WSW bottles (in triplicate) was prepared and compared to the
enriched WSW treatment for each experiment to assess the impact of nutrient addition on
phytoplankton growth rates. Also, an unenriched, <200 µm-filtered treatment was prepared to
investigate the grazing impact of the micro- and nanozooplankton in the absence of
mesozooplankton. Bottles were incubated in mesh bags composed of neutral density screening
(mimicking 50% incident light) and incubated in situ for 24 h.

1.2.3.2 Sample collection and analysis
Chlorophyll a (a proxy of total phytoplankton biomass) and cell abundances for specific
picoplankton were measured initially and at the end of the incubations to calculate growth and
grazing rates of the entire photosynthetic community and the three picoplankton groups
according to the modified dilution method (Landry et al. 1995). Triplicate samples for
chlorophyll a and picoplankton abundances were analyzed from the initial WSW, <200 µm
filtrate, and <0.2 µm filtrate. Duplicate chlorophyll a samples and singular flow cytometry
samples were analyzed from triplicate bottles within each treatment at the end of the incubation.
Chlorophyll a samples were prepared by filtering 50-500 mL of WSW onto 25 mm GF/F filters
and stored at -20°C until analysis within the week. Filters were extracted with 4 mL of 100%
acetone at -20°C overnight in the dark, and processed on a calibrated Trilogy Laboratory
Fluorometer (Turner Designs, San Jose, CA) using the non-acidification method (Welschmeyer
! 14!
1994). Samples collected and preserved for flow cytometry were processed as described in
section 2.2.
Net growth rates (nutrient-enriched growth (µ
n
); y-intercept of the regression) and
mortality rates (m; slope of the regression) of the total phytoplankton (based on chlorophyll a) or
specific assemblages (based on flow cytometric counts) were determined using Model I linear
regressions of plots of apparent growth rate (y-axis) versus dilution factor (x-axis) assuming
exponential growth of the phytoplankton (Landry & Hassett 1982). Intrinsic growth rates (µ
0
) of
the unenriched phytoplankton community were determined from growth in the unenriched
treatment and the mortality rate (Landry et al. 1995).
The daily percentage of total phytoplankton standing stock (%SS) consumed by grazers
was calculated according to the following equations,
%SS = G * (100/C
0
)
G = m * C
m
C
m
= C
0
[e
(µ0-m)t
– 1]/(µ
0
– m)t
where µ
0
= the unenriched growth rate of the phytoplankton (d
-1
), m = the mortality rate of the
phytoplankton (d
-1
), t = length of the incubation (d), C
0
= the initial Chlorophyll a concentration
(or picophytoplankton abundance), C
m
= the mean chlorophyll a concentration (or
picophytoplankton abundance) during the incubation, and G = the grazing impact of the
consumers (Landry et al. 2000, Landry & Calbet 2004).

1.2.4#Bacterivory#by#FLB#disappearance#
!
Fluorescently-labeled bacteria (FLB) were prepared from a monoculture of Dokdonia
donghaensis according to standard protocol (Sherr et al. 1987, Caron 2001). Briefly, D.
! 15!
donghaensis was cultured in Zobell medium, harvested by centrifugation, and resuspended and
incubated in 0.2 µm filtered seawater to starve the bacteria and induce shrinkage to better mimic
bacterial size in natural assemblages (rod-shaped, average length = 1 µm). Bacteria were then
stained with 5-(4,6-dichlorotriazin-2-yl) aminofluorescein (DTAF), heat-killed, rinsed three
times, aliquoted, and stored at -80°C until the time of the experiment. FLB aliquots for the
duration of the study were prepared in a single batch to ensure homogeneity across experiments.
FLB incubations were conducted in conjunction with the dilution experiments at each site
beginning in July 2012. Two sets of triplicate, 2.3 L, polycarbonate bottles were completely
filled with experimental water; one set received WSW to assess FLB disappearance related to
grazing, while the other set received 0.2 µm filtered seawater to serve as a control for losses
unrelated to grazing. The FLB stock was vortexed, filtered through a 3.0 µm polycarbonate filter
to remove clumps, and added to each bottle at ~10% of the natural bacterial abundance (average
T
0
FLB concentration = 1.52 x 10
5
cells mL
-1
; average natural bacterial abundance = 1.6 x 10
6

cells mL
-1
). Samples (2 mL) were removed from each bottle immediately after adding FLB and
at the end of the incubations, preserved with 1% formalin (final concentration), flash frozen in
liquid nitrogen, and stored -80
o
C until counted using flow cytometry. Changes in FLB
abundances in the WSW treatment were used to calculate removal of the bacterial community
due to grazing (taking non-grazing changes into account from the controls) according to the
following equation,
g = ln(F
t
/F
0
)*(-1/t)
where g = bacterial grazing mortality (d
-1
), F
0
and F
t
= the number of FLB at the beginning and
end of the incubation, respectively, and t = length of the incubation (d) (Marrasé 1992).
#
#
! 16!
1.2.5#Statistical#analysis#
!
All statistical analysis was carried out in R (R Core Team 2014), with a critical value of
0.05 denoting significance. Linear regression analysis (Model I) was used to relate dilution level
to phytoplankton growth rate, enabling the determination of m, µ
n
, and µ
0
for the phytoplankton
groups. Non-linear responses due to grazing saturation were not observed, even at the highest
chlorophyll a concentrations (Fig. A1). Differences in the phytoplankton growth rates between
the nutrient-enriched and unenriched treatments, as well as between the WSW and <200 µm
treatments were determined with a Welch Two-Sample T-Test. One-way analysis of variance
(ANOVA) paired with Tukey’s Honest Significant Difference Test was used to determine
significant differences in mean values between sites and seasons. A median percentile bootstrap,
a robust option for datasets containing tied-values, was employed to assess significant
differences in the median values between sites and seasons. Spearman’s rank order correlation
analysis was applied to determine significant relationships between environmental variables and
the population abundances, growth rates, and mortality rates of the prey populations studied.

1.3. Results
1.3.1!Environmental#and#biological#context#at#the#three#study#sites#
!
Temperatures ranged from 13.1 to 21.6
o
C throughout the annual cycle, with the Port of
LA values averaging ~1
o
C lower than those measured at the SPOT station and near Catalina on
all sampling dates (Table 1.1). Salinity was relatively constant for all sampling dates (!!(mean)
! 17!
Table 1. Temperature (T
0
temp), chlorophyll a (T
0
chl a), and nutrient concentrations (NH
3
– ammonia, NO
3
-
– nitrate, NO
2
-
–
nitrite, PO
4
3-
– phosphate, SiO
4
4-
– silicate) at the beginning of the experiments, and growth and mortality rates for the total
phytoplankton community (estimated from chlorophyll a) for each study date and site. Experiments were conducted using
surface water (0 to 1 m depth) and incubated in situ. Mortality rates (m) and nutrient-enriched growth rates (µ
n
) were
calculated using Model I Linear Regressions (significance denoted by asterisks next to the mortality rate values).
Unenriched growth rates (µ
0
) were calculated from µ
n
values to determine intrinsic population growth rates.

Table on following page.

! 18!
Experiment Temp (°C) T 0 chl a (µg L
-1
) NH 3 (µM) NO 3
-
(µM) NO 2
-
(µM) SiO 4
4-
(µM) PO 4
3-
(µM)
m (d
-1
) µ n (d
-1
) µ 0 (d
-1
)
Port of Los Angeles

13-Mar-12 13.1 8.03 1.39 7.23 0.44 10.2 0.68 0.54** 0.74
0.82
15-Jul-12 16.0 4.67 0.6 0.39 0.11 0 0.22 1.31** 1.74
1.21
19-Oct-12 20.0 2.06 0.71 0.14 0.1 3.36 0.24 0.31** 1.10
0.81
9-Jan-13 14.5 2.86 3.08 1.95 0.19 2.45 0.34 0.24** 0.38
0.42
24-Apr-13 14.5 12.70 0.78 0 0 1.68 0.2 0.59** 0.65
0.38
15-Jul-13 18.0 5.00 0.74 0.44 0.23 8.49 0.25 0.59** 0.92
0.32
16-Oct-13 17.2 4.53 0.53 0.65 0.15 7.39 0.16 1.11** 0.89
0.82
7-Jan-14 15.4 2.20 2.7 2.62 0.29 6.02 0.23 ns 0.29
0.32
8-Apr-14 13.2 15.04 0.1 2.11 0.2 2.46 0.14 0.87** 1.00
0.62
14-Jul-14 18.4 3.23 0.14 0.18 0.1 3.26 0.25 0.87** 1.92
0.44
6-Oct-14 20.4 2.81 0.18 0.37 0.11 3.14 0.24 0.20** 0.84
0.01
San Pedro Ocean Time Series

13-Mar-12 13.6 5.86 0.02 0.82 0.13 1.4 0.35 0.74** 0.75
0.72
15-Jul-12 18.0 0.45 0.87 0.38 0 1.22 0.23 -0.76** 0.58
-1.36
19-Oct-12 20.6 0.24 0.64 0.15 0 1.46 0 ns 0.63
0.27
9-Jan-13 15.2 1.29 0.58 0.23 0.11 0 0.12 0.21* 1.31
0.07
24-Apr-13 15.7 0.63 0 0 0 1.8 0.22 0.34** 0.94
0.17
15-Jul-13 19.9 0.29 0 0 0 2.75 0.1 0.93* 1.04
0.4
16-Oct-13 19.6 0.24 0 0 0 0 0.11 0.21** 1.03
0.59
7-Jan-14 16.7 0.22 0.3 0.08 0.12 1.73 0.17 ns 0.27
0.11
8-Apr-14 15.2 0.79 0 0 0 0 0.13 0.48** 1.87
0.51
14-Jul-14 21.4 0.34 0 0 0 1.09 0.16 0.22* 0.41
-0.31
6-Oct-14 21.6 0.24 0 0 0 1.36 0.16 ns 1.12
0.78
Big Fisherman's Cove

13-Mar-12 13.6 1.28 0.87 0.13 0.12 0 0.21 0.76** 2.20
1.53
15-Jul-12 19.0 0.78 0.12 0.03 0.21 1.06 0.17 ns 2.09
-0.45
19-Oct-12 20.6 0.20 1.16 0.24 0 1.66 0.22 ns 0.88
0.06
9-Jan-13 14.1 2.73 0.42 0.24 0.14 0 0.13 0.15* 0.53
-0.03
24-Apr-13 15.9 1.02 0.22 0.15 0.11 0 0.12 0.20** 0.77
-0.01
15-Jul-13 20.0 0.38 0.56 0 0 2.32 0.12 0.47** 1.13
0.36
16-Oct-13 19.4 0.68 0.34 0.2 0 0 0.11 0.58** 1.20
0.44
7-Jan-14 17.5 0.64 0.54 0.4 0 3.33 0.14 ns 0.42
0.18
8-Apr-14 15.4 0.37 0.17 0 0 0 0.12 0.38** 2.17
0.32
14-Jul-14 21.0 0.34 0.33 0.38 0 1.13 0.18 0.73** 1.64
0.73
6-Oct-14 21.5 0.24 0.52 1.23 0 0 0.17 0.37** 1.10
0.77
Significance level of the Model I regression: ns = not significant, *= significant at p ≤ 0.05, **= significant at p ≤ 0.01

! 19!
= 32.8 ppt), with no onshore-offshore salinity gradient observed (data not shown). Chlorophyll a
values were consistently highest at Port of LA on all dates (range 2.06 – 15.0 µg L
-1
). The ranges
of chlorophyll a values at the SPOT station and Catalina were 0.24 – 5.86 µg L
-1
and 0.20-2.73
µg L
-1
, respectively, and were <1 µg L
-1
on the majority of sampling dates (Table 1.1). Mean
concentrations of silicate, ammonium, nitrate, nitrite, and phosphate were significantly higher at
Port of LA than at the SPOT station and mean silicate, nitrite, and phosphate were significantly
higher at Port of LA than Catalina (p < 0.05, Table 1.1).
Port of LA had consistently higher standing stocks of microbial biomass than the SPOT
station or Catalina, and seasonal maxima at the latter sites were less than the seasonal minimum
at Port of LA (maxima were 65 and 61 µg C L
-1
at the SPOT station and Catalina, respectively,
versus a minimum of 73 µg C L
-1
at Port of LA; Fig. 1.2). All sites experienced seasonal maxima
in biomass during the spring, but the average value at Port of LA during spring was
approximately 4-fold greater than the average winter value at that site and >4-fold greater than
all seasons at the other sites, primarily due to diatom blooms in the harbor (Fig. 1.2). In
comparison, the seasonal ranges of total microbial biomass at the SPOT station and Catalina
varied by less than 2-fold (range = 34-65 µg C L
-1
). Seasonal minima in standing stocks at Port
of LA and Catalina occurred during winter, whereas the SPOT station exhibited seasonal minima
during summer and fall.
Heterotrophic bacteria (bacteria + archaea, excluding cyanobacteria) were a major
contributor to total community biomass at all sites, however bacterial biomass at Port of LA was
consistently greater (seasonal mean range = 25.9-40.3 µg C L
-1
) than that observed at the SPOT
station and Catalina (range = 14.3-26.2 µg C L
-1
and 15.6-27.4 µg C L
-1
, respectively; Fig. 1.2).
Mean bacterial biomass at Port of LA exhibited a winter minimum and fall maximum. Mean

! 20!

Figure 1.2: Average seasonal biomasses of the microbial assemblages (µg C L
-1
) at the
Port of Los Angeles (Port of LA), San Pedro Ocean Time-series (SPOT), and Santa
Catalina Island (Catalina) sampling sites. Carbon conversions from microbial
abundances are described in the Materials and Methods. Microbial assemblages include:
BACT (bacteria + archaea), SYN (Synechococcus), PRO (Prochlorococcus), PEUK
(photosynthetic picoeukaryotes), PNANO (photosynthetic/mixotrophic nanoplankton),
HNANO (heterotrophic nanoplankton), DIATOM (diatoms), DINO (dinoflagellates), and
CILIATE (loricate/aloricate ciliates).

! 21!
bacterial biomass also fluctuated seasonally at the SPOT station and Catalina, but exhibited
winter and summer minima and spring and fall maxima. Significant positive correlations were
found between chlorophyll concentration and bacterial abundance (rho = 0.492) and between
microzooplankton abundance and bacterial abundance (rho = 0.438) when all sites and sampling
dates were considered (p < 0.05, Fig. 1.3).
Diatoms were a significant proportion of the photosynthetic biomass at all sites during
the spring (70-90%, Fig. 1.4a). Diatom abundance in the San Pedro Channel was positively
correlated with phosphate concentration (rho = 0.343), chlorophyll concentration (rho = 0.729),
and the abundance of microzooplankton grazers (rho = 0.563), and negatively correlated with
water temperature (rho = -0.583; p < 0.05, Fig. 1.3). The taxonomic composition of the diatom
community was generally conserved across sites, with Chaetoceros spp. and Pseudo-nitzchia
spp. consistently dominant. Leptocylindrus, Eucampia, Skeletonema, Cylindrotheca, and
Guinardia were also common genera found during spring. Diatoms continued to be major
components of the phytoplankton community during winter and summer at Port of LA, while
photosynthetic pico- and nanoeukaryotes were prominent at that site in the fall (combined ~60%,
Fig. 1.4a). Photosynthetic picoeukaryote abundance was positively correlated with nitrate+nitrite
concentration (rho = 0.362), silicate concentration (rho = 0.597), and chlorophyll concentration
(rho = 0.386; p < 0.05, Fig. 1.3). Microbial eukaryotes were the primary component of the
photosynthetic community at Port of LA during all seasons.
Cyanobacteria were important constituents of phytoplankton biomass at the SPOT and
Catalina stations. Synechococcus and Prochlorococcus collectively comprised ~50% of the
phytoplankton biomass in the summer and fall months at those sites, and 30-40% of the
phytoplankton biomass during the winter (Fig. 1.4a). Synechococcus abundance positively
! 22!

Figure 1.3. Spearman’s rank correlation coefficients (rho) between environmental factors
and population abundances, unenriched growth rates, and grazing mortality rates of the
total phytoplankton and major picoplankton groups. Significant correlations are marked
by a * (p ≤ 0.05). Abbreviations represent: BACT (heterotrophic bacteria + archaea), SYN
(Synechococcus), PRO (Prochlorococcus), PEUK (photosynthetic picoeukaryotes),
DIATOM (diatoms), HNANO (heterotrophic nanoplankton), MICROZ (microzooplankton:
ciliates + dinoflagellates), PHYTO (total phytoplankton), and T
0
Chl (chlorophyll
concentration at the beginning of the experiments).
!
! 23!

Figure 1.4. Average seasonal percent contribution of major taxonomic groups to (a) total
phytoplankton biomass, and (b) total heterotrophic grazer biomass at the Port of Los
Angeles (Port of LA), San Pedro Ocean Time-series (SPOT), and Santa Catalina Island
(Catalina) sampling sites. Phytoplankton assemblages include: SYN (Synechococcus),
PRO (Prochlorococcus), PEUK (photosynthetic picoeukaryotes), PNANO
(photosynthetic/mixotrophic nanoplankton), DIATOM (diatoms), and DINO
(dinoflagellates). Grazer assemblages include: HNANO (heterotrophic nanoplankton),
DINO (dinoflagellates), and CILIATE (loricate/aloricate ciliates).

!
!
a
b
! 24!
correlated with temperature (rho = 0.727) and chlorophyll concentration (rho = 0.524), while
Prochlorococcus abundance positively correlated with temperature (rho = 0.548) and negatively
correlated with chlorophyll concentration (rho = -0.540) and microzooplankton abundance (rho =
-0.465; p < 0.05, Fig. 1.3). Synechococcus biomass surpassed Prochlorococcus biomass at all
sites. Prochlorococcus was nearly absent from the phytoplankton community at all sites during
the spring and during all seasons at Port of LA. The photosynthetic pico- and nanoeukaryotes
attained maximal contributions to phytoplankton biomass at the SPOT and Catalina stations
during the fall (40-45%), when diatom abundances were minimal (Figs. 1.2 & 1.4a).
Dinoflagellates were minor contributors to phytoplankton biomass at all sites and
seasons, never exceeding 5% of the biomass of the photosynthetic community (Fig. 1.4a).
The combined biomass of the heterotrophic grazer community ranged from 3-16% of the
total microbial biomass among all sites and seasons (! = 9%; Fig. 1.2). Microbial grazer biomass
was dominated by heterotrophic nanoplankton across all sites and sampling dates (65-98% of the
grazer community, Fig. 1.4b). Microzooplankton (ciliates, dinoflagellates) comprised
approximately one-third of the grazer community biomass at all sites during the spring and at
Port of LA during the fall. Ciliate biomass exceeded dinoflagellate biomass at the SPOT and
Catalina stations (except in the spring), while dinoflagellate biomass consistently surpassed
ciliate biomass at Port of LA.
!
1.3.2!Growth!and!grazing!mortality!of!the!photosynthetic!community!!
!
1.3.2.1!Phytoplankton!growth!rates!
!
Unenriched growth rates (µ
0
= intrinsic growth rate) of the total phytoplankton
community ranged from -1.36 to 1.53 d
-1
(! = 0.36 d
-1
; Table 1.1). Corresponding ranges and
! 25!
averages of µ
0
for Synechococcus, Prochlorococcus, and photosynthetic picoeukaryotes were -
0.09 to 2.56 d
-1
(! = 0.42 d
-1
), -1.94 to 2.5 d
-1
(! = 0.06 d
-1
), and -1.08 to 3.03 d
-1
(! = 0.25 d
-1
),
respectively (Table 1.2). No significant differences in median or mean values of µ
0
were
observed for the total phytoplankton community when the sites (all seasons included) or seasons
(all sites included) were compared (Fig. 1.5). However, median values of µ
0
trended higher at
Port of LA (0.44 d
-1
) than at the SPOT (0.27 d
-1
) or Catalina stations (0.32 d
-1
) for the total
phytoplankton community (Fig. 1.5a). Seasonally, higher median values of µ
0
were observed
during the spring and fall for the total phytoplankton community, and the lowest median value
occurred during winter (Fig. 1.5).
No significant differences in median or mean values of µ
0
were observed for the specific
picoplankton groups when compared by site (all seasons included, Fig. A2). Higher median

values, albeit not significantly, were observed at Port of LA for Synechococcus and
photosynthetic picoeukaryotes than at the SPOT or Catalina stations (Fig. A2). Conversely,
median µ
0
for Prochlorococcus was lowest for Port of LA, with negative growth rates in 5 of 10
experiments. When seasons were compared, all picoplankton groups exhibited their highest
median µ
0
values during summer (median growth rates were 0.46 d
-1
, 0.24 d
-1
, and 0.38 d
-1
for
Synechococcus, Prochlorococcus, and photosynthetic picoeukaryotes, respectively) (Fig. A3).
Overall, negative values of µ
0
for Prochlorococcus were obtained for 12 of 30 picoplankton
dilution experiments (Table 1.2).
Nutrient-enriched growth rates (µ
n
) of the total phytoplankton community were
significantly higher than the corresponding unenriched growth rates (µ
0
) at each site (p < 0.05,
all seasons included), and during the spring, summer, and fall seasons (p < 0.05, all sites
included; Table A2). However, nutrient addition did not consistently enhance the growth rates of

! 26!

Figure 1.5. Unenriched growth rates (µ
0
) (a,b) and grazing mortality rates (m) (c,d) of the
total phytoplankton community (based on chl a) for the three study sites (Port of Los
Angeles (Port of LA), San Pedro Ocean Time-series station (SPOT), and Santa Catalina
Island (Catalina)) (a,c) and the four seasons (b,d). Values at each site include all seasons,
and values in each season include all sites. The boxplot displays the distribution of the
data using the interquartile range (IQR) (rectangle; bottom is first quartile, top is third
quartile), the median (horizontal segment within the rectangle), and the minimum and
maximum values that fall within 1.5*IQR (bottom and top “whiskers”, or vertical lines).
Outliers (black points) were identified by R according to the Boxplot Rule (Wilcox 2011).
Pairs with significantly different mean values (▲) or median values ( ) are marked
above the corresponding pair (p ≤ 0.05).
! 27!
Table 1.2. Growth and mortality rates of three groups of photosynthetic picoplankton for each study date and site.
Experiments were conducted using surface water (0 to 1 m depth) and incubated in situ. Mortality rates (m) and nutrient-
enriched growth rates (µ
n
) were calculated using Model I Linear Regressions (significance denoted by asterisks next to the
mortality rate values). Unenriched growth rates (µ
0
) were adjusted from µ
n
values to calculate intrinsic population growth
rates.

Synechococcus Prochlorococcus Photosynthetic picoeukaryotes
Experiment m (d
-1
) µ n (d
-1
) µ 0 (d
-1
) m (d
-1
) µ n (d
-1
) µ 0 (d
-1
) m (d
-1
) µ n (d
-1
) µ 0 (d
-1
)
Port of Los Angeles
15-Jul-12 2.84** 1.12 2.56 2.22** 2.38 2.50 2.89** 1.51 3.03
19-Oct-12 0.69** 0.51 0.56 0.42** 0.11 0.06 0.67** 0.83 0.87
9-Jan-13 0.40** 0.22 0.24 ns 0.08 0.12 0.36** 0.13 0.10
24-Apr-13 0.71** 0.33 0.50 ns -0.12 -0.36 0.73** 0.57 0.58
15-Jul-13 0.84** 0.45 0.46 ns 0.32 0.05 0.75** 0.72 0.46
16-Oct-13 0.63** 0.27 0.30 0.97** -0.18 -0.05 0.68** 0.73 0.77
7-Jan-14 0.43** 0.37 0.39 0.52** -0.21 -0.11 ns 0.13 0.04
8-Apr-14 0.71** 0.16 0.20 ns -0.69 -1.04 0.32** 0.29 0.04
14-Jul-14 1.02** 0.56 0.58 ns 1.01 -1.94 0.75** 0.83 0.29
6-Oct-14 0.44** 0.42 0.52 0.55** 0.40 0.07 ns 0.22 -0.14
San Pedro Ocean Time-Series
15-Jul-12 0.49** 0.38 0.67 0.65** 0.24 0.57 ns 0.20 0.19
19-Oct-12 0.58** 0.41 0.22 0.55** 0.12 -0.10 0.45** 0.19 0.04
9-Jan-13 0.28** 0.19 0.18 0.36** -0.01 0.03 0.54* -0.19 -0.22
24-Apr-13 0.32** 0.42 0.11 ns -0.08 -0.31 ns -0.18 -0.41
15-Jul-13 0.92** 0.27 0.37 1.49** 0.05 0.23 0.99** 0.52 0.47
16-Oct-13 0.40** 0.37 0.42 0.32** 0.06 0.06 0.79** 0.58 0.68
7-Jan-14 0.13** 0.07 0.06 0.40** 0.21 0.31 ns -0.30 -0.30
8-Apr-14 0.49** 0.47 0.49 ns 0.30 0.14 ns 0.95 0.97
14-Jul-14 ns 0.15 0.15 1.09** -1.01 -0.86 ns -0.28 -0.21
6-Oct-14 0.18* 0.38 0.29 ns 0.44 0.32 ns -0.13 -0.34
Catalina Island
15-Jul-12 ns -0.21 -0.29 ns -0.58 -0.55 ns 0.33 -0.02
19-Oct-12 0.28* 0.22 0.23 ns 0.02 -0.01 ns 0.27 -0.10
9-Jan-13 ns -0.09 -0.09 0.70** 0.24 0.49 ns -0.57 -1.08
24-Apr-13 0.21** 0.21 0.02 ns -1.08 -0.93 0.61** 0.25 0.40
15-Jul-13 0.76** 0.36 0.43 0.85** 0.01 0.15 0.60** 0.29 0.38
16-Oct-13 0.40** 0.45 0.58 0.17** -0.04 -0.11 0.68** 0.58 0.40
7-Jan-14 0.26** 0.22 0.28 0.44** 0.13 0.28 ns 0.14 0.09
8-Apr-14 0.48** 0.51 0.69 ns 0.61 0.29 -0.62** -0.08 -0.56
14-Jul-14 0.24** 0.56 0.76 0.71** 0.95 0.91 0.53** 0.80 0.87
6-Oct-14 0.29** 0.60 0.61 0.64** 1.37 1.44 0.28** 0.33 0.31
Significance level of the Model I regression: ns = not significant, *= significant at p ≤ 0.05, **= significant at p ≤ 0.01
! 28!
Synechococcus, Prochlorococcus, or the photosynthetic picoeukaryotes, thus, averaged µ
0
and µ
n

for those assemblages were not significantly different when grouped by season or site. No
environmental parameters were found to correlate significantly with unenriched growth rates of
the total phytoplankton or the picoplankton groups (p < 0.05, Fig. 1.3).

1.3.2.2!Phytoplankton!grazing!mortality!rates!
!
Grazing mortality rates (m) of the total phytoplankton community ranged from -0.76 to
1.31 d
-1
throughout the study period, with an overall average of 0.38 d
-1
across 33 experiments
(Table 1.1). Ranges and averages for mortality rates of Synechococcus, Prochlorococcus, and
photosynthetic picoeukaryotes were 0 to 2.84 d
-1
(! = 0.51 d
-1
), 0 to 2.22 d
-1
(! = 0.44 d
-1
), and -
0.62 to 2.89 d
-1
(! = 0.40 d
-1
), respectively (Table 1.2).
Median grazing mortality rates for the total phytoplankton community were generally
greater at Port of LA (0.59 d
-1
) than at the SPOT (0.21 d
-1
) and Catalina (0.29 d
-1
) stations, but no
significant differences in the mean or median values were detected when data were grouped by
sites (Fig. 1.5c). The lowest median mortality rates for the total phytoplankton community were
observed in the winter, significantly lower than mortality rates observed during the summer (p <
0.05, Fig. 1.5d). Total phytoplankton grazing mortality rate was positively correlated with
chlorophyll concentration (rho = 0.497, p < 0.05, Fig. 1.3).
Median grazing mortality rates for Synechococcus and photosynthetic picoeukaryotes
were generally greater at Port of LA than the other two sites, but grazing mortality was lowest at
Port of LA for Prochlorococcus (Fig. A2). The mean and median grazing mortality rate of
Synechococcus was significantly higher at Port of LA than the SPOT station (mean of 0.87 d
-1

vs. 0.36 d
-1
) and Catalina (median 0.70 d
-1
vs. 0.27 d
-1
), respectively (p < 0.05). Neither the mean
! 29!
nor median grazing mortality rates for Prochlorococcus or photosynthetic picoeukaryotes were
significantly different when data were grouped by sites.
Highest median grazing mortality rates were observed for all groups during the summer,
but no significant differences in mean or median mortality rates were detected for Synechococcus
or photosynthetic picoeukaryotes when data were grouped by season (Fig. A3). However,
Prochlorococcus mean grazing mortality rate was significantly higher in the summer than in the
spring, and Prochlorococcus median grazing mortality rate was significantly higher in the winter
than the spring (p < 0.05).
Synechococcus grazing mortality rate was positively correlated with silicate
concentration (rho = 0.407), chlorophyll concentration (rho = 0.454), and heterotrophic
nanoplankton abundance (rho = 0.528; p < 0.05, Fig. 1.3), while the mortality rate of
Prochlorococcus was negatively correlated with microzooplankton abundance (rho = -0.375). No
significant correlations were found between environmental parameters and photosynthetic
picoeukaryote mortality rate.
Comparison of the net changes in phytoplankton growth rates between the treatment with
mesozooplankton and microzooplankton (WSW) and the treatment containing only
microzooplankton (<200 µm filtrate) yielded no significant differences for either the total
phytoplankton community or the three picoplankton groups. Mean apparent growth rate of total
phytoplankton for all dates was -0.041 and -0.039 d
-1
for the <200 µm and WSW treatments,
respectively, indicating no apparent effect of mesozooplankton removal (or a compensatory
effect by the microzooplankton).
!
!
!
! 30!
1.3.3!Bacterial!grazing!mortality!!
!
Rates of grazer-related bacterial mortality averaged 0.34 d
-1
for all experiments, with a
range of 0.00 to 0.72 d
-1
(Fig. 1.6). Median bacterial mortality rates grouped by study site were
slightly higher at Port of LA (0.39 d
-1
) and Catalina (0.38 d
-1
) than at the SPOT station (0.24 d
-1
),
but no significant differences among the mean or median bacterial grazing rates were detected
between sites (Fig. 1.6a). The mean or median bacterial mortality rates were not significantly
different between seasons, but lower median values were observed during the winter (0.23 d
-1
)
and higher median values during the summer (0.51 d
-1
; Fig. 1.6b), a similar trend to that
observed for total phytoplankton grazing mortality.

1.3.4!Carbon!flow!to!higher!trophic!levels!!
!
Daily removal rates of phytoplankton and bacterial standing stocks by grazers were
calculated in order to compare the amount of carbon grazed daily from each constituent of the
assemblage (Table 1.3). Photosynthetic nanoplankton biomasses were excluded from total
phytoplankton standing stock calculations due to limited data, as explained section 1.2.2. This
exclusion may have reduced the total phytoplankton standing stock calculations by an average of
7.7% ± 6%, which would proportionately underestimate the amount of daily phytoplankton
carbon grazed.
Mean values for the amount of phytoplankton standing stock grazed per day (all
sampling dates included) were 75.8, 8.78, and 8.27 µg C L
-1
d
-1
for Port of LA, the SPOT station,
and Catalina, respectively, while mean values for bacterial standing stock removal per day were
10.9, 4.83, and 5.47 µg C L
-1
d
-1
for each site. More phytoplankton carbon than heterotrophic
bacterial carbon was consumed daily by grazers during all seasons at Port of LA (1.3-18.5x; gray
! 31!

Figure 1.6. Heterotrophic bacterial grazing mortality rates (d
-1
) grouped (a) spatially
(including all seasons), and (b) seasonally (including all sites). No significant differences
in the mean or median values were identified by site or season (α = 0.05). The boxplot
displays the distribution of the data using the interquartile range (IQR) (rectangle; bottom
is first quartile, top is third quartile), the median (horizontal segment within the
rectangle), and the minimum and maximum values that fall within 1.5*IQR (bottom and
top “whiskers”, or vertical lines). Outliers (black points) were identified by R according to
the Boxplot Rule (Wilcox 2011).
!
!
a b
! 32!
Table 1.3. Daily standing stock removal by grazers (for calculation, see Materials and
Methods) for the total phytoplankton community (based on changes in chlorophyll a),
picoplankton groups (from flow cytometric measurements), and heterotrophic bacteria
(from fluorescently-labeled bacteria).

Daily Standing Stock Removal (%)
Experiment Total chl a Synechococcus Prochlorococcus P. picoeukaryotes Bacteria
Port of Los Angeles
13-Mar-12 63.0 - - - -
15-Jul-12 125 247 257 310 -
19-Oct-12 40.2 64.7 35.3 74.2 32.4
9-Jan-13 26.3 37.0 0.00 31.7 24.3
24-Apr-13 53.0 63.8 0.00 67.6 28.7
15-Jul-13 51.6 69.8 0.00 65.0 43
16-Oct-13 96.2 53.6 60.5 71.2 45.3
7-Jan-14 0.00 42.1 38.3 0.00 21.9
8-Apr-14 77.5 56.4 0.00 28.1 39.8
14-Jul-14 71.0 82.9 0.00 60.4 50.7
6-Oct-14 18.2 45.8 43.7 0.00 17.4
San Pedro Ocean Time-series
13-Mar-12 73.2 - - - -
15-Jul-12 0.00 53.6 62.5 0.00 -
19-Oct-12 0.00 49.0 40.9 37.2 26.5
9-Jan-13 19.6 26.7 30.8 38.1 28.5
24-Apr-13 31.2 28.8 0.00 0.00 21.4
15-Jul-13 71.8 70.4 83.9 76.8 45
16-Oct-13 25.5 40.4 28.2 74.9 10.7
7-Jan-14 0.00 12.5 38.2 0.00 10.7
8-Apr-14 48.7 49.0 0.00 0.00 42.5
14-Jul-14 17.1 0.00 48.9 0.00 16.3
6-Oct-14 0.00 19.0 0.00 0.00 20.8
Catalina Island
13-Mar-12 118 - - - -
15-Jul-12 0.00 0.00 0.00 0.00 -
19-Oct-12 0.00 27.3 0.00 0.00 35.9
9-Jan-13 13.8 0.00 63.3 0.00 19
24-Apr-13 18.1 19.2 0.00 55.1 0.5
15-Jul-13 44.4 64.4 60.6 53.7 38.3
16-Oct-13 54.0 43.9 14.8 59.2 34.4
7-Jan-14 0.00 26.3 40.6 0.00 11.4
8-Apr-14 36.9 53.5 0.00 0.00 32.7
14-Jul-14 - 31.5 78.6 63.1 24.2
6-Oct-14 45.6 34.2 98.5 28.4 31.8

! 33!
columns: Fig. 1.7a,b). Phytoplankton standing stocks and daily carbon removal were also higher
than that of the heterotrophic bacteria during the spring and summer at the SPOT station and
Catalina, although the magnitude of the difference between food sources was smaller (1.6-3.7x;
gray columns: Fig. 1.7c-f). During the winter, similar amounts of phytoplankton and bacterial
carbon were grazed at the SPOT station, whereas the amount of bacterial carbon grazed was 5.3x
higher than the phytoplankton carbon grazed at Catalina. More bacterial carbon was grazed daily
at the SPOT station than phytoplankton carbon during the fall (2.3x higher), while daily removal
of phytoplankton and bacterial carbon was similar at Catalina during this season (gray columns:
Fig. 1.7c-f).
The standing stock of the heterotrophic bacteria was the largest of the picoplankton
groups at all sites, exceeding the standing stock of the other groups by up to two orders of
magnitude (Table 1.4). The amount of carbon grazed daily from the heterotrophic bacteria (10.9,
4.83, and 5.47 µg C L
-1
d
-1
for Port of LA, the SPOT station, and Catalina, respectively) was
approximately equal to that grazed from the three picophytoplankton groups combined (13.49,
7.53, and 6.59 µg C L
-1
d
-1
for Port of LA, the SPOT station, and Catalina, respectively). The
average percentage of carbon grazed within a picoplankton group (mean daily carbon grazed
divided by mean standing stock) ranged from 25-59% of the total standing stock of that group
(Table 1.4).

1.4. Discussion
1.4.1!Microbial!growth!and!grazing!rates!were!similar!among!sites!
!
Despite differences in microbial community composition and standing stocks, rates of
phytoplankton growth and mortality revealed few statistically significant differences between the
! 34!

Figure 1.7: Average seasonal standing stock (µg C L
-1
; black columns, left axis) and daily
carbon grazed (µg C L
-1
d
-1
; gray columns, right axis) for the phytoplanktonic (a,c,e) and
heterotrophic bacterial (b,d,f) assemblages at the Port of Los Angeles (Port of LA, a,b),
San Pedro Ocean Time-series station (SPOT, c,d), and Catalina Island (Catalina, e,f). Note
the breaks in the y-axes and subsequent change of scale.
! 35!
Table 1.4. Means (and ranges) of the total standing stock (µg C L
-1
) and daily carbon removal by grazing (µg C L
-1
d
-1
) for
heterotrophic bacteria and three photosynthetic picoplanktonic assemblages (all seasons included). The percentage of
daily carbon grazed (mean daily carbon grazed divided by mean standing stock) within a given picoplankton group is
provided in parentheses in the ‘Daily Carbon Grazed’ column.

Port%of%Los%Angeles San%Pedro%Ocean%Time 6series Catalina%Island
%
Standing%Stock
(μg%C% L
61
)
Daily%Carbon%Grazed
(μg%C% L
61%
d
61
)
Standing%Stock
(μg%C% L
61
)
Daily%Carbon%Grazed
(μg%C% L
61%
d
61
)
Standing%Stock
(μg%C% L
61
)
Daily%Carbon%Grazed
(μg%C% L
61%
d
61
)
Heterotrophic%bacteria 33.1%(16.4% –% 45.3) 10.9%(5.13% –% 20.4)%(33 %) 19.1%(9.31% –% 28.8) 5.88%(1.47 611.7)%(25 %) 20.2%(6.53% –% 28.7) 5.47% (1.96% –% 9.89)%(27%)
Synechococcus 4.74%(0 .226% –% 14.3) 2.81%(0.21 5% –% 9.24)%(59%) 9.75%(1.43% –% 38.8) 4.74%(0 % –% 27.3)%(48%) 8.23%(0.286% –% 24.5) 3.13%( 0% –% 15.6)%(38%)
Prochlorococcus 0.536% (0.0243% –% 1.64) 0.174% (0% –% 0.628)%(32%) 2.97%(0.02 05% –% 13.3) 1.31%(0 % –% 5.08)%(44%) 2.09%(0.03 44% –% 5.87) 0.648% (0% –% 2.38)%(31%)
P.%picoeukaryotes 19.7% (3.58% –% 45.4) 10.5%(0 % –% 27.4)%(53%) 4.16%(1.11 69.10) 1.48%(0 % –% 6.82)%(36%) 7.03%(1.32% –% 19.5) 2.81%(0 % –% 11.5)%(40%)
! 36!
three study sites. The sole significant difference between sites was a greater Synechococcus
grazer-related mortality rate at the Port of LA, yet this relationship was not highly robust as it
depended on the metric used to evaluate statistical significance (i.e. mean or median). These
findings were surprising as they indicated that microbial communities of different composition
and standing stocks exhibited similar overall turnover rates. In our study, significantly higher
mean nutrient concentrations at the Port of LA supported microbial community standing stocks
that were 2-4.5x those found at the SPOT station or Catalina (Table 1.1 & Fig. 1.2). Also,
environmental differences between sites (Table 1.1) supported dominance of different
phytoplankton onshore and offshore (Figs. 1.2 & 1.4a). Photosynthetic microbial eukaryotes
(picoeukaryotes, nanoeukaryotes, diatoms, and dinoflagellates) comprised 77-99% of the
phytoplankton biomass during all seasons at the Port of LA (Fig. 1.4a), while coccoid
cyanobacteria constituted ~50% of the total phytoplankton biomass at the SPOT and Catalina
stations (with the exception of the spring season in which diatoms dominated the communities at
all sites). Bray Curtis community dissimilarity results from a parallel molecular survey of
microbial eukaryotic diversity and activity corroborate our findings, demonstrating that the Port
of LA clustered separately from the SPOT and Catalina stations (Hu et al. 2016). The
predominance of larger phytoplankton in the Port of LA presumably reflects more nutrient
replete conditions there, whereas the smaller phytoplankton dominance offshore may reflect the
lower half-saturation constants and greater diffusive efficiency characteristic of minute
phytoplankton (Williams & Follows 2011).
Similar growth and grazing rates within our geographically constrained study region
(~45km across, all at 118
o
W) are not surprising given that comparable findings have been
reported along longitudinal gradients spanning ~10
o
. In the California Current Ecosystem (CCE),
! 37!
standing stocks between coastal and offshore waters varied up to 30-fold, yet the rates of
phytoplankton growth and microzooplankton grazing did not differ substantially among locations
(Landry et al. 2009). Similarly, no significant differences in phytoplankton growth and mortality
rates were detected between a coastal river plume and Kuroshio current waters in the East China
Sea (Zheng et al. 2015), a gradient that spans ~7
o
longitude. Decreasing trends in phytoplankton
and picoplankton growth and grazing mortality rates have been reported along onshore-offshore
gradients in the North Atlantic Subtropical Gyre and East China Sea, however, those trends were
either not significant (Gutiérrez-Rodríguez et al. 2011) or the significance was not reported (Guo
et al. 2014). Our study, taken together with these findings, suggests that variability in
phytoplankton growth and mortality rates does not necessarily dictate differences in productivity
between sites.
Rates of grazer-mediated bacterial mortality were also not significantly different between
stations when all seasons were included (Fig. 1.6a). This finding is not surprising because
bacterial abundances did not vary markedly between locations (Fig. 1.2). To our knowledge, we
are the first study to address bacterial mortality rates across an onshore-offshore gradient. Our
study suggests that bacterial mortality rates, like phytoplankton growth and mortality rates, are
not necessarily linked to the biological productivity of an area at small spatial scales.
The microbial abundances and rates of growth and mortality reported in this study are in
good accordance with other microbial rate studies in the waters off California. In the San Pedro
Channel, we found phytoplankton growth rates and grazing mortality rates averaging 0.36 d
-1
and
0.38 d
-1
, respectively. These averages are similar to averages of growth (0.44 and 0.48 d
-1
) and
grazing mortality rates (0.2 and 0.5 d
-1
) reported in the CCE (Landry et al. 2009, Pasulka et al.
2015), as well as the median growth (0.53 d
-1
) and grazing mortality rates (0.36 d
-1
) reported for
! 38!
coastal Pacific regions in a meta-analysis (Schmoker et al. 2013). The average Synechococcus
growth rate (0.42 d
-1
) in the San Pedro Channel agreed well with other averages reported from
the CCE and at Scripps Pier (range 0.44-0.66 d
-1
), but our average mortality rate (0.51 d
-1
) was
approximately double that found at those sites (range 0.14-0.33 d
-1
) (Worden & Binder 2003,
Worden et al. 2004, Pasulka et al. 2015). Conversely, Prochlorococcus and photosynthetic
picoeukaryote average growth rates (0.06 and 0.25 d
-1
, respectively) were less than those
reported in the CCE and at Scripps Pier (range 0.33-0.99 and 0.40-0.99 d
-1
). Average grazing
mortality rates of Prochlorococcus (0.44 d
-1
) and the photosynthetic picoeukaryotes (0.40 d
-1
)
were within the range of mean values reported for the CCE and at Scripps Pier (range 0.36-0.49
and 0.39-0.45 d
-1
, respectively). The average bacterial standing stock removal (28% d
-1
) in our
study was less than the reported range from Santa Monica Bay (43-82% d
-1
)(Fuhrman & Noble
1995), but it was consistent with an average of 29% d
-1
reported by a seasonal study in the NW
Mediterranean (Boras et al. 2009). Our average bacterial mortality rate is also in agreement with
bacterial doubling times of ~5 days calculated for the SPOT station (Cram et al. 2014).

1.4.2%Microbial%growth%and%grazing%rates%revealed%low%seasonality%in%the%San%Pedro%Channel%
!
Relatively constant rates of growth and mortality across seasons were observed in our
study, which is consistent with the narrow range of environmental factors observed in the San
Pedro Channel (Table 1.1). Increases in phytoplankton and bacterial mortality rates in the
summer (Figs. 1.5, 1.6b, & A2), albeit non-significant for the bacteria, Synechococcus, or the
photosynthetic picoeukaryotes, suggest temperature may explain the somewhat elevated grazing
rates observed during the summer months. However, similarly high temperatures in the fall were
! 39!
accompanied by lower grazing rates, confounding our ability to determine whether temperature
is an important driver of protistan grazing in our region (Fig. 1.3).
Relatively constant growth and mortality rates between seasons were documented despite
seasonal shifts in the community composition. Several molecular studies have shown distinct
seasonality in bacterial community composition (Chow et al. 2013, Cram et al. 2014) and
eukaryotic community composition (Countway et al. 2010, Kim et al. 2014, Hu et al. 2016) in
the surface waters of the San Pedro Channel. Changes in the relative abundance of key microbial
groups between seasons were also documented in this study; for example, diatoms dominated the
photosynthetic community in the spring season, coincident with an increase of microzooplankton
in the heterotrophic grazer community (Fig. 1.4). Similar average growth and mortality rates
measured for seasonally distinct communities may reflect a tight coupling between the
progression of prey populations and their predator populations.
The seasonality of microbial rate processes has been shown to be stronger in truly
temperate and high-latitude regions than at our temperate-to-subtropical, coastal ocean site.
Forcing factors that vary with season include changes in light availability and day length, wind
and weather patterns, upwelling and nutrient availability, and temperature. The Southern
California Bight appears to function at the intersection of a subtropical and temperate regime: a
narrow temperature range and permanently stratified water column persist year-round (Caron et
al. in revision), yet episodic wind-driven upwelling in the spring and enhanced rainfall and
stormwater discharge in the winter drive seasonal fluctuations in community biomass and rate
processes characteristic of temperate regions (Nezlin & Li 2003, Nezlin et al. 2012). The
episodic nature of our region was reflected in the amount of variability observed for
phytoplankton growth and mortality rates within a season, between seasons, and between years.
! 40!
That is, intraseasonal variability—the variability observed in phytoplankton growth or mortality
rates within a season at all sites—demonstrated the highest level of variability (i.e. COV of 93%
for µ
0
and 113% for m in winter) as the timing of rainfall or upwelling events can vary between
years. Interseasonal variability—the variability observed between the average phytoplankton
growth and mortality rates for each season (all sites included)— was moderate (COV of 59% for
µ
0
and 52% for m), distinguishing each season from one another yet reflecting the mild seasonal
shifts experienced in our region. Interannual variability—the variability observed between the
average phytoplankton growth or mortality rates for each year (all sites included)—was the
lowest (COV of 10% for µ
0
and 21% for m), indicating a relatively consistent seasonal signature
between years (at least during this 3-year study) and reflecting the significance of episodic events
in the Southern California Bight as drivers of plankton biology.
Studies conducted in more northerly temperate environments have shown strong seasonal
trends in microbial growth and grazing rates. In Narragansett Bay, Rhode Island, USA,
phytoplankton growth and mortality rates peaked during the summer season, primarily due to
increased temperature, but also potentially due to changes in phytoplankton community
composition (Lawrence & Menden-Deuer 2012). Summer rates of phytoplankton growth and
mortality were also significantly higher in the East China Sea than those recorded in the winter at
that site, and paralleled increases in picoplankton abundance and temperature (Zheng et al.
2015); similar increases in picoplankton growth and mortality rates were documented, though the
significance of the relationship was not reported (Guo et al. 2014). Bacterial mortality was also
strongly affected by temperature in Vineyard Sound, Massachusetts, with the percentage of FLB
consumed steadily increasing from 3 to 41% per day from December to June (Marrasé 1992).
Conversely, bacterial mortality rates measured monthly in the NW Mediterranean (temp range
! 41!
15-25
o
C) showed no seasonal pattern (Boras et al. 2009). Therefore, the mild seasonal shifts in
growth and mortality rates observed in our study appear to reflect the temperate-to-subtropical
nature of the San Pedro Channel.

1.4.3%Community%carbon%flow%differed%between%sites%and%seasons%
!
The combination of our community biomass data and protistan grazing measurements
facilitated a unique comparison of carbon turnover rates between key prey populations. Our
measurements revealed that phytoplankton and heterotrophic bacteria (bacteria + archaea,
excluding Synechococcus and Prochlorococcus) were both important carbon sources to higher
trophic levels in the San Pedro Channel. Phytoplankton were the predominant carbon source for
protistan grazers at all three sites during the spring and summer seasons, when phytoplankton
standing stocks were highest (Fig. 1.7). However, carbon flow through the bacteria and
phytoplankton were of similar magnitude in the winter and fall seasons. Within the
picoplanktonic size class, the mean daily carbon consumption of the heterotrophic bacteria
equaled that of the photosynthetic component at all sites (Synechococcus, Prochlorococcus,
photosynthetic picoeukaryotes combined; Table 1.4).
Experiments investigating the removal rates of both the bacteria and phytoplankton are
rare and have generally been conducted using the dilution method to estimate growth and
mortality rates of all microbial groups (Putland 2000, Pearce et al. 2008, Pearce et al. 2011,
Pasulka et al. 2015). However, filtration processes can alter metabolite and organic matter
composition, and thereby impact bacterial growth rates unevenly across the dilution series
(Pasulka et al. 2015, Pree et al. 2016). For this reason we estimated bacterial mortality by the rate
of FLB disappearance. Bacterial mortality measurements based on fluorescently-labeled prey
! 42!
also have caveats (Caron 2001), although we note that our measured bacterial mortality rates
were in good accordance with calculated bacterial production rates in the region (noted in section
1.4.1; Cram et al. 2014).
Our study did not directly address other sources of mortality such as viral lysis (Brum et
al. 2014), therefore, estimates of community turnover rates, particularly for the bacteria, are
conservative estimates. No statistical difference was found between our microzooplankton
(<200µm) and micro- + mesozooplankton (WSW) treatments, demonstrating that pre-screening
of the seawater to exclude metazoan grazers does not alter phytoplankton mortality rates in our
region. However, as the collection method may not adequately sample the mesozooplankton
population, we cannot quantitatively assess the grazing impact of the mesozooplankton during
our study based on these experiments.

1.4.4%Implications%of%groupBspecific%carbon%turnover%for%biogeochemical%models%%
!
Comparisons of the standing stocks and carbon turnover of key microbial prey groups are
necessary for the construction of biogeochemical models. Despite our understanding of
heterotrophic bacteria as a fundamental part of microbial food webs (Azam et al. 1983, Ducklow
1983), only a few ecosystem models have explicitly represented bacterial biomass when
investigating ecosystem dynamics (Hasumi & Nagata 2014, Xiu & Chai 2014, Weitz et al.
2015). Our study highlights the importance of bacteria as a major, and sometimes dominant,
source of carbon to higher trophic levels, and our coincident measurements of bacterial biomass
and bacterial mortality enable their inclusion in model formulations. In addition, our study
supports the appropriateness of size-structured biogeochemical models (Ward et al. 2012, Ward
et al. 2014), indicating a small range of mean turnover rates among the different
! 43!
picophytoplankton populations (27%, 12%, and 9% at Port of LA, the SPOT station, and
Catalina Island, respectively; Table 1.4). The average standing stock removal for
picophytoplankton in the San Pedro Channel was 42.3% d
-1
, which corresponds to a turnover
time of 2-3 days. Our results agree with research showing comparable grazing susceptibility
amongst small phytoplankton groups (0.45-4.0 µm) in the Eastern Pacific (Taniguchi et al.
2014). Phytoplankton growth and mortality rates have been shown to vary by approximately a
factor of two throughout the euphotic zone (Landry et al. 2009, Landry et al. 2011, Landry et al.
2015), thus caution should be exercised when applying the surface rates measured in this study to
models of euphotic zone microbial community dynamics.
Modeling zooplankton dynamics requires the construction of mathematical relationships
between explanatory environmental parameters and grazing pressure. Ideal parameters are those
that provide high predictive power but are relatively easily collected, such as temperature and
chlorophyll concentration, but these parameters were only marginally important in the present
study. Several studies have shown a strong positive relationship between temperature and
microbial growth rates or grazing rates (Marrasé 1992, Rose & Caron 2007, Lawrence &
Menden-Deuer 2012, Schmoker et al. 2013, Taniguchi et al. 2014). Strong temperature
dependence was not detected in our study (Fig. 1.3), ostensibly due to the narrow annual
temperature range at our study site (13.1-21.6
o
C). Elevated (but non-significant) grazing rates on
all microbial assemblages during the summer suggest that temperature plays only a minor role in
the San Pedro Channel (Figs. 1.5, 1.6, & A2). We were able to detect a significant positive
correlation between the initial chlorophyll a concentration and phytoplankton mortality rate as
well as Synechococcus mortality rate (Fig. 1.3), a relationship that has been reported in other
studies (Verity 1986, Schmoker et al. 2013). However, chlorophyll concentration has not been a
! 44!
consistent predictor of protistan grazing rates in the literature as bulk chlorophyll does not
represent species composition, prey palatability, or predator-prey interactions (Menden-Deuer &
Fredrickson 2010, Lawrence & Menden-Deuer 2012, Guo et al. 2014, Zheng et al. 2015).
Integrating species composition, both of the grazers and the prey, into future studies of microbial
rate processes may help better elucidate spatiotemporal patterns and environmental predictors of
growth and grazing rates generally, and especially at our study site.

1.5. Conclusion
An investigation of microbial community composition and trophic activities across three
years, three sites, and four seasons in the surface waters of the San Pedro Channel revealed
similar averages in microbial rate processes across seasonal and spatial scales, despite
differences in standing stocks and the relative abundance of key microbial assemblages.
Microbial communities with substantially different standing stocks had similar rates of turnover.
Comparisons of carbon flow through several microbial prey assemblages highlighted the
importance of heterotrophic bacteria as a food source for protistan grazers. Carbon consumption
from the heterotrophic bacteria was of similar magnitude to the picophytoplankton in all seasons
and similar to the total phytoplankton community in some locations and seasons. The narrow
annual range of environmental parameters and episodic nature of the Southern California Bight
precluded the detection of meaningful relationships between many environmental factors and
microbial growth and mortality rates. The carbon fluxes reported in this study, combined with
the detailed community biomass values of Caron et al. (in revision), facilitate the creation and
parameterization of biogeochemical models at coastal ocean sites.

! 45!
Chapter One Literature Cited
Azam!F,!Fenchel!T,!Field!J,!Gray!J,!Meyer?Reil!L,!Thingstad!F!(1983)!The!ecological!role!of!
water?column!microbes!in!the!sea.!Mar!Ecol!Prog!Ser!10:257?263!
Boras!JA,!Sala!MM,!Vázquez?Domínguez!E,!Weinbauer!MG,!Vaqué!D!(2009)!Annual!changes!
of!bacterial!mortality!due!to!viruses!and!protists!in!an!oligotrophic!coastal!
environment!(NW!Mediterranean).!Environ!Microbiol!11:1181?1193!
Brum!JR,!Morris!JJ,!Décima!M,!Stukel!MR!(2014)!Mortality!in!the!oceans:!causes!and!
consequences.!Eco?DAS!IX!Chapter!2:16?48!
Buitenhuis!ET,!Li!WKW,!Lomas!MW,!Karl!DM!(2012)!Bacterial!biomass!distribution!in!the!
global!ocean.!Earth!Syst!Sci!Data!Disc!5:301?315!
Calbet!A,!Landry!MR!(2004)!Phytoplankton!growth,!microzooplankton!grazing,!and!carbon!
cycling!in!marine!systems.!Limnol!Oceanogr!49:51?57!
Caron!DA!(2001)!Protistan!herbivory!and!bacterivory.!In:!Paul!J!(ed)!Methods!in!
Microbiology,!Book!30.!Academic!Press!Ltd.!
Caron!DA!(2016)!Mixotrophy!stirs!up!our!understanding!of!marine!food!webs.!PNAS!
113:2806?2808!
Caron!DA,!Countway!PD,!Jones!AC,!Kim!DY,!Schnetzer!A!(2012)!Marine!protistan!diversity.!
Annu!Rev!Mar!Sci!4:467?493!
Caron!DA,!Dam!HG,!Kremer!P,!Lessard!EJ,!Madin!LP,!Malone!TC,!Napp!JM,!Peele!ER,!Roman!
MR,!Youngbluth!MJ!(1995)!The!contribution!of!microorganisms!to!particulate!
carbon!and!nitrogen!in!surface!waters!of!the!Sargasso!Sea!near!Bermuda.!Deep!Sea!
Res!I!42:943?972!
Casey!JR,!Aucan!JP,!Goldberg!SR,!Lomas!MW!(2013)!Changes!in!partitioning!of!carbon!
amongst!photosynthetic!pico?!and!nanoplankton!groups!in!the!Sargasso!Sea!in!
response!to!changes!in!the!North!Atlantic!Oscillation.!Deep!Sea!Res!II!93:58?70!
Chow!C?ET,!Sachdeva!R,!Cram!JA,!Steele!JA,!Needham!DM,!Patel!A,!Parada!AE,!Fuhrman!JA!
(2013)!Temporal!variability!and!coherence!of!euphotic!zone!bacterial!communities!
over!a!decade!in!the!Southern!California!Bight.!ISME!J!7:2259?2273!
Countway!PD,!Vigil!PD,!Schnetzer!A,!Moorthi!SD,!Caron!DA!(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!
Cram!JA,!Chow!C?ET,!Sachdeva!R,!Needham!DM,!Parada!AE,!Steele!JA,!Fuhrman!JA!(2014)!
Seasonal!and!interannual!variability!of!the!marine!bacterioplankton!community!
throughout!the!water!column!over!ten!years.!ISME!J!9:563?580!
del!Giorgio!PA,!Bird!DF,!Prairie!YT,!Planas!D!(1996)!Flow!cytometric!determination!of!
bacterial!abundance!in!lake!plankton!with!the!green!nucleic!acid!stain!SYTO!13.!
Limnol!Oceanogr!41:783?789!
Doney!SC,!Ruckelshaus!M,!Duffy!JE,!Barry!JP,!Chan!F,!English!CA,!Galindo!HM,!Grebmeier!JM,!
Hollowed!AB,!Knowlton!N,!Polovina!J,!Rabalais!NN,!Sydeman!WJ,!Talley!LD!(2012)!
Climate!change!impacts!on!marine!ecosystems.!Annu!Rev!Mar!Sci!4:11?37!
Ducklow!HW!(1983)!Production!and!fate!of!bacteria!in!the!oceans.!BioScience!33:494?501!
Field!CB,!Behrenfeld!MJ,!Randerson!JT,!Falkowski!P!(1998)!Primary!production!of!the!
biosphere:!integrating!terrestrial!and!oceanic!components.!Science!281:237?240!
Fuhrman!JA,!Cram!JA,!Needham!DM!(2015)!Marine!microbial!community!dynamics!and!
their!ecological!interpretation.!Nat!Rev!Microbiol!13:133?146!
! 46!
Fuhrman!JA,!Noble!RT!(1995)!Viruses!and!protists!cause!similar!bacterial!mortality!in!
coastal!seawater.!Limnol!Oceanogr!40:1236?1242!
Fukuda!R,!Ogawa!H,!Nagata!T,!Koike!I!(1998)!Direct!determination!of!carbon!and!nitrogen!
contents!of!natural!bacterial!assemblages!in!marine!environments.!Appl!Environ!
Microbiol!64:3352?3358!
Guo!C,!Liu!H,!Zheng!L,!Song!S,!Chen!B,!Huang!B!(2014)!Seasonal!and!spatial!patterns!of!
picophytoplankton!growth,!grazing!and!distribution!in!the!East!China!Sea.!
Biogeosciences!11:1847?1862!
Gutiérrez?Rodríguez!A,!Latasa!M,!Agustí!S,!Duarte!CM!(2011)!Distribution!and!contribution!
of!major!phytoplankton!groups!to!carbon!cycling!across!contrasting!conditions!of!
the!subtropical!northeast!Atlantic!Ocean.!Deep!Sea!Res!I!58:1115?1129!
Hasumi!H,!Nagata!T!(2014)!Modeling!the!global!cycle!of!marine!dissolved!organic!matter!
and!its!influence!on!marine!productivity.!Ecol!Modell!288:9?24!
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!
Karner!MB,!DeLong!EF,!Karl!DM!(2001)!Archaeal!dominance!in!the!mesopelagic!zone!of!the!
Pacific!Ocean.!Nature!409:507?510!
Kawasaki!N,!Sohrin!R,!Ogawa!H,!Nagata!T,!Benner!R!(2011)!Bacterial!carbon!content!and!
the!living!and!detrital!bacterial!contributions!to!suspended!particulate!organic!
carbon!in!the!North!Pacific!Ocean.!Aquat!Microb!Ecol!62:165?176!
Kiefer!D,!Chamberlin!W,!Booth!C!(1989)!Natural!fluorescence!of!chlorophyll!a:!relationship!
to!photosynthesis!and!chlorophyll!concentration!in!the!western!South!Pacific!gyre.!
Limnol!Oceanogr!34:868?881!
Kim!DY,!Countway!PD,!Jones!AC,!Schnetzer!A,!Yamashita!W,!Tung!C,!Caron!DA!(2014)!
Monthly!to!interannual!variability!of!microbial!eukaryote!assemblages!at!four!
depths!in!the!eastern!North!Pacific.!ISME!J!8:515?530!
Landry,!R.!M,!Kirshtein,!J.,!Constantinou,!J.!(1995)!A!refined!dilution!technique!for!
measuring!the!community!grazing!impact!of!microzooplankton,!with!experimental!
tests!in!the!central!equatorial!Pacific.!Mar!Ecol!Prog!Ser!120:53?63!
Landry!M,!Constantinou!J,!Latasa!M,!Brown!S,!Bidigare!R,!Ondrusek!M!(2000)!Biological!
response!to!iron!fertilization!in!the!eastern!equatorial!Pacific!(IronEx!II).!III.!
Dynamics!of!phytoplankton!growth!and!microzooplankton!grazing.!Mar!Ecol!Prog!
Ser!201:57?72!
Landry!MR,!Calbet!A!(2004)!Microzooplankton!production!in!the!oceans.!ICES!J!Mar!Sci!
61:501?507!
Landry!MR,!Hassett!RP!(1982)!Estimating!the!grazing!impact!of!marine!micro?zooplankton.!
Mar!Biol!67:283?288!
Landry!MR,!Ohman!MD,!Goericke!R,!Stukel!MR,!Tsyrklevich!K!(2009)!Lagrangian!studies!of!
phytoplankton!growth!and!grazing!relationships!in!a!coastal!upwelling!ecosystem!
off!Southern!California.!Prog!Oceanogr!83:208?216!
Landry!MR,!Selph!KE,!Décima!M,!Gutiérrez?Rodríguez!A,!Stukel!MR,!Taylor!AG,!Pasulka!AL!
(2015)!Phytoplankton!production!and!grazing!balances!in!the!Costa!Rica!Dome.!J!
Plankton!Res!
! 47!
Landry!MR,!Selph!KE,!Taylor!AG,!Décima!M,!Balch!WM,!Bidigare!RR!(2011)!Phytoplankton!
growth,!grazing!and!production!balances!in!the!HNLC!equatorial!Pacific.!Deep!Sea!
Res!II!58:524?535!
Lawrence!C,!Menden?Deuer!S!(2012)!Drivers!of!protistan!grazing!pressure:!seasonal!
signals!of!plankton!community!composition!and!environmental!conditions.!Mar!Ecol!
Prog!Ser!459:39?52!
Marrasé!CL,!Ee!Lin;!Caron,!David!A.!(1992)!Seasonal!and!daily!changes!in!bacterivory!in!a!
coastal!plankton!community.!Mar!Ecol!Prog!Ser!82:281?289!
Martiny!AC,!Talarmin!A,!Mouginot!C,!Lee!JA,!Huang!JS,!Gellene!AG,!Caron!DA!(2016)!
Biogeochemical!interactions!control!a!temporal!succession!in!the!elemental!
composition!of!marine!communities.!Limnol!Oceanogr!61:531?542!
McMahon!KW,!McCarthy!MD,!Sherwood!OA,!Larsen!T,!Guilderson!TP!(2015)!Millennial?
scale!plankton!regime!shifts!in!the!subtropical!North!Pacific!Ocean.!Science!
350:1530?1533!
Menden?Deuer!S,!Fredrickson!K!(2010)!Structure?dependent,!protistan!grazing!and!its!
implication!for!the!formation,!maintenance!and!decline!of!plankton!patches.!Mar!
Ecol!Prog!Ser!420:57?71!
Menden?Deuer!S,!Lessard!EJ!(2000)!Carbon!to!volume!relationships!for!dinoflagellates,!
diatoms,!and!other!protist!plankton.!Limnol!Oceanogr!45:569?579!
Nezlin!NP,!Li!B?L!(2003)!Time?series!analysis!of!remote?sensed!chlorophyll!and!
environmental!factors!in!the!Santa!Monica–San!Pedro!Basin!off!Southern!California.!
J!Mar!Syst!39:185?202!
Nezlin!NP,!Sutula!MA,!Stumpf!RP,!Sengupta!A!(2012)!Phytoplankton!blooms!detected!by!
SeaWiFS!along!the!central!and!southern!California!coast.!J!Geophys!Res!117!
Pasulka!AL,!Samo!TJ,!Landry!MR!(2015)!Grazer!and!viral!impacts!on!microbial!growth!and!
mortality!in!the!southern!California!Current!Ecosystem.!J!Plankton!Res!37:1?17!
Pearce!I,!Davidson!AT,!Thomson!PG,!Wright!S,!van!den!Enden!R!(2011)!Marine!microbial!
ecology!in!the!sub?Antarctic!Zone:!rates!of!bacterial!and!phytoplankton!growth!and!
grazing!by!heterotrophic!protists.!Deep!Sea!Res!II!58:2248?2259!
Pearce!I,!Davidson!AT,!Wright!S,!van!den!Enden!R!(2008)!Seasonal!changes!in!
phytoplankton!growth!and!microzooplankton!grazing!at!an!Antarctic!coastal!site.!
Aquat!Microb!Ecol!50:157?167!
Pree!B,!Kuhlisch!C,!Pohnert!G,!Sazhin!AF,!Jakobsen!HH,!Lund!Paulsen!M,!Frischer!ME,!
Stoecker!D,!Nejstgaard!JC,!Larsen!A!(2016)!A!simple!adjustment!to!test!reliability!of!
bacterivory!rates!derived!from!the!dilution!method.!Limnol!Oceanogr:!Methods!
14:114?123!
Putland!JN!(2000)!Microzooplankton!herbivory!and!bacterivory!in!Newfoundland!coastal!
waters!during!spring,!summer!and!winter.!J!Plankton!Res!22:253?277!
R!Core!Team!(2014)!R:!a!language!and!environment!for!statistical!computing.!!
Rose!JM,!Caron!DA!(2007)!Does!low!temperature!constrain!the!growth!rates!of!
heterotrophic!protists?!Evidence!and!implications!for!algal!blooms!in!cold!waters.!
Limnol!Oceanogr!52:886?895!
Schmoker!C,!Hernández?León!S,!Calbet!A!(2013)!Microzooplankton!grazing!in!the!oceans:!
impacts,!data!variability,!knowledge!gaps!and!future!directions.!J!Plankton!Res!
35:691?706!
! 48!
Sherr!BF,!Sherr!E,!Caron!DA,!Vaulot!D,!Worden!AZ!(2007)!Oceanic!protists.!Oceanography!
20:130?134!
Sherr!BF,!Sherr!EB,!Fallon!RD!(1987)!Use!of!monodispersed,!fluorescently!labeled!bacteria!
to!estimate!in!situ!protozoan!bacterivory.!Appl!Environ!Microbiol!53:958?965!
Sherr!E,!Sherr!B!(1994)!Bacterivory!and!herbivory:!key!roles!of!phagotrophic!protists!in!
pelagic!food!webs.!Microb!Ecol!28:223?235!
Sherr!EB,!Caron!DA,!Sherr!BF!(1993)!Staining!of!heterotrophic!protists!for!visualization!via!
epifluorescence!microscopy.!Handbook!of!methods!in!aquatic!microbial!
ecology:213?227!
Sherr!EB,!Sherr!BF!(2002)!Significance!of!predation!by!protists!in!aquatic!microbial!food!
webs.!Antonie!van!Leeuwenhoek!81:293?308!
Strom!SL!(2000)!Bacterivory:!interactions!between!bacteria!and!their!grazers.!In:!
Kirchman!DL!(ed)!Microbial!Ecology!of!the!Oceans.!Wiley?Liss,!Inc.,!New!York!
Suttle!CA!(2005)!Viruses!in!the!sea.!Nature!437:356?361!
Taniguchi!D,!Landry!M,!Franks!P,!Selph!K!(2014)!Size?specific!growth!and!grazing!rates!for!
picophytoplankton!in!coastal!and!oceanic!regions!of!the!eastern!Pacific.!Mar!Ecol!
Prog!Ser!509:87?101!
Utermöhl!H!(1958)!Zur!gewassertypenfrage!tropischer!seen.!Verh!Int!Ver!Limnol!13:236?
251!
Verity!PG!(1986)!Grazing!of!phototrophic!nanoplankton!by!microzooplankton!in!
Narragansett!Bay.!Mar!Ecol!Prog!Ser!29:105?115!
Ward!BA,!Dutkiewicz!S,!Follows!MJ!(2014)!Modelling!spatial!and!temporal!patterns!in!size?
structured!marine!plankton!communities:!top–down!and!bottom–up!controls.!J!
Plankton!Res!36:31?47!
Ward!BA,!Dutkiewicz!S,!Jahn!O,!Follows!MJ!(2012)!A!size?structured!food?web!model!for!
the!global!ocean.!Limnol!Oceanogr!57:1877?1891!
Weitz!JS,!Stock!CA,!Wilhelm!SW,!Bourouiba!L,!Coleman!ML,!Buchan!A,!Follows!MJ,!Fuhrman!
JA,!Jover!LF,!Lennon!JT,!Middelboe!M,!Sonderegger!DL,!Suttle!CA,!Taylor!BP,!Frede!
Thingstad!T,!Wilson!WH,!Eric!Wommack!K!(2015)!A!multitrophic!model!to!quantify!
the!effects!of!marine!viruses!on!microbial!food!webs!and!ecosystem!processes.!ISME!
J!9:1352?1364!
Welschmeyer!NA!(1994)!Fluorometric!analysis!of!chlorophyll!a%in!the!presence!of!
chlorophyll!b%and!phaeopigments.!Limnol!Oceanogr!39:1985?1992!
Wilcox!R!(2011)!Modern!statistics!for!the!social!and!behavioral!sciences:!a!practical!
introduction.!Taylor!&!Francis!
Williams!RG,!Follows!MJ!(2011)!Ocean!dynamics!and!the!carbon!cycle:!principles!and!
mechanisms.!Cambridge!University!Press!
Worden!AZ,!Binder!BJ!(2003)!Application!of!dilution!experiments!for!measuring!growth!
and!mortality!rates!among!Prochlorococcus!and!Synechococcus!populations!in!
oligotrophic!environments.!Aquat!Microb!Ecol!30:159?174!
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!
Worden!AZ,!Nolan!JK,!Palenik!B!(2004)!Assessing!the!dynamics!and!ecology!of!marine!
picophytoplankton:!the!importance!of!the!eukaryotic!component.!Limnol!Oceanogr!
49:168?179!
! 49!
Xiu!P,!Chai!F!(2014)!Connections!between!physical,!optical!and!biogeochemical!processes!
in!the!Pacific!Ocean.!Prog!Oceanogr!122:30?53!
Zheng!L,!Chen!B,!Liu!X,!Huang!B,!Liu!H,!Song!S!(2015)!Seasonal!variations!in!the!effect!of!
microzooplankton!grazing!on!phytoplankton!in!the!East!China!Sea.!Cont!Shelf!Res!
111:304?315!
! 50!
CHAPTER(TWO.(PHYTOPLANKTON(AND(BACTERIAL(DYNAMICS(ON(THE(CHUKCHI(SEA(SHELF(
DURING(THE(SPRING8SUMMER(TRANSITION(

By Paige E. Connell

Coauthors: Christine Michel, Guillaume Meisterhans, Kevin R. Arrigo, and David A. Caron

! 51!
Abstract
Climate warming is exerting significant change on the physical properties of the Arctic Ocean,
which in turn has marked consequences for the biology of the region. The Chukchi Sea is notable
for its species richness as a consequence of a nutrient-rich shelf region that supports substantial
primary production. However, little is known about the carbon transformations at the base of the
food web in the Chukchi Sea, and in particular the relative amounts of primary production that
are transferred directly to higher trophic levels or remineralized within the microbial loop. In this
study, we measured microbial standing stocks (bacteria to microplankton), phytoplankton growth
and mortality rates (dilution experiments), and bacterial production and mortality rates (
3
H-
leucine incubations and fluorescently-labeled bacteria disappearance experiments, respectively)
at ten stations in the Chukchi Sea and Bering Strait during the spring-summer transition. Our
study revealed that protistan grazers consumed substantially more phytoplankton carbon than
bacterial carbon. Phytoplankton growth rates were variable, but at times considerable (range -
0.06 to 0.71 d
−1
), with protistan grazing consuming an average of 46% of the daily primary
production. Heterotrophic protists exerted significant grazing pressure on phytoplankton despite
low environmental temperatures. Bacterial production and mortality rates were low (generally <
1 µg C l
−1
d
−1
), with production and grazer-mediated mortality more or less in balance. This
study improves our understanding of carbon cycling in the Chukchi Sea during the spring-
summer transition, demonstrating a significant transfer of primary production to heterotrophic
protists and highlighting the importance of the classical food web at this time of year.

! 52!
2.1 Introduction
The Chukchi Sea is a highly dynamic region of the Western Arctic Ocean that supports
some of the highest primary production rates and photosynthetic biomass reported in oceanic
environments (Springer & McRoy 1993, Arrigo et al. 2012, Arrigo et al. 2014). The delicate
interplay of several water masses in the Arctic Ocean and the seasonal progression of the
icescape define the biogeochemistry, species composition, and overall productivity of the region.
Nutrient-rich water enters the Chukchi Sea through the Bering Strait due to a sea-surface height
differential (Weingartner et al. 2005), fueling a seasonal cycle of primary production as light
availability increases during the spring-summer melt season. Additional nutrients for primary
production are brought to the surface by convective mixing on the Chukchi Shelf. This broad,
shallow continental shelf houses a rich benthic ecosystem that supports local marine bird and
mammal populations (Grebmeier 2012). Extensive regional riverine discharge (approximately
4000 km
3
yr
-1
; Shiklomanov 2000) also impacts microbial dynamics in the region by delivering
terrigenous dissolved organic matter, which attenuates light for the primary producers but
provides carbon for the heterotrophic bacterial assemblage (Holmes et al. 2008).
The Arctic Ocean has experienced marked change in recent decades, impacting the
ecosystem-level structure and processes of the Chukchi Sea. Documented changes include
increased atmospheric and water temperatures, decreased ice thickness and extent, reduced
persistence of multi-year ice, stronger wave activity resulting in enhanced coastal erosion, and
predicted increases in precipitation and riverine input (Arrigo 2015, Dickinson et al. 2016). The
increased heat-flux into Arctic waters stimulates a positive feedback loop, whereby warmer
waters further delay the formation of sea ice in the fall, reduce the regional albedo, and enhance
! 53!
the amount of solar energy entering polar waters (Kedra et al. 2015). Given the current warming
trend, the Arctic may become seasonally free as early as 2040 (Wang & Overland 2009).
Changes in the physical properties of the Chukchi Sea have potentially profound
implications for ecosystem function. Ice-free areas tend to have higher area-normalized carbon
fixation rates than ice-covered areas (Brown & Arrigo 2013), thus a decrease in sea-ice extent
may result in higher annual primary production in the Arctic (Arrigo & van Dijken 2015).
Massive phytoplankton blooms have been reported under the ice in recent years because more
light can penetrate the thinning ice pack, challenging the paradigm that polar primary
productivity is primarily confined to marginal ice zones and open waters (Arrigo et al. 2012,
Arrigo et al. 2014, Lowry et al. 2014). Phytoplankton productivity will also be influenced by
changes in nutrient availability, which may increase due to thawing of frozen tundra or enhanced
upwelling of nutrient rich, deep water (Kirchman et al. 2009b). However, nutrient availability in
the Chukchi Sea may also decrease should productivity in the Bering Sea increase, reducing
nutrient concentrations of the water entering the Chukchi Shelf through the Bering Strait
(Kirchman et al. 2009b). It is anticipated that warming will also result in a tighter coupling
between primary producers and herbivorous protists due to differential temperature effects on
phototrophic and heterotrophic protists (Rose & Caron 2007), as well as increased encounter
rates between predators and prey due to greater water column stratification (Behrenfeld & Boss
2014). Overall, climate change is expected to strengthen the ‘microbial loop’ in Arctic
ecosystems, with a greater proportion of primary production providing dissolved organic matter
for the growth of the heterotrophic bacterial assemblage (Kirchman et al. 2009b).
Documenting the present state of microbial community dynamics in the Chukchi Sea is
essential at this time in order to provide a benchmark for future shifts in ecosystem function
! 54!
resulting from climate change. To our knowledge, no studies directly comparing the rates of
growth and protistan grazing mortality for the phytoplankton and heterotrophic bacteria have
been conducted in the Arctic Ocean. In addition, few measurements of microbial rate processes
have been conduced during the spring-summer transition. Protists are the primary source of
mortality for phytoplankton and bacteria in the ocean (Sherr & Sherr 1994, Sherr & Sherr 2002,
Calbet & Landry 2004). Food webs in which large amounts of primary productivity are
consumed directly by protistan grazers (“classical” food webs) are more efficient at transferring
carbon to higher trophic levels than food webs in which a substantial amount of carbon is routed
through the heterotrophic bacterial assemblage (“the microbial loop”; Azam et al. 1983). Direct
comparisons of these trophic interactions thus provide important insight into the efficiency of
planktonic food webs.
In this study, we characterized microbial food web dynamics in the Chukchi Sea during
the spring-summer transition in order to provide a baseline for future climate change. Microbial
abundances (bacteria to microzooplankton), phytoplankton growth and mortality rates, and
bacterial production and mortality rates were measured during May-June 2014 on the Chukchi
Shelf and in the Bering Strait. Measurements were made as part of a larger sampling program,
SUBICE (Study of Under-ice Blooms In the Chukchi Ecosystem), which sought to characterize
the spatial distribution of under-ice phytoplankton blooms on the Chukchi Shelf and the physical,
chemical, and biological mechanisms that control them. Our study revealed that phytoplankton
were the primary food source for protistan grazers during the spring-summer transition in the
Chukchi Sea, as substantially more phytoplankton carbon was consumed than bacterial carbon.
Phytoplankton growth and mortality rates were highly variable in the region, showing no clear
relationship to phytoplankton standing stocks or other environmental variables. Protistan grazing
! 55!
removed an average of 46% of daily primary production, but was capable of removing >100% of
primary production at certain sites, suggesting that top-down control may hinder the formation of
under-ice blooms. Bacterial standing stocks were similar throughout the region, with rates of
bacterial production and grazer-mediated mortality in tight balance. This study highlights the
predominance of the classical food web in the Chukchi Sea during the spring-summer transition,
a food-web dynamic that may shift in our future oceans.

2.2 Materials and methods
2.2.1%Study%site%and%environmental%characterization%
!
Microbial community composition and rate processes were measured from May 16 - June
17, 2014 as part of the SUBICE expedition. Sampling was conducted at ten stations in the Pacific
Sector of the Arctic Ocean aboard the R/V USCG Healy: nine on the Chukchi Shelf and one in
the Bering Strait (Fig. 2.1). Station numbers were kept consistent with the original numbering
scheme from the SUBICE expedition and thus are nonconsecutive (total of 209 station numbers
designated on the cruise).
Water was collected using the Healy’s twelve-position, 30 l Niskin bottle system
mounted on a CTD rosette. This system was equipped with dual temperature and conductivity
sensors (SBE3/SBE4; Sea-Bird Electronics, Bellevue, WA), a 0-10,000 psi Digiquartz pressure
sensor (Paroscientific Inc., Redmond, WA), an oxygen sensor (SBE43; Sea-Bird Electronics), a
fluorometer (ECO-AFL/FL; WetLabs, Philomath, OR), a transmissometer (C-Star; WetLabs), a
PAR sensor (QSP-2300; Biospherical Instruments Inc., San Diego, CA), and an altimeter (PSA-
916; Teledyne Benthos, North Falmouth, MA). Water column depths ranged from 32-63 m (! =
45.7 m). The water column was isothermal at the majority of stations (n = 9), while salinity
! 56!

Figure 2.1. (a) Study location in the Chukchi Sea, Pacific Arctic Ocean, with bathymetry
lines (color bar), and (b) daily SSM/I satellite images of sea ice concentrations (color bar)
on four sampling dates (May 16
th
, June 4
th
, June 8
th
, and June 17
th
, 2014). Green boxes in
(b) denote the area covered by the map in (a). Samples for microbial community
composition and trophic activities were taken from one station in the Bering Strait
(station 8) and nine stations on the Chukchi Shelf. Station numbers follow the SUBICE
expedition site designations. Note the depletion in sea ice cover into early June, followed
by a heavy snowfall on June 8
th
, which recovered the ice-pack of central Chukchi Shelf.

! 57!
increased slightly with depth at six of the stations. A subsurface chlorophyll (Chl-a) maximum,
as visualized using real-time fluorescence data from the CTD downcast, was only detectable at
three stations (29, 132, and 209). Water was collected from the surface (! = 2.6 m) or at the
subsurface Chl-a maximum (station 29, 10 m) on the CTD upcast.
Seawater for the determination of microbial community composition, phytoplankton
growth and grazing mortality rates, and bacterial mortality rates was gently transferred from the
Niskin bottles into acid-rinsed (5% HCl), 23 l polycarbonate carboys using acid-rinsed, silicone
tubing, ensuring that the end of the tubing remained submerged to minimize bubbling that can
damage delicate microzooplankton. Carboys were then transported to a low-light (~10 µmol
photons m
-2
s
-1
), 0°C temperature-controlled room where initial (T
0
) sampling and experimental
set-up were conducted. Water used to determine bacterial production rates was collected directly
from the Niskin bottles into clean, sample-rinsed, 20 ml scintillation vials.
A suite of environmental parameters was measured at each station to examine population
abundances and rates of growth and mortality within the context of local physicochemical
conditions. Temperature and salinity were measured using dual temperature and conductivity
sensors (SBE3/SBE4; Sea-Bird Electronics) on the CTD rosette, with the data quality controlled
post-cruise. Dissolved oxygen concentrations were measured using an automated Winkler
system. Satellite sea ice concentrations were extracted for each station using SSM/I imagery at
25 km resolution as described in Arrigo and Van Dijken (2015). Spearman’s rank order
correlation analysis was conducted to determine significant relationships between the microbial
abundances, phytoplankton growth and mortality rates, and bacterial production and mortality
rates and these environmental parameters.
%
%
! 58!
2.2.2%Community%composition%and%biomass%%
!
Picoplankton abundances (prokaryotic and eukaryotic cells, 0.2-2.0 µm in size) were
determined for each station from triplicate samples collected at the beginning of the dilution
experiments (see section 2.2.3). Sample water was pre-filtered through 20 µm Nitex mesh,
preserved with 1% formalin (final concentration), flash-frozen in liquid nitrogen, and stored at
−80°C until flow cytometric analysis using a FACSCalibur flow cytometer (Becton Dickinson,
San Jose, CA). Abundances of phototrophic picoeukaryotes were enumerated and distinguished
from other picophytoplankton populations using the autofluorescence of the photosynthetic
pigments and forward scatter. Phycoerythrin-containing cells consistent with Synechococcus
were detected at some stations, which is in accordance with previous reports of Synechococcus in
the region (Cottrell & Kirchman 2009, Laney & Sosik 2014). However, abundances were too
low to reliably measure via flow cytometry and thus Synechococcus was not included in our
analyses. Abundances of heterotrophic bacteria (bacteria + archaea, although we considered
archaea to be minor contributors to prokaryotic abundances in our samples, as observed in Wells
and Deming (2003) and Garneau et al. (2006)) were measured by flow cytometry using a
standard SYTO13 (S7575; ThermoFisher Scientific, Waltham, MA) staining procedure (del
Giorgio et al. 1996).
Abundances of nanoplankton (microbial eukaryotes, 2-20 µm in size) were determined
for each station from samples collected at the beginning of the dilution experiments and
preserved with formalin (1% final concentration). Slides for microscopy were prepared using 30
ml aliquots of the preserved seawater filtered down to ~1 ml onto 25 mm diameter, 0.2 µm
blackened polycarbonate filters, and stained with 50 µl of a 1 mg ml
−1
working solution of 4’,6’-
diamidino-2-phenylindole (DAPI; D9542; Sigma-Aldrich, St. Louis, MO). Samples were then
! 59!
filtered down, rinsed with deionized water, and filters were placed onto glass slides with a drop
of immersion oil and a coverslip sealed with clear nail polish. Slides were prepared in triplicate
for each station and stored at −20°C until analysis by epifluorescence microscopy at 1000x
magnification. Phototrophic (possibly mixotrophic) and heterotrophic nanoplankton were
differentiated by the presence or absence of chlorophyll autofluorescence when viewed under
blue-light excitation.
Abundances of microplankton (microbial eukaryotes, 20-200 µm in size) were
determined using inverted light microscopy. Formalin preserved samples (1% final
concentration) of seawater were collected at the beginning of the dilution experiments for each
station. Aliquots (25-250 ml) of the preserved seawater were settled for 24-48 h in Utermöhl
chambers and the abundances of diatoms, dinoflagellates, and ciliates were enumerated at 400x
magnification (Utermöhl 1958).
Cell abundances were converted to carbon biomasses using carbon conversion factors
obtained from the literature for appropriate Arctic ecosystems. Bacterial abundance was
converted to carbon biomass using a conversion factor of 15.2 fg C cell
−1
(Ortega-Retuerta et al.
2012). The abundances of the other microbial groups were converted to carbon biomasses by
applying the conversion factors used in Terrado et al. (2008) for the Beaufort Sea: 0.49 pg C
cell
−1
for phototrophic picoeukaryotes, 5.8 pg C cell
−1
for phototrophic/mixotrophic and
heterotrophic nanoplankton, 242 pg C cell
−1
for ciliates and dinoflagellates, and 113 pg C cell
−1

for diatoms. Dinoflagellates can have a diverse array of nutritional modes (phototrophic,
mixotrophic, heterotrophic; Taylor et al. 2008) that were not distinguished in this study.
Dinoflagellate biomass was thus split evenly between phototrophic and heterotrophic nutritional
modes when calculating total phytoplankton biomass or total heterotrophic grazer biomass. The
! 60!
choice of these carbon conversion factors gave an average C:Chl-a ratio of 41 for our study,
which falls within the range of commonly used C:Chl-a factors for the Western Arctic Ocean (ie.
C:Chl-a of 30 used by Sherr et al. (2009); C:Chl-a of 88.5 used in Ortega-Retuerta et al. (2014)).
Size-fractionated Chl-a was also measured at each station as an additional means of
characterizing the photosynthetic community. Four size fractions (WSW, <200 µm, <20 µm and
<5 µm filtrate) were prepared by sequentially filtering water collected at the beginning of the
experiments through in-line filters equipped with 200 µm, 20 µm, and 5 µm Nitex mesh and
collecting the filtrate in darkened, polycarbonate bottles. Sample water (100 to 250 ml) was then
filtered in triplicate onto 25 mm, GF/F filters (nominal pore size 0.7 µm). Filters were extracted
in 5 ml of 90% acetone for 24 h, in the dark, at 0°C. Samples were processed using a calibrated
Turner 10-AU fluorometer (Turner Designs, San Jose, CA) and the acidification method (Holm-
Hansen et al. 1965, Arar & Collins 1997).

2.2.3%Protistan%growth%and%herbivory%by%dilution%experiments%%
!
The growth and grazing mortality rates of the total phytoplankton community (using Chl-
a as a proxy for phytoplankton biomass) and the phototrophic picoeukaryotes (using flow
cytometry) were determined using a modified dilution method (Landry & Hassett 1982, Landry
et al. 1995). The dilution method enables simultaneous measurement of the nutrient-enriched
growth rates (µ
n
), intrinsic (unenriched) growth rates (µ
0
), and grazing mortality rates (m) of the
phytoplankton community (as well as specific phytoplankton groups) through the sequential
dilution of unfiltered seawater (“whole seawater”, WSW) with 0.2 µm diluent prepared from the
same seawater. After collecting the experimental water in 23 l carboys (see section 2.2.1), the
WSW was joined into an acid-rinsed 50 l carboy to ensure homogeneity throughout the bottles.
! 61!
Diluent was prepared by filtering the WSW through an acid-rinsed 0.2 µm Acropak 1550
Capsule Filter with Supor Membrane (Pall Corporation, Port Washington, NY). A five-point
dilution series (100%, 80%, 60%, 40%, and 20% WSW) was prepared in triplicate in acid-rinsed
1.2 l polycarbonate bottles, with care taken to reduce bubbling that can harm delicate
microzooplankton. Nutrient stock (1 ml) was then added to the bottles in the dilution series at a
final concentration of 10 µM NaNO
3
, 1 µM NH
4
Cl, and 0.7 µM NaP
2
PO
4
·H
2
O to prevent
nutrient-limitation of phytoplankton growth in the more-dilute treatments (Landry et al. 1995). A
treatment of unenriched WSW was also prepared in triplicate to enable the calculation of
intrinsic phytoplankton growth rates (µ
0
) and to assess the impact of nutrient addition on the
phytoplankton assemblage. In addition, <200 µm filtrate was prepared using an acid-rinsed in-
line filter equipped with 200 µm Nitex mesh and incubated in triplicate 1.2 l bottles without the
addition of nutrients to determine the grazing impact of protistan grazers (nano- and
microzooplankton) in the absence of mesozooplankton.
Experimental treatments were prepared at low light in a 0°C cold room immediately
following water collection, and transferred to an on-deck, flow-through incubator maintained at
in situ temperatures. Bottles were incubated for 24-72 h, depending on the initial Chl-a
concentration (Table A3). Neutral-density screening covered the bottles to mimic in situ light
conditions (Table A3), as photo-adaptation of the phytoplankton can result in erroneous
estimations of µ when using Chl-a as a proxy of phytoplankton biomass. The extent of
photoadaptation experienced by the phytoplankton during the incubation period was monitored
using fast repetition rate fluorometry (FRRf) to monitor changes in the maximum efficiency of
photosystem II (Fv/Fm) (Kolber et al. 1998). At the beginning and end of each dilution
experiment, 45 ml aliquots of WSW were collected in darkened 50 ml conical tubes and dark-
! 62!
adapted for 30 min at 0°C. Samples were measured in duplicate on a custom made FRRf (Z.
Kolber). Fv/Fm measurements were corrected for blank effects, with blanks for each sample
prepared by filtering sample water through a 0.2 µm, polycarbonate, syringe-filter before
measurement (Cullen & Davis 2003). Fv/Fm values were relatively high and unwavering
throughout the incubation period in all experiments, indicating that phytoplankton growth rates
were not impacted by light-stress (Table A3).
Chl-a concentrations and phototrophic picoeukaryote abundances were measured in all
dilution experiments as described in section 2.3. Triplicate Chl-a samples and triplicate flow
cytometry samples were analyzed from the initial WSW, <200 µm filtrate, and <0.2 µm filtrate.
Duplicate Chl-a samples and a single flow cytometry sample were analyzed from each of the
triplicate bottles within each treatment at the end of the experiment.
Model I linear regressions of plots of the apparent growth rate (y-axis) against the
dilution factor (x-axis) were used to calculate the nutrient-enriched growth rates (µ
n
; y-intercept
of the regression) and mortality rates (m; slope of the regression) of the total phytoplankton and
phototrophic picoeukaryotic assemblages (Landry & Hassett 1982). Intrinsic growth rates (µ
0
) of
these assemblages were determined from growth in the unenriched treatment and the mortality
rate (Landry et al. 1995). Non-linear dilution curves indicative of grazing saturation were
detected for the total phytoplankton community at stations 29 and 186; for these stations,
regressions to determine µ and m were obtained from the three highest dilution levels (Gallegos
1989). Grazing saturation was not observed for the phototrophic picoeukaryotes. Differences in
the phytoplankton growth rates between the nutrient-enriched and unenriched treatments (µ
n
and
µ
0
), as well as differences in apparent growth rate between the WSW and <200 µm filtrate
unenriched treatments, were assessed using Welch Two-Sample T-Tests.
! 63!
The percentage of primary production removed daily was calculated as m:µ
0
ratio * 100
(Calbet & Landry 2004). The daily percent standing stock removal of the total phytoplankton
(Chl-a-based) and phototrophic picoeukaryotic assemblages (%SS) was determined according to
the following equations,
%SS = G * (100/C
0
)
G = m * C
m

C
m
= C
0
[e
(µ0-m)t
− 1]/(µ
0

− m)t
where µ
0

= the intrinsic growth rate of the phytoplankton (d
−1
), m = the mortality rate of the
phytoplankton (d
−1
), t = length of the incubation (d), C
0

= the initial Chl-a concentration (or
phototrophic picoeukaryote abundance), C
m

= the mean Chl-a concentration (or phototrophic
picoeukaryote abundance) during the incubation, and G = the grazing impact of the consumers
(Calbet & Landry 2004).
%
2.2.4%Bacterial%production%and%bacterivory%rates%%
!
Bacterial production was measured based on the incorporation rates of
3
H-leucine
according to the centrifugation protocol (Smith & Azam 1992).
3
H-leucine (specific activity, 60
Ci mmol
−1
) was added to triplicate, 1.2 ml subsamples at a final concentration of 10 nM,
incubating for 2 h at in situ temperature, and stopping the reaction by adding trichloroacetic acid
(TCA, 5% final concentration). Controls to determine background levels of
3
H-leucine were also
prepared by adding TCA (5% final concentration) to samples immediately to kill prokaryotes.
Samples were stored at −80°C until analysis. Samples were thawed at 4°C, twice centrifuged for
10 min at 14000 rpm to pellet cells, supernatant was decanted, and the pellet was rinsed with 5%
TCA. The pellet was resuspended in 1.5 ml of liquid scintillation cocktail (Ecolume; MP
! 64!
Biomedicals, Santa Ana, CA) and radioactivity was measured using a liquid scintillation counter
(PerkinElmer, Waltham, MA). Data were converted from pmol Leu l
−1
h
−1
to units of bacterial
carbon using a conversion factor of 1.5 kg C mol
−1
of
3
H-leucine, which has previously been
used in the Chukchi Sea and Western Arctic Ocean (Kirchman et al. 2009a, Ortega-Retuerta et
al. 2014).
Bacterial grazing mortality was determined by following the disappearance of
fluorescently-labeled bacteria (FLB) in natural seawater samples. Fluorescently-labeled bacteria
(FLBs) were prepared from a monoculture of Dokdonia donghaensis according to standard
protocol (Sherr et al. 1987, Caron 2001). D. donghaensis was cultured in Zobell medium, then
harvested by centrifugation and resuspended in 0.2 µm filtered seawater. Bacteria were
incubated in the 0.2 µm filtrate for two days to induce cell shrinkage, causing the FLB to better
mimic the bacterial size observed in natural assemblages (rod-shaped, average length = 1 µm).
Bacteria were then stained with 5-(4,6-dichlorotriazin-2-yl) aminofluorescein (DTAF), heat-
killed, rinsed three times, aliquoted, and stored at −80°C until the time of the experiment. FLB
aliquots were prepared in a single batch to ensure homogeneity across experiments
Seawater was collected and experiments prepared alongside the dilution experiments in a
low-light, 0°C cold room. FLB disappearance experiments consisted of triplicate 1.2 l
polycarbonate bottles containing WSW (to assess FLB disappearance attributable to grazing) and
triplicate 1.2 l polycarbonate bottles containing 0.2 µm filtrate (to serve as a control for non-
grazing losses of FLB in the bottles). FLB aliquots were vigorously vortexed, passed through a
3.0 µm polycarbonate filter to remove clumps, and added to each bottle at ~20% of the natural
bacterial abundance (mean T
0
FLB abundance = 1.07 x 10
5
FLB ml
−1
, mean T
0
natural bacterial
abundance = 5.78 x 10
5
cells ml
−1
). Samples for flow cytometry (2 ml, 1% formalin final
! 65!
concentration) were collected at the beginning of each experiment immediately after the addition
of FLB, and at the end of each experiment. Preserved samples were flash frozen in liquid
nitrogen and stored at −80°C until analysis by flow cytometry. FLB disappearance experiments
were incubated alongside the dilution experiments in the on-deck, flow-through incubator and
were exposed to the same incubation temperatures, light conditions, and incubation lengths
detailed in section 2.3. Changes in the FLB abundances in the WSW treatment, taking into
account the non-grazing changes in the control treatment, were used to calculate bacterial
removal due to grazing. Bacterial mortality rate (g; d
−1
) was calculated according to the
following equation,
g = ln(F
t
/F
0
)*(−1/t)
where F
0
= the number of FLB at the beginning of the incubation, F
t
= the number of FLB at the
end of the incubation, and t = the length of the incubation (d) (Marrasé et al. 1992). Bacterial
consumption rate (µg bacterial C consumed ml
−1
d
−1
) was calculated by multiplying the bacterial
grazing mortality rate (d
−1
) by the bacterial standing stock (µg C ml
−1
) at each station.

2.3 Results
2.3.1%Environmental%conditions%during%the%springBsummer%transition%
!
Water temperatures ranged from −1.74 to −0.84°C (! (mean) = −1.58°C) throughout the
study region, and salinities ranged from 31.90 to 32.49 (! = 32.2) (Fig. 2.2). The percentage of
ice cover at each station ranged from 46-100% at the time of sampling, with more than 90% ice
coverage at 7 of the 10 stations (Fig. 2.1b). A late season snowfall (Fig. 2.1; between June 4
th
&
8
th
) delayed the onset of the melt season compared to previous years, reducing the light
availability below the ice and prohibiting the formation of under-ice blooms.
! 66!

Figure 2.2. Mean (vertical black bars) and station-specific (black and gray symbols)
values for environmental parameters measured during the study. Units are denoted
below the parameter name on the y-axis. Environmental parameters include: temperature
(
0
C), salinity (PSS-78), dissolved oxygen (ml l
−1
), T
0
chlorophyll a (µg l
−1
), ammonium (µM),
nitrate (µM), phosphate (µM), and silicate (µM).

! 67!
Waters were relatively replete with macronutrients at most stations during the cruise,
with mean concentrations of 1.03 µM ammonium, 9.87 µM nitrate, 1.59 µM phosphate, and 40.9
µM silicate (Fig. 2.2). Nitrogen concentrations were highly variable, with ammonium
concentrations ranging from 0.05 to 2.75 µM and nitrate concentrations ranging from 0.01 to
13.4 µM between sites. Phosphate was less variable between sites (differences of 2-fold, 0.87 to
1.92 µM), while silicate ranged from 16.4 to 51.5 µM. Dissolved N:P ratios at all sites were low
(! = 6).
Initial

Chl-a concentration among the experiments ranged from 0.04 to 3.49 µg l
−1
(! =
1.10 µg l
−1
) and was positively correlated with oxygen concentration (ρ = 0.91, p < 0.001) and
negatively correlated with ammonium (ρ = −0.78, p < 0.01), phosphate (ρ = −0.81, p < 0.01), and
silicate (ρ = −0.83, p < 0.01) concentration. Chl-a and nitrate concentration were not correlated,
but nitrate drawdown was apparent at the stations with the highest Chl-a concentrations (stations
29 and 132; Fig. 2.2). Chl-a concentration was not correlated with the percentage of ice cover at
each station.
%
2.3.2%Microbial%community%structure%
!
The standing stock of the entire microbial community varied greatly by station, ranging
from 7.29 to 347 µg C l
−1
(Fig. 2.3a). Differences in microbial standing stocks were largely
driven by the high variability in diatom abundances between stations (C.O.V. of 151%); thus
microbial standing stock was lower when Chl-a concentrations were lower. Total microbial
standing stock was negatively correlated with the distance of the station from shore (ρ = −0.81, p
< 0.01). The highest microbial standing stocks observed during our study were located off of
Point Lay, Alaska, at stations 132 and 29. The lowest microbial standing stocks were recorded at
! 68!

Figure 2.3. (a) Absolute (µg C l
−1
) and (b) relative (%) contributions of the microbial
assemblages to living carbon biomass at each station. Stations are arranged by
descending chlorophyll a concentration (left to right). Microbial assemblages measured
include: bacteria, phototrophic picoeukaryotes, phototrophic/mixotrophic nanoplankton,
heterotrophic nanoplankton, ciliates (loricate + aloricate), dinoflagellates, and diatoms.

0.00
50.00
100.00
150.00
200.00
250.00
300.00
350.00
132 29 209 8 186 151 105 93 42 64
Carbon Biomass (µg C l
-1
)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
132 29 209 8 186 151 105 93 42 64
Percentage of Total Microbial Biomass
a"
b"
Bacteria
P. picoeukaryotes
P/M. nanoplankton
H. nanoplankton
Ciliates
Dinoflagellates
Diatoms
Station
Station
! 69!
the most northerly stations (64 and 93; Fig. 2.1). Total microbial standing stock was not
correlated with water column depth or the percentage of ice cover at each station.
Heterotrophic bacterial abundances ranged from 1.8 x 10
5
to 1.1 x 10
6
cells ml
−1
, which
corresponds to bacterial biomass values of 2.68 to 16.3 µg C l
−1
(Fig. 2.3a). Bacteria were
generally <20% of the total microbial biomass throughout our study, but bacteria constituted
more than half of the microbial community biomass at the low Chl-a stations (Fig. 2.3b).
Bacterial abundance was negatively correlated with water column depth (ρ = −0.71, p < 0.05;
Fig. A4).
Phytoplankton community biomass and composition varied greatly between sites. Diatom
biomass ranged from 0.31 to 316 µg C l
−1
, constituting between 4 and 93% of the total microbial
standing stocks among the stations (Figs. 2.3a & 2.3b). Phototrophic (possibly mixotrophic)
nanoplankton and picoeukaryote standing stocks were smaller and more consistent between sites,
ranging from 0.19 to 2.43 µg C l
−1
and 0.02 to 1.57 µg C l
−1
, respectively. Eukaryotic
phytoplankton <20 µm in size never exceeded more than ~5% of the total microbial biomass
(Fig. 2.3b). The predominance of large phytoplankton during our study was also highlighted by
the size-fractionated Chl-a concentrations. Phytoplankton >20 µm (predominantly diatoms)
constituted ≥50% of the total Chl-a at 9 of 10 stations (Fig. 2.4a). Twenty-three distinct diatom
genera were observed in the study region, with the dominant genera highlighted (Fig. 2.4b).
Fragillariopsis and Navicula septentrionalis, which are morphologically similar and were
grouped to reduce misidentification, were the most abundant diatoms at 9 of 10 stations.
Melosira varians was also abundant at half of the stations surveyed, especially those with high
Chl-a concentrations. The abundances of all phytoplankton groups were positively correlated
with oxygen and initial Chl-a concentration and negatively correlated ammonium concentration

! 70!

Figure 2.4. Phytoplankton community composition based on (a) size-fractionated
chlorophyll a (Chl-a), represented as percentage of the total Chl-a concentration detected
in each size fraction, and (b) diatom group abundances, represented as the percentage of
the total diatom community constituted by each group. Stations are arranged by
descending Chl-a concentration (left to right).

0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
132 29 209 8 186 151 105 93 42 64
Percentage of Total Chl a Concentration
n > 200 µm
20 µm < n < 200 µm
5 µm < n < 20 µm
n < 5 µm
Station
Nitzchia frigida
Thalassiosira
Cylindrotheca
Stephanopyxis
Pseudo-nitzschia
Other Pennate
Other Centric
Melosira varians
Fragilariopsis &
Navicula septentrionalis
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
132 29 209 8 186 151 105 93 42 64
Station
a"
b"
Percentage of Total Diatom Abundance
"
! 71!
(p < 0.05; Fig. A4). Diatom and phototrophic picoeukaryote abundances were also negatively
correlated with latitude and concentrations of phosphate and silicate.
The total biomass of heterotrophic grazers ranged from 2.55 to 11.1 µg C l
−1
and
constituted between 6.4 and 37% of total microbial community standing stocks across all stations
(Fig. 2.3). Heterotrophic nanoplankton biomass (range = 2.47 – 9.78 µg C l
−1
) exceeded
microzooplankton biomass (ciliates and dinoflagellates; range = 0.08 – 4.79 µg C l
−1
) at all
stations except station 29, where ciliate biomass was maximum (4.26 µg C l
−1
). Total
dinoflagellate biomass consistently fell below 1 µg C l
−1
(assuming half the cells were
phototrophic, heterotrophic dinoflagellate biomass was < 0.5 µg C l
−1
).
Total heterotrophic grazer biomass was positively correlated with both bacterial biomass
(ρ = 0.82) and phototrophic biomass (ρ = 0.85), as well as initial Chl-a concentration (ρ = 0.78)
(all p < 0.01). Neither heterotrophic nanoplankton biomass nor dinoflagellate biomass were
correlated with prey biomasses. However, ciliate biomass was positively correlated with
bacterial (ρ = 0.65), phototrophic picoeukaryote (ρ = 0.70), and diatom biomass (ρ = 0.78) (all p
< 0.05). Dinoflagellate abundance was also not correlated with any of the environmental factors
measured (Fig. A4). Conversely, heterotrophic nanoplankton abundance was positively
correlated with oxygen concentration (ρ = 0.65, p < 0.05) and negatively correlated with latitude
(ρ = −0.75, p < 0.05) and ammonium concentration (ρ = −0.83, p < 0.01). Ciliate abundance
showed positive correlations to oxygen (ρ = 0.82, p < 0.01) and Chl-a concentration (ρ = −0.68,
p < 0.05) and negative correlations to latitude (ρ = −0.86, p < 0.01), phosphate concentration (ρ =
−0.77, p < 0.01), and silicate concentration (ρ = −0.66, p < 0.05).
%
%
%
! 72!
2.3.3%Microbial%rate%processes%
!
2.3.3.1 Phytoplankton growth and mortality rates
Intrinsic growth rates (unenriched; µ
0
) of the total phytoplankton community (based on
Chl-a) ranged from −0.06 to 0.71 d
−1
(! = 0.47 d
−1
; Table 2.1). Nutrient addition significantly
enhanced phytoplankton growth rates only at stations 132 and 151 (Table A4). Intrinsic
phytoplankton growth rate was not correlated to any of the environmental factors tested,
however, nutrient-enriched growth rate showed a positive correlation with water column depth (ρ
= 0.76, p < 0.05)(Fig. A4).
Mortality rates of the total phytoplankton community ranged from 0 to 0.73 d
−1
(! = 0.25
d
−1
), and accounted for the removal of approximately half of the daily primary production (! =
46 % d
−1
; excluding stn 132 where value >>100%) and approximately one-third of the daily
phytoplankton standing stock (! = 29 % d
−1
). Grazers removed >100% of the daily primary
production at the stations with the highest Chl-a concentrations (stations 29 and 132), which
were both located closest to land near Point Lay, Alaska. Interestingly, the intrinsic
phytoplankton growth rate was high at station 29 (0.70 d
−1
) which showed moderate drawdown
of nitrate and silicate concentrations, but negligible at station 132 (−0.06 d
−1
) which had the
lowest nitrate concentration and highest Chl-a value (Table 2.1; Fig. 2.2). Phytoplankton
mortality rates were non-significant at two stations (93 and 209), which appeared unrelated to
Chl-a concentration at those stations (Table 2.1). Phytoplankton mortality rate was not correlated
to any of the environmental factors measured during this study (Fig. A4).
Intrinsic growth rates of the phototrophic picoeukaryotes were highly variable, ranging
from −0.41 to 0.92 d
−1
(! = 0.28 d
−1
; Table 2.1). Nutrient addition did not significantly affect the
growth rates of the phototrophic picoeukaryotes in any of the experiments conducted (Table A4).
! 73!
Table 2.1. Initial chlorophyll a (Chl-a) concentrations (µg l
−1
), temperatures (°C), grazing mortality rates (m), nutrient-
enriched growth rates (µ
n
), and intrinsic (unenriched) growth rates (µ
0
) for the total phytoplankton community (based on
Chl-a) and the phototrophic picoeukaryotes (based on flow cytometry) for each station sampled during the study period.
Grazing rates (m) and nutrient-enriched growth rates (µ
n
) were calculated using Model I Linear Regressions. Intrinsic
population growth rates (µ
0
) were adjusted from µ
n
values. Mean values (± standard deviations) of each parameter are
contained in the bottom row. Asterisks indicate significance of the Model I Linear Regression at the denoted p-value.

Total Phytoplankton (chl a) Phototrophic Picoeukaryotes
Station Date
T
0
Chl-a (µg
l
−1
)
Temp (°C) m (d
−1
) µ
n
(d
−1
) µ
0
(d
−1
) m (d
−1
) µ
n
(d
−1
) µ
0
(d
−1
)
8 16-May-14 1.67 −1.34 0.32** 0.44 0.71 ns 0.42 0.62
29 20-May-14 1.91 −1.74 0.73*** 0.33 0.70 ns −0.35 −0.41
42 23-May-14 0.04 −1.71 0.1*** 0.33 0.32 ns −0.05 0.06
64 27-May-14 0.04 −1.69 0.26*** 0.44 0.50 0.43* 0.12 0.36
93 01-June-14 0.06 −1.72 ns 0.54 0.57 0.66*** 0.63 0.92
105 04-June-14 0.34 −1.73 0.18*** 0.30 0.26 ns 0.06 0.09
132 08-June-14 3.49 −0.84 0.19** 0.20 −0.06 ns 0.21 0.51
151 10-June-14 0.85 −1.72 0.31*** 0.55 0.67 0.36*** 0.20 0.31
186 14-June-14 0.90 −1.69 0.41** 0.42 0.61 0.55** 0.17 0.17
209 17-June-14 1.71 −1.59 ns 0.40 0.45 ns 0.10 0.16

Mean
(± STDEV)
1.10
(± 1.11)
−1.58
(± 0.28)
0.25
(± 0.22)
0.40
(± 0.11)
0.47
(± 0.24)
0.20
(± 0.27)
0.15
(± 0.26)
0.28
(± 0.36)
Model I linear regression was significant at * = p ≤ 0.1, ** = p ≤ 0.05, *** = p ≤ 0.01, ns = non-significant
! 74!
Phototrophic picoeukaryotic growth rates (both µ
0
and
µ
n
) were positively correlated to water
column depth
(ρ = 0.73 & 0.76, respectively, p < 0.05) (Fig. A4).
Mortality rates of the phototrophic picoeukaryotes were largely non-significant

(Table
2.1). However, at the four stations with detectable phototrophic picoeukaryotic mortality rates,
mortality rates yielded an overall average of 0.20 d
−1
(0.36 to 0.66 d
−1
). Grazers removed
between 72 and 324 % of the daily phototrophic picoeukaryotic production and between 2.2 and
39 % of the phototrophic standing stock per day at the stations with detectable grazing mortality.

Mortality rates of the phototrophic picoeukaryotes exceeded those of the total phytoplankton
community (based on Chl-a) when detected (Table 2.1). Phototrophic picoeukaryotic mortality
rate was positively correlated with latitude (ρ = 0.73, p < 0.05) (Fig. A4).
Generally, removing metazoan grazers did not impact the apparent growth rates of the
phytoplankton of phototrophic eukaryotes; however, apparent growth rates in the <200 µm,
unenriched treatment (without metazoan grazers) were significantly higher than apparent growth
rates in the 100% WSW, unenriched treatment (with metazoan grazers) for the total
phytoplankton community at station 42 and for the phototrophic picoeukaryotes at stations 186
and 209 (Table A4). Ciliate biomass and phytoplankton mortality rate were positively correlated
(r
2
= 0.56, p < 0.05). No significant relationship could be detected between dinoflagellate
biomass or heterotrophic nanoplankton biomass and the mortality rates of the total phytoplankton
community or the phototrophic picoeukaryotes.

2.3.3.2 Bacterial production and mortality rates
Bacterial production rates ranged from 0.10 to 1.38 µg C l
−1
d
−1
(! = 0.52 d
−1
). Both
bacterial standing stocks and rates of bacterial production were significantly higher at stations
! 75!
that were closer to the shore (Figs. 2.5a & 2.5b; p < 0.05). Bacterial production rate also
correlated positively with Chl-a concentration (Fig. 2.5c; p < 0.001), with the highest production
rate measured at station 132 and the lowest rate measured at station 93.
Bacterial grazing mortality rates were persistently low throughout the study region,
ranging from 0.01 to 0.11 d
−1
(! = 0.03 d
−1
). The mortality rate in the Bering Strait (0.11 d
−1
;
station 8) was higher than the mortality rates measured on the Chukchi Shelf (0.00 to 0.04 d
−1
).
Bacterial grazing mortality rate was positively correlated with initial oxygen (ρ = 0.70, p < 0.05)
and Chl-a (ρ = 0.76, p < 0.05) concentrations, and negatively correlated with latitude (ρ = −0.6, p
< 0.05) and silicate concentration (ρ = −0.78, p < 0.01). Bacterial mortality rate was not
correlated with total heterotrophic grazer biomass or with the biomass of individual grazer
groups.
Rates of carbon production by bacteria and bacterial carbon consumption by grazers at
each station were relatively well-balanced, albeit, the positive relationship between production
and consumption was not significant (p = 0.08; Fig. 2.5d). Net changes in bacterial standing
stocks, based on a comparison of bacterial production and consumption at each station, were also
small, ranging from −1.9% to +9.9% at a given station (Fig. 2.6). This comparison indicated
increases in bacterial standing stocks at 6 of 10 stations, with the greatest percent increase
occurring at station 151. Bacterial standing stocks exhibited the greatest percent decrease in the
Bering Strait (station 8), where the rates of bacterial mortality were much higher than those
reported on the Chukchi Shelf.

! 76!

Figure 2.5. Linear regressions of (a) bacterial standing stock vs. minimum distance of the
station from shore, (b) bacterial production vs. minimum distance of the station from
shore, (c) bacterial production vs. chlorophyll a concentration, and (d) bacterial
production vs. bacterial consumption.

! 77!

Figure 2.6. Mean bacterial production (µg C l
−1
d
−1
; dark gray) and bacterial consumption
(µg C l
−1
d
−1
; light gray) rates at each station. Bacterial standing stocks (µg C l
−1
) are
shown above the histograms for each station, just below the station number. The
percentage of the bacterial standing stocks attributable to bacterial production and
bacterial consumption is denoted below each respective column, with changes in
standing stocks calculated as the sum of these values. Stations are arranged by
descending chlorophyll a concentration (left to right).

!
! 78!
2.3.4%Carbon%flow%to%higher%trophic%levels%
!
Daily phytoplankton carbon removal by grazers was quite variable between sites, ranging
from 0 to 76.7 µg C l
−1
d
−1
(! = 19.1 µg C l
−1
d
−1
), while daily bacterial carbon removal by
grazers was much more constant in comparison, ranging from 0.11 to 0.65 µg C l
−1
d
−1
(! = 0.31
µg C l
−1
d
−1
). Substantially more phytoplankton carbon was consumed daily than bacterial carbon
at most stations where phytoplankton grazing mortality rates could be detected (Fig. 2.7; Table
2.1). The average ratio of daily phytoplankton carbon consumption to bacterial carbon
consumption was 62 during the study.

2.4 Discussion
This study provides unique insight into microbial community composition and rate
processes during the spring-summer transition in the Chukchi Sea, a season which has been
historically under-sampled due to heavy ice cover. Phytoplankton intrinsic growth rates (µ
0
)
were highly variable (range of −0.06 to 0.71 d
−1
; Table 2.1) across the shelf during this period.
The mean µ
0
reported in our study (0.47 d
−1
) was higher than other mean µ
0
reported for the
Chukchi and Bering Seas during the spring-summer transition (means of 0.20 d
−1
(Sherr et al.
2009), and 0.19 d
−1
(Sherr et al. 2013) for growth rates measured during April-June), but were
well within the range of values reported in polar regions (Schmoker et al. 2013). Surprisingly, µ
0

was not correlated to any of the environmental parameters tested, such as nutrient concentration
or ice coverage (Fig. A4). Phytoplankton growth rates appeared not to be limited by
macronutrients on the shelf, as nutrient amendments did not significantly increase phytoplankton
growth rates in the majority of experiments (Table A4). In addition, F
v
/F
m
values measured at the
beginning of the dilution experiments were indicative of a healthy mixed phytoplankton
! 79!

Figure 2.7. Mean daily carbon removal (µg C l
−1
d
−1
) by protistan grazers from the
phytoplankton (gray columns) and heterotrophic bacterial (black columns) assemblages
at each station. Stations are arranged by descending chlorophyll a concentration (left to
right).

0
10
20
30
40
50
60
70
80
132 29 209 8 186 151 105 93 42 64
Station
Daily Carbon Removal (µg C l
-1
d
-1
)
Phytoplankton C
Bacterial C
0.0
0.2
0.4
0.6
0.8
1.0
132 29 209 8 186 151 105 93 42 64
! 80!
assemblage (Table A3) (Suggett et al. 2009). We speculate that light limitation was the primary
factor controlling phytoplankton growth during the spring-summer transition.
Massive under-ice phytoplankton blooms have been reported in the Chukchi Sea (Arrigo
et al. 2012, Arrigo et al. 2014), a phenomenon that has been detected annually in the satellite
record from 1998 to 2012 (Lowry et al. 2014). These blooms have been attributed to increased
light availability for phytoplankton in recent decades as a result of prevailing thin, first-year ice
cover and extensive melt pond formation both allowing greater light penetration into the water
column. Under-ice blooms were not detected in the present study. A late season snowfall during
the cruise delayed the formation of melt ponds and subsequent alleviation of light limitation that
may have resulted in the establishment of under-ice blooms, agreeing with previous results
(Fortier et al. 2002) (Fig. 2.1b). The highest Chl-a concentration reported in this study (3.49 µg
l
−1
) was in a marginal ice zone (station 132; 52% ice coverage) near Point Lay, Alaska. High
Chl-a and diatom abundances have been reported previously near Point Lay in May-June 2002
(Sukhanova et al. 2009), as coastal regions along the eastern border of the Chukchi Sea tend to
experience early season ice melt due to the heat content of the Alaska Coastal Current (Wood et
al. 2015). Phytoplankton growth rates were negative (−0.06 d
−1
) and protistan grazing rates were
positive (0.19 d
−1
) at station 132 (Table 2.1), suggesting we were measuring the phytoplankton
population during its decline. In addition, nitrate concentrations were below the detection limit at
this site (Fig. 2.2). Alaskan coastal waters are characteristically nutrient-poor due to the
biological drawdown of nutrients in the Bering Sea and input of river runoff (Arrigo et al. 2014);
thus, the low nutrient concentrations measured at this site presumably reflect a combination of
upstream and localized phytoplankton uptake. Nitrate drawdown below the limit of detection at
this site supports observations that nitrogen is the limiting nutrient in the Chukchi Sea (Sakshaug
! 81!
2004). It is possible that nitrogen availability will become the primary factor limiting
phytoplankton growth in the future should climate change continue to result in reduced ice
coverage and ice thickness. However, it is unclear whether nitrogen availability will increase in
the future Chukchi Sea due to increased upwelling of nutrient-rich, deep water, or if nitrogen
availability will decrease due to increased productivity upstream in the Bering Sea (Kirchman et
al. 2009b).
Phytoplankton growth rates were significantly higher than bacterial growth rates during
the spring-summer transition on the Chukchi Shelf (means of 0.47 d
−1
and 0.06 d
−1
,
respectively). Bacterial abundances were in good agreement with previous reports in Arctic
regions (! = 5.8 x 10
5
cells ml
−1
) (Garneau et al. 2008, Terrado et al. 2008, Ortega-Retuerta et al.
2014), as were bacterial production rates (range = 0.10-1.38 µg C l
−1
d
−1
; Fig. 2.6) and bacterial
growth rates (range = 0.01-0.12 d
−1
) (Kirchman et al. 2005, Garneau et al. 2008, Kirchman et al.
2009a, Kirchman et al. 2009b, Ortega-Retuerta et al. 2014). Low bacterial growth rates during
the spring-summer transition ostensibly resulted from low environmental temperatures, which
have been shown to limit bacterial production in polar regions compared to temperate or
subtropical regions (Kirchman et al. 2009b). In addition, extensive ice and snow cover limited
phytoplankton production and riverine input at this time of year, presumably restricting the
availability of labile organic matter for bacterial growth. A significant positive relationship
between bacterial production rate and Chl-a concentration, and significant negative relationship
between bacterial production rate and the distance of the station from shore (Fig. 2.5), support
the contention that dissolved organic matter limited bacterial growth rates during our study, as
was observed for under-ice Arctic microbial communities during spring in Niemi et al. (2014).
! 82!
Phytoplankton biomass greatly exceeded bacterial biomass at stations with Chl-a
concentrations >0.5 µg l
−1
(six stations; Fig. 2.3 & Table 2.1), while bacterial biomass was
somewhat greater than phytoplankton biomass at the stations with low Chl-a concentrations (1.4
to 9.8x higher). Diatoms persistently dominated the phytoplankton assemblage, as evidenced by
their contribution to community biomass (Fig. 2.3) and Chl-a in the larger size-fractions (Fig.
2.4). However, diatom absolute abundance varied 1000-fold between sampling locations,
contributing to a patchy distribution of microbial biomass across the shelf. The predominance of
diatoms on the Chukchi Shelf has been well documented in the literature (Sherr et al. 2009,
Sukhanova et al. 2009, Laney & Sosik 2014, Yang et al. 2014) and presumably reflects the influx
of nutrient-replete waters through the Bering Strait (Fig. 2.2) (Weingartner et al. 2005).
Phototrophic picoeukaryotes were always a minor component of the phytoplankton
community in our study. Cell abundances averaged 1000 cells ml
−1
and made meager
contributions to total photosynthetic biomass (! = 2.9%; Fig. 2.3). This study, to our knowledge,
provides the first early season (May-June) assessment of phototrophic picoeukaryote growth and
mortality rates in the Chukchi Sea. The mean phototrophic picoeukaryote intrinsic growth rate at
this time of year (0.28 d
−1
; Table 2.1) was similar to those reported for other Western Arctic
Regions (0.22 and 0.24 d
−1
) (Liu et al. 2002, Strom & Fredrickson 2008) as well as to the mean
phototrophic picoeukaryote growth rate found in late summer on the Chukchi Shelf (0.39 d
−1
)
(Yang et al. 2014). The low biomass of the phototrophic picoeukaryotes, coupled to non-
significant grazing rates at most stations (Table 2.1), indicate that these picophytoplankton did
not play a major role in food web dynamics of the Chukchi Sea during the spring-summer
transition of 2014.
! 83!
Phytoplankton mortality rates were generally higher (! = 0.25 d
−1
; Table 2.1) and more
spatially variable (range = non-significant to 0.73 d
−1
) than bacterial mortality rates during this
study (! = 0.03 d
−1
; range = 0.01 to 0.11 d
−1
). Other reported phytoplankton mortality rates
measured during the spring-summer transition in the Chukchi and Bering Sea are lower than
those reported in this study (! = 0.07 d
−1
(Sherr et al. 2009), and ! = 0.09 d
−1
(Sherr et al. 2013)).
One possible explanation for the low mortality rates reported in Sherr et al. (2009) was the
presence of senescent phytoplankton cells in their experiments, which may have significantly
reduced measured grazing rates (Calbet et al. 2011). Rates of phytoplankton mortality in polar
regions tend to be lower than those found in warmer oceanic regions (Calbet & Landry 2004,
Schmoker et al. 2013), a trend which may be due in part to the differential effect of temperature
on the maximal growth rates of phototrophic and herbivorous protists (Rose & Caron 2007). We
did, however, observe stations in which protistan grazing removed more than 100% of the daily
primary production (stations 29 and 132; Table 2.1). These findings suggest that protistan
grazers do have the potential to exert significant top-down control on phytoplankton production
during the spring-summer transition despite low environmental temperatures.
Relatively few studies have quantified bacterial mortality in polar regions. In Franklin
Bay, Northwest Territories, Canada, Vaqué et al. (2008) measured a mean bacterivory rate of
0.29 µg C l
−1
d
−1
in the spring, on the same order as the mean reported in this study (0.31 µg C l
−1
d
−1
). Sanders and Gast (2012) also measured low rates in the Beaufort Sea and Canada Basin,
where bacterivory removed an average of <5% of the bacterial standing stock per day. Thus, it
appears that bacterial mortality rates are depressed in polar regions relative to lower latitudes
during the spring-summer transition, as bacterial standing stock removal is often ~30% per day
in temperate-to-subtropical regions (Marrasé et al. 1992, Fuhrman & Noble 1995, Boras et al.
! 84!
2009, Connell et al. 2017). We speculate that a combination of temperature and bacterial
abundances that are close to the feeding threshold for the bactivorous protists resulted in the low
rates of bacterial mortality measured during the spring-summer transition in the Chukchi Sea.
The possibility of a threshold grazing response is supported by the relatively tight coupling
between bacterial production and bacterial mortality in our study (Figs. 2.5 & 2.6), as well the
observed increase in bacterial grazing mortality rates with increased bacterial abundance on the
Chukchi Shelf.
Phytoplankton were the dominant carbon source for protistan grazers in the Bering Strait
and Chukchi Sea during our study (Fig. 2.7). Protistan consumption of bacterial carbon remained
relatively low and constant among sites (<1 µg C l
−1
d
−1
; Figs. 2.6 & 2.7), while protistan
consumption of phytoplankton carbon was highly variable but generally much larger (0 – 77 µg
C l
−1
d
−1
; Fig. 2.7). The significant contribution of phytoplankton carbon to the overall diet of
protistan grazers was due to a combination of higher phytoplankton biomass and higher
phytoplankton mortality rates than those measured for the bacteria.
We demonstrated a minor contribution of bacterial carbon to protists during our study,
but hypothesize that bacterial carbon contributes a greater percentage of total carbon consumed
by protistan grazers as the season progresses into summer. Dilution experiment data compiled for
the Western Arctic Ocean indicate that the mean monthly phytoplankton intrinsic growth rate
does not change significantly from spring (April µ
0
! = 0.20 d
−1
) to mid-summer (June µ
0
! =
0.25 d
−1
) to late summer (August µ
0
! = 0.22 d
−1
) (this study; Sherr et al. 2009, Sherr et al. 2013,
Yang et al. 2014). A constant mean phytoplankton intrinsic growth rate may reflect a relatively
quick transition from a light-limiting environment for phytoplankton in the spring to a nutrient-
limiting environment in the summer. In contrast, Kirchman et al. (2009a) observed that bacterial
! 85!
production rates were three-fold greater during the summer than during the spring in the Chukchi
Sea. Increased bacterial production rates in the summer, coupled with relatively constant
phytoplankton growth rates, would result in bacterial prey carbon contributing a larger
proportion of consumer diet in the summer season.
Microzooplankton dominated phytoplankton and bacterial mortality, however, the
relative importance of the different protistan grazer groups in the consumption of phytoplankton
and bacteria during the spring-summer transition remains unclear. Total heterotrophic grazer
biomass constituted between 6.4 and 37% of the total microbial standing stock (Fig. 2.3) and was
positively correlated with phytoplankton prey biomass and ambient Chl-a concentration. Despite
the dominance of large, chain-forming diatoms on the shelf (Figs. 2.3 & 2.5), ciliates were the
only grazer assemblage whose biomass was significantly correlated with diatom biomass. Mean
ciliate biomass (1.01 µg C l
−1
) was approximately eight times greater than mean heterotrophic
dinoflagellate biomass (0.13 µg C l
−1
, assuming half of the dinoflagellates to be heterotrophic).
Heterotrophic dinoflagellates are generally considered to be the principal protistan consumer of
diatoms (Sherr & Sherr 2007), however, Sherr et al. (2013) and references therein have reported
episodic importance of large ciliates as grazers on diatom genera in both polar and temperate
seas. It is possible that ciliates were consuming diatoms during our study, potentially grazing
down single-cells and leaving behind the large diatom chains that were observed. Alternately,
dinoflagellates may have been the primary grazers of diatoms in the study, but their abundances
were lower than those of ciliates as a result of preferential grazing by mesozooplankton on the
dinoflagellates.
Mesozooplankton grazing and viral lysis are potentially important sources of mortality in
some marine environments. Growth rates of the phytoplankton and phototrophic picoeukaryotes
! 86!
were only significantly higher at one (station 42) and two (stations 186 and 209) stations,
respectively, when mesozooplankton were excluded from the incubation bottles (Table A4).
Thus, exclusion of mesozooplankton from the dilution experiments had little impact on the
phytoplankton mortality rates in our study, although avoidance by mesozooplankton of our
collection method (CTD) may have occurred. The contribution of mesozooplankton grazing to
overall phytoplankton mortality is thought to be low during the spring in the Western Arctic
ocean, with mesozooplankton demonstrating a prey preference for microzooplankton over
diatoms when available (Campbell et al. 2009). Viral lysis of bacteria has been found to be
comparable to protistan grazing in the late summer in the Chukchi Sea, removing ~10-25% of
bacterial production (Steward et al. 1996). In our study, bacterial production was almost
completely removed by protistan grazing (Figs. 2.5d & 2.6), suggesting that viruses were not a
dominant source of bacterial mortality at that time.
This study, to our knowledge, presents the first direct comparison of phytoplankton and
bacterial carbon consumption in a marine polar ecosystem. An understanding of these processes
is fundamental to our assessments of how much carbon is available to higher trophic levels or for
export out of the euphotic zone. Classical food webs, in which the majority of carbon goes
directly from the primary producers to herbivorous zooplankton, have greater trophic transfer
efficiency than microbially-dominated food webs, in which a considerable portion of the carbon
is remineralized and respired by the heterotrophic bacterial assemblage (Azam et al. 1983). The
prevalence of the classical food web in the Arctic provides the energy necessary to foster a wide
variety of invertebrates, fish, seabirds, and marine mammals (Grebmeier 2012). Nonetheless,
species size and composition may shift as a consequence of climate change. For example,
phytoplankton cell size decreased in the Canada basin in response to increased seawater
! 87!
temperatures, freshwater input, and stratification (Li et al. 2009). Our study provides a baseline
for Chukchi Sea food web dynamics, which may experience similar shifts in food web structure
in the coming years due to climate change.

2.5 Conclusion
An investigation of microbial predator-prey interactions in the Chukchi Sea revealed that
phytoplankton, particularly diatoms, were the primary source of prey carbon for higher trophic
levels during the spring-summer transition. Bacterial contributions to higher trophic levels were
small during our study (<1 µg C l
−1
d
−1
), as bacterial abundances and production rates were
restricted by a combination of temperature and dissolved organic matter availability early on in
the productive season. Measurements of phytoplankton growth and mortality rates revealed a
variable, yet sometimes active, phytoplankton community that was occasionally subject to
intense protistan grazing pressure (protistan grazing removing more than 100% of daily primary
production) despite low environmental temperatures; these findings suggest that protistan grazers
may play a role in top-down control on the formation of blooms in the marginal ice zones as well
as under heavy-ice cover. The bacterial community, conversely, exhibited relatively constant
abundances in our study region and bacterial mortality rates closely paralleled bacterial
production rates. Our study demonstrates that the classical food web, in which phytoplankton
carbon is directly transferred to higher trophic levels, dominates the Chukchi Sea early in the
productive season. However, food-web dynamics during the spring-summer transition in the
Chukchi Sea may shift as a result of climate change. The microbial biomass and carbon turnover
data presented here provide a baseline assessment of food-web structure in the Chukchi Sea and
! 88!
facilitate the creation of mathematical models aimed at understanding the impacts climate change
on the Arctic Ocean.

! 89!
Chapter Two Literature Cited
Arar!EJ,!Collins!GB!(1997)!Method!445.0:!In!vitro!determination!of!chlorophyll!a!and!
pheophytin!a!in!marine!and!freshwater!algae!by!fluorescence.!United!States!
Environmental!Protection!Agency,!Office!of!Research!and!Development,!National!
Exposure!Research!Laboratory!Ohio!
Arrigo!KR!(2015)!Impacts!of!Climate!on!EcoSystems!and!Chemistry!of!the!Arctic!Pacific!
Environment!(ICESCAPE).!Deep!Sea!Res!II!118,!Part!A:1X6!
Arrigo!KR,!Perovich!DK,!Pickart!RS,!Brown!ZW,!van!Dijken!GL,!Lowry!KE,!Mills!MM,!Palmer!
MA,!Balch!WM,!Bahr!F!(2012)!Massive!phytoplankton!blooms!under!Arctic!sea!ice.!
Science!336:1408X1408!
Arrigo!KR,!Perovich!DK,!Pickart!RS,!Brown!ZW,!van!Dijken!GL,!Lowry!KE,!Mills!MM,!Palmer!
MA,!Balch!WM,!Bates!NR!(2014)!Phytoplankton!blooms!beneath!the!sea!ice!in!the!
Chukchi!Sea.!Deep!Sea!Res!II!105:1X16!
Arrigo!KR,!van!Dijken!GL!(2015)!Continued!increases!in!Arctic!Ocean!primary!production.!
Prog!Oceanogr!136:60X70!
Azam!F,!Fenchel!T,!Field!J,!Gray!J,!MeyerXReil!L,!Thingstad!F!(1983)!The!ecological!role!of!
waterXcolumn!microbes!in!the!sea.!Mar!Ecol!Prog!Ser!10:257X263!
Behrenfeld!MJ,!Boss!ES!(2014)!Resurrecting!the!ecological!underpinnings!of!ocean!
plankton!blooms.!Annu!Rev!Mar!Sci!6:167X194!
Boras!JA,!Sala!MM,!VázquezXDomínguez!E,!Weinbauer!MG,!Vaqué!D!(2009)!Annual!changes!
of!bacterial!mortality!due!to!viruses!and!protists!in!an!oligotrophic!coastal!
environment!(NW!Mediterranean).!Environ!Microbiol!11:1181X1193!
Brown!ZW,!Arrigo!KR!(2013)!Sea!ice!impacts!on!spring!bloom!dynamics!and!net!primary!
production!in!the!Eastern!Bering!Sea.!J!Geophys!ResXOceans!118:43X62!
Calbet!A,!Landry!MR!(2004)!Phytoplankton!growth,!microzooplankton!grazing,!and!carbon!
cycling!in!marine!systems.!Limnol!Oceanogr!49:51X57!
Calbet!A,!Saiz!E,!Almeda!R,!Movilla!JI,!Alcaraz!M!(2011)!Low!microzooplankton!grazing!
rates!in!the!Arctic!Ocean!during!a!Phaeocystis!pouchetii!bloom!(Summer!2007):!fact!
or!artifact!of!the!dilution!technique?!J!Plankton!Res:fbq142!
Campbell!RG,!Sherr!EB,!Ashjian!CJ,!Plourde!S,!Sherr!BF,!Hill!V,!Stockwell!DA!(2009)!
Mesozooplankton!prey!preference!and!grazing!impact!in!the!western!Arctic!Ocean.!
Deep!Sea!Res!II!56:1274X1289!
Caron!DA!(2001)!Protistan!herbivory!and!bacterivory.!In:!Paul!J!(ed)!Methods!in!
Microbiology,!Book!30.!Academic!Press!Ltd.,!Cambridge,!UK!
Connell!PE,!Campbell!V,!Gellene!AG,!Hu!SK,!Caron!DA!(2017)!Planktonic!food!web!structure!
at!a!coastal!timeXseries!site:!II.!Spatiotemporal!variability!of!microbial!trophic!
activities.!Deep!Sea!Res!I!121:210X223!
Cottrell!MT,!Kirchman!DL!(2009)!Photoheterotrophic!microbes!in!the!Arctic!Ocean!in!
summer!and!winter.!Appl!Environ!Microbiol!75:4958X4966!
Cullen!JJ,!Davis!RF!(2003)!The!blank!can!make!a!big!difference!in!oceanographic!
measurements.!Limnol!Oceangr!Bulletin!12:29X34!
del!Giorgio!PA,!Bird!DF,!Prairie!YT,!Planas!D!(1996)!Flow!cytometric!determination!of!
bacterial!abundance!in!lake!plankton!with!the!green!nucleic!acid!stain!SYTO!13.!
Limnol!Oceanogr!41:783X789!
! 90!
Dickinson!I,!Walker!G,!Pearce!DA!(2016)!Microbes!and!the!Arctic!Ocean.!In:!Hurst!JC!(ed)!
Their!World:!A!Diversity!of!Microbial!Environments.!Springer!International!
Publishing,!Cham!
Fortier!M,!Fortier!L,!Michel!C,!Legendre!L!(2002)!Climatic!and!biological!forcing!of!the!
vertical!flux!of!biogenic!particles!under!seasonal!Arctic!sea!ice.!Mar!Ecol!Prog!Ser!
225:1X16!
Fuhrman!JA,!Noble!RT!(1995)!Viruses!and!protists!cause!similar!bacterial!mortality!in!
coastal!seawater.!Limnol!Oceanogr!40:1236X1242!
Gallegos!CL!(1989)!Microzooplankton!grazing!on!phytoplankton!in!the!Rhode!River,!
Maryland:!Nonlinear!feeding!kinetics.!Mar!Ecol!Prog!Ser!57:23X33!
Garneau!MXÈ,!Vincent!WF,!AlonsoXSáez!L,!Gratton!Y,!Lovejoy!C!(2006)!Prokaryotic!
community!structure!and!heterotrophic!production!in!a!riverXinfluenced!coastal!
arctic!ecosystem.!Aquat!Microb!Ecol!42:27X40!
Garneau!MÈ,!Roy!S,!Lovejoy!C,!Gratton!Y,!Vincent!WF!(2008)!Seasonal!dynamics!of!bacterial!
biomass!and!production!in!a!coastal!arctic!ecosystem:!Franklin!Bay,!western!
Canadian!Arctic.!J!Geophys!ResXOceans!113!
Grebmeier!JM!(2012)!Shifting!Patterns!of!Life!in!the!Pacific!Arctic!and!SubXArctic!Seas.!
Annu!Rev!Mar!Sci!4:63X78!
HolmXHansen!O,!Lorenzen!CJ,!Holmes!RW,!Strickland!JDH!(1965)!Fluorometric!
Determination!of!Chlorophyll.!Journal!du!Conseil!30:3X15!
Holmes!RM,!McClelland!JW,!Raymond!PA,!Frazer!BB,!Peterson!BJ,!Stieglitz!M!(2008)!
Lability!of!DOC!transported!by!Alaskan!rivers!to!the!Arctic!Ocean.!Geophys!Res!Lett!
35:1X5!
Kedra!M,!Moritz!C,!Choy!ES,!David!C,!Degen!R,!Duerksen!S,!Ellingsen!I,!Górska!B,!Grebmeier!
JM,!Kirievskaya!D,!van!Oevelen!D,!Piwosz!K,!Samuelsen!A,!Węsławski!JM!(2015)!
Status!and!trends!in!the!structure!of!Arctic!benthic!food!webs.!Polar!Research!34!
Kirchman!DL,!Hill!V,!Cottrell!MT,!Gradinger!R,!Malmstrom!RR,!Parker!A!(2009a)!Standing!
stocks,!production,!and!respiration!of!phytoplankton!and!heterotrophic!bacteria!in!
the!western!Arctic!Ocean.!Deep!Sea!Res!II!56:1237X1248!
Kirchman!DL,!Malmstrom!RR,!Cottrell!MT!(2005)!Control!of!bacterial!growth!by!
temperature!and!organic!matter!in!the!Western!Arctic.!Deep!Sea!Res!II!52:3386X
3395!
Kirchman!DL,!Morán!XAG,!Ducklow!H!(2009b)!Microbial!growth!in!the!polar!oceans—role!
of!temperature!and!potential!impact!of!climate!change.!Nat!Rev!Microbiol!7:451X459!
Kolber!ZS,!Prášil!O,!Falkowski!PG!(1998)!Measurements!of!variable!chlorophyll!
fluorescence!using!fast!repetition!rate!techniques:!defining!methodology!and!
experimental!protocols.!BBAXBioenergetics!1367:88X106!
Landry,!R.!M,!Kirshtein,!J.,!Constantinou,!J.!(1995)!A!refined!dilution!technique!for!
measuring!the!community!grazing!impact!of!microzooplankton,!with!experimental!
tests!in!the!central!equatorial!Pacific.!Mar!Ecol!Prog!Ser!120:53X63!
Landry!MR,!Hassett!RP!(1982)!Estimating!the!grazing!impact!of!marine!microXzooplankton.!
Mar!Biol!67:283X288!
Laney!SR,!Sosik!HM!(2014)!Phytoplankton!assemblage!structure!in!and!around!a!massive!
underXice!bloom!in!the!Chukchi!Sea.!Deep!Sea!Res!II!
Li!WKW,!McLaughlin!FA,!Lovejoy!C,!Carmack!EC!(2009)!Smallest!Algae!Thrive!As!the!Arctic!
Ocean!Freshens.!Science!326:539X539!
! 91!
Liu!H,!Suzuki!K,!Saino!T!(2002)!Phytoplankton!growth!and!microzooplankton!grazing!in!the!
subarctic!Pacific!Ocean!and!the!Bering!Sea!during!summer!1999.!Deep!Sea!Res!I!
49:363X375!
Lowry!KE,!van!Dijken!GL,!Arrigo!KR!(2014)!Evidence!of!underXice!phytoplankton!blooms!in!
the!Chukchi!Sea!from!1998!to!2012.!Deep!Sea!Res!II!105:105X117!
Marrasé!C,!Lim!EL,!Caron!DA!(1992)!Seasonal!and!daily!changes!in!bacterivory!in!a!coastal!
plankton!community.!Mar!Ecol!Prog!Ser!82:281X289!
Niemi!A,!Meisterhans!G,!Michel!C!(2014)!Response!of!underXice!prokaryotes!to!
experimental!seaXice!DOM!enrichment.!Aquat!Microb!Ecol!73:17X28!
OrtegaXRetuerta!E,!Fichot!CG,!Arrigo!KR,!Van!Dijken!GL,!Joux!F!(2014)!Response!of!marine!
bacterioplankton!to!a!massive!underXice!phytoplankton!bloom!in!the!Chukchi!Sea!
(Western!Arctic!Ocean).!Deep!Sea!Res!II!105:74X84!
OrtegaXRetuerta!E,!Jeffrey!WH,!Babin!M,!Bélanger!S,!Benner!R,!Marie!D,!Matsuoka!A,!
Raimbault!P,!Joux!F!(2012)!Carbon!fluxes!in!the!Canadian!Arctic:!patterns!and!
drivers!of!bacterial!abundance,!production!and!respiration!on!the!Beaufort!Sea!
margin.!Biogeosciences!9:3679X3692!
Rose!JM,!Caron!DA!(2007)!Does!low!temperature!constrain!the!growth!rates!of!
heterotrophic!protists?!Evidence!and!implications!for!algal!blooms!in!cold!waters.!
Limnol!Oceanogr!52:886X895!
Sakshaug!E!(2004)!Primary!and!secondary!production!in!the!Arctic!Seas.!!The!organic!
carbon!cycle!in!the!Arctic!Ocean.!Springer!
Sanders!RW,!Gast!RJ!(2012)!Bacterivory!by!phototrophic!picoplankton!and!nanoplankton!
in!Arctic!waters.!FEMS!Microbiol!Ecol!82:242X253!
Schmoker!C,!HernándezXLeón!S,!Calbet!A!(2013)!Microzooplankton!grazing!in!the!oceans:!
impacts,!data!variability,!knowledge!gaps!and!future!directions.!J!Plankton!Res!
35:691X706!
Sherr!BF,!Sherr!EB,!Fallon!RD!(1987)!Use!of!monodispersed,!fluorescently!labeled!bacteria!
! to!estimate!in!situ!protozoan!bacterivory.!Appl!Environ!Microbiol!53:958X965!
Sherr!E,!Sherr!B!(1994)!Bacterivory!and!herbivory:!key!roles!of!phagotrophic!protists!in!
pelagic!food!webs.!Microb!Ecol!28:223X235!
Sherr!EB,!Sherr!BF!(2002)!Significance!of!predation!by!protists!in!aquatic!microbial!food!
webs.!Antonie!van!Leeuwenhoek!81:293X308!
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!352:187X197!
Sherr!EB,!Sherr!BF,!Hartz!AJ!(2009)!Microzooplankton!grazing!impact!in!the!Western!Arctic!
Ocean.!Deep!Sea!Res!II!56:1264X1273!
Sherr!EB,!Sherr!BF,!Ross!C!(2013)!Microzooplankton!grazing!impact!in!the!Bering!Sea!
during!spring!sea!ice!conditions.!Deep!Sea!Res!II!94:57X67!
Shiklomanov!IA!(2000)!Appraisal!and!Assessment!of!World!Water!Resources.!Water!Int!
25:11X32!
Smith!DC,!Azam!F!(1992)!A!simple,!economical!method!for!measuring!bacterial!protein!
synthesis!rates!in!seawater!using!3HXleucine.!Mar!Microb!Food!Webs!6:107X114!
Springer!AM,!McRoy!CP!(1993)!The!paradox!of!pelagic!food!webs!in!the!northern!Bering!
Sea—III.!Patterns!of!primary!production.!Cont!Shelf!Res!13:575X599!
! 92!
Steward!GF,!Smith!DC,!Azam!F!(1996)!Abundance!and!production!of!bacteria!and!viruses!in!
the!Bering!and!Chukchi!Seas.!Mar!Ecol!Prog!Ser!131:287X300!
Strom!SL,!Fredrickson!KA!(2008)!Intense!stratification!leads!to!phytoplankton!nutrient!
limitation!and!reduced!microzooplankton!grazing!in!the!southeastern!Bering!Sea.!
Deep!Sea!Res!II!55:1761X1774!
Suggett!DJ,!Moore!CM,!Hickman!AE,!Geider!RJ!(2009)!Interpretation!of!fast!repetition!rate!
(FRR)!fluorescence:!signatures!of!phytoplankton!community!structure!versus!
physiological!state.!Mar!Ecol!Prog!Ser!376:1X19!
Sukhanova!IN,!Flint!MV,!Pautova!LA,!Stockwell!DA,!Grebmeier!JM,!Sergeeva!VM!(2009)!
Phytoplankton!of!the!western!Arctic!in!the!spring!and!summer!of!2002:!Structure!
and!seasonal!changes.!Deep!Sea!Res!II!56:1223X1236!
Taylor!FJR,!Hoppenrath!M,!Saldarriaga!JF!(2008)!Dinoflagellate!diversity!and!distribution.!
Biodivers!Conserv!17:407X418!
Terrado!R,!Lovejoy!C,!Massana!R,!Vincent!WF!(2008)!Microbial!food!web!responses!to!light!
and!nutrients!beneath!the!coastal!Arctic!Ocean!sea!ice!during!the!winter–spring!
transition.!J!Mar!Syst!74:964X977!
Utermöhl!H!(1958)!Zur!gewassertypenfrage!tropischer!seen.!Verh!Int!Ver!Limnol!13:236X
251!
Vaqué!D,!Guadayol!O,!Peters!F,!Felipe!J,!AngelXRipoll!L,!Terrado!R,!Lovejoy!C,!PedrósXAlioó!C!
(2008)!Seasonal!changes!in!planktonic!bacterivory!rates!under!the!ice covered!
coastal!Arctic!Ocean.!Limnol!Oceanogr!53:2427X2438!
Wang!M,!Overland!JE!(2009)!A!sea!ice!free!summer!Arctic!within!30!years?!Geophys!Res!
Lett!36!
Weingartner!T,!Aagaard!K,!Woodgate!R,!Danielson!S,!Sasaki!Y,!Cavalieri!D!(2005)!
Circulation!on!the!north!central!Chukchi!Sea!shelf.!Deep!Sea!Res!II!52:3150X3174!
Wells!LE,!Deming!JW!(2003)!Abundance!of!Bacteria,!the!CytophagaXFlavobacterium!cluster!
and!Archaea!in!cold!oligotrophic!waters!and!nepheloid!layers!of!the!Northwest!
Passage,!Canadian!Archipelago.!Aquat!Microb!Ecol!31:19X31!
Wood!KR,!Bond!NA,!Danielson!SL,!Overland!JE,!Salo!SA,!Stabeno!PJ,!Whitefield!J!(2015)!A!
decade!of!environmental!change!in!the!Pacific!Arctic!region.!Prog!Oceanogr!136:12X
31!
Yang!EJ,!Ha!HK,!Kang!SXH!(2014)!Microzooplankton!community!structure!and!grazing!
impact!on!major!phytoplankton!in!the!Chukchi!sea!and!the!western!Canada!basin,!
Arctic!ocean.!Deep!Sea!Res!II!
!

! 93!

CHAPTER(THREE.(NUTRITIONAL(REQUIREMENTS(AND(PREY(CELL(CYCLES(DICTATE(
TEMPORAL(FEEDING(STRATEGIES(OF(HETEROTROPHIC(AND(MIXOTROPHIC(
NANOPLANKTON(

By Paige E. Connell

Coauthors: François Ribalet, E. Virginia Armbrust, and David A. Caron
! 94!
Abstract:
Daily oscillations in photosynthetically active radiation control the timing of metabolic
processes in marine picocyanobacteria, but it is unclear how or if the light-dark cycle propagates
up marine food webs. We investigated the relationship between marine picocyanobacteria and
their consumers throughout the diel cycle in order to determine whether significant periodicity in
grazing pressure exists for heterotrophic or mixotrophic protists. Measurements of
Prochlorococcus and Synechococcus abundances and division rates, nanoplankton abundances,
and nanoplankton ingestion of fluorescently-labeled bacteria were conducted at regular time-
intervals (≤4 h) for four consecutive days in the North Pacific Subtropical Gyre. Heterotrophic
nanoplankton grazing was significantly greater at night and appeared to closely follow peak
cyanobacterial division at dusk. Conversely, mixotrophic nanoplankton grazing showed no diel
periodicity. We speculate that mixotrophs may consume prey to alleviate nutrient limitation,
rather than to acquire carbon, and therefore were less synchronized to cycles of prey cell
division. Estimates of protistan carbon demand indicated that nanoplankton were capable of
removing a substantial proportion of cyanobacterial production, contributing to temporally stable
abundances of Prochlorococcus and Synechococcus. These results highlight the importance of
accurately characterizing predator-prey relationships in the plankton in order to understand flows
of energy and elements within microbial food webs.

! 95!
3.1 Introduction
Marine picocyanobacteria are the most abundant phototrophic organisms on the planet,
with a global population abundance of ~10
27
cells (Flombaum et al. 2013). The marine
picocyanobacteria are comprised of two genera, Prochlorococcus and Synechococcus, which
are estimated to have diverged from a common ancestor 150 million years ago (Dufresne et al.
2005). Prochlorococcus and Synechococcus have since diversified into at least 12 major clades
(Biller et al. 2015) and 10 major clades (Ahlgren and Rocap 2012), respectively, which have
been defined by morphology, physiology, genomic similarity and environmental niche.
Synechococcus has a wider global distribution, owing to its’ greater capacity to tolerate cold
temperatures (Pittera et al. 2014) and metal contamination (Mann et al. 2002). However,
Prochlorococcus is the dominant organism in warm, oligotrophic regions, with cells inhabiting
all portions of the euphotic zone in the upper ocean (Partensky et al. 1999).
These abundant and diverse phytoplankton genera are critical components of the earth’s
biogeochemical cycles and marine food webs. Approximately 50% of global oxygen production
is a result of photosynthesis in the ocean (Field et al. 1998), with a substantial fraction of that
production attributable to picocyanobacteria. Synechococcus and Prochlorococcus biomass also
supports considerable secondary production by planktonic consumers capable of capturing and
ingesting pico-sized (0.2–2.0 µm diameter) prey in warm oceanic regimes (Landry and Calbet
2004). Therefore, characterizing the trophic relationship between the marine cyanobacteria and
the organisms that consume them is critical for understanding ecosystem dynamics and
biogeochemical fluxes in the marine environment.
Marine picocyanobacteria rely on photosynthetically active radiation (PAR) for cell
maintenance and growth. As PAR intensity fluctuates over a 24-h diel cycle, marine
! 96!
picocyanobacteria have evolved to tightly regulate cellular processes to maximize their
efficiency of carbon fixation. Synechococcus has a circadian clock involving three proteins
(KaiA, KaiB, and KaiC) that is synchronized with the light-dark cycle using the redox status of
the quinone pool and the ATP/ADP ratio as signals for the onset and duration of darkness
(Cohen and Golden 2015); this endogenous circadian clock enables Synechococcus to maintain
diel oscillations in cellular processes under continuous light conditions. Conversely, the timing
system in Prochlorococcus lacks the kaiA gene, preventing the continuation of diel periodicity
under continuous light conditions and causing this system to function as a diurnal oscillator
rather than an autonomous circadian clock (Axmann et al. 2009, Holtzendorff et al. 2008). Both
of these regulatory systems result in transcriptional oscillations in the picocyanobacteria. For
example, genes related to photosystem I/II and the Calvin cycle are upregulated at dawn and
genes related to the pentose phosphate pathway are upregulated at dusk (Ito et al. 2009, Zinser et
al. 2009).
There is presently little information as to what extent these diel rhythms in cyanobacterial
physiology influence the behaviors of higher trophic levels, or the other organisms dependent on
phytoplankton for organic substrates or cellular machinery (ie. heterotrophic bacteria and
viruses). Recent studies have demonstrated significant diel shifts in the transcriptional expression
profiles of heterotrophic bacteria in the North Pacific Subtropical Gyre (NPSG) and coastal
California upwelling region (Aylward et al. 2015, Ottesen et al. 2014). Aylward et al. (2015)
specifically found Prochlorococcus to be the primary determinant of the community
transcriptional response in the NPSG, with the maxima of Prochlorococcus carbon fixation and
photosynthesis transcripts preceding a cascade of heterotrophic transcript expression maxima.
! 97!
The relationship between phytoplankton diel activity levels and the activities of their
consumers throughout the light-dark cycle is enigmatic. Field observations that
picocyanobacterial division occurs at dusk, coupled with observations that indicate relatively
stable day-to-day population abundances, have led researchers to posit that cyanobacterial
mortality (viral- and/or grazer-mediated) is highest at night and negligible during the day
(Ribalet et al. 2015). Few studies have directly investigated changes in protistan grazing pressure
throughout the diel cycle and the few published reports on this behavior are somewhat
contradictory. For example, some studies looking at the ingestion of picoplanktonic prey by
nanoplankton have suggested grazing pressure is highest at night, when post-division prey cells
are smallest and most abundant (Christaki et al. 2002, Tsai et al. 2009). Other studies have
reported that grazing pressure was highest during the day, citing small cell size early in the day
(Dolan and Šimek 1999) or compensatory feeding on low quality prey (i.e. higher C:N ratios)
(Ng and Liu 2016) as possible explanatory factors. These results offer few consistent clues
regarding the diel activities of picoplankton consumers.
Protistan grazers are a predominant source of microbial mortality in marine environments
(Calbet and Landry 2004) and an important source of regenerated nutrients and dissolved organic
matter for the phytoplankton and bacterial assemblages (Caron and Goldman 1990). Protists
display a diverse array of nutritional strategies, including species that meet all growth
requirements via phagotrophy. Other species combine heterotrophic and phototrophic nutrition
(mixotrophy) to meet various nutritional requirements including carbon, macronutrient (N,P),
micronutrient (e.g. Fe), or vitamin acquisition (Liu et al. 2016, Stoecker et al. 2017). Thus, both
heterotrophic and mixotrophic protists may employ phagotrophy to meet different nutritional
! 98!
needs, resulting in different short-term responses of these grazers to diel fluctuations in prey size,
abundance, or cellular composition.
In this study, we investigated the relationship between marine picocyanobacteria and
their protistan consumers throughout the diel cycle. Measurements of population abundances,
cyanobacterial division rates, and nanoplankton grazing pressure (measured via the uptake of
fluorescently-labeled prey in short-term incubations) were conducted at regular time intervals
(4h or less) for four days in the NPSG. We specifically sought to determine whether
nanoplankton grazing pressure exhibited diel periodicity, and if so, whether this periodicity was
the same for heterotrophic and mixotrophic nanoplankton. Additionally, we examined how the
relative timing of consumption controls the day-to-day population abundances of the
picocyanobacterial assemblage. Heterotrophic and mixotrophic nanoplankton responded
differently to the diel periodicity of the picocyanobacterial cell cycle. Heterotrophic
nanoplankton exhibited significant diel periodicity in grazing pressure that closely followed the
cycle of prey division, while mixotrophic nanoplankton demonstrated no diel periodicity in their
grazing pressure. We speculate that the disparate nutritional requirements of heterotrophic and
mixotrophic predators may explain the different responses of the two consumer assemblages.
This study emphasizes the importance of understanding the community composition of the
consumer assemblage, and the spectrum of nutritional modes represented therein, to quantifying
the movement of carbon and other key elements through microbial food webs.

! 99!
3.2 Methods
3.2.1%Sampling%and%environmental%metadata%
!
Water collection occurred from July 26 to July 30, 2017 in the North Pacific Subtropical
Gyre (NPSG) (Fig. 3.1). A Lagrangian sampling scheme was conducted to ensure that the same
water mass was measured throughout the study period. World Ocean Circulation Experiment
Surface Velocity Profile drifters (WOCE SVPs) were deployed prior to sampling, with drogues
centered at a depth of 15 m (sampling depth), within an anticyclonic eddy northeast (~24.4 N,
156 W) of the long-term time-series location Station ALOHA (22.45 N, 158 W). The eddy had a
diameter of ~100 km and was targeted for Lagrangian sampling due to the high degree of
coherence of the water mass. Shipboard measurements were made alongside the drifters every 4
h throughout the study period to characterize the diel variability of microbial abundances and
activities at the study site. Water for incubations (section 3.2.2) was collected at a depth of 15 m
using a 24 x 12 L Niskin rosette equipped with a CTD (SBE 911Plus, Seabird Electronics,
Bellevue, WA) and fluorescence and oxygen sensors, calibrated to discrete chlorophyll a and
phaeopigment measurements and dissolved O
2
measurements, respectively.
%
3.2.2%Nanoplankton%abundances%and%relative%grazing%pressure%
!
The cellular abundances and relative grazing pressure of the nanoplankton (microbial
eukaryotes, 2–20 µm in size) were determined every 4 h during the study period. These
measurements were conducted in a series of incubations using the uptake of fluorescently-
labeled bacteria (FLB) as a tracer for the ingestion of picoplankton prey. FLB were prepared
from a monoculture of the rod-shaped bacterium, Dokdonia donghaensis, as previously

! 100!

Figure 3.1. Study location in the North Pacific Subtropical Gyre, in an anticyclonic eddy
northwest of Station ALOHA. Hawaiian Islands are shown in black. Insert: blue circles
indicate the sampling locations for nanoplankton abundance and grazing impact. The
black line shows the Seaflow sampling track, with black points indicating location of
hourly averages of picocyanobacterial abundance and division rates.
!
! 101!
described (Caron 2001, Sherr et al. 1987). Succinctly, D. donghaensis was grown in Zobell
medium for two days, harvested by centrifugation and resuspended in 0.2 µm filtered seawater,
and incubated in the filtrate for two days to starve the cells. Starving the cells caused cell
shrinkage, so that the labeled prey would better mimic the size of bacteria and cyanobacteria
observed in natural assemblages (average length ≈ 1 µm). Bacteria were then stained with 5-(4,6-
dichlorotriazin-2-yl) aminofluorescein (DTAF; Invitrogen, Carlsbad, CA) and heat-killed,
followed by three rinses in ultrapure water (Milli-Q Water Purification System, EMD Millipore
Corp., Billerica, MA). FLBs were aliquoted and stored at −80°C until the time of the
experiments. FLB used in the study were prepared in a single batch to ensure homogeneity
between experiments.
Natural seawater used in the incubation experiments was transferred directly from a
Niskin bottle into triplicate, 2.3 L polycarbonate bottles using acid-washed (5% HCl) silicone
tubing to prevent bubbling that harms delicate protistan grazers. FLB stock was vigorously
vortexed and added to each bottle at ~15% of the natural bacterial abundance. Bottles were
gently mixed to evenly distribute FLB and subsamples (T
0
= 90 mL) were removed from each
bottle and preserved with formaldehyde (final concentration 1%, final volume 100 mL). Bottles
were then incubated for 1 h in on-deck incubators maintained at in situ temperature and shaded
with blue plexiglass to approximate the light level experienced by the microbial assemblage at 15
m. After the incubation period, subsamples were again collected from each bottle (T
f
). The
preserved subsamples (T
0
and T
f
) were stored at 4°C until the preparation of slides for
epifluorescence microscopy (within 12 h of sampling). Slides were prepared by filtering the 100
mL subsamples down to ~1 mL onto blackened, 2.0 µm, 25mm polycarbonate filters and
staining the samples with 50 µL of a 4’-6’-diamidino-2-pheylindole (DAPI, Sigma-Aldrich, St.
! 102!
Louis, MO) working solution (1 mg mL
−1
) for 5–10 min in the dark (Sherr et al. 1993). Samples
were then filtered and rinsed, filters were placed onto glass slides with a drop of immersion oil
and a coverslip, and the coverslips were sealed with clear nail polish. Slides were stored at
−20°C until analysis using epifluorescence microscopy.
Nanoplankton abundances were enumerated for each time-point from the triplicate, T
0

slides by epifluorescence microscopy. Photo/mixotrophic nanoplankton were distinguished from
heterotrophic nanoplankton by the presence or absence of chlorophyll autofluorescence,
respectively, when viewed under blue-light excitation.
The relative grazing pressure of the nanoplankton assemblage was measured by
determining the percentage of nanoplankton with ingested FLB on the triplicate slides prepared
from the T
0
and T
f
time-points of each incubation. To quantify the grazing pressure of both the
photo/mixotrophic and heterotrophic nanoplankton, 100 cells on each slide were categorized as
having (1) no chlorophyll a, no ingested FLB, (2) no chlorophyll a, ingested FLB, (3)
chlorophyll a, no ingested FLB, or (4) chlorophyll a, ingested FLB. The percentage of total,
photo/mixotrophic, and heterotrophic nanoplankton with ingested FLB was then determined by
dividing the number of cells with FLB visualized inside within each category by the total number
of cells enumerated for that category. T
f
ingestion percentages for each bottle were corrected by
subtracting T
0
ingestion percentages from the same bottle, to account for FLB lying near
nanoplankton on the slides but not ingested by them, and the corrected values were averaged for
each time-point in the diel cycle.
%
%
%
%
! 103!
3.2.3%Cyanobacterial%abundances%and%rate%processes%derived%from%continuous%flow%cytometry%
!
Cyanobacterial abundances and hourly rates of cell division were determined as detailed
in Ribalet et al. (2015). Continuous measurements of Prochlorococcus and Synechococcus
abundances and cell size were made using the underway flow-cytometer Seaflow (Swalwell et al.
2011). Data files were created every 3 minutes and Prochlorococcus and Synechococcus
abundances were enumerated using a sequential bivariate manual gating scheme provided by the
R package Popcycle version 0.2. Hourly-averaged cell abundance was estimated from the mean
of cell abundances over a 1-h period (N=20). Light scattering was converted to cell size (Ribalet
et al. 2015) and hourly Prochlorococcus and Synechococcus size distributions were obtained by
aggregating cell size data over a 1-hr period.
Hourly division rates of Prochlorococcus and Synechococcus were estimated using a
size-structured matrix population model developed by Sosik et al. (2003). This model represents
changes in cell size over a 24-h period and can be fit to time-series of hourly cell size
distributions. A 1-h rolling window was used to determine the start of each 24-h period in the
time-series and hourly division rates were obtained by calculating the mean values for each hour
of the time-series (N=24). The model is based on two main assumptions: 1) increase in cell size
is caused by cell growth as determined only by light exposure, with other abiotic factors such as
nutrient availability and temperature operating at longer time scales, 2) decrease in cell size is
caused only by cell division, with the probability of a cell dividing depending on its size. The
model does not take into account size-selective grazing, which could alter size distribution
independently from cell growth and cell division. Extensive validation of the model for
Prochlorococcus and Synechoccoccus both in cultures and natural systems suggests that grazing
activity does not significantly alter the model results (Hunter-Cevera et al. 2016, Hunter-Cevera
! 104!
et al. 2014, Hynes et al. 2015, Ribalet et al. 2015, Sosik et al. 2003), but we acknowledge the
possibility that grazing effects may have occurred during the study period.
%
3.2.4%Statistical%analyses%
!
All statistics were computed using R software (R Core Team 2015). Significant changes
in population abundances throughout the 4-d period were determined using Welch Two-Sample
T-tests between the initial (July 26, 6:00) and final (July 30, 6:00) time-points. Significant diel
periodicity was determined using the RAIN package (Thaben and Westermark 2014). Peak times
were calculated using harmonic regression analysis for datasets with significant diel periodicity
as determined using the RAIN package. Statistical tests in which the p-value was <0.05 were
considered to be significant.

3.3 Results
Environmental conditions at 15 m remained relatively constant throughout the 4-d study
period (Table A5). The mean (± SE) water temperature at 15 m was 26.81 ± 0.02 °C and mean
salinity was 35.38 ± 0.001. Dissolved oxygen concentrations averaged 205.7 ± 0.13 µmol L
−1
,
while chlorophyll a concentrations averaged 0.19 ± 0.01 µg L
−1
.
Prochlorococcus was the dominant phytoplankter, with cellular abundances of
Prochlorococcus at ~200x that of Synechococcus (Figs. 3.2a and 3.2b). Mean (± SE)
Prochlorococcus abundance was 2.64 x 10
5
± 0.02 cells mL
−1
while mean Synechococcus
abundance was 1.24 x 10
3
± 0.01 cells mL
−1
. Overall, population abundances of both
cyanobacterial genera were relatively stable throughout the 4-d period. Final abundances of
Prochlorococcus and Synechococcus were significantly different from initial abundances (Welch
! 105!

Figure 3.2. Abundances of Prochlorococcus (PRO; a), Synechococcus (SYN; b),
heterotrophic nanoplankton (HNANO; c), and phototrophic/mixotrophic nanoplankton
(PMNANO; d) throughout the diel cycle. Points represent the mean measurements at
each time-point, while shading indicates the standard errors of the means. Day-night
cycles are shown, with gray boxes indicating nighttime. Note the change in y-axis scale
between plots.

a b
c d
! 106!
Two-Sample T-test; p <0.001), however changes in population abundances were minor.
Prochlorococcus abundances increased from 2.40 x 10
5
cells mL
−1
to 2.80 x 10
5
cells mL
−1
from
the initial to final time-point (17% increase), while Synechococcus abundances increased from
1.16 x 10
3
cells mL
−1
to 1.32 x 10
3
cells mL
−1
(14% increase). In addition, slight but significant
diel periodicities in the abundances of both cyanobacterial populations were detected (RAIN
analysis, p < 0.001). Population abundances peaked at different times of the day for each
cyanobacterial group, with Prochlorococcus abundances peaking at night (01:50) and
Synechococcus abundances peaking in the morning (09:20) (Fig. 3.2a and 3.2b, respectively).
The mean (± SE) total nanoplankton abundance was 313 ± 11 cells mL
−1
, with the
community comprised of approximately twice as many heterotrophic nanoplankton (! (mean) =
217 ± 6 cells mL
−1
) as photo/mixotrophic nanoplankton (! = 97 ± 8 cells mL
−1
) (Figs. 3.2c and
3.2d, respectively). The abundances of both consumer assemblages remained stable throughout
the 4-d period, showing a slight but insignificant decrease between the initial and final time-point
in the series (Welch Two-Sample T-test; p > 0.05). Neither the heterotrophic nor
photo/mixotrophic nanoplankton abundances exhibited significant diel periodicity (RAIN
analysis, p > 0.05).
Division rates of Prochlorococcus averaged 0.02 h
−1
(SE ± 6.5 x 10
−4
) and were ~3x
higher than those measured for Synechococcus (! = 0.007 h
−1
, SE ± 2.3 x 10
−4
) (Fig. 3.3). Both
Prochlorococcus and Synechococcus division rates showed significant diel periodicity (RAIN
analysis, p < 0.001). The peak division times for both cyanobacterial groups were similar, in
contrast to the peak abundance times of these groups, with Prochlorococcus division rate
peaking at 19:30 and Synechococcus division rate peaking at 18:20.
! 107!

Figure 3.3. Division rates of Prochlorococcus (PRO; a) and Synechococcus (SYN; b)
throughout the diel cycle. Hourly division rates were estimated from Seaflow data using a
size-structured matrix population model. Points represent the mean measurements at
each time-point, while shading indicates the standard errors of the means. Day-night
cycles are shown, with gray boxes indicating nighttime. Note the differences in y-axis
scales between plots.

a
b
! 108!

Figure 3.4. Percentage of heterotrophic nanoplankton (HNANO; a), photo/mixotrophic
nanoplankton (P/MNANO; b), and total nanoplankton (TNANO; c) with ingested
fluorescently-labeled bacteria (FLB) measured at 4-h time intervals throughout the study
period. Points represent the average percentages of nanoplankton cells with ingested
FLB from triplicate bottles, while shading depicts the standard errors of the means. Day-
night cycles are shown in this figure, with gray boxes indicating nighttime.

b
c
a
! 109!
The mean (± SE) percentage of nanoplankton grazers with ingested FLB was slightly
higher for heterotrophic nanoplankton (! = 12 ± 1.1%; Fig. 3.4a) than for mixotrophic
nanoplankton (! = 7.9 ± 0.8%; Fig. 3.4b). The grazing pressure of the heterotrophic
nanoplankton showed significant diel periodicity (RAIN analysis, p < 0.001), with peak
ingestion of FLBs occurring at 23:20. Conversely, no diel periodicity was detected for
mixotrophic nanoplankton grazing pressure (RAIN analysis, p > 0.05). The mean (± SE)
percentage of total nanoplankton with ingested FLB was 9.9 ± 0.9% (Fig. 3.4c). Patterns of
ingestion for the total nanoplankton were driven by the heterotrophic nanoplankton (which
comprised 69% of the total nanoplankton abundance; Figs. 3.2c and 3.2d); thus, total
nanoplankton ingestion was also found to show significant diel periodicity (RAIN analysis, p <
0.001) with a peak ingestion time at 23:30.

3.4 Discussion
3.4.1%Protistan%grazing%contributes%to%temporally%stable%picocyanobacterial%abundances%
!
Picocyanobacterial cell division rates at our sampling location in the NPSG peaked at
dusk during this study (Fig. 3.3). This finding is in agreement with other studies of
cyanobacterial division rates, which showed that both Prochlorococcus (Ribalet et al. 2015,
Vaulot and Marie 1999, Vaulot et al. 1995) and Synechococcus (Hunter-Cevera et al. 2014, Tsai
et al. 2009, Vaulot and Marie 1999) cell division occurred near dusk. Marine picophytoplankton
have been shown to upregulate genes related to photosynthesis and carbon fixation between
dawn and mid-day (Aylward et al. 2015), followed by DNA replication in the afternoon (Vaulot
et al. 1995), which prepares them for cell division in the early nighttime.
! 110!
Despite the strong diel pattern of picocyanobacterial division rates observed in this study,
the observed increases in Prochlorococcus and Synechococcus abundances were less than
predicted (Fig. 3.2a and 3.3a; Supplemental Text in Appendix B). These results imply that the
timing of picocyanobacterial cell removal was tightly coupled with the timing of
picocyanobacterial cell division, as cell division should have resulted in much larger increases in
picocyanobacterial abundances than we measured during the diel cycle. In support of this
observation, the measured hourly rates of FLB ingestion were significantly higher during the
night relative to ingestion rates measured during the day (Fig. 3.4), with the diel periodicity of
protistan grazing particularly pronounced for the heterotrophic nanoplankton assemblage (Fig.
3.4a).
We estimated the average number of picocyanobacterial cells removed daily during our
study, assuming that all removal was due to nanoplankton grazing, in order to assess the
potential for these consumers to constrain picocyanobacteria populations to their observed
changes in daily abundances. Daily removal of Prochlorococcus and Synechococcus cells was
calculated using changes in picocyanobacterial abundances over a 24-h period (Fig. 3.2), and
hourly division rates were calculated using the Seaflow data (Fig. 3.3) (for calculations, see
Supplemental Text in Appendix B). Our calculations indicate that, averaged over our 4-day diel
study, 6.7-13.3 x 10
4
Prochlorococcus cells mL
−1
and 92-184 Synechococcus cells mL
−1
should
have been ingested daily by nanoplanktonic protists. The number of cells grazed daily was
dependent on whether picocyanobacterial prey were removed immediately prior to division or
after cell division (see Supplemental Text in Appendix B). Typical ingestion rates for
nanoplanktonic consumers range from 10s to 100s of picoplanktonic prey hr
−1
(Fenchel 1986,
Nygaard and Tobieson 1993, Sherr and Sherr 2002). With a mean nanoplankton abundance of
! 111!
313 cells mL
−1
(Fig. 3.2), maintaining constant day-to-day picocyanobacterial abundances would
require the consumption of ≈1 Synechococcus nanoplankter
−1
d
−1
and 9-18 Prochlorococcus cells
nanoplankter
−1
hr
−1
. Therefore, complete removal of daily cyanobacterial production by
nanoplanktonic protists during our study is feasible.
While our calculations indicate that complete ingestion of picocyanobacterial daily
production by nanoplankton consumers is feasible, it is unclear whether these grazers could be
solely responsible for maintaining relatively constant cyanobacterial abundances if
cyanobacterial cell division was restricted to only a part of the day, at least in the case of
Prochlorococcus. For example, if all Prochlorococcus cell division was restricted to a 4-h
period in the evening, resulting in the appearance of 6.7-13.3 x 10
4
Prochlorococcus cells mL
−1
during that time period alone, then the nanoplankton ingestion rates required to prevent a
substantial increase in Prochlorococcus abundance during that period would be ≈106
Prochlorococcus nanoplankter
−1
hr
−1
(13.3 x 10
4
new Prochlorococcus cells mL
−1
÷ [313
nanoplankton mL
−1
x 4-h grazing period]). These rates are possible, but not probable, in this
highly oligotrophic environment. If, however, nanoplankton were consuming Prochlorococcus
prey prior to cell division, than the nanoplankton ingestion rates required to prevent a substantial
increase in Prochlorococcus abundance during that same time period would be ≈53
Prochlorococcus nanoplankter
-1
hr
-1
(6.7 x 10
4
new Prochlorococcus cells mL
−1
÷ [313
nanoplankton mL
−1
x 4-h grazing period]). Therefore, if nanoplankton grazing is the dominant
source of removal of Prochlorococcus daily production, we postulate that consumption of
dividing Prochlorococcus cells would be the most likely scenario for removing new
Prochlorococcus production.
! 112!
The estimates of potential grazer-mediated removal of picocyanobacteria by
nanoplankton consumers provided above indicate that grazer activities may account for the
relatively stable diel abundances of Synechococcus in the NPSG, but may not be completely
sufficient to explain the relatively stable diel abundances of Prochlorococcus if cyanobacterial
division is confined to a short period of the day. Viral lysis is another potentially important
source of picocyanobacterial mortality. It has been suggested that viral lysis may account for 10-
40% of total bacterial mortality in marine environments (Fuhrman 2000), with a similar impact
on cyanobacterial populations (Suttle 2000). Concurrent measurements of virus-induced
mortality were not conducted in this study, but Aylward et al. (submitted) demonstrated that viral
activity peaked in the afternoon during our study, prior to the peak in cyanobacterial division at
dusk (Fig. 3.3), indicating that viral lysis of cyanobacterial cells approaching cell division may
also play a role in reducing the magnitude of the observed picocyanobacterial production during
the evening. Cyanophage infection is affected by the cell cycle of the host (Ni and Zeng 2016)
and bacteriophage burst-size has been shown to be largest just prior to host cell division (Stárka
1962, Storms et al. 2014). Thus, it is possible that viral infection and lysis is also timed to the
diel cycle of cyanobacterial cell division in the NPSG. It is even possible that infected
cyanobacterial cells are more susceptible to consumption by protistan grazers. Viral infection has
been reported to increase host susceptibility to grazing for both eukaryotic (Evans and Wilson
2008) and prokaryotic prey cells (Zwirglmaier et al. 2009). For example, Zwirglmaier et al.
(2009) reported that two heterotrophic nanoflagellate species only consumed the phage-resistant
strain of Synechococcus WH7803 (designated as WH7803PHR) in culture studies. The coupling
of these two processes (viral infection and grazer ingestion) to remove Prochlorococcus cells as
! 113!
they prepare for cell division might explain the stable day-to-day abundances of
Prochlorococcus during our study period.
%
3.4.2%Potential%contribution%of%picoplankton%carbon%to%nanoplankton%growth%
!
We employed estimates of the removal of cyanobacterial biomass calculated above,
assuming that all of the carbon removal was due to protistan grazing, to estimate the carbon
potentially available for nanoplankton growth in the NPSG. The estimated total removal of
Prochlorococcus cells and Synechococcus cells was converted to carbon biomass using carbon
conversion factors from the literature (see Supplemental Text). These calculations indicate that
6.67 ng of Prochlorococcus C and 0.023 ng of Synechococcus C were removed daily (Table A6
and A7). One conclusion from these observations is that Synechococcus comprised a small
portion of the total nanoplankton diet (0.001%) and likely had little impact on the nutrition of the
phagotrophic nanoplankton assemblage.
Carbon demand (ng C mL
−1
d
−1
) for growth of the phagotrophic nanoplankton
assemblage was calculated according to the following equation,
((µ * B
v
* C) / GGE) * NANO
where µ = nanoplankton growth rate (d
−1
), B
v
= nanoplankton cell biovolume (µm
3
cell
−1
), C =
nanoplankton carbon content (ng C µm
−3
), GGE = nanoplankton gross growth efficiency, and
NANO = nanoplankton abundance (cells mL
−1
). Using the calculated removal of cyanobacterial
carbon (6.70 ng C mL
−1
d
−1
), a carbon content of 183 fg C µm
−3
(Caron et al. 2017), a gross
growth efficiency of 30% (Straile 1997), and a range of nanoplankton diameters (3-6 µm), we
estimate that the cyanobacterial new production could have supported nanoplankton growth rates
of 0.31 to 2.48 d
−1
(for the range of nanoplankton cell sizes considered). Maximal nanoplankton
! 114!
growth rates reported for nanoplankton cultured under rich conditions in the lab can be >2.0 d
−1
(Rose and Caron 2007), however, typical nanoplankton growth rates in the field are 0.2 to ~1.4
d
−1
(Karayanni et al. 2008, Neuer and Cowles 1994, Verity et al. 1993).

Thus, it appears that the
daily removal of cyanobacterial cells was sufficient to support realistic growth rates of the
nanoplankton community during our study and implies that protistan grazers might reasonably
remove a significant proportion of daily picocyanobacterial production.
Heterotrophic bacterioplankton (bacteria + archaea) are another source of prey carbon for
the nanoplankton assemblage. Nanoplankton grazing has been shown to be size-selective
(Jürgens and Massana 2008), with prey within the picoplankton size class (0.2–2.0 µm) thought
to be the optimal size for phagotrophic protists. For this reason we used fluorescently-labeled
bacteria (FLB, average length = 1.0 µm) as surrogate prey for Prochlorococcus (diameter = 0.7
µm) and Synechococcus (diameter = 1.0 µm) in our incubation experiments. The mean
abundance of heterotrophic bacterioplankton at Station ALOHA has been reported to be ~5.5 x
10
5
cells mL
−1
(Eiler et al. 2009), which we converted to 5.5 ng C mL
−1
of heterotrophic
bacterioplankton biomass using a carbon conversion factor of 10 fg C cell
−1
that is representative
of open ocean environments (Cermak et al. 2017, Fukuda et al. 1998, Kawasaki et al. 2011).
Assuming a mean growth rate of 0.08 d
−1
for open ocean bacterioplankton (Kirchman 2016), the
maximum production of heterotrophic bacterial carbon would be 0.44 ng C mL
−1
d
−1
. Thus, the
estimated maximum production of heterotrophic bacterial carbon was only 7% of the observed
production of cyanobacterial carbon (6.70 ng C mL
−1
d
−1
, see above), and by itself would have
supported very low nanoplankton growth rates (0.02 to 0.16 d
−1
) during our study. Even at
considerably higher growth rates of the heterotrophic bacteria, that assemblage would still
contribute a modest fraction of the total carbon available daily to nanoplankton consumers.
! 115!
Therefore, we speculate that Prochlorococcus was the primary carbon source for protistan
consumers in the NPSG during our study.
%
3.4.3%Disparate%responses%of%the%heterotrophic%and%mixotrophic%nanoplankton%may%indicate%different%
nutritional%needs%
!
The relative grazing pressure of the total nanoplankton community in the present study
was highest around midnight (Fig. 3.4c), however, heterotrophic and mixotrophic nanoplankton
responded differently to the cycles of cyanobacterial cell division. Heterotrophic nanoplankton
showed significant diel periodicity in grazing pressure (peak at 23:20) that repeatedly followed
the peak in cyanobacterial division rates at dusk (peaks at 18:00-20:00) (Fig. 3.5). Conversely,
no significant periodicity in grazing pressure was detected for mixotrophic nanoplankton (Figs.
3.4a and 3.4b).
We speculate that the observed difference between the timing of grazing by heterotrophic
and mixotrophic nanoplankton may reflect differences in the nutritional needs provided by prey
ingestion. The nutritional requirements met by phagotrophy in mixotrophic phytoflagellates vary
greatly (Liu et al. 2016, Stoecker et al. 2017). Mixotrophic prey ingestion may be a strategy to
obtain organic carbon or energy (Terrado et al. 2017), major nutrients (Arenovski et al. 1995,
Hartmann et al. 2011, Nygaard and Tobieson 1993, Unrein et al. 2007) or micronutrients
(Maranger et al. 1998, Stukel et al. 2011). Conversely, heterotrophic protistan grazers
phagocytize prey to fulfill all their nutritional needs. Thus, mixotrophic nanoplankton may be
less dependent than heterotrophic nanoplankton on efficient grazing to provide sufficient
nutrition to augment their phototrophic abilities.
! 116!
!
Figure 3.5. Timing of the division rate of Prochlorococcus (PRO; a) and Synechococcus
(SYN; b) relative to the heterotrophic nanoplankton grazing pressure (percentage of
nanoplankton with ingested FLB) throughout the study period. Note that peak
heterotrophic nanoplankton grazing pressure is in tight synchrony with the timing of
cyanobacterial cell division.
a
b
! 117!
The tight synchrony observed during our study between the timing of maximal grazing
pressure by the heterotrophic nanoplankton and the timing of picocyanobacterial cell division
(Fig. 3.5) may be a strategy for maximizing carbon and nutrient acquisition by those grazers,
although the mechanism by which it takes place is unclear. Mean population abundances of
Prochlorococcus and heterotrophic nanoplankton during our study were 2.64 x 10
5
cells mL
−1

and 217 cells mL
−1
, respectively, and neither abundance varied greatly throughout the day
(maximum deviation from the mean of 16% for Prochlorococcus and 34% for the heterotrophic
nanoplankton; Fig. 3.2). It is therefore unlikely that there was a meaningful difference in
encounter rate between predators and prey throughout the diel cycle. However, size-selective
removal by heterotrophic nanoplankton of Prochlorococcus cells entering cell division might
explain the synchrony between prey cell division and the peak in FLB ingestion rates during the
night. Size-selectivity is believed to be an important determinant of ingestion rates in small
phagotrophic protists (reviewed in Jürgens and Massana (2008)). Other processes that might
make dividing prey-cells more susceptible to grazing than non-dividing cells include chemical
composition (Chrzanowski and Foster 2014, John and Davidson 2001, Ng et al. 2017, Shannon
et al. 2007, Siuda and Dam 2010), cell surface properties (Monger et al. 1999, Wootton et al.
2007), the release of prey infochemicals (Breckels et al. 2011, Fenchel and Blackburn 1999,
Verity 1991), or viral infection, as noted above.
The lack of diel periodicity observed for mixotrophic nanoplankton grazing pressure
during our study, in combination with the persistently nutrient-deplete conditions found in the
NPSG (Karl and Church 2014), suggest that the mixotrophic nanoplankton in this environment
may be ingesting prey primarily as a means of nutrient acquisition. If mixotrophic nanoplankton
were ingesting prey as a primary means of organic carbon acquisition, one might expect their
! 118!
grazing activity to be higher at night when they cannot photosynthesize. The absence of diel
periodicity in mixotrophic nanoplankton grazing pressure may indicate that they are able to use
phototrophy (to varying degrees) to meet their carbon/energy demands, and that they
phagocytize prey primarily to gain nutrients. This hypothesis is substantiated by a recent
ecosystem model showing that mixotrophy is an important nutrient acquisition strategy for
phototrophic nanoplankton to survive in the subtropical gyres (Ward and Follows 2016). It is,
however, also possible that the mixotrophic nanoplankton assemblage contained a mixture of
species that were consuming prey to meet different nutritional requirements and were feeding at
different times of day to meet those needs, resulting in no apparent diel periodicity for the
assemblage. These differences would not have been detected using epifluorescence microscopy
because most nanoplankton are morphologically nondescript. Diel periodicity in phytoflagellate
grazing has been documented in the East China Sea, where mixotrophic phytoflagellates were
the primary consumers of the Synechococcus, however the strength of the periodicity changed
depending on the sampling month (Tsai et al. 2009). Thus, the composition of the mixotroph
assemblage—and the respective nutritional requirements of the species comprising it—may
dictate whether a synchronized diel cycle is detectable for the mixotrophic assemblage.
Mixotrophic flagellates were an important source of grazer-mediated picoplankton
mortality in the NPSG, although no diel periodicity of grazing activity was observed. We
estimate that mixotrophic grazers exerted ~25% of total bacterivory during our study (range 5–
76% for a given time-point), while heterotrophic grazers exerted the remaining 75%, based on
the mean grazing pressures and abundances of the photo/mixotrophic and heterotrophic
populations. Previous studies have reported that mixotrophic grazing may constitute ~50% of the
total bacterivory in equatorial (Hartmann et al. 2012, Stukel et al. 2011), subtropical (Hartmann
! 119!
et al. 2012, Sanders et al. 2000, Unrein et al. 2007, Zubkov and Tarran 2008), and temperate
regions (Hartmann et al. 2012, Sanders et al. 2000, Zubkov and Tarran 2008); however, the
contribution of mixotrophic phytoflagellates to total bacterivory in those studies varied greatly
with location, depth, and season (0-90%). This study, to our knowledge, provides the first
estimate of mixotrophic nanoplankton grazing impact in the NPSG, which is considered the
largest contiguous biome on the planet (Sverdrup et al. 1942).
%
3.4.4%The%impact%of%grazing%periodicity%on%other%microbial%assemblages%%
!
Protistan grazers are important sources of nutrients and dissolved organic matter for the
phytoplankton and bacterial communities and important prey items for mesozooplankton (Brum
et al. 2014, Caron and Goldman 1990). Thus, the diel periodicity in heterotrophic nanoplankton
grazing pressure observed during our study suggests that grazer excretion of dissolved organic
matter (DOM) may also be cyclical. Heterotrophic bacteria undergo a coordinated upregulation
of transcripts related to oxidative phosphorylation, metabolite transport, peptidases, amino acid
metabolism, and the citric acid cycle in the afternoon in the NPSG, suggesting that they are
responding to daily pulses of DOM (Aylward et al. 2015). Protistan grazers are thought to be a
major source of DOM production in oligotrophic ocean regimes due to the tight-coupling
between predator and prey populations, although phytoplankton leakage of excess fixed carbon
late in the light period each day may also be an important source of DOM when nutrients are
limiting (Nagata 2000). A combination of rhythmic cycles of phytoplankton exudation and
grazer excretion of DOM could explain the upregulation of bacterial metabolic transcripts in the
afternoon observed by Aylward et al. (2015).
! 120!
Intriguingly, heterotrophic nanoplankton abundances did not show significant diel
periodicity and remained constant throughout the 4-d study period (Fig. 3.2c). This presumably
indicates a synchrony between cycles of nanoplankton production and grazing by higher trophic
levels (micro- or mesozooplankton). Culture studies of microzooplankton grazing on
nanoplankton suggest that grazing by some species is higher during the day; however, results
vary between species (Jakobsen and Strom 2004, Ng et al. 2017). A significant number of
mesozooplankton taxa vertically migrate at night to feed in the surface waters at Station
ALOHA, with an average nighttime zooplankton biomass increase of ~40% in the euphotic zone
(Al-Mutairi and Landry 2001). Thus, it is possible that mesozooplankton were migrating to the
euphotic zone at night, consuming nanoplankton production, and contributing to the stable
nanoplankton population abundances observed during our study. Quantifying the relative
importance of microzooplankton and mesozooplankton to heterotrophic nanoplankton mortality,
as well as the level of synchrony between the activities of these populations, will provide greater
insight into what extent the diel cycle impacts biogeochemical fluxes in marine environments.

3.5 Conclusion
Our findings demonstrate a tightly-coordinated transfer of energy from the primary
producers to their consumers that is in synchrony with the light cycle. The observed coupling
between the cyanobacterial and nanoplanktonic assemblages has implications for the timing and
nature of the movement of elements and energy through the system. We observed daily peaks in
marine picocyanobacterial division at dusk that were consistently followed by peaks in
nanoplankton grazing pressure 4-5 h later. Estimates of daily prey carbon production, prey
carbon removal, and nanoplankton community carbon demand suggest that protistan grazing is
! 121!
capable of removing a significant portion of the daily new production in the system and
contributes to the remarkably stable population abundances observed for marine
picocyanobacteria in the NPSG (Ribalet et al. 2015). Interestingly, the imprint of the diel cycle
was only observed for the heterotrophic nanoplankton, which are solely dependent on prey
carbon to meet their nutritional needs, and not for the mixotrophic nanoplankton. Our results are
consistent with recent speculation that oceanic mixotrophic algae consume prey to alleviate
nutrient limitation. It is also possible that a mixture of mixotrophic species with different
phagotrophic behaviors and timings were present in the assemblage, masking individual diel
periodicities. This study demonstrates the complexity of predator-prey relationships in marine
food webs and emphasizes the importance of quantifying grazer community composition and
nutritional requirements when assessing the nature and timing of biogeochemical fluxes in the
marine environment.
!
! 122!
Chapter Three Literature Cited
Ahlgren!NA,!Rocap!G!(2012)!Diversity!and!distribution!of!marine!Synechococcus:!multiple!
gene!phylogenies!for!consensus!classification!and!development!of!qPCR!assays!for!
sensitive!measurement!of!clades!in!the!ocean.!Front!Microbiol!3:213!
AlXMutairi!H,!Landry!MR!(2001)!Active!export!of!carbon!and!nitrogen!at!Station!ALOHA!by!
diel!migrant!zooplankton.!Deep!Sea!Res!II!48:2083X2103!
Arenovski!AL,!Lim!EL,!Caron!DA!(1995)!Mixotrophic!nanoplankton!in!oligotrophic!surface!
waters!of!the!Sargasso!Sea!may!employ!phagotrophy!to!obtain!major!nutrients.!J!
Plankton!Res!17:801X820!
Axmann!IM,!Dühring!U,!Seeliger!L,!Arnold!A,!Vanselow!JT,!Kramer!A,!Wilde!A!(2009)!
Biochemical!evidence!for!a!timing!mechanism!in!Prochlorococcus.!J!Bacteriol!
191:5342X5347!
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.!PNAS!112:5443X5448!
Biller!SJ,!Berube!PM,!Lindell!D,!Chisholm!SW!(2015)!Prochlorococcus:!the!structure!and!
function!of!collective!diversity.!Nat!Rev!Micro!13:13X27!
Breckels!MN,!Roberts!EC,!Archer!SD,!Malin!G,!Steinke!M!(2011)!The!role!of!dissolved!
infochemicals!in!mediating!predator–prey!interactions!in!the!heterotrophic!
dinoflagellate!Oxyrrhis%marina.!J!Plankton!Res!33:629X639!
Brum!JR,!Morris!JJ,!Décima!M,!Stukel!MR!(2014)!Mortality!in!the!oceans:!causes!and!
consequences.!EcoXDAS!IX!Chapter!2:16X48!
Calbet!A,!Landry!MR!(2004)!Phytoplankton!growth,!microzooplankton!grazing,!and!carbon!
cycling!in!marine!systems.!Limnol!Oceanogr!49:51X57!
Caron!DA!(2001)!Protistan!herbivory!and!bacterivory.!In:!Paul!J!(ed)!Methods!in!
Microbiology,!Book!30.!Academic!Press!Ltd.,!Cambridge,!UK!
Caron!DA,!Connell!PE,!Schaffner!RA,!Schnetzer!A,!Fuhrman!JA,!Countway!PD,!Kim!DY!
(2017)!Planktonic!food!web!structure!at!a!coastal!timeXseries!site:!I.!Partitioning!of!
microbial!abundances!and!carbon!biomass.!Deep!Sea!Res!I!121:14X29!
Caron!DA,!Goldman!JC!(1990)!Protozoan!nutrient!regeneration.!In:!Capriulo!GM!(ed)!
Ecology!of!marine!protozoa.!Oxford!University!Press,!New!York!
Cermak!N,!Becker!JW,!Knudsen!SM,!Chisholm!SW,!Manalis!SR,!Polz!MF!(2017)!Direct!singleX
cell!biomass!estimates!for!marine!bacteria!via!Archimedes'!principle.!ISME!J!11:825X
828!
Christaki!U,!Courties!C,!Karayanni!H,!Giannakourou!A,!Maravelias!C,!Kormas!KA,!Lebaron!P!
(2002)!Dynamic!characteristics!of!Prochlorococcus!and!Synechococcus!consumption!
by!bacterivorous!nanoflagellates.!Microb!Ecol!43:341X352!
Chrzanowski!TH,!Foster!BLL!(2014)!Prey!element!stoichiometry!controls!ecological!fitness!
of!the!flagellate!Ochromonas%danica.!Aquat!Microb!Ecol!71:257X269!
Cohen!SE,!Golden!SS!(2015)!Circadian!rhythms!in!cyanobacteria.!Microbiol!Mol!Biol!Rev!
79:373X385!
Dolan!JR,!Šimek!K!(1999)!Diel!periodicity!in!Synechococcus!populations!and!grazing!by!
heterotrophic!nanoflagellates:!analysis!of!food!vacuole!contents.!Limnol!Oceanogr!
44:1565X1570!
! 123!
Dufresne!A,!Garczarek!L,!Partensky!F!(2005)!Accelerated!evolution!associated!with!genome!
reduction!in!a!freeXliving!prokaryote.!Genome!Biol!6:R14XR14!
Eiler!A,!Hayakawa!DH,!Church!MJ,!Karl!DM,!Rappé!MS!(2009)!Dynamics!of!the!SAR11!
bacterioplankton!lineage!in!relation!to!environmental!conditions!in!the!oligotrophic!
North!Pacific!subtropical!gyre.!Environ!Microbiol!11:2291X2300!
Evans!C,!Wilson!WH!(2008)!Preferential!grazing!of!Oxyrrhis%marina%on!virus!infected!
Emiliania%huxleyi.!Limnol!Oceanogr!53:2035X2040!
Fenchel!T!(1986)!The!Ecology!of!Heterotrophic!Microflagellates.!In:!Marshall!KC!(ed)!
Advances!in!Microbial!Ecology.!Springer!US,!Boston,!MA!
Fenchel!T,!Blackburn!N!(1999)!Motile!chemosensory!behaviour!of!phagotrophic!protists:!
mechanisms!for!and!efficiency!in!congregating!at!food!patches.!Protist!150:325X336!
Field!CB,!Behrenfeld!MJ,!Randerson!JT,!Falkowski!P!(1998)!Primary!production!of!the!
biosphere:!integrating!terrestrial!and!oceanic!components.!Science!281:237X240!
Flombaum!P,!Gallegos!JL,!Gordillo!RA,!Rincón!J,!Zabala!LL,!Jiao!N,!Karl!DM,!Li!WK,!Lomas!
MW,!Veneziano!D!(2013)!Present!and!future!global!distributions!of!the!marine!
cyanobacteria!Prochlorococcus%and!Synechococcus.!PNAS!110:9824X9829!
Fuhrman!JA!(2000)!Impact!of!viruses!on!bacterial!processes.!In:!Kirchman!DL!(ed)!
Microbial!ecology!of!the!oceans.!John!WIley!&!Sons,!Inc.,!New!York!
Fukuda!R,!Ogawa!H,!Nagata!T,!Koike!I!(1998)!Direct!determination!of!carbon!and!nitrogen!
contents!of!natural!bacterial!assemblages!in!marine!environments.!Appl!Environ!
Microbiol!64:3352X3358!
Hartmann!M,!Grob!C,!Scanlan!DJ,!Martin!AP,!Burkill!PH,!Zubkov!MV!(2011)!Comparison!of!
phosphate!uptake!rates!by!the!smallest!plastidic!and!aplastidic!protists!in!the!North!
Atlantic!subtropical!gyre.!FEMS!Microbiol!Ecol!78:327X335!
Hartmann!M,!Grob!C,!Tarran!GA,!Martin!AP,!Burkill!PH,!Scanlan!DJ,!Zubkov!MV!(2012)!
Mixotrophic!basis!of!Atlantic!oligotrophic!ecosystems.!PNAS!109:5756X5760!
Holtzendorff!J,!Partensky!F,!Mella!D,!Lennon!JXF,!Hess!WR,!Garczarek!L!(2008)!Genome!
streamlining!results!in!loss!of!robustness!of!the!circadian!clock!in!the!marine!
cyanobacterium!Prochlorococcus%marinus!PCC!9511.!J!Biol!Rhythms!23:187X199!
HunterXCevera!KR,!Neubert!MG,!Olson!RJ,!Solow!AR,!Shalapyonok!A,!Sosik!HM!(2016)!
Physiological!and!ecological!drivers!of!early!spring!blooms!of!a!coastal!
phytoplankter.!Science!354:326X329!
HunterXCevera!KR,!Neubert!MG,!Solow!AR,!Olson!RJ,!Shalapyonok!A,!Sosik!HM!(2014)!Diel!
size!distributions!reveal!seasonal!growth!dynamics!of!a!coastal!phytoplankter.!PNAS!
111:9852X9857!
Hynes!AM,!Rhodes!KL,!Binder!BJ!(2015)!Assessing!cell!cycle based!methods!of!measuring!
Prochlorococcus!division!rates!using!an!individual based!model.!Limnol!Oceanogr:!
Methods!13:640X650!
Ito!H,!Mutsuda!M,!Murayama!Y,!Tomita!J,!Hosokawa!N,!Terauchi!K,!Sugita!C,!Sugita!M,!
Kondo!T,!Iwasaki!H!(2009)!Cyanobacterial!daily!life!with!KaiXbased!circadian!and!
diurnal!genomeXwide!transcriptional!control!in!Synechococcus%elongatus.!PNAS!
106:14168X14173!
Jakobsen!HH,!Strom!SL!(2004)!Circadian!cycles!in!growth!and!feeding!rates!of!
heterotrophic!protist!plankton.!Limnol!Oceanogr!49:1915X1922!
! 124!
John!EH,!Davidson!K!(2001)!Prey!selectivity!and!the!influence!of!prey!carbon:nitrogen!ratio!
on!microflagellate!grazing.!J!Exp!Mar!Biol!Ecol!260:93X111!
Jürgens!K,!Massana!R!(2008)!Protistan!grazing!on!marine!bacterioplankton.!In:!Kirchman!
DL!(ed)!Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Hoboken,!NJ!
Karayanni!H,!Christaki!U,!Van!Wambeke!F,!Thyssen!M,!Denis!M!(2008)!Heterotrophic!
nanoflagellate!and!ciliate!bacterivorous!activity!and!growth!in!the!northeast!Atlantic!
Ocean:!a!seasonal!mesoscale!study.!Aquat!Microb!Ecol!51:169X181!
Karl!DM,!Church!MJ!(2014)!Microbial!oceanography!and!the!Hawaii!Ocean!TimeXseries!
programme.!Nat!Rev!Microbiol!12:699X713!
Kawasaki!N,!Sohrin!R,!Ogawa!H,!Nagata!T,!Benner!R!(2011)!Bacterial!carbon!content!and!
the!living!and!detrital!bacterial!contributions!to!suspended!particulate!organic!
carbon!in!the!North!Pacific!Ocean.!Aquat!Microb!Ecol!62:165X176!
Kirchman!DL!(2016)!Growth!rates!of!microbes!in!the!oceans.!Annu!Rev!Mar!Sci!8:285X309!
Landry!MR,!Calbet!A!(2004)!Microzooplankton!production!in!the!oceans.!ICES!J!Mar!Sci!
61:501X507!
Liu!Z,!Campbell!V,!Heidelberg!KB,!Caron!DA!(2016)!Gene!expression!characterizes!different!
nutritional!strategies!among!three!mixotrophic!protists.!FEMS!Microbiol!Ecol!
92:fiw106!
Mann!EL,!Ahlgren!N,!Moffett!JW,!Chisholm!SW!(2002)!Copper!toxicity!and!cyanobacteria!
ecology!in!the!Sargasso!Sea.!Limnol!Oceanogr!47:976X988!
Maranger!R,!Bird!D,!Price!N!(1998)!Iron!acquisition!by!photosynthetic!marine!
phytoplankton!from!ingested!bacteria.!Nature!396:248X251!
Monger!BC,!Landry!MR,!Brown!SL!(1999)!Feeding!selection!of!heterotrophic!marine!
nanoflagellates!based!on!the!surface!hydrophobicity!of!their!picoplankton!prey.!
Limnol!Oceanogr!44:1917X1927!
Nagata!T!(2000)!Production!mechanisms!of!dissolved!organic!matter.!In:!Kirchman!DL!(ed)!
Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.!,!New!York,!NY!
Neuer!S,!Cowles!TJ!(1994)!Protist!herbivory!in!the!Oregon!upwelling!system.!Mar!Ecol!Prog!
Ser!113:147X162!
Ng!WHA,!Liu!H!(2016)!Diel!periodicity!of!grazing!by!heterotrophic!nanoflagellates!
influenced!by!prey!cell!properties!and!intrinsic!grazing!rhythm.!J!Plankton!Res!
38:636X651!
Ng!WHA,!Liu!H,!Zhang!S!(2017)!Diel!variation!of!grazing!of!the!dinoflagellate!Lepidodinium%
sp.!and!ciliate!Euplotes%sp.!on!algal!prey:!the!effect!of!prey!cell!properties.!J!Plankton!
Res!39:450X462!
Ni!T,!Zeng!Q!(2016)!Diel!infection!of!cyanobacteria!by!cyanophages.!Front!Mar!Sci!2:123!
Nygaard!K,!Tobieson!A!(1993)!Bacterivory!in!algae:!a!survival!strategy!during!nutrient!
limitation.!Limnol!Oceanogr!38:273X279!
Ottesen!EA,!Young!CR,!Gifford!SM,!Eppley!JM,!Marin!R,!Schuster!SC,!Scholin!CA,!DeLong!EF!
(2014)!Multispecies!diel!transcriptional!oscillations!in!open!ocean!heterotrophic!
bacterial!assemblages.!Science!345:207X212!
Partensky!F,!Hess!W,!Vaulot!D!(1999)!Prochlorococcus,!a!marine!photosynthetic!prokaryote!
of!global!significance.!Microbiol!Mol!Biol!Rev!63:106X127!
Pittera!J,!Humily!F,!Thorel!M,!Grulois!D,!Garczarek!L,!Six!C!(2014)!Connecting!thermal!
physiology!and!latitudinal!niche!partitioning!in!marine!Synechococcus.!ISME!J!
8:1221X1236!
! 125!
R!Core!Team!(2015)!R:!A!language!and!environment!for!statistical!computing.!!
Ribalet!F,!Swalwell!J,!Clayton!S,!Jiménez!V,!Sudek!S,!Lin!Y,!Johnson!ZI,!Worden!AZ,!Armbrust!
EV!(2015)!LightXdriven!synchrony!of!Prochlorococcus!growth!and!mortality!in!the!
subtropical!Pacific!gyre.!PNAS!112:8008X8012!
Rose!JM,!Caron!DA!(2007)!Does!low!temperature!constrain!the!growth!rates!of!
heterotrophic!protists?!Evidence!and!implications!for!algal!blooms!in!cold!waters.!
Limnol!Oceanogr!52:886X895!
Sanders!RW,!Berninger!UXG,!Lim!EL,!Kemp!PF,!Caron!DA!(2000)!Heterotrophic!and!
mixotrophic!nanoplankton!predation!on!picoplankton!in!the!Sargasso!Sea!and!on!
Georges!Bank.!Mar!Ecol!Prog!Ser!192:103X118!
Shannon!SP,!Chrzanowski!TH,!Grover!JP!(2007)!Prey!food!quality!affects!flagellate!
ingestion!rates.!Microb!Ecol!53:66X73!
Sherr!BF,!Sherr!EB,!Fallon!RD!(1987)!Use!of!monodispersed,!fluorescently!labeled!bacteria!
to!estimate!in!situ!protozoan!bacterivory.!Appl!Environ!Microbiol!53:958X965!
Sherr!EB,!Caron!DA,!Sherr!BF!(1993)!Staining!of!heterotrophic!protists!for!visualization!via!
epifluorescence!microscopy.!Handbook!of!methods!in!aquatic!microbial!
ecology:213X227!
Sherr!EB,!Sherr!BF!(2002)!Significance!of!predation!by!protists!in!aquatic!microbial!food!
webs.!Antonie!van!Leeuwenhoek!81:293X308!
Siuda!ANS,!Dam!HG!(2010)!Effects!of!omnivory!and!predatorXprey!elemental!stoichiometry!
on!planktonic!trophic!interactions.!Limnol!Oceanogr!55:2107X2116!
Sosik!HM,!Olson!RJ,!Neubert!MG,!Shalapyonok!A,!Solow!AR!(2003)!Growth!rates!of!coastal!
phytoplankton!from!time series!measurements!with!a!submersible!flow!
cytometer.!Limnol!Oceanogr!48:1756X1765!
Stárka!J!(1962)!Cellular!Division!and!Reproduction!of!Bacteriophage!in!Synchronized!
Cultures!of!Escherichia!coli.!Microbiology!29:83X90!
Stoecker!DK,!Hansen!PJ,!Caron!DA,!Mitra!A!(2017)!Mixotrophy!in!the!marine!plankton.!
Annu!Rev!Mar!Sci!9:311X335!
Storms!ZJ,!Brown!T,!Cooper!DG,!Sauvageau!D,!Leask!RL!(2014)!Impact!of!the!cell!lifeXcycle!
on!bacteriophage!T4!infection.!FEMS!Microbiol!Lett!353:63X68!
Straile!D!(1997)!Gross!growth!efficiencies!of!protozoan!and!metazoan!zooplankton!and!
their!dependence!on!food!concentration,!predatorXprey!weight!ratio,!and!taxonomic!
group.!Limnol!Oceanogr!42:1375X1385!
Stukel!MR,!Landry!MR,!Selph!KE!(2011)!Nanoplankton!mixotrophy!in!the!eastern!
equatorial!Pacific.!Deep!Sea!Res!II!58:378X386!
Suttle!CA!(2000)!Cyanophages!and!their!role!in!the!ecology!of!cyanobacteria.!!The!ecology!
of!cyanobacteria.!Springer!Netherlands!
Sverdrup!HU,!Johnson!MW,!Fleming!RH!(1942)!The!oceans:!their!physics,!chemistry,!and!
general!biology,!Vol!7.!PrenticeXHall!New!York!
Swalwell!JE,!Ribalet!F,!Armbrust!EV!(2011)!SeaFlow:!a!novel!underway!flowXcytometer!for!
continuous!observations!of!phytoplankton!in!the!ocean.!Limnol!OceanogrXMeth!
9:466X477!
Terrado!R,!Pasulka!AL,!Lie!AAY,!Orphan!VJ,!Heidelberg!KB,!Caron!DA!(2017)!Autotrophic!
and!heterotrophic!acquisition!of!carbon!and!nitrogen!by!a!mixotrophic!chrysophyte!
established!through!stable!isotope!analysis.!ISME!J!
! 126!
Thaben!PF,!Westermark!PO!(2014)!Detecting!rhythms!in!time!series!with!RAIN.!J!Biol!
Rhythms!29:391X400!
Tsai!AXY,!Chin!WXM,!Chiang!KXP!(2009)!Diel!patterns!of!grazing!by!pigmented!
nanoflagellates!on!Synechococcus%spp.!in!the!coastal!ecosystem!of!subtropical!
western!Pacific.!Hydrobiologia!636:249X256!
Unrein!F,!Massana!R,!AlonsoXSáez!L,!Gasol!JM!(2007)!Significant!year round!effect!of!small!
mixotrophic!flagellates!on!bacterioplankton!in!an!oligotrophic!coastal!system.!
Limnol!Oceanogr!52:456X469!
Vaulot!D,!Marie!D!(1999)!Diel!variability!of!photosynthetic!picoplankton!in!the!equatorial!
Pacific.!J!Geophys!Res!104:3297X3310!
Vaulot!D,!Marie!D,!Olson!RJ,!Chisholm!SW!(1995)!Growth!of!Prochlorococcus,!a!
photosynthetic!prokaryote,!in!the!equatorial!Pacific!Ocean.!Science!268:1480X1482!
Verity!PG!(1991)!Feeding!in!planktonic!protozoans:!evidence!for!nonXrandom!acquisition!
of!prey.!J!Protozool!38:69X76!
Verity!PG,!Stoecker!DK,!Sieracki!ME,!Nelson!JR!(1993)!Grazing,!growth!and!mortality!of!
microzooplankton!during!the!1989!North!Atlantic!spring!bloom!at!47°N,!18°W.!Deep!
Sea!Res!I!40:1793X1814!
Ward!BA,!Follows!MJ!(2016)!Marine!mixotrophy!increases!trophic!transfer!efficiency,!mean!
organism!size,!and!vertical!carbon!flux.!PNAS!113:2958X2963!
Wootton!EC,!Zubkov!MV,!Jones!DH,!Jones!RH,!Martel!CM,!Thornton!CA,!Roberts!EC!(2007)!
Biochemical!prey!recognition!by!planktonic!protozoa.!Environ!Microbiol!9:216X222!
Zinser!ER,!Lindell!D,!Johnson!ZI,!Futschik!ME,!Steglich!C,!Coleman!ML,!Wright!MA,!Rector!T,!
Steen!R,!McNulty!N,!Thompson!LR,!Chisholm!SW!(2009)!Choreography!of!the!
transcriptome,!photophysiology,!and!cell!cycle!of!a!minimal!photoautotroph,!
Prochlorococcus.!PLoS!ONE!4:e5135!
Zubkov!MV,!Tarran!GA!(2008)!High!bacterivory!by!the!smallest!phytoplankton!in!the!North!
Atlantic!Ocean.!Nature!455:224X226!
Zwirglmaier!K,!Spence!E,!Zubkov!MV,!Scanlan!DJ,!Mann!NH!(2009)!Differential!grazing!of!
two!heterotrophic!nanoflagellates!on!marine!Synechococcus!strains.!Environ!
Microbiol!11:1767X1776!
!

! 127!

CHAPTER(FOUR.(MICROBIAL(CARBON(FLUXES(AT(A(COASTAL(OCEAN(SITE:(AN(INVERSE(
ECOSYSTEM(MODELING(ANALYSIS(

By Paige E. Connell

Coauthors: Thomas B. Kelly, Michael R. Stukel, Jed A. Fuhrman, and David A. Caron
! 128!
Abstract
The activities of microorganisms in marine environments shape the movement of energy
and elements through the ecosystem. Yet, major gaps still exist in our understanding of the
structure and function of microbial food webs due to our limited ability to sample and describe
the complex interactions occurring at these scales. Here, we constructed an inverse ecosystem
model based on data collected at the San Pedro Ocean Time-series (SPOT) station (a coastal
ocean site in the Southern California Bight, CA, USA) to investigate the carbon transformations
at that site. Specifically, our objectives were (1) to characterize the fate of primary production at
the SPOT station, (2) to examine the role that heterotrophic bacteria play in carbon flow, and (3)
to investigate how the inclusion or exclusion of viruses from the model structure impacted
carbon flow. Model results revealed that 8% of total phytoplankton mortality was due to viral
lysis (though viruses comprise ~25% of cyanobacterial mortality), whereas 90% and 2% of
phytoplankton mortality were due to zooplankton grazing and programmed cell death/necrosis,
respectively. The empirically measured ratio of bacterial production to net primary production
(BP:NPP) at the SPOT station was 0.53; however, modeled results indicated that lower BP:NPP
ratios (0.1-0.2) were possible at this coastal site and that our empirical measurements may have
overestimated bacterial production and/or underestimated net primary production. We found that
employing a higher bacterial production value (and thus shifting the BP:NPP ratio in the system
from 0.1 to 0.2) caused higher utilization of carbon by the smallest plankton, resulting in less
carbon consumption by grazers and less sinking flux of carbon out of the surface ocean.
Removing viral lysis from the model structure resulted in slightly lower higher bacterial
production values (~10%) and generally greater trophic transfer of carbon. This study leverages
our present understanding of microbial food web topology and trophic activities to gain new
! 129!
insights into the carbon transformations occurring at the SPOT station. Additionally, this study
highlights the importance of accurately characterizing heterotrophic bacterial processes, and the
processes that support or remove them (dissolved/particulate organic matter uptake and viral
lysis or zooplankton grazing), when investigating the fate of primary production in marine
environments.

! 130!
4.1. Introduction
Marine microbes are an exceedingly diverse group of organisms that span a considerable
range of sizes, forms, and functions (Fuhrman & Hagström 2008, Caron et al. 2012). Comprising
>95% of the biomass in marine environments (Caron et al. 2017), the abundance and diversity of
microorganisms makes them central players in the earth’s biogeochemical cycling of major
elements (C, N, P, S). Microbes represent a diverse array of nutritional strategies including
phototrophy, mixotrophy, and heterotrophy, which enable them to play multifarious roles in
marine food webs (Worden et al. 2015, Stoecker et al. 2017). Thus, characterizing the
biogeochemical pathways mediated by marine microbes is a necessary part of understanding the
ocean carbon cycle.
Acknowledgement of the pivotal roles that microbes play in marine food webs has led to
the creation of several time-series programs dedicated to understanding their activities in the
ocean (e.g. BATS (Steinberg et al. 2001), HOT (Karl & Church 2014), Station L4 (Southward et
al. 2004), CalCOFI (Bograd et al. 2003)). One such program is the San Pedro Ocean Time-series
station (SPOT) located equidistant between the coastline of greater Los Angeles and Santa
Catalina Island in the central Southern California Bight (33°33’N, 118° 24’W;
https://dornsife.usc.edu/spot/). This region has a Mediterranean-type, temperate-to-subtropical
climate characterized by a narrow annual range of temperatures and other physicochemical
properties (Hamersley et al. 2011, Caron et al. 2017). Episodic wind-driven upwelling and
rainfall events drive seasonal changes in regional productivity (Nezlin & Li 2003, Nezlin et al.
2012). Studies at the SPOT station have documented seasonal shifts in community composition
of viruses (Chow et al. 2013, Chow et al. 2014), prokaryotes (Chow et al. 2013, Cram et al.
2014), and single-celled eukaryotes (Countway et al. 2010, Kim et al. 2014, Hu et al. 2016) that
! 131!
are accompanied by modest seasonal shifts in community standing stocks (Caron et al. 2017,
Connell et al. 2017). In addition to measurements of microbial abudances and diversity,
significant efforts have been made to characterize the microbial rate processes at the SPOT
station. These measurements include bacterial production (Fuhrman et al. 2006, Cram et al.
2014), bacterial grazing mortality rates (Connell et al. 2017), phytoplankton growth and grazing
mortality rates (Connell et al. 2017), gross oxygen production rates (Haskell et al. 2017), net
community oxygen production rates (Haskell et al. 2016), sedimentation rates (Collins et al.
2011, Haskell et al. 2016), nitrogen fixation rates (Hamersley et al. 2011), and rates of iron flux
from the sediment pore-waters (John et al. 2012). The extensive information collected on
microbial processes at the SPOT station makes it a valuable location to study and model the
diverse roles microbes play in the ocean environment.
Understanding the fate of phytoplankton primary production in marine food webs is a
central theme in marine science. Photosynthetic fixation of inorganic carbon by marine
phytoplankton is believed to account for ~50% of the global oxygen production (Field et al.
1998) and thus phytoplankton are a major constituent of the global carbon budget. Phytoplankton
biomass directly supports the metabolism and growth of consumers in the food web (Calbet &
Landry 2004, Schmoker et al. 2013), and also contributes to the removal of carbon from the
surface to the deep ocean through fecal packaging by consumers or via particle aggregation and
sinking (Turner 2002). Alternatively, phytoplankton biomass may be converted to dissolved
organic matter by leakage of carbon from phytoplankton, viral lysis, and grazer exudation/sloppy
feeding (Nagata 2000); these pathways contribute dissolved organic matter that supports the
growth and metabolism of the heterotrophic bacterial assemblage. The environmental and
biological factors that govern these processes are complex and dynamic and the small scales on
! 132!
which these processes operate often prohibit precise characterization of individual rate processes.
Thus, while significant strides have been made to characterize the movement of phytoplankton
carbon through marine food webs during the past several decades, much still needs clarification.
We investigated the fate of primary production in surface waters at the SPOT station by
constructing a linear inverse model of the ecosystem and constraining it using empirical data
from the study site, as well as literature data. Linear inverse modeling is a data integration tool
that depicts an ecosystem as a linear function of its’ carbon transformations (flows) and
mathematically estimates these flows using empirical data (Soetaert & Van Oevelen 2009, van
Oevelen et al. 2010). Linear inverse models have been successfully applied to characterize
marine food webs (Vézina et al. 1997, Richardson et al. 2004, Stukel & Landry 2010). This
integrative tool does not require a mechanistic understanding of the processes governing flows,
and thus enables an investigation of the carbon transformations that cannot or have not been
observed. The process relies on quadratic and linear programming techniques to generate
mathematically driven model solutions that integrate the provided site-specific and literature-
derived data.
We also examined the role that heterotrophic bacteria play in the marine carbon cycle.
Specifically, we investigated the relationship between bacterial production and net primary
production (BP:NPP ratio) and how the fate of bacterial carbon (i.e. viral lysis or protistan
grazing) impacts carbon flow. The heterotrophic bacteria (bacteria + archaea) are a large and
active component of the microbial community at the SPOT station (Cram et al. 2014),
comprising more than one-third of the total biomass at the study site (Caron et al. 2017) and
contributing significant amounts of carbon to protistan grazers, especially in the fall and winter
seasons (Connell et al. 2017). A substantial amount of primary production supports the growth
! 133!
and metabolism of the bacterial assemblage, either directly (through phytoplankton exudation) or
indirectly (through phytoplankton mortality by viruses and zooplankton) (Nagata 2000). Bacteria
can be removed by viral lysis, a process which diverts carbon from high trophic levels and
redirects carbon to the dissolved and particulate organic matter pools (“the viral shunt”)
(Wilhelm & Suttle 1999, Breitbart et al. 2008). Alternatively, bacteria can be consumed by
protistan grazers and passed to higher trophic levels, recovering some of the carbon that would
have been lost from the system due to excretion and exudation (“the microbial loop”) (Azam et
al. 1983, Sherr & Sherr 2008). Therefore, characterizing the processes that support or remove
bacterial carbon at the SPOT station is important as they affect the amount of energy that is
recycled within the surface ocean or made available to higher trophic levels or for export (Brum
et al. 2014).
Model results revealed that zooplankton grazing was the primary cause of total
phytoplankton mortality at the SPOT station (90% of total phytoplankton mortality, ~75% of
cyanobacterial mortality). Additionally, model results suggested that empirical measurements at
the SPOT station might overestimate bacterial production and/or underestimate net primary
production at the study site. Imposing a higher bacterial production value on the model system
resulted in less trophic transfer of carbon and less sinking flux of organic matter from the surface
ocean. Excluding viral lysis from the model structure resulted in slightly lower values of
bacterial production (~10% less) and generally increased trophic transfer of carbon. This study
provides new insights into the fate of primary production at the SPOT station and emphasizes the
importance of accurately characterizing the relationship between the heterotrophic bacterial
assemblage and their sources of DOC and mortality when quantifying carbon transformations in
microbial food webs.
! 134!
4.2 Methods
4.2.1%Environmental%sampling%and%model%data%
!
In this study, we sought to represent the average movement of carbon between living and
non-living organic compartments in the mixed-layer at the San Pedro Ocean Time-series Station
(SPOT, 33° 33’N, 118° 24’ W). Microbial standing stocks and rate processes measured at the
SPOT station in previous studies were used as experimental inputs to the model or as biological
constraints on the model solution (Table 4.1). All measurements of pertinent experimental inputs
from the time-series were compiled to compute an average ecosystem state for the SPOT station.
These values and processes were measured during several cruises across several years. Seasonal
fluctuations in oceanographic conditions and biological response is muted in this region (Caron
et al. 2017), but parameters were evenly sampled throughout the annual cycle and thus equitably
capture the mean and range of values measured at the SPOT station. All experimental
measurements were made using water collected from the upper 5 m of the water column, with
the exception of gross primary production (see below).
Measurements of microbial standing stocks (bacteria – mesozooplankton) were calculated
by multiplying the abundance of each microbial group by an appropriate carbon conversion
factor from the literature (Caron et al. 2017, Connell et al. 2017). Briefly, picoplankton
abundances (heterotrophic and phototrophic bacteria + small eukaryotes, 0.2 – 2.0 µm in
diameter) were measured using flow cytometric analysis on a FACSCalibur flow cytometer
(Becton Dickinson, San Jose, California). Nanoplankton abundances (microbial eukaryotes, 2.0 –
20 µm in diameter) were measured using epifluorescence microscopy, differentiating between
photo/mixotrophic and heterotrophic nanoplankton by the presence or absence of chlorophyll
autofluorescence when viewed under blue-light excitation. Microplankton abundances (microbial
! 135!
Table 4.1. Experimental inputs to the SPOT linear inverse model. All standing stock measurements have units of mg C m
3
and all rate processes have units of mg C m
3
d
1
. Additional information for each measurement includes the number of
measurements taken (n), the time-period for sampling, the reference for any data that has been previously published, and
the carbon conversion factor used to calculate standing stocks from abundances (see Caron et al. 2017) or the model
equation for each rate process that was used as a model input. Bolded values indicate fluxes that are included in the model
scenarios containing viral lysis. Abbreviations include: total phytoplankton (PHYTO), Prochlorococcus (PRO),
Synechococcus (SYN), and phototrophic picoeukaryotes (PEUK).
Mean (± SD) n Years Reference C Conversion Factor or Model Equation
Standing Stock

H. bacteria 21.7 (± 13.5) 41 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 15 fg C cell
−1

Prochlorococcus 1.71 (± 2.42) 43 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 90 fg C cell
−1

Synechococcus 5.77 (± 6.67) 43 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 200 fg C cell
−1

P. picoeukaryotes 6.44 (± 10.6) 43 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 183 fg C µm
−3

P. nanoplankton 2.15 (± 1.58) 33 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 183 fg C µm
−3

Diatoms 7.23 (± 10.6) 44 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 138 pg C cell
−1

H. nanoplankton 7.29 (± 8.95) 36 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 183 fg C µm
−3

Dinoflagellates 1.05 (± 1.06) 45 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 138 pg C cell
−1

Ciliates 0.75 (± 0.78) 45 2000-2003; 2012-2014 Caron et al. 2017; Connell et al. 2017 138 pg C cell
−1

Mesozooplankton 5.17 (± 5.89) 86 2011-2016 CALCOFI Database 21 mg C cm
−3
DV

Rate Processes

Bacterial Production 4.96 (± 5.77) 145 2000-2015 This study
docTObact - bactTOresp
Bacterial Consumption 4.83 (± 3.29) 7 2012-2014 Connell et al. 2017
bactTOhnan + bactTOdino + bactTOcil
PHYTO Net Production 9.35 (± 11.7) 9 2012-2014 Connell et al. 2017
gppTOpro - proTOresp - proTOdocE - proTOdocV + gppTOsyn -
synTOresp - synTOdocE - synTOdocV + gppTOpeuk - peukTOresp -
peukTOdoc + gppTOpnan - pnanTOresp - pnanTOdoc + gppTOdiat -
diatTOresp – diatTOdoc
PHYTO Consumption 11.0 (± 16.4) 9 2012-2014 Connell et al. 2017
proTOhnan + proTOdino + proTOcil + synTOhnan + synTOdino + synTOcil
+ peukTOhnan + peukTOdino + peukTOcil + pnanTOdino + pnanTOcil +
diatTOdino
PRO Net Production 0.57 (± 1.82) 10 2012-2014 Connell et al. 2017
gppTOpro - proTOresp - proTOdocE - proTOdocV
PRO Consumption 1.60 (± 1.88) 10 2012-2014 Connell et al. 2017
proTOhnan + proTOcil + proTOdino
SYN Net Production 3.31 (± 4.52) 10 2012-2014 Connell et al. 2017
gppTOsyn - synTOresp - synTOdocE - synTOdocV
SYN Consumption 5.63 (± 10.8) 10 2012-2014 Connell et al. 2017
synTOhnan + synTOcil + synTOdino
PEUK Net Production 0.81 (± 2.46) 10 2012-2014 Connell et al. 2017
gppTOpeuk - peukTOresp - peukTOdoc
PEUK Consumption 1.74 (± 3.07) 10 2012-2014 Connell et al. 2017
peukTOhnan + peukTOcil + peukTOdino
Gross Primary Production 63.3 (± 70.4) 21 2013-2014 Haskell et al. 2016; Haskell et al. 2017
gppTOpro + gppTOsyn + gppTOpeuk + gppTOpnan + gppTOdiat
eukaryotes, 2.0-20 µm in diameter) were enumerated from settled, formalin-preserved samples 1"
using inverted light microscopy. Samples for mesozooplankton biomass (based on total 2"
displacement volume) were collected at the SPOT station by conducting oblique daytime tows 3"
(200 µm Nitex mesh) in the upper 100 m of the water column. This method of collection may 4"
underestimate mesozooplankton biomass due to net avoidance and diel vertical migration below 5"
the sampling depth. We therefore calculated a regional average mesozooplankton biomass using 6"
several California Cooperative Oceanic Fisheries Investigations stations near the SPOT station 7"
(CalCOFI, http://new.data.calcofi.org/index.php/reporteddata/hydrographic-data/zooplankton- 8"
biomass-reports). The average standing stock of each plankton assemblage was used to calculate 9"
minimum and maximum production constraints for each microbial compartment (growth rate × 10"
biomass), as well as maximum constraints on respiration and DOC uptake by the heterotrophic 11"
bacteria (Table 4.2). 12"
Microbial rate processes measured at the SPOT station were used as experimental inputs to 13"
the model. Bacterial production was measured by the incorporation of
3
H-thymidine into 14"
bacterial DNA in triplicate seawater samples, as detailed in Fuhrman et al. (2006). Bacterial 15"
mortality rates were quantified by assessing rates of removal of fluorescently-labeled bacterial 16"
particles by protistan grazers in 24-h, whole seawater incubations (Connell et al. 2017). 17"
Phytoplankton growth and mortality rates were measured using five-point, nutrient-amended 18"
dilution experiments (Connell et al. 2017). Total phytoplankton growth and mortality rates were 19"
calculated using changes in chlorophyll a as a proxy for phytoplankton biomass, while the 20"
growth and mortality rates of the individual picophytoplankton groups (Prochlorococcus, 21"
Synechococcus, and phototrophic picoeukaryotes) were calculated using cell abundances as 22"
determined by flow cytometry. Phytoplankton growth and mortality rates were multiplied by 23"
" 137"
Table 2. Minimum and maximum biological constraints on the model solution. 24"
Superscripts denote the reference from which the constraint was derived. 25"
Biological Group Process Minimum Maximum
Heterotrophic
Bacteria (BACT)
Gross Growth Efficiency
a
Fixed at 15% Ingestion
Respiration
b
20% Ingestion 1.7*W^(-0.25)*e^(0.0693(T-20))*biomass
DOC Uptake
b
- 63*W^(-0.25)*e^(0.0693(T-20))*biomass
Cell Death - 1% Production
Growth Rate
c
- 1.0 d
-1

Prochlorococcus
(PRO)
Respiration
b
5% GPP 30% GPP
Exudation
d
2% GPP 20% GPP
Growth Rate
c
- 1.0 d
-1

Cell Death - 1% Production
Synechococcus
(SYN)
Respiration
b
5% GPP 30% GPP
Exudation
d
2% GPP 20% GPP
Growth Rate
c
- 1.0 d
-1

Cell Death - 1% Production
Phototrophic
Picoeukaryotes
(PEUK)
Respiration
b
5% GPP 30% GPP
Exudation
d
2% GPP 22% GPP
Growth Rate
c
- 1.0 d
-1

Cell Death - 2% Production
Phototrophic
Nanoplankton
(PNAN)
Respiration
b
5% GPP 30% GPP
Exudation
d
2% GPP 20% GPP
Growth Rate
c
- 1.0 d
-1
Cell Death - 2% Production
Diatoms
(DIAT)
Respiration
b
5% GPP 30% GPP
Exudation
d
2% GPP 20% GPP
Growth Rate
c
- 1.0 d
-1

Cell Death - 5% Production
Export <10% NPP may directly sink
Heterotrophic
Nanoplankton
(HNAN)
Gross Growth Efficiency
f
Fixed at 25% Ingestion
Respiration 10% Ingestion 40% Ingestion
Excretion
d
10% Ingestion 40% Ingestion
Egestion
d
10% Ingestion 40% Ingestion
Growth Rate
g
0.1 d
-1
1.0 d
-1

Prey Preference By biomass within the picophytoplankton (PRO, SYN, PEUK)
Picophytoplankton consumption > Bacterial consumption (BACT)
Cell Death - 2% Production
Dinoflagellates
(DINO)
Gross Growth Efficiency
f
Fixed at 25% Ingestion
Respiration 10% Ingestion 40% Ingestion
Excretion
d
10% Ingestion 40% Ingestion
Egestion
d
10% Ingestion 40% Ingestion
Growth Rate
g
0.1 d
-1
1.0 d
-1

Prey Preference By biomass within the picophytoplankton (PRO, SYN, PEUK)
Picophytoplankton consumption > Bacterial consumption (BACT)
By biomass within large plankton (PNAN, HNAN, DIAT)
Large plankton consumption > Picoplankton consumption
Cell Death - 3% Production
Ciliates
(CIL)
Gross Growth Efficiency
f
Fixed at 25% Ingestion
Respiration 10% Ingestion 40% Ingestion
Excretion
d
10% Ingestion 40% Ingestion
Egestion
d
10% Ingestion 40% Ingestion
Growth Rate
g
0.1 d
-1
1.0 d
-1

Prey Preference By biomass within the picophytoplankton (PRO, SYN, PEUK)
Picophytoplankton consumption > Bacterial consumption (BACT)
By biomass within large plankton (PNAN, HNAN)
Large plankton consumption > Picoplankton consumption
" 138"
Cell Death - 3% Production
Mesozooplankton
(MESO)
Gross Growth Efficiency
f
Fixed at 25% Ingestion
Assimilation Efficiency
b
50% 90%
Respiration 10% Ingestion 40% Ingestion
Excretion
b,d
10% Ingestion 100% Respiration
Growth Rate
g
0.01 d
-1
0.5 d
-1

Prey Preference By biomass within the nanoplankton (PNAN, HNAN)
By biomass within the microplankton (DIAT, DINO, CIL)
Microplankton consumption > Nanoplankton consumption
Organic Matter

Export (LDET & DIAT)
h
0.84 mg C m
3
d
-1
(Trap)
1.24 mg C m
3
d
-1
(Th)
-
DOC Downward Mixing
h
- 15% of DOC produced
a
del Giorgio and Cole (2000), Robinson (2008) 26"
b
Stukel and Landry (2010) 27"
c
Kirchman (2016) 28"
d
Nagata (2000) 29"
e
Sarthou (2005) 30"
f
Straile (1997) 31"
g
Hansen et al. (1997)

32"
h
Haskell et al. (2016) 33"
" 139"
phytoplankton standing stocks to calculate rates of net primary production and carbon
consumption, respectively. Measurements of gross oxygen production (GOP), based on the triple
isotope composition of dissolved oxygen, and net oxygen production (NOP), based on O
2
:Ar
ratios, were conducted at the SPOT station and calculated for the mixed layer using a one-
dimensional, two-box, non-steady state model of the euphotic zone (Haskell et al. 2016, Haskell
et al. 2017); these measurements were then used to calculate gross primary production (GPP) for
the mixed layer according to the following equation: GPP = (NOP/1.4) + ((GOP – NOP)/1.1)
(Laws 1991, Bender et al. 1999). Integrated GPP values for the mixed layer (mg C m
–2
d
–1
) were
converted to depth-specific values (mg C m
–3
d
–1
) by dividing the integrated values by the
corresponding mixed layer depth.
!
4.2.2!Model!structure!
"
In total, the model structure was comprised of thirteen compartments and seventy-three
fluxes. The model structure consisted of three organic carbon compartments (dissolved organic
carbon (DOC) and two size classes of particulate matter (large and small detritus; LDET and
SDET)), a heterotrophic bacterial compartment, five phytoplankton compartments
(Prochlorococcus, Synechococcus, phototrophic picoeukaryotes, phototrophic nanoplankton, and
diatoms), three protistan grazer compartments (heterotrophic nanoplankton, dinoflagellates, and
ciliates), and one mesozooplankton compartment (Fig. 4.1; Table 4.3). We chose to only
incorporate dinoflagellates as a heterotrophic compartment for three reasons: (1) while
dinoflagellate taxa are comprised of phototrophic and non-phototrophic organisms in
approximately equal proportions, many of the phototrophic dinoflagellates have mixotrophic
capabilities (Taylor et al. 2008), (2) dinoflagellate prey preferences and feeding mechanisms
" 140"
Figure 4.1. Structure of SPOT model food web. Model compartments include:
heterotrophic bacteria (BACT), Prochlorococcus (PRO), Synechococcus (SYN),
phototrophic picoeukaryotes (PEUK), phototrophic nanoplankton (PNAN), diatoms
(DIAT), heterotrophic nanoplankton (HNAN), dinoflagellates (DINO), ciliates (CIL),
mesozooplankton (MESO), dissolved organic carbon (DOC), small detritus (SDET), and
large detritus (LDET). Energy enters the model through gross primary production (GPP)
and exits the model through downward mixing of DOC, sedimentation of DIAT and LDET,
and consumption by higher trophic levels (EXIT). (a) Schematic depicting the various
biological/chemical compartments and the carbon transformations occurring in the
model, with the arrow colors corresponding to the process being modeled (arrow
definitions at bottom of food web). Permanent fluxes included in the model structure are
depicted with solid lines, while fluxes that were included during a subset of the model
runs are depicted with dashed lines (viral lysis). (b) Schematic depicting the relative
importance of each carbon flux, with the weight of the arrow proportional to the value of
the carbon flux. Colors correspond to the compartment from which each flux originates.
The food web depicted in (b) corresponds to the mean model solution for the ‘Average’
scenario discussed in the manuscript text.

! 141!
Table 4.3. Steady-state equations for each model compartment in the SPOT linear inverse model. Bolded values indicate
fluxes that are included in the model scenarios containing viral lysis (‘Average’ and ‘+ BP’ scenarios).
Compartment Compartment
Abbreviation
Steady-state equation
H. bacteria BACT docTObact – bactTOresp – bactTOsdet – bactTOhnan – bactTOdino – bactTOcil – bactTOdocV
Prochlorococcus PRO gppTOpro – proTOsdet – proTOresp – proTOdocE – proTOhnan – proTOdino – proTOcil – proTOdocV
Synechococcus SYN gppTOsyn – synTOsdet – synTOresp – synTOdocE – synTOhnan – synTOdino – synTOcil – synTOdocV
P. picoeukaryotes PEUK gppTOpeuk – peukTOsdet – peukTOresp – peukTOdocE – peukTOhnan – peukTOdino – peukTOcil
P. nanoplankton PNAN gppTOpnan – pnanTOldet – pnanTOresp – pnanTOdocE – pnanTOdino – pnanTOcil – pnanTOmeso
Diatoms DIAT gppTOdiat – diatTOldet – diatTOresp – diatTOdocE – diatTOdino – diatTOmeso – diatTOexit
H. nanoplankton HNAN sdetTOhnano + bactTOhnano + proTOhnan + synTOhnan + peukTOhnan – hnanTOsdet – hnanTOresp –
hnanoTOdoc – hnanTOdino – hnanTOcil – hnanTOmeso – hnanToldet
Dinoflagellates DINO sdetTOdino + bactTOdino + proTOdino + synTOdino + peukTOdino + hnanTOdino + pnanTOdino + diatTOdino
– dinoTOdoc – dinoTOsdet – dinoTOresp – dinoTOmeso – dinoTOldet
Ciliates CIL sdetTOcil + bactTOcil + proTOcil + synTOcil + peukTOcil + hnanTOcil + pnanTOcil – cilTOdoc – cilTOsdet –
cilTOresp – cilTOmeso – cilTOldet
Mesozooplankton MESO ldetTOmeso + hnanTOmeso + pnanTOmeso + diatTOmeso + dinoTOmeso + cilTO meso - mesoTOdoc -
mesoTOldet - mesoTOresp - mesoTOexit
Dissolved Organic
Carbon
DOC sdetTOdoc + ldetTOdoc + bactTOdocV + proTOdocE + proTOdocV + synTOdocE + synTOdocV + peukTOdoc
+ diatTOdoc + hnanTOdoc + dinoTOdoc + cilTOdoc + mesoTOdoc – docTObact – docTOexit
Small Detritus SDET bactTOsdet + proTOsdet+ synTOsdet + peukTOsdet + hnanTOsdet + dinoTOsdet+ cilTOsdet – sdetTOhnano –
sdetTOdino – sdetTOcil – sdetTOdoc
Large Detritus LDET pnanTOldet + diatTOldet + hnanTOldet + dinoTOldet + cilTOldet + mesoTOldet – ldetTOexit – ldetTOmeso –
ldetTOdoc

! 142!
differ from other protistan grazers, making them critical consumers of diatoms in marine
ecosystems (Hansen & Calado 1999, Sherr & Sherr 2007), and (3) dinoflagellate biomass is a
small component of total microbial biomass at the SPOT station and would have little impact on
total phytoplankton production (Table 4.1). All phytoplankton compartments contributed to the
DOC pool through exudation and all grazer groups contributed to the DOC pool through
excretion/sloppy feeding. Grazer groups contributed to the detrital pool in two ways:
heterotrophic protists formed SDET through the egestion of undigested food particles and
mesozooplankton grazers formed LDET through the production of fecal pellets. In addition, all
groups of single-celled organisms contributed to the detrital pool by programmed cell
death/necrosis, with picoplanktonic compartments forming SDET and all compartments >2.0 µm
forming LDET. Both SDET and LDET contributed to DOC through solubilization, representing
both direct dissolution of detritus as well as microbial-mediated dissolution of detritus.
Biological compartments were allowed to die primarily through grazing and/or viral lysis.
Heterotrophic nanoplankton were allowed to consume all picoplanktonic compartments,
dinoflagellates were allowed to consume all phytoplanktonic compartments (including
phototrophic nanoplankton and diatoms) and the heterotrophic nanoplankton, ciliates were
allowed to consume all picoplanktonic and nanoplanktonic compartments, and mesozooplankton
were allowed to consume all nanoplanktonic and microplanktonic compartments. In addition, all
grazer groups were allowed to consume detritus smaller than themselves, with protistan grazers
consuming from the SDET pool and mesozooplankton consuming from the LDET pool. Viral
lysis was also included as an implicit mortality term for all prokaryotic compartments, directly
shunting organic carbon to the dissolved pool. Viral lysis was not included in all model scenarios
(see section 4.2.4) and thus is represented as a dashed line in Fig. 4.1a. Carbon was enabled to
! 143!
exit the ecosystem through four routes (EXIT; Fig. 4.1): consumption of mesozooplankton
biomass by higher trophic levels, sedimentation of large detritus and diatoms, downward
diffusion of DOC, and respiration by all biological compartments. The direct sedimentation of
live diatom particles is supported by molecular (Hu et al. 2016) and microscopic (Nelson et al.
1987, Caron et al. 2017) evidence of live diatoms at depth in our study region, ostensibly due to
the rapid sinking of these dense, large phototrophs during bloom conditions (Schnetzer et al.
2007).
In our effort to represent the average ecosystem structure at the SPOT station, we constrained
the standing stocks of the thirteen model compartments to steady-state conditions. The Southern
California Bight functions as a temperate-to-subtropical region, with the SPOT station
experiencing low seasonal variability in temperature and other physicochemical properties
(Hamersley et al. 2011, Caron et al. 2017). This results in low interseasonal and interannual
variability in microbial standing stocks and rate processes at the SPOT station (Caron et al. 2017,
Connell et al. 2017). Thus, we believe the assumption of steady-state conditions is appropriate
when representing the average ecosystem structure at the study site. Our choice to combine
biomass and rate measurements from multiple years and seasons in this study (Table 4.1) was
further validated by examining the variability between biomass measurements made in different
years, with no distinct outlier years detected (data not shown).
!
4.2.3!Model!constraints!from!the!literature!
!
Inequality equations were used to set minimum and maximum biological constraints on the
model solution (Table 4.2). DOC uptake and respiration by heterotrophic bacteria were
maximally constrained using allometric relationships from the literature (Stukel & Landry 2010).
! 144!
Phytoplankton respiration was allowed to vary from 5 – 30% of the gross primary production of
each assemblage (Stukel & Landry 2010) and phytoplankton exudation was allowed to vary from
2 – 20% of the gross primary production of each assemblage (Nagata 2000). Protistan grazers
were permitted to respire 10 – 40% of their ingested carbon, excrete 10 – 40% of their ingested
carbon, and egest 10 – 40% of their ingested carbon (Nagata 2000). Mesozooplankton were
given an assimilation efficiency of 50 – 90% and were allowed to respire 10 – 40% of their
ingestion and to excrete 10% of their ingestion up to 100% of their respiration (Nagata 2000,
Stukel & Landry 2010). Heterotrophic bacteria were given a fixed gross growth efficiency of
15% (del Giorgio & Cole 2000, Robinson 2008) and protistan and metazoan grazers were given
fixed gross growth efficiencies of 25% (Straile 1997). Prey preference was specified for each
grazer compartment based on the size and relative standing stock of each prey component (Table
2; Table 1) (Fenchel 1987). The maximum growth rate of all microbial compartments was
constrained to 1.0 d
–1
with the exception of the mesozooplankton, which were allowed to have a
maximum growth rate of 0.5 d
–1
(Hansen et al. 1997, Sarthou et al. 2005, Kirchman 2016).
Finally, while little is known about the relative importance of programmed cell death/necrosis in
marine ecosystems (Brum et al. 2014), we allowed a small fraction of cell production of single-
celled organisms (1-5%) to die and directly become detritus, with the amount of cell death
increasing with the size of the organism (Table 4.2).
Inequality constraints were also implemented to constrain the fluxes of detritus and organic
matter in the model (Table 4.2). Sedimentation of large organic material (large detritus and
diatoms) was constrained using sediment-trap data and
234
Th:
238
U dis-equilibrium data from the
SPOT station (Haskell et al. 2016). A maximum of 10% of diatom net primary production was
permitted to contribute to vertical flux in the model. Additionally, a maximum of 15% of the
! 145!
DOC produced was allowed to leave the system due to downward mixing (Haskell et al. 2016).
!
4.2.4!Model!solution!and!statistical!analyses!!
!
Despite the large amount of empirical data (Table 4.1) and literature data (Table 4.2)
available to constrain the solution space of the model, there remained an infinite number of
possible solutions to the model problem (i.e. the model is an under-determined system). We
implemented a Markov-Chain Monte Carlo (MCMC) sampling technique to thoroughly sample
the solution space and generate solution statistics for each flow in the model solution. To achieve
this goal, we utilized the R package limSolve (Soetaert et al. 2009), specifically implementing the
xsample() function to randomly sample the solution space and generate a marginal probability
density function for each unknown (Van den Meersche 2009). Mass balances of each
compartment were implemented as exact equalities (i.e. standing stocks fixed at steady-state),
while experimental inputs were included as approximate equalities in the model, which allowed
for deviation from the mean values as weighted by normal distributions generated from the
associated error of each input measurement (Table 4.1) (Soetaert & Van Oevelen 2009). The
MCMC technique has been shown to provide a more robust depiction of ecosystem processes
than the least-squares minimum norm (L
2
MN) solution (Stukel et al. 2012); the latter solution
method, which minimizes the sum of the squared flows, tends to select extreme solutions over
ecologically-likely solutions (Kones et al. 2006, Stukel & Landry 2010).
We constructed a linear inverse model to characterize the fate of primary production at the
SPOT station using our present knowledge of food web topology and the empirical
measurements (and errors) from the SPOT station (referred to henceforth as the ‘Average’
scenario). Model results from the ‘Average’ scenario suggested that the empirically measured
! 146!
ratio of bacterial production to net primary production (BP:NPP) at the SPOT station was much
higher than what could be reasonably supported in this system (see Results and Discussion).
Therefore, we conducted an additional model scenario in which we imposed the empirically
measured bacterial production value (4.96 mg C m
–3
d
–1
) as an exact equality on the model
solution (‘+ BP’ scenario). This model scenario shifted the BP:NPP ratio from ~0.1 to ~0.2 (see
Result and Discussion) and thus strengthened the role of the heterotrophic bacterial assemblage
in the model ecosystem. Both of these model scenarios contained viral lysis of the picoplankton
(cyanobacteria and heterotrophic bacteria) in the model structure. We conducted two additional
model scenarios to examine the role of viral lysis in carbon flow; these two scenarios excluded
viral lysis from the model structure and used the model-selected bacterial production value (‘–
Virus’ scenario) or imposed the higher, empirically-measured bacterial production value (‘+ BP
– Virus’). In summary, the four model scenarios were as follows (1) with viral lysis included,
using the SPOT means and SDs provided in Table 1 (scenario referred to in manuscript figures
as ‘Average’), (2) with viral lysis included, setting the bacterial production value at the empirical
measurement mean of 4.96 mg C m
–3
d
–1
(‘+ BP’), (3) removing viral lysis, using the SPOT
means and SDs provided in Table 1 (‘– Virus’), and (4) removing viral lysis and setting the
bacterial production value at the empirical measurement mean of 4.96 mg C m
−3
d
–1
(‘+ BP –
Virus’).
For each model scenario, we started the random walk by calculating the L
2
MN solution,
conducted a burn-in period of 100,000 iterations, then ran the model for an additional 1,000,000
iterations to generate a set of 1000 solutions for each model flow (n=73). The 1000 solutions
were subsampled at even intervals (1,000,000 iterations/1000 solutions = subsample every 1000
th

solution) to retain pseudo-independence of the solutions. Model convergence was visually
! 147!
verified and tests were conducted to confirm that increasing the number of iterations by an order
of magnitude did not impact model results (data not shown). Additionally, we aimed for an
acceptance ratio of ~0.234 as this has been shown to most efficiently achieve model convergence
(Roberts et al. 1997). The mean solution was then calculated for each model flow, as this has
been shown to provide the most robust representation of the flow (Saint-Béat et al. 2013). Model
scenarios were compared using the solution means for each flow.
In addition to quantifying the direct flows through the ecosystem (movement of carbon
directly from one compartment to another), we quantified the indirect flows of energy through
the ecosystem. Indirect flows are defined here as the amount of energy a given compartment
derives from another compartment without a direct connection between the compartments; for
example, mesozooplankton derive energy from picoplankton, but only through the consumption
of picoplankton by protistan grazers. We calculated the amount of energy each compartment
derives indirectly and directly from another compartment by multiplying the matrix (I-G)
–1
,
where G is the normalized production matrix and I is the identity matrix, by the total energy of
the compartments (Hannon 1973, Stukel & Landry 2010).

4.3 Results
The mean values for the seventy-three model flows in the ‘Average’ model scenario
(Scenario 1) ranged from 0.01 to 14.5 mg C m
−3
d
−1
(Fig. 4.2A). The largest flows were those
related to bacterial production (BP; docTObact and bactTOresp) and to the gross primary
production of Synechococcus, phototrophic picoeukaryotes, and diatoms (Figs. 4.1B & 4.2B
‘Average’); the largest flows were also the flows with the greatest uncertainty (Fig. 4.2A). Gross
primary production (GPP) of the phytoplankton was 29.7 mg C m
−3
d
−1
, with each compartment
!
! 148!
Figure 4.2. SPOT model solution values for individual carbon fluxes in the model. (a)
Solutions for every flux (n = 73) for the ‘Average’ model scenario. Boxplots display the
distribution of the data: the rectangle shows the interquartile range (IQR; bottom is the
first quartile, top is the third quartile), the horizontal segment within each box shows the
median, and the vertical lines show the minimum and maximum values that fall within
1.5*IQR. Pink points indicate outliers as identified by R according to the Boxplot Rule
(Wilcox 2011). (b) The ten largest mean fluxes for each of the three model scenarios
(columns). The four model scenarios include (1) ‘Average’: with viral lysis and using
SPOT measurement error, (2) ‘+ BP’: with viral lysis and setting the bacterial production
value to empirical measurement mean (3) ‘− Virus’: without viral lysis and using SPOT
measurement error, and (4) ‘+ BP – Virus’: without viral lysis and setting the bacterial
production value to empirical measurement mean. Colors broadly categorize each flux:
total gross primary production (gray), bacterial growth (peach), phytoplankton growth
(purple), grazer growth (blue), viral lysis (green).

! 149!
contributing to GPP in proportion to their standing stock at the SPOT station (Table 4.1): 34%
diatom, 26% Synechococcus, 24% phototrophic picoeukaryote, 8% phototrophic nanoplankton,
and 8% Prochlorococcus production. In the ‘Average’ model scenario, phytoplankton
compartments consistently respired ~21% of their GPP and exuded 14% of their GPP, resulting
in net primary production (NPP) values that were ~65% of GPP values. Imposing a higher
bacterial production value resulted in more exudation by the phytoplankton compartments, with
phytoplankton exuding 20% of their GPP in both the ‘+ BP’ (Scenario 4) and ‘+ BP – Virus’
(Scenario 4) model scenarios.
The identity of the top five model flows remained constant among all model scenarios,
however the magnitude of the flows was greater when higher bacterial production values were
used (Fig. 4.2B). In the ‘+ BP’ model scenario, where the imposed bacterial production value
(4.96 mg C m
−3
d
−1
) was approximately double the value given by the ‘Average’ model solution
(2.17 mg C m
−3
d
−1
), the ecosystem GPP was 56% higher with a value of 46.3 mg C m
−3
d
−1

(Fig. 4.2B, Table A8). Similarly, using a higher bacterial production value but removing the
flows related to viral lysis (‘+ BP – Virus’) resulted in a substantially higher GPP value (45.8 mg
C m
−3
d
−1
) than the ‘– Virus’ model solution. However, this scenario could not create sufficient
DOC to support bacterial production levels of 4.96 mg C m
−3
d
−1
, and a bacterial production
level of 4.50 mg C m
−3
d
−1
was designated instead for this model scenario, which resulted in the
slightly lower GPP value. Solely removing viral lysis of the prokaryotic compartments without
designating the bacterial production value decreased the bacterial production by 10% but
impacted GPP flows negligibly (‘– Virus’, Scenario 3; BP = 1.97 and GPP = 29.2 mg C m
−3
d
−1
;
Table A8).
! 150!
The ratio of bacterial production to net primary production (BP:NPP) was different
between our empirical measurements and the estimates made by the model scenarios. The mean
BP:NPP ratio measured at SPOT was 0.53 (Table 4.1). BP:NPP ratios of the model solutions
were lower in all scenarios tested, with values ranging from 0.10 to 0.21 (Table A8). BP:NPP
ratios were greater in the model scenarios with a higher imposed bacterial production value (0.21
in ‘+ BP’ scenario, 0.19 in ‘+ BP – Virus’ scenario) than in the model scenarios in which the
bacterial production value was chosen statistically by the model (0.11 in ‘Average’ scenario,
0.10 in ‘– Virus’ scenario).
The identity of the top five model flows was unchanged between model scenarios, but the
subsequent largest fluxes varied among scenarios (Fig. 4.2B). Processes related to heterotrophic
nanoplankton grazing and the creation of detritus/DOC were prominent fluxes in the ‘Average’
scenario, as well as both scenarios without viral lysis (Fig. 4.2B). Use of the higher bacterial
production value resulted in greater importance of flows related to DOC production for model
scenarios with and without viral lysis (‘+ BP’ & ‘+ BP – Virus’). Specifically, flows related to
the viral lysis of Synechococcus and heterotrophic bacteria became dominant flows in the
scenario with viruses (‘+ BP’), while the dissolution of detritus became a dominant flow in the
scenario without viruses (‘+ BP – Virus’).
The relative contribution of viral lysis and protistan grazing to prokaryotic cell mortality
varied greatly with model scenario (Fig. 4.3). In the ‘Average’ scenario, viral lysis was
responsible for 23% of Prochlorococcus mortality, 26% of Synechococcus mortality, and 32% of
bacterial mortality. Protistan grazing was responsible for 77% of Prochlorococcus mortality,
73% of Synechococcus mortality, and 68% of bacterial mortality in the same scenario, with
heterotrophic nanoplankton removing the greatest amount of prokaryotic production among
! 151!
Figure 4.3. Percentage of prokaryotic mortality attributable to viral lysis (virus), grazing
(heterotrophic nanoplankton (HNAN), dinoflagellates (DINO), ciliates (CIL)), and
programmed cell death/necrosis (death) for the four model scenarios (columns). The top
row is mortality of Prochlorococcus, the middle row is mortality of Synechococcus, and
the bottom row is mortality of heterotrophic bacteria. The four model scenarios include
(1) ‘Average’: with viral lysis and using SPOT measurement error, (2) ‘+ BP’: with viral
lysis and setting the bacterial production value to empirical measurement mean (3) ‘−
Virus’: without viral lysis and using SPOT measurement error, and (4) ‘+ BP – Virus’:
without viral lysis and setting the bacterial production value to empirical measurement
mean.

! 152!
protistan grazers (74-85% of grazing by HNAN). Imposing a higher bacterial production and
allowing for viral lysis substantially shifted the relative importance of these mortality processes
(‘+ BP’), with viral lysis removing 91% of Prochlorococcus, 98% of Synechococcus, and 98% of
bacterial production. This greater contribution of viral lysis to picoplankton mortality
corresponded to viral production of 12.0 mg C m
−3
d
−1
of DOC, in comparison to the 2.30 mg C
m
−3
d
−1
of DOC produced by viruses in the ‘Average’ scenario. Grazers in the ‘+ BP’ scenario
removed a much smaller portion of picoplankton production (9% of Prochlorococcus, 2% of
both Synechococcus and bacteria), with nanoplankton continuing to remove the greatest
proportion of prey among the grazers (49-59% of grazing by HNAN). Removing viral lysis,
understandably, caused protistan grazing to become the principal mortality agent for the
picoplanktonic prey. Heterotrophic nanoplankton were the dominant grazers in both scenarios
without viral lysis, with their relative importance becoming greater when a higher bacterial
production value was used (78-87% of grazing by HNAN in ‘– Virus’ scenario, compared to 89-
96% of grazing in ‘+ BP – Virus’ scenario). In addition to mortality caused by viruses and
grazers, a small percentage of cells died due to programmed cell death/necrosis and directly
became detritus (Fig. 4.1, Table 4.2). This value was tightly constrained for picoplankton (≤1%
of production) and did not change appreciably among model scenarios. Total phytoplankton
mortality (cyanobacteria + eukaryotes) was divided between 8% viral lysis, 90% zooplankton
grazing, and 2% programmed cell death/necrosis in the ‘Average’ model scenario.
The heterotrophic nanoplankton carbon consumption rate and dietary composition also
varied greatly between model scenarios (Fig. 4.4). Heterotrophic nanoplankton consumed 12.3
mg C m
−3
d
−1
of prey carbon in the ‘Average’ scenario, with their diet consisting of 9%
heterotrophic bacteria, 7% Prochlorococcus, 26% Synechococcus, 32% phototrophic
! 153!

Figure 4.4. Diet composition for the four protistan grazer compartments in the four model
scenarios (columns). (a) Carbon consumption (mg C m
−3
d
−1
) of each prey compartment,
and (b) relative contribution of each prey compartment to total carbon consumption. The
four model scenarios include (1) ‘Average’: with viral lysis and using SPOT measurement
error, (2) ‘+ BP’: with viral lysis and setting the bacterial production value to empirical
measurement mean (3) ‘− Virus’: without viral lysis and using SPOT measurement error,
and (4) ‘+ BP – Virus’: without viral lysis and setting the bacterial production value to
empirical measurement mean. Abbreviations include: heterotrophic bacteria (BACT),
Prochlorococcus (PRO), Synechococcus (SYN), phototrophic picoeukaryotes (PEUK),
phototrophic nanoplankton (PNAN), diatoms (DIAT), heterotrophic nanoplankton (HNAN),
dinoflagellates (DINO), ciliates (CIL), mesozooplankton (MESO), small detritus (SDET),
and large detritus (LDET).

! 154!
picoeukaryotes, and 26% small detritus. Employing a higher bacterial production value and
allowing viral lysis (‘+ BP’) resulted in a lower nanoplankton carbon consumption rate (6.40 mg
C m
−3
d
−1
; 48% reduction from ‘Average’) and significantly altered the nanoplankton diet, with
96% of the ingested prey carbon derived from phototrophic picoeukaryotic carbon.
Nanoplankton carbon consumption rates were higher in both scenarios without viral lysis, with
14.3 mg C m
−3
d
−1
ingested by nanoplankton in the ‘– Virus’ scenario and 17.4 mg C m
−3
d
−1
ingested in the ‘+ BP – Virus’ scenario. Removing viruses without imposing the higher bacterial
production value did not meaningfully change the nanoplankton diet (‘– Virus’). However,
imposing a higher bacterial production value and removing viral lysis caused the nanoplankton
diet to consist of 25% heterotrophic bacteria, 9% Prochlorococcus, 31% Synechococcus, 35%
phototrophic picoeukaryotes, and 1% small detritus. Thus, this model scenario was characterized
by less consumption of small detritus and more consumption of heterotrophic bacteria by the
heterotrophic nanoplankton.
Dinoflagellate carbon consumption rate and diet composition were similar with or without
viral lysis (ingestion of 3.69 mg C m
−3
d
−1
and 3.72 mg C m
−3
d
−1
for ‘Average’ and ‘– Virus’
scenarios, respectively), but were impacted by the use of higher bacterial production value (Fig.
4.4). In the ‘Average’ scenario, the dinoflagellate diet consisted of 27% picoplankton (6%
heterotrophic bacteria, 4% Prochlorococcus, 8% Synechococcus, 9% phototrophic
picoeukaryotes), 43% nanoplankton (14% phototrophic, 29% heterotrophic), 16% diatom, and
14% detritus. Employing a higher bacterial production value and retaining viral lysis in the
model structure (‘+ BP’) resulted in a greater dinoflagellate carbon consumption rate (4.12 mg C
m
−3
d
−1
; 11% greater than the ‘Average’ scenario) and altered the dinoflagellate diet, causing less
consumption of heterotrophic nanoplankton (from 29% of diet to 4% of diet) and small detritus
! 155!
(from 15% of diet to 1% of diet) and greater consumption of diatom carbon (from 16% of diet to
76% of diet). Dinoflagellate carbon consumption rate was also higher when a higher bacterial
production value was used in the absence of viruses (4.05 mg C m
−3
d
−1
; ‘+ BP – Virus’), but in
this model scenario dinoflagellates consumed a greater proportion of nanoplankton than in the ‘+
BP’ scenario (63% total, 18% phototrophic, 45% heterotrophic); the remaining 37% of the
carbon ingested by dinoflagellates in this model scenario was comprised of 23% diatom, 12%
picoplankton, and 2% small detrital carbon.
Ciliates consumed slightly more carbon in the model scenarios with a higher imposed
bacterial production value, though the changes were minor (~8% greater; Fig. 4.4). Ciliates
ingested 2.64 mg C m
−3
d
−1
in the SPOT ‘Average’ model scenario, 2.87 mg C m
−3
d
−1
when a
higher bacterial production value was imposed (‘+ BP’), 2.64 mg C m
−3
d
−1
when viral lysis was
removed (‘– Virus’), and 2.85 mg C m
−3
d
−1
when a higher bacterial production value was used
and viral lysis was removed (‘+ BP – Virus’; Fig. 4.4). The ciliate diet consisted of 33%
picoplankton (7% heterotrophic bacteria, 5% Prochlorococcus, 10% Synechococcus, 11%
phototrophic picoeukaryotes), 52% nanoplankton (19% phototrophic, 33% heterotrophic), and
15% small detritus in the ‘Average’ model scenario. Solely removing viral lysis did not impact
the composition of the ciliate diet (‘– Virus’; Fig. 4.4). Using a higher bacterial production value
resulted in greater ingestion of nanoplankton by ciliates, regardless of whether viral lysis flows
were retained in the model structure (diet consisting of 92% nanoplankton and 82%
nanoplankton in the ‘+ BP’ and the ‘+ BP – Virus’ scenarios, respectively). Consequently,
ciliates in these model scenarios consumed a lower percentage of picoplankton (5% ‘+ BP’, 13%
‘+ BP – Virus’) and small detritus (1% ‘+ BP’, 2% ‘+ BP – Virus’) in their diet.
! 156!
Mesozooplankton consumed 9.46 mg C m
−3
d
−1
in our SPOT ‘Average’ model scenario,
with a diet composed of 18% nanoplankton (6% phototrophic, 12% heterotrophic), 59% diatoms,
16% microzooplankton (9% dinoflagellates, 7% ciliates), and 7% large detritus (Fig. 4.4).
Imposing a higher bacterial production value and allowing for viral lysis (‘+ BP’ scenario)
resulted in a mesozooplankton carbon consumption rate that was approximately half the rate in
the ‘Average’ scenario (5.03 mg C m
−3
d
−1
). In this same scenario, mesozooplankton shifted so
that more diatom (61%) and microzooplankton carbon (34%) were consumed, and less
nanoplankton (5%) and large detritus (<1%). Removing viruses without imposing a higher
bacterial production value resulted in only a minor change in mesozooplankton carbon
consumption rate (9.62 mg C m
−3
d
−1
, ‘– Virus’) and did not impact diet composition. Using a
higher bacterial production value and removing viral lysis from the model structure resulted in a
less carbon consumption than the ‘– Virus’ scenario, but more carbon consumption than the ‘+
BP’ scenario (8.57 mg C m
−3
d
−1
consumed in ‘+ BP – Virus’). The mesozooplankton diet in this
scenario contained 62% diatoms, 20% microzooplankton, 18% nanoplankton, and 1% large
detritus.
While the direct contribution of the various prey compartments to mesozooplankton
ingestion varied with model scenario (Fig. 4.4), the ultimate source of that carbon was similar in
three of the four scenarios (Fig. A6). The source of the mesozooplankton carbon consisted of 3%
Prochlorococcus, 10% Synechococcus, 11% phototrophic picoeukaryote, 10% phototrophic
nanoplankton, and 66% diatom carbon in the ‘Average’ model scenario. These values changed
by <2% in both model scenarios without viruses (‘– Virus’ & ‘+ BP – Virus’). However,
employing the higher bacterial production value while permitting viral lysis resulted in a much
smaller contribution of cyanobacterial carbon to higher trophic levels (~2%) and a greater role
! 157!
for diatoms as a source of mesozooplankton carbon (77%). In all model scenarios, diatoms
contributed twice as much to mesozooplankton carbon (66-77%) as they contributed to total GPP
(34%, see above).
In the SPOT ‘Average’ model scenario, a total of 15.5 mg DOC m
−3
d
−1
was produced in
the system, 94% of which fueled bacterial processes (bacterial production and respiration) and
6% of which exited the model due to downward mixing (Fig. 4.5). The three compartments that
produced the most DOC were the heterotrophic nanoplankton (3.02 mg DOC m
−3
d
−1
),
Synechococcus (2.42 mg DOC m
−3
d
−1
; produced through exudation and lysis), and the
mesozooplankton (2.26 mg DOC m
−3
d
−1
). Overall, grazer excretion produced the largest
proportion of the DOC in this model scenario (44.3%), followed by phytoplankton exudation
(26.1%), viral lysis (15.1%), and solubilization of detritus (14.5%).
Imposing a higher bacterial production value and allowing for viral lysis caused a total of
33.1 mg DOC m
−3
d
−1
to be produced in the system, >99% of which fueled bacterial processes
and <0.1% of which exited the model (‘+ BP’; Fig. 4.5). The higher amount of DOC production
in this scenario (114% greater than the ‘Average’ scenario) was disproportionate to the increase
in GPP incurred in the same system (54% greater than the ‘Average’ scenario) (Fig. 4.2B; Table
A8). The three compartments that produced the most DOC were Synechococcus (7.91 mg DOC
m
−3
d
−1
; produced through exudation and lysis), small detritus (4.93 mg DOC m
−3
d
−1
), and the
heterotrophic bacteria (4.84 mg DOC m
−3
d
−1
; produced through lysis). In this model scenario,
viral lysis produced the greatest amount of DOC (36.3%), followed by phytoplankton exudation
(27.95), grazer excretion (18.1%), and detrital solubilization (17.7%).
Removing flows related to viral lysis resulted in the creation of 14.1 mg DOC m
−3
d
−1
, 93%
of which supported bacterial processes and 7% of which exited the model (‘– Virus’; Fig. 4.5).
! 158!

Figure 4.5. Production of dissolved organic carbon (DOC) by each model compartment in
the four model scenarios (columns). The four model scenarios include (1) ‘Average’: with
viral lysis and using SPOT measurement error, (2) ‘+ BP’: with viral lysis and setting the
bacterial production value to empirical measurement mean (3) ‘− Virus’: without viral
lysis and using SPOT measurement error, and (4) ‘+ BP – Virus’: without viral lysis and
setting the bacterial production value to empirical measurement mean. Bolded values at
bottom of each pie chart indicate the total amount of DOC produced, with the percentage
of the total DOC produced by grazing processes (Grazers), viral lysis (Viruses),
phytoplankton exudation (Phyto), and solubilization of detritus (Detritus) detailed below.
Black stars within each pie chart indicate the three largest producers of DOC in each
model scenario. Abbreviations include: heterotrophic bacteria (BACT), Prochlorococcus
(PRO), Synechococcus (SYN), phototrophic picoeukaryotes (PEUK), phototrophic
nanoplankton (PNAN), diatoms (DIAT), heterotrophic nanoplankton (HNAN),
dinoflagellates (DINO), ciliates (CIL), mesozooplankton (MESO), small detritus (SDET),
and large detritus (LDET).

! 159!
The total amount of DOC produced in this model scenario was similar to that in the ‘Average’
scenario, but the three compartments responsible for the greatest DOC production were the
heterotrophic nanoplankton (3.58 mg DOC m
−3
d
−1
), the mesozooplankton (2.27 mg DOC m
−3

d
−1
), and small detritus (1.96 mg DOC m
−3
d
−1
). Grazer exudation produced the majority of DOC
in this scenario (52.6%), followed by phytoplankton exudation (27.1%), and detrital
solubilization (20.3%). Thus, removing viruses did not significantly impact the total production
of DOC (9% decrease from the ‘Average’ scenario), but changed the major contributors to the
pool of DOC.
Using a higher bacterial production value and removing viral lysis yielded the production
of 30.0 mg DOC m
−3
d
−1
, >99% of which fueled bacterial processes and <0.1% of which exited
the system due to mixing (‘+ BP – Virus’; Fig. 4.5). Slightly less DOC was produced in this
model scenario than in the scenario with viruses (10% decrease from ‘+ BP’ model scenario).
This was due to model constraints, which only enabled us to impose a bacterial production value
of 4.50 mg C m
−3
d
−1
(compared to bacterial production values of 4.96 mg C m
−3
d
−1
in the ‘+
BP’ scenario)(Table A8). The three compartments that produced the most DOC in this model
scenario were small detritus (7.80 mg DOC m
−3
d
−1
), the heterotrophic nanoplankton (5.64 mg
DOC m
−3
d
−1
), and the mesozooplankton (2.78 mg DOC m
−3
d
−1
). Grazer exudation,
phytoplankton excretion, and detrital solubilization produced approximately equal amounts of
DOC in this model scenario (35.4%, 30.2%, and 34.4%, respectively).
While the direct contributions of the various compartments to DOC (and ultimately
bacterial processes) were distinct among model scenarios (Fig. 4.5), the ultimate source of that
carbon did not fluctuate among model scenarios (Fig. A6). The relative contribution (direct and
indirect) of each phytoplankton compartment to bacterial processes was comprised of 9%
! 160!

Figure 4.6. Vertical flux of particulate carbon out of the surface ocean in the four model
scenarios. (a) The mean carbon flux of diatoms (DIAT) and large detritus (LDET) out of
the surface ocean (mg C m
−3
d
−1
) . (b) The percentage of gross primary production (GPP)
that exits the model as vertical flux. The four model scenarios include (1) ‘Average’: with
viral lysis and using SPOT measurement error, (2) ‘+ BP’: with viral lysis and setting the
bacterial production value to empirical measurement mean (3) ‘− Virus’: without viral
lysis and using SPOT measurement error, and (4) ‘+ BP – Virus’: without viral lysis and
setting the bacterial production value to empirical measurement mean.

! 161!
Prochlorococcus, 32% Synechococcus, 25% phototrophic picoeukaryote, 7% phototrophic
nanoplankton, and 26% diatom carbon in the ‘Average’ model scenario. The relative
contribution of each phytoplankton group to bacterial production varied negligibly (<5%)
between model scenarios.
Imposing a higher bacterial production value in the model scenarios consistently resulted in
less vertical flux of carbon out of the system (Fig. 4.6). In the ‘Average’ model scenario, 1.94 mg
C m
−3
d
−1
exited the model through sedimentation (Fig. 4.6a), which corresponds to 6.5% of the
GPP in the system (Fig. 4.6b). Removing viruses (‘– Virus’) did not significantly impact the
vertical transport of carbon (2.02 mg C m
−3
d
−1
; 6.9% of GPP). Using a higher bacterial
production value resulted in less carbon export regardless of whether viral lysis was included (‘+
BP’, 1.25 mg C m
−3
d
−1
; 2.7% of GPP) or excluded (“+ BP – Virus”, 1.26 mg C m
−3
d
−1
; 2.8% of
GPP) from the model structure.

4.4 Discussion
4.4.1!Assessing!the!fate!of!primary!production!at!the!SPOT!station!
!
One of the main objectives of this study was to use linear inverse modeling to enhance our
understanding of the fate of primary production at the SPOT station. The ‘Average’ model
scenario provides the most statistically likely assimilation of the food web structure at the SPOT
station given our present state of understanding of microbial food web topology and
measurements of microbial standing stocks and rate processes. In this model scenario,
phytoplankton groups exuded ~14% of their gross primary production and created 26% of the
dissolved organic carbon that was utilized by the heterotrophic bacterial assemblage (Fig. 4.5).
These results are in excellent agreement with a theoretical model of microbial food web
! 162!
dynamics in the oligotrophic ocean that found phytoplankton exudation is responsible for 25% of
DOC production (Nagata 2000). Phytoplankton also respired ~21% of their gross primary
production, which resulted in phytoplankton net primary production being 65% of their gross
primary production.
Protistan grazing caused approximately three-fourths (73-77%) of the total
picocyanobacterial mortality (Fig. 4.3). Overall, zooplankton grazing caused 90% of the total
phytoplankton mortality, with viral lysis (8%) and programmed cell death/necrosis (2%)
responsible for the remainder. These model results are in good accordance with estimates that
viral lysis is responsible for ~10-40% of bacterial and cyanobacterial mortality in the marine
environment (Fuhrman 2000, Suttle 2000), and generally accounts for <10% of total
phytoplankton mortality during non-bloom conditions (Kimmance et al. 2007, Brum et al. 2014).
Viruses created 15% of the total DOC production the ‘Average’ model scenario, while protistan
grazing and subsequent detritus formation and solubilization created 44% and 15% of the DOC,
respectively (Fig. 4.5). DOC production by the various mortality sources was partitioned in a
similar fashion to the model of Nagata (2000), who reported that viruses contributed 10% and
grazers (and their feces) contributed 65% of the total DOC production.
A substantial percentage of daily net primary production was consumed by protistan (57%)
and mesozooplankton (31%) grazers. Microzooplankton are believed to consume approximately
60-70% of daily net primary production (Calbet & Landry 2004) and mesozooplankton are
believed to remove another 12% on average (Calbet 2001). Thus, model results are in good
accordance with estimates microzooplankton grazing impacts on primary production, but suggest
that mesozooplankton grazing at the SPOT station is higher than the mean value for the marine
environment. A comparison of phytoplankton growth rates at the SPOT station in the presence
! 163!
and absence of mesozooplankton suggested that mesozooplankton grazing impact is relatively
small at study site (Connell et al. 2017); however, it is possible that the sampling method in this
study did not adequately sample the mesozooplankton population and therefore underestimated
mesozooplankton grazing impact. Future characterization of the mesozooplankton population
and their grazing rates on phytoplankton at the SPOT station will enable further assessment of
these model results.
A small percentage of the gross primary production in the system exited the surface ocean
as sinking organic matter (7%, Fig. 4.6). This value is in good agreement with current estimates
that approximately 10% of gross primary production sinks out of the euphotic zone (De La
Rocha & Passow 2014). However the model may underestimate sinking flux if we compare this
value to the mean percentage of measured gross primary production in the mixed layer that
contributes to carbon flux based on sediment-trap data and
234
Th:
238
U dis-equilibrium data at
the SPOT station (12%) (Haskell et al. 2016, Haskell et al. 2017).
!
4.4.2!Model!BP:NPP!ratios!reveal!that!empirical!measurements!may!overestimate!bacterial!production!
and/or!underestimate!net!primary!production!!
!
Overall, the partitioning of carbon in the ‘Average’ model scenario is in good accordance
with literature reported values and provides a reasonable reconstruction of microbial food web
dynamics at the SPOT station. However, the absolute values of these flows in the model output
differ from the mean values of the empirical measurements conducted at the study site (Table
A8). The most noticeable difference between the empirical measurements and the model output
was the ratio of bacterial production to net primary production (BP:NPP) at the SPOT station.
The empirical mean BP:NPP ratio at the SPOT station was 0.53 (Table 4.1), while the model
! 164!
results indicated that much lower BP:NPP ratios can be supported at the study site (0.10-0.21;
Table A8). In the ‘Average’ model scenario, the model-selected bacterial production value was
2.3x lower than the measured average (2.17 vs. 4.96 mg C m
−3
d
−1
) and the model-selected net
primary production value was 2.1x higher than the measured average (19.6 vs. 9.35 mg C m
−3

d
−1
) (Table A8). The modeled BP:NPP ratios corroborate the findings of Ducklow (2000), who
synthesized several studies with contemporaneous measurements of bacterial production and
primary production and reported that BP:NPP ratios rarely deviate from 0.10-0.20 in the marine
environment. High BP:NPP values (>0.5) are possible in nature but require non-steady state
conditions to provide sufficient DOC to support the heterotrophic bacterial assemblage. For
example, Ducklow et al. (1993) observed BP:NPP ratios of 0.15-0.80 in response to the spring
phytoplankton bloom in the eastern North Atlantic Ocean, where an increase in bacterial
production lagged primary production by 3-4 days. However, our model construction assumes
steady-state at the SPOT station over long time-scales due to the subtropical nature of the region
and the low variability in microbial standing stocks and rate processes throughout the annual
cycle (Hamersley et al. 2011, Caron et al. 2017, Connell et al. 2017). If our assumption of
steady-state is valid for the SPOT station at these time-scales, our model results reveal that our
empirical measurements may have overestimated bacterial production and/or underestimated net
primary production at the SPOT station.
Three major model assumptions may affect our conclusion that empirical measurements
conducted at the SPOT station have overestimated bacterial production and/or underestimated
net primary production: (1) use of a mean of 15% as the fixed bacterial gross growth efficiency
(BGE) at the SPOT station, (2) the availability of all DOC produced by viral lysis to the
heterotrophic bacterial assemblage, and (3) the lack of an external source of DOC in the model.
! 165!
If BGE at the SPOT station was greater than 15%, then more DOC would be assimilated into
bacterial biomass, resulting in greater bacterial production per unit of primary production (i.e.
higher BP:NPP). A synthesis of 329 direct measurements of BGE in marine ecosystems revealed
that BGE values generally fall between 5-30%, with a mean BGE of 23% for all marine data (del
Giorgio & Cole 2000). BGE’s calculated for the SPOT station using the algorithm from the del
Giorgio and Cole (1998) meta-analysis ranged from 2-35%, with a mean value of 12% (Teel
2017). The BGE employed in the model (15%) was slightly greater than the mean BGE
calculated for the SPOT station by Teel (2017) and therefore it is unlikely that BGE is an
explanation for the observed difference between measured and modeled BP:NPP ratios.
The model assumption that all DOC produced by viral lysis is available to the heterotrophic
bacteria may also impact the assessment of BP:NPP ratios at the SPOT station. Model
construction assumes that viral lysis efficiently converts 100% of the host carbon to bioavailable
DOC, which overestimates the carbon availability in the ecosystem. This idealized representation
of viral lysis creates more DOC for the heterotrophic bacterial assemblage in the model
ecosystem than would be created in situ by viral lysis. In reality, viral lysis results in the release
of viral particles, cellular components (i.e. cell wall fragments), and a suite of dissolved and
particulate organic matter (Fuhrman 2000). In a culture study of marine bacterium and one of its’
associated phages, the identifiable composition of the lysate was comprised of 51-86% dissolved
combined amino acids, 2-3% dissolved free amino acids, 2-3% glucosamine, and 4-6% new viral
particles (Middelboe & Jørgensen 2006). Cell wall components were ~12% of the total identified
organic compounds, which may be fairly recalcitrant to degradation (Fuhrman 2000). Evidence
from several culture studies suggests that a substantial fraction of the organic matter made
available by viral lysis is eventually available to the heterotrophic bacterial assemblage (Bratbak
! 166!
et al. 1998, Middelboe et al. 2003, Middelboe & Jørgensen 2006), with labile compounds being
processed at more rapid time-scales and refractory compounds having longer turnover times in
the marine environment. Nonetheless, production of viral progeny will reduce the total amount
of DOC produced by viral lysis and made available to the heterotrophic bacterial assemblage,
whether these viruses proceed to infect new hosts or are removed by grazing of viruses or viral-
attached particles (Suttle & Chen 1992, González & Suttle 1993). Additionally, viruses utilize
the host’s cellular machinery and resources to produce viral progeny and thereby may decrease
host gross growth efficiency (Middelboe & Lyck 2002). The idealized portrayal of viral lysis in
the model structure caused the model to slightly overestimate, not underestimate, the BP:NPP
ratio; therefore, this assumption does not explain the differences between modeled and measured
BP:NPP ratios.
The third assumption that could impact the conclusion that the empirical measurements at
the SPOT station may overestimate bacterial production and/or underestimate net primary
production is that there was not an external source of DOC at steady-state conditions in our
model. Possible external sources of DOC include atmospheric deposition or advection from a
different location or depth. However, basic mass-balance calculations reveal that ~60% of the
DOC in the system would have to come from an external source to support a BP:NPP ratio of
0.53 at the SPOT station (see Supplemental Text in Appendix B). The magnitude of such a
contribution is highly unlikely. Based on an examination of these three possible factors affecting
bacterial production, we conclude that empirical measurements at the SPOT station
overestimated bacterial production and/or underestimated net primary production.
Measurements of microbial rate processes are notoriously problematic, and have high
levels of uncertainty due to methodological limitations and the assumptions made to calculate
! 167!
these values. For example, conversion factors used to calculate bacterial production vary
approximately three-fold (Robinson 2008) and cause these measurements to be uncertain by at
least a factor of two (Ducklow 2000). This uncertainty can be demonstrated by comparing the
mean bacterial production value calculated from
3
H-thymidine (4.96 mg C m
−3
d
−1
; Table 4.1)
and
3
H-leucine (6.03 mg C m
−3
d
−1
; data not shown) at the SPOT station. In addition to
differences due to choice of radiolabelled substrate or conversion factor, photoheterotrophic
organisms have been shown to artificially inflate measurements of heterotrophic bacterial
production using
3
H-leucine (Church et al. 2004, Björkman et al. 2015, Viviani & Church 2017).
Prochlorococcus is a photoheterotroph capable of assimilating a variety of organic compounds
including amino acids (Zubkov et al. 2003, Michelou et al. 2007), glucose (Muñoz-Marín et al.
2013), and DMSP (Vila-Costa et al. 2006) and has been shown to have higher cell-specific rates
of leucine incorporation than heterotrophic bacteria (Talarmin et al. 2011, Björkman et al. 2015).
Synechococcus also displays photoheterotrophic capabilities (Montesinos et al. 1997, Zubkov et
al. 2003). Cyanobacteria were 26% of the total prokaryotic standing stock at the SPOT station
(Table 1), so it is possible that photoheterotrophy by this component resulted in overestimation
of bacterial production at the study site. Cyanobacterial affinity for thymidine may be lower
(Fuhrman & Azam 1982), though this possibility has not been thoroughly investigated.
Alternatively, our calculations of net primary production at the SPOT station based on
dilution data may underestimate the true values due to methodological limitations and/or
mathematical assumptions.
14
C incubations are the gold standard for measuring net primary
production in the ocean, but dilution measurements have been shown to provide comparable
estimates of net primary production (Selph et al. 2015). Dilution-based methods can lead to
underestimation of growth rates due to manipulations of community interactions (Schmoker et
! 168!
al. 2013, Brum et al. 2014, Stoecker et al. 2015). Additionally, the carbon conversion factors
used to calculate phytoplankton standing stocks in this study were conservative (see discussion
in Caron et al. (2017)), and yielded somewhat lower estimates of total phytoplankton standing
stock (mean = 23 mg C m
−3
d
−1
; Table 4.1) than a comparable study of phytoplankton standing
stock at CALCOFI stations in the southern, inshore region of the Southern California Bight
(mean = 32 mg C m
−3
d
−1
) (Taylor et al. 2015). Both of these factors may contribute to an
underestimation of net primary production at the SPOT station. However, separate calculations
of net primary production at the SPOT station using observed chlorophyll concentrations and
satellite PAR and a regionally-tuned algorithm for the Southern California Bight (Jacox et al.
2015) showed that net primary production at SPOT varies from approximately 5-15 mg C m
−3

d
−1
in the surface waters (Teel 2017). Thus, the mean net primary production value of 9.35 mg C
m
−3
d
−1
calculated from dilution data is in good agreement with the satellite estimates.
The ratio between the measured net primary production and gross primary production
(NPP:GPP) was 0.16 at our study site (Table A8). Gross primary production values calculated
from H
2
18
O labeling experiments have been shown to be ~2x concurrent
14
C net primary
production measurements (Bender et al. 1999), suggesting that our measured ratio was low. The
empirically measured ratio of Bender et al. (1999) is much closer to the modeled ratio in the
‘Average’ model scenario (0.66, Table A8). As with other microbial rate measurements, gross
primary production values computed from triple oxygen isotopes and O
2
:Ar ratios require
estimation of several parameters (i.e. upwelling velocities, piston velocities, and vertical eddy
diffusivity) which can introduce large uncertainty into the calculated values (Haskell et al. 2016,
Haskell et al. 2017). Future characterization of bacterial production and net primary production
! 169!
rates at the SPOT station and examination of potential sources of error are needed to investigate
the seemingly unsustainable BP:NPP ratios measured at the study site.

4.4.3 Employing the empirically-measured bacterial production value resulted in less trophic
transfer of carbon and less sinking of organic matter

The second objective of this study was to assess the role that heterotrophic bacteria play
in carbon flow at the SPOT station. Several studies have demonstrated that a significant fraction
of the photosynthetically-fixed carbon is utilized by the heterotrophic bacteria to support cell
metabolism and growth (Azam et al. 1983, Ducklow 2000, Robinson 2008). The heterotrophic
bacteria were the largest planktonic assemblage at the SPOT station (Table 4.1; Caron et. al
2017), with fluxes related to bacterial processes (biomass production and respiration) dominating
movement of carbon in our SPOT ecosystem model (Fig. 4.2). However, the mean BP:NPP ratio
measured at the SPOT station (0.53) was considerably larger than the ratio predicted by the
model in the ‘Average’ scenario (0.11, Table S1; see section 4.4.3). Therefore, we tested how the
carbon transformations in the model changed when we imposed the empirically-measured, mean
bacterial production value on the model solution (‘+ BP’). The comparison of the ‘Average’ and
‘+ BP’ model scenarios not only respresents a shift in the absolute value of bacterial production
in the model system (2.17 to 4.96 mg C m
−3
d
−1
), but also a shift the BP:NPP ratio (0.21) and
thus the relative amount of photosynthetically-fixed carbon which is utilized to support bacterial
processes (Table A8).
A total phytoplankton net primary production rate of 23.3 mg C m
−3
d
−1
was required to
support a bacterial production value of 4.96 mg C m
−3
d
−1
in the ‘+ BP’ model scenario. This net
primary production rate corresponds to phytoplankton growth rates of 1.0 d
−1
, and thus was the
! 170!
maximum amount of primary production that could occur in the model system given the
maximum production constraints (max growth of 1.0 d
−1
; Table 4.2). Median phytoplankton
growth rates in the marine environment are approximately 0.5-0.7 d
−1
(Schmoker et al. 2013,
Kirchman 2016), while the median phytoplankton growth rate at the SPOT station was 0.27 d
−1

(Connell et al. 2017). Thus, phytoplankton growth rates would have to be substantially higher
than these median values to support a bacterial production value of 4.96 mg C m
−3
d
−1
at the
study site, assuming the carbon conversions used to estimate phytoplankton biomass at the study
site were reasonable. However, using the empirically measured bacterial production value
resulted in a gross primary production value that was much closer to the measured mean than did
the ‘Average’ scenario (Table A8).
Employing a higher bacterial production value in the model diverted carbon away from
higher trophic levels. The higher BP:NPP ratio in this model scenario was supported by lysing
the majority of bacterial and cyanobacterial production (91-98% of mortality caused by viruses;
Fig. 4.3), as well as by microbial-mediated dissolution of small detritus (Fig. 4.5). Consequently,
less picoplankton prey and detritus were available for consumption by higher trophic levels (Fig.
4.4). The largest impacts were observed for the carbon consumption rate and diet composition of
the heterotrophic nanoplankton. Higher levels of lysis in the ‘+ BP’ scenario resulted in 48% less
carbon consumption by nanoplankton and a shift in their diet to almost solely phototrophic
picoeukaryote prey (Fig. 4.4). The complete shift of the nanoplankton diet to picoeukaryotic prey
in the ‘+ BP’ scenario is probably exacerbated by model structure, which only allows for the
lysis of the prokaryotic assemblages (bacteria & cyanobacteria). In reality, viral lysis of
eukaryotic phytoplankton may also be an important mortality pathway (Bratbak et al. 1990,
Baudoux et al. 2006). The carbon consumed by the heterotrophic nanoplankton in the ‘+ BP’
! 171!
scenario supported growth rates of 0.22 d
-1
. These values are near the bottom of what has been
reported in marine environments, with typical in situ nanoplankton growth rates ranging from 0.2
to ~1.4 d
-1
(Verity et al. 1993, Neuer & Cowles 1994, Karayanni et al. 2008).
Interestingly, using a higher bacterial production value resulted in slightly more carbon
consumption by the microzooplankton (~10% more than ‘Average’) but significantly less carbon
consumption by the mesozooplankton (47% less than ‘Average’) (Fig. 4.4). Dinoflagellates
consumed a greater amount of diatom carbon (from 16% of diet to 76% of diet), removing 43%
of the diatom daily net primary production and consequently reducing the availability of diatom
prey for the mesozooplankton. Ciliates had a similar impact on mesozooplankton prey
availability, consuming 71% of the daily nanoplankton production (phototrophic and
heterotrophic) and causing nanoplankton to be a very small fraction (5%) of the
mesozooplankton diet. Thus, the nutritional flexibility of the microzooplankton grazers enabled
them to maintain similar growth rates at the expense of the mesozooplankton. Protistan grazing
has been shown to be size selective (Jürgens & Massana 2008), and grazing relationships in food
web models are often structured to mimic the 10:1 predator-prey size ratio demonstrated by
Fenchel (1987) (i.e. nanoplankton consume picoplankton, microplankton consume nanoplankton,
etc.). However, the diverse feeding strategies of protistan grazers enable them to consume a wide
variety of prey—including prey larger then themselves (Hansen & Calado 1999) —which causes
them to function at several trophic levels in marine food webs (Caron et al. 2012). If we had
created a simplified microzooplankton component that could not consume diatoms, then
microzooplankton production would have decreased and macrozooplankton production would
have decreased less drastically (or perhaps been unaffected) in the ‘+ BP scenario’. Our results
demonstrate the incorporating prey and predator diversity is essential to modeling the carbon
! 172!
transformations in marine food webs as it allows for inter-guild competition that has
consequences for carbon flow.
Both the composition and amount of sinking organic matter was different when the
absolute bacterial production value and the BP:NPP ratio in the model were greater. Vertical flux
of carbon out of the euphotic zone was 37% lower in the ‘+ BP’ model scenario than in the
‘Average’ model scenario (Fig. 4.6), primarily due to lower levels of mesozooplankton
production and thus lower production of fecal pellets in the ‘+ BP’ scenario (Figs. 4 and 6). Our
calculations of indirect and direct fluxes from the phytoplankton to the mesozooplankton also
suggest that the composition of the fecal pellets was altered in the ‘+ BP’ scenario (Fig. A6).
Specifically, viral lysis of most of the picoplankton daily production resulted in greater
utilization and respiration of picocyanobacterial carbon by the heterotrophic bacterial
assemblage, preventing cyanobacteria from contributing to mesozooplankton through trophic
transfer. Thus, while phytoplankton are believed to contribute to vertical flux in proportion to
their contribution to total gross primary production (Stukel & Landry 2010), these proportions
may change at higher BP:NPP ratios.

4.4.4 Removing viral lysis resulted in less bacterial production and generally more trophic
transfer of carbon
The production of DOC was 10% greater upon the inclusion of viral lysis regardless of
whether the model used a statistically-selected value of bacterial production (33.1 mg DOC m
−3
d
−1
in the ‘Average’ scenario vs. 30.0 mg DOC m
−3
d
−1
in the ‘– Virus’ scenario) or whether the
empirically measured, higher bacterial production value was imposed on the system (15.5 mg
DOC m
−3
d
−1
in the ‘+ BP’ scenario vs. 14.1 mg DOC m
−3
d
−1
in the ‘+ BP – Virus’ scenario)
! 173!
(Fig. 4.5). This resulted in 10% more bacterial production when viruses were included in the
model structure (Fig. 4.2; Table A8), as the model structure permitted all of the lysate to become
DOC and subsequently fuel bacterial production. The 10% increase observed in our study is
lower than a previously predicted increase in bacterial production of 27-33% due to inclusion of
viral lysis (Fuhrman 1992, Fuhrman & Suttle 1993). However, our model only considers the
release of carbon by viral lysis and does not consider potential nutrient dynamics in the system
which may impact this estimate (Weitz et al. 2015).
Excluding viral lysis from the model structure resulted in slightly higher carbon
consumption rates by the zooplankton assemblages, but had little impact on their diet
composition (‘– Virus’ vs. ‘Average’ scenario). Nanoplankton consumed 16% more carbon in
the ‘– Virus’ scenario, effectively removing the majority of the production that was previously
removed by viral lysis in the ‘Average’ scenario (Figs. 4.3 & 4.4). Microzooplankton
consumption and mesozooplankton consumption did not change appreciably when viruses were
removed (1-2% higher; Fig. 4.4). Therefore, while nanoplankton consumed a greater amount of
carbon, much of that carbon was lost to metabolic processes rather than passed on to higher
trophic levels. A mass balance model by Fuhrman (1992) demonstrated that permitting viral lysis
to enact equal amount of bacterial mortality as grazers led to a 37% decrease in carbon transfer
to nanoplankton, a 25% decrease in carbon transfer to microzooplankton, and a 7% decrease in
carbon transfer to larger zooplankton. Thus, while both models demonstrated decreased trophic
transfer upon the inclusion of viral lysis, our model results estimate that viruses have a slightly
weaker impact on zooplankton carbon consumption rates than the Fuhrman (1992) model.
If we assumed that bacterial production values are higher at the SPOT station (i.e.
imposed the empirically measured bacterial production mean value on the model), we found
! 174!
significantly higher trophic transfer of carbon in the scenario without viral lysis (‘+ BP – Virus’)
than the scenario with viral lysis (‘+ BP’). Nanoplankton consumed 171% more prey carbon
when viruses were removed, and ate a relatively even amount of the various picoplankton prey
groups rather than only picoeukaryotic prey carbon (Fig. 4.4). Microzooplankton consumed
relatively equal amounts of carbon in both scenarios; however, removing viruses shifted the diets
to contain more nanoplankton and picoplankton carbon. Consequently, more diatom carbon was
available for consumption by mesozooplankton, causing the mesozooplankton to consume 70%
more carbon in the ‘+ BP – Virus’ scenario than the ‘+ BP’ scenario. Our results confirm
evidence that viral lysis reduces the transfer of carbon to higher trophic levels (Fuhrman 1992,
Fuhrman 1999, Weinbauer 2004, Weitz et al. 2015), and suggest that this is true regardless of the
BP:NPP ratio in the ecosystem.
A comparison of both model scenarios that excluded viral lysis revealed competition
between the heterotrophic bacteria and the grazer assemblages for detrital carbon. The
heterotrophic nanoplankton consumed 18% more carbon and the microzooplankton consumed
10% more carbon in the ‘+ BP – Virus’ scenario than in the ‘– Virus’ scenario because more
bacterial production was available to the grazer assemblage (Figs. 4.3 & 4.4). Yet,
mesozooplankton carbon consumption rates were 9.4% less when a higher bacterial production
value was utilized, despite the increase in secondary production of the protistan grazer
assemblages in this scenario (Fig. 4.4). These results are surprising, as more prey consumption
by protistan grazers should theoretically lead to more food availability for the mesozooplankton.
However, detritus was an important source of organic matter for the heterotrophic bacterial
assemblage in the ‘+ BP – Virus’ scenario (Fig. 4.5), with a significant amount of bacterial-
mediated dissolution of detritus needed to support the higher, imposed bacterial production
! 175!
value. The utilization of fecal pellets by the bacterial assemblage prohibited coprophagy by the
mesozooplankton and resulted in less carbon consumption in the ‘+ BP – Virus’ scenario than in
the ‘– Virus’ scenario (Fig. 4.4). Greater utilization of large detritus by the heterotrophic bacteria
also resulted in less vertical flux of carbon out of the ecosystem (Fig. 4.6). Bacterial utilization of
detrital carbon also increased in the model scenario with the imposed bacterial production value
and viruses (‘+ BP’, Fig. 4.4 & 4.6), but this had a much smaller impact on mesozooplankton
consumption than the viral shunt.
Thus, if our empirical measurements of bacterial production at the SPOT are accurate, these
results suggest that an increased demand for carbon by the heterotrophic bacterial assemblage
may result in complete utilization of large detritus in the surface ocean and not only impact the
vertical flux of detrital particles but also impact higher trophic levels. This finding is based on
the assumption that detrital particles are an appreciable part of the mesozooplankton diet.
Evidence from laboratory studies suggests that copepods are important consumers of fecal pellets
(Paffenhöfer & Knowles 1979, Paffenhöfer & Van Sant 1985). However, Poulsen and Kiørboe
(2006) detected low levels of mesozooplankton coprophagy in the North Sea during the summer,
suggesting that microorganisms <200 µm were primarily responsible for the recycling of fecal
pellets in the surface ocean. Copepods may modify fecal pellets, effectively breaking them into
smaller pieces (with slower sinking rates) that can be more effectively used by single-celled
organisms (Turner 2002). Yet, it is unclear how the rate of bacterial-mediated detrital dissolution
compares to the time-scales of fecal pellet sedimentation out of the upper ocean and thus how
feasible the complete utilization of the detrital pool by the heterotrophic bacterial assemblage is
in the ‘+ BP – Virus’ model scenario. This study highlights the need to investigate the relative
partitioning of fecal pellets and other detritus between mesozooplankton, microzooplankton, and
! 176!
the heterotrophic bacterial assemblage, as these relationships may impact the transfer of carbon
to higher trophic levels and/or the vertical flux of carbon when the bacterial demand for DOC is
high (‘+ BP – Virus’ scenario; Figs. 4.4 & 4.6).
The different model scenarios compared in this study highlight the importance of
accurately characterizing heterotrophic bacterial processes when assessing the fate of primary
production in marine ecosystems. The absolute value of bacterial production and primary
production in the SPOT model ecosystem (and thus the BP:NPP ratio) had profound
consequences for the utilization of organic matter with the surface ocean and the transfer of
carbon to higher trophic levels or out of the euphotic zone. In addition, quantifying the relative
importance of viral lysis at the SPOT station is an integral part to understanding ecosystem
function as the inclusion or exclusion of viruses from the model structure impacted the
partitioning of energy in the system and ultimately the fate of primary production.

4.5 Conclusion
This linear inverse modeling analysis provides us with novel insights as to the fate of
primary production at the SPOT station by integrating our knowledge of food web topology with
our empirical datasets to reveal the most likely distribution of carbon in the ecosystem. The
‘Average’ model scenario indicated that ~20% of cyanobacterial mortality and 8% of total
phytoplankton mortality was due to viral lysis; thus, ~80% of cyanobacterial mortality and 90%
of total phytoplankton mortality was due to zooplankton grazing (with another 2% due to cell
death). Ultimately, 7% of gross primary production exited the surface ocean as sinking organic
matter. Modeled BP:NPP ratios at SPOT station (0.1 to 0.2) implied that current empirical
measurements from our study site may overestimate bacterial production and/or underestimate
net primary production values (mean value = 0.53). Imposing the empirically-measured bacterial
! 177!
production value (and subsequently increasing the BP:NPP ratio in the model) resulted more
viral lysis of picoplankton and increased utilization of carbon by the heterotrophic bacteria,
therefore decreasing trophic transfer of carbon and sinking flux of organic matter out of the
surface ocean. Removing viral lysis slightly increased bacterial production (~10%) and generally
increased zooplankton production. However, imposing the higher, empirically measured
production value and excluding viral lysis resulted in substantial utilization of large detritus by
the bacterial assemblage that prohibited coprophagy by the mesozooplankton and recycled
zooplankton fecal pellets prior to their export out of the surface ocean. Thus, accurately
characterizing bacterial processes, as well as the specific biological processes that support or
remove them, is essential for proper assessment of the fate of primary production in marine food
webs. The integration of several datasets through an inverse ecosystem modeling approach
improves our understanding of planktonic trophic interactions at the SPOT station and provides
fodder for both modelers and empiricists seeking to understand the fate of carbon in coastal
oceans.

! 178!
Chapter Four Literature Cited
Azam!F,!Fenchel!T,!Field!J,!Gray!J,!Meyer?Reil!L,!Thingstad!F!(1983)!The!ecological!role!of!
water?column!microbes!in!the!sea.!Mar!Ecol!Prog!Ser!10:257?263!
Baudoux!A?C,!Noordeloos!AA,!Veldhuis!MJ,!Brussaard!CP!(2006)!Virally!induced!mortality!
of!Phaeocystis!globosa!during!two!spring!blooms!in!temperate!coastal!waters.!Aquat!
Microb!Ecol!44:207!
Bender!M,!Orchardo!J,!Dickson!M?L,!Barber!R,!Lindley!S!(1999)!In!vitro!O2!fluxes!compared!
with!14C!production!and!other!rate!terms!during!the!JGOFS!Equatorial!Pacific!
experiment.!Deep!Sea!Res!I!46:637?654!
Björkman!KM,!Church!MJ,!Doggett!JK,!Karl!DM!(2015)!Differential!assimilation!of!inorganic!
carbon!and!leucine!by!Prochlorococcus!in!the!oligotrophic!North!Pacific!Subtropical!
Gyre.!Front!Microbiol!6!
Bograd!SJ,!Checkley!DA,!Wooster!WS!(2003)!CalCOFI:!a!half!century!of!physical,!chemical,!
and!biological!research!in!the!California!Current!System.!Deep!Sea!Res!II!50:2349?
2353!
Bratbak!G,!Heldal!M,!Norland!S,!Thingstad!TF!(1990)!Viruses!as!partners!in!spring!bloom!
microbial!trophodynamics.!Appl!Environ!Microbiol!56:1400?1405!
Bratbak!G,!Jacobsen!A,!Heldal!M!(1998)!Viral!lysis!of!Phaeocystis!pouchetii!and!bacterial!
secondary!production.!Aquat!Microb!Ecol!16:11?16!
Breitbart!M,!Middelboe!M,!Rohwer!F!(2008)!Marine!viruses:!community!dynamics,!
diversity!and!impact!on!microbial!processes.!In:!Kirchman!D!(ed)!Microbial!ecology!
of!the!oceans.!John!Wiley!and!Sons,!Inc.,!Hoboken,!New!Jersey!
Brum!JR,!Morris!JJ,!Décima!M,!Stukel!MR!(2014)!Mortality!in!the!oceans:!causes!and!
consequences.!Eco?DAS!IX!Chapter!2:16?48!
Calbet!A!(2001)!Mesozooplankton!grazing!effect!on!primary!production:!a!global!
comparative!analysis!in!marine!ecosystems.!Limnol!Oceanogr!46:1824?1830!
Calbet!A,!Landry!MR!(2004)!Phytoplankton!growth,!microzooplankton!grazing,!and!carbon!
cycling!in!marine!systems.!Limnol!Oceanogr!49:51?57!
Caron!DA,!Connell!PE,!Schaffner!RA,!Schnetzer!A,!Fuhrman!JA,!Countway!PD,!Kim!DY!
(2017)!Planktonic!food!web!structure!at!a!coastal!time?series!site:!I.!Partitioning!of!
microbial!abundances!and!carbon!biomass.!Deep!Sea!Res!I!121:14?29!
Caron!DA,!Countway!PD,!Jones!AC,!Kim!DY,!Schnetzer!A!(2012)!Marine!protistan!diversity.!
Annu!Rev!Mar!Sci!4:467?493!
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.!ISME!J!8:816?829!
Chow!C?ET,!Sachdeva!R,!Cram!JA,!Steele!JA,!Needham!DM,!Patel!A,!Parada!AE,!Fuhrman!JA!
(2013)!Temporal!variability!and!coherence!of!euphotic!zone!bacterial!communities!
over!a!decade!in!the!Southern!California!Bight.!ISME!J!7:2259?2273!
Church!MJ,!Ducklow!HW,!Karl!DM!(2004)!Light!dependence!of![(3)H]leucine!incorporation!
in!the!oligotrophic!North!Pacific!Ocean.!Appl!Environ!Microbiol!70:4079?4087!
Collins!LE,!Berelson!W,!Hammond!DE,!Knapp!A,!Schwartz!R,!Capone!D!(2011)!Particle!
fluxes!in!San!Pedro!Basin,!California:!A!four?year!record!of!sedimentation!and!
physical!forcing.!Deep!Sea!Res!I!58:898?914!
! 179!
Connell!PE,!Campbell!V,!Gellene!AG,!Hu!SK,!Caron!DA!(2017)!Planktonic!food!web!structure!
at!a!coastal!time?series!site:!II.!Spatiotemporal!variability!of!microbial!trophic!
activities.!Deep!Sea!Res!I!121:210?223!
Countway!PD,!Vigil!PD,!Schnetzer!A,!Moorthi!SD,!Caron!DA!(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!
Cram!JA,!Chow!C?ET,!Sachdeva!R,!Needham!DM,!Parada!AE,!Steele!JA,!Fuhrman!JA!(2014)!
Seasonal!and!interannual!variability!of!the!marine!bacterioplankton!community!
throughout!the!water!column!over!ten!years.!ISME!J!9:563?580!
De!La!Rocha!CL,!Passow!U!(2014)!The!Biological!Pump!In:!Turekian!KK!(ed)!Treatise!on!
Geochemistry!(Second!Edition).!Elsevier,!Oxford!
del!Giorgio!PA,!Cole!JJ!(1998)!Bacterial!growth!efficiency!in!natural!aquatic!systems.!Annual!
Review!of!Ecology!and!Systematics!29:503?541!
del!Giorgio!PA,!Cole!JJ!(2000)!Bacterial!energetics!and!growth!efficiency.!In:!Kirchman!D!
(ed)!Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.,!New!York!
Ducklow!H!(2000)!Bacterial!production!and!biomass!in!the!oceans.!In:!Kirchman!D!(ed)!
Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.,!New!York!
Ducklow!HW,!Kirchman!DL,!Quinby!HL,!Carlson!CA,!Dam!HG!(1993)!Stocks!and!dynamics!of!
bacterioplankton!carbon!during!the!spring!bloom!in!the!eastern!North!Atlantic!
Ocean.!Deep!Sea!Res!II!40:245?263!
Fenchel!T!(1987)!Ecology!of!protozoa:!the!biology!of!free?living!phagotropic!protists.!
Springer?Verlag!
Field!CB,!Behrenfeld!MJ,!Randerson!JT,!Falkowski!P!(1998)!Primary!production!of!the!
biosphere:!integrating!terrestrial!and!oceanic!components.!Science!281:237?240!
Fuhrman!J!(1992)!Bacterioplankton!roles!in!cycling!of!organic!matter:!the!microbial!food!
web.!Primary!productivity!and!biogeochemical!cycles!in!the!sea!Plenum!Press,!New!
York:361?383!
Fuhrman!JA!(1999)!Marine!viruses!and!their!biogeochemical!and!ecological!effects.!Nature!
399:541?548!
Fuhrman!JA!(2000)!Impact!of!viruses!on!bacterial!processes.!In:!Kirchman!DL!(ed)!
Microbial!ecology!of!the!oceans.!John!WIley!&!Sons,!Inc.,!New!York!
Fuhrman!JA,!Azam!F!(1982)!Thymidine!incorporation!as!a!measure!of!heterotrophic!
bacterioplankton!production!in!marine!surface!waters:!Evaluation!and!field!results.!
Mar!Biol!66:109?120!
Fuhrman!JA,!Hagström!Å!(2008)!Bacterial!and!archael!community!structure!and!its!pattern.!
In:!Kirchman!D!(ed)!Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.,!
Hoboken,!New!Jersey!
Fuhrman!JA,!Hewson!I,!Schwalbach!MS,!Steele!JA,!Brown!MV,!Naeem!S!(2006)!Annually!
reoccurring!bacterial!communities!are!predictable!from!ocean!conditions.!PNAS!
103:13104?13109!
Fuhrman!JA,!Suttle!CA!(1993)!Viruses!in!marine!planktonic!systems.!Oceanography!6:51?63!
González!JM,!Suttle!CA!(1993)!Grazing!by!marine!nanoflagellates!on!viruses!and!virus?sized!
particles:!ingestion!and!digestion.!Mar!Ecol!Prog!Ser!94:1?10!
Hamersley!MR,!Turk!K,!Leinweber!A,!Gruber!N,!Zehr!J,!Gunderson!T,!Capone!D!(2011)!
Nitrogen!fixation!within!the!water!column!associated!with!two!hypoxic!basins!in!the!
Southern!California!Bight.!Aquat!Microb!Ecol!63:193!
! 180!
Hannon!B!(1973)!The!structure!of!ecosystems.!J!Theor!Biol!41:535?546!
Hansen!PJ,!Bjørnsen!PK,!Hansen!BW!(1997)!Zooplankton!grazing!and!growth:!Scaling!
within!the!2?2.0!μm!body!size!range.!Limnol!Oceanogr!42:687?704!
Hansen!PJ,!Calado!AJ!(1999)!Phagotrophic!mechanisms!and!prey!selection!in!free living!
dinoflagellates.!J!Eukaryot!Microbiol!46:382?389!
Haskell!WZ,!Prokopenko!MG,!Hammond!DE,!Stanley!RHR,!Berelson!WM,!Baronas!JJ,!
Fleming!JC,!Aluwihare!L!(2016)!An!organic!carbon!budget!for!coastal!Southern!
California!determined!by!estimates!of!vertical!nutrient!flux,!net!community!
production!and!export.!Deep!Sea!Res!I!116:49?76!
Haskell!WZ,!Prokopenko!MG,!Hammond!DE,!Stanley!RHR,!Sandwith!ZO!(2017)!Annual!
cyclicity!in!export!efficiency!in!the!inner!Southern!California!Bight.!Global!
Biogeochem!Cycles!31:357?376!
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!
Jacox!MG,!Edwards!CA,!Kahru!M,!Rudnick!DL,!Kudela!RM!(2015)!The!potential!for!
improving!remote!primary!productivity!estimates!through!subsurface!chlorophyll!
and!irradiance!measurement.!Deep!Sea!Res!II!112:107?116!
John!SG,!Mendez!J,!Moffett!J,!Adkins!J!(2012)!The!flux!of!iron!and!iron!isotopes!from!San!
Pedro!Basin!sediments.!Geochim!Cosmochim!Ac!93:14?29!
Jürgens!K,!Massana!R!(2008)!Protistan!grazing!on!marine!bacterioplankton.!In:!Kirchman!
DL!(ed)!Microbial!Ecology!of!the!Oceans.!John!Wiley!&!Sons,!Inc.!
Karayanni!H,!Christaki!U,!Van!Wambeke!F,!Thyssen!M,!Denis!M!(2008)!Heterotrophic!
nanoflagellate!and!ciliate!bacterivorous!activity!and!growth!in!the!northeast!Atlantic!
Ocean:!a!seasonal!mesoscale!study.!Aquat!Microb!Ecol!51:169?181!
Karl!DM,!Church!MJ!(2014)!Microbial!oceanography!and!the!Hawaii!Ocean!Time?series!
programme.!Nat!Rev!Microbiol!12:699?713!
Kim!DY,!Countway!PD,!Jones!AC,!Schnetzer!A,!Yamashita!W,!Tung!C,!Caron!DA!(2014)!
Monthly!to!interannual!variability!of!microbial!eukaryote!assemblages!at!four!
depths!in!the!eastern!North!Pacific.!ISME!J!8:515?530!
Kimmance!S,!Wilson!W,!Archer!S!(2007)!Modified!dilution!technique!to!estimate!viral!
versus!grazing!mortality!of!phytoplankton:!limitations!associated!with!method!
sensitivity!in!natural!waters.!Aquat!Microb!Ecol!49:207?222!
Kirchman!DL!(2016)!Growth!rates!of!microbes!in!the!oceans.!Annu!Rev!Mar!Sci!8:285?309!
Kones!JK,!Soetaert!K,!van!Oevelen!D,!Owino!JO,!Mavuti!K!(2006)!Gaining!insight!into!food!
webs!reconstructed!by!the!inverse!method.!J!Mar!Syst!60:153?166!
Laws!EA!(1991)!Photosynthetic!quotients,!new!production!and!net!community!production!
in!the!open!ocean.!Deep!Sea!Research!Part!A!Oceanographic!Research!Papers!
38:143?167!
Michelou!VK,!Cottrell!MT,!Kirchman!DL!(2007)!Light?stimulated!bacterial!production!and!
amino!acid!assimilation!by!cyanobacteria!and!other!microbes!in!the!North!Atlantic!
Ocean.!Appl!Environ!Microbiol!73:5539?5546!
Middelboe!M,!Jørgensen!NO!(2006)!Viral!lysis!of!bacteria:!an!important!source!of!dissolved!
amino!acids!and!cell!wall!compounds.!J!Mar!Biol!Assoc!UK!86:605?612!
! 181!
Middelboe!M,!Lyck!PG!(2002)!Regeneration!of!dissolved!organic!matter!by!viral!lysis!in!
marine!microbial!communities.!Aquat!Microb!Ecol!27:187?194!
Middelboe!M,!Riemann!L,!Steward!GF,!Hansen!V,!Nybroe!O!(2003)!Virus?induced!transfer!of!
organic!carbon!between!marine!bacteria!in!a!model!community.!Aquat!Microb!Ecol!
33:1?10!
Montesinos!ML,!Herrero!A,!Flores!E!(1997)!Amino!acid!transport!in!taxonomically!diverse!
cyanobacteria!and!identification!of!two!genes!encoding!elements!of!a!neutral!amino!
acid!permease!putatively!involved!in!recapture!of!leaked!hydrophobic!amino!acids.!J!
Bacteriol!179:853?862!
Muñoz?Marín!MdC,!Luque!I,!Zubkov!MV,!Hill!PG,!Diez!J,!García?Fernández!JM!(2013)!
Prochlorococcus!can!use!the!Pro1404!transporter!to!take!up!glucose!at!nanomolar!
concentrations!in!the!Atlantic!Ocean.!PNAS!110:8597?8602!
Nagata!T!(2000)!Production!mechanisms!of!dissolved!organic!matter.!In:!Kirchman!DL!(ed)!
Microbial!ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.!,!New!York,!NY!
Nelson!JR,!Beers!JR,!Eppley!RW,!Jackson!GA,!McCarthy!JJ,!Soutar!A!(1987)!A!particle!flux!
study!in!the!Santa!Monica?San!Pedro!Basin!off!Los!Angeles:!Particle!flux,!primary!
production,!and!transmissometer!survey.!Cont!Shelf!Res!7:307?328!
Neuer!S,!Cowles!TJ!(1994)!Protist!herbivory!in!the!Oregon!upwelling!system.!Mar!Ecol!Prog!
Ser!113:147?162!
Nezlin!NP,!Li!B?L!(2003)!Time?series!analysis!of!remote?sensed!chlorophyll!and!
environmental!factors!in!the!Santa!Monica–San!Pedro!Basin!off!Southern!California.!
J!Mar!Syst!39:185?202!
Nezlin!NP,!Sutula!MA,!Stumpf!RP,!Sengupta!A!(2012)!Phytoplankton!blooms!detected!by!
SeaWiFS!along!the!central!and!southern!California!coast.!J!Geophys!Res?Oceans!117!
Paffenhöfer!G,!Knowles!S!(1979)!Ecological!implications!of!fecal!pellet!size,!production!and!
consumption!by!copepods.!J!Mar!Res!37!
Paffenhöfer!G?A,!Van!Sant!KB!(1985)!The!feeding!response!of!a!marine!planktonic!copepod!
to!quantity!and!quality!of!particles.!Mar!Ecol!Prog!Ser:55?65!
Poulsen!L,!K.!,!Kiørboe!T!(2006)!Vertical!flux!and!degradation!rates!of!copepod!fecal!pellets!
in!a!zooplankton!community!dominated!by!small!copepods.!Mar!Ecol!Prog!Ser!
323:195?204!
Richardson!TL,!Jackson!GA,!Ducklow!HW,!Roman!MR!(2004)!Carbon!fluxes!through!food!
webs!of!the!eastern!equatorial!Pacific:!an!inverse!approach.!Deep!Sea!Res!I!51:1245?
1274!
Roberts!GO,!Gelman!A,!Gilks!WR!(1997)!Weak!convergence!and!optimal!scaling!of!random!
walk!Metropolis!algorithms.!Ann!Appl!Probab!7:110?120!
Robinson!C!(2008)!Heterotrophic!bacterial!respiration.!In:!Kirchman!DL!(ed)!Microbial!
ecology!of!the!oceans.!John!Wiley!&!Sons,!Inc.,!New!York!
Saint?Béat!B,!Vézina!AF,!Asmus!R,!Asmus!H,!Niquil!N!(2013)!The!mean!function!provides!
robustness!to!linear!inverse!modelling!flow!estimation!in!food!webs:!A!comparison!
of!functions!derived!from!statistics!and!ecological!theories.!Ecol!Modell!258:53?64!
Sarthou!G,!Timmermans!KR,!Blain!S,!Tréguer!P!(2005)!Growth!physiology!and!fate!of!
diatoms!in!the!ocean:!a!review.!J!Sea!Res!53:25?42!
Schmoker!C,!Hernández?León!S,!Calbet!A!(2013)!Microzooplankton!grazing!in!the!oceans:!
impacts,!data!variability,!knowledge!gaps!and!future!directions.!J!Plankton!Res!
35:691?706!
! 182!
Schnetzer!A,!Miller!PE,!Schaffner!RA,!Stauffer!BA,!Jones!BH,!Weisberg!SB,!DiGiacomo!PM,!
Berelson!WM,!Caron!DA!(2007)!Blooms!of!Pseudo?nitzschia!and!domoic!acid!in!the!
San!Pedro!Channel!and!Los!Angeles!harbor!areas!of!the!Southern!California!Bight,!
2003–2004.!Harmful!Algae!6:372?387!
Selph!KE,!Landry!MR,!Taylor!AG,!Gutiérrez?Rodríguez!A,!Stukel!MR,!Wokuluk!J,!Pasulka!A!
(2015)!Phytoplankton!production!and!taxon?specific!growth!rates!in!the!Costa!Rica!
Dome.!J!Plankton!Res:fbv063!
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!
Sherr!EB,!Sherr!BF!(2008)!Understanding!roles!of!microbes!in!marine!pelagic!food!webs:!a!
brief!history.!In:!Kirchman!D!(ed)!Microbial!ecology!of!the!oceans.!John!Wiley!&!
Sons,!Inc.,!Hoboken,!New!Jersey!
Soetaert!K,!Van!den!Meersche!K,!van!Oevelen!D!(2009)!limSolve:!Solving!linear!inverse!
models.!R!package!version!1,5,4!
Soetaert!K,!Van!Oevelen!D!(2009)!Package!LIM,!implementing!linear!inverse!models!in!R.!!
Southward!AJ,!Langmead!O,!Hardman?Mountford!NJ,!Aiken!J,!Boalch!GT,!Dando!PR,!Genner!
MJ,!Joint!I,!Kendall!MA,!Halliday!NC!(2004)!Long?term!oceanographic!and!ecological!
research!in!the!western!English!Channel.!Adv!Mar!Biol!47:1?105!
Steinberg!DK,!Carlson!CA,!Bates!NR,!Johnson!RJ,!Michaels!AF,!Knap!AH!(2001)!Overview!of!
the!US!JGOFS!Bermuda!Atlantic!Time?series!Study!(BATS):!a!decade?scale!look!at!
ocean!biology!and!biogeochemistry.!Deep!Sea!Res!II!48:1405?1447!
Stoecker!DK,!Hansen!PJ,!Caron!DA,!Mitra!A!(2017)!Mixotrophy!in!the!marine!plankton.!
Annu!Rev!Mar!Sci!9:311?335!
Stoecker!DK,!Nejstgaard!JC,!Madhusoodhanan!R,!Pohnert!G,!Wolfram!S,!Jakobsen!HH,!
Šulčius!S,!Larsen!A!(2015)!Underestimation!of!microzooplankton!grazing!in!dilution!
experiments!due!to!inhibition!of!phytoplankton!growth.!Limnol!Oceanogr!
Straile!D!(1997)!Gross!growth!efficiencies!of!protozoan!and!metazoan!zooplankton!and!
their!dependence!on!food!concentration,!predator?prey!weight!ratio,!and!taxonomic!
group.!Limnol!Oceanogr!42:1375?1385!
Stukel!MR,!Landry!MR!(2010)!Contribution!of!picophytoplankton!to!carbon!export!in!the!
equatorial!Pacific:!A!reassessment!of!food!web!flux!inferences!from!inverse!models.!
Limnol!Oceanogr!55:2669?2685!
Stukel!MR,!Landry!MR,!Ohman!MD,!Goericke!R,!Samo!T,!Benitez?Nelson!CR!(2012)!Do!
inverse!ecosystem!models!accurately!reconstruct!plankton!trophic!flows?!
Comparing!two!solution!methods!using!field!data!from!the!California!Current.!J!Mar!
Syst!91:20?33!
Suttle!CA!(2000)!Cyanophages!and!their!role!in!the!ecology!of!cyanobacteria.!!The!ecology!
of!cyanobacteria.!Springer!Netherlands!
Suttle!CA,!Chen!F!(1992)!Mechanisms!and!rates!of!decay!of!marine!viruses!in!seawater.!
Appl!Environ!Microbiol!58:3721?3729!
Talarmin!A,!Van!Wambeke!F,!Catala!P,!Courties!C,!Lebaron!P!(2011)!Flow!cytometric!
assessment!of!specific!leucine!incorporation!in!the!open!Mediterranean.!
Biogeosciences!8:253?265!
! 183!
Taylor!AG,!Landry!MR,!Selph!KE,!Wokuluk!JJ!(2015)!Temporal!and!spatial!patterns!of!
microbial!community!biomass!and!composition!in!the!Southern!California!Current!
Ecosystem.!Deep!Sea!Res!II!112:117?128!
Taylor!FJR,!Hoppenrath!M,!Saldarriaga!JF!(2008)!Dinoflagellate!diversity!and!distribution.!
Biodivers!Conserv!17:407?418!
Teel!EN!(2017)!Multi?dataset!analysis!of!bacterial!heterotrophic!variability!at!the!San!
Pedro!Ocean!Time!Series!(SPOT):!an!investigation!into!the!necessity!and!feasibility!
of!incorporating!a!dynamic!bacterial!component!into!global!ecosystem!NPZD!
models.!Doctor!of!Philosophy:!Biological!Sciences,!University!of!Southern!California,!!
Turner!JT!(2002)!Zooplankton!fecal!pellets,!marine!snow!and!sinking!phytoplankton!
blooms.!Aquat!Microb!Ecol!27:57?102!
Van!den!Meersche!KS,!Karline;!Van!Oevelen,!Dick!(2009)!xsample():!an!R!function!for!
sampling!linear!inverse!problems.!J!Stat!Softw!30:1?15!
van!Oevelen!D,!Van!den!Meersche!K,!Meysman!FJ,!Soetaert!K,!Middelburg!JJ,!Vézina!AF!
(2010)!Quantifying!food!web!flows!using!linear!inverse!models.!Ecosystems!13:32?
45!
Verity!PG,!Stoecker!DK,!Sieracki!ME,!Nelson!JR!(1993)!Grazing,!growth!and!mortality!of!
microzooplankton!during!the!1989!North!Atlantic!spring!bloom!at!47°N,!18°W.!Deep!
Sea!Res!I!40:1793?1814!
Vézina!AF,!Demers!S,!Laurion!I,!Sime?Ngando!T,!Kim!Juniper!S,!Devine!L!(1997)!Carbon!
flows!through!the!microbial!food!web!of!first?year!ice!in!Resolute!Passage!(Canadian!
High!Arctic).!J!Mar!Syst!11:173?189!
Vila?Costa!M,!Simó!R,!Harada!H,!Gasol!JM,!Slezak!D,!Kiene!RP!(2006)!
Dimethylsulfoniopropionate!uptake!by!marine!phytoplankton.!Science!314:652?654!
Viviani!DA,!Church!MJ!(2017)!Decoupling!between!bacterial!production!and!primary!
production!over!multiple!time!scales!in!the!North!Pacific!Subtropical!Gyre.!Deep!Sea!
Res!I!121:132?142!
Weinbauer!MG!(2004)!Ecology!of!prokaryotic!viruses.!FEMS!microbiology!reviews!28:127?
181!
Weitz!JS,!Stock!CA,!Wilhelm!SW,!Bourouiba!L,!Coleman!ML,!Buchan!A,!Follows!MJ,!Fuhrman!
JA,!Jover!LF,!Lennon!JT,!Middelboe!M,!Sonderegger!DL,!Suttle!CA,!Taylor!BP,!Frede!
Thingstad!T,!Wilson!WH,!Eric!Wommack!K!(2015)!A!multitrophic!model!to!quantify!
the!effects!of!marine!viruses!on!microbial!food!webs!and!ecosystem!processes.!ISME!
J!9:1352?1364!
Wilhelm!SW,!Suttle!CA!(1999)!Viruses!and!nutrient!cycles!in!the!sea:!Viruses!play!critical!
roles!in!the!structure!and!function!of!aquatic!food!webs.!BioScience!49:781?788!
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!
Zubkov!MV,!Fuchs!BM,!Tarran!GA,!Burkill!PH,!Amann!R!(2003)!High!rate!of!uptake!of!
organic!nitrogen!compounds!by!Prochlorococcus!cyanobacteria!as!a!key!to!their!
dominance!in!oligotrophic!oceanic!waters.!Appl!Environ!Microbiol!69:1299?1304!

! !
! 184!
APPENDIX/A:/SUPPLEMENTAL/TABLES/AND/FIGURES/
/
Table A1. Microbial abundances (cells L
-1
) for each study site and date sampled.
Microbial assemblages include: BACT (bacteria + archaea), SYN (Synechococcus), PRO
(Prochlorococcus), PEUK (photosynthetic picoeukaryotes), PNANO
(photosynthetic/mixotrophic nanoplankton), HNANO (heterotrophic nanoplankton), DINO
(dinoflagellates), CIL (loricate/aloricate ciliates), and DIAT (diatoms).
Cell Abundances (cells L
-1
)

Experiment
BACT
x 10
9

SYN
x 10
7

PRO
x 10
7

PEUK
x 10
7

PNANO
x 10
5

HNANO
x 10
5

CIL
x 10
3

DINO
x 10
3

DIAT
x 10
4

Port of Los Angeles
13-Mar-12
- - - - 3.86 8.85 1.30 3.90 17.3
15-Jul-12
3.02 0.11 0.03 0.47 0.44 22.6 0.87 0.00 42.9
19-Oct-12
- 7.14 1.16 2.01 - - 3.90 5.20 3.77
9-Jan-13
1.90 0.90 0.20 0.92 - - 1.41 1.41 31.8
24-Apr-13
- 0.17 0.11 0.87 - - 10.4 62.4 201
15-Jul-13
1.10 2.18 0.16 5.20 - - 18.2 44.2 25.5
16-Oct-13
2.60 1.27 0.73 3.01 24.7 22.9 22.6 56.5 2.71
7-Jan-14
1.56 1.63 1.82 1.38 5.49 25.6 0.86 7.80 5.72
8-Apr-14
2.02 0.23 0.01 2.83 49.0 40.2 0.00 31.2 244
14-Jul-14
2.69 2.96 0.09 5.92 18.6 74.1 8.32 45.8 62.6
6-Oct-14 2.78 7.13 0.75 3.12 36.7 35.7 12.5 33.3 9.57
San Pedro Ocean Time Series
13-Mar-12
- - - - 6.78 7.66 0.00 5.20 3.64
15-Jul-12
1.06 6.08 6.82 0.58 - - 0.86 0.87 3.73
19-Oct-12
- 5.63 4.71 0.55 - - 0.00 1.30 0.13
9-Jan-13
1.39 2.56 0.42 0.15 - - 11.7 5.20 20.3
24-Apr-13
- 1.94 0.02 0.34 - - 0.87 2.60 3.47
15-Jul-13
0.62 19.4 0.88 1.02 - - 0.26 0.26 0.10
16-Oct-13
1.53 3.96 2.15 1.19 7.48 16.0 1.54 0.00 0.39
7-Jan-14
0.76 2.87 14.8 0.47 8.51 30.6 1.12 0.56 0.17
8-Apr-14
1.74 0.74 0.10 0.23 10.1 16.5 3.47 0.36 41.3
14-Jul-14
1.18 3.09 2.50 0.24 0.91 12.0 0.53 1.06 6.63
6-Oct-14 1.92 2.49 0.59 0.65 1.25 10.8 1.44 0.48 0.02
Big Fisherman's Cove
13-Mar-12
- - - - 4.15 11.5 1.30 0.00 18.3
15-Jul-12
1.45 2.36 3.09 0.44 - - 0.00 0.00 4.94
19-Oct-12
- 6.21 5.32 0.52 - - 0.06 0.00 0.0020
9-Jan-13
1.41 0.95 0.11 0.44 - - 0.00 0.00 0.91
24-Apr-13
- 4.64 0.11 1.03 - - 3.90 7.80 4.81
15-Jul-13
0.44 12.2 3.94 1.99 - - 0.52 0.00 0.16
16-Oct-13
1.92 4.35 1.90 2.54 9.88 10.1 1.89 0.00 3.59
! 185!
/
! /
7-Jan-14
1.14 2.85 6.52 0.65 4.59 7.78 1.21 1.51 2.14
8-Apr-14
1.46 0.14 0.04 0.17 8.65 12.6 4.46 13.4 35.5
14-Jul-14
1.24 3.03 1.45 0.40 3.19 22.5 0.66 1.97 3.35
6-Oct-14 1.74 4.37 0.68 0.98 5.07 11.6 1.97 1.18 0.12
! 186!
Table A2. The impact of nutrient-enrichment on total phytoplankton growth rate. Mean
nutrient-enriched growth rate (µ
n
) and mean unenriched growth rate (µ
0
) were compared
using Welch Two-Sample T-tests.
!
Mean µ
n
(d
-1
) Mean µ
0
(d
-1
) p-value
All sites and sampling dates 1.05 0.36 1.90E-06

Site

Port of LA 0.95 0.56 0.045
SPOT 0.91 0.17 0.005
Catalina 1.28 0.35 0.002

Season

Winter 0.53 0.18 0.082
Spring 1.23 0.56 0.023
Summer 1.27 0.15 0.003
Fall 0.98 0.51 0.002

! 187!
Table A3. The maximum efficiency of photosystem II (F
v
/F
m
) as measured from discrete
samples at the beginning (T
0
) and end (T
f
) of each dilution experiment. The amount of
incident light that reached the phytoplankton through the neutral density screening (%)
and length of the incubation (h) are also reported.

Station T
0
F
v
/F
m
T
f
F
v
/F
m

Percent Incident
Light (%)
Incubation
Length (h)
8 --- --- 60 24
29 0.558 0.577 15 24
42 0.303 0.393 15 72
64 --- --- 15 72
93 0.394 0.423 15 72
105 0.452 0.484 15 72
132 0.529 0.458 60 24
151 0.493 0.520 15 48
186 0.509 0.480 15 48
209 0.506 0.544 15 48

!
! 188!
Table A4. The impact of nutrient-enrichment and removal of metazoan grazers on the
apparent growth rates (µ; d
−1
) of the total phytoplankton (based on chlorophyll a) and the
phototrophic picoeukaryotes in the dilution experiments. Apparent growth rates were
calculated from three bottles and averaged for each treatment: the nutrient-enriched,
100% whole seawater (WSW) treatment, the unenriched, 100% WSW treatment, and the
unenriched, < 200 µm filtered treatment. Welch two-sample t-tests were used to compare
mean growth rates in the nutrient-enriched and unenriched treatments and in the WSW
and <200 µm filtered treatments. Bold values indicate a significant difference between the
apparent growth rates of the denoted treatments at p ≤ 0.05.

Mean apparent µ (d
−1
) of treatment T-test p-value
Station
100% WSW 100% WSW <200 µm Enriched vs. WSW vs.
enriched unenriched unenriched Unenriched <200 µm
Total phytoplankton (chlorophyll a)
8 0.17 0.40 0.46 0.08 0.28
29 −0.02 −0.03 0.04 0.77 0.19
42 0.24 0.22 0.36 0.76 0.01
64 0.19 0.25 0.23 0.48 0.77
93 0.46 0.47 0.47 0.68 1.00
105 0.16 0.08 0.21 0.35 0.21
132 0.03 −0.25 −0.11 0.05 0.20
151 0.25 0.37 0.32 0.04 0.44
186 0.21 0.20 0.09 0.69 0.10
209 0.32 0.35 0.22 0.71 0.12
Phototrophic picoeukaryotes
8 0.38 0.37 0.33 0.96 0.61
29 −0.48 −0.62 −0.28 0.40 0.06
42 −0.1 −0.12 −0.12 0.72 0.93
64 −0.34 −0.07 0.14 0.29 0.37
93 0.02 0.26 0.43 0.08 0.08
105 0.02 0.06 −0.05 0.61 0.18
132 0.16 0.42 0.30 0.13 0.27
151 −0.13 −0.05 −0.02 0.34 0.76
186 −0.31 −0.38 −0.08 0.57 0.03
209 0.11 0.13 0.28 0.72 0.01

! 189!
Table A5. GPS coordinates and environmental metadata for the 25 Lagrangian stations sampled at 15 m for nanoplankton
abundances and grazing pressure during the 4-d study period. Stations numbers are consistent with cruise designations to
allow comparison between studies. Unit abbreviations are as follows: decimal degrees (dec.deg) and hour:minute (HH:MM).

!
Date

Station

Latitude (°N)
dec.deg
Longitude (°W)
dec.deg
Time
HH:MM
Temperature
(°C)
Salinity

Dissolved Oxygen
(µmol L
−1
)
Chlorophyll a
(µg L
−1
)
26-Jul-15 6 24.35 156.80 6:00 26.59 35.38 204.90 0.22
26-Jul-15 7 24.37 156.80 10:00 26.63 35.39 205.20 0.17
26-Jul-15 8 24.39 156.77 14:00 26.63 35.39 206.10 0.14
26-Jul-15 11 24.37 156.79 18:00 26.68 35.38 206.60 0.20
26-Jul-15 14 24.42 156.76 22:00 26.73 35.37 205.70 0.22
27-Jul-15 15 24.46 156.74 2:00 26.71 35.38 205.30 0.23
27-Jul-15 16 24.48 156.72 6:00 26.68 35.38 205.20 0.22
27-Jul-15 17 24.50 156.69 10:00 26.68 35.38 205.70 0.15
27-Jul-15 18 24.53 156.65 14:00 26.92 35.39 205.60 0.13
27-Jul-15 19 24.54 156.64 18:00 26.82 35.39 206.10 0.25
27-Jul-15 20 24.54 156.63 22:00 26.90 35.38 205.00 0.23
28-Jul-15 21 24.55 156.58 2:00 26.81 35.38 205.00 0.23
28-Jul-15 22 24.55 156.58 6:00 26.73 35.38 205.20 0.24
28-Jul-15 23 24.56 156.56 10:00 26.81 35.37 205.10 0.15
28-Jul-15 24 24.60 156.52 14:00 26.86 35.38 205.90 0.15
28-Jul-15 26 24.60 156.51 18:00 26.94 35.37 206.10 0.20
28-Jul-15 28 24.60 156.50 22:00 26.90 35.37 205.90 0.23
29-Jul-15 29 24.59 156.44 2:00 26.72 35.39 207.60 0.03
29-Jul-15 30 24.58 156.42 6:00 26.88 35.38 205.50 0.24
29-Jul-15 31 24.57 156.43 10:00 26.88 35.38 205.60 0.15
29-Jul-15 32 24.58 156.41 14:00 26.93 35.37 205.90 0.15
29-Jul-15 33 24.58 156.39 18:00 26.83 35.38 207.10 0.24
29-Jul-15 34 24.58 156.38 22:00 26.98 35.37 205.40 0.23
30-Jul-15 35 24.58 156.36 2:00 26.92 35.37 205.20 0.23
30-Jul-15 37 24.58 156.36 6:00 26.86 35.38 205.00 0.23

Mean 26.80 35.38 205.68 0.19

Standard Error 0.11 0.01 0.67 0.05

Table A8. Mean measured values and model solutions for the rate processes used as
experimental inputs for the SPOT linear inverse model. All measurement units are mg C
m
−3
d
−1
. The four model scenarios include (1) ‘Average’: with viral lysis and using SPOT
measurement error, (2) ‘+ BP’: with viral lysis and setting the bacterial production value
to empirical measurement mean (3) ‘− Virus’: without viral lysis and using SPOT
measurement error, and (4) ‘+ BP – Virus’: without viral lysis and setting the bacterial
production value to empirical measurement mean. Abbreviations include: heterotrophic
bacteria (BACT), all phytoplankton (PHYTO), Prochlorococcus (PRO), Synechococcus
(SYN), phototrophic picoeukaryotes (PEUK), net primary production (NPP), and gross
primary production (GPP). Measurements of NPP were conducted using the dilution
method, which does not take into account viral mortality; NPP values are thus presented
as a direct comparison to the dilution value with the total NPP value in parentheses.
SPOT Model Scenario
Rate Process Measured Average + BP − Virus + BP − Virus
BACT Production 4.96 2.17 4.96 1.97 4.50*
BACT Protistan Grazing Mortality 4.83 1.47 0.10 1.97 4.48
PHYTO NPP 9.35 18.0 (19.6) 16.1 (23.3) 19.3 23.2
Total PHYTO Protistan Grazing Mortality 11.0 11.1 11.7 12.7 16.2
PRO NPP 0.57 1.28 (1.52) 0.15 (1.70) 1.46 1.69
PRO Protistan Grazing Mortality 1.60 1.17 0.15 1.45 1.69
SYN NPP 3.31 3.78 (5.14) 0.14 (5.76) 4.81 5.75
SYN Protistan Grazing Mortality 5.63 3.76 0.11 4.78 5.72
PEUK NPP 0.81 4.63 6.43 4.93 6.42
PEUK Protistan Grazing Mortality 1.74 4.58 6.34 4.88 6.35
GPP 63.3 29.7 46.3 29.2 45.8
*Set slightly below the to
3
H-thymidine average due to biological constraints on model solution (Table
2).
!
! 193!
Figure A1. Sample dilution curves from experiments conducted at Port of Los Angeles on
April 24, 2013 (first column) and April 8, 2014 (second column). Plots are shown for the
total phytoplankton community (based on chlorophyll a; first row), Synechococcus (SYN,
second row), Prochlorococcus (PRO, third row), and the photosynthetic picoeukaryotes
(PEUK, fourth row). The apparent growth rate (µ) of the phytoplankton in each bottle was
plotted against the dilution level of that bottle (fraction of the bottle that is whole,
unfiltered seawater (WSW)). Blue circles represent the nutrient-enriched, 5-point dilution
series (WSW enriched), red squares represent the unenriched, 100% WSW treatment
(WSW unenriched), and green triangles represent the unenriched, <200 µm treatment
(<200 µm unenriched). Model I regressions are shown when significant (p ≤ 0.05; black
lines). Non-linear responses due to grazing saturation were not detected, despite the
high chlorophyll concentrations measured during these experiments (April 2013: 12.70
µg L
-
1; April 2014: 15.04 µg L
-1
).

Port of Los Angeles
April 24, 2013
Chl a = 12.70 µg L
-1

Port of Los Angeles
April 8, 2014
Chl a = 15.04 µg L
-1

-1.5
-1
-0.5
0
0.5
1
0.0 0.2 0.4 0.6 0.8 1.0
Apparent Growth Rate (per day)
Dilution Level (Fraction WSW)
Port%of%Los%Angeles%040814,%Chl%a%
WSW enriched
WSW unenriched
<200um unenriched
Total
PHYTO
SYN
PRO
PEUK
! 194!
Figure A2. Unenriched growth rates (µ
0
) and grazing mortality rates (m) of
Synechococcus (SYN), Prochlorococcus (PRO), and photosynthetic picoeukaryotes
(PEUK) for the three study sites: Port of Los Angeles (Port of LA), San Pedro Ocean
Time-series station (SPOT), and Catalina Island (Catalina). Values at each site include all
seasons. The boxplot displays the distribution of the data using the interquartile range
(IQR) (rectangle; bottom is first quartile, top is third quartile), the median (horizontal
segment within the rectangle), and the minimum and maximum values that fall within
1.5*IQR (bottom and top “whiskers”, or vertical lines). Outliers (black points) were
identified by R according to the Boxplot Rule (Wilcox 2011). Sites with significantly
different mean values (▲) or median values ( ) are marked above each site in the pair
(p ≤ 0.05).

!
! 195!
Figure A3. Unenriched growth rates (µ
0
) and grazing mortality rates (m) of
Synechococcus (SYN), Prochlorococcus (PRO), and photosynthetic picoeukaryotes
(PEUK) for each season: Winter (January), Spring (March-April), Summer (July), and Fall
(October). Values in each season include all sites. The boxplot displays the distribution
of the data using the interquartile range (IQR) (rectangle; bottom is first quartile, top is
third quartile), the median (horizontal segment within the rectangle), and the minimum
and maximum values that fall within 1.5*IQR (bottom and top “whiskers”, or vertical
lines). Outliers (black points) were identified by R according to the Boxplot Rule (Wilcox
2011). Seasons with significantly different mean values (▲) or median values ( ) are
marked above each season in the pair (p ≤ 0.05).

!
! 196!
Figure A4. Spearman’s rank correlation coefficients (rho) between environmental factors
and population abundances, unenriched growth rates or production rates, and grazing
mortality rates of the microbial assemblages. Significant correlations are marked by an *
(p < 0.05). Microbial assemblages include: BACT (bacteria), PEUK (phototrophic
picoeukaryotes), PMNANO (phototrophic/mixotrophic nanoplankton), HNANO
(heterotrophic nanoplankton), CILIATE (ciliates), DINO (dinoflagellates), DIATOM
(diatoms), and PHYTO (total phytoplankton assemblage). Growth and mortality rates
have the units of “d
−1
”, while production and carbon consumption rates have the units of
“µg C l
−1
d
−1
”.
!
! !
! 197!

Figure A5. Inferring Prochlorococcus mortality using changes in Prochlorococcus cell
abundance over time. Points represent the hourly Prochlorococcus abundance, with
orange points denoting the 6:00 time-point and black dots denoting the 18:00 time-point.
Day-night cycles are shown, with gray boxes indicating nighttime. Subtracting the initial
Prochlorococcus abundance (Pi) from the final Prochlorococcus abundance (Pf) within a
24-h period provides the actual, net production of Prochlorococcus cells that day. The
expected production of cells, in the absence of mortality, can be calculated by
multiplying the initial Prochlorococcus abundance (Pi) by the daily division rate to obtain
the expected final Prochlorococcus abundance (Pe) (see Supplementary Text).
Subtracting the actual production from the expected production provides an estimate of
daily Prochlorococcus cell removal by grazers, viruses, and/or physical loss processes.

! !
! 198!
Figure A6. Indirect fluxes of carbon from the primary producers to (a) the heterotrophic
bacteria, and (b) to the mesozooplankton in the four model scenarios (columns). Values
do not represent direct fluxes of energy from a phytoplankton compartment to a
heterotrophic compartment, but rather represent the ultimate source of the carbon
transferred to each heterotrophic compartment (both directly and indirectly). The four
model scenarios include (1) ‘Average’: with viral lysis and using SPOT measurement
error, (2) ‘+ BP’: with viral lysis and setting the bacterial production value to empirical
measurement mean (3) ‘− Virus’: without viral lysis and using SPOT measurement error,
and (4) ‘+ BP – Virus’: without viral lysis and setting the bacterial production value to
empirical measurement mean.

!
! 199!
Figure A7. Mass balance calculations to determine the percentage of dissolved organic
carbon that would need to be externally sourced to support a bacterial production to net
primary production ratio of 0.53. Possible external sources of DOC include advection
from a different location or depth, atmospheric deposition, or allochthonous input from
nearby terrestrial ecosystems. Green values indicate known inputs to the mass balance
(given in Table 1), while the red value is the derived amount of external DOC required in
this scenario to support the bacterial assemblage. Mass balance calculations, and the
assumptions made to complete them, are described in the Supplemental Text.

! !
! 200!
APPENDIX!B:!SUPPLEMENTAL!TEXT!
!
Chapter Three Supplemental Text:
Daily changes in the population abundances of Prochlorococcus and Synechococcus were
used to calculate the mean amount of picocyanobacterial mortality occurring each day during our
study period (Fig. A5). The actual production of cells (net growth) was determined for each 24-h
period (6:00 to 6:00) by subtracting the final picocyanobacterial population abundance from the
initial picocyanobacterial abundance (Fig. A5; Table A6 and A7). The expected production of
cells—in the absence of removal from grazers, viruses, and/or physical processes—was
determined by multiplying the measured growth rate (d
−1
) by the initial picocyanobacterial
abundance. The growth rate for each day was calculated by summing the observed hourly
division rates (N=24). We then calculated the number of picocyanobacterial cells that were
removed by subtracting the actual production of cells from the expected production of cells. This
provided us with the number of cells that would need to be consumed by grazers, assuming that
grazing removes prey cells after the have completed cell division. Dividing this number in half
provides the number of picocyanobacterial cells that would need to be consumed before they
divide to maintain the population abundances observed during our study.
The estimated number of Prochlorococcus and Synechococcus cells removed was
converted to the amount of carbon biomass removed using carbon conversion factors from the
literature. If we assume that each cyanobacterial cell divides to create two, equally-sized
daughter cells exactly half of the size of the parent cell, then the amount of carbon ingested if
prey cells are consumed prior to cell division is equal to the amount of carbon ingested if the
prey cells are consumed after cell division. Therefore, for Prochlorococcus we used carbon
conversion factors of 100 fg C cell
−1
for pre-division cells and 50 fg C cell
−1
for post-division
! 201!
cells (Table A6) (Cermak et al. 2017). For Synechococcus, we used carbon conversion factors of
250 fg C cell
−1
for pre-division cells and 125 fg C cell
−1
for post-division cells (Table A7),
assuming a mean Synechococcus carbon biomass of 200 fg C cell
−1
(Caron et al. 2017).

Chapter Four Supplemental Text:
Simple mass balance calculations suggest that significant amounts of externally sourced
DOC would be required to support a BP:NPP ratio of 0.53 (Fig. S2). Empirical measurements
show that the mean bacterial production value is 4.96 mg C m
−3
d
−1
while the mean net primary
production value is 9.35 mg C m
−3
d
−1
(Table 1). Succinctly, 33 mg C m
−3
d
−1
of DOC must be
produced to support a bacterial production value of 4.96 mg C m
−3
d
−1
if BGE is 15% (4.96 ÷
0.15 = 33). If net primary production is 65% of gross primary production and 14% of gross
primary production is lost to phytoplankton exudation, then phytoplankton will produce 2.01 mg
C m
−3
d
−1
([9.35 ÷ 0.65] × 0.14 = 2.01). If we then allow viruses and grazers to enact equal
amounts of mortality on the heterotrophic bacteria and phytoplankton (assuming both eukaryotic
and prokaryotic phytoplankton are lysed), viruses would create 2.48 mg C m
−3
d
−1
due to lysis of
bacteria (4.96 × 0.5 = 2.48) and 4.68 mg C m
−3
d
−1
(9.35 × 0.5 = 4.68) due to lysis of
phytoplankton while grazers would create 1.24 mg C m
−3
d
−1
due to bacterial consumption ([4.96
× 0.5] × 0.5 = 1.24) and 2.34 mg C m
−3
d
−1
due to phytoplankton consumption ([9.35 × 0.5] × 0.5
= 2.34) (assuming 50% of ingestion is excreted or egested and becomes DOC). Thus, local
biological processes would create 12.8 mg C m
−3
d
−1
of DOC (2.01 + 2.48 + 4.68 + 1.24 + 2.34 =
12.8), which is only 39% of the DOC required to support the BP:NPP ratio of 0.53 (12.8/33 =
39%). Thus, ~60% of the organic matter in the system would have to come from external
sources to support a BP:NPP ratio of 0.53 at the SPOT station at steady-state conditions.

Abstract (if available)

Abstract Marine protists -- single-celled microbial eukaryotes -- are an extraordinarily diverse group of organisms that span a myriad of sizes, forms, and functions. These abundant organisms fulfill a wide array of ecological roles and are critical to the global cycling of key elements (e.g. C,N,P,S). Complex physical, chemical, and biological interactions structure these communities, which in turn sequester atmospheric carbon dioxide into biomass that is destined for higher trophic levels or export to the deep ocean. Understanding the abundances and trophic activities of microbes is therefore essential to accurately modeling carbon cycling. Significant attention has focused on the environmental factors controlling the production of phytoplankton and bacteria

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

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

PDF

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

PDF

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

PDF

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

PDF

Multi-dataset analysis of bacterial heterotrophic variability at the San Pedro Ocean Time-series (SPOT): an investigation into the necessity and feasibility of incorporating a dynamic bacterial c...

PDF

Genetic characterization of microbial eukaryotic diversity and metabolic potential

PDF

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

PDF

Physiological and ecological consequences of environmental temperature on Antarctic protists

PDF

Patterns of molecular microbial activity across time and biomes

PDF

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

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

The impact of mesoscale and submesoscale physical processes on phytoplankton biomass, community composition, and carbon dynamics in the oligotrophic ocean

PDF

The dynamic regulation of DMSP production in marine phytoplankton

PDF

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

PDF

Unexplored microbial communities in marine sediment porewater

PDF

Molecular ecology of marine cyanobacteria: microbial assemblages as units of natural selection

PDF

The impact of Lagrangian environmental variability on the growth of phytoplankton

PDF

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

Asset Metadata

Creator Connell, Paige Elizabeth (author)

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

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Biology

Degree Conferral Date 2017-12

Publication Date 09/26/2017

Defense Date 07/14/2017

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

Tag bacterivory,herbivory,marine food webs,microbial ecology,microzooplankton,OAI-PMH Harvest,plankton,protist,protistan grazing

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Caron, David A. (committee chair), Fuhrman, Jed A. (committee member), John, Seth G. (committee member), Levine, Naomi M. (committee member)

Creator Email paigeeconn@gmail.com,pconnell@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-437345

Unique identifier UC11263925

Identifier etd-ConnellPai-5782.pdf (filename),usctheses-c40-437345 (legacy record id)

Legacy Identifier etd-ConnellPai-5782.pdf

Dmrecord 437345

Document Type Dissertation

Rights Connell, Paige Elizabeth

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