University of Southern California Dissertations and Theses

Photophysiological parameters for CO2 and N2 fixation in trichodesmium spp. in natural populations and culture nutrient limitation experiments

(USC Thesis Other)

Photophysiological parameters for CO2 and N2 fixation in trichodesmium spp. in natural populations and culture nutrient limitation experiments

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PHOTOPHYSIOLOGICAL PARAMETERS FOR CO
2
AND N
2
FIXATION OF

TRICHODESMIUM SPP. IN NATURAL POPULATIONS AND IN CULTURE

NUTRIENT LIMITATION EXPERIMENTS

by

Juliette Anne Finzi

____________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(OCEAN SCIENCES)

August 2007

Copyright 2007 Juliette Anne Finzi
ii
Dedication

Pour Maman, Papa et notre Thalassa.
iii
Acknowledgments

So many people have helped me convert from art historian to biological
oceanographer…thank you to all of you. But, there are a special few, who deserve a
bit more recognition…
Of course, I wish to thank my mother and father for supporting me and
encouraging me to pursue this dream. I kept at it for all these years to make you proud.
Thank you for helping me stay on the path. I wish to thank my brother, Adrien, for
helping me with all the statistics, for reviewing this dissertation at the 11th hour, and,
most importantly, for inspiring me to become a scientist…just like him! And, I wish
to thank my other brother, Michel, for having been the first to have inspired in me the
love of the ocean. I still remember wandering around the tide pools at LIU when you
took that summer course and I still have that Invert Zoo textbook…I guess I still just
want to be just like my big brothers!
Thank you to Michael. You met me on the first day this madness began, stuck
with me through it all, and even married me! Thank you for still being here and for
being so kind. And, of course, thank you for teaching me how to use Illustrator, for
reminding me why I used whatever conversion factor however many years before, and
for always forcing me to question what I was doing and why I was doing it. I did this
with you and by you. Olive juice!
To the trifecta -- Jay Burns, Rachel Foster and Troy Gunderson. Rache - from
that first holographic lime at Chili's, while talking endlessly about how cool
zooplankton are…you are now one of my dearest friends. Through watching you
never give up and always believe in what you were studying, I began to understand
iv
what a true scientist is. You. And, in watching you always be kind, even when on a
tin can in the middle of the ocean, I learned what a true friend is. You. Rache, you're
my hero. Jay - you taught me literally everything I needed to know to figure out how
to do this stuff. How many times have I gone to the well? And, even though we
haven't seen each other in a year…seems I'm still turning to you to answer those
burning science questions! Thanks for showing me that it's always worth trying to fit
it in the bottle. And, Troy, thank you for always, always, being there to help me build
some odd contraption that I'd never use, for running my samples first, for always being
ready to go to lunch, for being the first to dance on the dancefloor, and for being one
of my favoritist friends - ever. You all made those cruises memorable and bearable.
Thank you to my officemates -- Isabel Romero and Lia Protopapadakis. It was
always good to have you two to turn to. And, thank you to the rest of the Capone Lab.
Sarah Govil and Reni Schimmoeller -- you made rad work fun and were great friends
along the way…as you both still are.
Thank you to those who helped me with science questions throughout the years:
Raleigh Hood, John Cullen, Sergio Sañudo-Wilhelmy, Jennifer Pett-Ridge, and Peter
Weber.
As a woman in science, it's most important to have good female mentorship.
Thank you to Linda Duguay, Phyllis Grifman, Jennifer Wolch and Stephanie Pincetl.
Thank you also to the SPO folks - Dan Farrow, Tom Culliton, Brent Ache and Lani
Watson (among others, of course) for supporting me as I tried to finish this

v
dissertation up, while working with you guys. Your understanding made this all
possible.
And, finally, thank you to my committee, Doug Capone, Doug Hammond, Ajit
Subramaniam, Dale Keifer and Tony Michaels. Thanks for making this process a lot
less painful than I thought it would be and thank you for all your help, advice, and just
for being a bunch of really good guys! Thank you!
vi
Table of Contents

Dedication ii

Acknowledgments iii

List of Tables vii

List of Figures viii

Abstract x

Chapter 1: Literature Review on Trichodesmium - The Enigmatic 1
N
2
-fixing Cyanobacteria

Chapter 2: Photophysiological Parameters in Natural Populations 18
of Trichodesmium in the Tropical N. Atlantic and
N. Pacific

Chapter 3: The Effect of Fe- and P-Limitation on the 58
Photophysiological Parameters of P
max
and α in
Trichodesmium IMS-101

Chapter 4: Temporal Segregation of CO
2
and N
2
Fixation in 86
Trichodesmium IMS-101 Using Nanometer Resolution
Secondary Ion Mass Spectrometry (nanoSIMS)

Chapter 5: Global Significance, Contributions and Synthesis 101

Bibliography 106

Appendices 119
Appendix A: Correlation plots of P
max
and α and [DFE] 119
[SRP] and MLD for CO
2
and N
2
fixation
Appendix B: Comments on Methodology 138
vii
List of Tables

Table 2.1: Previous Reported Values of P
max
and α for 22
Trichodesmium

Table 2.2: Station Characterization 28

Table 2.3: Basic Physiological Characteristics of 30
Trichodesmium

Table 2.4: P
max
and α Values for CO
2
fixation 31

Table 2.5: P
max
and α Values for N
2
fixation 33

Table 2.6: Units Table 35

Table 2.7: Carbon Fixed to Nitrogen Fixed Ratios 39

Table 2.8: Correlation Coefficients for CO
2
and N
2
Fixation 40
Parameters

Table 2.9: Comparison of Previous Values 43

Table 2.10: Comparison of Photosynthetron and On-Deck 45
Experiments

Table 3.1: Chl a cell
-1
measurements for the P and Fe 68
Manipulation Experiments

Table 3.2: P
max
and α for P Manipulation Experiments 72

Table 3.3: P
max
and α for Fe Manipulation Experiments 75

Table 3.4: Comparison of Results from Previous Study 79

Table 3.5: Differences Between High and Low Values 82

viii
List of Figures

Figure 2.1: Cruise charts for the N. Atlantic and N. Pacfic 24

Figure 2.2: Examples of P-E Curves for CO
2
and N
2
fixation

37

Figure 2.3: Examples of Curves from MP9 38

Figure 2.4: Concurrent On-Deck and Photosynthetron 46
Experiments

Figure 3.1: Curve fits for P Manipulation Experiments 70

Figure 3.2: Curve fits for Fe Manipulation Experiments 73

Figure 3.3: Dose-response Curve for P
chl
max
in Relation to Fe 84
Concentrations

Figure 4.1: Composite TEM and nanoSIMS Images 91

Figure 4.2:
15
N/
14
N Hotspots With Depth Through a Cell 92

Figure 4.3: Uptake of
15
N and
13
C for Individual Cells 93

Figure 4.4: Average rates of
15
N and
13
C Uptake Measured 95
by nanoSIMS and IRMS

Figure 4.5: SEM Images Demonstrating Burning-In Heterogeneity 100

Figure A.1: Regression Plots of P
max
and α for 120
CO
2
Fixation Against [DFe]

Figure A.2: Regression Plots of P
max
and α for 123
N
2
Fixation Against [DFe]

Figure A.3: Regression Plots of P
max
and α for 126
CO
2
Fixation Against [SRP]

Figure A.4: Regression Plots of P
max
and α for 129
N
2
Fixation Against [SRP]

ix
Figure A.5: Regression Plots of P
max
and α for 132
CO
2
Fixation Against MLD

Figure A.6: Regression Plots of P
max
and α for 135
N
2
Fixation Against MLD
x

Abstract

Trichodesmium is a globally significant cyanobacterium that plays an
important role in marine carbon (C) and nitrogen (N) cycling. Accordingly, global
ecosystem have begun to incorporate the critical role of diazotrophs, and
Trichodesmium specifically. Most commonly, this is done by incorporating the
photophysiological parameters P
max
(the rate at

which production is maximal) and
α (the linear relationship between production and light at low irradiance), which are
derived from production vs. irradiance (P-E) curves. Results of experiments in which
P
max
and α for CO
2
and N
2
fixation by Trichodesmium were measured in the tropical
N. Atlantic and tropical N. Pacific are presented. P
max
and α for CO
2
fixation were not
significantly different between basins, but they were reduced in the Atlantic compared
to the Pacific for N
2
fixation. It is hypothesized that differences in light history and
nutritional status contribute to these basin differences. Additionally, laboratory
culture experiments were conducted to elucidate the effect of phosphorus (P) and iron
(Fe) limitation on P
max
and α for CO
2
and N
2
fixation. All parameters were both
reduced under nutrient limiting conditions. Finally, Trichodesmium is capable of
fixing both CO
2
and N
2
concurrently throughout the day, but it does not contain
heterocysts. Nitrogenase, the key enzyme in N
2
fixation, is inhibited by the presence
of O
2
- a significant by-product from photosynthesis. Therefore, it still remains
unclear how these two functions co-occur. Hypotheses indicate that there may be
spatial and temporal segregation of the two processes. Using nanometer resolution
secondary ion mass spectrometry (nanoSIMS), uptake of
15
N and
13
C through N
2
and
xi
CO
2
fixation, were traced and measured throughout individual cells along a
Trichodesmium trichome at multiple time points throughout the day. CO
2
fixation
rates were maximal in the morning, while N
2
fixation rates were maximal in the early
afternoon indicating that there is indeed a temporal segregation between CO
2
and N
2

fixation. It could not be determined if there is also a spatial segregation of the two
processes.

CHAPTER 1
Literature Review on Trichodesmium - The Enigmatic N
2
-fixing Cyanobacteria

Global Significance of Trichodesmium
Trichodesmium is a nonheterocystous, diazotrophic, photosynthetic, marine
cyanobacterium that is found in tropical and subtropical seas (Carpenter and Romans
1991). It plays a critical role in marine carbon (C) and nitrogen (N) cycling. An early
analysis of N
2
fixation by Trichodesmium underestimated the contribution of
Trichodesmium to the marine N cycle as it relied upon historical records of
Trichodesmium that may have systematically underestimated population densities
(Carpenter 1999). Recent estimates, however, report that it contributes about 5.7 Tmol
new N y
-1
globally - a contribution that can be comparable to the rate of NO
3
-
that
diffuses from depth to the euphotic zone in oligotrophic oceanic ecosystems (Capone
et al. 2005). While it has recently become clear that Trichodesmium is not the sole or
predominant agent of oceanic marine N
2
fixation (Mague et al. 1977, Villareal 1991,
1992, Carpenter 1999, Zehr et al. 2001, Montoya et al. 2004, Falcon et al. 2004),
accumulating field evidence over the last decade clearly indicates that it does
contribute substantially to this process. Trichodesmium is therefore a globally
significant marine cyanobacterium (Capone et al. 1997).

Basic Ecology and Physiology of Trichodesmium
Trichodesmium grows best in waters ranging from 25 to 30 (°C) (La Roche
and Breitbarth 2005). Gas vesicles within Trichodesmium cells tend to keep these
2
organisms in the upper 50m of the water column, with biomass maxima located
between the surface and 20m depending on wind speed (Carpenter and Romans 1991).
They can form thick blooms that are visible on the ocean surface and from satellites
(Capone et al. 1998, Subramaniam et al. 1999b, Subramaniam et al. 2002, Westberry
et al. 2005). Because they reside primarily in the surface waters, Trichodesmium are
adapted to high light regimes (Capone et al. 1997, Subramaniam et al. 1999a,
discussed more extensively below).
In the North Atlantic, the most common species is Trichodesmium thiebautii,
which occurs predominantly as macroscopic aggregates composed of from 100 to 200
trichomes (Carpenter 2004), with densities ranging from about 10 to over 10,000
trichomes per liter (Capone 1983, Tyrrell 2003). As colonies, they form fusiform tuft
and spherical puff shaped colonies. In the N. Pacific, both T. thiebautii and
Trichodesmium erythraeum are common, and are commonly observed as free
filaments. Although self-shading occurs in the colonial morphology (Subramaniam et
al. 1999), it is unclear if either morphology is more efficient at CO
2
or N
2
fixation over
the other.
Trichodesmium has a slow growth rate with a reported doubling time of 5.2
days based on carbon (C) and 37.5 days based on nitrogen (N) (Capone et al. 1994).
Despite this slow growth rate, it is able to remain competitive in the oligotrophic seas
because it can fix N
2
gas, converting it to more bioavailable forms of N. It therefore
has the advantage over other organisms present in these seas that are N-limited. Yet,
between 30-50% of the newly fixed N can leak to the external environment (Glibert
3
and Bronk 1994, Capone et al. 1994, Mulholland et al. 2004) thus providing a vital
source of N to the environment. Trichodesmium is therefore a critical link in the
cycling of C and N in the oligotrophic seas.
Several studies have demonstrated an endogenous diel cycle in N
2
fixation of
Trichodesmium (Capone et al. 1990, Zehr et al. 1993, Wyman et al. 1995, Chen et al.
1998). The NifH transcripts that code for nitrogenase - the key structural enzyme in
N
2
fixation - are formed a few hours before sunrise (~ 4 am) daily. The levels increase
throughout the day, start to decrease in the afternoon and disappear by dusk (Wyman
et al. 1995, Chen et al.1998). The iron (Fe) protein of the nitrogenase enzyme has
been shown to alternate between an active, unmodified state during the day and
inactive, modified state during the night. It has been suggested that modification of the
nitrogenase enzyme may be dependent on internal nitrogen pools – perhaps within the
glutamine and glutamate pools whose concentrations follow the pattern of nitrogenase
activity (Zehr et al. 1993).
Though we understand quite a bit about the physiology and processes of
Trichodesmium, one great mystery remains. Trichodesmium fixes both CO
2
and N
2

concurrently. Nitrogenase, however, is oxygen (O
2
) sensitive – it is immediately
inhibited in its presence. Other cyanobacteria have therefore developed mechanisms
with which to segregate the two processes. Some spatially segregate nitrogenase from
O
2
by sequestering the N
2
fixation in specialized cells called heterocysts which have
thick cell walls, lack photosystem II (PSII), have fewer phycobiliproteins and have
decreased rubisco activity (Bergman and Carpenter 1991, Siddiqui and Carpenter
4
1992). Bergman et al. (1997) and Wolk et al. (1994) review several mechanisms by
which reductant and energy can be delivered to the nitrogenase within the heterocysts.
From photosynthesis, NAPDH generated by the oxidative pentose phosphate pathway
or by the Entner-Doudoroff pathway may supply the reductant necessary to reduce
ferredoxin. Non-photosynthetically, NADPH-dependent isocitrate dehydrogenase and
pyruvate:Fd oxidoreductase could also generate reduced ferredoxin. The reduced
ferredoxin can then donate electrons to nitrogenase. The energy required by
nitrogenase, in the form of ATP, may be provided by cyclic or linear
photophosphorylation by the PSI within the heterocyst. Other cyanobacteria segregate
N
2
fixation and photosynthesis temporally. They photosynthesize during the day and
fix N
2
at night, thus protecting the nitrogenase by separating the two processes
temporally. Photosynthate produced during the day is stored and is oxidized at night to
support N
2
fixation.
Trichodesmium, however, is unique in that it lacks heterocysts, but fixes both
CO
2
and N
2
during the day (Saino and Hattori 1978). How this is accomplished
remains as yet unresolved. Early hypotheses suggested that there are anoxic
microzones within the colonies in which the N
2
fixation occurs. Carpenter and Price
(1976) used microautoradiography to characterize CO
2
fixation in Trichodesmium and
they argued that Trichodesmium colonies must be intact in order for N
2
fixation to
occur. They found that disrupted cells demonstrated decreased N
2
fixation under oxic
conditions whereas they did not demonstrate decreased N
2
fixation under anoxic
conditions. Paerl (1994) conducted more microautoradiographic experiments and
5
found that some regions are devoid of photosynthesis. He argued that these cells are
low in pO
2
and may be sites of N
2
fixation. In another study, Kana (1993) reported
high rates of respiration in Trichodesmium colonies primarily due to enhanced Mehler
activity. Trichodesmium's increased respiration may also serve to decrease internal
[O
2
] thereby protecting nitrogenase.
Another hypothesis suggests that nitrogenase may be located in only a few
cells. Paerl et al. (1989) were the first to examine this by using immunogold labeling
to characterize nitrogenase abundance in Trichodesmium colonies. They found
nitrogenase dispersed throughout the cells and colonies. Bergman and Carpenter
(1991) repeated these experiments with colonies collected from waters southeast of
Bermuda. They instead found nitrogenase limited to only a few cells clustered along a
trichome (with ~100- 200 cells/trichome) and nitrogenase was located in electron
dense cytoplasm distributed throughout those cells. They found increasing amounts of
nitrogenase labeling as one moved towards the center of the colony. These findings
seemed to corroborate the contemporaneous argument that there are microzones of N
2

fixation within colonies. Bergman et al. (1996) combined nitrogenase labeling and
microautoradiography and found that 87% of the cells demonstrated
14
C uptake and
these cells did not contain nitrogenase. Moreover, Fredriksson and Bergman (1995)
found a daily fluctuation of nitrogenase with a minimum in nitrogenase labeling in the
morning (at 0600 hours) and in the afternoon (at 1600 hours) and a maximum in
labeling at midday (1200 hours). They argued that the variation in nitrogenase
6
synthesis leads to daily fluctuations of N
2
fixation with maximum N
2
fixation
occurring at midday.
Fredriksson and Bergman (1997) reported morphological differences in
nitrogenase labeled cells. They found that labeled cells had a denser appearance than
non-labeled cells, they have a tighter thylakoid network, less extensive gas vacuoles
and smaller vacuole-like spaces between the thylakoids. They termed the nitrogenase
labeled sites as "diazocytes" defining them as "structurally modified nitrogenase
containing cells, capable of cell division, but lacking additional cell envelopes." Lin et
al. (1998) used immunofluorescence labeling on Trichodesmium and found similar
results. They reported that ~10% only of the cells in a trichome express nitrogenase
and that the labeled cells were clustered at the center of the colony. They further
corroborated Fedriksson and Bergman's (1997) work by noting that nitrogenase
containing cells had a denser thylakoid network, smaller cyanophycin granules and
found the nitrogenase-labeled cells to be shorter than non-nitrogenase labeled cells.
They did, however, find that the nitrogenase cells did have some CO
2
uptake activity -
although it was reduced compared to activity in the non-nitrogenase containing cells.
Finally, Berman-Frank et al. (2001) used Fast Repetition Rate Fluorescence
(FRRF), O
2
production,
14
CO
2
uptake and acetylene reduction assays (for N
2
fixation)
and found a diel pattern in CO
2
and N
2
fixation. They argued that CO
2
fixation
increases in morning and decreases at midday as N
2
fixation activity increases. This
was observed in field and lab samples. This study implies therefore that there may be
a temporal segregation between CO
2
and N
2
fixation.
7
Thus, current studies suggest that N
2
and CO
2
fixation in Trichodesmium may
be both spatially and temporally segregated. In Chapter 4, I describe experiments
using the stable isotopes
15
N and
13
C to trace CO
2
and N
2
fixation over eight discrete
time points. Using nanometer resolution secondary ion mass spectrometry
(nanoSIMS), I was able to observe the location of these isotopes within a cell and to
determine uptake rates. I was not able to confirm spatial segregation of the two
processes but did observe temporal segregation.

Photophysiology of Trichodesmium
The major light harvesting antenna complexes in cyanobacteria are the
phycobilisome structures (PBS) comprised of the phycobiliproteins, phycocyanobilin
(PCB), phycobiliviolin (PVB), phycourobilin (PUB) and phycoerythrobilin (PEB)
(Grossman 1994, Lantoine 1999, Glazer 1999). Cyanobacteria can live in a dynamic
marine environment with fluctuations in light intensities (changing with weather and
depth in the water column) and light quality (changing mainly with depth). They can
exhibit plasticity in their pigment composition and concentration and are able to adjust
their pigment concentrations in response to these variable conditions. For instance,
Kana & Glibert (1987) demonstrated that Synechococcus isolate WH7803 adjusted
their PE:PC ratios in response to different irradiances. Similarly, Foy and Gibson
(1982b) found that the freshwater cyanobacteria Oscillatoria redekei increased their
PC:chl a ratios under low light.
8
In Trichodesmium, specifically, Shimura & Fujita (1975) showed that PUB and
PEB were active in trapping light energy for photosynthesis. Subramaniam et al.
(1999) found that Trichodesmium alter their pigment concentrations in response to
changing irradiance. Notably, Trichodesmium have high PUB:PEB ratios that have
their maxima at midday. Normally, the increased levels of PUB (that absorb at blue
light wavelengths) ensure that cyanobacteria are able to take advantage of the
primarily blue light found at depth (Shimura 1975). Trichodesmium have been found
as deep as 120m and can migrate through the water column (Villareal 1990, Villareal
and Carpenter 2003). However, Trichodesmium also have gas vesicles that largely
maintain the organisms in the upper 50m of the water column with biomass maxima
located between the surface and 20m depth (Villareal 1990, Romans 1994, Letelier
1996). Trichodesmium form extensive blooms in which the organisms remain at the
surface for extended periods of time; and, despite this prolonged exposure at the sea
surface, they are still able to fix N
2
(Capone 1998). Subramaniam et al. (1999)
hypothesize that, apart from aiding in light acquisition at depth, the PUB may act as a
photoprotectant at high light, allowing Trichodesmium to be high light adapted.

Production vs. Irradiance (P-E) in Trichodesmium
CO
2
and N
2
fixation in Trichodesmium are both dependent on light availability.
This relationship between CO
2
or N
2
fixation and light can be defined through
experimentally derived CO
2
and N
2
fixation versus irradiance relationships (P-E
curves).
9
Blackman (1905) first described the relationship between CO
2
fixation and
light, in which he suggested that there is a linear relationship in phytoplankton
production with light intensity until a point at which production stabilizes. Since then,
there have been a number of formulations based on this initial principle aimed at
refining the relationship between fixation and light (Smith 1936, Riley 1946, Webb et
al. 1974, 1976, Jassby and Platt 1976, Platt et al. 1977, Laws and Bannister 1980, Platt
et al. 1980). The formulations utilized in the experiments described in this
dissertation are those of Webb et al. (1976) and Platt et al. (1980):
(1)
!
P =P
s
(1"e
"#E/P
s
)e
"$E/P
s
(Platt et al. 1980)
(2)
!
P =P
s
(1"e
"#E/P
s
) (Webb et al. 1976)
where P
s
is defined as:
(3)
!
P
max
=P
s
"
(" +#)
$
%
&
'
(
)
#
(# +")
$
%
&
'
(
)
$
%
&
'
(
)
#
"
(Hood et al. 1991)
and where P
max
is the value at which production is maximal and the curve stabilizes, α
is the light utilization coefficient (mg C/N mg chl a
-1
h
-1
/µmol quanta m
-2
s
-1
), E is the
incident scalar irradiance (µmol quanta m
-2
s
-1
), P is the instantaneous production rate
(mg C/N mg chl a
-1
h
-1
) and β is a photoinhibition parameter (mg C/N mg chl a
-1
h
-
1
/µmol quanta m
-2
s
-1
). Alpha, the light utilization coefficient, is the product of the
optical absorption cross-section (a*, the average target area presented by chlorophyll
and accessory pigments for photon absorption) and quantum yield (φ, number of
atoms fixed per number of photons absorbed, Sakshaug et al.1997). It is therefore a
function of light-harvesting efficiency and photosynthetic energy conversion
10
efficiency (Falkowski et al. 1992, Henley 1994). All curves were first fit with Platt et
al. (1980). For those experiments where no photoinhibition was observed, the curves
were refit using Webb et al. (1976).
Several previous studies have characterized the photosynthetic (CO
2
)
parameters of Trichodesmium in the Caribbean and Sargasso Seas (Li 1980, Lewis
1988, Carpenter et al. 1993, Carpenter and Roenneberg 1995). These studies are
limited to the subtropical N. Atlantic. In Chapter 2, these previous studies are
discussed more extensively and I refine these estimates for the tropical N. Atlantic and
provide new estimates of P
max
and α for the tropical N. Pacific.
Recently, it has been demonstrated that N
2
fixation vs. irradiance curves can be
described by the same formulations as for CO
2
fixation vs. irradiance (Hood et al.
2002, Staal et al. 2002). Using data collected on a cruise in the subtropical N.
Atlantic, Hood et al. (2001) fit the equation of Platt et al. (1980) to N
2
fixation vs.
irradiance experiments conducted on Trichodesmium . These experiments are
described fully in Chapter 2.

The Effect of Light Limitation on the Photophysiological Parameters
In comparisons between light-limited and light-saturated cultures of
phytoplankton, low-light adapted cells have been demonstrated to have higher α than
cells exposed to higher irradiances (Falkowski 1994). Additionally, cells acclimated to
lower light intensity may become photoinhibited more readily than those adapted to
higher light (Falkowski 1994). Bouman et al. (2000) found that P
max
is indirectly
correlated to optical depth. Indeed, in examining the effect of low and high light on
11
twenty different freshwater strains of cyanobacteria, Foy and Gibson (1982)
demonstrated that all Oscillatoria species showed a uniform depression of P
max
under
low light. They also found evidence of increased α.
Though Trichodesmium has its biomass maxima at 20m, it has been argued
that Trichodesmium migrate through the water column due to ballasting of their
internal carbohydrate stores (Villareal 1990, Villareal and Carpenter 2003). In this
migration, Trichodesmium colonies may spend time at different depths, exposed to
variable light quantity and quality, and may adjust their pigment function and
concentrations, thereby altering their photophysiological parameters accordingly.

The Effect of Nutrient Limitation on the Photophysiological Parameters
The oligotrophic open ocean is, by definition, nutrient limited (Falkowski
1992, Falkowski 1997). There is continued debate between biologists and geochemists
as to whether N or P is the ultimate limiting nutrient for oceanic primary productivity
(Falkowski 1997, Tyrrell 1999). On the timescales of phytoplankton growth, oceanic
phytoplankton are generally thought to be N limited (Falkowski 1998, Tyrrell 1999).
On longer time scales, Tyrell (1999) concludes that P is the key limiting nutrient in the
ocean, as N
2
fixation can make up any potential N deficits. There is no analog to N
2

fixation for the P cycle: the only processes that add P to the ocean are weathering of P-
bearing strata off the continent and river discharge.
Diazotrophs are more successful than non-diazotrophs in oligotrophic regions
due to their ability to utilize the largest reservoir of N on the planet, N
2
gas, however
12
their growth is constrained by other factors, such as P- and Fe-limitation. The
evidence strongly supports the hypothesis that they are Fe-limited in the subtropical N.
Pacific where aeolian dust inputs, the primary distributor of Fe to the open ocean, is
very low (Falkowski 1997, Falkowski et al. 1998, Rueter et al. 2000). Although more
recently, Karl et al. (2001) have suggested a shift to a more P-limited environment due
to increased N
2
fixation during the last decade. The subtropical N. Atlantic receives
two to ten times higher Fe influx from aeolian dust deposition and also receives
tropical river run-off rich in micronutrients, including Fe. Based on cellular quotas,
Sañudo-Wilhelmy et al. (2001) argue that Trichodesmium in the central Atlantic are P-
limited rather than Fe-limited, specifically for N
2
fixation. Wu et al. (2000) compared
DNN (dissolved nitrate and nitrite) to DIP (dissolved inorganic phosphorus)
concentrations in the subtropical N. Atlantic and subtropical N. Pacific. They
measured DNN:DIP ratios of 20 - 32 in the Atlantic and less than 17 in the Pacific,
implying P limitation in the subtropical N. Atlantic. Equally, it has been demonstrated
that Trichodesmium can induce the enzyme alkaline phosphatase (PhoA) to cleave
inorganic phosphorus from dissolved organic phosphorus; and, that this induction is
monitored by internal inorganic P status. Thus increased PhoA activity can be an
indicator of P-stress in Trichodesmium cells. Dyrhman et al. (2002) used an enzyme-
labeled fluorescence method to monitor PhoA activity and found P-stressed
Trichodesmium cells in four stations in the western N. Atlantic.
Thus current studies indicate P-limitation in the subtropical N. Atlantic and Fe-
limitation in the subtropical N. Pacific. There are also indications of P and Fe co-
13
limitation. Mills et al. (2004) conducted nutrient addition experiments at three sites in
the subtropical N. Atlantic and examined bulk phytoplankton and diazotroph response
to additions of N, Fe, P and both Fe/P released from Saharan Dust (one of the sources
for the aeolian dust deposition over the N. Atlantic). CO
2
fixation was most enhanced
after additions of N and either Fe, P, or both Fe and P. In two of the three stations, N
2

fixation was only enhanced if both Fe and P were added together. That Fe was
necessary to enhance N
2
fixation contradicts the notion that the subtropical N. Atlantic
is primarily P-limited, and not Fe-limited. Mills et al. (2004) observe that the
experiments conducted by Sañudo-Wilhelmy (2001) were conducted during a time of
high dust deposition, while the experiments of Mills et al. (2004) were conducted
during a period of low dust deposition. Thus, which nutrient(s) is ultimately limiting
in the subtropical N. Atlantic may be seasonally, and depositionally, driven.
Nonetheless, Mills et al. (2004) provide evidence for co-limitation of the two
nutrients.
Kirk (2000) suggests that inorganic nutrition is the most important indirect
factor affecting P-E characteristics. For instance, it has been demonstrated that nutrient
stress can lower P
max
(Welschmeyer 1981, Falkowski 1992, Sakshaug 1997, Sosik
1996, Bouman 2000, Behrenfeld 2002). Equally, Sosik (1996) demonstrated that
distance from the nutricline correlates with a decrease in α. Most of the literature
focuses on the effects of inorganic N-limitation on P-E characteristics of eukaryotes.
Several studies have addressed the effect of Fe limitation on phytoplankton cells.
Behrenfeld (2002) found that Fe limitation causes a stoichiometric change in
14
photosystem II to photosystem I (PSII:PSI) ratios. Fe limitation may also slow down
turnover rates (Falkowski 1992). Both lower turnover rates and decreased PSII
efficiency lead to decreased overall efficiency, thus yielding lower P
max
values. There
have been far fewer studies that address P limitation (Kromkamp et al. 1987, Foy
1993). Kromkamp et al. (1987) compared light-limited and P-limited cyanobacterial
cultures of Microcsytis aeruginosa (a freshwater cyanobacterium). In their
comparisons, they found that the light harvesting capacity of the P-limited cultures
was reduced compared to the light-limited cultures, and that P
max
per unit chlorophyll
was the same between limitation experiments, although P
max
per unit protein was
reduced in the P-limited cultures. They concluded that M. aeruginosa adapted to P-
limitation by decreasing its light harvesting capacity. Thus, the effect of nutrient
limitation on the photophysiological parameters is fairly well documented in the
literature with respect to eukaryotes and non-diazotrophic cyanobacteria.
For Trichodesmium specifically, there have been a few papers that have
demonstrated reduced N
2
fixation in P-limited cultures. Fu and Bell (2003a) incubated
Trichodesmium GBRTRL101 in P-reduced media for 7 days. Cultures incubated in
the higher concentrations of P had higher rates of N
2
fixation. Mulholland and
Bernhard (2005) grew Trichodesmium IMS101 in continuous culture under P-replete
and P-deplete conditions and determined that cultures grown under P-replete
conditions fixed more N
2
than those grown in the depleted media.
There is a longer history of descriptive papers that demonstrate reduction in
CO
2
and N
2
fixation in Fe-limited cultures. Rueter (1988) incubated Trichodesmium
15
colonies collected from Barbados in Fe-enriched seawater ranging in concentration
from 10nM - 1µM and observed increased chlorophyll, N
2
fixation and, to a lesser
extent, CO
2
fixation, in the colonies incubated at the higher Fe concentrations. Paerl et
al. (1994) also incubated Trichodesmium colonies collected in waters off North
Carolina and the Caribbean in Fe-enriched seawater and found increases in CO
2
in
both populations and in N
2
fixation in the North Carolina waters. Rueter et al. (1990)
conducted experiments on laboratory cultures Trichodesmium NIBB 1067; those
cultures grown in higher concentrations of Fe had higher rates of both CO
2
and N
2

fixation. Berman-Frank et al. (2001) looked at the effect of Fe-limitation on laboratory
cultures of IMS101 and found increased N
2
fixation correlated with highest Fe-
availability. Finally, Fu and Bell (2003b) measured an 8-fold increase in N
2
fixation in
cultures grown in 450 nM Fe compared to those grown in 0 nM Fe media (0.002 vs.
0.014 fmol N cell
-1
h
-1
).
To my knowledge, there have been no previous studies that have examined the
effect of P-limitation on the CO
2
and N
2
fixation parameters P
max
and α by
Trichodesmium and only one that has examined the effect of Fe limitation on CO
2

fixation (Fu and Bell 2003b). They observed increased P
max
and α with increasing Fe
concentrations. In Chapter 3, I report results of experiments that examined the effect
of P- and Fe-limitation on CO
2
and N
2
fixation in cultures of Trichodesmium IMS101.
The results of my Fe experiments corroborate and refine the findings of Fu and Bell
(2003b).

16
Modeling of CO
2
and N
2
Fixation by Trichodesmium
Ecosystem modeling has become an ever more important means of
synthesizing our observations, testing our mechanistic understanding of systems, and
providing predictions of system behavior. There have been a number of recent efforts
to explicitly incorporate N
2
fixation in ecosystem ocean models in order to accurately
represent the N dynamics of particular marine ecosystems (Spitz et al. 2001, Hood et
al. 2001, 2004, Fennel et al. 2002, Moore et al. 2002a, 2002b, 2004, Coles et al 2004).
Rates of N
2
fixation are often incorporated into these models based on the two key
parameters, α and P
max
, which should be derived from empirical production
measurements. Of the models mentioned above, only Hood et al. (2001) utilized
production parameters derived from current empirical data. There is a current and
pressing need for accurate estimates of light and nutrient limited rates of
photosynthesis and nitrogen fixation.

Summary
The photophysiological parameters describe the fundamental physiological
response of the organism to light and have been related to the structure of the
photosystem and its response to differing light and nutrient regimes. In Chapter 2, I
describe the photophysiological parameters of field populations of Trichodesmium in
both the Atlantic and the Pacific. There are no significant differences in P
max
and α
for CO
2
fixation between basins, but P
max
and α for N
2
fixation are significantly
reduced in the Atlantic compared to the Pacific. It is hypothesized that nutritional
17
status and previous light history may contribute to these observed differences. In
Chapter 3, I report the results of studies addressing the effect of P- and Fe-limitation
on P
max
and α for CO
2
and N
2
fixation in laboratory cultured populations. These
experiments provide insight into the photophysiology of Trichodesmium and provide
values for parameters critical to effective ecosystem modeling. In Chapter 4, I
describe the results from nanoSIMS experiments in which I demonstrate a temporal
segregation of CO
2
and N
2
fixation within Trichodesmium colonies.
18

CHAPTER 2

Photophysiological Parameters in Natural Populations of Trichodesmium in the
Tropical N. Atlantic and N. Pacific

ABSTRACT
Trichodesmium is a significant contributor to global C and N cycling in
tropical and subtropical seas and it is an important component of global ecosystem
models. Many of these models utilize the photophysiological parameters P
max
and α
to incorporate CO
2
and N
2
fixation by Trichodesmium. Yet, to date, there have been
limited empirical estimates of these parameters, and of those, they are mainly limited
to descriptions of CO
2
fixation in the subtropical N. Atlantic. This study provides
photosynthetron-derived estimates of P
max
and α for CO
2
and N
2
fixation by
Trichodesmium in the tropical N. Atlantic and tropical and subtropical N. Pacific. For
CO
2
fixation, there was no significant difference in P
max
and α between the Atlantic
and the Pacific. However, for N
2
fixation, P
max
and α were significantly reduced in
the Atlantic compared to the Pacific. Soluble reactive phosphorus (SRP)
concentrations were significantly lower in the Atlantic stations compared to the Pacific
stations and mixed layer depths (MLD) were significantly deeper in the Atlantic
compared to the Pacific. It is hypothesized that these two factors may contribute to
19
the lower rates of N
2
fixation by Trichodesmium colonies in the Atlantic compared to
those in the Pacific.

INTRODUCTION
Trichodesmium is a non-heterocystous, diazotrophic, marine cyanobacterium
distributed ubiquitously in tropical and subtropical seas. It is capable of fixing both
CO
2
and N
2
concurrently during the day. Current reports estimate that Trichodesmium
contributes about 5.7 Tmol new N y
-1
(Capone et al. 2005) - a contribution that is at
times comparable to the rate of NO
3
-
that diffuses from depth in oligotrophic oceanic
ecosystems (Capone et al. 2005, Karl et al. 1997). Trichodesmium is thus now
recognized as a globally significant cyanobacterium (Capone et al. 1997, Carpenter
1999, Mahaffey et al. 2005). As such, CO
2
and N
2
fixation by Trichodesmium are
now key components in global ecosystem models (Spitz et al. 2001, Hood et al. 2001,
2002, 2004, Fennel et al. 2002, Moore et al. 2002a, 2002b, 2004, Coles et al. 2004).
Certainly Trichodesmium is not the sole significant N
2
fixing marine organism (Zehr
et al. 2001, Falcon et al. 2004, Montoya et al. 2004), but it is an important source of
new N. It is also the most conspicuous and there is a vast body of literature on
Trichodesmium from which the modelers can draw.
CO
2
fixation in photoautotrophic cyanobacteria is dependent on light
availability (Falkowski and Raven 1997). There have been a number of formulations
developed that have attempted to describe this relationship by describing the
production vs. irradiance (P-E) response (Smith 1936, Riley 1946, Jassby and Platt
20
1976, Platt et al. 1977, 1980, Laws and Bannister 1980, Webb et al. 1974, 1976,
among others); this is discussed in more detail in Chapter 1. The formulations used in
this study were those of Webb et al. (1976) and Platt et al. (1980) (equations presented
in methods). The main parameters derived from P-E curves are P
max
, the rate at which
production is maximal, and α, the slope which describes production at low light.
N
2
fixation in Trichodesmium is also dependent on light availability in that
light drives photosynthesis and the energy captured and reductant thus formed is
utilized in N
2
fixation. Hood et al. (2002), using Trichodesmium field data,
demonstrated that the Platt et al. (1980) formulation can also describe N
2
fixation vs.
irradiance and that these curves have a similar shape to those representing CO
2

fixation vs. irradiance. These parameters describe the fundamental physiological
response of the organism to light and have been related to the structure of the
photosystem and its response to differing light regimes.
Several studies have previously characterized the CO
2
P-E parameters of
Trichodesmium in the Caribbean and Sargasso Seas (Table 2.1, Li et al. 1980, Lewis et
al. 1988, Carpenter and Roenneberg 1993 and Carpenter et al. 1995). Staal et al.
(2002) provided estimates of these parameters for N
2
fixation in heterocystous,
diazotrophic cyanobacteria. However, only the study of Hood et al. (2002) described
the relationship between light and N
2
fixation specifically in Trichodesmium in the
subtropical N. Atlantic. They have provided the only currently published estimates of
α and P
max
for N
2
fixation by Trichodesmium.
21
Ocean ecosystem models that represent CO
2
and N
2
fixation rely on these
empirically derived relationships, notably by utilizing the parameters α and P
max
.
Therefore accurately characterizing α and P
max
for CO
2
and N
2
fixation in
Trichodesmium is critical for effective ecosystem modeling. This study provides
photosynthetron-derived estimates of P
max
and α for CO
2
and N
2
fixation in
Trichodesmium in the tropical N. Atlantic, and the first estimates of these parameters
for both CO
2
and N
2
fixation for Trichodesmium in the tropical N. Pacific.

METHODS
Sample collection:
Experiments were conducted on three cruises in the subtropical N. Atlantic
(MP1, MP3 and MP8, Figure 2.1a) and two cruises in the subtropical N. Pacific (MP6
and MP9, Figure 2.1b). Trichodesmium colonies were collected by gently towing a
202 µm mesh plankton net at 10-30 m depth behind the research vessels. Colonies
were handpicked out of the tow using plastic inoculating loops and were rinsed in 0.7
µm GF/F filtered seawater (FSW). Samples were set aside for both NaH
14
CO
3
-

uptake (
14
C uptake) (Steemann-Nielson 1952) and acetylene reduction (AR) analysis
(Capone and Montoya 2001) as described below.

22
Paper Location
P
chl
max
!
chl
CO
2
Fixation
mg C (mg chl a)
-1
h
-1
mg C (mg chl a)
-1
h
-1
/
µmol quanta m
-2
s
-1
Lewis et al. (1980) Eastern Caribbean Sea ~3.8 -
Li et al. (1988) Western Sargasso Sea 1.3 - 6.0 0.004 - 0.011
Carpenter et al. (1993)
Bahama Islands & Eastern Caribbean
Sea
11.1 - 12.6 0.020 - 0.032
Carpenter and
Roenneberg (1995)
Barbados, Sargasso Sea, Eastern
Caribbean Sea
1.9 - 13.4 0.017 - 0.041
N
2
Fixation
µg N (mg chl a)
-1
h
-1
µg N (mg chl a)
-1
h
-1
/
µmol quanta m
-2
s
-1
Hood et al. (2002) Tropical N. Atlantic 9 - 140 ~2.3
Table 2.1. This table lists the previously reported values of P
max
and α for field populations of Trichodesmium
for both CO
2
and N
2
fixation.

23
Production-Irradiance (P-E) Experiments:
Depending on biomass availability, 5-10 colonies were picked into 20 ml glass
scintillation vials pre-filled with 10 ml Whatman GF/F filtered seawater (FSW)
inoculated with 10 µl EDTA (Burns et al. 2006). For
14
C uptake experiments, 2 µCi
NaH
14
CO
3
-
(~50 µl by volume) was added to each sample. Samples were incubated in
the photosynthetron for 2-3 hours at 27°C. After incubation, samples were placed on
ice in the dark to halt
14
C uptake. The samples were filtered immediately onto 0.2 µm
polycarbonate filters to measure particulate organic carbon (POC). To ensure that all
excess inorganic
14
C was removed, 150 µl of 5% HCl was added to the filters, which
were then left to aerate for 24 hours. The filtrate was also collected to measure any
dissolved organic carbon (DOC) that may have leaked from the cells during the
cooling process. 250 µl of 10% HCl was added to the filtrate and left to aerate for 24
hours. After full aeration, scintillation fluid cocktail was added and samples were
counted on a Beckman LS-6500 scintillation counter.
For N
2
fixation estimates, 5-10 Trichodesmium colonies were picked into 14
ml serum vials with 10 ml GF/F FSW and 10 µl EDTA. The vials were crimp-sealed
and 1 ml of C
2
H
2
was injected into each sample through a silicone rubber stopper.
Samples were incubated in the photosynthetron for 2 - 3 hours at 27°C. After
incubation, 100 µl of 1M NaOH was added to the samples to stop any further AR.
Samples were measured on an FID Shimadzu GC-9A gas chromatograph.
The photosynthetrons were bottom-lit with high intensity halogen lights and
irradiance was varied using grey neutral density filters (Hood et al. 1991, 1995). Light
24
A
B
Figure 2.1a-b. Cruise charts indicating locations of stations in a) the North
Atlantic (MP1, MP3 and MP8) and b) the North Pacific (MP6, MP9).

25
intensity was measured in each of the 21 wells at the start of every experiment using a
Biospherics Model QSL-100 PAR sensor. Water was circulated through the
incubation chamber to maintain the cultures at a constant temperature of 27°C.

Curve fitting:
Two different equations were used to fit the data:
(1)
!
P =P
s
(1"e
"#E/P
s
)e
"$E/P
s
(Platt et al. 1980)
(2)
!
P =P
s
(1"e
"#E/P
s
) (Webb et al. 1976)
where P
s
is defined as:
(3)
!
P
max
=P
s
"
(" +#)
$
%
&
'
(
)
#
(# +")
$
%
&
'
(
)
$
%
&
'
(
)
#
"
(Hood et al. 1991)

and P
max
(mg C/N mg chl a
-1
h
-1
) is the value at which production is maximal and the
curve reaches an asymptote, α (mg C/N mg chl a
-1
h
-1
/µmol quanta m
-2
s
-1
) describes
the linear P-E relationship at low light, E (µmol quanta m
-2
s
-1
) is the incident scalar
irradiance, P (mg C/N mg chl a
-1
h
-1
) is the instantaneous production rate and β (mg
C/N mg chl a
-1
h
-1
/µmol quanta m
-2
s
-1
) is a photoinhibition parameter. An additional
parameter derived from this equation, E
k
(µmol quanta m
-2
s
-1
), the light saturation
parameter, describes the light level at which photosynthesis begins to become
saturated by the incident irradiance. It is defined as P
max
/α.

26
Statistical analysis:
Statistical analysis was conducted using Kaleidagraph version 3.6.2. One-way
ANOVA was used to analyze variance between parameters. Differences were
considered significant at p < 0.05. Regression analysis was utilized to determine
correlation, described by the correlation coefficient, r
2
. To determine the quantitative
significance of r
2
, the probability (Prob
N
(r≥ r
0
) or Prob
N
) that N measurements of
two uncorrelated variables will give a coefficient r larger than any particular r
0
. Prob
N

< 5% denotes significant correlation and Prob
N
<1% denotes highly significant
correlation (Taylor 1982). Finally, coefficient of variance (C
v
):
!
C
v
=
"
µ

where σ is the standard deviation and µ is the mean, was utilized to examine the
variance within a dataset.

RESULTS
Location and basic water body characteristics for each of the stations in this study are
presented in Table 2.2. The Atlantic cruises focused on sampling in and around the
Amazon River outflow plume. Waters were classified as either "plume" (salinity
below 34.8) or "oceanic" (salinity equal to or above 34.8). These stations are
indicated by asterisks in Table 2.2. The mixed layer depths (MLD) at these stations
were significantly more shallow than those at the oceanic stations with means values
of 11 ± 4 m within the plume waters compared to 68 ± 6 m outside of the plume.
MLDs in the Atlantic cruise stations were also significantly deeper than the Pacific
27
stations, which averaged 52 ± 2 m. Concentrations of soluble reactive phosphorous
(SRP) and dissolved iron (DFe) varied considerably between stations with [SRP]
ranging from 0.008 - 0.054 µM in the Atlantic and 0.002 - 0.081 µM in the Pacific.
Similarly, [DFe] ranged from 0.09 - 3.34 nM (only samples from the Atlantic cruises
were available at the time of publication).
Basic physiological characteristics of field Trichodesmium populations are
presented in Table 2.3. It should be noted that these are means for all samples
analyzed on the cruise and are not limited to the stations where photosynthetron
experiments were conducted. All characteristics varied significantly between stations
and ocean basins with coefficients of variance (C
v
) ranging from 28% - 103%. Chl a
content per trichome was the most variable characteristic, ranging from 0.019 - 0.490
ng chl trichome
-1
in the Atlantic and 0.050 - 0.466 ng chl trichome
-1
in the Pacific. In
the Atlantic cruises, chl a trichome
-1
was highly significantly reduced at the plume
stations (n=21) compared to the oceanic stations (n=19) with mean values of 0.056 ±
0.005 ng chl a trichome
-1
vs. 0.153 ± 0.035 ng chl a trichome
-1
(one-way ANOVA, p <
0.001), respectively (data not shown). On average, the number of trichomes per
colony in the Atlantic Trichodesmium samples was greater than in the Pacific samples
(187 ± 18 vs. 100 ± 10 trichomes per colony). Conversely, carbon content per
trichome and chl a content per trichome were higher in the Pacific samples (0.056
±0.011vs. 0.022 ± 0.002 µg C trichome
-1
and 0.176 ± 0.031 vs. 0.122 ± 0.025 ng chl a

28
cruise station location date latitude longitude [SRP] µM [DFe] nM MLD (m)
23 tropical N. Atlantic 2-Feb-01 9 18.08N 49 36.07W 0.013 0.13 70
24 tropical N. Atlantic 3-Feb-01 9 34.98N 47 51.62W 0.020 0.14 58
25 tropical N. Atlantic 4-Feb-01 7 46.04N 48 39.28W 0.011 0.09 77
37 tropical N. Atlantic 9-Feb-01 9 18.83N 42 00.97W 0.054 0.42 86
41 tropical N. Atlantic 11-Feb-01 10 01.29N 44 56.92W 0.035 0.32 53
44 tropical N. Atlantic 14-Feb-01 9 42.08N 49 35.42W 0.018 0.27 60
49 tropical N. Atlantic 17-Feb-01 10 59.69N 56 16.76W 0.030 1.30 106
51 tropical N. Atlantic 18-Feb-01 11 29.91N 55 00.09W 0.026 0.22 68
17 tropical N. Atlantic 10-Jul-01 11 10.20N 50 55.10W - - 84
MP3 48* tropical N. Atlantic 9-Aug-01 11 40.50N 54 34.00W 0.008 0.54 6*
49* tropical N. Atlantic 10-Aug-01 10 01.37N 5 558.44W 0.016 - 10*
30* tropical N. Atlantic 11-May-03 10 38.74N 54 16.55W 0.039 3.34 18*
34 tropical N. Atlantic 13-May-03 10 33.44 N 50 04.37W 0.028 1.99 38
38 tropical N. Atlantic 15-May-03 8 00.14N 48 29.45W 0.025 1.23 50
3 subtropical N. Pacific 24-Sep-02 23 29.99N 156 59.89W 0.008 - 58
3b subtropical N. Pacific 25-Sep-02 23 29.99N 156 59.89W 0.008 - 58
5 subtropical N. Pacific 27-Sep-02 25 00.02N 154 29.96W 0.007 - 60
7 subtropical N. Pacific 29-Sep-02 23 44.15N 156 21.31W 0.002 - 57
8 subtropical N. Pacific 1-Oct-02 22 44.06N 157 59.80W 0.048 - 61
14 tropical N. Pacific 7-Aug-03 18 30.01N 157 00.00W 0.081 - 41
16 tropical N. Pacific 10-Aug-03 19 29.59N 157 00.10W 0.071 - 61
17 tropical N. Pacific 12-Aug-03 19 32.14N 158 58.75W 0.045 - 46
18 tropical N. Pacific 14-Aug-03 20 32.11N 160 59.87W 0.038 - 47
19 tropical N. Pacific 15-Aug-03 21 00.78N 158 59.70W 0.058 - -
20 tropical N. Pacific 17-Aug-03 20 14.46N 156 30.01W 0.059 - 37
21 tropical N. Pacific 18-Aug-03 20 15.18N 159 11.32W 0.031 - 48
22 tropical N. Pacific 19-Aug-03 19 06.72N 162 13.06W 0.026 - 50
MP1
MP8
MP6
MP9
Table 2.2. Station characterization including latitude and longitude, soluble reactive phosphorus concentration
[SRP], dissolved iron concentrations [DFe], and mixed layer depth. Stations which were located within the Amazon
River outflow plume are indicated with an asterisk (*).

29
trichome
-1
). Yet, chl a:C content in the colonies was significantly greater in the
Atlantic than in the Pacific (0.006 ± 0.001 vs. 0.003 ± 0.001, one-way ANOVA, p <
0.001).
P
max
and α values for CO
2
and N
2
fixation normalized to mg chl a (P
chl
max
and
α
chl
), mg C (P
c
max
and α
c
)

and trichome (P
b
max
and α
b
) for each station and cruise
means are presented in Tables 2.4 and 2.5 and units are presented in Table 2.6. For
CO
2
fixation, P
max
values were not significantly different between ocean basins (Table
2.4) regardless of which normalization factor was used. However, in the Atlantic,
P
chl
max
was significantly lower on MP1 compared to MP3 and MP8 (one-way ANOVA
p < 0.01), which were not significantly different from each other. There was no
significant difference within the Pacific stations. Similarly, there were no significant
differences in α for CO
2
uptake between and within ocean basins (Table 2.4). On
average, there was more variation between stations on the N. Atlantic cruises
compared to the N. Pacific cruises (as illustrated by C
v
in Tables 2.4 and 2.5). In the
Atlantic, the majority of P-E curves (~80%) indicated photoinhibition, whereas in the
Pacific less than half (~43%) indicated photoinhibition.
In contrast to CO
2
fixation, the results from the N
2
fixation experiments do
reveal interbasin differences, but this was dependent on normalization factor (Table
2.5). P
chl
max
and P
b
max
values in the Atlantic were significantly lower in the Atlantic
than in the Pacific (one-way ANOVA, p < 0.01 for both). But, there was no statistical
difference in P
C
max
within and between ocean basins. A similar pattern holds for α

30

Table 2.3. Averages and standard error (in parentheses) of basic physiological characteristics of Trichodesmium for
all cruises. n = the number of stations and C
v
is the coefficient of variation.

ocean basin cruise trichomes per colony
µg C trichome
-1
ng chl a trichome
-1
chl a C
-1
(g g
-1
)
145 (29) 0.026 (0.003) 0.275 (0.049) 0.010 (0.001)
n=6 n=9 n=8 n=5
249 (25) - 0.037 (0.007) -
n=9 - n=14 -
160 (25) 0.019 (0.002) 0.069 (0.008) 0.004 (0.001)
n=11 n=15 n=11 n=11
N. Atlantic Avg 187 (18) 0.022 (0.002) 0.122 (0.025) 0.006 (0.001)
N. Atlantic C
v
49% 103% 38% 63%
99 (13) - 0.231 (0.050) -
n=10 - n=10 -
104 (14) 0.056 (0.010) 0.124 (0.022) 0.003 (0.001)
n=10 n=10 n=10 n=9
N. Pacific Avg 100 (10) 0.056 (0.011) 0.176 (0.031) 0.003 (0.001)
N. Pacific C
v
42% 75% 62% 68%
MP6
MP9
N. Atlantic
N. Pacific
MP1
MP3
MP8

31
Table 2.4. Average and standard error (in parentheses) P
max
and α for CO
2
fixation in Trichodesmium at all
stations. An asterisk (*) indicates an outlier value and a plus sign (+) indicates significant difference of p < 0.01.
C
v
is the coefficient of variation.

ocean
basin
cruise station
P
chl
max
!
chl
P
b
max
!
b
P
c
max
!
c
E
k r
2
23 1.4 0.874* 14 9.10* 12.0 7.92* 2* -
24 2.0 (0.6) 0.093 (0.415)* 36 (11) 1.70 (0.52)* 16 (5) 0.77 (0.22)* 21* 0.2
25 2.8 (0.5) 0.020 (0.004) 72 (12) 0.52 (0.10) 31 (5) 0.22 (0.04) 139 0.8
37 1.6 (0.5) 0.016 (0.006) 40 (5) 0.41 (0.15) 14 (2) 0.14 (0.05) 99 0.5
41 0.1 (0.02) 0.001 (0.001) 5 (1) 0.04 (0.03) 2 (0.3) 0.01 (0.01) 160 0.1
44 1.0 (0.3) 0.006 (0.002) 21 (7) 0.11 (0.03) 7 (2) 0.04 (0.01) 178 0.6
49 1.9 (0.7) 0.006 (0.002) 23 (11) 0.08 (0.03) 15 (7) 0.05 (0.02) 298 0.5
51 0.7 (0.2) 0.008 (0.007) 11 (3) 0.13 (0.11) 7 (2) 0.08 (0.06) 84 0.3
MP1 avg
1.4 (0.3)
+
0.010 (0.003) 28 (8) 0.22 (0.08) 13 (3) 0.19 (0.10) 159 (31) -
MP1 C
v
58% 150% 77% 93% 67% 142% 48% -
17 28.7 (4.4)* 0.340 (0.128)* 12 (2) 0.14 (0.05) 8 (1) 0.09 (0.03) 84 0.6
48 1.8 (0.2) 0.035 (0.016) 10 (1) 0.19 (0.09) 10 (1) 0.19 (0.09) 52 0.4
49 3.7 (0.8) 0.017 (0.004) 20 (4) 0.09 (0.02) 13 (3) 0.06 (0.01) 212 0.6
MP3 avg
2.8 (1.0)
+
0.026 (0.009) 14 (3) 0.14 (0.03) 10 (1) 0.11 (0.04) 116 (49) -
MP3 C
v
49% 49% 38% 36% 24% 60% 73% -
30 4.3 (1.1) 0.026 (0.012) 12 (3) 0.07 (0.04) 16 (4) 0.10 (0.05) 166 0.6
34 3.2 (2.1) 0.007 (0.005) 15 (10) 0.03 (0.02) 12 (8) 0.03 (0.02) 462 0.5
38 3.8 (0.9) 0.025 (0.012) 25 (6) 0.16 (0.08) 17 (4) 0.11 (0.05) 154 0.4
MP8 avg 3.8 (0.3)
+
0.019 (0.006) 17 (4) 0.09 (0.04) 15 (2) 0.08 (0.01) 260 (100) -
MP8 C
v
15% 55% 39% 77% 18% 54% 67% -
N. Atlantic avg 2.2 (0.4) 0.022 (0.007) 23 (5) 0.16 (0.04) 13 (2) 0.15 (0.05) 174 (32) 112.1
N. Atlantic C
v
59% 114% 77% 91% 52% 136% 65% -
N. Atlantic
MP1
MP3
MP8

32
ocean
basin
cruise station
P
chl
max
!
chl
P
b
max
!
b
P
c
max
!
c
E
k r
2
3 2.6 (0.5) 0.023 (0.009) 24 (5) 0.20 (0.08) 8 (2) 0.07 (0.03) 115 0.6
5 1.8 (0.2) 0.016 (0.005) 27 (3) 0.24 (0.08) 10 (1) 0.09 (0.03) 112 0.6
MP6 avg. 2.2 (0.4) 0.020 (0.003) 26 (2) 0.22 (0.02) 9 (1) 0.08 (0.01) 114 (2) -
MP6 C
v
26% 25% 8% 13% 16% 18% 2% -
14 4.2 (1.0) 0.030 (0.013) 56 (14) 0.40 (0.17) 19 (5) 0.13 (0.06) 141 0.4
14b 2.5 (0.8) 0.019 (0.011) 23 (7) 0.17 (0.10) 7 (2) 0.06 (0.03) 132 0.4
16 2.7 (0.4) 0.010 (0.002) 19 (3) 0.07 (0.08) 6 (1) 0.02 (0.03) 281 0.7
17 1.5 (0.3) 0.007 (0.002) 37 (7) 0.18 (0.04) 10 (2) 0.05 (0.01) 204 0.8
21 2.6 (0.3) 0.013 (0.004) 27 (3) 0.14 (0.05) 11 (1) 0.06 (0.02) 198 0.6
MP9 avg. 2.7 (0.4) 0.016 (0.004) 32 (7) 0.19 (0.06) 11 (2) 0.06 (0.02) 191 (27) -
MP9 C
v
36% 58% 46% 65% 48% 63% 31% -
N. Pacific avg 2.6 (0.3) 0.017 (0.003) 30 (5) 0.20 (0.04) 10 (2) 0.07 (0.01) 169 (23) -
N. Pacific C
v
34% 47% 41% 51% 42% 50% 37% -
MP9
N. Pacific
MP6
Table 2.4. Continued.

33
Table 2.5. Average and standard error (in parentheses) P
max
and α for N
2
fixation in Trichodesmium at all
stations. An asterisk (*) indicates an outlier value and a plus sign (+) indicates significant difference of p <
0.01. C
v
is the coefficient of variation.
ocean
basin
cruise station
P
chl
max
!
chl
P
b
max
!
b
P
c
max
!
c
E
k r
2
24 14 (5) 0.1 (0.1) 0.2 (0.1) 0.002 (0.001) 0.1 (0.04) 0.001 (0.001) 100 0.4
25 35 (5) 0.3 (0.2) 0.8 (0.1) 0.008 (0.005) 0.3 (0.1) 0.003 (0.002) 100 0.3
37 49 (9) 0.6 (0.6) 1.1 (0.2) 0.013 (0.013) 0.4 (0.1) 0.005 (0.004) 80 0.2
41 27 (6) 0.1 (0.1) 1.0 (0.2) 0.004 (0.003) 0.3 (0.1) 0.001 (0.001) 300 0.3
44 20 (13) 0.2 (0.2) 0.4 (0.2) 0.004 (0.004) 0.1 (0.1) 0.001 (0.001) 100 0.2
49 136 (29) 1.1 (0.4) 1.5 (0.3) 0.011 (0.004) 0.9 (0.2) 0.007 (0.003) 129 0.6
51 149 (52) 0.8 (0.5) 2.0 (0.7) 0.011 (0.007) 1.2 (0.4) 0.006 (0.004) 200 0.2
MP1 avg 61 (21) 0.5 (0.1) 1.0 (0.2) 0.008 (0.002) 0.5 (0.2) 0.003 (0.001) 144 (29) -
MP1 C
v
92% 85% 62% 56% 89% 75% 55% -
17 43 (43) 0.1 (0.1) 0.02 (0.02) 0.00004* 0.01 (0.01) 0.00003* 333 0.8
48 56 (9) 1.1 (0.9) 0.3 (0.04) 0.005 (0.004) 0.3 (0.04) 0.005 (0.004) 53 0.2
49 427 (33) 7.8 (3.0)* 2.0 (0.2) 0.036 (0.014) 1.3 (0.1) 0.023 (0.009) 57 0.5
MP3 avg 175 (126) 0.6 (0.5) 0.8 (0.6) 0.021 (0.016) 0.5 (0.4) 0.014 (0.009) 147 (92) -
MP3 C
v
124% 140% 139% 107% 126% 91% 109% -
30 105 (12) 4.2 (26.9)* 0.3 (0.3) 0.010 (0.066) 0.3 (0.04) 0.013 (0.085) 23 0.2
34 93 (34) 0.2 (0.1) 0.4 (0.1) 0.001 (0.0002) 0.3 (0.1) 0.001 (0.0002) 495 0.6
38 20 (4) 0.2 (0.2) 0.1 (0.02) 0.001 (0.001) 0.1 (0.01) 0.001 (0.001) 94 0.2
MP8 avg 73 (27) 0.2 (0.0) 0.3 (0.1) 0.004 (0.003) 0.2 (0.1) 0.005 (0.004) 203 (147) -
MP8 C
v
63% 0% 57% 130% 49% 139% 93% -
N. Atlantic avg
90 (31)
+
1.1 (0.6)
+
0.8 (0.2)
+
0.009 (0.003)
+
0.4 (0.1) 0.006 (0.002) 159 (38) -
N. Atlantic C
v
123% 206% 89% 108% 98% 117% 87% -
N. Atlantic
MP1
MP3
MP8

34
ocean
basin
cruise station
P
chl
max
!
chl
P
b
max
!
b
P
c
max
!
c
E
k r
2
3 82 (27) 1.2 (0.8) 0.6 (0.2) 0.009 (0.006) 0.2 (0.1) 0.003 (0.002) 71 0.1
3b 136 (17) 5.0 (3.3) 1.8 (0.2) 0.066 (0.044) 0.6 (0.1) 0.023 (0.016) 28 0.3
5 149 (18) 18.7 (25)* 2.0 (0.2) 0.247 (0.339)* 0.7 (0.1) 0.087 (0.120) 8* 0.4
7 596 (217) 2.8 (1.0) 3.5 (1.3) 0.016 (0.005) 0.8 (0.3) 0.004 (0.001) 219 0.5
8 94 (48) 1.0 (0.5) 3.1 (1.6) 0.033 (0.017) 0.2 (0.1) 0.002 (0.001) 95 0.5
MP6 avg 211 (97) 2.5 (0.9) 2.2 (0.5) 0.031 (0.013) 0.5 (0.1) 0.024 (0.016) 103 (41) -
MP6 C
v
103% 74% 52% 82% 57% 79% -
14 118 (188)* 0.7 (1.2)* 1.2 (2.2)* 0.008 (0.013) 0.5 (0.7)* 0.003 (0.005)* 152 0.2
14b 161 (93) 1.4 (1.4) 1.3 (0.7) 0.011 (0.011) 0.4 (0.2) 0.004 (0.004) 103 0.1
16 207 167.0* 1.3 1.001* 0.4 0.315* 1* 0.6
16b 339 (103) 0.9 (0.3) 2.5 (0.8) 0.007 (0.002) 0.6 (0.2) 0.002 (0.001) 323 0.7
17 127 (39) 1.4 (1.4) 2.6 (0.8) 0.028 (0.028) 0.7 (0.2) 0.008 (0.008) 88 0.3
17b 257 (60) 1.7 (1.0) 2.3 (0.6) 0.015 (0.010) 0.4 (0.1) 0.003 (0.002) 142 0.6
18 316 (45) 4.3 (1.4) 1.4 (0.2) 0.019 (0.006) 0.1 (0.02) 0.002 (0.001) 74 0.6
19 325 107.2* 2.8 0.916* 0.3 0.093* 3* 0.6
19b 392 (35) 5.1 (2.4) 1.4 (0.1) 0.018 (0.008) 0.4 (0.04) 0.006 (0.003) 77 0.5
21 213 78.3* 1.9 0.709* 0.8 0.284* 3* 0.4
22 95 96.9* 0.8 0.815* 0.2 0.160* 1* 0.4
MP9 avg 232 (31) 2.2 (0.7) 1.8 (0.2) 0.015 (0.003) 0.4 (0.1) 0.004 (0.001) 137 (33) -
MP9 C
v
44% 79% 38% 48% 49% 56% 63% -
N. Pacific avg
225 (35)
+
2.3 (0.5)
+
1.9 (0.2)
+
0.021 (0.005)
+
0.5 (0.1) 0.012 (0.007) 125 (25) -
N. Pacific C
v
62% 73% 44% 81% 50% 198% 66% -
MP9
MP6
N. Pacific
Table 2.5. Continued.
35

Table 2.6. Units for measurement abbreviations.
CO
2
Fixation
P
chl
max mg C (mg chl a
-1
) hr
-1
!
chl
mg C (mg chl a
-1
hr-1) (µE m
-2
s
-1
)
-1
P
b
max pmol C trichome
-1
hr
-1
!
b
pmol C trichome
-1
hr
-1
(µE m
-2
s-
1
)
-1
P
c
max
mmol C
f
(mmol C)
-1
hr
-1
!
c
mmol C
f
(mmol C)
-1
hr
-1
(µE m
-2
s
-1
)
-1
E
k µE m
-2
s
-1
N
2
Fixation
P
chl
max
µg N (mg chl a
-1
) hr
-1
!
chl
µg N (mg chl a
-1
hr-1) (µE m
-2
s
-1
)
-1
P
b
max pmol N trichome
-1
hr
-1
!
b
pmol N trichome
-1
hr
-1
(µE m
-2
s-
1
)
-1
P
c
max
mmol N
f
(mmol C)
-1
hr
-1
!
c
mmol N
f
(mmol C)
-1
hr
-1
(µE m
-2
s
-1
)
-1
E
k µE m
-2
s
-1
36
(Table 2.4). Both α
chl
and α
b
were significantly lower in the Atlantic than in the
Pacific (one-way ANOVA, p < 0.01 for both). But again, there was no statistical
difference between and within ocean basins for α
C
. Half of the results indicated
photoinhibition in the Atlantic and half did not; and in the Pacific, slightly more than
the majority indicated photoinhibition (~63%).
Figures 2.2a-b depict examples of CO
2
and N
2
fixation P-E curves. In several stations
on MP9 (n=7), for the N
2
fixation experiments, production was almost instantaneously
maximal even at very low light conditions and α values were exceedingly high. If
P
max
and α from these curves were five times greater than the standard deviation, the
data were considered outliers and were not included in the cruise means. Figures 2.3a-
b provide two examples of N
2
fixation P-E curves from
MP9 that reached maximal production rates very rapidly but that still had P
max
and α
values that fell within the cruise means.
Mean E
k
values for CO
2
fixation and N
2
fixation were not statistically different
between ocean basins though there was greater range in E
k
values in the Atlantic than
in the Pacific (Tables 2.4 and 2.5). On average, the E
k
for N
2
fixation was lower than
for CO
2
fixation (119 ± 17 m for N
2
fixation vs. 172 ± 22 m for CO
2
fixation), but the
difference was not statistically significant. Newly fixed C to newly fixed N (C
f
:N
f
)
ratios were higher in the Atlantic than in the Pacific (Table 2.7) but the difference was
again not statistically significant. There was however more variance in the Atlantic
samples than in the Pacific samples (note C
v
. ≥ 90% in most samples in the Atlantic
vs. C
v
≈ 55% in the Pacific, Table 2.7). A data point was considered an outlier if the
37
0
20
40
60
80
100
0 500 1000 1500 2000 2500
CO
2
Fixation (pmol C trichome
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.5
1
1.5
2
2.5
0 500 1000 1500 2000
N
2
Fixation (pmol N trichome
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
Figure 2.2a-b. Examples of P-E curves for a) CO
2
fixation and b) N
2
fixation.

A

B

38
0
0.5
1
1.5
2
2.5
3
3.5
0 500 1000 1500 2000 2500
N
2
Fixation (pmol N trichome
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 500 1000 1500 2000 2500
N
2
Fixation (pmol N trichome
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)

Figure 2.3a-b. Examples of curves from MP9 where N
2
fixation was
immediately saturated at low light.
A
B
39
Table 2.7. Carbon fixed to nitrogen fixed ratios (C
f
:N
f
). Values in parentheses are
standard errors, n is the number of samples and C
v
is the coefficient of variance.
ocean basin cruise
C
chl
f
:N
chl
f
C
b
f
:N
b
f
C
C
f
:N
C
f
MP1 47 (19) 55 (24) 57 (22)
c.v. (n=7) 108% 114% 102%
MP3 20 (12) 22 (12) 20 (12)
c.v. (n=2) 81% 76% 81%
MP8 88 (51) 109 (70) 88 (41)
c.v. (n=2) 99% 112% 81%
N. Atlantic average 40 (12) 46 (15) 49 (14)
N. Atlantic c.v. 101% 111% 98%
MP6 22 (10) 27 (13) 22 (12)
c.v. (n=2) 108% 108% 81%
MP9 18 (5) 22 (6) 20 (4)
c.v. (n=5) 58% 66% 51%
N. Pacific average 19 (4) 23 (5) 22 (4)
N. Pacific c.v. 54% 61% 53%
N. Atlantic
N. Pacific
40
Table 2.8. Correlation coefficients for CO
2
and N
2

fixation parameters P
max
and α
with [DFe], [SRP] and mixed layer depth (MLD). r
2
represents the correlation
coefficient, n is the number of data points, and Prob
N
is the percent probability that
the two parameters (CO
2
or N
2
vs. [DFe], [SRP] or MLD) are not correlated.
CO
2
Fixation
r
2
n Prob
N r
2
n ProbN
r
2
n Prob
N
P
chl
max
0.6 11 5.1% 0.1 19 69% 0.3 17 24%
!
chl
0.1 10 78% 0.0 17 100% 0.2 16 46%
P
b
max
0.1 11 77% 0.3 19 21% 0.1 17 70%
!
b
0.2 10 58% 0.1 17 70% 0.1 16 71%
P
C
max
0.2 11 56% 0.0 19 100% 0.0 20 100%
!
C
0.0 10 78% 0.1 18 68% 0.0 18 100%
N
2
Fixation
r
2
n Prob
N r
2
n Prob
N r
2
n Prob
N
P
chl
max
0.2 11 56% 0.3 23 18% 0.0 22 100%
!
chl
0.0 9 100% 0.1 17 70% 0.1 16 71%
P
b
max
0.1 11 77% 0.2 20 40% 0.0 24 100%
!
b
0.1 9 80% 0.2 12 40% 0.0 17 100%
P
C
max
0.0 10 100% 0.0 24 100% 0.0 23 100%
!
C
0.0 10 100% 0.1 18 69% 0.0 18 100%
[DFe] [SRP] MLD
41
value was five times greater than the standard deviation. No outliers were used in
determining the C
f
:N
f
ratios. There were no significant correlations between P
max
and
α for CO
2
and N
2
fixation parameters with either [DFe], [SRP] and mixed layer depth
(Table 2.8 and Appendix A).

DISCUSSION
Comparison to previously published estimates:
This study provides photosynthetron-derived estimates of P
max
and α for CO
2

and N
2
fixation in Trichodesmium in the tropical N. Atlantic and provides the first
estimates of both these parameters for CO
2
and N
2
fixation in Trichodesmium for the
subtropical and tropical N. Pacific. Based on multiple cruises, P
max
and α for CO
2

fixation was not significantly different between the N. Atlantic and the N. Pacific
stations. To the contrary, P
max
and α for N
2
fixation in the tropical N. Pacific stations
was significantly greater than in the tropical N. Atlantic stations.
In general, P
chl
max
and α
chl
values for both CO
2
and N
2
fixation in this study generally
corroborate previously published data for Trichodesmium from the N. Atlantic (Table
2.9). Neither P
max
nor α for either CO
2
or N
2
fixation of natural populations of
Trichodesmium in the N. Pacific have been reported. In our studies, P
max
for CO
2

fixation ranged from 1.0 - 4.3 mg C (mg chl a)
-1
h
-1
in the Atlantic and 1.5 - 4.2 mg C
(mg chl a)
-1
h
-1
in the Pacific. These values fall well within the range of previous
studies that have measured P
max
in natural populations of Trichodesmium in the N.
Atlantic, which report values over a much greater range of 0.1 - 13.4 mg C (mg chl a)
-
42
1
h
-1
(Table 2.9). The highest measurements of P
max
are from Carpenter et al. (1993)
and Carpenter and Roenneberg (1995) (11.9 ± 1.1 and 13.4 mg C (mg chl a)
-1
h
-1
,
respectively) and were measured during winter cruises in the subtropical Atlantic.
Other measurements from the same Carpenter and Roenneberg (1995) study, from
Trichodesmium collected in different regions of the subtropical N. Atlantic and during
the autumn, are ~6 times lower (1.9 - 2.5 mg C (mg chl a)
-1
h
-1
) and are within the
same range as the other previously reported values and this dissertation. Carpenter et
al. (1993) attribute their high values to natural variability in chl a concentrations
within cells from different locations and to the fact that the cells were rapidly growing.
The α
chl
values for CO
2
fixation presented in this dissertation study ranged
from 0.006 - 0.035 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
. Again, these values
match previously reported values of α
chl
, which range from 0.004 - 0.041 mg C (mg
chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
(Table 2.9).
P
max
and α values for N
2
fixation in this study showed significant differences
between the Atlantic and the Pacific with an average P
chl
max
of 90 ± 31 µg N (mg chl
a)
-1
h
-1
in the Atlantic and 225 ± 35 µg N (mg chl a)
-1
h
-1
in the Pacific and mean α
values of 1.1 ± 0.6 µg N (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
in the Atlantic and 2.3 ±
0.5 µg N (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
in the Pacific.
Only one paper has previously reported P
max
values for field populations of
Trichodesmium with respect to N
2
fixation. Hood et al. (2002), reported lower mean
P
chl
max
values of 45 µg N (mg chl a)
-1
h
-1
, with a range from 9 - 140 µg N (mg chl a)
-1

43
Paper Location
P
chl
max
!
chl
CO
2
Fixation
mg C (mg chl a)
-1
h
-1
mg C (mg chl a)
-1
h
-1
/
µmol quanta m
-2
s
-1
Lewis et al. (1980) Eastern Caribbean Sea ~3.8 -
Li et al. (1988) Western Sargasso Sea 1.3 - 6.0 0.004 - 0.011
Carpenter et al. (1993)
Bahama Islands & Eastern Caribbean
Sea
11.1 - 12.6 0.020 - 0.032
Carpenter and
Roenneberg (1995)
Barbados, Sargasso Sea, Eastern
Caribbean Sea
1.9 - 13.4 0.017 - 0.041
Tropical N. Atlantic 1.0 - 4.3 0.006 - 0.035
Tropical N. Pacific 1.5 - 4.2 0.007 - 0.030
N
2
Fixation
µg N (mg chl a)
-1
h
-1
µg N (mg chl a)
-1
h
-1
/
µmol quanta m
-2
s
-1
Hood et al. (2002) Tropical N. Atlantic 9 - 140 ~2.3
Tropical N. Atlantic 14 - 149 0.1 - 1.1
Tropical N. Pacific 82 - 596 1.0 - 5.1
This Study
This Study
Table 2.9. Comparison of previous studies and this study for P
chl
max
and α
chl
for CO
2
and N
2
fixation in
Trichodesmium. Values determined to be outliers likely due to erroneous values in this study were not
included in the ranges.
44
in Hood et al. correlate most closely with the Atlantic values from this study, but the α
values correlate more closely with this study's Pacific values.
In this study, a 21-well photosynthetron was used (Hood et al. 1991,1995)
whereas in the previous studies, experiments were conducted in on-deck, open,
Plexiglass incubators with temperature maintained by surface seawater pumped
through the incubator. In these incubators, multiple layers of window screening were
utilized to create five irradiance levels: 100%, 50%, 25%, 10% and 1% of incident
irradiance (Carpenter et al. 2004). During the cruises MP1 (tropical N. Atlantic) and
MP9 (tropical N. Pacific), concurrent on-deck (n=13) N
2
fixation experiments were
conducted on the same Trichodesmium populations as those that were utilized in the
photosynthetron experiments (n=18). There was no statistical difference in P
chl
max

and P
b
max
values between the two different methods, nor was there significant
difference in α
chl
values (Table 2.10). However, there were significant differences
between α
b
values measured on the photosynthetrons compared to those measured on-
deck (one-way ANOVA p<0.05); mean α
b
values measured on the photosynthetron
were much higher than those measured in the on-deck incubations for both cruises
combined: 0.015 ± 0.003 vs. 0.003 ± 0.001 pmol N trichome
-1
h
-1
/µmol quanta
-1
m
-2

s
-1
, respectively. The on-deck measurements underestimated α by ~5-fold.
The on-deck experiments are subject to a greater range of environmental
variability such as appearing and disappearing clouds and the natural cycle of
irradiance. Thus, the on-deck Trichodesmium colonies were not receiving the same
amount and quality of light throughout the experiment, whereas the phytoplankton

45
Cruise P
max
Ptron P
max
On Deck ! Ptron ! On Deck
MP1 0.061 (0.023) 0.071 (0.013) 0.5 (0.1) 0.4 (0.1)
MP9 0.244 (0.045) 0.453 (0.256) 2.2 (0.7) 0.8 (0.4)
MP1 1.0 (0.3) 1.3 (0.2) 0.008 (0.002)* 0.007 (0.002)*
MP9 1.8 (0.3) 2.1 (0.8) 0.015 (0.003)*^ 0.003 (0.001)*^
normalized to chl
normalized to
trichome
Table 2.10. Comparison of photosynthetron and on-deck experiments for P
max
and α. * indicates
p < 0.05 significance between cruises and ^ indicates p < 0.05 significance between methods.

46

0
20
40
60
80
100
120
140
0 500 1000 1500 2000 2500
on deck exp
photosynthetron exp
N
2
fixation (pmol N col
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
Figure 2.4. Example of a concurrent on-deck experiment and photosynthetron
experiment on colonies collected during the same new tows.

47
Trichodesmium were. This may explain the difference in P
max
and α values in
Trichodesmium which were collected from presumably the same population.
The difference may also be due to more methodological considerations. Data
from an on-deck N
2
fixation experiment and a concurrent photosynthetron experiment
are plotted with the respective curve fits in Figure 2.4. In the on-deck experiment,
there are only three points below 50% irradiance. In contrast, in photosynthetron
experiment, there are 16 data points at the lower light levels. The extra data points in
the photosynthetron experiment allow the curve to be characterized more robustly,
whereas with only three points, the curve-fit routine misses key data. Quite simply,
there are not enough data points in the on-deck experiments to accurately predict
α, thereby leading to an underestimation of α.
Ultimately, where published data are available, there is good correlation
between the values of P
max
and α for CO
2
and N
2
fixation presented in this study and
those previous studies.

Abiotic controls on photophysiological parameters in Trichodesmium:
The data from this study indicate a significant difference in N
2
fixation rates
between the N. Atlantic cruise stations and the N. Pacific cruise stations. There are a
number of potential causes of variability in the photophysiological parameters within
and between ocean basins, such as life cycle stage (Welschmeyer and Lorenzen 1981),
nutritional status (Kirk 2000) or previous light history (Falkowski 1994).
Welschmeyer and Lorenzen (1981) measured α and P
max
in six different species of

48
non-diazotrophic phytoplankton. Within species, there were no significant differences
in α and P
max
in cultures during exponential growth phase; there was, however,
significant variability among cells during stationary growth phase.
In this study, when sampling in the field, it is probable that cells at different
growth stages were collected and assayed, however cell cycle stage was not
determined in this study and therefore it cannot be ascertained if this was a cause of
the differences within stations. However, the marked reductions in P
max
and α were
basin-wide, and it is not likely that colonies of one age were systematically sampled in
the Pacific compared to the Atlantic. It is therefore doubtful that the differences
observed can be attributed to system-wide differences in cell cycle stage.
Reduction of P
max
for CO
2
fixation in eukaryotic phytoplankton in nutrient
limiting conditions has been well documented (Welschmeyer and Lorenzen 1981,
Falkowski 1992, Sosik 1996, Sakshaug 1997, Bouman et al. 2000, Behrenfeld et al.
2002). Equally, Sosik (1996) demonstrated that distance from the nutricline correlates
with a decrease in α. Similarly, reduction in N
2
fixation capacity by both P- and Fe-
limitation has been reported (Rueter 1988, Rueter et al. 1990, Paerl et al. 1994,
Berman-Frank et al. 2001, Kustka et al. 2003, Fu and Bell 2003b). Nutrient stress can
both directly and indirectly decrease N
2
fixation rates. Directly, Fe is a primary
component in nitrogenase, the key enzyme in N
2
fixation; under conditions of Fe-
limitation, synthesis of nitrogenase can be reduced, thereby directly reducing N
2

fixation capacity. Indirectly, P- and Fe-limitation reduce CO
2
fixation capacity

49
(Rueter 1990, Rueter et al. 1990, Berman-Frank et al. 2001, Kustka et al. 2003). N
2

fixation is dependant on CO
2
fixation in that the energy and reductant formed during
photosynthesis accumulate in an internal pool from which N
2
fixation derives the
energy required for this energetically expensive process. Thus, if CO
2
fixation is
reduced, leading to a reduction in internal ATP stores, so too N
2
fixation will be
reduced. Trichodesmium IMS101 had significantly reduced P
max
and α values for both
CO
2
and N
2
fixation when cultured under P- and Fe-limiting nutrient conditions
(Chapter 3).
The current literature suggests that the subtropical N. Atlantic is limited by P
availability while the subtropical N. Pacific is Fe-limited (Wu et al. 2000, Sañudo-
Wilhelmy et al. 2001). Soluble reactive phosphorus (SRP) (Table 2.2, K. Bjorkmann
pers. comm.) and dissolved iron (DFe, Table 2.2, S. Sañudo-Wilhelmy 2001, 2004)
were measured on cruises in both the N. Atlantic and N. Pacific. There was no
statistical significance between either of these nutrients and P
max
and α for either CO
2

or N
2
fixation. However, mean [SRP] at stations where photosynthetron experiments
were conducted were lower in the Atlantic than in the Pacific (0.023 ± 0.003 µM vs.
0.043 ± 0.008 µM, one-way ANOVA, p < 0.05). P
max
and α values for CO
2
fixation
were not significantly different between basins, however for N
2
fixation, P
max
and α
were significantly reduced in the Atlantic compared to the Pacific. It is possible then
that the Atlantic Trichodesmium populations were slightly more P-limited than those
in the Pacific, resulting in reduced N
2
fixation rates. One should be careful and not

50
overstate this correlation, however, as water column nutrients may not be an indicator
of nutritional status of an organisms. Internal nutrient concentrations may provide a
better indication (Sañudo-Wilhelmy, pers. comm.).
Previous light history can have a significant effect on the photophysiological
parameters P
max
and α (Falkowski 1994, Bouman et al. 2000). In studies on
eukaryotic phytoplankton, Bouman et al. (2000) found that P
max
is indirectly correlated
to optical depth. Model results indicate that wind speed (Hood et al. 2002) and depth
and duration of mixed layer depth (Hood et al. 2004) control spatial extent of
Trichodesmium. In the field, Sañudo-Wilhelmy (2001) observed higher rates of N
2

fixation with shallower mixed-layer depth, hypothesizing that inadequate mean
irradiance for photosynthesis may directly reduce the energetically expensive N
2

fixation process. Most recently, Breitbarth (2004) examined the effect of light
intensity on growth rate, CO
2
fixation, and N
2
fixation in Trichodesmium. His results
indicate that growth rate increases up to ~180 µmol quanta m
-2
s
-1
, and remained
constant up to ~1100 µmol quanta m
-2
s
-1
, at which point they became photoinhibited.
The stations in the Atlantic (n=13) had significantly deeper mixed layer depths
than those in the Pacific (n=8, one-way ANOVA, p<0.01), with mean mixed layer
depths in the Atlantic of 68 ± 6 m and mean depths in the Pacific of 52 ± 2 m.
Accordingly, we observed higher N
2
fixation rates in the Pacific. Thus, light
availability and light history may be a contributing factor in explaining the much
higher rates of N
2
fixation in the tropical Pacific compared to the tropical Atlantic.

51
For four stations in the Pacific cruises, the α values estimated yielded the
extremely high α values of 167, 107, 78 and 97 µg N (mg chl a)
-1
h
-1
/µmol quanta m
-2

s
-1
compared to the cruise average of 2 µg N (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
.
Figures 2.2a and 2.2b are examples of reasonable PE curve fits for N
2
fixation in
Trichodesmium while Figures 2.3a and 2.3b are examples of anomalous PE curve fits
for two of the four anomalous curves. Alpha values as high as these suggest that the
Trichodesmium colonies fixed N
2
at a maximal rate with a minimal amount of light.
Note also that not only do these colonies respond to very little light, but they are also
not photoinhibited. These anomalously high α values were not included in the cruise
mean, yet the cruise mean for MP9 is still significantly higher than the means for the
Atlantic cruises. Equally, although MP6 (the other Pacific cruise) did not have
examples of these anomalously high α values, the cruise average was not significantly
different from MP9 (1.6 ± 0.7 vs. 2.2 ± 0.7 µg N (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
)
and as such was again significantly higher than the means for the Atlantic cruises.
No immediate explanation is available for these high α values; i.e. there was
no correlation to any of the environmental parameters analyzed in this study. It is not
clear if these α values are higher because of genetic differences between the Atlantic
and Pacific Trichodesmium or if the α values indicate some sort of environmental
stress in the Atlantic populations. The main conclusion that can be drawn is that there
is a significant difference in α values between the two basins, which is consistent over

52
five cruises, suggesting that this may be a real and persistent difference and should
therefore be incorporated when modeling Trichodesmium in ecosystem models.
It is hypothesized that organisms adapted to a shallower MLD and/or that are
nutritionally stressed are photoinhibited more readily (Grossman et al. 1994).
However, there was no correlation between photoinhibition and any of the factors
discussed above ([SRP], [DFe], mixed layer depth). Similarly, there was no
correlation between photoinhibition and P
max
or α for CO
2
and N
2
fixation.

P
max
and α in ecosystem modeling:
In addition to the information can be gleaned about an organism's
photophysiology, determining accurate values of α and P
max
is critical as these are
often the key parameters used in global ocean ecosystem models.
The first models to incorporate diazotrophs did so by including their
contribution to CO
2
fixation in a distinct compartment for diazotroph production. For
instance, Fennel et al. (2002) assume that α for diazotrophs would be lower than α for
eukaryotic, non-diazotrophic phytoplankton and thus utilize an α value for eukaryotes
of 0.032 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
and for diazotrophs of 0.004 mg C
(mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1.1
Cullen et al. (1992) present values for eukaryotic

1
Fennel et al. report that they based their model parameters on previously reported values from Spitz et
al. (2001), Oschlies and Garcon (1999), Doney et al. (1996) and Fasham et al. (1990). Neither Fasham
et al. (1990), Doney et al. (1996) nor Oschlies and Garcon (1999) discuss diazotroph photophysiology.
Spitz et al. (2001) used α values reported in Geider et al. (1997) who report cyanobacterial values
ranging from 0.034 - 0.086 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
. Yet, these values are much
higher than values reported for Trichodesmium in this report and others, and ultimately also much

53
phytoplankton that range from 0.021 - 0.057 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1

and Peterson et al. (1987) report values of 0.037 mg C (mg chl a)
-1
h
-1
/mmol quanta m
-
2
s
-1
for low light adapted organisms and 0.056 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2

s
-1
for high light adapted organisms. Platt et al. (1992) present eukaryotic α values in
the Sargasso Sea from before, during and after a spring bloom in the oligotrophic
ocean, which range from 0.011 - 0.030 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
. The
values of Platt et al. are lower than the other values, but are essentially equivalent to
the α value for Trichodesmium (0.011 - 0.026 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2

s
-1
). It appears then that Trichodesmium α values are, at the least, equivalent to α
values for non-diazotrophic phytoplankton or, at most, are reduced by a factor of 2.
The α values in Fennel et al. (2002) are therefore too low, which may result in a
reduced contribution of CO
2
fixation by Trichodesmium at the lower light range,
leading to an underestimation of its overall productivity in the models.
Moore et al. (2002a, 2002b) developed another model aimed at describing N
2

fixation globally and they cite the work of Fennel et al. (2002) and the citations therein
as guides for determining what α value to use for their diazotrophic component.
Moore et al. (2002a, 2002b) use the values of 0.027 mg C (mg chl a)
-1
h
-1
/µmol quanta
m
-2
s
-1
for eukaryotic phytoplankton and 0.009 mg C (mg chl a)
-1
h
-1
/µmol quanta m
-2

s
-1
for diazotrophic phytoplankton. These α values for Trichodesmium are higher than
those presented in Fennel et al. (2002) and are closer to the values for Trichodesmium

higher than even the values used in Fennel et al. (2002). It therefore remains unclear how the values in
the Fennel model were determined.

54
presented in this dissertation. Yet, in Moore et al. (2004), an expanded and revised
version of the Moore et al. (2002) model, they kept the same value for eukaryotic
phytoplankton but dropped the value for diazotrophic phytoplankton to 0.003 mg C
(mg chl a)
-1
h
-1
/µmol quanta m
-2
s
-1
which is far too low. Only Hood et al. (2002,
2004) utilize previously published values of α for both CO
2
and N
2
fixation in
modeling Trichodesmium.
Concurrently, in efforts to describe N
2
fixation by Trichodesmium, some
models use the standard carbon to nitrogen (C:N) Redfield ratio of 6.6 to work
between C and N units (e.g. Moore et al. 2002, 2004). In this study, both CO
2
and N
2

fixation in the same populations were measured, thereby providing newly fixed C to
newly fixed N (C
f
:N
f
) comparisons. As apparent in Table 2.7, C
f
:N
f
ratios in these
experiments were ~4-5 times higher than the canonical Redfield C:N ratio of 6.6 with
an average C
f
:N
f
= 40 ± 12 in the Atlantic and 19 ± 4 in the Pacific. Orcutt et al.
(2001) reported similar findings for Trichodesmium experiments conducted at the
Bermuda Atlantic Time-series Study (BATS). For N
2
fixation measured utilizing
15
N
2

gas uptake, average C
f
:N
f
= 128 (with a range from 13 - 437), and for samples
measured utilizing the acetylene reduction technique, average C
f
:N
f
= 198 (with a
range from 15 - 703). Yet, the C:N ratios determined in natural abundance biomass
measurements (C
b
:N
b
) of Trichodesmium by isotope ratio mass spectrometry (IRMS)
are closer to Redfield ratios. Carpenter et al. (2004) report average C
b
:N
b
ranging
from 5.64 ± 0.56 to 7.25 ± 0.39.

55
Orcutt et al. (2001) provide several potential explanations for this large
discrepancy. First, in measuring N
2
fixation by acetylene reduction of
15
N
2
(g) uptake,
one only measures the capacity of Trichodesmium to fix N
2
(Mulholland et al. 2004).
However, Trichodesmium has the capacity to take up other forms of N present in
seawater, such as nitrate or ammonium, (Mulholland and Capone 1999, Holl and
Montoya 2005); this would increase N uptake and thereby decrease C
f
:N
f
.
Assimilation of these other forms of N however has only been demonstrated to
contribute to between 1-6% of N
2
fixation in actively fixing Trichodesmium.
Similarly, Trichodesmium have been demonstrated to leak as much as 50% of the
newly fixed N (Glibert and Bronk 1994, Capone et al. 1994, Mulholland et al. 2004),
which would have the effect of reducing N
f
and increasing the C
f
:N
f
. Conversely, it is
also plausible that carbon turns over more quickly than nitrogen, in Trichodesmium,
whether through ballasting of internal carbohydrate stores (Villareal and Carpenter
2003) or increased respiration and Mehler activity (Kana 1993).
Orcutt et al. (2001) also discuss the observed decoupling of CO
2
and N
2

fixation during different seasons, primarily due to fluctuations in water temperatures.
In their studies, N
2
fixation rates remained fairly constant between summer and winter,
but CO
2
fixation increased in the summer, with increased water temperatures. C
f
:N
f

ratios therefore increased during the summer and decreased during the winter.
Carpenter et al. (2004) provide some evidence that this can even be observed in
biomass C
b
:N
b
ratios. In the two cruises during the spring average C:N = 5.6 ± 0. 6

56
(May 1994) and 6.6 ± 0.1 (October 1996) and in the fall, C:N = 7.3 ± 0.4 (October
1994). Again, however, the differences in biomass C
b
:N
b
are still not as dramatic as
in the C
f
:N
f
ratios reported here and elsewhere.
Lastly, time and duration of sampling may also play a role. Berman-Frank et
al. (2001) describe a temporal segregation of CO
2
and N
2
fixation in Trichodesmium,
with maximal CO
2
fixation rates in the morning and maximal N
2
fixation rates in the
afternoon. Results from nanoSIMS experiments (study presented in Chapter 4) also
demonstrated a temporal segregation in the two processes. In this study, the
photosynthetron experiments were conducted and completed by local noon, during
maximal CO
2
fixation and before peak N
2
fixation. This may serve to bias the C
f
:N
f
for C
f
.
Because C:N ratios are commonly utilized in global ecosystem models, it is
necessary to determine if these are methodological artifacts or if there is indeed
decoupling of CO
2
and N
2
fixation. If the former is the case, then it would be
appropriate for modelers to continue using the Redfield ratios of 6.6 to transfer
between C and N units. However, if the latter is true, the context in which the C:N
ratio is being utilized thus becomes very important. If modelers wish to compare
biomass units of C and N, then 6.6 is the most accurate number. However, if modelers
wish to move between units of CO
2
and N
2
fixation, a C:N closer to 20 may be more
accurate.

57
Conclusion:
In summary, this study provides estimates of P
max
and α for CO
2
and N
2

fixation in Trichodesmium in both the tropical N. Atlantic and tropical N. Pacific.
These data are the first values reported of the photophysiological parameters for CO
2

and N
2
fixation in the tropical N. Pacific. Though there is no significant difference in
P
max
and α for CO
2
fixation between basins, both P
max
and α are reduced in the
Atlantic compared to in the Pacific. It is hypothesized that reduced SRP concentrations
and deeper MLDs contribute to overall energetic reduction in Trichodesmium colonies
of the tropical N. Atlantic compared to the tropical N. Pacific.
As a significant component of global ecosystem models, it is critical to
determine accurate values of P
max
and α for Trichodesmium. Estimates of P
max
and α
for CO
2
fixation in this study corroborate those of previous studies and are therefore
should be considered accurate estimates for use in models.
The difference in P
max
and α for N
2
fixation between the Atlantic cruise
stations and the Pacific cruise stations is a significant finding and is an important
consideration for modelers and ecologists alike. Though it is unclear if the difference
in rates between the two basins are due to systematic difference in populations or are
examples of chance environmental stress in the Atlantic at the time of sampling, it is
clear that one P
max
and α value for all ocean basins is good enough. The use of one
over the other would lead to a two-fold under- or over-estimation of total N
2
fixation;
this would be a significant miscalculation.

58
Chapter 3
The Effect of Fe- and P-Limitation on the Photophysiological Parameters P
max

and α in Trichodesmium IMS101.

ABSTRACT
Laboratory cultures of Trichodesmium IMS101 were grown under P- and Fe-
limiting conditions. CO
2
and N
2
fixation vs. irradiance experiments were conducted to
determine the effect of nutrient limitation on the photophysiological parameters P
max

and α. For CO
2
and N
2
fixation, both P
max
and α were reduced under both P- and Fe-
limiting conditions. These are the first data to: report the effect of nutrient limitation
for the photophysiological parameters for N
2
fixation under P- and Fe-limiting
conditions; report the effect of P-limitation on the photophysiological parameters for
CO
2
fixation; and, review and refine estimates of Fe-limitation on the
photophysiological parameters for CO
2
fixation for Trichodesmium. These results
provide needed insight into Trichodesmium photophysiological responses and are
critical parameters necessary for effective ecosystem modeling.

INTRODUCTION
Trichodesmium is a nonheterocystous cyanobacterium that fixes both carbon
dioxide (CO
2
) and dinitrogen (N
2
) gas and is ubiquitous in oligotrophic tropical and
subtropical seas (Karl et al. 1997, Capone et al. 2005). Recent estimates report that it

59
contributes about 5.7 Tmol new N y
-1
globally (Capone et al. 2005) - a contribution at
times comparable to the diffusion of NO
3
-
from depth in oligotrophic oceanic
ecosystems (Karl et al. 1997, Capone et al. 2005). Trichodesmium is therefore a
critical player in global carbon and nitrogen cycling (Karl et al. 1997, Capone et al.
2005). As such, CO
2
and N
2
fixation by Trichodesmium have begun to be
incorporated in global ecosystem models that model carbon (C) and nitrogen (N)
cycling in the oceans (Spitz et al. 2001, Fennel et al. 2002, Hood et al. 2002, 2004,
2006, Moore et al. 2002a, 2002b, 2004, Coles et al. 2004).
CO
2
and N
2
fixation in photoautotrophic cyanobacteria are both dependent on
light availability; light drives the light reactions of photosynthesis and the energy
captured and reductant generated are utilized in both CO
2
and N
2
fixation. The
relationships between light and CO
2
and N
2
fixation can be defined experimentally
through production versus irradiance response curves (P-E curves) (Falkowski and
Raven 1997, Hood et al. 2002, Staal et al. 2002). The major features of such P-E
curves can be described by two parameters, P
max
(the maximum specific production
rate) and α (the rate of production at low light). These parameters describe the
fundamental physiological response of the organism to light and have been related to
the structure of the photosystem and its response to differing light regimes (for more
extensive discussion see chapter 2).
The physiological effects of Fe-limitation on marine phytoplankton have been
well documented in the literature (Martin et al. 1990). Fe is an essential cofactor in

60
photosynthetic proteins (Rueter 1988, Geider and La Roche 1994), as well as nitrogen
fixation proteins, and oxygen cycling, a critical process for the non-heterocystous
Trichodesmium (Geider and La Roche 1994). In Trichodesmium, about 20-35% of the
cellular Fe may be bound in photosynthetic components (including the relatively iron
rich photosystem I), while an additional 20-50% may be bound in nitrogenase and
dinitrogenase reductase proteins (Kustka et al. 2003). The ratio of photosystem I to
photosystem II (PSI:PSII) was correlated to Fe availability in Trichodesmium IMS101
cultures (Berman-Frank et al. 2001). Similarly, N
2
fixation may both directly and
indirectly affected by Fe-limitation. Utilizing immunochemical methods to identify
cells with or without nitrogenase, the key enzyme in N
2
fixation, Berman-Frank et al.
(2001) demonstrated that in Fe-depleted cells, 11 ± 5% of cells expressed nitrogenase
while in Fe-replete cells, 20 ± 8% of cells expressed nitrogenase. This directly affects
the cell's capacity to fix N
2
. Equally, N
2
fixation is powered by drawing from an
internal pool of energy and ATP generated through the light reactions. Thus, if
photosystem function is reduced under conditions of Fe stress, less ATP is generated,
the reserve pool of ATP shrinks, and N
2
fixation is reduced. N
2
fixation is thus
indirectly reduced by lowered photosynthetic capacity (Rueter 1988).
The literature on the role of P in CO
2
and N
2
fixation is not as extensive as it is
for Fe, but P also plays a significant role in phytoplankton cells. It is a major
component of ATP (Falkner et al. 1984), which is the primary carrier of chemical
energy in the cell and may be involved in post-translational modification of proteins

61
by phosphorylation, critical for adaptive response (Scanlan and Wilson 1999). P is
also a crucial component of nucleic acids and membranes.
It is believed that vast areas of the subtropical oceans are nutrient limited, most
notably by bioavailable nitrogen (N) (Falkowski 1997, Tyrell 1998). As diazotrophs,
Trichodesmium do not experience N-limitation, but it is believed that they do
experience both P- and Fe-limitation (Wu et al. 2000, Sañudo-Wilhemy et al. 2003).
There is strong evidence that supports Fe limitation in the subtropical N. Pacific where
aeolian dust inputs, the primary source of Fe to the upper layers of the open ocean, are
very low (Rueter 1991, Falkowski 1997, Falkowski et al. 1998). More recently, Karl
et al. (2001) have suggested a shift to a more P-limited environment due to increased
N
2
fixation during the last decade. The subtropical N. Atlantic receives 2 to 10 times
higher Fe Aeolian flux than the N. Pacific and also receives tropical rivers’ run-off
rich in Fe. Based on the analysis of cellular quotas, Sañudo-Wilhelmy et al. (2001)
argue that Trichodesmium populations they sampled in the central Atlantic are
primarily P-limited rather than Fe-limited (Tovar-Sanchez 2006).
Wu et al. (2000) compared dissolved nitrate and nitrite (DNN) to dissolved
inorganic phosphorus (DIP) ratios in the subtropical N. Atlantic and subtropical N.
Pacific and observed DNN:DIP = 20 - 32 in the Atlantic and less than 17 in the
Pacific, implying P limitation in the subtropical N. Atlantic. Mills et al. (2004)
provided evidence for co-limitation of both Fe and P in the subtropical N. Atlantic.

62
Therefore, while there is no current consensus on which ocean basin is limited
by what nutrient, nutrient limitation can have significant physiological effects on
marine phytoplankton, and will certainly affect their photophysiological performance
and parameters. Fu and Bell0 (2003b) analyzed the photophysiological response in Fe-
stressed cultures of Trichodesmium isolate IMS101. This study, reviews the results of
Fu and Bell (2003b) and provides photosynthetron-derived estimates of the
photophysiological parameters, P
max
and α, in Trichodesmium under P- and Fe-
limiting conditions. Both parameters are significantly reduced under nutrient
limitation.

METHODS
Culture growth conditions:
Unialgal non-axenic cultures of Trichodesmium IMS-101 were grown at 27 ºC
under a 12:12 h light: dark cycle at 80 µmol quanta m
-2
s
-1
supplied by fluorescent
tubes. Cultures were grown in YBCII media (Chen et al. 1996) with nitrogen,
phosphorous (for the P-limitation experiments) and Fe (for the Fe-limitation
experiments) excluded.
For the P manipulation experiments, KH
2
PO
4
was added to the YBCII media
to yield two final concentrations of 0.2µM and 30µM PO
4
3-
. 500ml of cultures were
grown in 1L polycarbonate media bottles. 250 ml of culture was discarded and

63
replaced with 250ml fresh media every 4 days. Before use, all bottles were cleaned in
10% OmniTrace HCl and were rinsed four times with milli-Q water.
For Fe manipulation experiments, trace metal clean protocols were followed
for media preparation and for handling the cultures and all steps were conducted under
a laminar flow hood. A modified version of YBCII media was used (Sunda et al.
2005). The major salt components of this media were prepared and passed through a
column packed with Chelex-100 (BioRad) resin to decrease the background
concentrations of iron and other trace metals. After chelexing and microwave
sterilization, trace nutrients (Zn, Mn, Co, Mo, Se) and disodium EDTA were added as
a precomplexed mixture at final total concentrations of 20 µM. Fe-limited and replete
cultures were grown in 30 and 200 nM, respectively. These are similar to the
Fe:EDTA ratios used by Kustka et al. (2003a) for limited and replete cultures of
Trichodesmium. pH was measured regularly in the cultures to ensure that pH = 8.1;
this is critical for maintaining Fe chemistry (Sunda et al. 2005).
The cultures for both the P- and Fe-limitation experiments were grown for ~90
days before physiological parameters were measured. With an doubling time of 3 - 5
days, 90 days allowed ample time for the cultures to go through 10 generations under
nutrient limitation to be sure that they were indeed acclimated to the experimental
conditions (MacIntyre and Cullen 2005).

64
Measurements of chl a and cell counts for both P and Fe cultures:
Subsamples were taken prior to each set of P-E measurements for chl a and
cell count determination. For chl a, 25-50 ml of culture was filtered onto a 25 mm
Whatman GF/F filter, and extracted in methanol for chl a analysis (Porra et al. 1989).
For cell counts, 5 ml of culture were fixed with 2% glutaraldehyde and stored at 4 ºC
until counting. Subsamples (500 µl) were filtered on a 1.0µm pore size nucleopore
filter and placed onto a slide. The entire filter was then counted.

Production vs. Irradiance Experiments
CO
2
Fixation Experiments:
10ml of culture were pipetted into 20 ml glass scintillation vials. 2 µCi
NaH
14
CO
3
-
(~50 µl by volume) was added to each sample. Samples were incubated in
the photosynthetron for 2 hours at 27°C. After incubation, samples were placed on ice
to halt any further
14
C uptake. The samples were filtered onto 0.2 µm polycarbonate
filters to measure particulate organic carbon (POC). The filtrate was also collected for
measurement of any
14
C, which could be found in the dissolved organic carbon (DOC)
pool. To ensure that all excess inorganic
14
C was removed, 150µl of 5% HCl was
added to the filters and 250 µl 5% HCL was added to the filtrate, and these were left to
aerate for 24 hours. After full aeration, Ultima Gold LLT (Perkin-Elmer) scintillation
fluid cocktail was added and samples were counted on a Beckman LS-6500
scintillation counter.

65
N
2
Fixation Experiments
Similarly, 10 ml of culture was pipetted into 14 ml serum vials. The vials were
crimp-sealed and 1 ml of C
2
H
2
was injected into each sample through a rubber
stopper. Samples were incubated in the photosynthetron for 2 hours at 27°C. After
incubation, 100 µl of 1 M NaOH was added to the samples to stop any further
acetylene reduction (AR). Samples were measured on an FID Shimadzu GC-9A gas
chromatograph by injecting 100 µL of headspace gas (Capone and Montoya 2001).

Description of photosynthetrons:
The photosynthetrons were bottom-lit with high intensity halogen lights and
irradiance was varied using grey filters from 0 - 2500 µmol quanta m
-2
s
-1
(Hood et al.
1991). Light intensity in each of the 21 wells was measured at the start of every
experiment using a Biospherics Model QSL-100 PAR sensor.

Curve fitting:
Two different equations were used to fit the data:
(1)
!
P =P
s
(1"e
"#E/P
s
)e
"$E/P
s
(Platt et al. 1980)
(2)
!
P =P
s
(1"e
"#E/P
s
) (Webb et al. 1976)
where P
s
is defined as:
(3)
!
P
max
=P
s
"
(" +#)
$
%
&
'
(
)
#
(# +")
$
%
&
'
(
)
$
%
&
'
(
)
#
"
(Hood et al. 1991)

66

and P
max
(mg C/N mg chl a
-1
h
-1
) is the value at which production is maximal and the
curve reaches an asymptote, α (mg C/N mg chl a
-1
h
-1
/µmol quanta m
-2
s
-1
) describes
the linear P-E relationship at low light, E (µmol quanta m
-2
s
-1
) is the incident scalar
irradiance, P (mg C/N mg chl a
-1
h
-1
) is the instantaneous production rate and β (mg
C/N mg chl a
-1
h
-1
/µmol quanta m
-2
s
-1
) is a photoinhibition parameter. An additional
parameter derived from this equation, E
k
(µmol quanta m
-2
s
-1
), the light saturation
parameter, describes the light level at which photosynthesis begins to become
saturated by the incident irradiance. It is defined as P
max
/α.

Statistical analysis:
Statistical analysis was conducted using Kaleidagraph version 3.6.2. One-way
ANOVA was used to analyze variance between parameters. Differences were
considered significant at p < 0.05. Regression analysis was utilized to determine
correlation, described by the correlation coefficient, r
2
. To determine the quantitative
significance of r
2
, the probability (Prob
N
(r≥ r
0
) or Prob
N
) that N measurements of
two uncorrelated variables will give a coefficient r larger than any particular r
0
. Prob
N

< 5% denotes significant correlation and Prob
N
<1% denotes highly significant
correlation (Taylor 1982). Finally, coefficient of variance (C
v
):

67
!
C
v
=
"
µ

where σ is the standard deviation and µ is the mean, was utilized to examine the
variance within a dataset.

RESULTS
Growth rate and chl a content per cell:
Cultures were successfully grown 0.2 µM and 30 µM PO
4
3-
and 30 nM and 200
nM Fe. Based on visual observation (Webb et al. 2001), it was observed that growth
rates were reduced in the lower P and Fe concentration cultures. Chl a content per cell
was also significantly reduced in the cultures grown at lower concentrations compared
to those grown at the higher concentrations (one-way ANOVA, p<0.05, Table 3.1).
Both the reduced chl a cell
-1
and the reduced growth rates in the cultures grown at
lower concentrations indicate that these cultures were nutrient-limited.

Photophysiological parameters of the P cultures:
The P-E curves for CO
2
and N
2
fixation vs. irradiance in the P cultures are
presented in Figures 3.1a-d and the photophysiological parameters are listed in Table
3.2. P
b
max
and α
b
(parameters normalized to cell number) were significantly reduced in
the cultures grown at the lower P concentrations for CO
2
fixation (one-way ANOVA,
p < 0.001) and for N
2
fixation (one-way ANOVA, p < 0.05, Figures 3.1a and 3.1b,
Table 3.2). Similarly, P
chl
max
and α
chl
(parameters normalized to mg chl a) were

68

0.2 µM 30 µM 30 nM 200 nM
0.3 1.9 0.4 0.8
0.4 1.2 0.8 0.4
0.4 1.3 0.5 1.0
0.5 1.4 0.5 1.0
0.4 - 0.7 1.0
0.5 - 0.9 1.0
- - - 0.9
avg (s.e.) 0.4 (0.04)* 1.4 (0.2)* 0.6 (0.1)* 0.9 (0.1)*
n 6 4 6 7
C
v
21% 21% 31% 24%
Fe experiments
PO
4
3-
experiments
Table 3.1. Chl a cell
-1
(pg chl a cell
-1
) measurements for the P and Fe
manipulation experiments.

* p < 0.05 (one-way ANOVA), n is the number of samples, C
v
is the coefficient
of variation.

69
significantly reduced for N
2
fixation (one-way ANOVA, p < 0.05 for both) but only
P
chl
max
was significantly reduced for CO
2
fixation (one-way ANOVA, p<0.05, Figures
3.1c and 3.1d, Table 3.2). E
k
values (the light level at which production starts to
become light saturated) for the P-replete cultures were lower than those for P-deplete
cultures, but the difference was not significant (Table 3.2). Overall, E
k
for N
2
fixation
was significantly lower than for CO
2
fixation. The fixed C to fixed N (C
f
:N
f
) ratios
were reduced in the 30 µM P cultures compared to the 0.2 µM P cultures (19 ± 6 vs.
34 ± 2).

Photophysiological Parameters of the Fe cultures:
The patterns for the Fe-limitation experiments were similar to those found in
the P-limitation experiments. The P-E curves for CO
2
and N
2
fixation vs. irradiance in
the Fe cultures are presented in Figures 2a-d and the photophysiological parameters
are listed in Table 3.3. In the Fe experiments, P
b
max
and α for CO
2
and N
2
fixation
were significantly reduced (one-way ANOVA, p < 0.001 for CO
2
fixation and p <
0.05 for N
2
fixation) and α
chl
for CO
2
fixation was significantly reduced (one-way
ANOVA, p < 0.05, Figures 3.2c and 3.2d, Table 3.3). For E
k
, the Fe-deplete cultures
had greater E
k
values than the Fe-replete cultures, though this difference was only
significant in the case of the cell number normalized values (Table 3.3). This pattern
was opposite to that observed in the P-limitation experiments. As with the P-limitation
experiments, the E
k
values for CO
2
fixation were greater than those for N
2
fixation.

70
0
1
2
3
4
5
0 500 1000 1500 2000
30uM
0.2uM
CO
2
Fixation (mg C mg chl a
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 500 1000 1500 2000
30uM
0.2uM
N
2
Fixation (mg N mg chl a
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)

Figures 3.1a-d. Data points and curve fits for five photosynthetron runs for both the 0.2 and 30 µM
[PO
4
3-
] experiments, normalized to mg chl a for a) CO
2
fixation and b) N
2
fixation, and runs normalized
to cell number for c) CO
2
fixation and d) N
2
fixation. Included on these plots are five photosynthetron
runs at both the low and high [Fe].

A

B

71
0
200
400
600
800
1000
0 500 1000 1500 2000
30uM P
0.2uM P
CO
2
Fixation (fmol C cell
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.1
0.2
0.3
0.4
0.5
0.6
0 500 1000 1500 2000
30uM P
0.2uM P
N
2
Fixation (fmol cell
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)

Figures 3.1a-d. Continued.

72
[PO
4
3-
] P
b
max !
b
E
k P
b
max !
b
E
k
C:N
0.2µM (n=5) 100 (25)*** 0.4 (0.1)*** 255 (15) 4 (1)*** 0.03 (0.01)* 139 (10) 34 (2)
30µM (n=5) 675 (150)*** 3.0 (0.5)*** 228 (17) 33 (4)*** 0.70 (0.06)* 182 (31) 19 (6)
[PO
4
3-
] P
chl
max !
chl
E
k P
chl
max !
chl
E
k
0.2µM (n=5) 3.1 (0.6)* 0.012 (0.003) 259 (16) 0.090 (0.015)* 0.7 (0.2)* 135 (13) 38 (1)
30µM (n=5) 5.5 (2.2)* 0.024 (0.009) 223 (50) 0.335 (0.047)* 1.9 (0.3)* 183 (31) 19 (6)
N
2
Fixation CO
2
Fixation
* P < 0.1 (one-way ANOVA)
** p < 0.01 (one-way ANOVA)
**** p < 0.0001 (one-way ANOVA)

CO
2
Fixation Units:
P
b
max
(fmol C cell
-1
h
-1
)
α
b
(fmol C cell
-1
h
-1
/µE m
-2
s
-1
)
P
chl
max
(mg C (mg chl a
-1
) h
-1)

α
chl
((mg C (mg chl a
-1
) h
-1
)
/
(µE m
-2
s
-1
))
E
k
(µmol quanta m
-2
s-1)

N
2
fixation Units:
P
b
max
(fmol N cell
-1
h
-1
)
α
b
((fmol N cell
-1
h
-1
)

/ (µE m
-2
s
-1
))
P
chl
max
(mg N (mg chl a
-1
) h
-1
)
α
chl
((µg N (mg chl a
-1
) h
-1
)

/ (µE m
-2
s
-1
))
E
k
(µmol quanta m
-2
s-1)

Table 3.2. P
max
and α for the P manipulation experiments. Standard error is in parentheses and n is the
number of samples.

73

0
0.5
1
1.5
2
2.5
3
0 500 1000 1500 2000 2500
200nM Fe
30nM Fe
CO
2
Fixation (mg C mg chl a
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.05
0.1
0.15
0.2
0.25
0.3
0 500 1000 1500 2000 2500
200nM Fe
30nM Fe
N
2
fixation (mg N mg chl a
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)

Figures 3.2a-d. Data points and curve fits for the Fe manipulation experiments, normalized to
mg chl a for a) CO
2
fixation and b) N
2
fixation, and normalized to cell number for c) CO
2

fixation and d) N
2
fixation. Included on these plots are six photosynthetron runs at the low [Fe]
and seven runs at the high [Fe].
A B

74
0
50
100
150
200
250
0 500 1000 1500 2000 2500
200nM Fe
30nM Fe
CO
2
Fixation (fmol C cell
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)
0
0.05
0.1
0.15
0.2
0 500 1000 1500 2000 2500
200nM Fe
30nM Fe
N
2
Fixation (fmol N cell
-1
h
-1
)
Irradiance (µmol quanta m
-2
s
-1
)

Figures 3.2a-d. Continued.

C D

75
[Fe]
P
b
max !
b
E
k P
b
max !
b
E
k
C:N
30nM (n=6) 108 (17)** 0.6 (0.1)**** 200 (17)* 7 (1)* 0.06 (0.01)** 140 (30) 16 (4)
200nM (n=7) 175 (8)** 1.4 (0.1)**** 128 (10)* 9 (1)* 0.11 (0.01)** 93 (14) 16 (2)
[Fe]
P
chl
max
!
chl
P
chl
max
!
chl
30nM (n=6) 2.4 (0.4) 0.012 (0.001)* 213 (7) 0.154 (0.018) 1.2 (0.2) 139 (28) 17(4)
200nM (n=7) 2.7 (0.6) 0.021 (0.004)* 129 (9) 0.169 (0.039) 1.9 (0.5) 92 (13) 17 (1)
N
2
Fixation CO
2
Fixation
* P < 0.1 (one-way ANOVA)
** p < 0.01 (one-way ANOVA)
**** p < 0.0001 (one-way ANOVA)

CO
2
Fixation Units:
P
b
max
(fmol C cell
-1
h
-1
)
α
b
(fmol C cell
-1
h
-1
/µE m
-2
s
-1
)
P
chl
max
(mg C (mg chl a
-1
) h
-1)

α
chl
((mg C (mg chl a
-1
) h
-1
)
/
(µE m
-2
s
-1
))
E
k
(µmol quanta m
-2
s-1)

N
2
fixation Units:
P
b
max
(fmol N cell
-1
h
-1
)
α
b
((fmol N cell
-1
h
-1
)

/ (µE m
-2
s
-1
))
P
chl
max
(mg N (mg chl a
-1
) h
-1
)
α
chl
((µg N (mg chl a
-1
) h
-1
)

/ (µE m
-2
s
-1
))
E
k
(µmol quanta m
-2
s-1)

Table 3.3. P
max
and α for the Fe manipulation experiments. Standard error is in parentheses and n is the
number of samples.

76
C
f
:N
f
ratios were not significantly different between nutrient treatments: 16 ± 4 and 16
± 2 (normalized to cell number) and 17 ± 4 and 17 ± 1 (normalized to mg chl a) for
the 30nM and 200nM Fe cultures, respectively.

DISCUSSION
This study presents the first photosynthetron-derived measurements of P
max

and α for both CO
2
and N
2
fixation in P- and Fe-limited cultures of Trichodesmium
IMS101. The cultures from this study were successfully grown using semi-continuous
culture methods (MacIntyre and Cullen 2005) and were Fe- and P-limited as indicated
by both reduced chl a cell
-1
ratios and growth rates in the low nutrient cultures (Sunda
et al. 2005). While reduction in chl a, can be an indicator of nutrient stress (Geider et
al. 1993, Scanlan and Wilson 1999), it can also be due to variations in irradiance
(Geider et al. 1987). In these experiments, all cultures were grown in the same
incubator and were not exposed to different light intensities during growth, thus the
observed reduced chl a cell
-1
is most likely an indication of nutrient limitation.
In general, both P
max
and α for CO
2
and N
2
fixation were reduced under
nutrient limited conditions. Statistically significant reductions in all parameters were
observed when the data were normalized to cell number for both the P- and Fe-limited
cultures, but were only observed in some instances when the data were normalized to
mg chl a (Tables 3.2 and 3.3). This dependence on normalization factor had been
reported previously in Rueter (1988) when he noted that colony enrichment was not

77
detected if the data were normalized to mg chl a, but they were detected if normalized
to cell volume. Equally, Rueter et al. (1990) observed a 2-fold increase in N
2
fixation
in cultures grown at 50µM Fe compared to those grown in lower concentrations when
normalized to mg chl a, but they observed a 4-fold increase when normalized to cell
number. Phytoplankton can adjust their chl a concentrations within a cell in a matter
of hours in response to an environmental variable or stressor (Geider et al. 1987). Yet
as in the case with Trichodesmium, which has an estimated doubling time of 3 - 5 days
(Capone et al. 1994), cell biomass takes longer to respond to environmental stressors.
There is therefore a decoupling between chl a content within a cell and true biomass
concentration. With the abundance of ocean color data collected by remote sensing
and due to the relative ease in measuring chl a, it is often the normalization parameter
of choice. As demonstrated in the results from this study and in the previous studies
reported, data normalized to chl a can underestimate productivity contributions. These
are therefore not the optimal data to utilize in ecosystem models. Whenever possible,
it is recommended that data normalized to cell number (or carbon, which has more
direct relationship and less fluctuating relationship to cell number (Falkowski and
Raven 1997)), be utilized instead.
There have been several papers that have demonstrated reduced N
2
fixation in
P-limited and Fe-limited cultures (Rueter 1988, Paerl et al. 1994, Rueter et al. 1990,
Berman-Frank et al. 2001, Fu and Bell 2003a, Kustka et al. 2003, Mulholland and
Bernhard 2005, see Chapter 1 for more detailed discussion). But none have expressly

78
examined the effect of P-limitation on the CO
2
and N
2
fixation photophysiological
parameters P
max
and α by Trichodesmium. The data presented in this dissertation are
therefore the first to do so. Recently, Fu and Bell (2003b) cultured the Trichodesmium
isolate GBBTRL101 in Fe concentrations ranging from 0 – 450 nM Fe and examined
the production parameters for CO
2
fixation. The P
max
and α values reported there are
2-fold lower than the P
max
and α values reported in this study (Table 3.4). This
difference in absolute P
max
and α values between these two studies may either be due
to differences in the strains or species of the two Trichodesmium cultures or due to
methodological considerations. Carpenter et al. (1993) measured P
max
, α, E
k
for
Trichodesmium thiebautii and Trichodesmium erythraeum in the Sargasso and
Caribbean Seas. They report higher P
max
values in T. erythraeum than in T. thiebautii,
but lower α values. Additionally, E
k
was twice as deep for T. thiebautii (687m) than
for T. erythraeum (324 m). And, these are differences for two species that share the
same environment and habitat. Trichodesmium IMS101 (used in this study) was
isolated from the Atlantic Ocean in 1992, while Trichodesmium GBBTRLI101 (used
in Fu and Bell) was isolated from the Low Isles in the Northern Great Barrier Reef
Lagoon. These isolates have been in culture for a number of years and have cycled
through numerous laboratories. Thus, is most probable that the differences in absolute
values of P
max
and α are due to interspecific differences.
There were some notable methodological differences between this study and
that of Fu and Bell. Fu and Bell measured P
max
and α using an O
2
-electrode to measure

79

Table 3.4. Results of Fu and Bell (2003b) compared to the results from this
study.

CO
2
fixation units:
P
chl
max
(mg C (mg chl a)
-1
h
-1
)
α
chl
(mg C (mg chl a)
-1
h
-1
) / (µmol quanta m
-2
s
-1
)
E
k
(µmol quanta m
-2
s
-1
)
[Fe]
P
chl
max
!
chl
E
k
0 0.9 0.007 133
9 1.2 0.007 176
45 1.5 0.007 214
450 2.0 0.008 248
30 2.4 (0.4) 0.012 (0.001) 200
200 2.7 (0.6) 0.021 (0.004) 129
Fu and
Bell
(2003b)
This Study

80
O
2
evolution as a proxy for photosynthesis, whereas in this study,
14
C uptake was
measured as a proxy for CO
2
fixation. Trichodesmium has high rates of respiration
(Carpenter et al. 1993) and Mehler activity (Kana 1993). With such high rates
of O
2
consumption, the correlation between O
2
-evolution and CO
2
fixation may
therefore not be direct and O
2
-evolution may overestimate both P
max
and α.
Additionally, in Fu and Bell, the highest irradiance to which the colonies were
exposed was 300 µmol quanta m
-2
s
-1
, while in this study, cultures were exposed to
irradiances as high as 2000 µmol quanta m
-2
s
-1
. This may lead to an underestimation
of P
max
and α in my experiments as the curve fits may be biased by photoinhibition
that occurs at the higher light intensities.
Additionally, it is not clear if the cultures of Fu and Bell were (or were
intended to be) Fe-limited. Their cultures were grown for one generation in 0nM Fe
media before beginning experiments. In contrast, the cultures in this study were
grown at each nutrient concentration for 10 generations before commencing
experiments. MacIntyre and Cullen (2005) argue that this is critical for ensuring that
the cultures are indeed acclimated to the different nutrient regimes and have utilized
whatever nutrient stores they may have, as many organisms are capable of luxury
uptake and storage of nutrients. Fu and Bell were essentially measuring the effect of
Fe-enrichment on healthy populations of Trichodesmium, while this study was
examining the effect of Fe-limitation on Trichodesmium. This is a subtle, yet
important distinction.

81
These methodological considerations would should actually lead to an
overestimation of P
max
and α in Fu and Bell, however this was not observed when
their values were compared to those of this study. Therefore, it remains most
plausible that the differences in absolute values are due to differences between the two
strains of Trichodesmium.
Nonetheless, the P
max
and α values for these two studies fall within the same
order of magnitude and both demonstrate increased P
max
and α values in cultures
exposed to higher media Fe concentrations. Ultimately, to advance the modeling of
the effect of nutrient limitation, it is most critical for these studies to describe the
effect and response of organisms to nutrient limitation. These two studies are then a
good start towards defining this CO
2
fixation in Trichodesmium.
Choice of nutrient concentration is an important consideration when designing
nutrient limitation experiments. In the experiments described in this study, P
max
and α
were reduced by differing amounts depending on which nutrient was limiting (Table
3.5). For instance, there was a 6.7-fold reduction in the P
b
max
of the P-limited culture
experiments compared to a ~1.6-fold reduction in the Fe-limited culture experiments.
However, there was a 150-fold difference between high and low P concentrations (0.2
vs. 30 µM), compared to a 7-fold difference in Fe concentrations (30 vs. 200 nM). In
normalizing the difference in P
b
max
between high and low cultures to the difference in
the nutrient concentrations, the trend reverses, and one observes a ~4% reduction in
the P
b
max
of the P-limited cultures compared to a ~23% reduction in the Fe-limited

82

Nutrient
!
[nutrient]
!(P
b
max
)
!(P
b
max
) /
! [nutrient]
!("
b
)
!("
b
) / !
[nutrient]
!(P
b
max
)
!(P
b
max
) /
! [nutrient]
!("
b
)
!("
b
) / !
[nutrient]
P 150 6.7 4% 7.5 5% 8.3 6% 2 1%
Fe 7 1.6 23% 2.3 33% 1.3 19% 1.8 26%
Nutrient
!
[nutrient]
!(P
chl
max
)
!(P
chl
max
) /
! [nutrient]
!("
chl
)
!("
chl
) / !
[nutrient]
!(P
chl
max
)
!(P
chl
max
) /
! [nutrient]
!("
chl
)
!("
chl
) / !
[nutrient]
P 150 1.8 1% 2 1% 3.7 2% 2.7 2%
Fe 7 1.1 16% 1.8 26% 1.1 16% 1.6 23%
N2 fixation CO
2
fixation
Table 3.5. This table lists the differences between high and low experimental nutrient concentrations; differences
between average P-E parameter values (P
max
and α) for both CO
2
and N
2
fixation and normalized to both mg chl a
and to cell number; and, the differences in parameter values normalized to the difference in nutrient concentration.

83
cultures. These results imply that Fe-limitation had a more reductive effect on both
CO
2
and N
2
fixation than P-limitation. Yet, from the data presented here, it cannot be
determined if the differences observed are truly due to differential physiological
limitations imposed by P- and Fe-limitation or if it they are caused by the wide
difference between nutrient concentrations.
This becomes more relevant as these data become utilized in global ecosystem
models. In the Moore et al. (2002a) model, CO
2
fixation is regulated by whether or not
a cell is able to take up enough of a particular nutrient to attain their maximum cell
quota (this was first described in Geider et al. 1998 for eukaryotic phytoplankton). As
an example, Moore et al. describe a scenario in which a cell has only enough of one
nutrient to fulfill 50% of its cell quota, the model then forces the cell to fix at 50% of
its total capacity. In Figure 3.3, P
chl
max
from Fu and Bell and from this study are
plotted against the corresponding media Fe concentrations. (Because the data from
this study are approximately half of the values in Fu and Bell, the data from this study
were halved for scaling purposes.) The data can be fit by both a linear equation and a
logarithmic equation, and both yield essentially equal correlation coefficients (r
2
=0.76
and r
2
=-0.73, respectively). Thus, the relationship between nutrient availability and
production cannot be predicted. If the response of nutrient uptake in nutrient limited
cultures is linear to rate of production, then a 50% reduction in nutrient availability

84

Figure 3.3. P
chl
max
for CO
2
fixation from Fu and Bell (closed circles ) and this study
(open triangles ) plotted against media [Fe] (nM) concentration. The data are fit
with both a linear curve-fit and a logarithmic curve fit and correlation of fit (r
2
) are
presented in the legend.

r
2
= 0.76
r
2
= 0.73

85
would yield a 50% in production. However, if the response is non-linear (perhaps as
in a Monod-type response), the reduction in production could be less than 50% and
therefore the model would yield an underestimation of production. It is critical
therefore that more experiments be conducted with a wider range of nutrient
concentrations to accurately determine the true relationship between P
max
and nutrient-
limitation.

Summary
This study provides the first descriptions of the effects of P-limitation on the
photophysiological parameters for CO
2
and N
2
fixation in Trichodesmium; provides
the first descriptions of the effects of Fe-limitation on N
2
fixation; and, it contributes
to the observations of the effect of Fe-limitation on CO
2
fixation. The
photophysiological parameters of CO
2
and N
2
fixation were found to be reduced under
nutrient-limiting conditions. As yet, there have been no attempts to model the effect of
nutrient limitation on N
2
fixation specifically. These results provide a critical first step
for ecosystem modelers to describe nutrient limitation in N
2
fixation.

86
Chapter 4

Temporal Segregation of CO
2
and N
2
Fixation in Trichodesmium IMS-101 Using
Nanometer Resolution Secondary Ion Mass Spectrometry (nanoSIMS)

ABSTRACT
Trichodesmium is an enigmatic organism in that it is a nonheterocystous
cyanobacteria that can fix both CO
2
and N
2
concurrently during the day. As O
2
is
inhibitory to nitrogenase, one of the key enzymes in N
2
fixation, it remains a mystery
how this is accomplished. Several current hypotheses suggest that CO
2
and N
2

fixation are both spatially and temporally segregated. Using the stable isotopes
15
N
and
13
C to trace N
2
and CO
2
fixation, respectively, data generated from nanometer
resolution secondary ion mass spectrometry (nanoSIMS) were used to determine if
there is indeed spatial and/or temporal segregation. From the data, it is apparent that
CO
2
and N
2
fixation are temporally segregated with CO
2
fixation maximal in the
morning and N
2
fixation maximal in the afternoon. Whether or not spatial segregation
occurs remains undetermined. In addition, these analyses also provided unique views
into the movement and location of newly fixed C and N within a Trichodesmium cell.

INTRODUCTION
Diazotrophic cyanobacteria are capable of both CO
2
and N
2
fixation. Yet, the
O
2
produced from CO
2
fixation is inhibitory to nitrogenase, the key enzyme in N
2

87
fixation. Diazotrophs have therefore developed different methods to protect
nitrogenase from the O
2
evolved during photosynthesis. Heterocystous cyanobacteria
which have terminally differentiated cells (heterocysts) with thick walls and reduced
PS II and rubisco activity (Wolk et al. 1994), spatially segregate the two processes: N
2

fixation occurs in the heterocysts while oxygenic photosynthesis and CO
2
fixation
occurs in vegetative cells. Some non-heterocystous cyanobacteria temporally
segregate the processes by fixing CO
2
during the day and fixing N
2
at night (Gallon
1992).
The marine cyanobacterium, Trichodesmium, is unique however in that it is a
non-heterocystous cyanobacteria that apparently fixes both CO
2
and N
2
concurrently
during the day (Capone et al. 1997). Several hypotheses have been put forth to
explain this paradoxical behavior. Immunochemical studies indicate that only ~15%
of cells along a trichome contain nitrogenase and that these cells occur in clusters
termed diazocytes along a trichome (Bergman 1999, El-Shehawy et al. 2003). The
diazocytes have the same density of carboxysomes as other cells along a trichome, but
appear less granulated, have fewer gas vesicles, have fewer or no cyanophycin
granules and have additional membranes (Fredriksson and Bergman 1997, El-
Shehawy et al. 2003). These studies suggest there is spatial segregation of N
2
fixation
and CO
2
fixation along a trichome. Berman-Frank et al. (2001) demonstrated that the
two processes are also temporally segregated. In fine-scale temporal experiments
measuring both CO
2
and N
2
fixation throughout the day, Berman-Frank et al. found

88
that CO
2
fixation reaches maxima in the morning and afternoon while N
2
fixation is
maximal in the late-morning to early afternoon, whilst CO
2
fixation is down-regulated.
Using nanometer resolution Secondary Ion Mass Spectrometry (nanoSIMS),
we sought to elucidate the timing and spatial patterns of N
2
and CO
2
fixation in
diazotrophic cyanobacteria by measuring both
13
C-labeled NaHCO
3
and
15
N
2
-labeled
N
2
gas uptake in individual cells of Trichodesmium isolate IMS-101 at eight time
intervals throughout the day. Our results provide support for the hypothesis that CO
2

and N
2
fixation in Trichodesmium may be temporally segregated, but neither support
nor refute the diazocyte hypothesis of spatial segregation.

METHODS
Subsamples of Trichodesmium isolate IMS-101 were incubated in 165ml
serum vials crimped sealed with silicone rubber stoppers to prevent gas exchange. The
cultures were inoculated with 0.07 ml of a 0.47 M
13
C- labeled NaHCO
3
(final 4.8% of
total DIC) and 0.3 ml of 99% enriched
15
N-labeled N
2
gas (final enrichment 16% of
total N
2
) as tracers for CO
2
and N
2
fixation, respectively.
Experiments were initiated at 10am. Eight time points were taken at 0 min, 15
min, 30 min, 1hr, 2hrs, 4 hrs, 8 hrs, and 24 hrs. For each time point 1ml of sample
was preserved in 0.1 ml 25% glutaraldehyde for SIMS analysis.
Secondary ion mass spectrometry (SIMS) was performed at Lawrence
Livermore National Laboratory using a Cameca NanoSIMS 50. A ~1 pA Cs+ primary
beam was focused to a nominal spot size of ~100 nm and stepped over the sample in a

89
256 x 256 pixel raster to generate secondary ions for quantification. Dwell time was 1
ms/pixel, and raster size was 5 to 10 µm square. The secondary mass spectrometer was
tuned for ~6800 mass resolving power to resolve isobaric interferences. Five
secondary ions [
12
C,
13
C,
12
C
14
N,
12
C
15
N and
31
P] were detected in simultaneous
collection mode by pulse counting on electron multipliers to generate 10 to 20 serial
quantitative secondary ion images (layers) on the same spot. Samples were also
imaged simultaneously by secondary electrons detected by a photomultiplier.
The data were processed as quantitative isotopic ratio images using L’Image
software. The data were corrected for detector deadtime and image shift from layer to
layer. Each cell was defined as a region of interest (ROI), and the isotopic composition
for each ROI was calculated by averaging over all of the replicate layers. SIMS data
were normalized to sample splits run on our CF-IRMS. For each cell, ~1400 spots
were analyzed. The cell data presented are therefore the average of these 1400 spots.
Depending on biomass concentrations, 50 - 80 ml samples were filtered onto
pre-combusted GF/F filters (25 mm) and dried for isotope ratio mass spectrometry.
Samples were run at the USC Stable Isotope Facility in continuous flow mode on a
Micromass IsoPrime mass spectrometer interfaced with a EuroVector elemental
analyzer (samples run by T. Gunderson).

RESULTS
NanoSIMS provided the unique opportunity to study the chemical composition
within individual cells in unprecedented detail. Figure 4.1a is a composite image of

90
several TEM images from a Trichodesmium trichome after 8 hours of incubation with
enriched
13
CO
2
and
15
N
2
. Figures 4.1b - d present composite images generated by
nanoSIMS, which show relative concentrations for δ
15
N (Figure 4.1b), δ
13
C (Figure
4.1c) and
31
P:
12
C ratios (Figure 4.1d), within individual cells along the same trichome.
In Figure 4.1b, discrete discernible hotspots of
15
N/
14
N are likely cyanophycin
granules, nitrogen-rich storage products consisting of a polymer of asparagine and
aspartate found in many cyanobacteria (Simon 1987, Sherman et al. 2000), including
Trichodesmium (Fredriksson and Berman 1997). These hotspots correspond to
vacuoles visible in the TEM images (see arrows between Figures 4.1a and 4.1b for
examples). In both experiments, the abundance of
15
N hotspots increased with
increasing incubation time suggesting that these granules appear throughout the course
of the day as more N
2
is fixed and are then consumed overnight after the cell has
stopped fixing new N
2
(data not shown). High concentrations of
13
C are also associated
with the
15
N hotspots, but the enrichment is not as apparent (Figure 4.1c). Other
vacuoles are present in the TEM images that are not enriched with
15
N but exhibit high
31
P:
12
C ratios (Figure 4.1d). These may be either polyphosphate granules,
carboxysomes enclosed in protein shells, or other storage vacuoles. Spatial
heterogeneity throughout an individual cell was also observed; as the nanoSIMS
analyzed sections deeper into the cell, hotspots appear and disappear (Figures 4.2a-d).
In addition to these detailed views of the isotopic composition of granules
within individual cells, rates for CO
2
and N
2
fixation were also determined. δ
13
C and
δ
15
N increased in all cells along the trichomes throughout the day (Figures4.3a-d).

91

Figure 4.1a-d. Composite TEM images (A) and nanoSIMS images of δ
15
N (B), δ
13
C (C) and P:C ratios (D)
along the same trichome (from experiment 3) after 8 hours incubation. Scale bar is located to the right of
the images. Arrows link
15
N hotspots to granules visible in the TEM image; these are thought to be
cyanophycin granules.

92

Figure 4.2a-d.
15
N/
14
N "hotspots" that appear and disappear within an IMS-101 cell as one travels deeper into the
cell. For instance, observe that the “hotspot” on the right in Figure 4.2a disappears in Figure 4.2b as a new
"hotspot" begins to emerge on the left, which persists and intensifies through Figures 4.2c and 4.2d.

A
B
C D

93

Figures 4.3a-d. Uptake in individual cells along trichomes at increasing time intervals for: a)
13
C uptake (experiment 1),
b)
15
N uptake (experiment 1), c)
13
C uptake (experiment 2), and d)
15
N uptake (experiment 2). Because each time point
required destructive sampling, each time point measures both
13
C and
15
N in the same trichome. But different trichomes
were analyzed from one time point to the next. i.e. in (A) and (B), data from the 2 hour time point are from the same
trichome, however data from the 8 hour time point are from an entirely different trichome.

A
B

94

Figures 4.3a-d. Continued.

C
D

95
0
0.05
0.1
0.15
0
0.005
0.01
0.015
0.02
0.025
0.03
0 5 10 15 20 25
15N uptake (exp 1)
15N uptake (exp 2)
13C uptake (exp 1)
13C uptake (exp 2)
15
N uptake (atom % excess l
-1
h
-1
)
13
C uptake (atom % excess l
-1
h
-1
)
Time (h)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0
0.004
0.008
0.012
0.016
0.02
0.024
0.028
0.032
0 5 10 15 20 25
15N uptake (exp 1)
15N uptake (exp 2)
13C uptake (exp 1)
13C uptake (exp 2)
15
N uptake (atom % excess cell
-1
min
-1
)
13
C uptake (atom % excess cell
-1
min
-1
)
Time (h)
A
B
C
D
Figures 4.4a-b. Average rates of
15
N (closed circles and triangles) and
13
C
(open circles and triangles) uptake in cells along a trichome of
Trichodesmium from (A) nanoSIMS data and (B) IRMS data.

A
B

96

Rates of CO
2
and N
2
uptake were determined both directly using parallel
samples harvested analyzed by isotope ratio mass spectrometry (Figure 4.4b, Capone
and Montoya 2001) as well as by averaging the per cell enrichment in nanoSIMS
analysis along a trichome for each time point (Figure 4.4a). Results from both
approaches indicate that CO
2
fixation rates were highest in the morning while N
2

fixation rates were highest in the afternoon (Figure 4.4a-b).

DISCUSSION

The theory of temporal segregation
N
2
fixation in Trichodesmium is chronically inhibited by O
2
to some degree
(Saino and Hattori 1982). However, Trichodesmium clearly exhibits nitrogenase
activity during the light period (Capone et al. 1990). As proposed by Berman-Frank et
al. (2001), temporal segregation of CO
2
and N
2
fixation may occur through the down
regulation of photosystem II (PSII), which provides a window during which O
2

production is reduced, stray O
2
is consumed and N
2
fixation rates increase. However,
Berman-Frank et al. (2001) also reported a subsequent resurgence of photosynthesis in
the latter part of the light period. As a result, N
2
fixation activity slowed. While the
initial downregulation of photosynthesis and the upregulation of N
2
fixation was
observed, the ensuing upregulation of photosynthetic activity later in the day was not
(Figure 4.4a-b); the temporal resolution of this study may have been insufficient to

97
observe this. In spite of this, the results from this study provide new evidence that
corroborates the hypothesis of temporal segregation of CO
2
and N
2
fixation within
individual cells along Trichodesmium IMS-101 trichomes.

The theory of spatial segregation
Despite the increasing concentrations of newly fixed
13
C and
15
N in all cells
along a trichome throughout the day (Figures 4.3a-d), it cannot be determined from
these data if all cells were indeed fixing CO
2
and N
2
simultaneously. Wolk et al.
(1974), using autoradiography with
13
N-N
2
gas demonstrated in the cyanobacteria
Anabaena variabilis that the N
2
fixing heterocyst cell was able to assimilate, fix and
redistribute fixed N metabolites to the neighboring vegetative cells in less than 1.5
minutes. Moreover, in experiments conducted on the heterocystous cyanobacteria
Anabaena oscillariodes, after 4 hours of incubation with enriched
15
N
2
tracer, the
heterocysts were in fact depleted in
15
N/
14
N relative to the vegetative cells (Popa et al.
2007, in press) despite the fact that the heterocysts are the indisputable sites of N
2

fixation. It was only after 8 hours of incubation, once the vegetative cells had
completed their diel growth cycle and had reduced need of newly fixed N, that
accumulation of
15
N in the heterocysts was observed (Popa et al. 2007, in press).
Thus, presence of newly fixed isotope at a location in a cell does not necessarily
indicate that that is where the N
2
fixation occurs.
In this study, the first time point after incubation was 15 minutes and isotope
concentrations were not significantly different from the time zero concentrations (data

98
not shown); after two hours of incubation, isotope accumulations became apparent and
were significantly higher than initial background concentrations (Figures 4.3a-d). If
the Wolk model for A. variabilis described above holds for Trichodesmium IMS-101,
two hours would provide ample time for the cells to fix and redistribute any newly
fixed N to neighboring potentially non-fixing cells thus dispersing the
15
N signal
throughout all cells along the trichome. Thus, although the images appear to suggest
that all cells fix both CO
2
and N
2
, with this potentially rapid redistribution of
13
C and
15
N along a trichome after two hours, it is not possible to discern if all cells were
fixing both CO
2
and N
2
or if only a subset of cells were fixing N
2
while the others
fixed CO
2
. Pulse-chase tracer experiments with higher enrichment and a shorter time
course may help resolve this question.
It is important to note, in this study only a small subset of samples were run
(one trichome per time point, for a total of two complete experiment runs). In
experiment 1, the trichomes were analyzed just below the surface of the cell wall; i.e.
they were "burned in" within nanometers of the cell surface. To the contrary, in
experiment 2, trichomes were burned in much deeper, however the depth to which
they were burned in was not uniform (J. Pett-Ridge, pers. comm.). Figures 4.5a-d
show examples of SEM images of cell after they were burned in. It is clear from these
images that the cells were not uniformly sampled, but it is not clear if this affects the
results. As discussed above, there is remarkable concurrence in both the ratios of
13
C
and
15
N uptake (Figures 4.3a-d) and in the rates and timing of uptake (Figure 4.4b),
however there are also different patterns in
15
N hotspot accumulations between the

99
two experiments. If a section with many hotspots was sampled this could lead to an
overestimation of N fixed if compared to a section in which few hotspots were present.
More analysis is needed to determine if the depth to which a sample is burned-in leads
to significant under- or overestimation of the presence of newly fixed C or N.

SUMMARY
This study provides further support to the theory that the maxima in CO
2
and
N
2
fixation by Trichodesmium are temporally separated during the day with the
highest rates of CO
2
fixation occurring in the morning and highest rates of N
2
fixation
occurring later in the day. Our data do not permit us to determine if there is also
spatial segregation of CO
2
and N
2
fixation along a trichome. Studies are currently
underway to combine TEM work with more nanoSIMS measurements to probe this
question in more detail.

100
Figures 4.5a-d. SEM images demonstrating burning-in heterogeneity.

A B

C
D

101

Chapter 5

Global Significance, Contributions and Synthesis

Contributions and global significance
Trichodesmium is a globally significant marine cyanobacterium (Capone et al.
1997, Karl et al. 1997, Karl et al. 1998, Carpenter et al. 2004, Capone et al 2005).
Though Trichodesmium is not the only significant source of new N provided by N
2

fixation (Zehr et al. 2001, Montoya et al. 2004), it is the most conspicuous and has
been the subject of many studies over the last two decades. As such, global ecosystem
modelers have begun incorporating N
2
fixation by Trichodesmium into their models,
most commonly by utilizing the empirically derived parameters, P
max
and α. As these
parameters are essential for ecosystem models, it is critical, therefore, to develop
accurate and precise estimates of these parameters. Previous estimates of P
max
and α
for CO
2
fixation of Trichodesmium are relatively limited in their spatial extent and
have not considered the effects of nutrient limitation on these key parameters. Lewis et
al. 1980, Li et al. 1988, Carpenter et al. 1993 and Carpenter and Roenneberg 1995
have reported values of P
max
and α for CO
2
fixation by Trichodesmium in the N.
Atlantic (Table 2.1). To my knowledge, there have been no previous studies that have
considered these values for populations in the N. Pacific subtropical gyre.
Therefore an important contribution of this dissertation is the development of a
photosynthetron-derived data set of P
max
and α for CO
2
and N
2
fixation for
C D

102
Trichodesmium populations from both the tropical N. Atlantic and N. Pacific. In
Chapter 2, I report the first values for N
2
fixation by Trichodesmium populations in
both the tropical N. Atlantic and the tropical N. Pacific, as well as the first reported
values for CO
2
fixation in the tropical N. Pacific and present refined estimates for the
tropical N. Atlantic. For CO
2
fixation there were no significant difference between
basins for P
max
or α. However, for N
2
fixation, both P
max
and α were significantly
reduced in the Atlantic compared to the Pacific. I report that this may be due to both
potential P-stress and deeper mixed layer depths in the Atlantic compared to the
Pacific. I also demonstrate that photosynthetron-derived rates may provide more
accurate estimates of P
max
and α compared to the standard five irradiance, on-deck
experiments.
Additionally, as described in Chapters 1 and 3, large portions of the subtropical
N. Atlantic and N. Pacific are N-limited (Falkowski 1997, Tyrell 1997, Wu et al.
2000, Sañudo-Wilhelmy et al. 2001). Though N-limitation is not a concern for the N
2
-
fixing Trichodesmium, P- and Fe-limitation are. Ecosystem models are beginning to
incorporate the effects of nutrient limitation on global C and N cycling (Moore et al
2002a). Thus, accurately characterizing the effect of nutrient limitation on P
max
and α
for both CO
2
and N
2
fixation is critical for effective modeling. In the studies
presented in this dissertation, I have successfully cultured P- and Fe-limited
populations of Trichodesmium IMS101 and I have demonstrated that both P
max
and
α are reduced under P- and Fe-limiting conditions. These experiments refine the
results of a previous study that describes the effect of Fe-limitation on CO
2
fixation

103
and present the first results of the effect of P-limitation on CO
2
fixation. Moreover,
these are also the first to provide any observations on the effect of both P- and Fe-
limitation on the photophysiological parameters of N
2
fixation in Trichodesmium.
Finally, since intensive study began on Trichodesmium, researchers have been
interested in its ability to fix both CO
2
and N
2
concurrently, during the day, without
heterocysts. In the nanoSIMS tracer experiments described in Chapter 4, I provide
new evidence to corroborate the theory that Trichodesmium temporally segregate CO
2

and N
2
fixation (Berman-Frank et al. 2001). This dissertation therefore advances our
understanding of how Trichodesmium is able to undertake two potentially conflicting
processes, without the aid of specialized cells.

Synthesis
The observations from the field studies described in Chapter 2 inspired the
nutrient limitation experiments described in Chapter 3. Given that the tropical and
subtropical oceans are hypothesized to be P- and Fe-limiting for diazotrophic
cyanobacteria, laboratory experiments isolating the effect of single nutrient limitation
on the photophysiological parameters of Trichodesmium cultures may provide clues to
the observed interbasin differences. In the field studies, no statistically significant
correlations between water column dissolved iron, soluble reactive phosphorous or
mixed layer depth and P
max
and α for CO
2
and N
2
fixation were observed. However, at
stations where photosynthetron experiments were conducted, SRP concentrations were
generally lower in the Atlantic than in the Pacific, as were P
max
and α for N
2
fixation.

104
Therefore, it could be hypothesized that the populations at the Atlantic stations were
P-stressed such that their CO
2
fixation capacity was not affected, but their N
2
fixation
capacity was reduced. The laboratory experiment results from Chapter 3 do therefore
provide some further insights into the differences observed in the field experiments of
Chapter 2.
A potentially interesting and logical next step in the study of how nutrient-
limitation affects the photophysiological parameters, would be to study the effect of
multiple concurrent nutrient limitation. Mills et al. (2004) describe experiments
conducted with phytoplankton populations of the eastern subtropical N. Atlantic in
which they observed co-limitation of CO
2
fixation by N, P and Fe. Similarly, nitrogen
fixation was at times co-limited by P and Fe. Co-limitation of P and Fe by
Trichodesmium could be easily tested in the laboratory and would provide further
insights into the field studies.
Finally, the results from the nanoSIMS experiments, demonstrating temporal
decoupling of CO
2
and N
2
fixation, can provide insight into the CO
2
fixation to N
2

fixation (C
f
:N
f
) ratios presented in Chapters 2 and 3. The lab and field experiments
conducted in this dissertation were usually begun in the morning hours (around 10am)
and were completed by noon or 1pm. Therefore, my experiments coincided with the
time when CO
2
fixation was maximal and N
2
fixation was reduced. This may provide
another explanation (in addition to those discussed extensively in Chapter 2) for why
C
f
:N
f
ratios measured in the study, which ranged from 18 - 108, were greater than
those described by the Redfield ratio (generally ~6.6.:1). This may be an important

105
consideration for other researchers who also conduct experiments at sea and who run
their experiments in the early part of the day.

106
BIBLIOGRAPHY

Behrenfeld, M.J., E. Maranon, D.A. Siegel, S.B. Hooker, S.B. (2002)
Photoacclimation and nutrient-based model of light-saturated photosynthesis for
quantifying oceanic primary production. Marine Ecology Progress Series 228:103-
117.

Bergman, B. (1999) Distribution of nitrogenase in the marine non-heterocystous
cyanobacterium Trichodesmium: a review. In: Marine Cyanobacteria (L.Charpy and
A.W.D. Larkum, eds), Bulletin de l'Institut oceanographique, Monaco, no. 19, pp.
223-228.

Bergman, B. and E.J. Carpenter (1991) Nitrogenase confined to randomly distributed
trichomes in the marine cyanobacterium Trichodesmium thiebautii. Journal of
Phycology 27:158-165.

Bergman, B., C. Fedriksson, S, Janson, E.J. Carpenter, H. Paerl, C. Lugomela (1996)
Trichodesmium has cells specialized for nitrogen fixation but lacks heterocysts.
Proceedings of the "2nd European Nitrogen Fixation Conference", Pozham, Poland,
Sept. 1996.

Bergman, B., J.R. Gallon, A.N. Rai, L.J. Stal (1997) N2 fixation by non-heterocystous
cyanobacteria. FEMS Microbiological Review 19:139-185.

Berman-Frank, I., P. Lundgren, Y. Chen, H. Kupper, Z. Kolber, B. Bergman, P.
Falkowski (2001) Segregation of nitrogen fixation and oxygenic photosynthesis in the
marine cyanobacterium Trichodesmium. Science 294:1534- 1537.

Blackman, F.F. (1905) Optima and limiting factors. Annals of Botany 9:281-295.

Bouman, H.A., T. Platt, G.W. Kraay, S. Sathyendranath, B.D. Irwin (2000) Bio-
optical properties of the subtropical N. Atlantic I. Vertical variability. Marine Ecology
Progress Series 200:3-18.

Breithbarth, E. (2004) Ecophysiology of the marine cyanobacterium Trichodesmium.
Dissertation.

Burns, J.A., J.P. Zehr, J.P. Montoya, A.B. Kustka, and D.G. Capone (2006) Effect of
EDTA additions on Natural Populations of Trichodesmium spp. (Cyanobacteria)
Populations. Journal of Phycology 42:900-904.

107
Capone, D.G., J.A. Burns, J.P. Montoya, A. Subramaniam, C. Mahaffey, T.G.
Gunderson, A.F. Michaels, E.J. Carpenter (2005) Nitrogen fixation by Trichodesmium
spp.: An important source of new nitrogen to the tropical and subtropical North
Atlantic Ocean. Global Biogeochemical Cycles 19, GB2024.

Capone, D.G., M.D. Ferrier, E.J. Carpenter (1994) Amino acid cycling in colonies of
the planktonic marine cyanobacterium Trichodesmium thiebautii. Applied
Environmental Microbiology 60(11):3989-3995.

Capone, D.G. and J.P. Montoya (2001) Nitrogen fixation and denitrification. Methods
in Microbiology 30:501-515.

Capone, D.G., J.M. O'Neil, J. Zehr, E.J. Carpenter (1990) Basis for diel variation in
nitrogenase activity in the marine planktonic cyanobacterium Trichodesmium
thiebautii. Applied Environmental Microbiology 56:3532-3536.

Capone, D.G., A. Subramaniam, J.P Montoya, M. Voss, C. Humborg, A.M. Johansen,
R.L. Siefert, E.J. Carpenter (1998) An extensive bloom of the N2-fixing
cyanobacterium Trichodesmium erythraeum in the central Arabian Sea. Marine
Ecology Progress Series 172:281-292.

Capone, D.G., J. Zehr, H. Paerl, B. Bergman, and E.J. Carpenter (1997)
Trichodesmium: a globally significant marine cyanobacterium. Science 276:1221-
1229.

Carpenter, E.J. (1983) Physiology and ecology of marine Oscillatoria
(Trichodesmium). Marine Biology Letters 4:69-85.

Carpenter, E.J., J.P. Montoya, J. Burns, M.R. Mullholland, A. Subramaniam, D.G.
Capone (1999) Extensive bloom of a N2-fixing diatom/cyanobacterial association in
the tropical Atlantic Ocean. Marine Ecology Progress Series 185:273-283.

Carpenter, E.J., J.M. O’Neil, R. Dawson, D.G. Capone, P.J.A. Siddiqui, T.
Roenneberg, B. Bergman (1993) The tropical diazotrophic phytoplankter
Trichodesmium: biological characteristics of two common species. Marine Ecology
Progress Series 95:295-304.

Carpenter, E.J. and C.C. Price (1976) Marine Oscillatoria (Trichodesmium):
explanation for aerobic nitrogen fixation without heterocysts. Science 191: 1278-1280.

Carpenter E.J. and T. Roenneberg (1995) The marine planktonic cyanobacteria
Trichodesmium spp.:photosynthetic rate measurements in the SW Atlantic Ocean.
Marine Ecology Progress Series 118:267-273.

108

Carpenter, E.J. and K. Romans (1991) Major role of the Cyanobacterium
Trichodesmium in nutrient cycling in the North Atlantic Ocean. Science 254: 1356-
1358.

Carpenter, E.J., A. Subramaniam, D.G. Capone (2004) Biomass and primary
productivity of the cyanobacterium Trichodesmium spp. in the North tropical Atlantic
Ocean. Deep-Sea Research I 51:173-203.

Chen, Y., B. Dominic, M.T. Mellon, J.P. Zehr (1998) Circadian rhythm of nitrogenase
gene expression in the diazotrophic filamentous nonhetercystous cyanobacterium
Trichodesmium sp. Strain IMS 101. Journal of Bacteriology 180:3598-3605.

Chen, Y. B., J. P. Zehr, M. M. Mellon (1996) Growth and nitrogen fixation of the
diazotrophic filamentous nonheterocystous cyanobacterium Trichodesmium sp IMS
101 in defined media: Evidence for a circadian rhythm. J. Phycol. 32: 916–923

Coles, V.J., R.R. Hood, M. Pascual, D.G. Capone (2004) Modeling the impact of
Trichodesmium and nitrogen fixation in the Altantic Ocean. Journal of Geophysical
Research 109:C06007.

Corno, G., R.M. Letelier, M.R. Abbott (2005) Assessing primary production
variability in the North Pacific Subtropical Gyre: A Comparison of Fast Repetition
Rate Fluorometry and 14C Measurements. Journal of Phycology 42:51-60.

Cullen, J. J., M. R. Lewis, C. O. Davis, and R. T. Barber (1992), Photosynthetic
characteristics and estimated growth rates indicate grazing is the proximate control of
primary production in the equatorial Pacific, Journal of Geophysical Research 97(C1),
639–654.

Doney, S.C., D.M. Glover, R.G. Najjar (1996) A new coupled, one-dimensional
biological-physical model for the upper ocean; Applications to the JGOFS Bermuda
Atlantic Time-Series Study (BATS) site. Deep Sea Research (II) 43:591-624.

Dyrhman, S.T., E.A. Webb, D.M. Anderson, J.W. Moffett, J.B. Waterbury (2002)
Cell-specific detection of phosphorus stress in Trichodesmium from the western North
Atlantic. Limnology and Oceanography 47:1832-1836.

El-Shehawy, R., C. Lugomela, A. Ernst, and B. Bergman (2003) Diurnal expression of
hetR and diazocyte development in the filamentous non-heterocystous cyanobacterium
Trichodesmium erythraeum. Microbiology Sgm. 149:1139-1146.

109
Falkowski P.G., (1994) The role of phytoplankton photosynthesis in global
biogeochemical cycles. Photosynthesis Research 39:235-258

Falkowski, P.G. (1997) Evolution of the nitrogen cycle and its influence on the
biological sequestration of CO
2
in the ocean. Nature 387:272- 275.

Falkowski, P.G., R.M. Greene, R.J. Geider (1992) Physiological limitations on
phytoplankton productivity in the ocean. Oceanography 5(2):84-91.

Falkowski, P.G. and J.A. Raven (1997) Aquatic Photosynthesis. Blackwell Science,
Massachusetts, 375pp.

Fasham, M.J.R., H.W. Ducklow, S.M. McKelvie (1990) A nitrogen-based model of
plankton dynamics in the oceanic mixed layer. Journal of Marine Research 48:591-
639.

Fennel, K., Y.H. Spitz, R.M. Letelier, M.R. Abbott, D.M. Karl (2002) Deep Sea
Research II 49:149-174.

Foy, R.H. and C.E. Gibson (1982) Photosynthetic characteristics of planktonic blue-
green algae: changes in photosynthetic capacity and pigmentation of oscillaotria
redekei van Goor under high and low light. British Journal of Phycology 18:267-273.

Fredriksson, C. and B. Bergman (1995) Nitrogenase quantity varies diurnally in a
subset of cells within colonies of the non-heterocystous cyanobacteria Trichodesmium
spp. Microbiology 141:2471-2478.

Fredriksson, C. and B. Bergman (1997) Ultrastructural characterization of cells
specialized for nitrogen fixation in a non-heterocystous cyanobacterium,
Trichodesmium spp. Protoplasma 197:76-85.

Fu, F.-X. and P.R.F. Bell (2003a) Factors affecting N
2
fixaiton by the cyanobacterium
Trichodesmium sp. GBRTRL101. FEMS Microbiology Ecology 45:203-209.

Fu, F.-X. and P.R.F. Bell (2003b) Growth, N
2
fixation and photosynthesis in a
cyanobacterium, Trichodesmium sp., under Fe stress. FEMS Microbiology Ecology
45:203-209

Fu, F-X, Y. Zhang, F.R.F. Bell and D.A.Hutchens (2005) Phosphate uptake and
growth kinetics of Trichodesmium (Cyanobacteria) isolates from the North Atlantic
Ocean and the Great Barrier Reef, Australia. Journal of Phycology 41:62-73.

110
Gallon, J.R. (1992) Reconciling the incompatible: N
2
fixation and O
2
. Tansley Review
No. 44. New Phytologist 122:571-609.

Geider, R.J. (1987) Light and temperature dependence of the carbon to chlorophyll a
ration in microalgae and cyanobacteria: implications for physiology and growth of
phytoplankton. New Phytologist 106:1-34.

Geider, R.J., J. LaRoche, R.M. Greene, M. Olaziola (1993) Response of the
photosynthetic apparatus of Phaeodactylum tricornutum (Bacillariophycaea) to nitrate,
phosphate, or iron starvation. Journal of Phycology 29:755-766.

Geider, R.J., H.L. MacIntyre, T.M. Kana (1997) Dynamic model of phytoplankton
growth and acclimation: response of the balanced growth rate and the chlorophyll a:
carbon ration to light, nutrient-limitation and temperature. Marine Ecology Progress
Series 148:187-200,

Glazer, A.N. (1999) Cyanobacterial photosynthetic apparatus: an overview. Marine
Cyanobacteria suppl. Bulletin de l'Institut Oceanographic, Monaco, pp.419-434.

Glibert, P.M. and D.A. Bronk (1994) Release of dissolved organic nitrogen by marine
diazotrophic cyanobacteria, Trichodesmium spp. Applied and Environmental
Microbiology 60:3996-4000.

Grossman, A.R., M.R. Schaefer, G.G. Chiang, J.L. Collier (1994) The responses of
cyanobacteria to environmental conditions: light and nutrients. In: The Molecular
Biology of Cyanobacteria (D.A. Bryant, ed), Kluwer Academic Publications,
Dordrecht, pp.641-675.

Henley, W.J. (1994) Measurement and interpretation of photosynthetic light response
curves in algae in the context of photoinhibition and diel changes. Journal of
Phycology 29:729-739.

Hood, R.R. (1995) Light response of phytoplankton in the South Atlantic Ocean:
Interpretation of observartions and application to remote sensing. Journal of
Geophysical Research 100:10927-10942.

Hood, R.R., M.A. Abbott, and A. Huyer (1991) Phytoplankton and photosynthetic
light response in the coastal transition zone off Northern California in June 1987.
Journal of Geophysical Research 96:14769-14780.

Hood, R.R., N.R. Bates, D.G. Capone, D.B. Olson (2001) Modeling the effect of
nitrogen fixation on carbon and nitrogen fluxes at BATS. Deep-Sea Research II
48:1609-1648.

111

Hood, R.R., V.J. Coles, D.G. Capone (2004) Modeling the distribution of
Trichodesmium and nitrogen fixation in the Atlantic Ocean. Journal of Geophysical
Research 109:C06006.

Hood, R.R., E.A. Laws, R.A. Armstrong, N.R. Bates, C.W. Brown, C.A. Carlson, F.
Chai, S.C. Doney, P.G. Falkowski, R.A. Feely, M.A.M. Friedrichs, M. R. Landry,
J.K. Moore, D.M. Nelson, T. L. Richardson, B.Salihoglu, M. Schartau, D.A. Toole,
J.D. Wiggert (2006) Pelagic functional group modeling: Progress, challenges, and
prospects. Deep-Sea Research 11 53:459-512.

Hood, R.R., A. Subrmaniam, L.R. May, E.J. Carpenter, D.G. Capone (2002) Remote
estimation of nitrogen fixation by Trichodesmium. Deep-Sea Research II 49:123-147.

Holl, C.M. and J.P. Montoya (2005) Interactions between nitrate uptake and nitrogen
fixation in continuous cultures of the marine diazotroph Trichodesmium
(cyanobacteria). Journal of Phycology 41:1178-1183.

Jassby, A.D. and R. Platt (1976) Mathematical formulation of the relationship between
photosynthesis and light for phytoplankton. Limnology and Oceanography 21:540-
541.

Kana, T.M. (1993) Rapid oxygen cycling in Trichodesmium thiebautii. Limnology and
Oceanography 38(1):18-24.

Kana, T.M. and P.M. Glibert (1987) Effect of irradiances up to 2000µE m
-2
s
-1
on
marine Synechococcus WH7803. W. Photosynthetic responses and mechanisms.
Deep-Sea Research 34(4a):497-516.

Karl, D.M., R. Letelier, D. Hebel, L. Tupas, J. Dove, J. Christian, C. Winn (1995)
Ecosystem changes in the N. Pacific subtropical gyre attributed to the 1991-1992 El
Nino. Nature 373:230-234.

Karl, D., Letelier, R., Tupas, L., Dore, J., Christian, J., Hebel, D. (1997) The role of
nitrogen fixation in biogeochemical cycling in the subtropical North Pacific Ocean.
Nature 388:533-538.

Kirk, J.T.O (2000) Light and photosynthesis in aquatic ecosystems. Cambridge
University Press, Cambridge, 509pp.

Kromkamp, J., A. van den H., L.R. Mur (1987) Phosphorus uptake and photosynthesis
by phosphate-limited cultures of cyanobacterium Microcystis aeruginosa. British
Phycological Journal 24:347-355.

112

Kustka, A.B., S.A. Sañudo-Wilhelmy, E.J. Carpenter, D.G. Capone and J.A. Raven
(2003a) A revised estimate of the iron use efficiency of nitrogen fixation with special
reference ot the marine cyanobacterium Trichodesmium spp. (Cyanophyta). Journal of
Phycology 39:12-25.

Kustka, A.B., S. A. Sañudo-Wilhelmy, E.J. Caprenter, D.G. Capone, J.A. Burns, W.G.
Sunda (2003b) Iron requirements for dinitrogen- and ammonium-supported growth in
cultures of Trichodesmium (IMS101): Comparison with nitrogen fixation rates and
iron:carbon rations of field populations. Limnology and Oceaongraphy 48:1869-1884.

Kyewalyanga, M.N., T. Platt, S. Sathyendranath, V.A. Lutz, V. Stuar (1998) Seasonal
variations in physiological parameters of phytoplankton across the North Atlantic.
Journal of Plankton Research 20:17-42.

Lantoine, F.and J. Neveux (1999) Phycoerythrins in the sea: abundance and spectral
diversity. In: Marine Cyanobacteria (L.Charpy and A.W.D. Larkum, eds), Bulletin de
l'Institut oceanographique, Monaco, no. 19, pp. 443-450.

La Roche, J. and E. Breitbarth (2005) Importance of the diazotrophs as a source of
new nitrogen in the ocean. Journal of Sea Research 53:67-91.

Laws, E.A. and T.T. Bannister (1980) Nutrient- and light-limited growth of
Thalassiosira fluviatus in continuous culture, with implications for phytoplankton
growth in the ocean. Limnology and Oceanography 25:457-473.

Letelier R.M. and D.M. Karl (1996) Role of Trichodesmium spp. in the productivity
of the subtropical North Pacific Ocean. Marine Ecology Progress Series 133:263-273.

Lewis, M.R. , O. Ulloa, T. Platt (1988) Photosynthetic action, absorption, and
quantum yield spectra for a natural population of Oscillatoria in the North Atlantic.
Limnology and Oceanography 33(1):92-98.

Li, W.K.W., H.E. Glover, I. Morris (1980) Physiology of carbon photoassimilation by
Oscillatoria thiebautii in the Caribbean Sea. Limnology and Oceanography 25(3):447-
456.

Lin, S., Henze, S., Lundgren, P., Bergman, B., Carpenter, E.J. (1998) Whole-cell
immunolocalization of nitrogenase in marine diazotrophic cyanobacteria,
Trichodesmium spp. Applied Environmental Microbiology 64(8):3052-3058.

113
MacIntyre, H.L. and J.J. Cullen (2005) Using cultures to investigate the physiological
ecology of microalgae. In: Algal Culturing Techniques, R.A. Anderson (ed). Elsevier
Academic Press, Massachusetts, pp. 287-326.

Mague T.H., F.C. Mague, O. Holm-Hansen (1977) Physiology and chemical
composition of nitrogen-fixing phytoplankton in the Central North Pacific Ocean.
Marine Biology 41:212-227.

Mahaffey, C., A.F. Michaels, D.G. Capone (2005) The conundrum of marine N
2

fixation. American Journal of Science 305:546-595.

Mills, M.M., C. Ridame, M. Davey, J. LaRoche, R.J. Geider (2004) Iron and
phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature
429:292-232.

Montoya, J.P., C.M. Holl, J.P. Zehr, A. Hansen, T.A. Villareal, D.G. Capone (2004)
High rates of N
2
fixation by unicellular diazotrophs in the oligotrophic Pacific Ocean.
Nature 430:1027-1032.

Moore, J.K., S.C. Doney, J.A. Kleypas, D.M. Glover, I.Y, Fung (2002a) An
intermediate complexity marine ecosystem model for the global domain. Deep-Sea
Research II 49:403-462.

Moore, J.K., S.C Doney, D.M. Glover, I.Y. Fung (2002b) Iron cycling and nutrient-
limitation patterns in surface waters of the World Ocean. Deep Sea Research II
49:463-507.

Moore, J.K., S.C. Doney, K. Lindsay (2004) Upper ocean ecosystem dynamics and
iron cycling in a global three-dimensional model. Global Biogeochemical Cycles
18:GB4028.

Morel, F.M.M. (1987) Kinetics of nutrient uptake and growth in phytoplankton.
Journal of Phycology 23:137-150.

Mulholland, M.R. and P.W. Bernhart (2005) The effect of growth rate, phosphorus
concentration, and temperature on N
2
fixation, carbon fixation and nitrogen release in
continuous cultures of Trichodesmium IMS101. Limnology and Oceanography
50:839-849.

Mulholland, M.R., D.A. Bronk, and D.G. Capone (2004) Dinitrogen fixation and
release of ammonium and dissolved organic nitrogen by Trichodesmium IMS101.
Aquatic Microbial Ecology 37:85-94.

114
Mulholland, M.R. and D.G. Capone (1999) Nitrogen fixation, uptake and metabolism
in natural and cultured populations of Trichodesmium spp. Marine Ecology Progress
Series 188:33-49.

Mullholland, M.R., S. Floge, E.J. Carpenter, D.G. Capone (2002) Phosphorus
dynamics in cultures and natural populations of Trichodesmium spp. Marine Ecology
Progress Series 239:45-55.

Orcutt, K.M., F. Lipschultz, K. Gundersen, R. Arimoto, A.F. Michaels, A.H. Knap,
J.R. Gallon (2001) A seasonal study of the significance of N2 fixation by
Trichodesmium spp. at the Bermuda Atlantic Time-series Study (BATS) site. Deep-
Sea Research II 48:1583-1608.

Oschlies, A and V. Garcon (1999) Eddy-induced enhancement of primary production
in a model of the North Atlantic Ocean. Nature 394:266-269.

Paerl, H.W. (1994) Spatial segregation of CO
2
fixation in Trichodesmium spp.:
linkage to N
2
fixation potential. Journal of Phycology 30:790-799.

Paerl, H.W., J.C. Priscu, D.L. Brawner (1989) Immunochemical localization of
nitrogenase in marine Trichodesmium aggregates: relationship to N
2
fixation potential.
Applied Environmental Microbiology 55(11):2965-2975.

Paerl, H.W., L.W. Prufert-Bebout, C. Guo (1994) Iron-stimulated N2 Fixation and
Growth in Natural and Cultured Populations of the Planktonic Marine Cyanobacteria
Trichodesmium spp. Applied Environmental Microbiology 60(3):1044-1047.

Peterson, D.H., M.J. Perry, K.E. Bencala, M.C. Talbot (1987) Phytoplankton
productivity in relation to light intensity: a simple equation. Estuarine and Coastal
Shelf Science 24:813-832.

Platt T. and C.L. Gallegos (1980) Modelling primary production. In: Primary
Production in the Sea (ed. Falkowski), pgs. 339-362.

Platt, T., S. Sathyendranath, O. Ulloa, W.G. Harrison, N. Hoepffner, J. Goes (1992)
Nutrient control of phytoplankton photosynthesis in the western North Atlantic.
Nature 356:229-231.

Platt T. and C.L. Gallegos (1980) Modeling primary production. In: Primary
Production in the Sea (ed. Falkowski), Plenum Press, NY, pgs. 339-362.

115
Rueter, J.G. (1988) Iron stimulation of photosynthesis and nitrogen fixation in
Anabeana 1720 and Trichodesmium (cyanophyceae). Journal of Phycology 24:249-
254.

Rueter, J.G., K. Ohki, Y. Fujita (1990) The effect of iron nutrition on photosynthesis
and nitrogen fixation in cultures of Trichodesmium (Cyanophyceae). Journal of
Phycology 26:30-35.

Riley, G.A. (1946) Factors controlling phytoplankton populations on Georges Bank.
Journal of Marine Research 6:54-73.

Saino, T. and A. Hattori (1978) Diel variation in nitrogen fixation by a marine blue-
green alga, Trichodesmium thiebautii. Deep-Sea Res. 25:1259-1263.

Saino, T. and A. Hattori (1982) Aerobic nitrogen fixation by the marine non-
heterocystous cyanobacterium Trichodesmium (Oscillatoria) spp.: its protective
mechanism against oxygen. Marine Biology (Berlin) 70:251-254.

Sakshaug, E., Bricaud, A., Dandonneau, Y. Falkowski, P.G., Kiefer, D.A., Legendre,
L., Morel, A., Parslow, J., Takahashi, M. (1997) Parameters of photosynthesis:
definitions, theory and interpretation of results. Journal of Plankton Research 19:1637-
1670.

Sañudo-Wilhelmy, S.A., A.B. Kustka, C.J. Gobler, D.A. Hutchins, M. Yang, K.
Lwize, J. Burns, D.G. Capone, J.A. Raven, E.J. Carpenter (2001) Phosphorous
limitation of nitrogen fixation by Trichodesmium in the central Atlantic Ocean. Nature
411:66-69.

Sañudo-Wilhelmy, S.A., A. Tovar-Sanchez, F.-X. Fu, D.G. Capone, E.J. Carpenter,
D.A. Hutchens (2004) The impact of surface-adsorbed phosphorus on phytoplankton
Redfield stoichiometry. Nature 432:897-901.

Scanlan, D.J. and W.H. Wilson (1999) Application of molecular techniques to
addressing the role of P as key effector in marine ecosystems. Hyrdobiologia 401:149-
175.

Sherman et al (2000) Heterocyst development and localization of cyanophycin in N
2
-
fixing cultures of Anabaena sp. PCC7120 (Cyanobacteria). Journal of Phycology
36:932-941.

Shimura S. and Y. Fujita (1975) Phycoerythrin and photosynthesis of the pelagic
blue-green alga Trichodesmium thiebautii in the waters of the Kuroshio, Japan. Marine
Biology 31:121-128.

116

Siddiqui P.J.A. and E.J. Carpenter (1992) Ultrastructure and immunolocalization of
phycobiliproteins and ribulose 1,5-bisphosphate carboxylase/oxygenase in the marine
cyanobacterium Trichodesmium thiebautii. Journal of Phycology 28:320-327.

Simon, R.D. (1971) Cyanophycin granules from the blue-green alga Anabaena
cylindrical: a reserve material consisting of copolymers of aspartic acid and arginine.
Proceedings of the National Academy of Science 68:265-267.

Smith, E.L. (1936) Photosynthesis in relation to light and carbon dioxide. Proceedings
of the National Academy of Science USA 22:504-511.

Sosik, H.M. (1996) Bio-optical modeling of primary production: consequences of
variability in quantum yield and specific absorption. Marine Ecology Progress Series
143:225-238.

Spitz, Y.H., J.R. Moisan, M.R. Abbott (2001) Configuring an ecosystem model using
data from the Bermuda Atlantic Time Series (BATS). Deep Sea Research II
48(8/9):1733-1768.

Staal, M., L.J. Stal, S. te Linter Hekkert, F.J.M. Harren (2002) Light action spectra of
N
2
fixation by heterocystous cyanobacteria from the Baltic sea. Journal of Phycology
39:668-677.

Steemann-Nielsen, E. 1952. The use of radioactive carbon (
14
C) for measuring organic
production in the sea. J. Cons. Int. Explor. Mer 18: 117-140.

Subramaniam, A., E.J. Carpenter, D. Karentz, P.G. Falkowski (1999a) Bio-optical
properties of the marine diazotrophic cyanobacteria Trichodesmium spp. I. Absorption
and photosynthetic action spectra. Limnology and Oceanography 44(3):608-617.

Subramaniam, A., E.J.,Carpenter, P.G. Falkowski (1999b) Bio-optical properties of
the marine diazotrophic cyanobacteria Trichodesmium spp. II. A reflectance model for
remote sensing. Limnology and Oceanography 44:618-627.

Subramaniam, A., C.W. Brown, R.R. Hood, E.J. Carpenter, D.G. Capone (2002)
Detecting Trichodesmium blooms in seaWIFS imagery. Deep-Sea Research, Part II,
49:107-121.

Suggett, D., G. Kraay, P. Holligan, M. Davey, J.Aiken, R.Geider (2001) Assessment
of photosynthesis in a spring cyanobacterial bloom by use of a fast repetition rate
fluorometer. Limnology and Oceanography 46:802-810.

117
Sunda, W.G, N.M. Price, F.M.M.Morel (2005) Trace metal ion buffers and their use in
culture studies. In: Algal Culturing Techniques, R.A. Anderson (ed). Elsevier
Academic Press, Massachusetts, pp. 35-64.

Taylor, J.R. (1997) An introduction to error analysis: the study of uncertainties in
physical measurements. University Science Books, Sausalito, California, 327pp.

Tyrell, T. (1999) The relative influences of nitrogen and phosphorous on ocean
primary productivity. Nature 400:525 531.

Tyrrell, T., E. Maranon, A.J. Poulton, A.R. Bowie, D.S. Harbour, E.M.S. Woodward
(2003) Large-scale latitudinal distribution of Trichodesmium spp. in the Atlantic
Ocean. Journal of Plankton Research 25:405-416.

Villareal T.A. (1991) Nitrogen-fmation by the cyanobacterial symbiont of the diatom
genus Hemiaulus. Marine Ecology Progress Series 76:201-204

Villareal T.A. (1992) Marine nitrogen-fixing diatom-cyanobacteria symbioses. In:
Carpenter EJ, Capone DG. Rueter JG (eds) Marine pelagic cyanobacteria:
Tnchodesmium and other diazotrophs Kluwer Academic Publishers, Dordrecht,
p 163-175

Villareal, T.A. and Carpenter, E.J. (1990) Diel buoyancy regulation in the marine
diazotrophic cyanobacterium T. thiebautii. Limnology and Oceanography 35(8):1832-
1837.

Villareal, T.A. and E.J. Carpenter (2003) Buoyancy regulation and the potential for
vertical migration in the oceanic cyanobacterium Trichodesmium. Microbial Ecology
45:1-10.

Wu, J., W. Sunda, E.A. Boyle, D.M. Karl (2000) Phosphate depletion in the western
North Atlantic Ocean. Science 289:759-762.

Watkinson, A.J., J.M. O'Neil, W.C. Dennison (2005) Ecophysiology of the marine
cyanobacterium, Lyngbya majuscula, (Oscillatoriaceae) in Moreton Bay, Australia.
Harmful Algae 4:697-715.

Webb, W.L., M. Newton, D. Starr (1974) Carbon dioxide exchange of Alnus rubra; a
mathematical model. Oecologia 17:281-291.

Webb, E.A., J.W. Moffett and J.B. Waterbury (2001) Iron stress in open-ocean
cyanobacteria (Synechococcus, Trichodesmium, and Crocosphaera spp): identification
of the idiA protein. Applied and Environmental Microbiology. 67:5444-5452.

118

Welschmeyer, N.A. and C.J. Lorenzen (1981) Chlorophyll-specific photosynthesis
and quantum efficiency at subsaturating light intensities. Journal of Phycology
17:283-293.

Wolk, C.P., S.M. Austin, J. Bortins, A. Galonsky (1974) Autoradiographic
localization of
13
N after fixation of
13
N-labeled nitrogen gas by a heterocyst-forming
blue-green alga. Journal of Cell Biology 61:440-453.

Wolk C. P., A. Ernst and J. Elhai (1994) In: The Molecular Biology of Cyanobacteria,
D.A. Bryant (ed.) Kluwer, Dordrecht, pp. 769-823.

Wyman, M., J.P. Zehr, D.G. Capone (1996) Temporal variability in nitrogenase gene
expression in natural populations of the marine cyanobacterium Trichodesmium
thiebautii. Applied and Environmental Microbiology 62:1073-1075.

Zehr, J.P., M. Wyman, V. Miller, L. Duguay, D.G. Capone (1993) Modification of the
Fe protein of nitrogenase in natural populations of Trichodesmium thiebautii. Applied
and Environmental Microbiology 59:669-767.

119

Appendix A:
Correlation plots of P
max
and α and dissolved iron [DFe], soluble reactive
phosphorous [SRP] and mixed layer depth (MLD) for
CO
2
fixation and N
2
fixation

120

A
B
Figures A.1a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
, and
(f) α
C
max
vs. [DFe] for CO
2
fixation.

121

C D
1e

Figures A.1a-f. Continued.

122

Figures A.1a-f. Continued.
E
F

123
A B
Figures A.2a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
,
and (f) α
C
max
vs. [DFe] for N
2
fixation.

124
Figures A.2a-f. Continued.
C
D

125
Figures A.2e-2f. Continued.
E
F

126
A
B
Figures A.3a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
, and
(f) α
C
max
vs. [SRP] for CO
2
fixation.

127
Figures A.3a-f. Continued.
C
D

128
E
F
Figures A.3a-f. Continued.

129
A B
Figures A.4a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
,
and (f) α
C
max
vs. [SRP] for N
2
fixation.

130
Figures 4a-f. Continued.
C
D

131
E
F
Figures A.4a-f. Continued.

132
A
B
Figures A.5a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
, and
(f) α
C
max
vs. mixed layer depth (MLD) for CO
2
fixation.

133
Figures A.5a-f. Continued.
C
D

134
E
F
Figures A.5a-f. Continued.

135
A
B
Figures A.6a-f. Regression plots of (a) P
chl
max
, (b) α
chl
max,
(c) P
b
max
, (d) α
b
max
,

(e) P
C
max
,
and (f) α
C
max
vs. mixed layer depth (MLD) for N
2
fixation.

136
Figures A.6a-f. Continued.
C
D

137
E
F
Figures A.6a-f. Continued.

138
Appendix B: Comments on Methodology

Limitations of the research and future perspectives
I learned a tremendous amount throughout the course of my dissertation. Yet,
given the opportunity to repeat the experiments or improve upon them, there are a
number of issues I would attempt to resolve. With the laboratory experiments, the
main concerns are choice in nutrient concentration and culturing method. With
respect to the field experiments, I would have (of course) liked the opportunity to
conduct more experiments in more locations, and with the efficiency that I developed
by the last set of experiments. Moreover, with the knowledge gained from the
laboratory experiments, as described in the synthesis above, it would be very useful to
conduct a suite of physiological measurements to the same population of colonies as
those used in the photosynthetron experiments: e.g. the measurement of cellular Fe
and P quotas to correlate directly nutritional status with photophysiological responses.
These limitations are discussed in further detail below.

A. Laboratory Experiments: Nutrient Concentration Limitations
For several reasons, I would have chosen different ranges of P and Fe
concentrations for the laboratory experiments of Chapter 3. One significant reason for
the choice of nutrient concentration was purely methodological. A large initial
component of my doctoral research was spent learning how to grow cultures at low
nutrient concentrations. This was particularly an issue for the Fe cultures in which it

139
was necessary for me to go to the laboratory of Dr. Francois Morel at Princeton
University to learn trace metal clean methodology with the help of Dr. Adam Kustka.
All the Fe experiments were conducted using the trace metal clean technique, but the P
experiments were not. Without using the trace metal clean techniques, I was not able
to get concentrations of P lower than 0.2 µM. Thus, this set the lower limit for my P
experiments. I was finally able to grow Trichodesmium cultures at 10 nM Fe, but was
not able to conduct any P-E experiments. Therefore, my choice in nutrient
concentration was driven more by practical considerations than by literature-based
theoretical considerations. Given the opportunity to repeat these experiments under
trace metal clean conditions, I would have chosen and would have been able to
experiment with lower concentrations.
Moreover, the lowest concentrations used in the laboratory P and Fe
experiments were substantially greater than the highest concentrations measured in the
field: 0.081µM SRP in nature vs. 0.200 µM SRP in the laboratory and 3.34nM DFe in
nature vs. 30nM DFe in the laboratory. Prior to beginning these experiments, the
cultures were growing in 50 µM SRP and 1 µM Fe concentrations and were
acclimated to these concentrations. Therefore compared to the nutrient concentrations
to which they were acclimated, the low Fe and P concentrations in the laboratory
experiments were drastic reductions. The reduction in chl a cell
-1
and growth rate did
indicate that these cultures were nutrient-limited. Similarly, the reduced P
max
and α for

140
both CO
2
and N
2
fixation indicates that even though these are not naturally limiting
nutrient concentrations, there was indeed an effect in the lower nutrient cultures.
However, these cultures were isolated from the field, and were originally
acclimated to the much lower field nutrient concentrations. Thus, it is possible that
even over the many years of culturing and despite moving between one laboratory to
another, even at the lower concentrations used, the cultured Trichodesmium cells have
been living in the lap of nutrient luxury for all these years, and were never actually
limited. Therefore, it would be useful to examine this possibility by growing the
cultures at concentrations more similar to those that would be observed in nature.
Additionally, one of the main results from Chapter 3 is that I observed a larger
effect from Fe-limitation than from P-limitation on the cultures. Cultures grown at the
low Fe concentrations had P
max
and α values reduced by between ~19-33% whereas
cultures grown at the low P concentrations had P
max
and α values reduced by between
~4-6%.
It is not clear whether Fe-limitation does indeed lead to more severe
physiological reductions than P-limitation or if my choice in concentrations biased
these findings. The percentages presented above were determined by dividing the
difference in parameter value (e.g. P
max
high/ P
max
low) by the concentration difference
([X]high/[X]low). For the reasons discussed above, the difference between low and
high P concentrations was 150-fold while the difference between low and high Fe
concentrations were 7-fold. Therefore, the difference in parameter values for the P

141
(7), and this may have masked the real effect of P-limitation. In a linear relationship,
if the difference in P concentration was only 50-fold (i.e. conducting experiments on
Trichodesmium cultures grown in 6µM P rather than 30µM P), the percentage
difference for the P experiments may have then been between ~14 - 46% difference,
on par or greater than the Fe experiments. Thus, to address these concerns, it would
be worthwhile to conduct these experiments with nutrient concentrations closer in
values, over a wider range in nutrients, and more similar to natural field populations.

B. Laboratory Experiments: Culturing Method Limitations
Another large component of this research was determining how best to culture
the Trichodesmium cultures in order to answer the appropriate questions. Cultures are
often grown either in continuous culture or batch culture (MacIntyre and Cullen 2005).
When grown in batch cultures, the organisms go through the different stages of lag
phase with minimal growth, an exponential growth phase until a nutrient is limited,
and stationary growth (MacIntyre and Cullen 2005). Thus these stages are essentially
constrained by availability of nutrients. Conducting experiments at different stages of
growth yield different results. For instance, Welschmeyer and Lorenzen (1981)
measured α and P
max
in six different species of non-diazotrophic phytoplankton.
Within species, there were no significant differences in α and P
max
in cultures during
exponential growth phase; there was, however, significant variability among cells
during stationary growth phase. Therefore, when utilizing batch experiments, growth

142
stage is a significant consideration and may bias results in experiments aiming at
studying the effect of nutrient limitation.
On the contrary, in continuous culture, the cultures are maintained at balanced
growth and the cells are fully acclimated to their conditions; biomass in the culture is
determined by the initial concentration in the medium (MacIntyre and Cullen 2005).
Continuous cultures are ideal for assessing the effect of nutrient limitation on an
organism. In the studies of this dissertation, however, I grew the cultures under semi-
continuous growth. In these, the culture was diluted regularly to maintain the same
biomass concentrations. But, these experiments are mainly utilized to keep cultures at
constant nutrient-replete concentrations (MacIntyre and Cullen 2005).
There were two main reasons for choosing the semi-continuous method and
both were methodological. Until very recently, Trichodesmium have not been
maintained in continuous cultures (M. Mulholland pers. comm., C. Holl pers. comm.,
J. Sohm pers. comm.). Knowing that the batch culture method would lead to some
considerable constraints in interpreting the data, I chose the next best method, semi-
continuous cultures. Our lab has now successfully grown cultures in continuous
culture (J. Sohm pers. comm.). Thus, the next logical and important step would be to
repeat these experiments with the cultures grown in this manner to compare the results
to the cultures grown in semi-continuous culture to determine if differences in culture
method yield differences in experimental observations.

143
However, one significant drawback of utilizing the continuous culture method
is that in these types of experiments, one can be potentially limited by biomass. For
each experiment, I would utilize 500ml of sample. In most continuous culture
experiments, this would be a sizable portion of the culture and therefore, after each
experiment, the cultures would need time to build back up. To the contrary, with the
semi-continuous method, I set the dilution rate by splitting the cultures every four
days. Once the cultures were established, there was always enough culture to run
experiments. (In fact, so much so, I could not keep up!)
Ultimately, to eliminate any possible question, it would be best to run
experiments with Trichodesmium cultures at stationary and exponential growth stage
from batch culture methods, cultures grown under differing nutrient concentrations
from semi-continuous culture methods and cultures grown under differing nutrient
concentrations from continuous culture methods.

Abstract (if available)

Abstract Trichodesmium is a globally significant cyanobacterium that plays an important role in marine carbon (C) and nitrogen (N) cycling. Accordingly, global ecosystem have begun to incorporate the critical role of diazotrophs, and Trichodesmium specifically. Most commonly, this is done by incorporating the photophysiological parameters Pmax (the rate at which production is maximal) and alpha

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The connection of the phosphorus cycle to diazotrophs and nitrogen fixation

PDF

The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

PDF

Biological nitrogen fixation associated with living and decomposing macroalgae

PDF

Iron-dependent response mechanisms of the nitrogen-fixing cyanobacterium Crocosphaera to climate change

PDF

Carbonate dissolution at the seafloor: fluxes and drivers from a novel in situ porewater sampler

PDF

Future impacts of warming and other global change variables on phytoplankton communities of coastal Antarctica and California

PDF

Understanding the role of population and place in the dynamics of seasonal influenza outbreaks in the United States

PDF

Modeling deep ocean water and sediment dynamics in the eastern Pacific Ocean using actinium-227 and other naturally occurring radioisotopes

PDF

Design and modification of electrocatalysts for use in fuel cells and CO₂ reduction

Asset Metadata

Creator Finzi, Juliette Anne (author)

Core Title Photophysiological parameters for CO2 and N2 fixation in trichodesmium spp. in natural populations and culture nutrient limitation experiments

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Ocean Sciences

Degree Conferral Date 2007-08

Publication Date 08/06/2007

Defense Date 08/28/2006

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

Tag CO2 fixation,Fe limitation,N2 fixation,nanoSIMS,OAI-PMH Harvest,P limitation,photophysiology,Trichodesmium

Language English

Advisor Capone, Douglas (committee chair), Hammond, Douglas E. (committee member), Keifer, Dale (committee member), Michaels, Anthony (committee member), Subramaniam, Ajit (committee member)

Creator Email finzi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m758

Unique identifier UC1418805

Identifier etd-Finzi-20070806 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-538403 (legacy record id),usctheses-m758 (legacy record id)

Legacy Identifier etd-Finzi-20070806.pdf

Dmrecord 538403

Document Type Dissertation

Rights Finzi, Juliette Anne

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu