~D

UNIVERSITY OF HAWAI'I LIBRARY

DISTRIBUTIONS AND DIVERSITY OF AEROBIC ANOXYGENIC PHOTOTROpmC BACTERIA ALONG ONSHORE OFFSHORE TRANSECTS NEAR

PACIFIC ISLANDS

A THESIS SUBMITTED TO THE GRADUATE DMSON OF THE UNIVERSITY OF HA WAl'I IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF

MASTER OF SCIENCE

IN

OCEANOGRAPHY

AUGUST 2008

By Anna E. Ritchie

Thesis Committee:

Zackary Johnson, Chairperson Susan Brown

Matthew Church

II

We certify that we have read this thes is and that , in our opinion, it is satisfactory in scope

and quality as a thes is for the degree o f Master of Sc ience in Oceanography.

THESIS COMMITTEE

iii

This thesis is dedicated to Jan, Paul, Drew, Heidi, and Baby Ritchie.

iv Acknowledgments

I would like to thank the crew of the RN Kilo Moana for assistance during the WP2 cruise and mini-WP2 cruise. Also Kathy Barbeau. crew of the RN New Horizon, and Yajuan Lin for assistance during the California DCM! cruise. I am grateful to Dan Reineman, Dan Barschis, and Eric Bachman for assistance during sampling in Kaneohe Bay. Also matlab codes for Kaneohe Bay and graphics help from Jerome Aucan, Thomas Decloedt, and Michael Romano. Thanks to NSF for financial assistance.

v

Abstract

Aerobic anoxygenic phototrophic bacteria (AAP's) popuIation abundances are not well characterized, in spite of their potential importance for the microbial loop and primary production. AAPs have a photoheterotrophic metabolism, thus both inorganic and organic nutrients, as well as physical variables likely drive their ecological distributious. To explore AAP abundances and the environmental variables that may be regulating them, we quantified AAP abundance by counting the ptifM biomarker along onshore/offshore transects near 5 distinct islands in the Pacific Ocean (Oahu, Molokai, Futuna. Aniwa, and Lord Howe) with steep environmental gradients. Abundance patterns are further explored by investigating the genetic diversity of the AAP community using plffM clone libraries and QPCR dissociation curves. Overall we report small but sometimes significant AAP populations that increase near shore and are comprised of distinct genetic clades.

vi Table of Contents

Abstract ..........•.......................................................................•............. v List of Tables ...................................................................................... viii List of Figures ................................................................•..................... ix Chapter 1: Introduction ......•...................................................................... l Chapter 2: Aerobic Anoxygenic Phototrophic Bacterial Abundance •........................ 4

Methods ......................................•.........................•..................... 7 Sampling ........................................................................... 7 DNA Extraction ................................................................... 8 QPCR Standards .....•............................................................ 9 QPCR Enumeration of AAPs ....•.•....•.••..•.•....••••...••....•••••...•..... 1 0 QPCR Negative Controls ........................................................ 11 Enumeration of Total Bacteria and Picophytoplankton

using Flow Cytometry ••...•••.....•....••.••....•......•.....•.•...•... 12 Nutrient Analyses ................................................................ 12 Statistical Analyses .............................................................. .13

Results •.•........•.•.•.•••....••••...•.....•..•..••......•••...••...••••.....•....•••••......• 13 AAP Abundance ...................•............................................... 13 EnvironmentaI Variables: Salinity and Temperature ........................ 15 Environmental Variables: Chlorophyll, Photosynthetic

PicopIankton, and Total Bacteria ..................................... .15 Environmental Variables: Nutrients ................•.......................... 16 Environmental Variables Correlated with AAPs Abundance .............. 16

Discussion .................................•................................................. 17 Chapter 2 Figures and Tables ............................................................ 22

Chapter 3: Aerobic Anoxygenic Phototrophic Bacterial Diversity ........................... 30 Methods ..................................................•...................•.............. 32

Sampling ........................................................................... 32 Preliminary AAP's Diversity Study at Kaneohe Bay ........................ 33 PufM Clone Libraries ............................................................ 34 Comparing Clone Libraries using SONS and

Confirming Clone Library Saturation ................................. 35 Examining Diversity of AAPs using Dissociation Curves .....••........... 35

Results ..•.•..••..•..••••....•....•••....••••...•.•.....•..........•.•.•....•••..•••••...•.••. 36 Kaneohe Bay Preliminary Diversity Study Results ...............•......... 36 Clone LlDraries Results ......................................................... 37 Dissociation Curves of Selected Clone Library Sequences ................ 39 Clone Library Community Structure ........................................ .41 Near shore and Offshore Diversity and Community

Structure ....................•............................................ .41 Broad AAP Diversity Determined by Dissociation Curves ............... .43

vii

I>isctJssi(gl •••••••••••••••••••••••••••••••••••••••••••••..•...•••••••••...•••••••.••••••••••. ~ Chapter 3 Figures and Tables ...... ............................................................... 48 Ftef~CI:S ................................................. ......................................... 5ti

viii List of Tables

1. List of Locations and Chlorophyll Size Fractions ...........................................•...................... 28

2. List of Correlation and Model II Linear Regression Results .................. 29

3. List of SONS Results for Aniwa, California, Oahu, and Onshore/OffShore Libraries .................................................. 55

ix List of Figures

1. Map of Sampled Locations ......................................................... 22

2. QPCR AAP Standard Curve ....................................................... 23

3. Depth Profiles for AAP Concentrations ......................................... 24

4. AAP Concentrations Versus Distance from Shore for Pacific Islands and California Transects ..................................... 25

5. AAP Concentrations Versus Distance from Shore for Kaneohe Bay Transects ....................................................... 26

6. Positive Correlations of AAP Concentrations with Chlorophyll, Synechococcus, and Bacteria .................................. 27

7. Clone Libraries Dissociation Curves ............................................ .48

8. Neighbor-joiniog Trees for Aniwa, California, and Oahu Libraries ........ .49

9. Rarefraction Curves ................................................................. 50

10. Neighbor-joiniog Tree for Onshore and Offshore Sequences ................. 51

11. Neighbor-joiniog Tree for Onshore and Offshore Sequences With Chlorophyll, Synechococcus, and Bacteria ............................ 52

12. Diversity of AAP Community for Aniwa, Futuna, and Kaneohe Bay Transects ......................................................................... .53

13. Diversity of AAP Community for California, Lord Howe, Oahu, and Molokai ...................................................................... 54

Chapter 1: Introduction

Microbial oceanography is a field that focuses on understanding the distribution,

community composition, and interactions of bacteria, viruses, archaea, and protists with

each other and the environment. The terminology "microbial" is an interchangeable term

in microbial oceanography literature; here the term "microbial" will refer to the marine

bacterial commnnity.

1

Microbial oceanography dates back to 1677 with Antonie van Leewenhook's

description of small animals in a drop of water (Van Leeuwenhook, 1677 referenced in

Karl, 2007). In recent years this field has exponentially expanded due to technological

advancements such as genetics and flow cytometry to enumerate and identify key players

in the microbial community that elude cultivation. In the past thirty years scientists have

identified marine microbes as an alternative pathway to facilitate recycling of nutrients by

utilizing DOM, then being grazed upon by protists; this is known as the microbial loop

(pomeroy, 1974, Azam et at, 1983). The importance of marine microbes can vary from

very small scales of individual microbial interactions to the global scale. On the global

scale marine microbes influence the degradation and production of DMS, which is an

important greenhouse gas more potent than carbon dioxide in its control on global

climate (Andreae et at, 1997, Zemmelink et at, 2005). While some microbes directly

impact the global climate, some such as Prochlorococcus can be used as a model to help

us understand what primary production could be like if the oceans warmed. stratification

increased, and waters become nutrient depleted due to anthropogenic induced global

warming (Behrenfe1d, et at, 2006). Almost daily our perceived knowledge of how the

ocean and the microbial community functions is challenged and expanded to new

dimensions.

2

While these examples are a few of the periodic large discoveries of the field there

are many other achievements addressing the current questions and problems of microbial

oceanography. The driving questions of our field are: How abundant are marine

microbes? What is the diversity of marine microbes? What are the community's

metabolic capabilities? How do microbes interact and how are they distributed through

the world's ocean? An additional goal in this field is to link microbial communities to

geochemical cycles and physical processes in the ocean.

The first order questions of how abundant and who makes up the microbial

community are one of the most challenging questions due to the sheer abundance of

marine bacteria. One approach is to investigate a part of the community using biomarkers

unique to a given population. A few examples of commonly used biomarkers are

pigments, functional genes, or 16S rDNA. Genetic biomarkers descnbing the diversity of

marine microbes has advanced dramatically but is limited by our ability to only search for

wbat we already know genetically exists. Therefore much of microbial diversity is

uncharacterized and leaves scientist questioning the feasibility of determining the

diversity of marine bacteria (O'Malley, 2007, Whitfield, 2005, Pedros-Alio, 2006,

Pommier et al., 2007, Longhurst, 1998). While knowing how many different types of

bacteria are present in the ocean would help us to comprehend how feasible it is to

descnbe microbial diversity, it is not the only important aspect of understanding

microbial diversity. A majority of what we have learned about marine microbial diversity

has been through non-cultivated techniques due to the inability to cu1ture up to 99% of

marine bacteria in the ocean. While molecular information has widened our

understanding of the types of microorganism in the sea, we nmst keep in mind this

molecular information must be put into ecological context through cultivation of the

organisms (Giovannoni and Sting, 2005). For example, cuhivated representatives and

environmental studies of Prochlorococcus strains MED4 and MIT9312 have revealed

different horizontal and vertical distributions within the ocean while their genetic

similarity for the 16S rRNA would suggest they are the same strain (partensky et al.,

1999). The linking of diversity data and bacterial distributions with cultured strains

proves to be critical information for the forward movement of this field.

The following chapters use a genetic biomarker to describe the diversity and

distnbution of a group of marine bacteria termed aerobic anoxygenic phototrophic

bacteria (AAP's). AAPs are a group of mixotrophic bacteria containing the pigment

bacteriochlorophyll a (Bchl a). In chapter 2, I describe the distributions of AAPs along

onshore and offshore transects near distinct Pacific Islands using quantitative PCR

targeting the functional gene pujM. The pufM gene encodes for a reaction centre

associated with the bacteriochlorophyll a photosynthetic system In chapter 3, I use this

same functional gene to investigate the diversity of AAPs along the same onshore and

offshore transects. The overall purpose of these two chapters is to understand AAPs

ecological distnbution through their abundance, diversity, and the environmental

variables which drive these patterns

3

4 Cbapter 2: Aerobic Anoxygenic Phototropbic Bacterial Abundance

AAPs are a group of mixotrophic bacteria which contain the pigment

bacteriochlorophyll a (Bchl a). Oxygen, light, and organic carbon sources are known to

influence the formation of this pigment (Yurkov and Beatty, 1998a). Generally AAPs

have low Bchl a concentrations per cell (Kolber et aL, 2(01) and previous research

suggests they cannot survive on a sole photoautotrophic metabolism (Yurkov and Beatty,

1998a). AAPs have multiple metabolic pathways that may confer an advantage in

inhabiting a variety of environments or environments in constant flux. Laboratory

experiments and whole genome sequencing have revealed a wide range of metabolic

capabilities such as utilization of low weight organic carbon, growth optimization on

complex organic carbon, nitrification, carbon dioxide fixation, light enhanced cellular

growth, preferred suboxic environments, aggregate furmations, and high amounts of

caretnoids for UV protection (Fuchs et aL, 2007, Denner et aL, 2002, Gich and

Overmann, 2006, Qiang. et aL, 2006). Due to the broad metabolic capabilities of AAPs it

is expected and has been fuund that they live in a variety of environments ranging from

freshwater, marine environments, soil, hot springs, Antarctic lakes, and hyrdrothermal

vents (Yurkov and Beatty., 1998a, Rathgeber et aL, 2004, Labrenz et aL, 2(05).

AAPs anoxygenic phototrophy is thought to have been obtained from their

anaerobic counterparts (Kolber, 2007). This furm of phototrophy uses reduced sulfur,

iron, or unknown electron donors and does not produce oxygen as a byproduct. The lack

of oxygen evolution is one of the main reasons AAPs have not been included in some

estimateg of photosynthesis. In addition, AAPs have not been included in estimates of

5 photosynthetic biomass because they do not contain chlorophyl1 Therefore, AAPs have

historically been included within heterotrophic bacteria community and bacterial

production measurements. Photoheterotrophy within the bacterial connnunity has been

shown to be important in that light stimulated leucine uptake increased significantly

compared to conventional methods of measuring bacterial production in the dark (Church

et al., 2004). Follow up studies on this phenomenon reported that up to 40% of the

increase could be due to unidentified photoheterotrophs such as AAPs (Michealou et al.,

2007). These results suggest that photoheterotrophy is a common strategy that some

current measurements of photosynthesis and respiration do not resolve.

In addition to understanding what proportion of phototrophy is carried out by

AAPs, another major goal is to understand how this new form of phototrophy affects the

carbon cycle. While the abundance of AAPs within the oligotrophic ocean suggests they

do not contnbute significantly to the carbon cycle. other AAP characteristics such as

growth rates suggest AAPs contribution can be significant. For example, Koblizek and

collaborators (2007) predicted AAPs would account for 4- 50% of total bacterial

production in the oligotrophic, if growth rates were I division per day for AAPs and an

abundance of 2-4% of the total bacterial community. Thus I would argue other AAP

characteristics such as growth rate, amount of Bchl a per strain, and cell size should be

considered before concluding they are not important contnbutors to the carbon cycle.

Understanding the ecology of AAPs will facilitate answering these bigger issues

discussed above. The current study is focused on describing the distribution of AAPs

along onshore/offshore transects and determining the relationship with enviromnental

6 variables driving their ecology. Our current understanding of the distribution of AAPs

is that they are most abundant near coastal waters, ranging from 0.5 % to 34% of the total

community (Zhang and Jian, 2007, Do et al., 2006, Cottrell et al., 2006, Sieracki et al.,

2006, Waidner and Kircbman, 2007, Schwalbach and Fuhrman, 2005). In the

oligotrophic ocean abundance estimates are usually less than I to 6% of the total

community (Cottrell et al., 2006, Jiao et al., 2007, Yutin et al., 2007) excluding two

studies that estimate AAPs to be as nmch as 11 % (Kolber et al., 2000) and 24% (Lami et

al., 2007) of the total comnmnity. These periodic reports of high AAP abundance in the

oligotrophic ocean challenge our understanding of what controls the distribution of

AAPs, and prompted this study near Pacific Islands. Here I use gradients in a variety of

physical and chemical variables of these Pacific Islands to determine what are the

proximal variables responsible for determining AAP abundance. I hypothesized that

controls on AAPs distributions along natural environmental gradients could be light,

nutrients, salinity, and photosynthetic community composition. I report that AAPs

generally comprise 1 % of the total bacterial community, except for the most eutrophic

location where they accounted for more than 4 % of the total bacterial community. The

abundance of AAPs was most closely correlated with chlorophyll (photosynthetic

biomass), Synechococcus, and total bacteria.

7 METHODS:

Sampling

Eight onshore/offshore transects were conducted in the Pacific Ocean at the

following locations: California (280 36'N, 1250 4'W), Aniwa (ISO 10'S,169° 15'E), Lord

Howe (310 33'S, 1590 05'E), Futuna (190 32'S,170° 13'E), Oahu (210 28'N, 1570 59'W),

Molokai (21 0 OS'N, 1570 02'W), and two in Kaneohe Bay Iocated on the island of Oahu

(210 27' 35''N, 1570 48' 15" W). Surface water samples were collected in dark amber

bottles for chlorophyll, salinity, flow cytometry, nutrients (NO:z, N0J, P04, and N'l4),

and DNA for abundance measurements. Two of the onshore/offshore transects located in

Kaneohe Bay were sampled three times (June 13th 2006, August 4th, 2006, and June 21'"

2007) (FJgUre 1). In addition to surface samples, two depth profiles were measured to

explore depth distribution of AAPs. Depth profiles were conducted off of the island of

Oahu and sampled using oceanographic Diskin bottles mounted to a CID equipped

rosette.

Along Aniwa, Futuna, and Kaneohe Bay transects, temperature was recorded

using a digital thermometer (Traceable). Temperature and salinity measurements for

California, Oahu, and Molokai were made with a CID mounted on the rosette. Kaneohe

Bay salinity measurements were made in duplicate using a refractometer (Vista).

Total (0.22 !JlD) or size fractionated (O.S, 2.0, or 5.0 !JlD filters, see Table 1 for

details) chlorophyll measurements were made by fihering 100 m1 of sample onto

polycarbonate fihers (GE Water and Process Technologies) under low fihration pressure

« 15 mm Hg). Filters were extracted in 100% methanol for 24 hours in the dark at -4 "C,

8 and then read on a Turner 10AU fluorometer calibrated according to Porra et a1. (1989).

Chlorophyll a measurements were calculated following Welschmeyer (1994).

DNA was collected by filtering 50 or 100 ml of water through 0.22 f.LID

po1ycarbonate filter (GE Water and Process Technologies) followed by 3 ml of

preservation solution (l0mM Tris,lOOmM EDT A,0.5 NaCI at a pH of 8.0). Filters were

placed in bead beating tubes and stored at -80°C freezer until extraction.

DNA Extraction

DNA was extracted from duplicate filters using physical and chemical procedures

following the Gentra DNA extraction kit, with the following modifications: Briefly, bead

beating tubes containing filters were removed from -80°C freezer and 0.25 g of sterile

0.1 mm zirconium beads (Biospec) were added. Seven hundred fifty microliters of cell

lysis solution (Gentra systems) was added to each tube and samples were put on ice for 1

minute, then placed in bead beaters for 1 minute at 4800 RPM's. Tubes were incubated

for 5 minutes at 80 OC on a heating block, after which 4 )11 of RNAse A was added and

tubes were inverted 25 times before incubating for 30 minutes at 37 oc. After incubation

tubes were cooled on ice for 1 minute before adding 250 f.Il of Protein Precipitation

solution (Gentra Systems) and vortexed for 20 seconds at a maximum speed of 3200

RPM's (VWR). Tubes were placed on ice for 5 minutes before centrifuging at 16,000 x g

for 3 minutes. Nine hundred microliters of the supernatant were transferred to a new 1.5

ml microcentrifoge tube and centrifuge as above. Seven hundred fifty microliters of the

supemantant was transferred to a new microcentrifoge tube containing 750 )1l of 100%

9 isopropanol Samples were inverted 50 times before centrifuging for 1 minute at 16,000

x g. Supernant was poured off and 300 !Jl of 70% ethanol was added and tubes were

inverted several times. Centrifuging was repeated for 1 minute at 16,000 x g. Afterwards

supernatant was poured off and tubes were allowed to dry in a laminar flow hood for 15

minutes before lOO!Jl of DNA Hydration Solution (Gentra Systems) was added. Tubes

were stored overnight at room temperatnre to allow DNA to rehydrate and stored in -80

·C freezer.

OPCR Standards Four sets of QPCR standards were made from quadruplicate cultures of

Erythrobacter longus Strain NJ3Y (Koblizek et al 2003) cultivated in rich media (F/2

media (Guillard and Ryther, 1962, Guillard, 1975),0.5 g peptone!L, and O.lg yeast

extractlL) at 30 IlJDOI quanta/m2/sec on a 12112 light cycle. Cultures were grown until the

late log phase of growth and a maximum concentration of at least 1 x Iff cellslml as

determined by epi fluorescent microscopy (Nikon) ofDAPI (Sigma Aldrich) stained cells

(Porter and Feig Y., 1980), and flow cytometric (F ASCaliber) cell counts of SYBR green

(Invitrogen) stained cells. Cultures were diluted with sterile seawater collected from

Station ALOHA (22" 45' N, 158· OO'W) and DNA was extracted as described above. A

serial dilution of the extracted DNA was performed to create standards that ranged

between 1 x 107 to 1 X 10-\ cellslrnl.

The average for all QPCR threshold cycle values (CT) at a given ptifM standard

concentration were calculated and plotted against the known ptifM standard

concentrations (Figure 2). A model I linear regression resulted in a slope of -2.13, which

10 supports near a 100% amplification efficiency. The average r-squared value for

standard curves from all QPCR runs was 0.99. The standard detection limit was 3.7 x 101

pufM copieslml, which was determined by tbe lack of resolution between the lowest

standard and tbe no template control

OPCR Enumeration qfAAPs

Enumeration of AAPs was achieved using primers specific to the functional gene

pufM, which encodes a central photosystem reaction center protein that is unique to

AAPs. The abundance of AAPs was calculated assuming one copy of pufM per genome.

This is consistent with all genomic data available, including the AAP strain Roseobacter

dinitrificans (Swingley et al., 2007). Primers (pufM557F (5'TAC GGS AAC CfG TWC

TAC 3') and pujMW A WR (5'GCR AAC CAC CAN acc CA 3 '» amplified 240 base

pairs of the pufM gene (totalpufM gene is approximately 1000 bp) as described by Yutin

et al. (2005) and Waidner and Kirchman (2007).

Duplicate QPCR reactions were analyzed for each standard dilution,

environmental DNA extract, and for all negative controls. For each station a total of four

QPCR reactions were performed (duplicate QPCR reaction per filter plus duplicate filters

per station). Each 25 !Jl QPCR reaction contained tbe following: 2.5 !Jl of DNA template,

IX SYBR iTaq supermix (BioRad), and 0.1 !JlIlO1 L-1 of forward primer pufM557F and

reverse primer pufm-WA WR.

QPCR assay was performed following Kirchman and Waidner (2007) PCR

protocol (95 "C for 10:00 min, tben 40 cycles of 95 "C for 15 sec, 560 C for 45 sec, 72 "C

11 for 45 sec) using a 7300 Real time PCR system (Applied Biosystem). Enumerations

were calculated using manufacture's software to create a 1inear regression of fluorescence

for each standard versus the known concentration of cells determined by flow cytometry.

OPCR Negative Controls

In my original evaluation of various suppliers SYBR-Green QPCR mastermixes, I

observed significant non-specific amplification of non-target template DNA using pujM

primers. As a resuh, I used the following negative controls to monitor non-specific

binding during the QPCR assay: water, Altermonoas UH00601, and Prochlorococcus

MIT9312. Each negative control was extracted similar to enviromnentaI samples and

standards. Negative controls were extracted at 10e6 cellsfml concentration as determined

by DAPI stained epi-fluorescent microscopy.

Negative controls concentrations (UH06001 or MIT9312), which gave

statistically similar results, were averaged together for all QPCR assays. Negative

controls that were added at 1 X 106 cellsfml, resuked in amplification equivalent to 276

pufM copieS/ml for UH06OO1 and 225 pufM copies/ml for MIT9312. Therefore,

amplification of negative controls was four orders of magnitude below the concentrations

added to each reaction.

12 Enumeration Q[Total Bacteria and Picophytqplankton using Flow Cytometry

Flow cytometry samples were preserved by collecting 1 ml of sample water in a

cryovial containing 6,.d of 25% glutaraldehyde (Tousimis) for a final concentration of

0.125% . Samples were mixed and placed in the dark for 10 minutes before storing in -80

°C freezer or liquid nitrogen.

A FASCah"ber flow cytometer equipped with a syringe pump (flow rate 10

J.IlImin) collected the following variables: forward light scatter. side light scatter. 530/30

run, 670 nm LP. and 585/42 run. Prochlorococcus. Synechococcus. and small pico

eukaryote populations were enumerated based on unique combinations of the 5 variables

following Olson et aI. (1990). Total bacteria were processed similarly. but stained with

IX Sybr Green I (Molecular Probes) for 30 minutes in the dark before processing (Marie

et a1.. 1997). All samples were run with 1 J.Lm yellow-green microspheres (polysciences.

Inc.) as an internal standard.

Nutrient Analyses

Nutrient samples for Kaneohe Bay. Oahu. Molokai, Aniwa, Lord Howe. and

Futuna were processed following Hernandez-Lopez and Vargas-AIbores (2003). Nitrite

standard curves were made up of 17 concentrations ranging between 0.7 - 50 J.LM. Nitrate

standard curves were made up of 17 concentrations ranging between 0.7 - 40 J.LM.

Phosphate standard curves were made up of 17 concentrations ranging between 0.5 - 30

J.LM. Samples were analyzed in triplicate for all measurements except for phosphate for

the islands of Aniwa, Oahu. Lord Howe. and Futuna, which were analyzed in duplicate.

13 Nutrient concentrations were calculated using model I linear regressions fitted

to the standard concentrations against absorbance values. Based on environmental

nutrient concentrations the ranges of standards were restricted to 15 JlM as the highest

standard. This provided a better linear regression with r-squared values ranging between

0.98-0.99 for nitrate, nitrite, and phosphate. The limit of detection for nitrite, nitrate, and

phosphate nutrient assays were I JlM.

Statistical ~ses

Statistical calculations such as correlations, model I regressions, and non -

parametric calculations were made using MlNITAB 14. Model II regressions were

calculated using a Matlab code created by Edward T. Peltzer. Plots of the island transects

were accomplished with Matlab. Images of Futuna, Aniwa, Lord Howe, Oahu,

California, and Molokai were obtained from Google Earth then digitized with the z graph

digitizer 1.9. Tables of latitude and longitude obtained from the digitized image were

then imported into Matlab to graph the islands. Kaneohe Bay plots were formed in

Matlab via bigh resolution bathymetry data supplied by Dr. Jermone Aucan.

RESULTS:

AAP Abundance

Two vertical profiles of AAPs concentrations were measured at stations along the

Oahu transect (Figure 3). AAP abundance was highest at the surface and decreased with

depth (FIgUre 3). AAPs were detectable to about 100 meters.

14 Most island and California surface samples had AAP concentrations that were

less than 6 x 16' cellslml and less than 1 % of the total bacterial community (Figure 4).

Kaneohe Bay had the highest concentration of AAPs at 1.35 x 10' cellslml (Figure 5A).

Kaneohe Bay also had the highest AAP's percentage (4.3%) of the total bacterial

community (Figure 5B). Offshore stations in Kaneohe Bay had AAP concentrations

similar to those reported for samples collected offshore in Califurnia and the other island

transects.

Most locations had the highest AAP concentrations near shore. AAP abundances

at the offshore stations were consistently lower than those observed in the near shore

stations. Although there was significant variability between locations, the overall patterns

of AAP abundance and as a percentage of the total bacterial community show the highest

values near shore.

Variability seen in the overall offshore AAP abundance pattern can be explained

by analyzing each location/transect separately. For Aniwa, California, and all of Kaneohe

Bay transects a decrease in AAP abundance towards the offshore stations was observed.

AAP abundance along the Molokai and Futuna transects generally increased with

distance from shore. AAP abundance along the Oahu and Lord Howe transects varied

little between stations (Figure 4).

While no seasonal study was performed in this study it is surprising the

consistency of AAP abundance between the three sampling dates in Kaneohe Bay.

Enumeration of AAPs for Kaneohe Bay's north and south channel transects were similar

between sampling dates. While the total bacterial estimates were less consistent,

resulting in change in the AAP to total bacterial abundance (Figure 5).

Environmental Variables: Salinity and Te111j!erature

15

Salinity ranged from 31 to 38. Temperature ranged from 17.4 °C (California) to

28°C (Futuna and Aniwa). Temperatures varied little along the onshore/offshore

transects with the exception of Kaneohe Bay (6.21.07) where temperatures ranged

between 21.9 °C to 26.6 °C. For a majority of the transects high salinity and lower

temperatures were seen offshore, while lower saJinity and higher temperatures were seen

near shore.

Environmental Variables: Chlorophyll. Photosynthetic Picoplankton. and Total Bacteria

Total chlorophyll a was greatest in Kaneohe Bay (6.13.06) at 3.1 J.Lg/L and

(8.4.06) 2.8 J.Lg/L. Of the other locations, the highest chlorophyll a concentration was 0.5

J.Lg/L (California transect). Chlorophyll a tended to decrease as distance from shore

increased, similar to patterns observed for AAP abundance. Total bacteria concentrations

ranged from 3.2 x 106 to 1.26 x 1(i1 cellslml with an average of9.4 x 1(i1 (+1- 5.3 x 1(i1

cellslml). Prochlorococcus abundances varied between 2.4x HY to 1.2 x l(i1 cellslml.

Prochlorococcus composed less than 1 to 28 percent of the total bacterial community.

Synechococcus abundances varied between 7 x 101 to 6.7 x 1 (ii, comprising less than 1 to

21.9 percent of the total bacterial community. Synechococcus concentrations were

greatest in Kaneohe Bay and lowest for Futuna, Aniwa. Oahu, and Molokai.

16 Picoeukaryotes were quantified by flow cytometry analysis, but our flow cytometry

setting cannot capture the whole community while simultaneously enumerating

Prochlorococcus, as a result picoeukaryotes were excluded from this analysis.

Correlations and Model II linear regressions were calculated to asses whether distance

from shore, a proximity for naturaI environmental gradients, explained the variability in

Prochlorococcus, Synechococcus, and total bacterial concentrations. Distance from shore

did not explain any variability for Prochlorococcus (r =0.193, p =0.293, n=77),

Synechococcus (r =-0.209, p =0.065, n =77), and total bacteria (r =-0.12, p =0.293, n =77)

concentrations.

Environmental Variables: Nutrients

Phosphate and nitrite concentration were below detection limits (l JiM) for all

samples. Nitrate was detectable in 25 out of 59 samples at concentrations ranging from

I J.!.M to 3.3 J.!.M. No significant relationships were found with nitrate and

Prochlorococcus. Synechococcus, or total bacteria concentrations.

Environmental Variables Correlations with AAPs Abundgnce

Model II linear regressions were calculated to assess the relationship between

environmental variables and AAP abundances (Table 2). In spite of significant variability

in environmental variables, no relationships were found between AAP cellslml and

distance from shore (um) (r =-0.181 ,p =0.111, n =79), nitrate (r =0.03, p =0.965, n =25),

chlorophyll greater than 2 J.l.m (r =0.0518, n =77) and Prochlorococcus (r =-0.220, p

17 =0.033, n =94). However, AAPs were correlated with total bacteria (r =0.784, p

=<0.005. n= 94), Synechococcus (r =0.871. p =<0.005, n =94). total chlorophyll

(r,=O.802. p =<0.005,94). and 0-2 J.UIl chlorophyll a (r =0.743. p =<0.005, n =77). In

general, the relationships between AAP abundance and Synechococcus. total bacteria, 0-2

!lm chlorophyll a, and total chlorophyll were largely driven by the variability observed in

Kaneohe Bay. At the other sampling sites. the regressions were consistent with the

patterns observed in Kaneohe Bay. but the low variability in AAPs and environmental

variables reduced the statistical power of comparison (FIgUre 6).

Outliers (2 standard deviations above the mean) were identified by box and

whisker plots and removed from the analyses to determine if the linear regression were

heavily influenced by extreme values. Further statistical analyses were performed on

positive relationships to confirm that linear regression assumptions of normality were

met. Transformations and non-parametric analysis were also made. Data was also fourth

root transformed and model II regressions were recalculated. For all of these analyses. the

r values changed slightly but the relationships were still consistent. The Spearman's rank

correlation non-parametric analysis was performed.

DISCUSSION:

Overall AAP concentrations were highest near shore, particularly in Kaneohe Bay

where concentrations were greatest near the river input to the central part of the bay.

Kaneohe Bay has two river inputs (southern and central bay), neither of which

significantly affected the salinity at the sampling site, but the central bay was visually

18 more turbid and murky. AAPs have been reported to be influenced by river inputs due

to increased particle load, decreased light attenuation, freshwater source of AAPs, and

changes in saIinity (Waidner and Kirchman, 2007 ). For example, Waidner and

Kirchman's study (2007) of the Delaware and Chesapeake estuaries reported the highest

AAP concentrations at the turbidity maximum and AAP concentrations were correlated

to light attenuation and particle associated AAPs. While my study did not measure light

attenuation or particle associated AAPs. I can expect that our high AAP concentrations

near these river inputs were influenced by light and particle associated AAPs.

Kaneohe Bay was the only location with a documented river source, so I also

considered other environmental characteristics such as benthic habitat and coastal water

residence time as factors controlling AAP bacterial variability. Interestingly, a common

trend appeared when grouping islands into those with lagoons protected by barriers reefs

(Kaneohe Bay, Lord Howe, and Aniwa) and those with steep cliffs and fringing reefs

(Futuna and Molokai) (Whittaker, 1998). AAP concentrations decreased with distance

from shore for islands with barrier reefs, in contrast islands with steep cliffs had no

patterns or had the highest concentrations furthest from shore. These trends maybe linked

to benthic habitats that influence the residence time of coastal waters and circulation

patterns.

While there were some weak relationships between AAPs and the physical

variables, the strongest correlations were observed for biological variables. Positive

relationships were documented for total chlorophyll, Synechococcus, total bacteria, and

the 0 - 2 !lJl1 chlorophyll fraction (Table 2 and Figure 6). Positive relationships between

19 total chlorophyll and AAP abundance have been well documented since Kolber and

collaborators (2000) open ocean description of AAPs. The actual mechanisms as to why

AAPs are correlated with total chlorophyll are unknown. One current idea is that AAPs

are similar to heterotrophic bacteria, whose relationship to chlorophyll is tightly coupled

due to the dependence on DOM derived from leaky phytopJankton cells. Another

possibility is AAPs have similar requirements as phytoplankton, therefore are in direct

competition fur these resources; hence the co-occurrence.

DOM derived from phytoplankton is complex and biological communities that

utilize freshly produced DOM vary (\ependiog on DOM pool constituents (Teira et al.,

2008, Hold et al., 2001, Schafer et al., 2002). Experiments by Pinhassi and collaborators

(2004) demonstrated that dominance by a specific species of phytoplankton can alter

bacteriopJankton community composition due to the availability of preferred DOM

sources. The most eutrophic site Kaneohe Bay was dominated by Synechococcus and

small picoeukaryotes (Cox et al., 2006), whereas all other locations with concentrations

of AAPs at 1 % or less were dominated by Prochlorococcus. While it is difficult to infer

if AAPs are influenced by DOM produced from different picocyanobacteria1

communities, we do know that AAPs will be influenced by the composition of the DOC

pool due to their dependence on organic carbon for growth.

Of all the relationships reported here I find that the most revealing and excitiog

was the positive relationship between AAPs and Synechococcus. Synechococcus is one

of the most abundant picocyanobacteria in the ocean and peak abundances are observed

near coastal upwelling regions, as well as periodically in oligotrophic waters

20 (Zwirglmaier et ai, 2008, ZwirgImaier et ai, 2007). Similar patterns have been

described for AAPs, and suggest that similar mechanisms are controlling these two

bacterial groups. In addition to similar distribution patterns, whole genome analysis and

QPCRI fish probe targeted Synechococcus ecotypes reveals that genetically different

types of Synechococcus are capable of using phototrophic and heterotrophic metabolisms

(Palenik et ai, 2006). The mechanisms that control the whole Synechococcus genus as

well as the different ecotype distributious are still unknown. I speculate that better

understanding of Synechococcus distributions could reveal what controls AAP

distributions due to their shared resources and genetic evidence of open ocean and coastal

types.

In conclusion, I have reported Pacific Island transects with AAP estimates similar

to oligotrophic locations and other volcanic islands (Schwalbach and Fuhrman, 2005).

The low nutrients, chlorophyll a concentrations, and the dominance of Prochlorococcus

at these sites suggest they are oligotrophic environments, which is not surprising due to

their remote locations within the Pacific Ocean. Low abundance of AAPs reported here,

and in other studies, suggests AAPs playa less critical role in the oligotrophic ocean

carbon cycle. However, even at < I % of the total bacterial popu1ation, AAPs may still

contribute significantly. Further research is needed on AAP bacterial growth rates, strain

specific amount of Bchl a pigment, Variables affecting photosynthetic efficiency, and

susceptibility to grazing. In contrast to these oligotrophic locations, I was able to sample

Kaneohe Bay that represented a mesotrophic body of water, and AAP estimates increased

with increasing total chlorophyll a. While this relationship is not new (Kolber et ai,

2(01) it raises the question of AAPs importance in coastal waters. Previous studies

report that AAP's comprise up to 34% of the total bacterial community, but their

importance are not considered important to coastal water carbon cycling due to the

dominance of large phytopJankton (Moran. 2007).

21

J

laonQ"tLoCkl

· ltQ

C . F6 -'''II ·'i 5 ."

_'8.5\

"" · 111.53

-III.S.

-11156 170" 110 1' 170.2 170.21 170.22 17'On 1702~ 110~

~

". A

21 .2

" , " ~.

~.

. OM . MOL1

' MOl2 ... 1582 -1M ·1$1.8 .'57~$7_ · 1572 ·157 ·1S6.f1

""r A ~ . • ~,'~"""'-" " CD _e: '

• ,,>

r ..... ,

,~ D ' ANO

",

I. ,~.

". , .... , .. .... -

" 8 " "

I" '" ' CA7

~

.. -CM

• CAS • CAB

CAB

28 -125 . 1~ 123 122 .121 .121) _I'll ·118 _\17 · lIe

.:11,5

.31.$2 E

"'~

~1 .5I! LHl

'->1 ,.51 .

"" "'~

Ltt2

-

LH3 LHS&B. · lH4 LH7

~ ... 158.&5 151.0 lM,1IS ISO 15005 l !it I -FIGURE 1: Locations sampled within the Pacific Ocean. A.) Island of Oahu and Molokai with an expanded view of

Kaneohe Bay.B. ) Californ ia transect C.) Island of Futuna D.) Island of Aniwa E.) Lord Howe Island

22

23

45

40

35 30

... 25 (,)

20

15 10

5

0 1.00E+00 1.00E+02 1.00E+04 1.00E+06 1.00E+08

AAP~ Standard Concentration (cellS/mI)

Figure 2: Average cycle threshold (Ct) for all QPCR assays versus AAP standard concentrations determined by flow cytometry. Average standard detection limit (- - ).Standard deviation bars are +/- one deviation.

y=-2.2(ln(x»+4B.2 with R2=O.99.

A) AAP cells/ml B.)

200 400 600 800

50

..-E 100 ...... .c -Do

~150

I

200 I 200

250

o

•

% AAP/Bacterla

0.03 0.06 0.09

24

0.12

Figure 3: Depth profiles of AAP abundance determined by QPCR of the pufM gene at Oahu Stations OA 1 ( diamonds) and OA4 ( squares). A.) AAP cellslml. Dashed black line represents the no template control. Standard deviations

are +1- one deviation. B.) % AAP/Bacteria.

A. 25

7 .00E+03 7.ooE+03 I

~ 6.ooE+03 ,

6 .00E+03 .§ 5.ooE+03 j ~ 4.OOE+03

5 .00E+03 i .. u

3.ooE+03 j Q.

iJ-~ ... 2.ooE+03 4 .00E+03 " ...

B 3 .00E+03 J

1.ooE+03 II -n. 0.00E+00

~ ~

0 2 4 6 8 10

2 .00E+03 T

Nm's from shore f .1.

f 1.00E+0 3

!: II • O.OO E+OO

0 100 200 3 00 400

Nm's from shore

B.

1 l til 0 .7: ]

I f <> •

0 .5

t.:~ '. (1j 0 .75

t ·c ~ 0 .25 2 0 &

U & .. (1j

t <> • & &

aJ 0.5 • 0 • ~

0 2 4 6 8 10 a.. « Nm's fro m sho re « ~ 0.25 0 •

• • • 0 ---,------- --- --,

0 100 200 300 400

Nm's from shore

Figure 4: AAP abundance for Aniwa (+), Futuna (- ), Lord Howe (A), Molokai (.0.), Oahu (. ), and California ( ) estimated by QPCR of the pufM gene. A. ) AAP celislml. Orange line ind icates no template control. Standard deviations are +1- one deviation . B.) % AAP/Bacteria

26 A.

E ~

.!!1 Qi u 0..

~

B.

2.000.05

1 1.S0E+OS

1.00E+OS "

S.OOE+04 ~ O.OOE+OO

6

'" .~ 4.S ti '" (;fl 3

0.. <t <t ~ 1.S o

0

Iii

o f-­o

1

... • <>

~ ~ <.:>

Q.

:::

2.00E+OS ]

1.S0E+OS I

1.00E.OS .:

S.OOE+04

O.OOE.OO

0

2

Nm from s hore

6

'" ·c Q) 4.5 • tl

'" ~ 3 • ~ 1.5

2

/'tn's f rom shore

3

• •

•

• >'!. 0

0 j-__ ~ __ ,, __ ..c:.~

0 2 3

Nm's from shore

• ...

<>r- . --~

2 3

Nm from shore

•

3

4

• 4

4

Figure 5: AAP abundance estimated with QPCR of the pufM gene for Kaneohe Bay north and south channel samples taken on 6.13.06, 8.4.06, and 6.21 .07. North channel samples are indicated as 6.13.06 ( . ), 8.4.06 ( • ), and 6.21 .07 ( ). South channel samples are indicated as 6.13.06( <» , 8.4.06 ( ... ), and 6.21 .07 ( . ) . A. ) AAP cells/ml. Orange line is the no template control. Standard deviations are +1- one deviation B.) % AAP/Bacteria.

4

A.

B.

C.

1.60E+05

'§ 1.20E+05

~ ~ 8.00E+04

Q. <I: <I: 4.00E+04

O.OOE+OO

O.OOE+OO

1.60E+05

E 1.20E+05

:!l "il 8.00E+04 Q. <[

<I: 4.00E+04

O.OOE+OO

O.OOE+OO

1.60E+05

E 1.20E+05

:!l "il 8.00E+04

Q. " <[ <I: 4.00E+04

O.OOE+OO

0.000

27

r = 0.87

2.00E+05 4.00E+05 6.00E+05 8.00E+05

Synechococcus (cells/ml)

r = 0.78

1.50E+06 3.00E+06 4.50E+06

Total Bacteria (cells/ml)

r = 0.80

0.750 1.500 2.250 3.000 3.750 4.500

Total Chlorophyll (1l9!L)

Figure 6: Relationships between AAPs and Synechococcus (A), Total bacteria (B) , and Total ch lorophyll (C). Relationships were calculated based on the whole

dataset but data here is grouped into locations: Aniwa (e) , Futuna (. ). Lord Howe (~ ). California ( ), Molokai (~ ), Oahu (e), Kaneohe Bay 6.13 .06 (0). 8.4.06 (. ) 6.21 .07 (a)

Table 1: List of sampled locations, dates sampled, and the length of the transects conducted.Filter sizes used for size 28 fractionated chlorophyll data are listed. Kaneohe Bay was only sampled once on 6.21.07 with the 5 J.lm size.

Chlorophyll N02• N03 P04+ FIlter Sizes

Location Sample Date (mmlddlyyyy) Length of Transect (nm) Salinity Nutrients Nutrients (Jun)

Califomia 07117/2007-08/0212007 332.48 x nla nla 0.22,0.8 Oahu 11/5/2006 31.44 x n/a x 0.22,0.8

Molokai 11/6/2006 31.08 x x x 0.22,0.8 Aniwa 1/21/2007 3 nla nla x 0.22,2,5 Futuna 1/21/2007 1.59 n/a n/a x 0.22,2

Lord Howe 216/2007 15.79 nla x x 0.22,2 Kanoehe Bay 06/13/2006,08/04/2006,06/21/2007 3.7,3.4 x x x 0.22,2,5

Table 2: Regressions analysis of AAPs. n= number of samples, r = r value, 29

p= p-value, m = slope of line, and lastly spearman's rank value determined with minitab.

Variables n r p m spearman's rank value

Synechococcus (cells/ml) 94 0.871 <0.005 0.1859 0.832 Total chlorophyll a (J.Lg/L) 94 0.802 <0.005 3.53E+04 0.852 Total bacteria (cells/ml) 94 0.784 <0.005 0.0365 0.823 0-2 urn chlorophyll a (J.Lg/L) 77 0.743 <0.005 6.08E+04 0.758 Temperature (DC) 94 0.344 <0.005 7.2SE+03 0.335 Nitrate (J.LM) 25 0.03 0.964 4.98E+04 0.238 Salinity 72 -0.101 0.399 -1.41E+04 0.424 Nautical miles 79 -0.181 0.111 -268.27 -0.334

30 Chapter 3: Aerobic Anoxygenic Phototrophic Bacterial Diversity

Photosynthetic bacteria are found within 5 bacterial phyla (Cyanobacteria,

Proteobacteria, Chlorobi, Chlorof1exi, and Fl11Ilicutes); aerobic anoxygenic phototrophic

bacteria (AAPs) cluster within the proteobacteria phyla (Bryant and Frigaard, 2006).

Earlier diversity studies of AAPs found them all clustered within the a proteobacteria

subclass, closely aligning with their anaerobic counterparts and other chemotaxic bacteria

(Yurkov and Beatty. 1998b). However. recent evidence from culture and culture­

independent studies have shown AAPs to be more broadly distributed, with members in

the y and p proteobacterial subclasses (Beja et al.. 2002). Based on 168 rRNA sequences.

AAP bacteria cluster with Roseobacter (a- 3). Rhodobacter (a-I). Erythrobacter (a-4).

OM60 (y). and NOR5 (y) genera. Phylogenetic relationships among 168 and

photosynthetic genes are incongruent suggesting that lateral transfer of the photosynthetic

superoperon has produced a large diversity of organisms with the same photosynthetic

system and pigments. in spite of more conserved photosynthetic genes (Nagashlma et al..

1997. Yurkov and Beatty. 1998). The photosynthetic superoperon is commonly used

genetically to describe diversity of AAPs due to the conserved regions and the difficulty

to target AAPs diverse sequences using 168 rONA.

Our understanding of how these different AAP groups are distrIbuted through the

ocean has been changing drastically as more AAPs are found in different proteobacteria

subclasses. Originally the majority of AAPs were thought to be related to a

proteobacteria including Erythrobacter strains. but in recent genetic studies most of the

diversity is found to fall outside of this group (Beja et al.. 2002). Further advances in

primer design have revealed y and a proteobacterial groups dominate the AAP

31 community (yutin et aI., 2(05). For example, in the Baltic Sea AAP '1 proteobacteria

were detected with FISH probes and dominated the AAP community (Masin et aI., 2006).

In the global ocean sampling (GOS) expedition the Roseobacter-like group dominated

the oligotrophic AAP community. In addition. the GOS revealed new unknown groups

throughout the Atlantic and Pacific Ocean previously missed by current primers (Yutin et

aI., 2007). Thus, AAPs are far more diverse than originally thought.

Evidence from GOS and other studies have suggested that this diversity is

structured along near shore/offshore gradients. For example, coastal waters sampled by

the GOS expedition demonstrated AAP groups, like p..proteobacteria, were consistently

detected near freshwater sources, as well as other groups that seemed more abundant in

coastal waters. The GOS dataset for AAP is the first evidence of coastal and open-ocean

types and further research is still needed on a smaller spatial scale to see if these patterns

are general and if so, what environmental variables are driving them. In support of the

patterns found by the GOS sampling, another global survey reported a strong relationship

between AAP diversity and chlorophyll concentrations (Jiao et aI., 2(07). High diversity

was reported when chlorophyll concentrations were the lowest and vise versa for high

chlorophyll concentrations. Thus, there is evidence that diversity is structured with

respect to environmental gradients and that specific clades maybe ecotypes. are

associated with particular ecologies.

While both of these global scale surveys contribute to the understanding of AAP

ecology on very large scales, my study bridges the gap between relating the abnndance of

AAPs to environmental variables and linking these patterns to changes in the

phylogenetic diversity. Using natural near shore and offshore environmental gradients

32 near Pacific Islands, I explored the diversity of AAPs as distance from shore increased

using dissociations curves and clone hbraries. Abundance and environmental data

described in chapter 2 were used to understand how these variables influence AAPs

diversity. I hypothesize there will be broad diversity changes between near shore and

offshore sites and AAP genetic types will be structured along environmental gradients.

METHODS

Sampling

Eight near shore/offshore transects were conducted within the Pacific Ocean at

the following locations: California (280 36'N, 1250 4'W), Aniwa (1So IO'S,169° 15'E),

Lord Howe (31 0 33'S, lS9° OS'E), Futuna (190 32'S, 1700 I3'E), Oahu (210 2S'N, lS7"

S9'W), Molokai (21 0 OS'N, IS7° 02'W), and 2 within Kaneohe Bay located on the island

of Oahu (21 0 27' 3S"N, lS7° 4S' IS" W). Surface water samples were collected in dark

amber bottles for environmental DNA. Two of the offshore transects which are located

within Kaneohe Bay were sampled three times (6.13.06, S.4.06, & 6.21.07) (Figure 1).

DNA was collected by filtering 100 or SO mI of water through 0.22 J.I.Ill

polycarbonate fihers (GE Water and Process Technologies) under low vacuum pressure

followed by 3 mI of preservation solution (10mM Tris,IOOmM EDTA,O.S NaCI at a pH

of S.O). Filters were placed in bead beating tubes and stored in a -SO °C freezer. For

details on DNA extraction and QPCR techniques see chapter 2 methods.

33 Preliminary AAP Diversity Studv at Kaneohe Bay

Due to genetic variability in AAPs, it is important to use primers that capture their

full diversity. To asses if I chose primers that would adequately represent the diversity in

the Pacific islands, three preliminary clone libraries were performed from 3 stations in

Kaneohe Bay (21 0 27' 3S"N, IS7° 48' IS" W). Partial pufM products were amplified

using the primersputM(I669)F (S'OON AAY YTNTWYTAY AAY CCNTIY CA3')

and putM(I840)R (S'CCA TSG TCC AGK GKC AGA A 3') and the following PCR

conditions: 9S °C for 10 min, then 40 cycles of 9S °C for 1 min, 48.6 °C for 1 min, 72 °C

for S seconds, and 10 min at 72 °C (modified from Waidner and Kirchman, 200S) .

Twenty five microliter reactions were performed using the following solutions: 17 f.Il of

water, 2.S J.Ll of lOX buffer (Fermentas), 2 f.Il of 25 mM MgC~ (Fermentas), o.S f.Il of 10

mM dNTP's (Promega) , 0.5 f.Il of25 ~ stock for putM(l669)F and putM(l840)R

primers, 0.1 J.Ll of Hot Start Taq (Fermentas), and 1111 of DNA template. For each hllrary,

triplicste PCR reactions were run and pooled together when visualizing on a 1 % agarose

gel stained with ethidium bromide. PCR products were extracted from the gel using a

Qiagen gel extraction kit via manufacture's instructions. Once re-snspended, DNA was

inserted into a pCR4-TOPO (Invitrogen) vector and then transformed into E-coli using a

TOPO Cloning kit. Clones were grown overnight on LB plates with SO Ilg/L of

kanamycin (Shelton Scientific) and randomly picked for sequencing. Sequences were

cleaned and oriented using Sequencher and aligned in ARB (Ludwig, W. et aI., 2004).

Neighbor-joining trees were made in ARB.

34

PJifM Clone Libraries

Six clone libraries were created to explore the difference between offshore and

near shore pufM containing bacterial communities. The criteria for selecting samples to

perform clone h"braries on were to have near shore and offshore stations that had distinct

dissociation curves (Figure 7). Two clone libraries each were constructed from the

islands of Aniwa, Oahu. and the coast of California. To determine how many clones

would be needed to see a change in diversity structure, sequences obtained from Kaneohe

Bay preliminary clone libraries were used to create rarefraction curves nsing DOTUR

(Schloss and Handelsman, 2005). Based on those preliminary rarefractions curves, 48

clones per library were chosen for each of the new libraries. Six libraries were

constructed as described above, except different primers and PCR program were used to

amplify the partial pufM PCR product. PCR primers differed from the Kaneohe Bay

PCR reactions in that pufM557F and pufMW A WR were used in the QPCR reaction to

determine AAP abundance. The new primers amplify a smaller portion of the pufM gene

but are inside of the amplified product used for the Kaneohe Bay libraries. Partial pufM

PCR product (240bp) was amplified usingpufM557F (5'TAC GGS AAC CTO TWC

TAC 3') and pufM-W A WR (5'OCR AAC CAC CAN GeC CA 3') primers and the

following PCR conditions: 95 ·C for 10:00 min, then 40 cycles of 95·C for 15 sec, 56· C

for 45 sec, 72 ·C for 45 sec. Twenty five microliter reactions were conducted in triplicate

with the following reagents (17.771Jl water, 2.51JlI0X buffer with 15mM MgC12

(Sigma), 0.241Jl of 10 pM primer stock, 0.25 IJl of Jump Start Taq (Sigma), and 2.5 IJl of

DNA template).

35

Phylogenetic clusters observed in the clone hllraries results were related back to

original dissociation curves to understand which clusters represented observed

dissociation peaks. Dissociation curves were created for 10 clones from each library. The

clones picked were representatives of sequences from major clusters.

Comoaring Clone libraries using SONS and confirming clone library saturation

To determine the level of sampling coverage, rarefraction curves were constructed

for each hbrary following Schloss and Handelsman (2005). Near shore and offshore

libraries at each location were compared using SONS (Schloss and Handelsman, 2006).

Near shore and offshore libraries were compared within an island, and were also pooled

among all islands and compared. Comparisons within pooled near shore or offshore

sequences confirmed that community structure was comparable. Kaneohe Bay was

excluded from this analysis due to a lack of an offshore library.

Examining Diversity ofAAPs using Dissociation Curves

Dissociation curves were measured on p!f/M amplicons obtained from the AAP

enumeration protocol (as descnDed in Chapter 1 methods) using the following protocol:

95°C for 15 sec, 56°C for 30 sec, then 95 °C for 30 sec. Between 56°C and 95 °C the

heating rate was at a constant 0.2 ° C per second. The change in fluorescence with change

in temperature was calculated using the 7300 SDS software and plotted versus

temperature. Raw data of the absolute change in fluorescence with change in temperature

was imported into Excel for further analysis. Each station's dissociation curves was

averaged and nonnalized to the change in fluorescence with change in temperature

over proportional abundance. Normalized data was plotted against temperature using

Mat1ab.

36

Dissociation curves were composed of multiple temperature peaks per sample and

to quantify the number of peaks a Matlab code created by Kevin Bartlett was used. The

program identified peaks when the slope of the line changed signs. After applying the

matlab code, major temperature peaks were manually identified and used for diversity

analysis.

RESULTS

Kaneohe B~ Preliminary Diversity Study Results:

For Kaneohe Bay 89 sequences were blasted in the NCBI website and all of the

top scores were similar to p!f/M sequences with e-values ranging from 2e-17 to ge-87. Only

2 of the Kaneohe Bay sequences were similar to cultured AAP p!f/M sequences.

Phylogenetic analysis of Kaneohe Bay sequences revealed a diverse assemblage of

sequences with two main clusters containing the majority of the sequences (22, 37

sequences) (Data not shown). Sequences did not cluster with known cultured AAPs but

instead clustered with sequences obtained from metagenomics, BAC's, and clone

libraries analysis.

37 Clone Libraries Results:

Clone libraries were selected based on dissociation curves that would have broad

diversity difference based on the concept that different dissociation temperature peaks

represent sequences with different base pair combinations for a targeted gene. Aniwa,

California, and Oahu's near shore and offshore dissociation curves were composed of

varying number of temperature peaks, as well as different temperatures where the peaks

occurred (Figure 7). For example, Aniwa's near shore (ANI) dissociation curve had one

peak at 86 °C. Whereas Aniwa's offshore (AN6) dissociation curve had three peaks at 83,

85, 88.5 °C. Thns, suggesting one AAP genetic population should be present for the near

shore (ANI) clone hDrary and three AAP genetic populations for the offshore library

(AN6). California's near shore (CA2) dissociation curve has three temperature peaks at

82, 84, and 86 oc, whereas the offshore (CM) curve has one temperature peak at 83OC.

Oahu's near shore and offshore dissociation curves overlapped significantly suggesting

similar AAP genetic populations will be found for both near shore and offshore clone

libraries.

All sequences obtained from clone libraries were blasted in the NCBI database

and the majority of the top scores were identified as pujM genes with e-values ranging

from 2 e-!? to e-H. Only California (total sequences 89) and Oahu (total sequences 92) had

one sequence that did not have a top match with the pufM gene and these were removed

from the analysis. A majority of sequences were identified with pufM sequences obtained

from clone horaries, BAC's, or metagenomics. Aniwa sequences had the largest number

of sequences (32%) similar to pufM sequences derived from cultures within the a

proteobacteria subclass such as Rhodobacter (a-I), Roseobacter (a-3), and Erythrobacter

38 (a-4) . For Oahu and California less than 5% of all sequences matched with culture

pufM sequences, respectively.

High diversity was observed for all libraries and distances between clusters

ranged from 94% to -80% similar. In general within an island. clone libraries grouped

into phylogenetic clusters that were dominated by either near shore or offshore

sequences. As well as the differentiation between near shore and offshore sequences

occurring at the subclass leveL For example, California and Aniwa had distinct clusters

that were dominated by either near shore or offshore sequences (Figure 8A,B). Oahu's

tree formed 5 distinct clusters each mixed with near shore and offshore sequences within

a cluster (Figure 8C).

California's neighbor-joining tree for CA2 and CAS had clusters separated by

genetic distances of up to 85% similar; these distances suggest genetic difference of at

least class or subclass levels (Figore 8A). Near shore sequences (CA2,red in Figure 8A)

arranged into five clusters, with three of the clusters containing only near shore

sequences. Offshore sequences (CAS, blue in Figure 8B) arranged into four clusters, with

two of the clusters containing only offshore sequences. While the other CAS clusters

were dominated by offshore sequences.

Aniwa's neighbor-joining tree was segregated between ANI and AN6 sequences.

These clusters were genetically separated by a distance of 0.1 to 0.15 substitutions

(Figure 8B). Near shore sequences (ANI, red in Figure 8B) arranged into three clusters

with two of the clusters dominated by near shore sequences. Offshore sequences (AN6,

blue in Figure 8B) were represented by one major cluster, which contained approximately

80% of all offshore sequences. The offshore cluster could be bifurcated into two

clusters with a genetic distance of 0.1 substitutions separating them.

39

Oahu's neighbor-joining tree had overlapping sequences from both stations within

clusters. Five major clusters composed Oahu's neighbor joining tree and four of these

clusters were composed of near shore and offshore sequences. Only one cluster was

composed of near shore sequences (OAl, red in Figure 8C). Genetic distances between

clusters were generally 0.08 to 0.20 substitutions (Figure 8C).

Dissociation Curves of Selected Clone Library Sequences:

Dissociation curves produced from QPCR of puj'M clones were used to confirm

which clusters from the phylogenetic trees were contributing to the original dissociation

curve temperature peaks (FJgIlIll7). Almost all of the clusters had distinct temperature

ranges as reflected by the original dissociation curves (Figure 7, Figure 8). In general. a

majority of clusters could be related back to the original dissociation peaks, while a few

clusters had dissociation temperature peaks that were not represented in the original

dissociation curve, or vise versa. Therefore, more clones would need to be processed to

resolve these differences.

California near shore station (CA2) had four phylogenetic clusters with

dissociation temperatures ranging from 84.5 to 86.3 "C. This is consistent with the major

peak in the original dissociation curve at 85. °C (Figure 78, Figure 8A). Two near shore

peaks (82, 83"C) from the original dissociation curve were not identified by puj'M clones.

California offshore station (CAS) ouly had one phylogenetic cluster with temperatures

representing the dissociation peak of 83.3 ·C. While two offshore clusters had

temperatures ranging from 84.5 to 87.8"C, which was not identified as peaks.

40

Aniwa's offshore station (AN6) had two dissociation peaks at 84.8 and 88.4·C

and appropriate temperature ranges were identified to belong to two different clusters

(Figure 7 A, Figure 8B). No offshore sequences had a dissociation temperature peak that

matched the peak of 83 "C from the original dissociation curve. Aniwa near shore station

(ANI) had no partial pujM clones that had dissociation temperatures that matched the

original dissociation peak at 86.1 ·C. Higher or lower dissociation temperatures were

observed for four main clusters that were composed of near shore sequences. One cluster

had dissociation temperatures ranging between 84.1 to 85.5 ·C and the other three

clusters had higher dissociation temperatures ranging from 87.4 to 88.4 ·C.

For Oahu's near shore and offshore stations, clusters represented dissociation

peaks and had distinct temperature ranges. Four clusters were identified to have

dissociation temperature peaks of 86-87 ·C, while one cluster represented the lower

dissociation temperature peak (84 .C). The near shore station (OAl) lacks a cluster

dominated by near shore sequences, which would have represented the original

dissociation temperature peak of 88.5 ·C.

10 general, multiple clusters for each station had temperature peaks that

represented one major peak in the original dissociation curve. While individual AAP

clusters, at a particular genetic distance, should represent a single dissociation peaks it is

possible that these different clusters have the same dissociation temperature due to

various combinatious of genetic deletions and insertions. As well as some genetic

distances between clusters are small, therefore small changes in genetic sequences mean

similar dissociation temperatures. Thus genetic identification of the various clusters is

needed to determine at what genetic level dissociation curves identify differences

between AAP populations.

Clone Library Community Structure

41

Aniwa was the most diverse AAP commnnity based on the high nnmber of OTUs,

at the 80% similarity level (Table 3). Aniwa libraries had different community structures

and very little overlap occurred between OTUs. In contrast, 43% of Oahu's OTU's were

shared between near shore and offshore sequences. As well as 75% of the community

structure was similar for Oahu's near shore and offshore commnnities, at the 80%

similarity level (FIgure 8e). California had the lowest diversity compared to the rest of

the locations, based on the number of OTU's. Similar to Oahu, California had the same

percentage of shared OTU's, but the similarity between near shore and offshore

communities was 52%, at the 80% similarity level.

At the 80% similarity level, all rarefraction curves appeared to be approaching

saturation. except for Aniwa station ANl (Figure 98). AN1lack of saturation is

explained by the higher number ofOTU's observed.

Nearshore and Offshore Diversity and Community Structure:

A global dataset exploring near shore and offshore diversity among all stations

was made by combining all of the sequences obtained from Aniwa, Oahu, and California

clone libraries. A neighbor joining tree rooted to Rhodobacter veldkampii was created to

explore if a separation between near shore and offshore sequences observed within a

location would occur when all locations were pooled together (Figure 10). A

bifurcation separates a majority of near shore and offshore sequences at a genetic

distance of 0.17 substitutions.

42

To further explore the separation of near shore and offshore sequences, I

statistically compared the similarities between near shore/offshore community structures

(Table 3). Prior to this comparison, I determined if the combined libraries of near shore

or offshore had similar commnnity structure, to then allow for a comparison between near

shore and offshore hbraries. The near shore lihrary was the most diverse based on

number of OTUs, and community structure was only similar between Aniwa and Oahu

sequences. California sequences were quite different from the rest of the locations, which

influenced the neighbor-joining trees' topology. For the offshore hl!rary there were fewer

OTUs and the community structure between locations was similar. The variability

observed for near shore and offshore libraries community structure suggests that some of

the variability is due to location, as well as distance from shore affecting the AAP

community composition.

At the 80% sequence similarity level, the composition of the combined near shore

and offshore libraries were only were 52% similar. Therefore, 50% of the AAP

community did not overlap and was unique to either near shore or offshore sequences.

Rarefraction curves saturated at the 85 % similarity level

Environmental variables from chapter 2 were included to determine if the

observed separation of near shore/offshore sequences in the combined hllrary was

explained by total chlorophyll, total bacteria, and Synechococcus concentrations (FIgUre

11). Variables from the same location clnstered together, suggesting enviromnental

43 variables explained the lack of AAP community overlap within a location, as well as

between locations. For example. Synechococcus concentrations have a near

shore/offshore separation, where high Synechococcus concentrations relate to near shore

AAP diversity (top of circle tree) (Figure 11C). Whereas. chlorophyll concentrations

cluster with AAP sequences associated with a specific locations and environment (Figure

l1A). The patterns of AAP types clustering with environmental variables suggest AAPs

diversity could be structured along environmental gradients. as well as AAP genetic types

maybe associated with particular environments.

Broad AAP Diversity Determined by Dissociation Curves

Based on clone library described above. the formation and disappearance of

dissociation peaks can be used as tools to screen for broad changes in community

structure. Thus. I used all of the dissociation curves created in chapter 2 and analyzed

changes in diversity as a function of distance from shore (a proxy for environmental

gradients) (Figure 12. Figure 13).

Each station had one to multiple (maximnm 3) dissociation temperature peaks.

Dissociation curve temperatures were between 81°C to 8SOC. Dissociation curves varied

within a location and between locations by the number of temperature peaks observed, as

well as at which temperature these peaks occurred at. There were two alternate significant

trends that were present: 1.) one dominate peak for stations closest to shore with the

number of peaks increasing as distance from shore increased; 2.) multiple peaks were

observed fur stations closest to shore and number of peaks decreased as distance from

shore increased; Aniwa, Califurnia, and Lord Howe fullowed the first trend, Oahu

followed the second trend. Futuna's near shore dissociation curves were inconsistent,

but the offshore stations had a distinct signature.

DISCUSSION

44

In this study, AAPs diversity was found to change within an island, as well as

between locations. For example, nmltiple peaks from dissociation curves were observed

to change within an island, as well as between locations. Using clone hDraries, the

observed changes in dissociation temperature peaks between stations were shown to be

related to diverse phylogenetic clusters of AAPs. Previous studies using dissociation

curves report similar temperature ranges for AAP pufM amplicons, but mnJtiple

dissociation peaks have not been observed (Schwa1bach and Fuhrman, 2005, Du et al.,

2(06). The nmltiple peaks I detected maybe unique to these locations, or derived from the

selection of primers used in this study.

The changes in AAP diversity, observed with both clone libraries and dissociation

curves, occurred at the subclass or genera level, suggesting the differences in community

structure between near shore and offshore locations occurred at the a, p, and r

proteobacterla1 subclasses leveL While genetic distances between clone horaries clusters

were inferred from genetic distances calculated from the neighbor joining tree, the

specific subclasses a, Ii, and r proteobacterial were not identified in this study. While a,

Ii, and r subclass clusters were not identified in this study, Kaneohe Bay's dissociation

curves near freshwater sources had no major changes in AAP community composition

compared to other Kaneohe Bay stations, implying no freshwater p proteobacteria1

45 subclass was present near river stations. Within this dataset observed diversity changes

were speculated to be between the y and a proteobacteria subclasses.

Previous studies have demonstrated that y and a proteobacterial subclasses

dominate AAPs diversity (Y utin et ai, 2005, Y utin et al., 2007, Masin et al., 2007). Due

to the observed separation between near shore and offshore communities it is speculated

the changes in CODJlIDlnity structure was between open ocean and coastal y and a

proteobacteria AAP types. For example, y proteobacteria have been shown to dominate

AAP diversity in most marine environments; this subclass contains two phylotypes

obtained from both coastal and open ocean ecosystems (Hu et al., 2006, Cho et al., 2(07).

For a proteobacteria, Roseobacter and Roseobacter-like sequence types contribute most

substantially to oligotrophic environments compared to other a proteobacteria genera

(Y utin et al., 2007, Oz et al., 2005). In the GOS expedition, similar dominance of a and y

proteobacteria was observed, as well as AAP genera presence varying depending on the

environment (Y utin et al., 2007).

Along these offshore transects environmental variables chlorophyll, total bacteria,

and Synechococcus were shown to explain the variability in AAPs' diversity. Therefore,

suggesting AAPs diversity is structured along environmental gradients. In a previous

study of chlorophyll concentrations, AAP abundance, and AAP diversity, AAP

abundance and chlorophyll concentrations were shown to have a positive relationship,

where as AAP diversity and chlorophyll concentrations were negatively correlated (Jiao

et al., 2(07). This study observed that chlorophyll concentrations grouped with specific

AAP phylogenetic clusters, while having a positive relationship with AAP abundance. In

contrast to Jiao and collaborators, the relationship between AAP CODJlIDlnity structure and

46 chlorophyll concentration varied within a 10ca1ions as well as between loca1ions and

suggests other environmental variables other than chlorophyll inflnence AAP diversity.

For example. Kaneohe Bay's diversity was consistent along both transects, even though

Kaneohe Bay had the largest range of chlorophyll concentrations. Elevated AAP

abundance observed at the closest station to the river may infer that AAP abundance and

diversity is influence by particles, therefore increasing the number of AAP particle

associated species. which are contained within the r proteobacteria subclass. Whole

genome analyses of AAP r proteobacteria strain Congregibacter litoralis KT71. suggests

these organisms prefer suboxic environments with potential enzymes production common

to particle attached bacteria (Fuchs et al.. 2(07). The increased particle load near

Kaneohe Bay's river station suggests a suitable habitat for coastal r proteobacteria.

For all the other locations AAP abundance and AAP diversity were influenced by

loca1ion and types of benthic habitat. For example. Aniwa and Lord Howe had one

dissociation peak near shore where AAP abundance was high, and diversity became more

complex as AAP abundance decreased. As mentioned in Chapter 2 both of these islands

had barrier reefs and lagoons. In addition to chlorophyll clone hbrary results

demonstrated a separation of near shore and offshore AAP diversity correlating to

Synechococcus concentrations similar to the relationship described with AAP abundance.

As in chapter 2, I suggested this relationship is due to a common unidentified

environmental variable influencing both populations because they compete for similar

resources. or that Synechococcus produces DOM that influences AAP diversity and

abundance.

47 In conclusion, this study reports that AAP are phylogenetically diverse and that

some phylogenetic clusters appear to be associated with certain environments. For

example, the diversity of AAPs is correlated with chlorophyll, AAP abundance,

Synechococcus, and total bacterial concentrations. Different AAP diversity types suggest

these groups of AAPs perform different ecological roles depending on the environment

they occupy. Obtaining some of these sequence types as culture representatives will help

investigate how different AAPs might have different amounts of bacteriochlorophyll a

pigment, growth rates, or rates of carbon dioxide fixation. These details are imperative to

understanding AAPs importance in the oligotrophic and coastal carbon and energy

cycles.

A. I -ANI

0.8 - --AN6

0.6 ~~ ------- .. _ ....

Aniwa Clone Library Stations .......• ,,.,., :", , ~

~~ ...... --_ ...

0.4~_--,-_____ -

0.2

~,

.,., ., , " , , . , . ,

48

\

OL-____ L-____ L-____ L-____ ~ ____ ~ ____ ~ ____ ~ ____ _L ____ ~

B. t -CA2

0.8 __ -CA5

c. t -OAt

0.8 __ -OAJ

0.6

0.4 ... , .... ,-.............. 0.2

~O 81 82

California Clone Library Stations

~, , , .,' \

Oahu Clone Library Stations

83 84 85 86 87 88 89

Figure 7: Dissociation curves from the Island of Aniwa (A), California (B), and Oahu (C) used for clone library studies. Squares represent peaks in dissociation curves.

A. --- 0.1

Rhodobacter veldkampii B. --01

Rhodobacter veldkampii

c. --01

3 seq .

r----CA00078 00040

Rhodobac/er veldkampii

r-_-iIMW00039

.-___ CA00001 MW00027

I MW00029 5 seq. 86 'C (1)

l G ""00:00007 10 seq. 87 'C (5)

.r-CA00041

CA00005

ACOD11

~I 'M'00031

12 seq. 85 -88 'C (4)

CA00021

CA00027

CA00033

...e---4 4 seq. 85 'C (2)

'--I~ seq. 86 'C (2)

CAOOQ45

.----l CAOOOS3

.,~ 21 seq. 86 'C (4)

12 seq . 85-86 'C (3)

16 seq. 83-84 'C (3)

·'M'OOO15

v.>OOOS1

T • 6 seq. 88 'C (1)

00024

'M'00022 88 'C (1)

13 seq . 85 , 86'C (2)

r--'M>00043

1M'OOO" ...----WOOO35

00056

3 seq. 87 'C (1)

6 seq. 85, 89 'C (2)

34 seq . 88-89 'C (7)

U MW00020

I MW00028

.----MW00084

.----Ir MW00004 MW00021

9 seq . 84 'C (3)

7 seq .86 .- 87 'C (2)

MWO0033

23 seq. 87 'C (4)

5 seq. 88 'C(1)

49

Figure 8: Phylogenetic relationships (neighbor-joining) among AAP clones from (A) California, (8) Aniwa, and (C) Oahu inferred from partial pufM sequences (240 bp) . Pie charts represent the proportion of sequences that contribute to nearshore (red) or offshore (blue) . Temperature ranges represent the peaks of the dissociation curves found in those clusters, with the number of observations in parentheses.

A.) 50

40 "0 OJ 30 2: OJ til .0 20 0

::J I-0 10

0 0 5 10 15 20 25 30

B.) 40 "0 OJ 30 2: OJ til .0 20 0

::J I-0 10

0 0 10 20 30 40 50

C.) 40 "0 OJ 30 2: OJ til .0 20 0

::J I-

10 0

0 0 10 20 30 40 50

Number of sequences sampled

Figure 9: Rarefractions curves for Kaneohe 8ay station SC2 (A) , Aniwa AN1 (8), and Oahu OA3(C) . Dark blue line 100% similarity line, green line 99 % similar, red Ine 94% similar, and light blue 80% similar.

Figure 10: Neighbor joining tree of clones inferred from partial pufM (240 bp) sequences from Aniwa , Oahu ,and California . Green bars indicate clones from near shore, blue bars indicate clones from offshore.

A.

Y \l1IiIIlliN"",..". # " \ ,~ '" ',"'" "". \

1"'

- ~ l~\ \

\ .. ~7' ,.. ~' ~ ; ? 'f 'el

l;& ~ ., ...

- 0 01

oft-. /' --1 _J:-

I "I

,~ \

52

Figure 11 : Relative proportion of environmental variables total chlorophyll (A), total bacteria (B), or Synechococcus (C) overlaided onto a neighbor joining tree of clones inferred from partial pufM sequences from Aniwa , Oahu, and California , Blue branches indicate offshore sequences, green branches indicate near shore sequences,

0-0.5 nm

0.5 -1 nm

1-2nm

Aniwa Futuna KB-SC 1 1 1

I ~ I~ j~ 0.6 0.6 I 0.6

u u u

KB-NC

/'v~. \

1°·, 1°" I'" o 0 0' '-!

ro ~ M ~ ~ ~ ~ M ~ ~ M ~ M M W -_\"C) T_\"C) _\"C)

1 1

J 0.8 h. j 0.6 j 0.6

1o.' 1°"" 02 02

° • ~ ~ M ~ ~ ro ~ M ~ -Temperature (OC) Temperature (OC)

1

10.8

0..

10 ••

'.2 0 1 'J 8D 82 84 86 88

_('C)

I ,__ 1 '(\' 1 I

1·.8 \ .1··8 /.\. 1'.8 Vl' .... JO.8 0.6· \ ',I j 0.6 -.J \: 0.6 ;\~/.., .. :\',: j 0.6

10.4 1°·4 1°'" \ 1°" 02 0.2 0.2 '\ 0.2

o 0 0 0' 'J M ~ M M ~ ~ ~ M 86 88 M ~ 84 ~ ~ ~ 82 84 M ~

_\"C) __ \"C) T_\"C) __ \"C)

1 1 1

I ~ I~ 1M 2-4nm M M M

I~ I~ I~ 0.2 0.2 0.2 I .

° 0. ro ~ 84 U ~ M ~ 84 M ~ ~ ~ 84 86 88 _\"C) T_\"C) T_('C)

Figure 12: Dissociation curves for Aniwa, Futuna, Kaneohe Bay southern channel (KB-SC) and northern channel (KB-NC) seperated into categories of distance from shore (0- 0.5, 0.5-1,1-2, and 2-4 nautical miles). Squares represent temperature peaks.

53

0 -5 nm

5 - 10 nm

10- 100 nm

100 - 300 nm

> 300 nm

California j , i 0.'

0.'

J 0 .. ~ 0.'

•

It 0 IL--:":_-::-:-_,.,----,:~ 80 82 84 86 88

; 3 0.8

~ 0.' I 0.4

Temperature re)

•• . ~ .

•

l 0.2 ----:::;--;;,~88-­< 0 ' 80 II::! " 86 88

j 0: II 0.6

! 0.4

8" 0.2

Temperature re)

,~

\ .t 0 ' -\ >-

80 82 84 86 88 TemperallJl"fI (oc1

i O.s l • •

• ~ 0' , . i 0.4

! O! • \ <

80 112 84 86 88 Temperature (ee)

Lord Howe

i ~ 0'

0.'

J 04 0.2

0 80 " 84 " gg

Temperature ("C)

i 0'

• • <. or ,

\ \ 0.' , \ ..

104 . " '(. 0.2

< 0 80 " " 8' " TemperatlSe re) , ,

j 0.8

~ 0.' • r' 0.2 < 0

80 " 8' 86 88 Temperature ("C)

Oahu Molokai ! , -! 0.8

~ 0.' , .

.1 0.4

• •

to' 0 80 " 84 86 " TemperalUf1l COC)

; , •• J 0: .,,-,\ x\ i 08

0.' ' J l \ 0.' • .~ /' ' \ f 0.4 . . 10

.4 0.2 ell 0.2 < 0 < 0

• • • \

~ \ •

80 " 84 ,. 88 80 " 84 86 88 Temperature cae)

1 ' ~ 08

0.' • ~ 0.' ~ 0.2 • 0

80 " 84 86 " TemperatlXfI ("C)

Figure 13: Dissociation curves for California , Lord Howe, Oahu, and Molokai transects seperated into categories of distance from shore (0-5 , 5-10 , 10-100, 100-300, and> 300 nautical miles). Squares represent temperature peaks.

54

Table 3: Comparing onshore and offshore clone libraries community structure for Califomia,Aniwa, and Oahu. Number of OTU's were determined using DOTUR for both stations. Number of sequences, OTU's sharedfTotal OTU's, and Community Structure were calculated using SONS. Community Structure Similiarity is calculated based on overlapping OTUs, as well as the relative abundances among the shared OTUs between the communities.

55

OTU's shared! Community Structure Community Structure Locations Distance leyel # of sequences #ofOTU's TotalOTU's Similarity Similarity SE

CaIHomia 78% 88 7,7 0.43 0.52 0.13

Oahu 78% 91 7,7 0.43 0.75 0.12

Aniwa 80% 93 14,14 0.14 0.042 0.03

All 80% 272 22,22 0.41 0.52 0.083

Nearshore CAYs.OA 80% 91 19,19 0.13 0.0059 0.0049 CAys. AN 80% 93 19,19 0.11 0.059 0.038 ANYs.OA 80% 92 19,19 0.3 0.21 0.076

Offshore CAvs.OA 80% 90 9,9 0.25 0.3 0.095 CAys. AN 80% 88 9,9 0.17 0.99 0.0051 ANYs.OA 80% 84 9,9 0.29 0.32 0.01

56 References:

1. Andreae, M., and Crutzen, P. (1997) Atmospheric aerosols: biogeochemical

sources and role in atmospheric chemistry. Science 276: 1052-1058.

2. Azam, F., Fenchel, T., Field, J., Gray, J., Meyer-Rei!, L., and Thingstad, F. (1983)

The Ecological role of water-column microbes in the sea. Marine Ecology

Progress Series 10: 257-263.

3. Behrenfeld, M., O'Malleyl, R., Siegel, D., McClain. C., Sarmiento, J., Feldman,

G., Milligan, A, Falkowski, P., Letelier, R., and Boss, E. (2006) Climate­

driven trends in contemporary ocean productivity. Science 444: 752-755.

4. Beja, 0., Suzuki, M., Heidelberg, J., Nelson, W., Preston, C., Hamada, T., Eisen,

J., Fraser, C., and DeLong, E. (2002) Unsuspected diversity among marine

aerobic anoxygenic phototrophs. Nature 415: 630-633.

5. Bryant, D., and Frigaard, N. (2006) Prokaryotic photosynthesis and phototrophy

illuminated. Trends in Microbiology 14: 488-496.

6. Cho, J., Stapels, M., Monis, R., Vergin, K., Schwalbach, M., Givan, S., Barofsky,

D., and Giovannoni, S. (2007) Polyphyletic photosynthetic reaction centre

genes in oligotrophic marine Gammaproteobacteria. ETfVironmental

Microbiology 9: 1456-1463.

7. Church, M., Short, M., Jenkins, B., Karl, D., and Zehr, J. (2005) Temporal

patterns of nitrogenase (nifH) gene expression in the oligotrophic North

Pacific Ocean. Applied ETfVironmental Microbiology 71: 5362-5370.

8. -------- Ducldow, H., and Karl, D. (2004) Light Dependence of eH]Leucine

Incorporation in the Oligotrophic North Pacific Ocean. Applied and

Environmental Microbiology 70:4079-4087.

9. Cottrell, M., Mannino. A, and Kirchman, D. (2006) Aerobic anoxygenic

phototrophic bacteria in the Mid-Atlantic Bight and North Pacific Gyre.

Applied and Environmental Microbiology 72: 557-564.

57 10. Cox, E., Ribes, M., and Kinzie m, R. (2006) Temporal and spatial scaling of

planktonic responses to nutrient inputs into a subtropical embayment. Marine

Ecology Progress Series 324:19-35.

11. De Wit, R. and Bouvier, T. (2006) 'Everything is everywhere, but, the

environment selects'; wbat did Baas Becking and Beijerinck really say?

Environmental Microbiology 8: 755-758.

12. Denner, E., Vybiral, D., Koblizek, M., Kampfer, P., Busse, H., and VeIimirov, B.

(2002) Erythrobacter citreus sp. Nov., a yellow pigmented bacterium that

lacks bacteriochlorophyll a, isolated from the western Mediterranean Sea.

International Journal of Systematic and Evolutionary Microbiology 52:1655-

1661.

13. Didier, R., Audie, S., Robert, C., Abergel, C., Renesto, P., Ogata, H., La Scola,

B., Suzan, M., and Claverie, J. (2004) The 1.2-Megabase Genome Sequence

of Mimivirus. Science 306: 1344-1350.

14. Du, H., Jiao, N., Hu, Y., and Zeng, Y. (2006) Real-time PCR for quantification of

aerobic anoxygenic phototrophic bacteria based on ptifM gene in marine

environment. Journal of Experimental Marine Biology and Ecology 329: 113-

121.

15. Fuchs, B., Spring, S., Telling, Ii, Quast, c., Wulf, 1., Schattenhofer, M., Yau, S.,

Ferriera, S., Johnson, J., Glockner, F., and Amann, R. (2007) Characterization

of a marine gammaproteobacterium capable of aerobic anoxygenic

photosynthesis. Proceedings of the National Academy of Science 104: 2891-

2896.

16. Gich, P., and Overmanu, J. (2006) Sandarakinorhabdus limnophila gen. nov., sp.

nov., a novel bacteriochlorophyll a-containing, obligately aerobic bacterium

isolated from freshwater lakes. International Journal of Systematic and

Evolutionary Microbiology 56: 847-854.

17. Giovannoni, S., Tripp, J., Givau, S., Podar, M., Vergin, K., Baptista, D., Bibbs,

L., Eads, J., Eads,R., Toby, H., Noordewier, M., Rappe, M., Short, J.,

Carrington, J., and Mathur, E. (2005) Genome streamlining in a

cosmopolitan oceanic bacterium. Science 309: 1242-1245.

58

18. Giovannoni, S. and Sting, U. (2005) Molecular diversity and ecology of microbial

pIankton. Nature 437: 343-348.

19. Goericke, G. (2002) Bacteriochlorophyll a in the ocean: is anoxygneic bacterial

photosynthesis important? limnology and Oceanography 47: 290-295.

20. Guillard, R and Ryther, J. (1%2) Studies ofmarinepIanktonic diatoms. 1.

cyctotella nana hustedt, and Detonula confervacea (cleve) Gran. Canadian

Journal of Microbiology 8: 229-239.

21. Hern8ndez-L6pez, J., Vargas-Albores, F. 2003. A microplate technique to

quantify nutrients (N02-, N03-, NH4+ and P043-) in seawater. Aquaculture

34: 1201-1204.

22. Hold, G., Smith, E., Rappe, M., Maas, E., Moore, E., Stroempi, C., Stephen, J.,

Prosser, J., Birkbeck, T., and Gottschal, J. (2000) Characterization of bacterial

communities associated with toxic and non-toxic dinoflagellates. FEMS

Microbiology Ecology 37: 161-173.

23. Hu, Y., On, H., Jiao, N., and Zeng, Y. (2006) Abundant presence of the 'Y-Jike

Proteobacterialp/ifM gene in oxic seawater. FEMS Microbiology Letters 263:

200-206.

24. Jiao, N., Zhang, Y., Zeng, Y., Hong, N., Lin, R., Chen, F .• and Wang, P. (2007)

Distinct distribution pattern of abundance and diversity of aerobic anoxygenic

phototrophic bacteria in the global ocean. Environmental Microbiology 9:

3091-3099.

25. Karl, D. (2007) Microbial oceanography: paradigms, processes and promise.

Nature Reviews 5: 759-769.

26. ------. (2002) Hidden in a sea of microbes. Nature 415: 590-591.

27. Koblizek, M, Masin, M., Ras, J., Poulton, A., and Prasil, O. (2007) Rapid growth

rates of aerobic anoxygenic phototrophs in the ocean. Environmental

Microbiology 9: 2401-2406.

59 28. ------., Beja, 0., Bidigare, R., Christensen, S., Benitez-Nelson, B., Vetriani,

C., Kolber, M., Falkowski, P., and Kolber, Z. (2003) Isolation and

characterization of Erythrobacter sp. strains from the upper ocean. Archives of

Microbiology 180: 327-338.

29. Kolber, Z. 2007. Energy cycle in the ocean: powering the microbial world.

Oceanography 20:79-88.

30. ----- Plumley, F., Lang, A., Beatty, J., Blankenship, R., VanDover, C.,

Vetriani, C., Koblizek, M., Rathgeber, C., and Falkowski, P. (2001)

Contribution of aerobic photoheterotrophic bacteria to the carbon cycle in the

ocean. Science 292: 2492-2495.

31. ------ Van Dover, C., Niederman, R., and Falkowski, P. (2000) Bacterial

photosynthesis in surface waters of the open ocean. Nature 407: 177-179.

32. Labrenz .• M .• Lawson.. P .• Tindall. B .• Collins .• M .• and Hirsch.. P. (2005)

Roseisalinus antarcticus gen. nov., sp. nov .• a novel aerobic

bacteriochlorophyll a-producing a-proteobacterinm isolated from hypersaline

Ekho Lanke, Antarctica. International Journal of Systematic and Evolutionary

Microbiology 55:41-47.

33. Lami, R., Cottrell. M., Ras, J., Ulloa, 0., Obernosterer, I., Clanstre, H., Kirchman,

D .• and Lebaron, P. (2007) High abundances of aerobic anoxygenic

photosynthetic bacteria in the South Pacific ocean. Applied and

Environmental Microbiology 73: 4198-4205.

34. Longhurst, A. (1998) Ecological Geography of the Sea. Academic Press. San

Diego and London.

35. Ludwig, W., Strunk, 0., Westram, R., Richter, L., Meier, H., Yadhukumar,

Buchner, A., Lai. T., Steppi, S., Jobb, G., Forster, W., Brettske, L, Gerber, S.,

Ginhart, A., Gross, 0., Gmmann, S., Hermann, S., Jost, R., Konig, A., Liss,

T., LuBmann, R., May, M., Nonhoff, 8., Reichel, B., Strehlow, R.,

Stamatakis, A., Stuckmann, N., Vilbig, A., Lenke, M., Ludwig, T., Bode, A.,

and Schleifer, K. (2004) ARB: a software environment for sequence data.

Nucleic Acids Research 32: 1363-1371.

36. Marie.D .• Brusaard.C .• Thyrhaug.R.. Bratbak. G .• and Vanlot,D. (1999)

Enumeration of marine viruses in culture and natura1 samples by flow

cytometry. Applied and Environmental Microbiology 65: 45-52.

37. Masin, M .• Zdun. A.. Ston-Egiert. J., Nausch. M .• Labrenz. M .• Moulisova, V .•

60

and Koblizek. M. (2006) Seasonal changes and diversity of aerobic

anoxygenic phototrophs in the Baltic Sea. Aquatic Microbial Ecology 45: 247-

254.

38. Michelou, V .• Cottrell, M .• and Kirchman, D. (2007) Light-stimulated bacterial

production and amino acid assimiIation by cyanobacteria and other microbes

in the North Atlantic Ocean. Applied and Environmental Microbiology 73:

5539-5546.

39. Moore, L. and Chisholm, S. (1999) PhotophysioIogy of the marine

cyanobacterium Prochlorococcus:ecotypic differences among cultured

isolates. Limnology and Oceanography 44: 628-638.

40. Moran, X. (2007) Annual cycle of picophytoplankton photosynthesis and growth

rates in a temperate coastal ecosystem: a major contnbution to carbon fluxes.

Aquatic Microbial Ecology 49:267-279.

41. Nagashima, K.. Hiraishi, A.. Shimada, K.. and Matsuura, K. (1997) Horizontal

transfer of genes coding for the photosynthetic reaction centers of purple

bacteria. Journal of Molecular Evolution 45: 131-136.

42. O·Malley. M. (2007) The nineteenth century roots of 'everything is everywhere'.

Nature Reviews 5: 647-651.

43. Oz, A.. Sabehi, G .• Koblizek. M .• Massana, R .• and Beja, o. (2005) Roseobacter-

like bacteria in Red and Mediterranean Sea aerobic anoxygenic photosynthetic

populations. Applied and Environmental Microbiology 71: 344-353.

44. Partensky. P .• Hess. W .• and Vaulot, D. (1999) Prochlorococcus. a marine

photosynthetic prokaryote of global significance. Microbiology and Molecular

Biology Reviews 63: 106-127.

45. Palenik,B .• Reo. Q .• Dupont,C .• Myers.G .• Heidelberg) .• Badger. J .• Madupu,R.. Nelson, W .• Brinkac. L.. Dodson.R.. Durkin,A.. Daugherty.S .• Sullivan,S .•

61 Khouri,H., Mohamoud, Y., Halpin,R., and Paulsen,I. 2006. Genome sequence of Synechococcus CC9311: insights into adaptation to a coastal environment. Proceedings of the National Academy of Science. 103: 13555-13559.

46. Pedros-Alio, C. (2006) Marine microbial diversity: can it be determined?

TRENDS in Microbiology 14: 257-263.

47. Pinhassi, J., Sala, M., Havskum, H., Peters, F., Guadayol, 0., Malits, A., and

Marrase, C. (2004) Changes in bacterioplankton composition under different

phytoplankton regimens. Applied and Environmental Microbiology 70: 6753-

6766.

48. Pomeroy, L.R. (1974) The ocean's food web, a changing paradigm. Bioscience

24: 499-504.

49. Pommier, T., Canaback, B., Riemann, L., Bostrom, H., Simu, K., Lundberg,

P.,Tunlid, A., and Hagstrom, A. (2007) Global patterns of diversity and

community structure in marine bacterioplankton. Molecular Ecology 16: 867-

880.

50. Porra, R., Thompson, W., and Kriedemann, P. (1989) Determination of accurate

extinction coefficients and simultaneous equations for assayiog chlorophylls a

and b extracted with four different solvents: verification of the concentration

of chlorophyll standards by atomic absorption spectroscopy. Biochimica et

biophysica acta 975: 384-394.

51. Porter, K. and Feig, Y. (1980) The use ofDAPI for identifyiog and counting

aquatic microflora. Limnology and Oceanography 25: 943-948.

52. Qiang., L., Nianzhi., J., and Zaiqing., P. (2006) Environmental control growth and

BChl a expression in an aerobic anoxygenic phototrophic bacterium,

Erythrobacter longus (DSMZ6997). Acta Oceanologica Sinica 25: 138-144.

53. Rathgeber, C., Beatty, T., and Yurkov, V. (2004) Aerobic phototrophic bacteria:

new evidence for the diversity, ecological importance and applied potential of

this previously overlooked group. Photosynthesis Research 81: 113-128.

54. Schafer, H., Abbas, 8., Witte, H., and Muyzer, G. (2002) Genetic diversity of

"satellite" bacteria present in cultures of marine diatoms. FEMS Microbiology

Ecology 42: 25-35.

62 55. Schloss, P. and Handelsman, J. (2006) Introducing SONS, a tool for operational

taxonomic unit based comparisons of microbial community memberships and

structures. Applied and Environmental Microbiology 72: 6773-6779.

56. ------ and Handelsman, J. (2005) Introducing DOTUR, a computer program for

defining operational taxonomic units and estimating species richness. Applied

and Environmental Microbiology 71: 1501-1506.

57. Schwalbach, M., and Fuhrman, J. (2005) Wide-ranging abundances of aerobic

anoxygenic phototrophic bacteria in the world ocean revealed by

epifluorescence microscopy and quantitative PCR. Limnology and

Ocemwgraphy 50: 620-628.

58. Sieracki, M., Gilg, I., Thier, E., Poulton, N., and Goericke, R. (2006) Distribution

of planktonic aerobic anoxygenic photoheterotrophic bacteria in the northwest

Atlantic. Limnology and Oceanography 51: 38-46.

59. Swingley WD, Sadekar S, Mastrian SD, Matthies HJ, Hao J, Ramos H, Acharya

CR,Comad AL, Taylor HL, Dejesa LC, Shab MK, O'hnallachain ME, Lince

MT, Blankenship RE, Beatty IT, Touclunan JW. (2007) The complete

genome sequence of Roseobacter denitrificans reveals a mixotrophic rather

than photosynthetic metabolism Journal of Bacteriology. 189: 683-690.

60. Teira, E., Gasol, J., Aranguren-Gassis, M., Fernandez, A, Gonzalez, J.,

Lekunberri, I, and Alvarez-Saigado, X. (2008) I .inkages between

bacterioplankton community composition, heterotrophic carbon cycling and

environmental conditions in a highly dynamic coastal ecosystem

Environmental Microbiology. In press.

61. Waidner, L. and Kirchman, D. (2007) Aerobic anoxygenic phototrophic bacteria

attached to particles in turbid waters of the Delaware and Chesapeake

estuaries. Applied and Environmental Microbiology 73: 3936-3944.

62. We1schmeyer, N. (1994) Floorometrlc analysis of chlorophyll a in the presence of

chlorophyll b and pheopigments. Limnology and Oceanography 39: 1985-

1992.

63 63. Whitfield, J. (2005) Biogeography: is everything everywhere? Science 310:

%0-961.

64. Whittaker, RJ. Island biogeography: ecology, evolution, and conservation.

Oxford University Press. Oxford. 1998.

65. Yurkov, V. and Beatty, T. (1998a) Aerobic aerobic anoxygenic phototrophic

bacteria. Microbiology and Molecular Biology Reviews 62: 695-724.

66. ------- and-----. (1998b) Isolation of aerobic anoxygenic photosynthetic

bacteria from black smoker plume waters of the Juan de Fuca ridge in the

Pacific Ocean. Applied and Environmental Microbiology 64: 337-341.

67. Yutin, N., Suzuki, M., Teeling, H., Weber, M., Venter, J., Rusch, D., and Beja, O.

(2007) Assessing diversity and biogeography of aerohic anoxygenic

phototrophic bacteria in surface waters of the Atlantic and Pacific oceans

using the global ocean sampling expedition metagenomes. Environmental

Microbiology 9: 1464-1475.

68. ------., Suzuki, M, and Beja, O. (2005) Novel primers reveal wider diversity

among marine aerobic anoxygenic phototrophs. Applied and Environmental

Microbiology 71: 8958-8962.

69. ZemrneJink, H., Houghton, L., Sievert, S., Frew, N., and Dacey, J. (2005)

Gradients in dimethylsulfide,dimethylsulfoniopropionate, dimethylsulfoxide,

and bacteria near the sea surface. Marine Ecology Progress Series 295: 33-42.

70. Zhang, T., and Fang, H. (2005) 16S rDNA clone library screening of

environmental sample using melting curve analysis. Journal of the Chinese

Institute of Engineers 28: 1153-1155.

71. Zhang, Y. and Jiao, N. (2007) Dynamics of aerobic anoxygenic phototrophic

bacteria in the East China Sea. FEMS Microbiology Ecology 61: 459-469.

72. Zinser, E., Coe, A, Johnson, z., Martiny, A., Fuller, N., Scanlan, D., and

Chisholm, S. (2006) Prochlorococcus ecotype abundances in the North

Atlantic Ocean revealed by an improved quantitative PCR method. Applied

and Environmental Microbiology 72: 723-732.

73. Zwirglmaier, K., Jardillier, L., Ostrowski, M., Mazard, S., Garczarek, L.,

Vaulot, D., Not, F., Massan3, R., Ulloa, 0., and Scanlan, D. (2008) Global

phylogeography of marine Syneclwcoccus and Prochlorococcus reveals a

distinct partitioning of lineages among oceanic biomes. Environmental

Microbiology 10: 147-161.

74. ------, Heywood, J •• Chamberlain, K., Woodward, M., ZUbkov, M., and

Scanlan, D. (2007) Basin-scale distribution patterns of picocyanobacterial

lineages in the Atlantic Ocean. Envirorunental Microbiology 9: 1278-1290.

64