University of Southern California Dissertations and Theses

Marine bacterioplankton biogeography over short to medium spatio-temporal scales

(USC Thesis Other)

Marine bacterioplankton biogeography over short to medium spatio-temporal scales

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

MARINE BACTERIOPLANKTON BIOGEOGRAPHY OVER SHORT TO MEDIUM
SPATIO-TEMPORAL SCALES

by
Joshua Adam Steele
______________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOLOGY)
May 2010

Copyright 2010 Joshua Adam Steele

ii

Dedication

This work is dedicated to J.M.S. for all of her help, love, and support and for keeping
me sane and laughing during this journey. It is also dedicated to my family, L.S.S.,
J.H.S., and N.A.S., for their support, humor, and for keeping me grounded.

iii

Acknowledgements

This work could not have been done without the help, support and guidance of many
people. I cannot possibly list them all, and I am indebted to each and every person who
encouraged, inspired, and helped me along the way. I first must acknowledge my
advisor Jed Fuhrman, who provided guidance and motivation, while letting me find my
own way. I also must acknowledge my committee Doug Capone, Dale Kiefer, Joe
Devinny, and Wiebke Ziebis. They provided support, guidance, and helpful comments
and edits on this dissertation. J. Sayre also provided editing and useful comments. All
of the errors found within this dissertation are entirely my own. I also would like to
thank the Fuhrman Lab past and present. Especially M. Schwalbach, I. Hewson, C.
Chow, A. Patel, J. Cram, D. Needham, A. Parada, J. Griffith, R. Noble, J.M. Beman, M.
Brown, R. Sachdeva, M. Kakajiwala, S. Obrien, L. Gilbane, X. Liang, X. Hernandez, L.
Sargsyan, T.M. To, E. Royker, V. Work, K. McKissick, S. Thakkar, A. Guggenheim, C.
Teuscher. Thanks also to L. Duguay, P. Griffman, J. Finzi, J. Burns, T. Gunderson, A.
Close, G. Smith, R. Smith, T. Oudin, L. Garske, L. Czarnecki, K. Chvostal, P. Lopez, J.
Aguilar. I also wish to thank my cohort, J. Sohm, P.Yu, C. Purcell. Finally, I wish to
thank C. Viggiani, I. Cetinic, N. Steele, J. Sayre, & J. Cohen for putting up with me
during this process.

iv

Table of Contents

Dedication

Acknowledgements

List of Tables

List of Figures

Abstract

Chapter 1: Brief Introduction to Biogeography of Bacteria in the
Ocean: Community Variation and its Relationship to Habitat.
Introduction
Chapter 1 References

Chapter 2: Spatial Scales of Marine Bacterioplankton Assemblages in
the Southern California Bight
Chapter 2 Abstract
Introduction
Materials and Methods
Results
Discussion
Conclusions
Chapter 2 References

Chapter 3: Stability in Bacterial Assemblages Over Hour-Day
Temporal Scales
Chapter 3 Abstract
Introduction
Materials and Methods
Results
Discussion
Conclusions
Chapter 3 References

ii

iii

vi

vii

x

1
1
13

22
22
24
29
34
62
74
76

81
81
83
88
95
114
124
125

v

Chapter 4: Bacterial Community Composition in the Sea Surface
Microlayer and Surface Waters in the Southern California Bight.
Chapter 4 Abstract
Introduction
Results & Discussion
Experimental Procedures
Chapter 4 References

Chapter 5: Three-Domain Microbial-Environmental Networks from
Ocean Time Series Data
Chapter 5 Abstract
Introduction
Results & Discussion
Methods
Chapter 5 References

Chapter 6: Synthesis and Conclusions
Introduction
Summary of Results
Synthesis and Conclusions
Chapter 6 References

Alphabetized Bibliography

131
131
133
139
162
167

174
174
176
179
194
196

201
201
202
206
216

219

vi

List of Tables

Table 2-1 Locations and measured environmental and biotic parameters
for the 2004 study.

Table 2-2 Locations and measured environmental and biotic parameters
for the 2005 study.

Table 2-3 Distance-Decay regression model output and statistics.

Table 2-4 Rank Correlations between Bacterial Community Scores,
Geographic Distance, and Environmental Parameters

Table 2-5 DISTLM multivariate multiple regression model output and
statistics.

Table 3-1 Sampling time, location, depth, and environmental
parameters from the August 2000 study.

Table 3-2 Biotic parameters from August 2000 study.

Table 3-3 Sampling time, location, depth, and environmental
parameters from May 2004.

Table 3-4 Biotic parameters from May 2004.

Table 3-5 Sample date, environmental, and biotic parameters for
Catalina sites.

Table 3-6 Mantel-type Spearman rank correlations (r
s
) of community
similarity values with time and with environmental similarity.

Table 4-1 Site number, sampling type, date, location, and measured
environmental parameters

Table 4-2 Location, depth and biological parameters measured at the
sampling sites.

Table 4-3 Results from PERMANOVA test.

Table 4-4 Correlations between environmental variables, similarity
indices, and biological parameters.

35

39

57

58

61

96

96

97

99

100

104

142

143

150

158
vii

List of Figures
Figure 2-1. Map of the Southern California Bight showing the average
sea surface temperature for May 2004.

Figure 2-2 Map of the Southern California Bight showing the average
sea surface temperature for May 2005.

Figure 2-3 Maps of interpolated environmental parameters measured at
the 18.5°C isotherm in May 2004.

Figure 2-4 Maps of interpolated environmental parameters measured
during the 2005 study.

Figure 2-5 Dendrograms showing whole community similarity from
nested grid studies using the Bray-Curtis Index

Figure 2-6 Dendrograms showing the whole community similarity of
nested grid studies using the Sorenson Index.

Figure 2-7 The 20 most abundant OTUs identified from ARISA peaks
in the 2004 18.5°C isotherm sites

Figure 2-8 The 20 most abundant OTUs identified from ARISA peaks
from the 2005 12.1°C isotherm sites.

Figure 2-9 Dendrograms showing the community similarity at 5m in
the 2005 study.

Figure 2-10 Map of the bacterial assemblage from 5m depth during the
2005 study

Figure 2-11 The 20 most abundant OTUs from ARISA fingerprinting at
5m from the 2005 study.

Figure 2-12 Distance in km vs community similarity for the 2004 and
2005 studies.

Figure 3-1 Map of sampling locations for the drogue and daily studies.

Figure 3-2 Bacterial and Viral Abundance during the 2000 study.

27

28

36

37

41

42

45

46

49

50

52

55

89

98

viii

Figure 3-3 Dendrograms of bacterial assemblage similarity at the 19 ºC
isotherm depth during the 2000 drogue study.

Figure 3-4 Graphs of bacterioplankton OTUs during the 2000 drogue
study and at 5m depth in the 5 months following.

Figure 3-5 Dendrograms of bacterial assemblage similarity and
abundance curves of the 6 most abundant OTUs during the 2004
drogue study at the 5m depth and the 18.5°C isotherm depth.

Figure 3-6 Dendrograms of bacterial assemblage similarity and
abundance curves of the 6 most abundant OTUs during the 2004
drogue study at the Chlorophyll maximum depth.

Figure 3-7 Rare (<5%) bacterioplankton abundance curves.

Figure 3-8 Daily bacterial assemblage similarities over one week,
separated by one year in June 2002 and June 2003.

Figure 3-9 Daily rare bacterioplankton ARISA abundance curves over a
week in June 2002 and 2003

Figure 4-1 Map of sampling locations in the Southern California Bight.

Figure 4-2 Photographs of the rotating drum sampler.

Figure 4-3 Total viruses and bacteria, bacterial production and richness
in the SML and the corresponding ULW, along with the
SML:ULW enrichment factors.

Figure 4-4 Dendrograms showing whole community similarities for
free living bacteria from the SML and ULW in 2004.

Figure 4-5 Dendrograms showing whole community similarities
between SML and ULW communities in May 2005 and October
2006.

Figure 4-6 ARISA-derived rank-abundance of free-living bacterial
OTUs from 2004.

Figure 4-7 ARISA-derived rank-abundance of free living bacterial
OTUs from 2005 and 2006.

102

103

106

107

109

111

112

137

138

140

148

151

154

155

ix

Figure 5-1 Subnetwork showing SAR11 OTUs as the central nodes and
their “nearest neighbors.”

Figure 5-2 Subnetwork with stramenopiles as the central nodes.

Figure 5-3 Alveolate Subnetworks.

Figure 5-4 Circular network showing all significant correlations
(p≤0.01, q≤0.063) with nodes sorted by the number of
correlations.

Figure 5-5 Subnetwork with γ-proteobacteria as central nodes.

Figure 5-6 Subnetwork with ciliates and flagellates as central nodes.

Figure 5-7 Subnetwork with cyanobacteria as central nodes.

180

181

183

186

187

190

191

x

Abstract

Microbes are central to ocean food webs and global biogeochemical processes.
Understanding their distribution and interactions over space and time is fundamental to
understanding their ecology. Even with the advent of molecular techniques, it is
difficult to observe their interactions with each other and with their environment. We
are in the early stages of expanding our focus from simply identifying microbes, to
discovering what they are doing (i.e. how each microorganism interacts and fits within
the functioning of the ecosystem). In the ocean, which is in constant lateral and vertical
motion, these community patterns and interactions may change on short to medium
spatial and temporal scales. The aim of this study was to determine the community
patterns in bacterioplankton in the ocean surface waters on hour-day temporal scales
and kilometer spatial scales, patterns in bacterial communities on µm to m depth scales,
and interactions among bacteria, protists and Archaea on monthly scales. Molecular
fingerprinting techniques coupled with clone libraries for identification were used to
study changes in bacterioplankton communities in the surface waters from 2000-2005,
in the sea surface microlayer and surface layer from 2004-2006, and interactions
between bacterioplankton and protists from 2000-2004 in the ocean near Southern
California. Investigating bacterioplankton community patterns over km spatial scales
and hour-day temporal scales, we found coherent communities at 2-97km
2
area scales
and at distances along transects of 0-15km, and the major taxa appeared remarkably
consistent throughout, while at distances from 50-255km, there were strong differences
xi

in the communities which correlated to environmental parameters. Temporally, we
found stable communities within a water patch and in the surface mixed layer over 20-
30 hours, and also high community similarity between days over a week at a single
geographic location. Together, these results suggest that in the Southern California
Bight, a typical 10L oceanographic sample can represent a bacterioplankton community
on the scale of 10-100km
2
and 4-6 days. Investigating the communities in the sea
surface microlayer compared to the communities at 0.5m depth, our results showed a
sea surface microlayer community which was distinct from the bacterial community at
0.5m depth in some locations, but showed no difference between the communities at
others. We found low and variable enrichment and depletion of bacteria and viruses
and strong depletion of bacterial production in the microlayer. The wide range of
similarities found between the sea surface microlayer and the underlying water
communities and the high number of shared taxa between the communities, suggests a
patchy microlayer community that is likely formed by bacteria being transported from
the surface waters to the microlayer. Using a network to visualize the statistical
correlations between co-occurrence of bacteria, archaea, and protists, along with viruses
and other environmental factors in a microbial community over 3 years suggested a
succession of microbial communities and identified possible local ecotypes of groups
like the SAR11 cluster, Stramenopiles, Alveolates, Cyanobacteria, and
microphagotrophs (e.g. ciliates and choanoflagellates). This new approach provides a
feel for the natural history of microbes, and should facilitate the inclusion of complex
microbial assemblages in community ecology studies; it also points to biogeochemical
xii

roles of “unknown” microorganisms, and indicates possible keystone species. These
studies provide a solid addition to a more complete understanding of bacterioplankton
biogeography and their ecological roles; and will provide insight into the function and
response of the bacterioplankton ecosystem in the surface waters.

1

Chapter 1: Brief Introduction to Biogeography of Bacteria in the Ocean:
Community Variation and Its Relationship to Habitat.

Introduction

We live now in the "Age of Bacteria." Our planet has always been in the "Age of
Bacteria."
Stephen Jay Gould
Since it was recognized roughly 30 years ago, the dominance of bacteria in
productivity, abundance, and biomass in the ocean has become increasingly clear over
time (Fuhrman & Azam 1982, Azam et al. 1983, Cho & Azam 1988, Azam 1998,
Ducklow 2000, Kirchman 2000). Not long after this revolution in marine microbial
ecology, was a molecular revolution when genes sequenced from the environment
began to be used to characterize microbial communities (Olsen et al. 1986, Pace et al.
1986, Giovannoni et al. 1995, Diez et al. 2001). This led to a new appreciation of
microbial diversity across the planet, and it is now clear that microbes are ubiquitous
and involved in every ecosystem process imaginable (Whitman et al. 1998).

Considering just bacterial taxonomic richness in the marine environment, estimates
have converged on millions of distinct bacterial “species” (Curtis et al. 2002); the
further marine ecosystems are examined, the more species are discovered. Studies
using cloning and sequencing have found as many as 1600 different ribotypes within an
2

estuary (Acinas et al. 2004), and metagenomic shotgun sequencing reported around
1800 taxa from one environment (Venter et al. 2004). It should be noted that whole-
genome metagenomics may underestimate diversity due to the low number of
taxonomically informative sequences relative to total sequence in these samples.
Recent investigations into the rare biosphere using high throughput tag sequencing
(Sogin et al. 2006, Huber et al. 2007) show that we still do not know the full extent of
the microbial diversity.

Biogeography (the study of community diversity patterns over space and time and the
environmental factors which drive them) for microbes would ideally start with
straightforward observation of the relationships among species as well as between
species and their environment, as the earliest ecology had done with animals and plants.
Within an observationally based system, the functional role of many of these organisms
would be obvious. With microbes, however, direct observation of species has been
impossible. Microbial biogeography has been revolutionized by development of
molecular genetic techniques which allow for the detection of identity and sometimes
suggest the functional roles of microbes in their environment. But direct or even indirect
characterization of the function of particular naturally occurring microbes in their
habitat has been especially challenging, as has investigation of microbial interactions.

One way to examine function and interactions is to correlate the distributions of
microbial communities with ecological parameters and infer functions from these
3

relationships. Microbial ecologists who study soils have shown changes over space and
time that are related to distance and environment (e.g. relationships between genetic
dissimilarities and geographic distances (Cho & Tiedje 2000) or a taxa–area
relationship (see below, and reviewed by (Green & Bohannan 2006)), and the patterns
emerging are similar to what is found for eukaryotic organisms (Hughes-Martiny et al.
2006, Prosser et al. 2007, Ramette & Tiedje 2007a, b). Relatively few broad scale
ecological patterns have been looked at in marine bacterioplankton, although the
decrease of diversity with latitude has been shown to be detectable in marine
bacterioplankton (Pommier et al. 2007, Fuhrman et al. 2008) unlike soil bacteria where
no such pattern was detected (Fierer & Jackson 2006). There are a number of recent
studies that examine seasonal patterns in marine bacterioplankton (see below).

Oceanographic studies rely on a few representative samples from a continuously
moving environment. This adds an extra layer of difficulty to separating spatial features
from temporal features, and confounds the interpretation of community change over
time - perhaps accounting for marine microbial ecology’s lag behind soil microbial
ecology in terms of describing biodiversity. The first steps are to simply look for
patterns in distributions in space and time, see the appropriate scales of variability, and
learn what a particular sample might represent (Fuhrman & Steele 2008, Fuhrman
2009).

4

Spatial and Temporal Scales: The adequacy of typical oceanographic sampling to
describe a microbial habitat (i.e. an environment that contains a coherent group of
microbes) and the geographic and temporal scale of that habitat is unclear.
Investigations into spatial variation of marine bacterioplankton community composition
have tended to focus on the millimeter to meter scales (Schauer et al. 2000, Kirchman et
al. 2001, Long & Azam 2001, Casamayor et al. 2002, Seymour et al. 2004) or hundreds
to thousands of kilometers (Hewson et al. 2006a, b, Pommier et al. 2007, Fuhrman et al.
2008). Typically, investigations into temporal changes in bacterioplankton community
composition are on the order of months to years, e.g. in the Antarctic Ocean (Murray et
al. 1998), the Santa Barbara Channel (Murray et al. 1999), the Aegean and Adriatic
Seas (Moeseneder et al. 1999), the Southern California Bight (Fuhrman et al. 2006) the
Sargasso Sea (Morris et al. 2005, Carlson et al. 2009, Treusch et al. 2009), the Mid
Atlantic Bight (Nelson et al. 2008), and the Chesapeake Bay (Campbell et al. 2009).

Only a few studies have examined the diversity patterns on daily to weekly scales. Lee
& Fuhrman (1991) found high similarity between communities collected one week
apart from Long Island Sound, New York, and low similarity seven months apart. They
also found low similarity between communities from the same location collected two
weeks apart were only about 40% similar (Lee & Fuhrman 1991). The authors also
reported that, of open ocean samples at one geographic location (500km west of San
Diego) over three consecutive days, were highly similar the first two days and changed
significantly by the third day. Acinas et al. (1997) found little variation among
5

assemblages in a water patch of a few kilometers over 48 hours in the Mediterranean
Sea. Hewson et al (2006b) reported low rates of change from 24-36h following a water
mass in the Gulf of Mexico, the North Tropical Pacific and the North tropical Atlantic
and across a few km in the North Tropical Pacific. The authors suggested that a
coherent bacterial community could be found at this scale (Hewson et al. 2006b).

While bacteria vary on scales of millimeters to centimeters and months-years, the
patterns of bacterioplankton communities described above suggest structure on a larger
scale in the ocean. We hypothesize that the relevant size of a habitat for
bacterioplankton is a cohesive water mass or water patch near 10km diameter, the size
of mesoscale eddies (Hewson et al. 2006b, Fuhrman & Hagström 2008). The obvious
timeframe to look for is near the doubling time of bacteria (around one day – one
week); and, since bacterioplankton show little variation in abundance at a given depth,
this implies growth and death rates are well-balanced in terms of top down (i.e.
predation) and bottom up (i.e. resource availability) processes (Fuhrman 1992,
Thingstad & Lignell 1997, Ducklow 2000, Simek et al. 2001, Torsvik et al. 2002). It is
still an open question which environmental factors control the composition of the
bacterioplankton at the km or hour-day scale.

In addition to determining the limits of variation of communities in space and time,
spatial studies allow for observation of the relationship of species (or taxa) to area
which has not yet been reported for marine bacterioplankton (Fuhrman 2009). The
6

species- area relationship, first described formally by (Arrhenius 1921), is an ecological
pattern that can be extrapolated beyond the area studied and compared across species
(Rosenzweig 1995, Green & Bohannan 2006, Fierer 2008). This relationship between
the number of observed species (S) and the sampled area (A), is generally assumed to
be best represented by a power-law relationship, S = cA
z
. In this relationship, the
location specific constant (c) and the rate of change of species per unit area (z) are
determined empirically. Z values are typically 0.1–0.3 for animals and plants within
contiguous habitats and 0.25–0.35 for islands (Rosenzweig 1995).

Studies of microbes including marine planktonic diatoms and saltmarsh bacteria, have
found a range of z values, often lower than 0.1, although some recent studies have
reported z values that overlap the canonical ranges (0.1-0.3), especially for functional
“islands” such as tree holes and bioreactors (Horner-Devine et al. 2004b, Bell et al.
2005, van der Gast et al. 2005, Green & Bohannan 2006, van der Gast et al. 2006). The
z-value can be lowered by the ease at which the organisms disperse, the ecological
redundancy of the organisms (Horner-Devine et al. 2004a, Prosser et al. 2007), the
phylogenetic level at which the organisms are compared (Horner-Devine et al. 2004b),
the sampling technique (Green & Bohannan 2006, Woodcock et al. 2006), the
environmental heterogeneity (Horner-Devine et al. 2004a, Hughes-Martiny et al. 2006,
Ramette & Tiedje 2007b), and, in some cases, the scale at which the taxa-area
relationship is studied (Crawley & Harral 2001, Horner-Devine et al. 2004b). Thus, the
z values reported may underestimate the relationship. However, because z-values are
7

greater than 0 (indicating that there is spatial variation) and, in some cases, overlap with
the canonical values, the spatial variability in microbes may follow the same universal
rules that have been found for larger organisms.

Small Vertical Scales: The sea surface microlayer is operationally defined as the top
millimeter of the ocean (Hardy 1982, Liss & Duce 1997). Because of its location at the
air-sea interface, the sea surface microlayer is a fundamental part of regulating
biogeochemical and geochemical processes between the ocean and the atmosphere (Liss
& Duce 1997). 1mm is not a small scale to a bacterium, which may still have thousands
of individuals in a single microliter. In terms of ocean processes, the distinction
between the sea surface microlayer and the sea surface layer (1m) is much smaller than
the water patches discussed above, however, since it is often considered a distinct depth,
there is every reason to think it may display lateral community variation on a scale
similar to the water below it (i.e. the km-day scale discussed above) or because of the
influence of wave and wind energy on the ocean surface, it may display an entirely
different pattern.

Irrespective of lateral changes in the microlayer, there are strong differences between
the sea surface microlayer and the surface layer. It has been shown to be enriched in
organic compounds, carbohydrates, amino acids and humic substances (Williams et al.
1986, Gasparovic et al. 1998, Zhang et al. 1998, Yang et al. 2001, Upstill-Goddard et al.
2003, Franklin et al. 2005, Yang et al. 2005, Wurl & Holmes 2008). Due to its location
8

at the air-sea interface, it also receives intense UV radiation and tends to be enriched in
pollutants and heavy metals (Hardy et al. 1985, Hardy et al. 1987). Studies have found
virtually no enrichment in bacteria or bacterial production to high (200 fold) enrichment
(Sieburth 1971, Carty & Colwell 1975, Sieburth et al. 1976, Carlucci et al. 1985,
Carlucci et al. 1991, Agogue et al. 2004, Obernosterer et al. 2005) and depending on the
location, have shown completely different bacteria when compared with the surface
waters (Franklin et al. 2005, Cunliffe et al. 2008, Cunliffe et al. 2009), and nearly
identical communities (Agogue et al. 2005, Joux et al. 2006, Obernosterer et al. 2008).
It is still an open question if bacteria in the sea surface microlayer (SML) are a unique
community thriving at the air-water interface, or a collection of microbes brought up
from the ocean (Bezdek & Carlucci 1972) or deposited from the atmosphere (Norkrans
1980, Liss & Duce 1997).

Comparing the community similarity patterns between the surface microlayer bacteria
and the bacterioplankton in the surface waters will not only provide a sense of how
bacterioplankton distinguish their habitat in the surface of the ocean but it will also
provide insight into the scales of community variation at the immediate surface of the
ocean. Investigating scales of variation in bacterioplankton communities over
kilometers and hours-days will not only provide a sense of what a typical 10L sample
represents, but will also begin to describe the bacterial definition of habitat within a
9

slightly deeper mixed layer depth and within a single water mass. Correlating the
community patterns with environmental patterns should at least partially reveal the
factors driving the spatial and temporal variation.

Microbial Interactions and Networks:Comparing community similarities with
environmental similarities is powerful, but necessarily limited in scope. These patterns
provide insight into the relationship between the bacterial communities and their
environment, but it is nearly impossible, except in the smallest communities to
distinguish the ecological roles at a more specific level. In order to more closely
examine these roles we must look to more specific statistical relationships.

Pairwise correlations among the presence and abundance of bacteria, protists and
Archaea, and between these microbes and other parameters over time, allows for
differentiation of the organisms in terms of their preferred conditions. It also allows for
assignment of microbes into groups that tend to co-occur and those that do not tend to
occur together. Using the data from the USC Microbial Observatory which includes
bacterioplankton, protists, Archaea, virus and bacterial abundance, and a suite of
environmental variables (described in Brown et al. 2005, Fuhrman et al. 2006, Vigil et
al. 2009) from the chlorophyll maximum depth (a biologically defined feature) we can
characterize patterns of particular taxa.

10

Ruan et al. (2006) developed a Local Similarity Analysis (LSA) to evaluate such time-
lagged relationships, and showed, using examples from the San Pedro Ocean Time-
Series, how LSA can detect significant relationships that would be missed if time lags
(i.e. when one organism tends to follow another, or tends to decline after another one
increases) are ignored. These relationships are best viewed as a co-occurrence network,
rather than a massive correlation table, that indicates the positive and negative
mathematical relationships among microorganisms and between microbes and
environmental parameters (Ruan et al. 2006). In a sense, this is a diagram of the niche
space of the various organisms, and we believe that it provides a very powerful tool for
the examination of the ‘natural history’ of microbes in their complex wild habitats
(Fuhrman & Steele 2008).

Networks are not new to ecology, and there is considerable literature describing
ecological interaction networks (particularly food webs as reviewed in (Bascompte
2009, Ings et al. 2009). The microbial loop itself was a vital addition to the broad
interaction network describing the energy flow within the planktonic environment
(Azam et al. 1983). Technical and theoretical advances have made networks a
particularly interesting tool to describe and compare not only complete food webs
(Banasek-Richter et al. 2009), but also systems as far ranging as the connection of pages
on the internet (Barabasi & Albert 1999), the spread of emergent diseases through a
population (Pastor-Satorras & Vespignani 2001, Aparicio & Pascual 2007), the
interaction of molecules in a cell (Barabasi & Oltvai 2004) and reviewed in (Proulx et
11

al. 2005), mutualistic (i.e. pollen-pollenator) networks (Olesen et al. 2007), and the
response of populations to global change (Ledger et al. 2008). Looking at the
interaction networks of the microbes will allow us to determine the robustness or
fragility of the community (Albert et al. 2000, Montoya et al. 2006, Banasek-Richter et
al. 2009) and may identify relationships for further study or provide insight into how
future changes may affect the relationships present.

Techniques for measuring scales of variation: Many of the studies described in this
introduction used the tool of molecular fingerprinting. This has a benefit of describing
in one “snapshot” a large proportion of the community (Fuhrman 2009). In this study
we used Automated Ribosomal Intergenic Spacer Analysis (ARISA) to determine the
bacterioplankton community (Fisher & Triplett 1999) for the following reasons:
ARISA can distinguish taxa (actually operational taxonomic units, or OTUs) with 16S
rRNA sequence similarity of about 98% or less, similar to what many consider near the
“species” level for Bacteria (Brown & Fuhrman 2005). By standardizing the amount of
DNA between samples, the version of the method presented in Brown & Fuhrman
(2005) is generally quantitative, especially when comparing how the amount of a
particular OTU changes between different samples (as opposed to comparing the
amount of one OTU vs. another). Brown et al (2005) compared flow cytometric and
ARISA-based estimates of the monthly abundance of Prochlorococcus at the SPOT
station over three years. The linear regression of these variables had an r
2
value of 0.86,
indicating that the ARISA abundance correctly predicted Prochlorococcus abundance
12

(Brown et al. 2005). This method is not perfect and might miss certain taxa that do not
have linked 16S and 23S rRNA genes (e.g. most Planctomycetes and close relatives).
However it provides a reasonably comprehensive picture of the bacterial community in
most marine samples (e.g. Fuhrman et al. 2006).

The studies presented in this dissertation make a modest attempt to add to the
understanding of the spatial, temporal and environmental distribution patterns of marine
bacterioplankton. They do so by answering the following: 1) What is the scale of
variation in the bacterioplankton community across kilometer spatial scales and how do
bacterial community spatial patterns relate to environmental gradients? 2) What is the
scale of bacterioplankton community variation over hours-days within a water patch
and at a single geographic location and how do bacterial community temporal patterns
relate to environmental gradients? 3) Is the free-living bacterial community in the sea
surface microlayer distinct from the community in the 0.5m depth water and how do
these communities relate over space and time? 4) Can correlations among individual
bacterial and eukaryotic taxa and environmental factors be used to infer ecological roles
across three years at the chlorophyll maximum depth? These studies provide insight
into the ecology of the bacterioplankton as well as the microbiological scales relevant to
oceanographic sampling.

13

Chapter 1 References

Acinas SG, Klepac-Ceraj V, Hunt DE, Pharino C, Ceraj I, Distel DL, Polz MF (2004)
Fine-scale phylogenetic architecture of a complex bacterial community. Nature
430:551-554

Acinas SG, Rodriguez-Valera F, Pedros-Alio C (1997) Spatial and temporal variation in
marine bacterioplankton diversity as shown by RFLP fingerprinting of PCR
amplified 16S rDNA. Fems Microbiol Ecol 24:27-40

Agogue H, Casamayor EO, Bourrain M, Obernosterer I, Joux F, Herndl GJ, Lebaron P
(2005) A survey on bacteria inhabiting the sea surface microlayer of coastal
ecosystems. Fems Microbiol Ecol 54:269-280

Agogue H, Casamayor EO, Joux F, Obernosterer I, Dupuy C, Lantoine F, Catala P,
Weinbauer MG, Reinthaler T, Herndl GJ, Lebaron P (2004) Comparison of
samplers for the biological characterization of the sea surface microlayer.
Limnol Oceanogr-Meth 2:213-225

Albert R, Jeong H, Barabasi AL (2000) Error and attack tolerance of complex networks.
Nature 406:378-382

Aparicio JP, Pascual M (2007) Building epidemiological models from R-0: an implicit
treatment of transmission in networks. P R Soc B 274:505-512

Arrhenius O (1921) Species and Area. Journal of Ecology 9:95-99

Azam F (1998) Microbial control of oceanic carbon flux: The plot thickens. Science
280:694

Azam F, Fenchel T, Field JG, Gray JS, Meyerreil LA, Thingstad F (1983) The
Ecological Role of Water-Column Microbes in the Sea. Mar Ecol-Prog Ser
10:257-263

Banasek-Richter C, Bersier LF, Cattin MF, Baltensperger R, Gabriel JP, Merz Y,
Ulanowicz RE, Tavares AF, Williams DD, De Ruiter PC, Winemiller KO,
Naisbit RE (2009) Complexity in quantitative food webs. Ecology 90:1470-1477

Barabasi AL, Albert R (1999) Emergence of scaling in random networks. Science
286:509-512

Barabasi AL, Oltvai ZN (2004) Network biology: Understanding the cell's functional
organization. Nature Reviews Genetics 5:101-U115

14

Bascompte J (2009) Disentangling the Web of Life. Science 325:416-419

Bell T, Ager D, Song JI, Newman JA, Thompson IP, Lilley AK, van der Gast CJ (2005)
Larger islands house more bacterial taxa. Science 308:1884

Bezdek HF, Carlucci AF (1972) Surface Concentration of Marine Bacteria. Limnol
Oceanogr 17:566-&

Brown MV, Fuhrman JA (2005) Marine bacterial microdiversity as revealed by internal
transcribed spacer analysis. Aquat Microb Ecol 41:15-23

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show
marine microbial diversity: development and application to a time series.
Environ Microbiol 7:1466-1479

Campbell BJ, Yu L, Straza TRA, Kirchman DL (2009) Temporal changes in bacterial
rRNA and rRNA genes in Delaware (USA) coastal waters. Aquat Microb Ecol
57:123-135

Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ, Vergin K (2009)
Seasonal dynamics of SAR11 populations in the euphotic and mesopelagic
zones of the northwestern Sargasso Sea. Isme J 3:283-295

Carlucci AF, Craven DB, Henrichs SM (1985) Surface-Film Microheterotrophs -
Amino-Acid Metabolism and Solar-Radiation Effects on Their Activities.
Marine Biology 85:13-22

Carlucci AF, Craven DB, Wolgast DM (1991) Microbial-Populations in Surface-Films
and Subsurface Waters - Amino-Acid-Metabolism and Growth. Marine Biology
108:329-339

Carty C, Colwell RR (1975) A microbiological study of air and surface water
microlayers in the open ocean. J Wash Acad Sci 65:148–152

Casamayor EO, Pedros-Alio C, Muyzer G, Amann R (2002) Microheterogeneity in 16S
ribosomal DNA-defined bacterial populations from a stratified planktonic
environment is related to temporal changes and to ecological adaptations. Appl
Environ Microb 68:1706-1714

Cho BC, Azam F (1988) Major role of bacteria in biochemical fluxes in the ocean's
interior. Nature 332:441-443

15

Cho JC, Tiedje JM (2000) Biogeography and degree of endemicity of fluorescent
Pseudomonas strains in soil. Applied and environmental microbiology 66:5448-
5456

Crawley MJ, Harral JE (2001) Scale dependence in plant biodiversity. Science 291:864-
868

Cunliffe M, Schafer H, Harrison E, Cleave S, Upstill-Goddard R, Murrell JC (2008)
Phylogenetic and functional gene analysis of the bacterial and archaeal
communities associated with the surface microlayer of an estuary. Isme Journal
2:776-789

Cunliffe M, Whiteley AS, Newbold L, Oliver A, Schafer H, Murrell JC (2009)
Comparison of Bacterioneuston and Bacterioplankton Dynamics during a
Phytoplankton Bloom in a Fjord Mesocosm. Appl Environ Microb 75:7173-
7181

Curtis TP, Sloan WT, Scannell JW (2002) Estimating prokaryotic diversity and its
limits. Proc Natl Acad Sci U S A 99:10494-10499

Diez B, Pedros-Alio C, Massana R (2001) Study of genetic diversity of eukaryotic
picoplankton in different oceanic regions by small-subunit rRNA gene cloning
and sequencing. Appl Environ Microb 67:2932-2941

Ducklow HW (2000) Bacterial production and biomass in the oceans. In: Kirchman DL
(ed) Microbial ecology of the oceans. Wiley-Liss, New York, p 85-120

Fierer N (2008) Microbial biogeography: patterns in microbial diversity across space
and time. In: Zengler K (ed) Accessing Uncultivated Microorganisms: from the
Environment to Organisms and Genomes and Back. ASM Press, Washington
DC, p 95-115

Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial
communities. P Natl Acad Sci USA 103:626-631

Fisher MM, Triplett EW (1999) Automated approach for ribosomal intergenic spacer
analysis of microbial diversity and its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636

Franklin MP, McDonald IR, Bourne DG, Owens NJP, Upstill-Goddard RC, Murrell JC
(2005) Bacterial diversity in the bacterioneuston (sea surface microlayer): the
bacterioneuston through the looking glass. Environ Microbiol 7:723-736

16

Fuhrman JA (1992) Bacterioplankton roles in cycling of organic matter: the microbial
food web. In: Falkowski PG, Woodhead AD (eds) Primary productivity and
biogeochemical cycles in the sea. Plenum Press, New York, p 361-383

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Azam F (1982) Thymidine incorporation as a measure of heterotrophic
bacterioplankton production in marine surface waters: Evaluation and field
results. Marine Biology 66:109

Fuhrman JA, Hagström Å (2008) Bacterial and archaeal community structure and its
patterns. In: Kirchman DL (ed) Microbial Ecology of the Oceans. Wiley

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocean
conditions. P Natl Acad Sci USA 103:13104-13109

Fuhrman JA, Steele JA (2008) Community structure of marine bacterioplankton:
patterns, networks, and relationships to function. Aquat Microb Ecol 53:69-81

Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, Brown JH
(2008) A latitudinal diversity gradient in planktonic marine bacteria. Proc Natl
Acad Sci U S A 105:7774-7778

Gasparovic B, Kozarac Z, Saliot A, Cosovic B, Mobius D (1998) Physicochemical
characterization of natural and ex-situ reconstructed sea-surface microlayers. J
Colloid Interf Sci 208:191-202

Giovannoni SJ, Mullins T, Field KG (1995) Microbial diversity in marine systems:
rRNA approaches to the study of unculturable microbes. In: Joint I (ed)
Molecular Ecology of Aquatic Microbes. Springer-Verlag, Berlin-Heidelberg-
New York-Tokyo

Green J, Bohannan BJ (2006) Spatial scaling of microbial biodiversity. Trends Ecol
Evol 21:501-507

Hardy JT (1982) The Sea-Surface Microlayer - Biology, Chemistry and Anthropogenic
Enrichment. Prog Oceanogr 11:307-328

Hardy JT, Apts CW, Crecelius EA, Fellingham GW (1985) The Sea-Surface Microlayer
- Fate and Residence Times of Atmospheric Metals. Limnol Oceanogr 30:93-
101
17

Hardy JT, Crecelius EA, Antrim LD, Broadhurst VL, Apts CW, Gurtisen JM, Fortman
TJ (1987) The Sea-Surface Microlayer of Puget Sound .2. Concentrations of
Contaminants and Relation to Toxicity. Mar Environ Res 23:251-271

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006a) Remarkable heterogeneity in
meso- and bathypelagic bacterioplankton assemblage composition. Limnol
Oceanogr 51:1274-1283

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006b) Temporal and spatial scales of
variation in bacterioplankton assemblages of oligotrophic surface waters. Mar
Ecol-Prog Ser 311:67-77

Horner-Devine MC, Carney KM, Bohannan BJM (2004a) An ecological perspective on
bacterial biodiversity. P Roy Soc Lond B Bio 271:113-122

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJ (2004b) A taxa-area
relationship for bacteria. Nature 432:750-753

Huber JA, Mark Welch DB, Morrison HG, Huse SM, Neal PR, Butterfield DA, Sogin
ML (2007) Microbial population structures in the deep marine biosphere.
Science 318:97-100

Hughes-Martiny JB, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL,
Horner-Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S,
Ovreas L, Reysenbach A-L, Smith VH, Staley JT (2006) Microbial
biogeography: putting microorganisms on the map. Nat Rev Micro 4:102-112

Ings TC, Montoya JM, Bascompte J, Bluthgen N, Brown L, Dormann CF, Edwards F,
Figueroa D, Jacob U, Jones JI, Lauridsen RB, Ledger ME, Lewis HM, Olesen
JM, van Veen FJF, Warren PH, Woodward G (2009) Ecological networks -
beyond food webs. Journal of Animal Ecology 78:253-269

Joux F, Agogue H, Obernosterer I, Dupuy C, Reinthaler T, Herndl GJ, Lebaron P
(2006) Microbial community structure in the sea surface microlayer at two
contrasting coastal sites in the northwestern Mediterranean Sea. Aquat Microb
Ecol 42:91-104

Kirchman DL (2000) Uptake and regeneration of inorganic nutrients by marine
heterotrophic bacteria. In: Kirchman DL (ed) Microbial Ecology of the Oceans.
Wiley, New York, p 261-288

Kirchman DL, Yu LY, Fuchs BM, Amann R (2001) Structure of bacterial communities
in aquatic systems as revealed by filter PCR. Aquat Microb Ecol 26:13-22
18

Ledger ME, Harris RML, Armitage PD, Milner AM (2008) Disturbance frequency
influences patch dynamics in stream benthic algal communities. Oecologia
155:809-819

Lee S, Fuhrman JA (1991) Spatial and temporal variation of natural bacterioplankton
assemblages studied by total genomic DNA cross-hybridization. Limnol.
Oceanogr. 36:1277-1287

Liss PS, Duce RA (eds) (1997) The Sea Surface and Global Change, Vol. Cambridge
University Press, Cambridge, UK

Long RA, Azam F (2001) Microscale patchiness of bacterioplankton assemblage
richness in seawater. Aquat Microb Ecol 26:103-113

Moeseneder MM, Arrieta JM, Muyzer G, Winter C, Herndl GJ (1999) Optimization of
terminal-restriction fragment length polymorphism analysis for complex marine
bacterioplankton communities and comparison with denaturing gradient gel
electrophoresis. Appl Environ Microbiol 65:3518-3525

Montoya JM, Pimm SL, Sole RV (2006) Ecological networks and their fragility. Nature
442:259-264

Morris RM, Vergin KL, Cho JC, Rappe MS, Carlson CA, Giovannoni SJ (2005)
Temporal and spatial response of bacterioplankton lineages to annual convective
overturn at the Bermuda Atlantic Time-series Study site. Limnol Oceanogr
50:1687-1696

Murray AE, Blakis A, Massana R, Strawzewski S, Passow U, Alldredge A, DeLong EF
(1999) A time series assessment of planktonic archaeal variability in the Santa
Barbara Channel. Aquat Microb Ecol 20:129-145

Murray AE, Preston CM, Massana R, Taylor LT, Blakis A, Wu K, DeLong EF (1998)
Seasonal and spatial variability of bacterial and archaeal assemblages in the
coastal waters near Anvers Island, Antarctica. Appl Environ Microbiol 64:2585-
2595

Nelson JD, Boehme SE, Reimers CE, Sherrell RM, Kerkhof LJ (2008) Temporal
patterns of microbial community structure in the Mid-Atlantic Bight. Fems
Microbiol Ecol 65:484-493

Norkrans B (1980) Surface microlayers in aquatic environments. In: Alexander M (ed)
Advances in microbial ecology. Plenum Press, New York and London, p 51–83

19

Obernosterer I, Catala P, Lami R, Caparros J, Ras J, Bricaud A, Dupuy C, van
Wambeke F, Lebaron P (2008) Biochemical characteristics and bacterial
community structure of the sea surface microlayer in the South Pacific Ocean.
Biogeosciences 5:693-705

Obernosterer I, Catala P, Reinthaler T, Herndl GJ, Lebaron P (2005) Enhanced
heterotrophic activity in the surface microlayer of the Mediterranean Sea. Aquat
Microb Ecol 39:293-302

Olesen JM, Bascompte J, Dupont YL, Jordano P (2007) The modularity of pollination
networks. P Natl Acad Sci USA 104:19891-19896

Olsen GJ, Lane DL, Giovannoni SJ, Pace NR (1986) Microbial ecology and evolution:
A ribosomal RNA approach. Ann. Rev. Microbiol. 40:337-365

Pace NR, Stahl DA, Lane DL, Olsen GJ (1986) The analysis of natural microbial
populations by rRNA sequences. Adv. Microbiol. Ecol. 9:1-55

Pastor-Satorras R, Vespignani A (2001) Epidemic spreading in scale-free networks.
Physical Review Letters 86:3200-3203

Pommier T, Canback B, Riemann L, Bostrom KH, Simu K, Lundberg P, Tunlid A,
Hagstrom A (2007) Global patterns of diversity and community structure in
marine bacterioplankton. Molecular Ecology 16:867-880

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green
JL, Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ,
Young JPW (2007) The role of ecological theory in microbial ecology. Nat Rev
Micro 5:384-392

Proulx SR, Promislow DEL, Phillips PC (2005) Network thinking in ecology and
evolution. Trends in Ecology & Evolution 20:345-353

Ramette A, Tiedje JM (2007a) Biogeography: an emerging cornerstone for
understanding prokaryotic diversity, ecology, and evolution. Microbial ecology
53:197-207

Ramette A, Tiedje JM (2007b) Multiscale responses of microbial life to spatial distance
and environmental heterogeneity in a patchy ecosystem. Proc Natl Acad Sci U S
A 104:2761-2766

Rosenzweig ML (1995) Species diversity in space and time. Cambridge University
Press, Cambridge p.463
20

Ruan QS, Dutta D, Schwalbach MS, Steele JA, Fuhrman JA, Sun FZ (2006) Local
similarity analysis reveals unique associations among marine bacterioplankton
species and environmental factors. Bioinformatics 22:2532-2538

Schauer M, Massana R, Pedros-Alio C (2000) Spatial differences in bacterioplankton
composition along the Catalan coast (NW Mediterranean) assessed by molecular
fingerprinting. Fems Microbiol Ecol 33:51-59

Seymour JR, Mitchell JG, Seuront L (2004) Microscale heterogeneity in the activity of
coastal bacterioplankton communities. Aquat Microb Ecol 35:1-16

Sieburth JM (1971) Distribution and activity of oceanic bacteria. Deep Sea Res
18:1111-1121

Sieburth JMN, Willis PJ, Johnson KM, Burney CM, Lavoie DM, Hinga KR, Caron DA,
French FW, Johnson PW, Davis PG (1976) Dissolved Organic-Matter and
Heterotrophic Microneuston in Surface Microlayers of North-Atlantic. Science
194:1415-1418

Simek K, Pernthaler J, Weinbauer MG, Hornak K, Dolan JR, Nedoma J, Masin M,
Amann R (2001) Changes in bacterial community composition and dynamics
and viral mortality rates associated with enhanced flagellate grazing in a
mesoeutrophic reservoir. Appl Environ Microb 67:2723-2733

Sogin ML, Morrison HG, Huber JA, Welch DM, Huse SM, Neal PR, Arrieta JM,
Herndl GJ (2006) Microbial diversity in the deep sea and the underexplored
"rare biosphere". P Natl Acad Sci USA 103:12115-12120

Thingstad TF, Lignell R (1997) Theoretical models for the control of bacterial growth
rate, abundance, diversity and carbon demand. Aquat. Microb. Ecol. 13:19-27

Torsvik V, Ovreas L, Thingstad TF (2002) Prokaryotic diversity - Magnitude,
dynamics, and controlling factors. Science 296:1064-1066

Treusch AH, Vergin KL, Finlay LA, Donatz MG, Burton RM, Carlson CA, Giovannoni
SJ (2009) Seasonality and vertical structure of microbial communities in an
ocean gyre. Isme J 3:1148-1163

Upstill-Goddard RC, Frost T, Henry GR, Franklin M, Murrell JC, Owens NJP (2003)
Bacterioneuston control of air-water methane exchange determined with a
laboratory gas exchange tank. Global Biogeochemical Cycles 17:-

21

van der Gast CJ, Jefferson B, Reid E, Robinson T, Bailey MJ, Judd SJ, Thompson IP
(2006) Bacterial diversity is determined by volume in membrane bioreactors.
Environ Microbiol 8:1048-1055

van der Gast CJ, Lilley AK, Ager D, Thompson IP (2005) Island size and bacterial
diversity in an archipelago of engineering machines. Environ Microbiol 7:1220-
1226

Venter JC, Remington K, Heidelberg JF, Halpern AL, Rusch D, Eisen JA, Wu DY,
Paulsen I, Nelson KE, Nelson W, Fouts DE, Levy S, Knap AH, Lomas MW,
Nealson K, White O, Peterson J, Hoffman J, Parsons R, Baden-Tillson H,
Pfannkoch C, Rogers YH, Smith HO (2004) Environmental genome shotgun
sequencing of the Sargasso Sea. Science 304:66-74

Vigil P, Countway PD, Rose J, Lonsdale DJ, Gobler CJ, Caron DA (2009) Rapid shifts
in dominant taxa among microbial eukaryotes in estuarine ecosystems. Aquat
Microb Ecol 54:83-100

Whitman WB, Coleman DC, Wiebe WJ (1998) Prokaryotes: the unseen majority. Proc
Natl Acad Sci U S A 95:6578-6583

Williams PM, Carlucci AF, Henrichs SM, Vanvleet ES, Horrigan SG, Reid FMH,
Robertson KJ (1986) Chemical and Microbiological Studies of Sea-Surface
Films in the Southern Gulf of California and Off the West-Coast of Baja-
California. Marine Chemistry 19:17-98

Woodcock S, Curtis TP, Head IM, Lunn M, Sloan WT (2006) Taxa-area relationships
for microbes: the unsampled and the unseen. Ecol Lett 9:805-812

Wurl O, Holmes M (2008) The gelatinous nature of the sea-surface microlayer. Marine
Chemistry 110:89-97

Yang GP, Tsunogai S, Watanabe S (2005) Biogeochemistry of
dimethylsulfoniopropionate (DMSP) in the surface microlayer and subsurface
seawater of Funka Bay, Japan. Journal of Oceanography 61:69-78

Yang GP, Watanabe S, Tsunogai S (2001) Distribution and cycling of dimethylsulfide
in surface microlayer and subsurface seawater. Marine Chemistry 76:137-153

Zhengbin Z, Liu LS, Wu ZJ, Li J, Ding HB (1998) Physicochemical studies of the sea
surface microlayer - I. Thickness of the sea surface microlayer and its
experimental determination. J Colloid Interf Sci 204:294-299

22

Chapter 2: Spatial Scales of Marine Bacterioplankton Assemblages in the
Southern California Bight

Chapter 2 Abstract

Planktonic marine bacteria interact with each other and with their environment on sub-
millimeter scales. Yet oceanographic studies rely on relatively few representative
samples, typically of a few liters sampled most often kilometers or more apart, and the
adequacy of depicting the microbial community is unclear. To determine the boundaries
of a cohesive habitat (i.e. water patch) for bacterioplankton assemblages in the San
Pedro Channel, bacterioplankton communities were sampled in May 2004 at 18.5 °C
isotherm depths and in 2005 from 5m and at 12.1°C isotherm from a sample grid with
nested sampled areas increasing from 2 km
2
to 35km
2
, and 4 km
2
to 100 km
2
areas.
Bacterial assemblages were also sampled from 5m depth along a 260 km transect along
the Southern California Bight from the Santa Barbara to San Diego. 10 liter samples
were collected by Niskin and the 0.2-1.2um size fraction was extracted and investgated
using Automated Ribosomal Intergenic Spacer Analysis. Communities were fairly
uniform at 2-97km
2
area scales and at distances of 0-15km and the major taxa appeared
remarkably consistent throughout. Variations among the communities appeared to
respond to chlorophyll a and virus abundance gradients at 35km
2
in 2004, but did not
show any correlation with distance or with measured environmental parameters in at
97km
2
in 2005. At distances from 50-255km, there were strong differences in the

23

community similarity values which correlated to temperature and chlorophyll a which
was measured at the sites, though there were still some remarkable similarities with
communities as distant as 45 km. This coherence of the bacterioplankton community in
the San Pedro channel suggests that despite small variations between sites, typical
collection of water by niskin bottle can provide a representative sample of the
bacterioplankton over tens of square kilometers. At greater distances (>100km)
environmental parameters, and differences in water mass drove the variation in
community composition. Taxa-area analysis revealed a scale-dependence for the
bacterioplankton communities with z-values near zero at shorter distances but with
values comparable to soil bacteria or diatoms along a 255km transect.

24

Introduction

Numbering between a few hundred thousand and one million per milliliter, planktonic
marine bacteria undoubtedly interact with each other and with their environment on
sub-millimeter scales. Yet, oceanographic studies rely on relatively few representative
samples, often at km scales or greater, in a patchy, continuously moving environment.
The adequacy of typical oceanographic sampling to describe a microbial habitat and the
geographic scale of that habitat is unclear. Prior to the discovery of the vast number of
unculturable microbes, the similarity of microbial taxa which were grown in vitro, led
to the Baas-Becking (1934) hypothesis of “everything is everywhere, but the
environment selects.”

More recently, investigations into spatial variation of marine bacterioplankton at fine
scales (millimeter to meter) has been attempted using in vivo fluorescence (Seymour et
al. 2004), molecular fingerprinting techniques (Long & Azam 2001), and clone libraries
(Schauer et al. 2000, Kirchman et al. 2001, Casamayor et al. 2002). These studies
suggest that the bacterioplankton communities exhibit remarkable heterogeneity on a
mm-scale, but are increasingly homogeneous on a centimeter-scale, meter scale and
even at the scale of a few kilometers (Acinas et al. 1997, Schauer et al. 2000, Kirchman
et al. 2001, Long & Azam 2001, Casamayor et al. 2002). We have extended these
studies here.

25

Bacterial communities in other environments been investigated using molecular
techniques, and basic ecological properties such as the taxa-area relationship (the rate at
which the communities change with regard to distance) have been used to compare
microbes to animals and plants, and to see to what extent these follow apparently
universal “laws” (Horner-Devine et al. 2004b, Bell et al. 2005, van der Gast et al. 2005,
Green & Bohannan 2006, van der Gast et al. 2006, Fuhrman 2009). Other ecological
patterns, such as the decrease of diversity with latitude, have been shown for marine
bacterioplankton from studies using clone library data (Pommier et al. 2007) and
Automated Ribosomal Intergenic Spacer Analysis (ARISA) (Fuhrman et al. 2008).

Defining microbial habitats and the scales at which they operate is now a major
challenge. Hughes-Martiny et al. (2006) hypothesized that a microbial province can be
described by the relationship of the microbial community to geographic distance and
microbial habitat can be described by the relationship of the microbial community to
environmental parameters. Although it is difficult to separate provinces from habitats
in the ocean, it is reasonably certain that if differences in microbial assemblage
composition correlate with the environmental parameters, there is likely a change in
habitat. The marine bacterioplankton habitat is likely a patchy water mass on the order
of a few km (Hewson et al. 2006b), defined by “fronts” or edge effects or convective
overturn (Morris et al. 2005). Although it is difficult to imagine a correlation of
community composition without a correlation to any environmental variables, a

26

microbial province could be defined operationally by water mass, current system, ocean
basins, or climate zones, though that is beyond the scope of this study.

This study used molecular fingerprints generated through ARISA to investigate the
spatial distribution of bacterioplantkton and to determine the size of the water patch
relevant to those communities. To assist in the comparison, relevant peaks produced by
ARISA were assigned identifications using a 16S-ITS clone library. Bacterial
assemblages were sampled in the San Pedro Channel, off the coast of Los Angeles, in a
nested Grid design in May 2004 at the 18.5° C isotherm over areas increasing from 2
km
2
to 35km
2
(Figure 2-1). Another nested Grid study was performed in May 2005 at
5m and at the depth of the 12.1° C isotherm over areas increasing from 4 km
2
to 100
km
2
areas (Figure 2-2). Bacterial assemblages were also sampled from a 260 km
transect along the Southern California Bight from the Santa Barbara Channel to San
Diego. The collected bacterioplankton assemblages were compared using community
similarity indices as well as through the relative abundances of major OTUs. To test the
relationship between the bacterioplankton, an exploratory analyses was performed
which compared assemblage composition to the environmental variables measured
during sampling. The nested design of the experiment also permitted a taxa-area
anaylsis (through decay of community similarity with distance) which allowed for the
Southern California Bight ecosystem to be put in a broad ecological context.

27

B
A
Figure 2-1. Map of the Southern California Bight showing the average sea surface
temperature for May 2004. The map of average sea surface temperature
shows major water masses for the whole Southern California Bight (A) and a
detail shows the sites and nested grid sampling design (B). (Sea surface
temperature map is from NASA Giovanni visualization tool)

B A

Figure 2-2 Map of the Southern California Bight showing the average sea surface
temperature for May 2005. The map of the entire Southern California Bight
shows the major watermasses and locations of the Santa Barbara-San Diego
transect(A) and a detail shows the sites sampled with the nested grid design (B).
(Sea surface temperature data and map from NASA Giovanni
visualization tool).

28

29

Materials and Methods

Sample Collection: We sampled the bacterioplankton assemblage in the Southern
California Bight during May 2004 (fig 2-1) and May 2005 (fig 2-2). In May 2004 the
sampling took place over 36 hours while the ship followed two submerged drogues (at
the 18.5°C isotherm) in order to track the water mass in lagrangian fashion. The
sampling consisted of boxes increasing in area (2km
2
, 9.5km
2
, and 35km
2
, fig. 2-1b)
centered around these lagrangian drogues (thus the box presumably moved with the
“water mass”). In 2005, the sampling covered a 4km
2
and a 97km
2
area (Figure 2-2b), at
5m and the 12.5°C isotherm at each station. This was done over 6 hours in a
geographically fixed location (ignoring the movement of the water mass). In addition, in
2005 9 samples were taken at 5m depth in a transect from Santa Barbara to San Diego
(Figure 2-1a). Seawater collected by Niskin bottle (10 L) on a CTD array. Parameters
determined using the CTD array included temperature, salinity, beam attenuation,
chlorophyll a fluorescence, and dissolved O2 concentration. Seawater samples were
pre-filtered through Gelman A/E glass fiber filters (nominal pore size 1.2 μm). The A/E
filtered seawater, containing the free-living bacterioplankton that have been shown to be
about 85% of the total bacteria (Lee et al. 1995), was then filtered through a 0.22 μm
Durapore filter (Millipore) to collect the bacteria. Filters were frozen at –80
o
C prior to
analysis at the University of Southern California.
DNA Extraction and Amplification:- DNA was extracted from frozen filters following
(Fuhrman et al. 1988). Briefly, frozen filters were crushed, cells were lysed with hot

30

1% SDS STE buffer, and DNA was purified by phenol:chloroform extraction. DNA
was stored frozen at -80
o
C in TE buffer. Automated rRNA Intergenic Spacer Analysis
(ARISA, Fisher & Triplett 1999, Hewson & Fuhrman 2004) was conducted on 2.5 ng
DNA as measured by PICO Green fluorescence. A standard amount of template
genomic DNA was used in each PCR reaction, with the intention of analyzing the same
amount of bacteria from each sample. PCR reactions (50 μl) contained 1X PCR buffer,
2.5 mM MgCl
2
, 250μM of each deoxynucleotide, 200 nM each of universal primer 16S
– 1392F (5’-G[C/T]ACACACCGCCCGT-3’) and a TET labeled bacterial primer 23S –
125R (5’-TET-GGGTT[C/G/T]CCCCATTC(A/G)G-3’), 2.5U Taq polymerase
(Promega), and BSA (Sigma # A-7030; 40 ng/ μl final conc.). This primer set
specifically targets bacteria, and we know of no major group of marine bacteria in
surface waters whose DNA these primers fail to amplify (Brown et al. 2005, Hunt et al.
2006). Thermocycling was preceded by a 3 min heating step at 94
o
C, followed by 30
cycles of denature at 94
o
C for 30 s, anneal at 56
o
C for 30 s, extend at 72
o
C for 45s,
with a final extension step of 7 min at 72
o
C. Amplification products were cleaned using
Clean & Concentrator-5 (Zymo Research), and purified DNA was measured by PICO
Green fluorescence. Purified DNA diluted to a standardized amount (5 ng µl
-1
) was
loaded in the fragment analysis. Standardization prevented fragment differences
arising from different amounts of loaded DNA. Products were then run for 5.5 h on an
ABI 377XL automated sequencer operating as a fragment analyzer (Avaniss-Aghajani
et al. 1994) with a custom-made ROX-labeled 1500bp standards (Bioventures Inc.). The
sequencer electropherograms were then analyzed using ABI Genescan software.

31

Outputs from the ABI Genescan software were transferred to Microsoft Excel for
subsequent analysis. Peaks less than 5 times the baseline fluorescence intensity were
discarded since they were judged not clearly distinguishable from instrument noise
(Hewson & Fuhrman 2004). With this criterion, the practical detection limit for one
operational taxonomic unit (OTU) is ca. 0.09% of the total amplified DNA (Hewson &
Fuhrman 2004).
Bacterioplankton production:-Bacterioplankton production was estimated by
incorporation of
3
H- leucine into protein as described previously (Kirchman et al. 1985,
Simon & Azam 1989). Briefly, triplicate 10 ml seawater samples in sterile, sample
rinsed, polypropylene centrifuge tubes had
3
H- leucine added to 5 nmol L
-1
final
concentration. The samples were then incubated 1 h at ambient seawater temperature in
the dark. Before
3
H- Leucine addition, one replicate sample was killed by adding 5%
formalin. After incubation, samples were filtered onto 25 mm-diameter 0.45 μm
Millipore (Type HA) nitrocellulose filters. Proteins were treated by incubation for 2 min
with 2 ml ice-cold trichloroacetic acid (TCA). TCA was then filtered through, the
towers were rinsed 3 times with TCA onto the filters, the filters were then rinsed 3
times with TCA, then placed into 6 ml scintillation vials containing 5 ml Ultima Gold
scintillation fluid. After incubating vials for at least 2 hours at room temperature to
allow clarification of filter membranes in scintillation fluid, radiolabel incorporation
was measured in a Beckman-Coulter LS6500 scintillation counter. We used a
conversion factor of 1.5 x 10
17
cells/mol leucine incorporated (Simon & Azam 1989) to
estimate bacterial production.

32

Bacterial abundance- Bacterial abundance was determined by SYBR Green I staining
and epifluorescence microscopy described in (Noble & Fuhrman 1998), detailed
protocol in (Patel et al. 2007). Briefly, samples (50 ml) were fixed with 0.02 μm-
filtered formaldehyde, to a final concentration of 2%, and kept at 4
o
C in the dark until
processing, which occurred within 24 h of sampling. Aliquots (2-5ml) of the samples
were filtered onto 0.02 μm Anodisc Al
2
O
3
filters, drying the filter on tissue paper, then
staining on a 100 μl, 1:2500-diluted drop of SYBR Green I. After staining, the filters
were re-dried on tissue paper and mounted on a glass slide with a solution of 50:50:0.01
Glycerol: Phosphate Buffered Saline: p-phenylenediamine as mountant. Slides were
observed under blue light excitation at 1,000 X magnification on an Olympus BH-60
microscope. More than 200 viruses and bacteria were counted in 10-20 fields.
Statistical analyses: ARISA peaks were standardized by the total area in each sample
in order to compare relative abundances of ARISA peaks. Prior to analysis, the peaks in
replicate samples were binned according to the dynamic binning protocol described in
Ruan, et al 2006. The ARISA peak areas were transformed by ln(x+1), to account for
skew in the distribution. For some analyses, the peak areas were transformed to the
binary presence-absence values.
Bacterial counts, virus counts, and bacterial production were also log transformed. The
environmental parameters were then normalized according to their mean in order to
remove differences in scale. Euclidean distance was calculated among the
environmental parameters in order for comparisons between similarity matrices.

33

Similarity matrices on the ARISA samples were created using the Sorenson Index of
Similarity (compares presence-absence) and Bray-Curtis similarity (compares relative
abundances) (Legendre & Legendre 1998). The 2004 sites and the 2005 sites were
compared through hierarchical clustering using unweighted pair group method with
arithmetic mean (UPGMA) and these relationships were visualized through
dendrograms using PRIMER (Clarke & Gorley 2006).
Tests for correlation between environmental parameters (e.g. temperature, salinity,
chlorophyll a, dissolved oxygen, beam attenuation, density, bacterioplankton
production, bacterial counts and virus counts) and bacterioplankton assemblage
similarity values were carried out using Mantel-type tests using the BIO-ENV routine in
PRIMER v6. To test the Spearman correlation between distance and similarity the
RELATE routine was used. Correlations between environmental parameters other than
distance was examined using the BIOENV routine in PRIMER v6 (Clarke 1993, Clarke
& Gorley 2006) and distance-based multivariate multiple regression was performed
using the DISTLM routine as a part of the PERMANOVA+ package (McArdle and
Anderson 2001). In each of these cases, p-values were calculated by permutation tests.
The relationship of the community similarity versus the geographic distance was tested
directly by fitting a regression model which best predicted the natural logarithm of
similarity values by the natural logarithm of distance values. The z value reported is the
(regression coefficient)/2 as described in Horner-Devine et al. 2004.

34

Results

Site Characterization
In the 2004 study, samples at the 18.5°C isotherm taken at sites A, B, C, and D make up
the corners of the 2km
2
area, with site a in its center, sites E, F, G, and H make up the
corners of the 9.5km
2
area with site e in its center, and sites I, J, and K make up the
furthest corners of the 35km
2
box which is inclusive of all other sites (Figure 2-1). The
average sea surface temperature from the MODIS satellite data (Figure 2-1a) and from
temperatures measured in situ (Figure 2-3a) show a cohesive water mass, though there
was a slight increase in temperature near the East corner (site J), reflecting the accuracy
in sampling the isotherm. Most of the environmental parameters measured at the sites
show no obvious spatial gradients (Table 2-1). However, chlorophyll a fluorescence
(Figure 2-3b), dissolved oxygen (Figure 2-3c), and virus abundance (Figure 2-3d) show
an increase from West to East.

In the 2005 grid study A, B, C, and D defined the corners of the 4 km
2
area and sites E,
F, and G defined the furthest corners of the 97km
2
area (Figure 2-2b); and sites 1-9
show the transect from the Santa Barbara Channel down to San Diego (Figure 2-2a).

35

Table 2-1. Locations and measured environmental and biotic parameters for the 2004 study. Total Bacteria, total viruses,
and bacterial production are reported as averages ± standard deviation.
Site Latitude Longitude
Depth
M
Chloro-
phyll a
µgL
-1

Light Trans-
mission
(%)
Oxygen
mL/L
Temp-
erature
°C
Salinity
ppt
Total
Bacteria
(x10
6
ml
-1
)
Total Virus
(x 10
7
ml
-1
)
Bacterial
Production
leucine
( x10
6
cells*m
-1
*day
-1
)
A 33.55176 -118.401 10 0.31 87.5 4.71 18.48 33.30 1.64 ± 0.03 2.90 ± 0.08 1.59 ± 0.25
A 33.54628 -118.408 7 0.21 88.6 4.69 18.50 33.30 1.38 ± 0.04 3.23 ± 0.25 2.10 ± 0.78
B 33.55708 -118.409 8 0.27 87.5 4.70 18.53 33.31 1.41 ± 0.24 3.04 ± 0.59 3.17 ± 0.50
C 33.5579 -118.391 9 0.44 86.3 4.75 18.62 33.31 1.75 ± 0.19 3.51 ± 0.21 3.37 ± 0.30
D 33.5461 -118.392 11 0.28 88.1 4.71 18.45 33.29 1.39 ± 0.06 3.58 ± 0.08 3.06 ± 0.19
E 33.5438 -118.381 11 0.38 87.3 4.75 18.53 33.31 1.09 ± 0.15 3.72 ± 0.51 3.61 ± 1.97
E 33.5312 -118.398 10 0.26 88.6 4.68 18.54 33.30 1.38 ± 0.06 3.40 ± 0.13 2.50 ± 0.04
F 33.55838 -118.399 8 0.30 87.0 4.73 18.56 33.31 1.57 ± 0.04 3.80 ± 0.04 3.61 ± 0.56
G 33.55674 -118.366 12 0.49 86.6 4.83 18.57 33.30 1.49 ± 0.06 3.69 ± 0.09 2.90 ± 1.40
H 33.52722 -118.368 10 0.42 87.5 4.71 18.60 33.31 1.33 ± 0.04 4.23 ± 0.26 3.27 ± 0.44
I 33.51576 -118.37 10 0.41 87.8 4.73 18.32 33.30 1.65 ± 0.08 4.19 ± 0.12 3.23 ± 0.44
I 33.50326 -118.403 12 0.31 89.2 4.65 18.49 33.30 1.32 ± 0.11 3.65 ± 0.18 3.05 ± 1.03
J 33.55618 -118.338 14 0.90 84.3 5.03 18.58 33.32 1.78 ± 0.12 4.47 ± 0.25 4.12 ± 0.51
K 33.50402 -118.341 13 0.46 88.5 4.69 18.45 33.31 1.25 ± 0.11 4.03 ± 0.17 3.27 ± 0.35
L 33.5287 -118.353 12 0.44 86.9 4.72 18.51 33.31 1.47 ± 0.18 4.25 ± 0.01 8.86 ± 0.98
M 33.54004 -118.364 9 0.54 86.3 4.70 18.88 33.32 1.53 ± 0.16 4.26 ± 0.07 2.65 ± 0.55

36

Figure 2-3 Maps of interpolated environmental parameters measured at the 18.5°C
isotherm in May 2004. In situ tempe e(A), chlorophyll a concentration (B),
dissolved oxygen (C), and virus abundance (D)

D
A
B
C
ratur

37

igure 2-4 Maps of interpolated environmental parameters measured during the
entrations
(B) from the 5m Santa Barbara-San Diego transect sites and in situ temperatures
.

A
B
D C

F
2005 study. Maps show situ temperatures (A) and chlorophyll a conc
(C) and chlorophyll a concentrations (D) from the 12.1°C isotherm grid sites

38

ng
h there is some variation in
easured temperature shown (table 2-2), reflecting the accuracy of isotherm sampling.
showing
s

ey were effectively identical over short distances
(figure 2-5a). Furthermore, among the assemblages that showed the largest differences
from the other sites (sites a, E, H, and J, figure 2-5a), only site J came from an outside
corner of the 35km
2
grid (Fig 2-1b).
The average sea surface temperature from May 2005, from MODIS satellite
measurements, showed a cohesive water mass for the grid study sites, and a strong
gradient indicating different water masses for the northern section of the transect, though
the southern section was more uniform (Figure 2-2b). In situ temperatures at 5m alo
the transect showed a strong North-South gradient (Figure 2-4a). The chlorophyll a
gradient (Figure 2-4b), though patchier than the temperature gradient, suggests multiple
environments for the bacteria within this transect. At the 12.1°C isotherm sites, the
temperature remains relatively constant (Figure 2-4c) althoug
m
There is, however, a chlorophyll a gradient to the Northwest (figure 2-4d)
environmental changes at this isotherm over the 97km
2
area.

Comparison of bacterial assemblage composition in 2km
2
to 97km
2
areas
Comparing the whole assemblages at the 2004 18.5°C isotherm sites, there were no
consistent differences in assemblage composition with distance (Figures 2-5, 2-6). The
clustering of average Bray-Curtis similarity in the 12 highly similar (>80%) communitie
did not correspond to geographic location (Figure 2-5a). Assemblages which were
statistically indistinguishable from each other did tend to come from locations that were
near to one another, suggesting that th

39
Table 2-2 Locations and measured environmental and biotic parameters for the 2005 study. Total Bacteria, total viruses, and
bacterial production are reported as averages ± standard deviation. ND indicates missing data.
Site Latitude Longitude
Depth
(m)
Temp-
erature
(°C)
Chlor-
ophyll a
(µg*L
-1
)
Light
Trans-
mission
(%)
Oxygen
(mL/L)
Salinity
(ppt)
Total
Bacteria
(x10
6
ml
-1
)
Total
Viruses
(x 10
7
ml
-1
)
Bacterial Production
leucine (x10
5
cells*
ml
-1
*day
-1
)
A 33.5488 -118.402 5 16.5 0.74 86.8 5.94 33.36 4.84 ± 0.20 5.54 ± 0.11 7.46 ± 0.34
24 12.0 0.66 93.2 4.04 33.41 2.48 ± 0.19 2.21 ± 0.07 2.39 ± 0.03
B 33.56767 -118.400 5 16.5 0.65 86.2 5.81 33.35 6.21 ± 0.39 4.96 ± 0.03 8.05 ± 1.09
22 12.1 1.02 92.6 4.12 33.42 2.03 ± 0.14 1.81 ± 0.05 2.61 ± 0.17
C 33.56938 -118.377 5 16.4 0.89 86.4 5.95 33.37 4.51 ± 0.43 5.10 ± 0.20 10.2 ± 0.49
23 12.1 0.94 92.6 4.11 33.42 3.01 ± 0.29 2.92 ± 0.02 3.29 ± 0.10
D 33.5507 -118.379 5 16.7 0.54 88.1 5.67 33.35 4.64 ± 0.19 5.49 ± 0.09 5.04 ± 0.81
31 12.1 0.89 93.7 4.09 33.42 2.23 ± 0.14 3.15 ± 0.59 3.06 ± 0.49
E 33.47703 -118.383 5 17.3 0.23 90.9 5.39 33.28 2.30 ± 0.10 4.02 ± 0.24 3.93 ± 0.05
27 12.0 0.81 92.6 4.08 33.37 ND ND 2.50 ± 0.01
F 33.56209 -118.486 5 16.3 0.56 89.0 5.73 33.33 4.83 ± 0.41 4.91 ± 0.39 6.28 ± 0.01
29 12.1 1.57 91.6 4.07 33.42 2.67 ± 0.09 2.86 ± 0.07 4.99 ± 0.39
G 33.47409 -118.494 5 17.1 0.65 89.0 5.50 33.27 2.05 ± 0.10 2.08 ± 0.03 3.86 ± 1.34
42 12.0 0.86 92.8 4.21 33.39 2.37 ± 0.04 2.36 ± 0.11 2.18 ± 0.09
1 34.24018 -119.898 5 13.5 4.11 83.3 6.13 33.51 2.32 ± 0.08 2.61 ± 0.08 14.2 ± 0.42
2 34.11972 -119.428 5 14.4 0.91 92.5 5.69 33.42 3.40 ± 0.06 3.56 ± 0.05 5.96 ± 0.07
3 33.90845 -119.110 5 15.2 1.24 91.4 5.51 33.32 ND ND 8.54 ± 0.13
4 33.73438 -118.763 5 16.2 0.27 91.7 5.44 33.25 ND ND 5.44 ± 0.57
5 33.54812 -118.401 5 16.2 0.89 88.5 5.83 33.39 ND ND 12.2 ± 0.83
6 33.14985 -118.312 5 16.9 0.19 90.8 5.32 33.29 2.37 ± 0.08 4.06 ± 0.06 3.28 ± 0.06
7 33.10553 -117.651 5 17.6 0.08 91.4 5.45 33.29 3.29 ± 0.20 4.13 ± 0.16 7.22 ± 0.33

40
Site Latitude Longitude
Depth
(m)
Temp-
erature
(°C)
Chlor-
ophyll a
(µg*L
-1
)
Light
Trans-
mission
(%)
Oxygen
(mL/L)
Salinity
(ppt)
Total
Bacteria
(x10
6
ml
-1
)
Total
Viruses
(x 10
7
ml
-1
)
Bacterial Production
leucine (x10
5
cells*
ml
-1
*day
-1
)
8 33.14901 -117.930 5 17.5 0.24 90.5 5.26 33.34 1.55 ± 0.46 3.59 ± 0.26 2.60 ± 0.40
9 32.75827 -117.645 5 16.7 0.13 92.6 5.34 33.28 2.72 ± 0.54 3.85 ± 0.18 4.02 ± 0.54
Table 2-2 (cont’d)

J
H
C
D
B
A
F
e
G
i
m
l
I
K
a
E
A

100 90 80 70 60

41

D
G
A
E
F
B
C

B

100 90 80 70 60
Similarit

y

Figure 2-5 Dendrograms showing whole community similarity from nested grid studies
using the Bray-Curtis Index. 2004 nested sites a-m at the 18.5° isotherm (A) and
2005 nested sites A-G at the 12.1°C isotherm (B). Clusters in black show a
significant difference (p<0.05), red clusters show no significant difference.

42

igure 2-6 Dendrograms showing the whole community similarity of nested grid studies
using the Sorenson Index. 2004 nested sites a-m at 18.5° isotherm (A) and 2005

D
E
A
G
B
C
F
J
a
E
C
D
H
G
i
B
m
e
A
F
l
I
K
A
100 90 80 70 60 50
Similarity
B
100 908070 60 50
Similarity

F
nested sites A-G at 12.1° C isotherm (B). Clusters in black show a significant
difference (p<0.05), red clusters show no significant difference.

43

sing the Sorenson Index (and ignoring relative abundances of OTUs) there was a
but a

he whole assemblages at the 2005 12.1°C isotherm sites showed no consistent
the
B,

dex,

U
greater variation of similarity among the bacterioplankton assemblages (55%-94%),
majority of the communities still showed at least 80% similarity to each other (Figure 2-
6a). Assemblages that decreased in similarity when measured by Sorenson index,
particularly at sites E and J, suggests that the rare organisms (i.e. those at the lower
detection limits) are likely driving the decrease in similarity.

T
difference in composition over the distances sampled (figure 2-5, 2-6). Although
areas measured were greater, up to 97km
2
, similarities, 5 of 7 communities more than
80% similar to each other by Bray-Curtis index (Figure 2-5b), and assemblages at sites
C, E, and F were statistically indistinguishable from one another. Included in this cluster
of statistical replicates, were sites from both the 4km
2
area and the 97km
2
area, showing a
wide range of distances that correspond to a single community. The Sorenson similarity
of the 2005 12.1°C isotherm assemblages showed similarities to one another which
ranged from 70% to 83%, with only 2 of 7 communities showing greater than 80%
similarity (Figure 2-6b). Despite the decrease in similarity shown by the Sorenson in
only the assemblage at site D was significantly different from the other sites (Figure 2-
6b). This again implies that the rare organisms are variable among these sites, but these
assemblages share most of the abundant OTUs.

44

2004, the abundant OTUs at 18.5°C isotherm were present across nearly all samples
hich
nces
p
red
n 2005, the major OTUs in the assemblage at the 12.1°C isotherm sites (Figure 2-8),
e

In
(figure 2-7). The 20 most abundant OTUs accounted for an average of 87% of the
relative ARISA abundance. The assemblages were dominated by 5 SAR11 OTUs w
made up an average of 60.8% of the ARISA abundance across all sites (figure 2-7). The
remaining abundant taxa included OTUs from the Bacteroidetes group (formerly known
as CFB), γ-Proteobacteria, α-Proteobacteria (excluding SAR11), Synechococcus,
Actinobacteria, and Marine Group A/SAR406 Group 1 (Figure 2-7). Most differe
among the assemblages at the 18.5°C isotherm sites were driven by fluctuations in the
relative ARISA abundance of the top 5-7 dominant OTUs (Figure 2-7). At site J, the to
3 OTUs (all from the SAR11 clade) become more dominant making up 45% of the
relative ARISA abundance. SAR 92 754 became as much as 5% of the ARISA
abundance in sites I and K, possibly signaling a patch where the community diffe
between the Southeast corner of the 9km
2
and the 35m
2
area.

I
showed greater consistency over a greater area than the 2004 18.5°C isotherm sites. Th
25 most abundant OTUs made up an average of 90% of the relative ARISA abundance of
each of the sampled bacterial assemblages. The SAR11 OTUs were still the most well
represented group among abundant OTUs making up an average of 36%.

e
a A B C D
E G H
i I J K l m
F
SAR 11 Surface 1 686
SAR11 Surface 1 662
SAR11 Surface 1 665
SAR11 Surface 1 669
SAR11 680
Cytophagales 621
Cytophagales/Plastid 538
Marine Group A/SAR406 Grp I 627
Alpha-Proteobacterium 740
SAR92 748
Bacteroidetes 726
Actinobacteria 434
Gamma-Proteobacterium 3 838
OTU 965
Gamma-Proteobacterium 943
Synechococcus MarA Grp 3 1052
Bacteroides 707
Bacteroidetes 593
SAR92 754
Alpha-Proteobacteria-4 703

Figure 2-7 The 20 most abundant OTUs identified from ARISA peaks in the 2004 18.5°C
isotherm sites. Putative IDs shown are assigned from clone library identification.

45

A B
C D
E F G
Actinobacteria 434
SAR11 Surface 1 686
Flavobacterium 854
SAR11 Surface 1 665
Cytophogales/Plastid 538
SAR11 Surface 1 669
Plastid 562
SAR11 Surface 1 662
SAR92 750
Actinobacteria 423
SAR86 IIB 527
Cytophogales 730
OTU 1186
SAR11 683
OTU 551
SAR86 IIA 532
Actinobacteria 419
Alphaproteobacteria 489
Plastid (Prasinophyte) 570
Bacteroides 593

Figure 2-8 The 20 most abundant OTUs identified from ARISA peaks from the 2005
12.1°C isotherm sites. Putative IDs shown are assigned from clone library
identification.

46

47

Actinobacteria, Bacteroidetes (CFB), γ-Proteobacteria, α-Proteobacteria, and Plastids
(which amplify with our bacteria-specific primers) were abundant in the sampled
assemblages as well (figure 2-8). Although overwhelmingly similar, the few differences
between the 12.1°C sites came from changes in the relative abundance of the common
and abundant OTUs (Figure 2-8). For example, Flavobacterium 854 and Plastid 562
decreased sharply in relative ARISA abundance at site D, and SAR86 IIB 527 decreased
to below detection. Prasinophyte Plastid 570, Bacteroides 593, Actinobacter 419 and
SAR11 680 increased at site D compared to other sites.

Rare OTUs did not occur at every site in either grid study, and varied widely in
abundance. In 2004, α-Proteobacterium 677 which occurred only at sites a, C, and J
ranging from 0.1% to 1.7% ARISA abundance while Alteromonas-like Group 4 555
occurred at sites C, K, and m increasing from 0.09% to 1.8% ARISA abundance. In
2005, SAR11 683 occurred at sites A, B and F where it ranged from 0.36-1.5% of ARISA
abundance and OTU 540 occurred at sites B, F, and G where it ranged from 1.13-3.72%
of ARISA abundance. These OTUs were not specific to any area group and likely fell
below the ARISA detection limit at some of the sites sampled.

48

Comparison of assemblage diversity along a 255km transect
There were distinct differences found between the 5m depth bacterioplankton
communities North and South from the San Pedro Channel, although the pattern was not
strictly one of decreasing similarity with distance (Figures 2-9, 2-10). The Sorenson
Index (focusing on presence and absence only, Figure 2-9a) varied among
sites from 79% to 38%. The Bray-Curtis index (comparing) relative abundance, revealed
almost the same pattern, although there was greater variation among sites
ranging from 93% to 27% (Figure 2-9b). The assemblages were split into the same two
major clusters (Figure 2-9) with the exception of Site 1 clustering separately and sharing
only 27% similarity to all other sites by Bray Curtis index (Figure 2-9b). This showed a
completely distinct community at site 1 which, when the stark environmental differences
are considered (table 2-2) suggests a response of the community to habitat as well as
distance. The 2005 5m Grid (sites A-F) and sites 6 and 8 from the southern section of the
2005 transect clustered together, as did sites 2-5 and 7 and 9. This could indicate a
coherent bacterioplankton community as far as 60km apart in the San Pedro Channel
towards San Diego, and closely related communities at sites 3 and 4 or 7 and 9.

This pattern is evident when the Bray-Curtis similarities to site C are interpolated using
the Data Interpolating Variational Analysis (DIVA) gridding across geographic space,
(Figure 2-10) and suggests at least three different communities from North to South along
the Southern California bight. The bacterioplankton assemblage similarities to site C
were depicted spatially.

49

G
D
8
F
B
C
E
6
9
3
4
7
5
1
2
A
1
2
5
3
4
9
7
D
8
F
A
B
C
G
E
6
B
Figure 2-9 Dendrograms showing the community similarity at 5m in the 2005 study. The
100 9080 706050 40 30 20
Similarity
2005 5m nested samples A-G and the 2005 5m Southern California Bight
Transect sites 1-9 using Sorenson Index (A) on presence-absence data and using
Bray-Curtis index (B) on relative abundance data. Clusters in black show a
significant difference (p>0.05), red clusters show no significant difference.

Figure 2-10 Map of the bacterial assemblage similarity from 5m depth during the 2005
study. Nested grid sites are labeled A-F and Santa Barbarba-San Diego 5m
transect sites are labeled 1-9. The contours are interpolated from the Bray-Curtis
similarity to the bacterial assemblage at site C.

50

51

This shows the high similarity (near 80% by the Bray-Curtis Index) of the nested samples
and the decrease in similarity with distance. There is a clear gradient from the Santa
Barbara Channel to San Pedro, while sites 7 and 9 are less similar to the San Pedro
Channel samples than sites 6 and 8.

The pattern shown by the clustering is reflected through the relative abundance of the
major OTUs (Figure 2-11). For most of the sites sampled, the differences among them
were primarily among the relative abundance of the major OTUs, and are best described
by noting exceptions from the sites from the San Pedro Channel (sites A-G). Sites A-G
are dominated by Actinobacterium 434, which makes up an average of 14% of ARISA
abundance across all communities, and by 5 SAR11 Surface-1 group OTUs which make
up an average of 30.5% of ARISA abundance across all communities. The remaining
abundant taxa are from the α-Proteobacteria , γ-proteobacteria, eukaryotic plastids,
Bacteroidetes and three unidentified OTUs: OTU 1186, OTU 570, and OTU 573. There
was an ambiguous identification for OTU 538 due to similar ITS sizes between a
Cytophogales and a Plastid.

F
E G
1 3 4 5
9 8 7 6
2
A B C D
Actinobacterium 434
SAR11-Surface 1 686
SAR11-Surface 1 665
SAR86 IIB 532
SAR11 Surface 1 669
Cytophogales 538
OTU 1186
Alpha-Proteobacterium 489
Plastid 562
Flavobacterium 854
SAR11 Surface 1 662
SAR11 Surface 1 681
OTU 573
Bacteroidetes 593
OTU 570
OTU 476
OTU 479
Alteromonas-like Grp4 555
Sphingobacter 621
Sphingobacter 640
SAR11 Surface I 664
Bacteroidetes 768
OTU 805
Gamma-Proteobacteria 838
OTU 876
Figure 2-11 The 20 most abundant OTUs from ARISA fingerprinting at 5m depth from
the 2005 study. Santa Barbara-San Diego transect 5m sites are labeled 1-9 and 5m
nested grid sites are labeled A-G. The putative identities of the OTUs are
assigned through sequence identification from clone libraries from the San Pedro
Channel.

52

53

Site 1 represents the greatest break from the other bacterioplankton assemblages sampled
where there is a shift to being dominated by unidentified OTUs, with OTU 479, OTU
805, OTU 476, OTU 1186 and OTU 876 making up 40% of ARISA abundance (Figure
2-11). With the exception of OTU 1186, these OTUs were not detected at sites A-G. The
SAR11 OTUs made up less than 7% of ARISA abundance at site 1. Since this site differs
in both OTUs present and in the relative abundance of shared OTUs it is clear that the
assemblage at this site represents a bacterioplankton community distinct from the San
Pedro Channel.

The intermediate sites to the north (sites 2 to 5) and the sites to the south (sites 6 to 9)
suggest mixing between at least two different communities between Santa Barbara and
San Diego. Sites 2 to 5, between the Santa Barbara Channel and the San Pedro Channel
contain OTUs present in site 1 and in sites A-F, but the relative abundance of OTUs more
closely resemble the San Pedro Channel sites (Figure 2-11). Sites 6-9, between the San
Pedro Channel and San Diego contain similar abundant OTUs to the San Pedro Channel,
but share some of the OTUs found in sites 2-5 (Figure 2-11). The most abundant OTUs
are nearly identical between these sites and the 5m grid sites: Actinobacterium 434 and
SAR11 Surface 1 686, SAR11 Surface 1 662, and SAR11 Surface 1 665 become the most
abundant averaging of 40% of ARISA abundance in sites 2-5 and 45% of ARISA
abundance at sites 6-9 (Figure 2-11). The unidentified OTUs which were abundant at
Site 1 are present in sites 2-5( averaging 1.75% of ARISA abundance) but only OTU 476,

54

and OTU 479 were present in sites 6-9 (averaging 0.5% ARISA abundance). The
similarity shown between sites 7 and 9 and sites 3 and 4 seems to be driven largely by
these shared, less abundant OTUs.

Relationship of Bacterioplankton Similarity to Distance and Environmental Parameters
To find a pattern between the similarity of bacterioplankton assemblages and the
distances between them, all pairwise distances and similarity values between
assemblages, were plotted (Figure 2-12) against each other. The data are noisy, but show
that between 0 and 15km, the slope is close to 0, but between 15 and 220km, there is a
distinct downward slope. This relationship holds true for both the Sorenson Index
similarities and the Bray-Curtis similarities. The distance-decay relationship was
modeled for the nested Grid areas at 5m, 18.5°C isotherm depth, the 12.1°C isotherm
depth, and the 5m transect to yield better comparison between the nested sites and the
transect sites as well as for comparison with other studies.

To model the distance-decay relationship a line was fitted of the form
ln(S
C
)= C + 2z(ln(D)).
Where ln(S
C
) is the natural log of the assemblage similarity based on Sorenson Similarity
for presence-absence data or on the Bray-Curtis Similarity for relative abundance, ln(D)
the natural log of the geographic distance between sites, C is a constant, and the z–value
represents the rate of change in species per unit distance (after Horner-Devine et al.
2004a).

DIstance (km )
0 5 10 15 50 100 150 200 250
Bray Curtis Similarity
0
20
40
60
80
100
2004 18.5 D eg. C Isotherm
2005 5m D epth
2005 12.1 D eg. C Isotherm
A
Distance (km)
0 5 10 15 50 100 150 200 250
Sorenson Sim ilarity
0
20
40
60
80
100
2004 18.5 degreee C Isotherm
2005 5m Transect plus Grid
2005 12.1 degree C Isotherm
B
Figure 2-12 Distance in km vs community similarity for the 2004 and 2005 studies.
Similarities measured using the Bray-Curtis index (A) and the Sorenson Index (B)
showing the variability and scale of the bacterioplankton community over
distance in space. Closed circles show data from the 2004 18.5°C isotherm,
triangles are from the 2005 12.1°C isotherm, and open circles include both grid
data and transect data from the 2005 5m.
55

56

The z-values and regression diagnostics are displayed in Table 2-3. Among the nested
sites, only the 2004 18.5°C isotherm assemblages yielded a significant z-value for both
Sorenson similarity (z=0.0305, p=0.007, r
2
= 0.06) and for Bray-Curtis Similarity
(z=0.0285, p= 0.002, r
2
=0.16). Neither the 2005 12.1°C isotherm depth nor the 2005 5m
Grid sites showed a significant pattern of community similarity with distance. The 2005
transect (sites 1-9) showed a significant relationship with distance with a z value of 0.065
(r
2
=0.156, p=0.02) for Sorenson Similarity and a z value of 0.15 (r
2
=0.303, p=0.0007) for
Bray-Curtis Similarity. Combining all 2005 5m depth sites (A-G and 1-9) caused the z
value to decrease to 0.04 (p<0.0001, r
2
= 0.129) for Sorenson Similarity and 0.07
(p<0.001, r
2
=0.29) for Bray-Curtis similarities.

To bolster this analysis, a rank-correlation Mantel test comparing the Bray-Curtis
community similarity to the geographic distance matrix with Spearman rank correlation
(r
s
), was generated using the RELATE procedure in PRIMER v.6. A significant
correlation to geographic distance was found in the 2004 18.5°C isotherm sites (r
s
=0.361,
p=0.02), in the 2005 Transect sites (r
s
= 0.545, p=0.005), and in the 2005 Transect sites
with the 5m Grid included (r
s
=0.598 , p=0.002; Table 2-4). There was no significant
correlation between the bacterial assemblages and distance at either the 2005 12.1°C
isotherm depths or the 2005 5m Grid sites (Table 2-4). The lack of correlation with
distance provides further evidence that the 2005 Grid studies took place within a coherent
bacterial community.

57

Table 2-3 from 5m depth during the 2005 study. The z-values are calculated from
the slope of the regression line.
Site z-value
Standard Error
of z-value T p-value r
2

2004 Isotherm Grid
Sorenson0.031 0.011 -2.7330.007 0.06
Bray-Curtis0.029 0.0075 -3.8540.0002 0.112
2005 Isotherm Grid
Sorenson-0.01 0.0105 0.897 0.3809 0.041
Bray-Curtis-0.01 0.0145 0.697 0.4943 0.025
2005 5m Grid
Sorenson0.036 0.024 -1.5030.1492 0.106
Bray-Curtis0.02 0.0145 -1.4020.177 0.046
2005 5m Transect
Sorenson0.066 0.026 -2.5220.0165 0.158
Bray-Curtis0.15 0.039 -3.8440.0005 0.303
2005 5m Transect + Grid
Sorenson0.041 0.0095 -4.181<0.0001 0.129
Bray-Curtis0.069 0.01 -6.935<0.0001 0.29

58

Table 2-4 Rank Correlations between Bacterial Community Scores, Geographic Distance,
and Environmental Parameters
Sites
Correlation
with
Distance
(r
s
) p-value
Environmental
Parameters
included
Correlation
with
Environmental
Parameters (r
s
) p-value
2004 18.5°C
Grid
0.361 0.027
Chlorophyll
Light Transmission
Density
Virus
Bacterial
Production
0.546 0.065
2005 12.1°C
Grid
-0.029 0.477
Light Transmission
Oxygen
Density
0.426 0.709
2005 5m Grid 0.226 0.231
Virus
Bacterial
Production
0.512 0.39
2005 5m
Transect
0.545 0.005
Oxygen
Density
0.84 0.001
2005 5m
Transect +
Grid
0.598 0.002
Chlorophyll
Light Transmission
Density
Bacterial
Production
0.679 0.0002

59

Environmental factors that may have influenced the bacterial assemblage were
investigated as well using the BIO-ENV routine to test which the rank correlations
between combinations of the measured environmental factors and bacterioplankton
assemblage similarity. No significant correlation between environmental variables and
the bacterioplankton assemblages from the 2005 5m and isotherm depth grid sites (Table
2-4) provides more evidence that this 97km
2
constitutes a coherent environment and
bacterioplankton community. The bacterioplankton assemblages from the 2005 transect
correlated with dissolved oxygen and density (r=0.84, p=0.001) showing the influence of
a physico-chemical parameters on the bacterial assemblages (Table 2-4). When all 2005
5m (transect + Grid) sites were included in the analysis, the bacterioplankton
assemblages correlated with density, chlorophyll a, beam attenuation, and bacterial
production (r=0.664, p=0.0002), showing that biological variables also account for the
differences in bacterioplankton assemblages from Santa Barbara to San Diego (Table 2-
4).

The DISTLM multivariate multiple regression model with forward selection was
performed to test the conditional influence (i.e. partial regression) of combinations of
distance and environmental variables in predicting the bacterioplankton assemblage
similarities. These conditional regression values are chosen beginning with the variable
which describes the largest proportion of the variance first, then searching for the next
largest proportion of the variance, that doesn’t overlap with the variance described by the
first variable, and it continues adding until it no longer increases the adjusted r
2
. In all

60

Grid samples, biological parameters (e.g. chlorophyll a concentration or bacterial
production) were necessary to significantly predict the bacterial assemblage similarities
(Table 2-5).

The 2004 18.5°C isotherm assemblage similarities were predicted best by chlorophyll a
and distance, accounting for 41% of the variance. The 2005 12.1°C isotherm assemblage
similarities were best predicted by a combination of dissolved oxygen, light transmission,
and chlorophyll a accounting for 76% of the variance. Bacterial production significantly
predicted the 2005 5m Grid assemblage similarities explaining 36% of the variance.
Density best predicted the bacterial assemblages at the 2005 5m transect sites explaining
42% of the variance, and density and dissolved oxygen predicted the assemblage
similarity of all 2005 5m (Grid + Transect) sites, explaining 56% of the variance (Table
2-5). Taken together, this provided evidence for a biological influence on the bacterial
assemblage composition within the nested Grid samples and environmental influence on
the assemblage composition over the greater distances in the transect.

61

Table 2-5 DISTLM multivariate multiple regression model output and statistics.
Values shown are from the forward-stepping partial multiple regression model adding
the variables shown increasing the adjusted r
2
.
Sites
(Variables) Adjusted r
2
Pseudo-F P-value
Proportion
al Variance
Cumulative
Variance

2004 Isotherm
Chlorophyll a
0.21808 5.1836 0.0016 0.27021 0.27021
+Distance
0.3194 3.0841 0.0023 0.13994 0.41014
+Viruses
0.34521 1.5125 0.1366 6.60E-02 0.47617
+Oxygen
0.34799 1.0511 0.4141 4.57E-02 0.52186
2005 Isotherm
Oxygen
6.30E-02 1.4031 0.254 0.21913 0.21913
+ Light Transmission
0.14172 1.4588 0.2675 0.20868 0.42781
+Chlorophyll a
0.52993 4.3034 0.0415 0.33715 0.76496
+Viruses
0.57228 1.2971 0.3587 9.25E-02 0.85743
2005 5m
Bacterial Production
0.23269 2.8195 0.009 0.36057 0.36057
+Viruses
0.39823 2.3755 0.1594 0.23825 0.59882
+Chlorophyll a
0.53177 2.1408 0.148 0.16706 0.76588
+Oxygen
0.67917 2.3783 0.1681 0.12717 0.89306
+ Light Transmission
0.7171 1.2682 0.4134 5.98E-02 0.95285
2005 TX
Density
0.34168 5.1521 0.0004 0.42397 0.42397
+Bacterial Production
0.40056 1.6875 0.0891 0.12645 0.55042
+Distance
0.49039 2.0576 0.0583 0.13108 0.68149
+ Light Transmission
0.5087 1.1864 0.3698 7.29E-02 0.75435
+Bacterial Counts
0.55302 1.3966 0.336 7.80E-02 0.83238
2005 TX +G
Density
0.27986 6.8293 0.0001 0.32787 0.32787
+Oxygen
0.44921 5.3045 0.0009 0.19478 0.52265
+Distance
0.47906 1.745 0.0922 6.06E-02 0.58325
+Bacterial Production
0.49003 1.2581 0.2423 4.28E-02 0.62602
+Bacterial Counts
0.50719 1.383 0.1866 4.54E-02 0.67146
+Chlorophyll a
0.51128 1.0836 0.3495 3.53E-02 0.70677

62

Discussion

Evidence for a coherent community and a cohesive habitat at 2-97 km2 scales.
The assemblages from the nested Grid samples in May 2004 and 2005, though variable,
appear to form a coherent community over distances of 2-15km, or 2km
2
to 97km
2
areas.
The isotherm depth samples, which are taken below the mixed layer, were chosen
specifically to attempt to sample within a water mass. The cluster analysis shows that
certainly within the nested 2-35km
2
areas from the 2004 18.5°C isotherm and the 4-97
km
2
areas from the 12.1°C isotherm, the large majority of the assemblages (12 of 16 in
2004 and 5 of 7 in 2005) exhibit more than 80% similarity to one another by Bray-Curtis
similarity which takes abundance into account (Fig 2-5). The 2005 5m depth assemblages
displayed a similar pattern to the 12.1°C depths below them (Figure 2-9). A surprising
finding is that the fixed-depth 5 m assemblages were more similar to one another than the
isotherm depths, even though the temperature range was greater (16.5-17.2°C) at the 5m
depth compared to the isotherm depth (12.0-12.1°C) (Table 2-2, Figure 2-4). Although
this may be due to the higher mixing of the surface waters down to 5m. We noted a slight
decrease in similarity between assemblages using the Sorenson Index (though there was
still a majority of samples from both sites greater than 80% similar) (Figures 2-5, 2-9a),
which we believe is due to the fact that the presence-absence transformation elevates the
less abundant OTUs to equal footing with the abundant OTUs. This increases the
likelihood that stochastic differences from sampling in the “tail” of the distribution will

63

account for the increase in variance seen. This is quite certain to be the case for the 2005
12.1°C isotherm nested samples where the decrease in similarity is coupled with less
likelihood to statistically distinguish sites (Figure 2-5).

The assemblages in the nested studies which did differ must be considered as well, and
particular attention paid to insights offered by the comparison to environmental
parameters and spatial variation. Variance in the community must be caused by
environmental factors (i.e. bottom-up control), top-down factors (e.g predation),
biological interactions (e.g. competition), or stochastic factors (e.g. chance occurrence).
In spite of the high similarity shown in the dendrograms, there is variation in the
assemblages taken from the nested Grid sites particularly sites J, a, and E from the 2004
18.5°C isotherm depths (figs 5,6), sites A, G, and D in the 2005 12.1 °C isotherm depths
(Figures 2-5, 2-6) and sites G, D, and E in the 5m depths (Figure 2-9). Variation among
sites is often attributable to underlying factors such as correlation to environmental
parameters (Hughes-Martiny, et al 2006; Ramette and Tiedje 2007). At the 12.1°C
isotherm depth and the 5m depth for the 2005 nested Grid sites, there was no significant
correlation between assemblage composition and geographic distance, or assemblage
composition and environmental variables (Table 2-4). The direct analysis, modeling the
assemblage by environmental and spatial variables, showed bacterial production was the
only significant environmental descriptor of the 2005 5m assemblage composition and

64

the combination of oxygen, beam attenuation, and chlorophyll predicted assemblage
composition at 12.1°C isotherm sites, although none of these descriptors individually
predicted the assemblage composition (Table 2-5).

The lack of correlation of assemblage composition with environmental and spatial
similarity suggests that the 2005 nested Grid sites represent a single microbial province
(i.e. water mass) and habitat (i.e. coherent environment) (Hughes-Martiny et al. 2006).
Due to the environmental uniformity found at these grid sites (table 2-2, figure 2-4), the
main factors influencing the assemblage composition must be interactions with other
organisms (e.g. predation or competition). In fact, the 2005 Grid 5m and isotherm
assemblage similarities are best predicted biological factors (Table 2-5). Bacterial
production can predict the assemblage composition at 5m. At the 12.1°C isotherm,
including the biological factor chlorophyll a with the physico-chemical factors of
dissolved oxygen and beam attenuation, significantly predicts the assemblage
composition (Table 2-5) and none of the factors alone are able to predict the assemblage
similarities. Though it should be noted that these biological factors do not perfectly
predict the assemblages, and there is unexplained variance in the model that may come
from stochastic or unmeasured environmental or biological factors.

The bacterial assemblage composition at the 2004 18.5°C isotherm depths, although
correlated to both phyisico-chemical and biological factors, was best explained by
biological factors. Temperature appeared patchy with a small increase near the NE

65

corner of the 2004 18.5°C Grid, while dissolved oxygen, chlorophyll a, and virus
abundance showed strong gradients over the areas sampled (Figure 2-3). These gradients
seemed to correspond with the difference in bacterial assemblage composition at sites J,
E, and a which set up on opposite ends of the chlorophyll and virus gradients. Indeed,
assemblage composition was correlated to distance and to chlorophyll, beam attenuation,
density, virus counts, and bacterial production (Table 2-4). Yet only chlorophyll a and
distance were able to predict the assemblage composition using the DISTLM model
(Table 2-5). The correlation to light transmission, density, and chlorophyll, along with
distance may be indicative of a transition to a different habitat, the influx of a different
water mass altering the environment or assemblage composition, or simply the normal
patchiness observed in the water column. The influx of a different water mass or the
influence of the biological and environmental changes in the surface mixed layer at the
18.5°C isotherm is likely in this case, since the isotherm was near (~10m) to the surface
(table 2-1). The correlation with bacterial production could suggest niche competition
and the correlation to virus counts could imply a top-down effect (i.e. removing abundant
OTUs (Thingstad & Lignell 1997). The combination of chlorophyll and virus counts may
indicate a bloom or the end of a bloom among phytoplankton that had an effect on the
composition of the bacterioplankton assemblage. Here also, the prediction is not perfect,
and particularly in the DISTLM model, distance may be serving as a proxy for
unmeasured environmental or biological factors.

66

The high level of similarity among the majority of assemblages in the nested Grid studies
makes it likely that we successfully sampled a coherent microbial community. This
coherent community spanned the 2005 4km
2
and 97km
2
areas at both 5m and 12.1°C
isotherm depths which seemed to come from within a distinctive province and habitat
(Figures 2-5, 2-9). A coherent community also spans the 2km
2
area, the 9km
2
area, and
the majority of the 35km
2
area of the 2004 18.5°C isotherm depth sites (Figure 2-5).
Taken together, this suggests that taking a typical oceanographic sample of 10 liters
within a ten kilometer patch of water will give provide a reasonably representative
sample of the local community. The change in habitat and the possible influence of a
different assemblage towards the eastern edge of the 35km
2
area (Figures 2-3, 2-5) shows
that changes in the bacterial community will likely follow a change in habitat.
Furthermore, features routinely measured, such as chlorophyll a fluorescence can be used
to help determine if a new environment has been reached.

A scale up to 97km
2
(nearly 10 by 10 km) is similar to the few kilometers in distance that
Hewson et al. (2006) hypothesized to be a coherent microbially-relevant patch size in the
oligotrophic ocean. Coherence over several km may seem to contradict studies showing
variability over millimeter to centimeter spatial scales (Long & Azam 2001, Seymour et
al. 2004, Seymour et al. 2005), but we believe these smallest scales are affected by
stochastic and highly localized effects that tend to average out at scales of several meters
to several km. This is analogous to showing that adjacent pinches of beach sand can vary
considerably in composition, while the differences would average out in adjacent buckets

67

Evidence for a variable community at 35-220km scales in response to different microbial
provinces and habitats.
Bacterioplankton assemblages collected at the 5m mixed layer depth in a transect from
Santa Barbara to San Diego exhibit a decrease in similarity over 45-255 km distances
(Figure 2-12). Clustering analysis shows low similarity among the transect sites with site
1 clearly the furthest apart by Bray-Curtis Analysis (Figure 2-9). This nearly distinct
community (not completely distinct, since there is still 27% Bray-Curtis similarity to the
adjacent site) shows a clear break from the San Pedro Channel assemblages, and this site
has the biggest leverage in the distance-decay relationship (Figure 2-12). Temperature
and chlorophyll a show strong gradients (Figure 2-4) and even sea surface temperature
averaged over the whole month shows site 1 is located in a completely different water
mass (Figure 2-2). The decay in similarity from the nested Grid sites (Figures 2-10, 2-
12), and in particular the gradual nature of the decrease in similarity until the dramatic
drop at site 1 (Figure 2-10), suggests a mixing of different assemblages, e.g. through a
mixing of water masses. Interestingly, the assemblages from the 2005 5m sites D and E,
on the eastern boundary of the nested Grid study, were 78% and 85% similar and
statistically indistinguishable from sites 8 (located 60km away) and 6 (located 45 km
away), respectively (Figure 2-9). This relationship, mapped in figure 2-10, may reflect a
mixing front between the stable community in the San Pedro Channel and populations
further south in the Southern California Bight.

68

The decay of assemblage composition with distance implies environmental change. The
variation in similarity values was strongly correlated to distance and to dissolved oxygen
and density (table 2-4). The multivariate model showed bacterial production, density,
and distance best predicted bacterial assemblage composition (Table 2-5). This suggests
that over the 255km distances sampled in the Southern California Bight, both distinctly
different habitats (i.e. environments) and provinces (i.e. water masses) were encountered.
This combination of environmental differences and distance are likely the result of both
physical mixing as well as some historical influence on the assemblage composition. The
addition of the nested samples to the transect illustrates the difference between provinces
and habitats. As the samples from within a single habitat are added, assemblage
composition becomes more correlated to environmental parameters increase (Table 4).

Taxa-Area relationship
A species-area relationship was first generalized by Arrhenius (1921) and Gleason (1922)
to take the form of a power law S=cA
z
, where S is the species richness, A is the area
sampled, c is a constant in log-log space, and z is the rate of change of species per unit
area. Empirical evidence shows z-values ranging from 0.1 to 0.3 for animals and plants
within contiguous habitats and z values between 0.25-0.35 for discrete (non-contiguous)
islands (Rosenzweig 1995, Green & Bohannan 2006). A high z-value shows high
turnover of taxa as you move from habitat to habitat (Horner-Devine et al. 2004b). The
z-value can be lowered by the ease at which the organisms disperse, the ecological
redundancy of the organisms (Horner-Devine et al. 2004a, Prosser et al. 2007), the

69

phylogenetic level at which the organisms are compared (Horner-Devine et al. 2004b),
the sampling technique (Green & Bohannan 2006), the environmental heterogeneity
(Horner-Devine et al. 2004b, Hughes-Martiny et al. 2006, Ramette & Tiedje 2007b), and,
in some cases, the scale at which the taxa-area relationship is studied e.g. z-value
increases with distance studied for plants (Crawley & Harral 2001, Horner-Devine et al.
2004b).

Organisms with low barriers to dispersal tend to have z-values at the lower end of the
range (e.g. 0.15 for birds, 0.1 for butterflies, and 0.08 for ants), while plants can be at the
higher range 0.2-0.3 (Rosenzweig 1995, Horner-Devine et al. 2004b). Microbes,
including bacteria, have z values that overlap the canonical ranges, especially for
noncontiguous islands (Horner-Devine et al. 2004b, Green & Bohannan 2006, Hewson et
al. 2006b, Woodcock et al. 2006, Prosser et al. 2007, Fierer 2008). A distance–decay
relationship has been found for mesopelagic bacteria over 3500 km at 3000m depth in the
Pacific Ocean and over 1000km at 1000m depth in the Atlantic (Hewson et al. 2006a).

Including both the Grid samples and the transect samples from the 2005 study allows for
comparison between different scales (Horner-Devine et al. 2004b, Green & Bohannan
2006). Looking at the nested grid studies and the transect on a single plot, a clear pattern
emerges where the slope is small at short (i.e. 2-15km) distances and larger at greater
distances (fig 2-12). The transect assemblages compared using the Sorenson index
yielded a z-value of 0.06 and the 2005 5m transect + Grid assemblages yielded a z-value

70

of 0.04 (table 3). This is near the z-value of 0.066 reported by (Azovsky 2002) for
benthic diatoms on scales up to 10
-4
to 10
12
km
2
, the bacteria in estuarine sediments up to
0.09 km
2
(Horner-Devine et al. 2004b), and above the z-value of 0.03 soil bacteria 10
2
-
10
8
km
2
(Fierer & Jackson 2006). Although it is less appropriate to use Bray-Curtis
values to calculate the z-value, it is worth noting that the 2005 transect study and the
transect + Grid study z-value inceased to 0.15 in the transect and to 0.066 in the transect
+ Grid from Bray-Curtis analysis (table 2-3). This indicates that there is change in the
assemblage composition, and suggests that the values based on the Sorenson index, may
provide conservative estimates of assemblage change.

For the 2004 and 2005 nested Grid studies, only the z-values of 0.03 for the 2004 18.5°C
isotherm assemblages was significantly different from zero (table 2-3). The change in
taxa over space in the 2004 study was most likely caused by the observed gradients in
chlorophyll a, bacterial production and virus counts, showing both top-down and bottom
up drivers of community change. The 2004 z-value is comparable to the range reported
by (Fierer & Jackson 2006) for soil bacteria on scales of 10
2
-10
8
km
2
, and for the studies
of bacteria from salt marsh sediment at high taxonomic resolutions (z=0.02 for 97% 16S
rRNA gene identity and 0.04 for 99% identity) reported by (Horner-Devine et al. 2004b).
This difference seen in z-values on the small <100km
2
area and <45km distance (table 2-
3) scale, does show a dependence on the distance sampled. This shows an apparently
coherent community within a few km (2-9km
2
) water patch and suggests that this
coherent community has few barriers to dispersal. At larger distances (35km
2
and 97km
2

71

it appears the rate of change in taxa are driven by the heterogeneity (figure 2-3, table 2-1)
or homogeneity (figure 2-4, table 2-2) of the environment and the interactions between
organisms (e.g. bacterial production, virus abundance figure 2-3). This taxa-area
relationship may be analogous to soil bacteria (Fierer & Jackson 2006) and salt marsh
sediment bacteria (Horner-Devine et al. 2004b), and larger organisms such as diatoms or
ants (Rosenzweig 1995, Azovsky 2002) suggested to be driven by environmental
heterogeneity and history of the community (Ramette & Tiedje 2007b, a).
Unlike the z-values reported by Horner-Devine et al. (2004) and by Azovsky (2002),
there does appear to be a scaling of turnover with distance (i.e. the z-value increased at
greater distances) which has been reported in plants (Crawley & Harral 2001).

Changes in abundant OTUs within and among microbial habitats and provinces.
Although the OTUs which make up the bacterial community at the San Pedro Ocean
Time Series site has been described in previous studies, the focus was on the seasonality
(Brown et al. 2005), microdiversity (Brown & Fuhrman 2005), and the monthly and
annual reoccurrence patterns (Fuhrman et al. 2006). Previous studies focusing on spatial
scale found coherent communities over distances of a few km in the oligotrophic open
ocean (Hewson et al. 2006b) but did not describe the OTUs present in those communities.
Here we described the abundant OTUs which account for the coherence observed in the
whole-community comparisons in the nested Grid samples (figures 2-5,2-6,2-9) and the
differences between the different habitats observed over the 255km transect (figure 2-9).

72

The SAR11 Surface 1 group which we report as dominant in the nested Grid samples in
May 2004 and May 2005 (figures 2-7, 2-8, 2-11) has been shown to display seasonality at
this location (Brown et al. 2005) and is has been suggested that this group forms distinct
ecotypes that occur more commonly in oligotrophic waters in the Mediterranean and the
Sargasso Sea (Carlson et al. 2009). The coherence of this community within the
measured 35km
2
and 97km
2
habitats suggests that other OTUs abundant across these
areas, e.g. Actinobacterium 434 in May 2005 (Figures 2-8, 2-11) or Cytophogales 621 in
May 2004 (figure 2-7) may also be associated with these oligotrophic conditions. The
high ARISA abundance of the SAR11 Surface 1 group in May 2004 and May 2005 also
provides further evidence for the seasonality described by Brown and Fuhrman (2005)
and the annual reoccurrence described by Fuhrman et al (2006). As suggested by
Fuhrman et al. (2006) it appears that these coherent communities and their similarity year
to year is likely driven by the seasonal current pattern in Southern California which shifts
to Northward flow of oligotrohpic waters in late spring and early summer (Di Lorenzo
2003).

The differences in abundant OTUs in the 2005 transect and along the edges of the 35km
2

grid samples and the 97km
2
grid samples provide insight into their ecology. The sharp
decrease from 40% to 7% of the community in Actinobacter 434 and SAR11 Surface 1
group ARISA abundance at the northernmost transect sites (Figure 2-11) along with the
increase in temperature and chlorophyll in the northernmost sites (table 2-2) suggests that
the SAR11 Surface 1 group and Actinobacterium 434 are adapted to warmer, more

73

oligotrophic waters, while the major OTUs at Site 1 are likely adapted to colder,
eutrophic waters. The shared OTUs between sites 2-5 and the 5m nested grid (Figure 2-
11) along with the temperature and community similarity gradients (figures 2-4, 2-10)
provide evidence for mixing between patches and suggests that physical mixing is
important in the surface waters of the Southern California coastal current system. This is
similar to the reports of (Hewson et al. 2006b) in oligotrophic surface waters that
suggested a correlation between the community composition and chlorophyll. A
correlation between community composition and chlorophyll may also explain the sharp
increase in Synechococcus at the edge of the 35km
2
Grid in 2004 (figure 2-7) which
appears to be driven by chlorophyll a concentration and virus abundance (tables 2-4,2-5).
The increase in Flavobacterium 854 at the edge of the 97km
2
grid 2005 5m depth (fig 2-
11) does not appear to correlate with any measured environmental characteristics, and
may be due to unmeasured environmental factors or stochastic effects in the environment
or bacterioplankton community.

74

Conclusions

Our results demonstrate that coastal bacterioplankton assemblages in the Southern
California Bight are often uniform over short spatial scales (e.g. 2-97km
2
), particularly
when sampled within a single habitat. However, these communities do vary in response
to environmental gradients as well as spatial distance. The taxa-area analysis yielded a z-
value that scaled with distance, from a near zero to 0.066. A scaling of z-value with
distance in plant communities has been shown to reflect different processes acting on the
communities at different distances (Crawley & Harral 2001). In the Southern California
Bight, also, there are different processes acting on the bacterioplankton community at
smaller scales, such as the 35km
2
area, which is driven by gradients in biological
features (e.g. chlorophyll a and virus abundance) and the larger scales, such as the 255km
transect which is driven primarily by gradients in temperature (although there are also
strong gradients in chlorophyll a). The appearance of transition communities which share
taxa between the Santa Barbara Channel and the San Pedro Channel and between the San
Pedro Channel and San Diego, further suggests physical mixing plays a prominent role in
determining the bacterioplankton community structure. This adds to the evidence for
spatial structure in oceanic bacterioplankton at larger distance scales, with mixing driven
by physical features, such as eddies, and small-scale patchiness shown by Hewson et al.
(2006b). Correlation of community similarity with temperature at larger distances
supports the hypothesis put forward by Fuhrman et al. (2008) that temperature is a major
driver of bacterioplankton diversity on a global scale. Finally, these results also have

75

clear implications for oceanographic sampling of bacterioplankton communities.
Collection of at least 10L of water seems to provide a representative sample of a
bacterioplankton community on the order of tens of square kilometers. Furthermore, the
response of the community to environmental gradients (e.g. temperature, salinity, and
chlorophyll a), at least in the coastal Southern California Bight, means that it is possible
to measure changes in bacterioplankton habitat, and to predict locations to sample, while
a ship is underway.

76

Chapter 2 References

Acinas SG, Rodriguez-Valera F, Pedros-Alio C (1997) Spatial and temporal variation in
marine bacterioplankton diversity as shown by RFLP fingerprinting of PCR
amplified 16S rDNA. Fems Microbiol Ecol 24:27-40

Avaniss-Aghajani E, Jones K, Chapman D, Brunk C (1994) A Molecular Technique For
Identification of Bacteria Using Small Subunit Ribosomal Rna Sequences.
Biotechniques 17:144-149

Azovsky AI (2002) Size-dependent species-area relationships in benthos: is the world
more diverse for microbes? Ecography 25:273-282

Baas-Becking LGM (1934) Geobiologie of Inleiding Tot de Milieukunde The Hague:
Van Stockkum & Zoon

Bell T, Ager D, Song JI, Newman JA, Thompson IP, Lilley AK, van der Gast CJ (2005)
Larger islands house more bacterial taxa. Science 308:1884

Brown MV, Fuhrman JA (2005) Marine bacterial microdiversity as revealed by internal
transcribed spacer analysis. Aquat Microb Ecol 41:15-23

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show marine
microbial diversity: development and application to a time series. Environ
Microbiol 7:1466-1479

Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ, Vergin K (2009)
Seasonal dynamics of SAR11 populations in the euphotic and mesopelagic zones
of the northwestern Sargasso Sea. Isme Journal 3:283-295

Casamayor EO, Pedros-Alio C, Muyzer G, Amann R (2002) Microheterogeneity in 16S
ribosomal DNA-defined bacterial populations from a stratified planktonic
environment is related to temporal changes and to ecological adaptations. Appl
Environ Microb 68:1706-1714

Clarke KR (1993) Non-parametric multivariate analyses of changes in community
structure. Australian Journal of Ecology 18:117-143.

Clarke KR, Gorley RN (2006) PRIMER v6: User Manual/Tutorial, Plymouth, UK

Crawley MJ, Harral JE (2001) Scale dependence in plant biodiversity. Science 291:864-
868

77

Di Lorenzo E (2003) Seasonal dynamics of the surface circulation in the Southern
California Current System. Deep Sea Research II 50:2371-2388

Fierer N (2008) Microbial biogeography: patterns in microbial diversity across space and
time. In: Zengler K (ed) Accessing Uncultivated Microorganisms: from the
Environment to Organisms and Genomes and Back. ASM Press, Washington
DC, p 95-115

Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial
communities. P Natl Acad Sci USA 103:626-631

Fisher MM, Triplett EW (1999) Automated approach for ribosomal intergenic spacer
analysis of microbial diversity and its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Comeau DE, Hagstrom A, Chan AM (1988) Extraction of DNA suitable for
molecular biological studies from natural planktonic microorganisms. Appl.
Environ. Microbiol. 54:1426-1429

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocean
conditions. P Natl Acad Sci USA 103:13104-13109

Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, Brown JH
(2008) A latitudinal diversity gradient in planktonic marine bacteria. P Natl Acad
Sci USA 105:7774-7778

Green J, Bohannan BJ (2006) Spatial scaling of microbial biodiversity. Trends Ecol Evol
21:501-507

Hewson I, Fuhrman JA (2004) Richness and diversity of bacterioplankton species along
an estuarine gradient in Moreton Bay, Australia. Appl Environ Microb 70:3425-
3433

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006a) Remarkable heterogeneity in
meso- and bathypelagic bacterioplankton assemblage composition. Limnol
Oceanogr 51:1274-1283

78

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006b) Temporal and spatial scales of
variation in bacterioplankton assemblages of oligotrophic surface waters. Mar
Ecol-Prog Ser 311:67-77

Horner-Devine MC, Carney KM, Bohannan BJM (2004a) An ecological perspective on
bacterial biodiversity. P Roy Soc Lond B Bio 271:113-122

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM (2004b) A taxa-area
relationship for bacteria. Nature 432:750-753

Hughes-Martiny JB, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL,
Horner-Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S,
Ovreas L, Reysenbach A-L, Smith VH, Staley JT (2006) Microbial biogeography:
putting microorganisms on the map. Nat Rev Micro 4:102-112

Hunt DE, Klepac-Ceraj V, Acinas SG, Gautier C, Bertilsson S, Polz MF (2006)
Evaluation of 23S rRNA PCR primers for use in phylogenetic studies of bacterial
diversity. Appl Environ Microb 72:2221-2225

Kirchman D, Knees E, Hodson R (1985) Leucine Incorporation and Its Potential as a
Measure of Protein-Synthesis by Bacteria in Natural Aquatic Systems. Applied
and Environmental Microbiology 49:599-607

Kirchman DL, Yu LY, Fuchs BM, Amann R (2001) Structure of bacterial communities in
aquatic systems as revealed by filter PCR. Aquat Microb Ecol 26:13-22

Lee S, Kang YC, Fuhrman JA (1995) Imperfect Retention of Natural Bacterioplankton
Cells By Glass Fiber Filters. Mar Ecol-Prog Ser 119:285-290

Legendre P, Legendre L (1998) Numerical Ecology. Developments in Environmental
Modelling 20. Elsevier, Amsterdam, p 853

Long RA, Azam F (2001) Microscale patchiness of bacterioplankton assemblage richness
in seawater. Aquat Microb Ecol 26:103-113

Morris RM, Vergin KL, Cho JC, Rappe MS, Carlson CA, Giovannoni SJ (2005)
Temporal and spatial response of bacterioplankton lineages to annual convective
overturn at the Bermuda Atlantic Time-series Study site. Limnol Oceanogr
50:1687-1696

Noble RT, Fuhrman JA (1998) Use of SYBR Green I rapid epifluoresence counts of
marine viruses and bacteria. Aquatic Microbial Ecology 14:113-118

79

Patel A, Noble RT, Steele JA, Schwalbach MS, Hewson I, Fuhrman JA (2007) Virus and
prokaryote enumeration from planktonic aquatic environments by epifluorescence
microscopy with SYBR Green I. Nat Protoc 2:269-276

Pommier T, Canback B, Riemann L, Bostrom KH, Simu K, Lundberg P, Tunlid A,
Hagstrom A (2007) Global patterns of diversity and community structure in
marine bacterioplankton. Molecular Ecology 16:867-880

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green JL,
Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ, Young
JPW (2007) The role of ecological theory in microbial ecology. Nat Rev Micro
5:384-392

Ramette A, Tiedje JM (2007a) Biogeography: an emerging cornerstone for understanding
prokaryotic diversity, ecology, and evolution. Microb Ecol 53:197-207

Ramette A, Tiedje JM (2007b) Multiscale responses of microbial life to spatial distance
and environmental heterogeneity in a patchy ecosystem. Proc Natl Acad Sci U S
A 104:2761-2766

Rosenzweig ML (1995) Species diversity in space and time. Cambridge University Press,
Cambridge, UK p.460

Schauer M, Massana R, Pedros-Alio C (2000) Spatial differences in bacterioplankton
composition along the Catalan coast (NW Mediterranean) assessed by molecular
fingerprinting. Fems Microbiol Ecol 33:51-59

Seymour JR, Mitchell JG, Seuront L (2004) Microscale heterogeneity in the activity of
coastal bacterioplankton communities. Aquat Microb Ecol 35:1-16

Seymour JR, Patten N, Bourne DG, Mitchell JG (2005) Spatial dynamics of virus-like
particles and heterotrophic bacteria within a shallow coral reef system. Mar Ecol-
Prog Ser 288:1-8

Simon M, Azam F (1989) Protein content and protein synthesis rates of planktonic
marine bacteria. Marine Ecology Progress Series 51:201-213

Thingstad TF, Lignell R (1997) Theoretical models for the control of bacterial growth
rate, abundance, diversity and carbon demand. Aquat. Microb. Ecol. 13:19-27

van der Gast CJ, Jefferson B, Reid E, Robinson T, Bailey MJ, Judd SJ, Thompson IP
(2006) Bacterial diversity is determined by volume in membrane bioreactors.
Environ Microbiol 8:1048-1055

80

van der Gast CJ, Lilley AK, Ager D, Thompson IP (2005) Island size and bacterial
diversity in an archipelago of engineering machines. Environ Microbiol 7:1220-
1226

Woodcock S, Curtis TP, Head IM, Lunn M, Sloan WT (2006) Taxa-area relationships for
microbes: the unsampled and the unseen. Ecol Lett 9:805-812

81

Chapter 3: Stability in Bacterial Assemblages over Hour-Day Time Scales

Chapter 3 Abstract

Although bacterioplankton double on the order of days to weeks, and it is clear that
physiology can respond to diel cycles, most oceanographic studies focus on seasonal,
annual, or even decadal timeframes. Understanding the ecology of marine
bacterioplankton is confounded both by the difficulty in separating the movement of
water from the biological changes with time as well as a lack of knowledge about the
time-frames at which these communities react to biological and environmental change.
Using lagrangian drifters to follow a water patch over 20-30 hours, we examined the
bacterioplankton assemblages in the San Pedro Channel off Los Angeles California in
August 2000 and May 2004, at 19.5ºC and 18.5ºC isotherm depths, respectively.
Samples were also collected at 5m depth and at the Chlorophyll Maximum depth in
2004. A daily time-series at a single geographic location of the bacterioplankton
assemblages at 5m depth was taken near Catalina Island in June 2002 and 2003. The
bacterioplankton community was examined using Automated Ribosomal Intergenic
Spacer Analysis (ARISA). The assemblages within a water patch in 2000 and 2004 and
in the surface mixed layer in 2004 remained remarkably stable (75-90% similarity) over
20-30 hours, while the assemblages at the chlorophyll maximum depth showed more
variability (68-85% similarity). Bacterioplankton assemblages within a water patch
showed no correlation with similarities calculated from measured environmental
82

variables or elapsed time. During daily sampling for one week on Catalina, assemblage
similarity was 78-90% between days, but less than 50% between successive years.
Assemblages sampled daily correlated to time and environmental similarity in 2002 and
2003 and Catalina showed a correlation to environmental similarity. Rare taxa were
more variable than abundant taxa, perhaps responding to transitioning ecological
conditions. Collectively, these results suggest that in the Southern California Bight, a
sample can represent a bacterioplankton community on the scale of 4-6 days.

83

Introduction

Bacterioplankton are the most abundant organisms in the ocean and are responsible for
cycling 20-50% of primary production (Fuhrman & Azam 1982, Azam et al. 1983, Cho
& Azam 1988, Azam 1998, Ducklow 2000, Kirchman 2000). Studying the
bacterioplankton community patterns is critical to understanding the ecology of the
oceans (Fuhrman 2009). To do this it is essential to understand the timeframe at which
these communities change. Temporal changes in the community structure of
bacterioplankton have recently been investigated at oceanic time-series (e.g. Bermuda-
Atlantic Time Series (BATS), Hawaii Ocean Time-Series (HOT), and San Pedro Ocean
Time Series (SPOTS)), some of which have been collecting environmental data for
decade. Often, due to the sampling scheme, these studies look at processes at time
scales of months to years (Murray et al. 1998, Riemann et al. 1999, Morris et al. 2005,
DeLong et al. 2006, Fuhrman et al. 2006, Carlson et al. 2009, Treusch et al. 2009).
Most marine bacterioplankton doubling times are on the order of days to weeks and that
their abundance varies little at a given depth range (Fuhrman 1992, Ducklow 2000,
Fuhrman 2000). However, the relevant timeframe for bacterial assemblage patterns has
not been fully explored and the steady abundance implies that growth and death rates
are well-balanced. Changes in bacterial community composition would be expected to
result from differential growth rates and death rates of the various taxa, i.e. both
“bottom up” and “top down” processes

84

In nearly every location that has been studied, bacterioplankton communities respond
on a seasonal scale. Murray et al. used DNA hybridization methods to show seasonal
changes of Archaea in the Antarctic Ocean (Murray et al. 1998) and the Santa Barbara
Channel (Murray et al. 1999), (Moeseneder et al. 1999) observed seasonal patterns in
spring and summer in the bacterioplankton in the Aegean and Adriatic, and seasonal
patterns of SAR11 ecotypes have been shown in the surface waters BATS (Carlson et
al. 2009), and in the whole bacterioplankton community throughout the water column
(Treusch et al. 2009). Morris et al. (2005) demonstrated a response to the annual spring
deep mixing events within the mesopelagic zone at BATS. At the San Pedro Ocean
Time-Series site, (Brown et al. 2005) showed monthly changes for several lineages
within the SAR11, Bacteroidetes, and Prochlorococcus and Synechococcus groups.
Looking at the whole bacterioplankton community measured monthly over four years,
Fuhrman et al. (2006) reported predictable annual reoccurrence of this community with
cyclical changes, and suggested that it was driven by seasonal environmental factors.
Recently, (Nelson et al. 2008) showed distinct temporal patterns in a coastal two-year
time-series over a 34-km transect in the Mid-Atlantic Bight. (Campbell et al. 2009)
reported no seasonal patterns over in major members of the bacterioplankton
community in the Chesapeake Bay over 2 years, but did find correlation between
bacterial communities and temperature, light, and bacterial production in the summer
and dissolved organic carbon in the winter.

85

Bacterioplankton physiologically respond on a short time-scale, even on diel cycles
(reviewed in (Azam & Malfatti 2007, Gasol et al. 2008). Following drifters over two
days in spring and fall, (Fuhrman et al. 1985) showed diel cycles in bacterioplankton
abundance and production that mirrored primary production. Over 36 hours, (Winter et
al. 2005) found that both viral production and bacterial production were positively
correlated to each other and correlated to diel cycles.

Short term studies focusing on community change on the order of a day or week are less
frequent, but several have been done. (Lee & Fuhrman 1991) using DNA/DNA
hybridization of whole communities, found that communities collected 1 week apart
from Long Island Sound, New York, were 92 to 94% similar to each other, while
samples collected 7 months apart were only 33 to 57% similar. Communities from the
same location collected 2 wk apart were only about 40% similar (Lee & Fuhrman
1991). The authors also reported that, of open ocean samples collected from 25 m water
depth at one geographic location over 3 consecutive days, those from Days 1 and 2 were
>90% similar to each other, while samples collected on the third day were only 50 to
70% similar to those collected on the other 2 days, respectively. Acinas et al (1997)
showed little variation among assemblages in a water patch of a few km over 48 hours.
Thus, these few studies suggest that communities collected within days of one another
are likely but not certain to be very similar in community composition.

86

The ocean is continuously moving and the difficulty of separating spatial features from
temporal features confounds the interpretation of community change over time. Thus
to follow a particular “patch” of water as it moves with currents, the standard
oceanographic approach is to use lagrangian drifters. Hewson et al (2006) performed 7
such drifter studies of 24 to 360 h duration in the oligotrophic Gulf of Mexico, the
North Pacific and the West Tropical Atlantic, showing an average change by Sorensen
index of 12% per day (comparing the presence/absence of taxa and a Whittaker index of
17% per day (comparing proportions in various taxa) across all surface drifter studies.

In this study, we attempted to describe the change in bacterioplankton community
composition over a short time-frame relevant to rates of marine bacterial growth. To
minimize environmental or spatial differences, we sampled the bacterial community of
a single patch of water at an isotherm depth in the San Pedro Channel in August 2000
and May 2004. We also measured daily bacterial community composition at a single
geographic location nearshore to Catalina Island June 2002 and 2003.

We found that the communities showed little change within an offshore water patch
over 20-30 hours in August 2000 and May 2004 and remained stable days to weeks
before and after the studies were performed, and showed no correlation to elapsed time
or environmental similarity. At the nearshore site, the community was also stable, with
adjacent days showing high similarity, but there was a clear relationship to time over 6-
87

7 days and showed 50% community turnover from one year to the next. Across all sites,
rare taxa showed much greater variability over short timeframes compared to abundant
taxa, and nearly all sites showed no correlation to environmental variables.

88

Materials and Methods

Drifter sample collection: - Two drifters with 1-m high and wide cruciform sea-anchors
(constructed of PVC and nylon) adjusted to the initial depth of the isotherm were set out
simultaneously in the San Pedro Channel, CA. Samples were collected over 1 to as
many as 8 days in August 2000 and May 2004 (Figure 3-1). Samples were collected by
Niskin bottle from the water at 19.5ºC, and 18.5ºC isotherm depths between the two
drogues to attempt to follow a single water mass. Environmental parameters collected
by CTD included temperature, depth, salinity, chlorophyll a fluorescence, beam
attenuation, dissolved oxygen, density ( σ- θ). Subsamples were taken for bacterial and
viral enumeration and bacterial production (described below).
Daily Sampling Collection: Daily samples were taken to investigate short term temporal
changes on nearshore mixed layer bacterioplankton over a week in June 2002 and 2003
at 33° 27’ 02” N and 118° 29’ 08” W (fig 1) in the channel between Bird Rock and the
North shore of Catalina Island near the Wrigley Institute of Environmental Studies
(Catalina Island, CA). 20 liter samples were collected at an estimated 5m depth by
hand-Niskin. In contrast to the drogue sampling, these samples were taken in a Eulerian
design, (i.e. sampled from a single geographic site). Subsamples for bacterial and viral
enumeration

2000 ‐45h
2002 &2003 Daily
2000 +98h
2004 Drogue
2000 Drogue

Figure 3-1 Map of sampling locations for the drogue and daily studies. The black dots
are the sampling locations for August 2000 with the drogue sites connected
sequentially by a black arrow. The red dots are the May 2004 drogue study sites
connected sequentially with a red arrow. The green dot is the location of the
Catalina Island site. A plot of temperature and salinity measured
simultaneously with the Drogue samples in 2000 and 2004 showing the
groupings of water masses sampled (inset).

89

90

were also taken. Although temperature was not measured at the time of sampling, sea
surface temperatures were taken from CDIP buoy 092 (33° 37.07’ N, 118° 19.02’ W,
NDBC 46222).
Monthly Sampling: Samples from 5-m and the chlorophyll maximum depth were
collected at the San Pedro Ocean Time Series (SPOTS) Microbial Observatory site (33°
33’ N, 118° 24’ W) on board the R/V Sea Watch, using Niskin bottles approximately
monthly from August 2000 to December 2004 (as described in Fuhrman et al 2006,
Figure 3-1).
Bacterioplankton Collection: At each sampling location, seawater (20 L) was pre-
filtered through Gelman A/E glass fiber filters (nominal pore size 1.2 μm) to remove
eukaryotic cells (containing plastids which complicate the interpretation of ARISA
fingerprints). The A/E filtered seawater, containing the free-living bacterioplankton that
have been shown to be about 85% of the total bacteria (Lee et al. 1995), was then
filtered through a 0.22 μm Durapore filter (Millipore) to collect the bacteria. Filters
were frozen at –80
o
C prior to analysis at the University of Southern California.
DNA Extraction and Amplification:- DNA was extracted from frozen filters following
(Fuhrman et al. 1988). Briefly, frozen filters were crushed, cells were lysed with hot
1% SDS STE buffer, and DNA was purified by phenol:chloroform extraction. DNA
was stored frozen at -80
o
C in TE buffer. Automated rRNA Intergenic Spacer Analysis
(ARISA)(Fisher & Triplett 1999, Hewson & Fuhrman 2004) was conducted on 2.5 ng
DNA as measured by PICO Green fluorescence. A standard amount of template
genomic DNA was used in each PCR reaction, with the intention of analyzing the same
91

amount of bacteria from each sample. PCR reactions (50 μl) contained 1X PCR buffer,
2.5 mM MgCl
2
, 250μM of each deoxynucleotide, 200 nM each of universal primer 16S
– 1392F (5’-G[C/T]ACACACCGCCCGT-3’) and a TET labeled bacterial primer 23S –
125R (5’-TET-GGGTT[C/G/T]CCCCATTC(A/G)G-3’), 2.5U Taq polymerase
(Promega), and BSA (Sigma # A-7030; 40 ng/ μl final conc.). This primer set
specifically targets bacteria, and we know of no major group of marine bacteria in
surface waters whose DNA these primers fail to amplify (Brown et al. 2005, Hunt et al.
2006). Thermocycling was preceded by a 3 min heating step at 94
o
C, followed by 30
cycles of denature at 94
o
C for 30 s, anneal at 56
o
C for 30 s, extend at 72
o
C for 45s,
with a final extension step of 7 min at 72
o
C. Amplification products were cleaned using
Clean & Concentrator-5 (Zymo Research), and purified DNA was measured by PICO
Green fluorescence. Purified DNA was diluted to a standardized amount (5 ng µl
-1
) was
loaded in the fragment analysis. This prevented differences arising from different
amounts of loaded DNA. Products were then run for 5.5 h on an ABI 377XL automated
sequencer operating as a fragment analyzer (Avaniss-Aghajani et al. 1994) with a
custom-made ROX-labeled 1500bp standards (Bioventures Inc.).
The sequencer electropherograms were then analyzed using ABI Genescan software.
Outputs from the ABI Genescan software were transferred to Microsoft Excel for
subsequent analysis. Peaks less than 5 times the baseline fluorescence intensity were
discarded since they were judged not clearly distinguishable from instrument noise
(Hewson & Fuhrman 2004). With this criterion, the practical detection limit for one
92

operational taxonomic unit (OTU) is ca. 0.09% of the total amplified DNA (Hewson &
Fuhrman 2004).
Bacterioplankton production:-Bacterioplankton production was estimated by
incorporation of
3
H-thymidine incorporation into DNA and
3
H- leucine incorporation
into protein as described previously (Fuhrman & Azam 1982, Kirchman et al. 1985,
Simon & Azam 1989). Briefly, 10 ml seawater samples in sterile, sample rinsed,
polypropylene centrifuge tubes were inoculated with either [
3
H]thymidine or
[
3
H]leucine (5 nmol L
-1
final concentration) and incubated 1 hour at ambient SW
temperature in the dark. One replicate was killed before inoculation by addition of 5%
formalin. After incubation, samples were filtered onto 25 mm-diameter 0.45 μm
Millipore (Type HA) nitrocellulose filters to dryness. Proteins were then extracted by
incubation for 2 min with 2 ml ice-cold trichloroacetic acid (TCA). After extraction,
TCA was filtered through, the towers were rinsed 3 times with TCA onto the filters, the
filters were then rinsed 3 times with TCA, then placed into 6 ml scintillation vials
containing 5 ml Ultima Gold scintillation fluid. After incubating vials for at least 2 h at
room temperature to allow solubilization of filter membranes in scintillation fluid,
radiolabel incorporation was measured in a Beckman-Coulter LS6500 scintillation
counter. We used conversion factors of 2 x 10
18
cells/mol thymidine and 1.5 x 10
17
cells/mol leucine and incorporated (Fuhrman & Azam 1982, Simon & Azam 1989) to
estimate bacterial production.
Bacterial abundance: Bacterial abundance was determined by SYBR Green I staining
and epifluorescence microscopy (Noble & Fuhrman 1998, Patel et al. 2007) detailed
93

protocol in Patel et al 2007. Briefly, samples (50 ml) were fixed with 0.02 μm-filtered
formaldehyde, to a final concentration of 2%, and kept at 4
o
C in the dark until
processing, which occurred within 24 h of sampling. Aliquots (2-5ml) of the samples
were filtered onto 0.02 μm Anodisc Al
2
O
3
filters, drying the filter on tissue paper, then
staining on a 100 μl, 1:2500-diluted drop of SYBR Green I. After staining, the filters
were re-dried on tissue paper and mounted on a glass slide with a solution of 50:50:0.01
Glycerol: Phosphate Buffered Saline: p-phenylenediamine as mountant. Slides were
observed under blue light excitation at 1,000 X magnification on an Olympus BH-60
microscope. More than 200 viruses and bacteria were counted in 10-20 fields.
Statistical analyses: ARISA peaks were standardized by the total area in each sample in
order to compare relative abundances of ARISA peaks. Prior to analysis, the peaks in
replicate samples were binned according to the dynamic binning protocol described in
Ruan et al. (2006). The ARISA peaks were log-transformed (by ln(x+1)). The ARISA
peaks were also transformed by presence-absence. Bacterial counts, virus counts, and
bacterial production were transformed by their natural logarithm. The environmental
parameters were then normalized according to their mean in order to remove differences
in scale. Euclidean distance was calculated among the environmental parameters in
order for comparisons between similarity matrices. Similarity matrices on the ARISA
samples were created using the Sorenson Index of Similarity (presence-absence data)
and Bray-Curtis similarity (compares relative abundances) (Legendre & Legendre
1998). The drifter and daily samples were compared through hierarchical clustering
using unweighted pair group method with arithmetic mean (UPGMA) and these
94

relationships were visualized through dendrograms using PRIMER (Clarke & Warwick
2001, Clarke & Gorley 2006). Differences between clusters were tested using the
SIMPROF permutation routine in PRIMER (Clarke 1993, Clarke & Warwick 2001,
Clarke & Gorley 2006). Tests between environmental parameters and the
bacterioplankton assemblages were carried out using Mantel-type tests in PRIMER v6.
The RELATE routine was used to test the Spearman correlation between seriation, i.e.
correlation of rank similarity values with sample order (in this case time). Correlations
between environmental parameters were examined using the BIOENV routine in
PRIMER v6 (Clarke 1993, Clarke & Warwick 2001, Clarke & Gorley 2006).

95

Results

San Pedro Channel
Environmental parameters collected during the 2000 Drogue study show a very stable
environment, with the exception of an increase in temperature at the 14 hour sample
(Table 3-1). The water movement in the San Pedro Channel was N and E during 26
hours in August 2000 and moved in an L pattern going from SW to W to N during 29.5
hours in 2004, possibly from a surface current eddy (Figure 3-1). Plotting measured
temperature vs. salinity (Figure 3-1, inset) shows the distinct water masses that were
sampled in August 2000 and May 2004. It is worth noting that the warmer (19.5°C)
August 2000 isotherm was roughly 10m deeper than the 18.5°C isotherm in May 2004
(tables 3-1, 3-3). Viral abundance increased during the 2000 drogue study (table 3-2,
Fig 3-2A) and correlated strongly (r=0.88, p=0.008) with elapsed time. Although
bacterial abundance also increased over time (Figure 3- 2B), it was not significantly
correlated to either viral abundance (r = 0.51, p=0.24) or to time (r = 0.45, p= 0.32).
Bacterial growth rates in 2000 averaged 0.28 ± 0.06 d
-1
, and in 2004 averaged 0.44 ±
0.13 d
-1
at 5m depth, 0.48 ± 0.10 d
-1
at 18.5°C isotherm depth and 0.31 ± 0.09 d
-1
at the
chlorophyll maximum depth (tables 3-2, 3-4).

Table 3-1 Sampling time, location, depth, and environmental parameters from the August 2000 study.
Sampling
Time (h) Latitude Longitude
Depth
(m)
Temperature
(°C)
Chlorophyll
Fluorescence
(ug/L)
Beam
Attenuation
(percent)
Dissolved
Oxygen
(ml/L)
Salinity
(ppt)
Density
(kg/L)
-45 33.51 -118.44 21 19.78 0.10 88.17 7.83 33.65 23.79
0 33.55 -118.43 20 19.63 0.12 88.1 8.07 33.64 23.82
6.5 33.58 -118.45 22 19.75 0.13 88.56 7.83 33.64 23.79
14 33.58 -118.48 20 20.52 0.11 88.25 7.69 33.64 23.59
20 33.58 -118.5 21 19.65 0.11 88.08 7.91 33.64 23.81
26 33.61 -118.54 19 19.1 0.12 88.49 8.11 33.63 23.95
98 33.55 -118.38 18 19.2 0.13 88.2 8.21 33.65 23.94

Table 3-2 Biotic parameters from August 2000 study. Bacterial production and bacterial and virus
count values ± standard deviation are shown; n.d.- no data

Sampling
Time (h)
Bacterial Production
Thymidine
(x 10
5
cells ml
-1
day
-1
)
Bacterial Production
Leucine
(x 10
5
cells ml
-1
day
-1
)

Bacteria
Abundance
(X10
6
cells ml
-1
)
Virus
Abundance
(x 10
7
particles ml
-1
)
-45 2.61 ± 0.15 1.81 ± 0.02 0.93 ± 0.22 2.74 ± 0.31
0 2.14 ± 0.48 2.19 ± 0.05 0.81 ± 0.01 2.75 ± 0.04
6.5 2.86 ± 0.03 1.6 ± 0.26 1.14 ± 0.03 3.09 ± 0.09
14 2.3 ± 0.15 1.79 ± 0.98 1.02 ± 0.01 3.02 ± 0.03
20 2.69 ± 0.08 3.48 ± 1.12 0.78 ± 0.03 3.24 ± 0.02
26 n.d. n.d. 1.21 ± 0.06 3.64 ± 0.05
98 n.d. n.d. 1.16 ± 0.74 3.73 ± 0.15

96

97
Table 3-3 Sampling time, location, depth, and environmental parameters from May 2004.
Sampling
Time (h) Latitude Longitude
Depth
(m)
Temperature
(°C)
Chlorophyll
Fluorescence
(ug/L)
Beam
Attenuation
(percent)
Dissolved
Oxygen
(ml/L)
Salinity
(ppt)
Density
(kg/L)
0 33.55 -118.40 5.00 18.74 0.25 87.33 4.68 33.31 23.79
10.00 18.49 0.31 87.46 4.71 33.29 23.85
34.00 11.09 1.29 89.66 3.85 33.34 25.47
6 33.54 -118.38 5.00 19.25 0.39 87.40 4.65 33.31 23.67
10.00 18.62 0.41 87.05 4.74 33.30 23.82
38.00 11.70 1.36 89.47 4.44 33.30 25.32
12 33.52 -118.37 5.00 19.03 0.38 87.56 4.67 33.32 23.73
8.00 19.01 0.38 87.61 4.68 33.32 23.73
32.00 11.98 1.24 89.87 4.36 33.31 25.28
24 33.53 -118.35 5.00 19.15 0.24 86.65 4.66 33.33 23.71
11.00 18.98 0.42 86.73 4.68 33.32 23.75
32.00 11.95 1.35 89.81 4.31 33.30 25.28
30 33.54 -118.36 5.00 19.33 0.53 86.16 4.64 33.33 23.66
10.00 18.15 0.57 86.45 4.75 33.30 23.94
32.00 11.90 1.48 89.47 4.20 33.30 25.29

B
A

Figure 3-2 Bacterial and Viral Abundance during the 2000 study. Viral abundance vs.
time (A) and Bacterial abundance vs. time (B) at 19.5°C isotherm depth during
the 2000 drogue study (0-26 hours) and at 45 hours before and 72 hours after the
drogue study. The lines shown are least-squares regression fit to the data.
98

Table 3-4 Biotic parameters from May 2004. Bacterial production and bacterial and virus abundance are reported as
average values ± standard deviation.

Sample
Time (h) Depth(m)
Bacterial Production
thymidine
(x 10
5
cells ml
-1
day
-1
)
Bacterial Production
Leucine
(x 10
6
cells ml
-1
day
-1
)

Bacteria
Abundance
(X10
6
cells ml
-1
)
Virus
Abundance
(x 10
7
cells ml
-1
)
0 5.00 7.14 ± 0.45 3.37 ± 0.08 1.75 ± 0.15 4.62 ± 0.26
10.00 6.13 ± 1.1 1.59 ± 1.10 1.64 ± 0.03 2.90 ± 0.08
34.00 3.25 ± 0.31 1.24 ± 0.01 1.17 ± 0.02 1.23 ± 0.01
6 5.00 5.07 ± 0.1 2.61 ± 0.24 1.35 ± 0.19 3.79 ± 0.05
10.00 5.74 ± 0.12 3.61 ± 1.97 1.09 ± 0.15 3.72 ± 0.51
38.00 5.45 ± 0.52 2.38 ± 0.13 1.57 ± 0.20 1.83 ± 0.09
12 5.00 5.09 ± 0.45 2.76 ± 0.05 1.61 ± 0.09 4.24 ± 0.07
8.00 2.40 ± 0.83 1.77 ± 0.05 1.65 ± 0.07 4.19 ± 0.12
32.00 4.51 ± 0.07 2.09 ± 0.17 1.61 ± 0.10 2.41 ± 0.03
24 5.00 5.56 ± 0.57 10.99 ± 0.10 1.20 ± 0.14 4.06 ± 0.02
11.00 6.68 ± 0.35 8.51 ± 0.26 1.47 ± 0.17 4.25 ± 0.01
32.00 3.79 ± 0.07 4.48 ± 0.19 1.54 ± 0.02 2.48 ± 0.05
30 5.00 7.11 ± 0.74 3.17 ± 0.26 1.40 ± 0.14 3.77 ± 0.16
10.00 6.54 ± 0.04 2.65 ± 0.55 1.53 ± 0.16 4.26 ± 0.07
32.00 2.62 ± 0.01 0.85 ± 0.03 1.39 ± 0.03 1.83 ± 0.10

99
Catalina Island Nearshore Site
Sampling for the daily studies took place at a 30m deep nearshore site. Although
temperatures were not directly measured, sea surface temperatures taken from CDIP
buoy 092 (NBDC 46222) averaged 18.2 in 2002 and 18.7 in 2003 (Table 3-5). Viral and
bacterial abundance were quite stable in daily samples in both 2002 and 2003 (Table 3-
5).

Table 3-5 Sample date, environmental, and biotic parameters for Catalina sites.
Values listed are averages ± standard deviation.
Sample Dates
Virus
Abundance
(x 10
7
cells ml
-1
)
Bacteria
Abundance
(X10
6
cells ml
-1
)
Sea Surface
Temperature
(°C)
2002 June 3-9, 2002 4.83 ± 0.51 3.03 ± 0.26 17.8 ± 0.7
2003 June 14-19, 2003 6.79 ± 0.46 3.36 ± 0.48 18.2 ± 0.2

100

101

Drogue Studies
In the 2000 study, the bacterioplankton assemblage showed minor changes over the
drifter time showing 84-92% similarity when measured by Bray-Curtis similarity
(Figure 3-3A) and, although the Sorenson similarity values were slightly less (78%-
85%), the communities were statistically indistinguishable (Figure3-3B). The Bray-
Curtis Clustering into 3 main groups with communities at 0-20 hours in one cluster and
28 & 96 hours in another, and the community at -45 hours clustering separately (Figure
3- 3A). Sorenson similarities revealed 3 clusters with similarities that reflected the
sampling time (Figure 3-3B). The abundant members of the bacterioplankton
community remained remarkably stable during the 26 hour drifter study and at 45 hours
before and 72 hours after, with SAR11 Surface 1 666 and Prochlorococcus 828
maintaining their high abundance (Figure 3-4A). Taxa averaging less than 5% showed
greater variability (e.g. Cytophogales 728 or CHABI-7 402, Figure 3-4B), although
some remained steady around 1% ARISA abundance (e.g. Actinobacter 434 or SAR11
703, Figure 3-4B). The abundant taxa showed greater variability in the ensuing
monthly samples (Figure 3-4A).

At the 2000 19.5°C isotherm depth, there was no correlation between assemblage
similarity and environmental similarity or between assemblage similarity and time
during the 0-26 hour drogue study (Table 3-6). This implies that any changes in the
communities shown during the 26 hours were not driven by their environment, or at
least not by the environmental parameters measured.

102

98 hours
26 hours
20 hours
6.5 hours
14 hours
0 hours
-45 hours
26 hours
98 hours
20 hours
6.5 hours
14 hours
-45 hours
0 hours
A B
100 95 90 85807
100 95 90 85 80 75
Similarit
5
y
Similarity

Fig 3-3 Dendrograms of bacterial assemblage similarity at the 19 ºC isotherm depth
during the 2000 drogue study. Assemblages sampled during the drogue (0-26
hours) and 45 hours before and 72 hours after the drogue study. Clustering using
Bray-Curtis similarities (A) and Sorenson similarities (B) are shown. Black
clusters are significantly different based on a SIMPROF test; red clusters
indicate no significant difference between the sites.

0
5
10
15
20
25
30
‐45 h0 h6.5 h14 h20 h26 h98 h Aug Sep Oct Dec Jan
A
Percent ARISA Abundance
SAR11 Surface 1 666 Prochlorococcus 828 SAR11 Surface 1 662
SAR11 Surface 1686 SAR11 680 Actinobacter 421
0
1
2
3
4
‐45 hours 0 hours 6.5 hours 14 hours 20 hours 26 hours 98 hours
B
Cytophogales/Plastid 538 Cytophogales 728 SAR11 677
Actinobacter 434 Actinobacter 419 CHABI‐7 402
α‐Proteo grp‐4 701 SAR11 703
Figure 3-4 Graphs of bacterioplankton OTUs during the 2000 drogue study and at 5m
depth in the 5 months following. Abundant OTUs during the 2000 drogue study
(underlined), the 5m depth samples taken 45 hours before and 72 hours after the
drogue, and the same OTUs from 5m depth in monthly samples from the USC
Microbial Observatory at the San Pedro Ocean Time-Series Station (separated
by the break) (A). Taxa averaging less than 5% of the ARISA abundance at the
19.5°C isotherm depth during the 2000 drogue study and samples taken 45 hours
before and 98 hours after the drogue (B).

103

104

Table 3-6 Mantel-type Spearman rank correlations (r
s
) of community
similarity values with time and with environmental similarity.
Sample
Correlation
with time (r
s
) p-value
Correlation with
environment (r
S
) p-value
2000
drogue 0.21 0.25 -0.17 0.93
drogue
+days 0.48 0.02 0.21 0.35
2004
5m 0.20 0.31 0.79 0.30
18.5°C
isotherm 0.29 0.18 0.86 0.30
chlorophyll
maximum 0.02 0.50 0.89 0.30
2002 0.69 0.002 0.33 0.09
2003 0.78 0.003 0.03 0.39

105

When the communities sampled roughly two days before (-45 hours) the drogue study
and the sample 3 days (98 hours) after the drifter study was included in a Mantel-type
test, the community similarities did correlate with time (Table 3-6). Since these
communities were taken before and after the drogue and from sites further South
(Figure 3-1) they are not necessarily from the same water mass, although they appear
similar on the temperature-salinity plot (Figure 3-1 inset).

In the 2004 study, the bacterioplankton communities sampled over 29.5 hours at the 5m
depth and at the 18.5°C isotherm depth were 72-90% similar using either the Bray-
Curtis index or the Sorenson index (Figure 3-5A, B). Clustering by Bray-Curtis
similarities showed 4 clusters 2 of which were 83% similar to one another, only the 0 hr
isotherm community, and the 12 and 24 hour 5m community clustered out separately at
75% and 72% similar, respectively (Figure 3-5A). The Sorenson index rearranged the
clusters slightly, bringing the 12hr and 24hr 5m community into the same grouping as
the 12 and 24hr isotherm communities and the 29.5hr 5m depth community (Figure 3-
5B). These indices show that the 5m and isotherm depth communities were stable and
nearly interchangeable in both taxa present and abundances. The chlorophyll maximum
depth communities were more variable showing 68-85% similarity by Bray-Curtis
index and 60-75% similarity by Sorenson index, with the 0 and 6hr communities
showing the greatest divergence (Figure 3-6 A, B).

106

Figure 3-5 Dendrograms of bacterial assemblage similarity and abundance curves of the
6 most abundant OTUs during the 2004 drogue study at the 5m depth and the
18.5°C isotherm depth. Dendrograms shown are based on Bray-Curtis
similarities (A) and Sorenson similarities (B). Abundance curves of the 6 most
abundant OTUs from ARISA abundance in the 5m depth (C), the 18.5°C
isotherm depth (D) during the 2004 drogue and the 5m depth months before and
after the drogue (C&D).
0
5
10
15
20
25
12 hours 12 hours
24 hours 24 hours
0 hours 12 hours
12 hours 24 hours
29.5 hours 0 hours
0 hours 6 hours
6 hours 6 hours
29.5 hours 29.5 hours
24 hours 0 hours
29.5 hours
100 9590 8580 75 70
Similarity
6 hours
depth
Surface Mixed Layer
18.5 Degree C isotherm depth
100 9590 8580 7570
Similarity
0
10
20
30
Mar Apr May 0 6 12 24 29.5 Jun Jul Aug
SAR11 Surface 1 685 SAR11 Surface 1 666 SAR11 Surface 1 662
SAR11 Surface 1 669 SAR11 680 Bacteroidetes 725
Percent ARISA Abundance
A B
C
D
107

e
6 most abundant OTUs during the 2004 drogue study at the Chlorophyll
maximum depth. Dendrograms are based on Bray Curtis similarities (A) and
Sorenson similarities (B). Abundance curves of the 6 most abundant OTUs are
derived from ARISA abundance at the chlorophyll maximum depth during the
2004 Drogue and the months coming before and after the drogue (C).

Figure 3-6 Dendrograms of bacterial assemblage similarity and abundance curves of th

0
5
10
20
25
15
Mar Apr May 0 6 12 24 29.5 Jun Jul Aug
SAR11 Surface 1 685 SAR11 Surface 1 666 SAR11 Surface 1 669
SAR11 Surface 1 662 Cytophogales 740 SAR 92 748
0 hours
6 hours
12 hours
24 hours
29.5 hours
6 hours
0 hours
29.5 hours
12 hours
24 hours
100 908070 60 50
Similarit
100 90 80706050
Similarity y
A
B
C
Percent ARISA ce dan n Abu
108

e

e taxa in the 5m depth and the chlorophyll maximum depth communities (Fig 3-7).
hyll
depths, but the high p-values, suggest that these correlations are not real (Table 3-
).
The similarity between the 5m and 18.5°C isotherm depth communities is reflected in
the most abundant taxa, which are shared between the two depths (Figure 3-5). Thes
taxa remained stable throughout the 29.5 hour study (Figure 3-5C, D). Among the
abundant taxa at the isotherm depth, only SAR11 Surface 1 666 changed over the 29.5
hours (Figure 3-5D). There was more volatility in abundant taxa at 5m depth with only
Bacteroidetes 725 remaining steady (Figure 3-5C). These taxa remained abundant in
adjacent months at 5m depth (outside the break in Figure 3-5C and D), but they were
clearly more variable on a month to month scale. The chlorophyll maximum depth
communities showed a similar volatility in its abundant taxa to the 5m depth
communities throughout the 29.5hours (Figure 3-6C). Although when these abundant
taxa were compared to adjacent months they were much more variable than in either the
5m or 18.5°C isotherm depth (Figure 3-6C). The rare taxa (i.e. taxa averaging less than
5% ARISA abundance) showed more variability at all depths, and less stability among
th

The community similarity values did not correlate to elapsed time at any depth (Table
3-6). Even with the increased variability of the between communities at the chlorop
maximum depth, there was no clear temporal pattern. The rank correlations values
between the community similarity and environmental similarity were quite high at all
three
6

109

igure 3-7 Rare (<5%) bacterioplankton abundance curves. ARISA abundance curves
measuring <5% average ARISA abundance during the 2004 drogue study at 5m
depth (A), 18.5°C isotherm depth (B), and chlorophyll maximum depth (C).
0
2
4
6
8
0 6 12 24 29.5
SAR11/SAR116 659 Bacteroidetes 593
OTU 479 Bacteroidetes 652
OTU 965 Cytophogales/Plastid 538

0
1
2
3
4
5
0 6 12 24 29.5

OTU 965 SAR11/SAR116 659 SAR92 755

Synechococcus 1052 Cytophogales/Plastid 538 Cytophogales 740
0
2
4
6
0 6 12 24 29.5

F

Cytophogales 621 Cytophogales/Plastid 538 SAR11 683
γ‐Proteobacterium 943 OTU 627 Bacteroidetes 768
Percent ARISA Abundance
A
B
C
110

aily Studies
lustering analysis showed a stable assemblage over days in June 2002 and June 2003
igure 3-8). In 2002, there were 3 clusters which separated almost sequentially with
ays 0 &1 and days 2 &3 separating from days 4, 5, & 6 by 75% and from one another
y 78% in both the Bray-Curtis (Figure 3-8A) and Sorenson (Figure 3-8B)
endrograms. Within the clusters, the communities were statistically indistinguishable
n adjacent days and were 85% similar by Bray-Curtis and by Sorenson index (Figure
-8). In 2003 the Bray-Curtis index showed 3 main clusters, which separated the
acterioplankton community on days 3&4 from days 1&2 by 83% with days 0 and 5
lustering separately (Figure 3-8A). The Sorenson index showed only 2 main clusters
with th
he communities also clustered by year showing low (50%) interannual similarity by
Bray-Curtis and Sorenson indices (Figure 3-8 A, B).

Considering only the OTUs with high abundances (Figure 3-8 C, D), the major
bacterioplankton were remarkably stable. The abundant taxa were mainly from the
SAR11 group, with Actinobacterium 434 and Cytophogales 622 also abundant in 2002
and 2003, respectively (Figure 3-8C, D). SAR11 Surface 1 685 and SAR11 Surface 1
666 show a drop from day 1 to day 2 going from 12%-6% ARISA abundance but by
day 5 and 6 they are back to 10% abundance (Figure 3-8C).

D
C
(F
d
b
d
o
3
b
c
e communities from days 0 & 1 separated by 77% from days 2-5 (Figure 3-8B).
T
111

Figure 3-8 Daily bacterial assemblage similarities over one week, separated by one year
in June 2002 and June 2003. Dendrograms showing the relationship of the
community group average determined by Bray-Curtis Similarities (A) and by
Sorenson Similarities (B). Relative abundance of the top 6 abundant OTUs for
the 2002 daily coastal samples(C) and the 2003 daily coastal samples (D) and
the ARISA abundance of the same OTUs from the San Pedro Ocean Time-
Series for the months surrounding the daily samples.

Day 5
Day 0
Day 3
Day 4
Day 1
Day 2
Day 6
Day 4
Day 5
Day 0
Day 1
Day 2
Day 3
100 90 8070 6050 40
Similarity
Day 0
Day 1
Day 5
Day 2
Day 3
Day 4
Day 5
Day 4
Day 6
Day 0
Day 1
Day 2
Day 3
100 90 80 70 6050
year
2002
2003
40
Similarity
0
5
10
15
20
25
Apr
May
0
1
2
3
4
5
6
Jun
Jul
Aug
Sep
SAR11 Surface 1 685
SAR11 Surface 1 666
SAR11 Surface 1 669
SAR11 Surface 1 662
Actinobacterium 434
SAR11 680
April
May
Jun
0
1
2
3
4
5
Jul
Aug
Sep
SAR11 Surface 1 669
SAR11 Surface 1 666
SAR11 Surface 1 685
SAR11 Surface 1662
SAR11 680
Cytophogales 622
A
B
D C
Percent ARISA Abu ce ndan
0
2
4
6
alpha‐Proteobacteria‐4 702
Cytophogales/Plastid 538
Cytophogales/Plastid 538
Bacteroidetes 725
OTU 750 Alteromonas‐like grp 2b 647
OTU 577 Actinobacter 421
Roseobacter 987
OTU 570
Bacteroidetes 593
gamma‐proteobacterium 942
0
1
2
3
112

Figure
% in
ived relative abundance.
3-9 Daily rare bacterioplankton ARISA abundance curves over a week in June
2002 and 2003. Members averaging <5% of the community in 2002 (A) and
2003 (B), averaging <3% of the community in 2002(C) and 2003(D) and <1
2002 (E) and 2003 (F). All abundances are ARISA der

Cytophogales 965
OTU 617
gamma‐protebacteria 943
Bacteroidetes/SAR92 761
SAR92 748
OTU 619
0
0.2
0.4
0.6
0.8
1
012 345 6
Plastid 562
OTU 479
SAR92 754
Synechococcus MarA grpI 1035
SAR11 Surface 3 718
Prochlorococcus loba grpI 828
Cytophogales 727
OTU 573
OTU 479
gamma/beta‐proteobacterium 837
Prochlorococcus loba grp1 828
beta‐proteobacteria 849
012 345
Days Days
ARISA Abundance Percent
Alteromonas‐like grp 2b 647
Mar GrpA/SAR406 626
Bacteroidetes 593
SAR86 IIB 530
OTU 489
SAR116 654
113

In 2003, these abundant bacterioplankton remained steady with only SAR11 Surface 1
666 varying more than a few percent in 2003 (Figure 3-8D). In general, the abundant
taxa in the daily samples are also abundant in the San Pedro Channel in months
immediately before and after the daily sampling (Figure 3-8C, D), but there is volatility
on a monthly scale (e.g. Actinobacterium 434 in 2002, Figure 3-8C, and Cytophogales
622, Figure 3-8D).
ee if this stability existed at all levels of the community, the less abundant members
a with ARISA abundance averaged between 5% and 0.1%) were also plotted against
maintaining their abundance, particularly
those averaging 5% ARISA abundance (Figure 3-9A, B) and 2% ARISA abundance
(Figure 3-9C,D). However, there was greater volatility in the abundance of these rare
taxa compared to the abundant taxa especially those averaging less than 1% ARISA
abundance (Figure 3-9E,F). There are intriguing abundance changes in rare taxa that
show a steady increase or decrease from rare or undetectable to abundant (e.g.
Unknown OTU 577 in 2002 and Prochlorococcus 828 in 2003 (Figure 3-9 A, D)).

These communities displayed a clear temporal pattern (Figure 3-8) and both the 2002
nd 2003 community similarities correlated to time (Table 3-6). The 2002 communities
abunda
ental similarities (Table 3- 6).

To s
(tax
time (Figure 3-9). Many taxa were steadily
a
also correlated to environmental similarity, i.e. the combination of temperature, virus
nce, and bacteria abundance. The 2003 communities did not correlate to
environm
114

Discussion

Our results demonstrate a bacterial communities are remarkably stable in offshore
waters over short periods (hours to days), but can be more variable over longer
timeframes. Within a water patch, and in surface waters, the bacterial communities were
stable from 20-30 hours in the San Pedro Channel, and 2-3 days at Catalina Island.
Bacterial communities changed over time after 4-7 days in surface waters both offshore
and nearshore locations. At each site, rare taxa showed much greater variability
compared to abundant taxa, with some taxa increasing from undetectable to abundant.
Environmental and biological factors likely act on these communities over4-7 days, yet
at all but the 2002 Catalina Island sites, the bacterial communities were uncorrelated to
measured changes in environmental and biological factors. Thus, the communities must
be responding to unmeasured factors or these factors do not drive all members of the
community equally.

Community composition within a water patch
Within a water patch, bacterial communities were stable from 20-30 hours in the San
Pedro Channel (Figures 3-3, 3-5). The bacterioplankton community was the least
variable in the 2000 isotherm depth showing 86-92% Bray Curtis similarity over 26
hours (fig 3A) and slightly more variable 75-90% Bray-Curtis similarity over 29.5h at
the isotherm depth in 2004 (Figure 3-4A). The isotherm depth community in 2000
showed an average difference of 13 ± 1.4% per day and the isotherm depth communities
115

in 2004 showed an average difference of 14 ± 4.1% similarity per day. This was
ightly lower than the average 17% community difference per day reported by Hewson
as
hours
ere
4). These growth rates are averages across complex assemblages
ontaining cells growing at different rates (Luna et al. 2004). Therefore, rare, fast
before they are influencing the major
x
,
r
sl
et al. (2006) in open ocean waters, although the change in the community in 2004 w
capable of reaching 17-18% difference. Interestingly, the communities taken 45
before and 96 hours after the drogue in August 2000 showed similar difference (16%
and 15% respectively, Figure 3-3) from the isotherm community even though they w
taken south and south east of the drogue (Figure 3-1).

A possible reason for the lower variability in August 2000 is the lower average growth
rate (0.28 d
-1
) in August compared to 0.48 d
-1
in May 2004. This suggests that May
2004 is a more dynamic community than August 2000 and implies higher loss rates
from the bacteria in 2004 from predation by viruses or grazers (Thingstad & Lignell
1997, Simek et al. 2001), since the total number of bacteria is not increasing over the
time studied (Table 3-
c
growing taxa may influence growth rates, even
abundance. The lower values for the Sorenson index compared to the Bray-Curtis inde
in the 2004 drogue (Figure 3-5), is likely a result of the influence of these rare taxa
which are given the same weight as abundant taxa for Sorenson similarities.

Hewson et al. (2006) suggest that variability may be introduced due to mixing wate
masses even when following a drifter, since drifters cannot account for small-scale
116

e
t
Pedro Channel current system,
at exhibits strong seasonality with southward flow in winter changing to a weak
5).

st twice as
st as the fastest rate of change in the community composition at the isotherm depth.
eddies, mixing, and shear (O’Donnell et al. 1997). The L-shaped movement of th
water mass in 2004 compared to the straighter northward flow in 2000 makes greater
mixing more likely in 2004 and suggests that the August 2000 assemblages may be par
of a much bigger habitat (or similar water patch) in August 2000 (Figure 3-1). In
August, there was also greater water column stratification (the 19.5°C isotherm was
10m deeper than 18.5°C isotherm), fewer total bacteria, and higher water temperature
(Tables 3-1, 3-3). This is most likely driven by the San
th
northward flow in early summer and a stronger northward flow in late summer and fall
that tends to produce more oligotrophic conditions (Di Lorenzo 2003).

The community at 5m depth above the 29.5 h drogue in 2004 showed 72-87% Bray
Curtis similarity (Figure 3-5). Although slightly more variable, the 5m depth
community is quite similar to the community at the isotherm depth and appears
interchangeable with that community at all but the 12 and 24 hour mark (Figure 3-
The similarity is likely driven by mixing between the surface layer community and the
isotherm just below it (sometimes only 3m apart, Table 3-3). Their growth rates are
similar 0.44 d
-1
in 5m depth and 0.48 d
-1
in the isotherm depth. These growth rates
would allow for a turnover of the community near in 48-72 hours which at lea
fa

117

5m
and
aphic
In
hift
-8) and were more similar to one another at 6 and 7 days than the chlorophyll
maximum depth was in May at 29.5 hours. This result is reminiscent of the high
The community at the chlorophyll maximum depth showed 67-83% Bray-Curtis
similarity (Figure 3-6). These changes could be driven by biological interactions (e.g.
allelopathy) or the sampling of different water patches containing different
communities. Since the average growth rate was smaller (0.31 d
-1
) compared to the
and isotherm depths (0.44-0.48 d
-1
), it is less likely that productivity was driving this
interaction. While biological interactions must play a role, it also should be noted that
although this feature was sampled following a drogue, the drogue was not tracking the
chlorophyll maximum depth (which is biologically rather than physically defined)
may not be a sample of the identical water patch as should be happening at the isotherm
depth.

Community composition over days at a single location
Clustering analysis showed a stable bacterioplankton assemblage in one geogr
location over 7 days in June 2002 and 6 days June 2003 (Figure 3-8, Table 3-5).
2002, the communities were very similar on adjacent days, (85%), but were only 78%
similar after 2 days and 75% similar after 3 days. In 2003 the communities sampled
showed 83% similarity even at 3 days apart in the middle of the sampling, where the
communities were 88-90% similar on adjacent days, but the community seemed to s
on day 0 and day 5 which were 82% and 78% similar to the other days. These day to
day similarities are as high as the surface water in the drifter studies (Figures 3-3, 3-5,
3
118

nas et
ted by Lee
%
n
ely driven by changes in the abundant taxa (Figure 3-8C,D). These changes
ere most likely driven by physical movement of the water by the sampling location.
d 2003 but showed greater
not
02

e
similarities along a coastal-offshore transect in the Mediterranean reported by Aci
al. (1997) in the Mediterranean Sea. Although not quite as drastic, these changes in
community similarity over days are close to the high daily similarities repor
and Fuhrman (1991) at an open ocean site in the Pacific and the steep drop to 50-70%
similar after 3 days. However, the interannual similarity of these communities was 50
(Figure 3-8), which was the low end reported for 3 day differences by Lee and Fuhrma
in the Pacific (1991).

In spite of the high day to day similarity, there was a temporal pattern (Figure 3-8). This
was larg
w
The sea surface temperature was similar in both 2002 an
variability in 2002, with a range of 17 to 18.5°C. Although bacterial production was
measured, both bacterial abundance and viral abundance were remained steady in 20
and 2003 (Table 3-5). Additional stabilizing influences may come from influence of the
coastline which can cause the formation of small scale eddies. Previous studies on th
currents in the wake of islands in California have shown that the surface waters near
Catalina Island, in spite of being in the lee of the island, are well mixed due to their
exposure to San Pedro Channel currents, wave, and tidal action (Dong & McWilliams
2007, Dong et al. 2009).

119

t,

y 0 to 1% at day 5 in 2003, (Figures 3-9F) over 6 days. Within a
ater mass over 20-30 hours, many of these bacterioplankton remain steady at lower

e
stable and can sustain something akin to a growth curve. This
me Prochlorococcus 828 taxon is also present in the 2002 nearshore in low
bundance (Figure 3-9C) and is one of the most abundant in offshore waters August
Role of rare bacterioplankton
Our results show that bacterioplankton that are rare can rapidly change to be abundan
on the order of days (Figures 3-4B, 3-7, 3-9). In the daily samples at one location,
unknown OTU 577 increased from less than 1% to more than 5% ARISA abundance in
2002 (Figures 3-9A). Even for bacteria which averaged less than 1% abundance, there
were taxa which showed a steady increase in abundance: e.g. SAR116 653 increasing
from undetectable at da
w
abundance, but for those that change, the patterns seem to be more chaotic, e.g.
Bacteroidetes 593 at 5m depth or SAR11 683 at the chlorophyll maximum depth in
2004 (Figure 3-7) or Cytophogales 728 in 2000 (Figure 3-4B). It is important to note,
however since these rare taxa are being measured at the limits of ARISA sensitivity,
some of these changes are likely real and others may be caused by random noise.

One intriguing example is the rapid increase in relative abundance of Prochlorococcus
828 in the 2003 daily samples (Figure 3-9D) which displays almost a classical
exponential growth curve. The year before, this same Prochlorococcus 828 decreases
from a few percent to undetectable (Figure 3-9C). This would imply not only that th
environment sampled was favorable the growth of Prochlorococcus 828, but that the
environment is relatively
sa
a
120

ococcus 828 has been previously been shown to increase
ton
009).
ts
anipulation on timeframes short enough to avoid bottle effects.
e
ity is to
2000 (Figures 3-4A). Prochlor
from rare to abundant on monthly scales at the San Pedro Ocean time-series (Brown et
al. 2005, Fuhrman et al. 2006) to become seasonally important in this bacterioplank
community.

With molecular techniques probing further into the rare biosphere (Sogin et al. 2006,
Huber et al. 2007), it has become important to investigate the ecology of the rare
bacterioplankton in situ. Rare taxa may represent part of the community that is
especially responsive to shifts in the environment (Pedros-Alio 2006, Fuhrman 2
This is not limited to bacterioplankton: (Countway et al. 2005) described drastic shif
in the abundance of protists from undetectable to abundant over a period of days in
bottles and in the water column. Microbes that can increase in abundance on short
timeframes as described here, make good candidates for ecologically relevant
m

Community-environment interactions
Community changes over time are driven by environmental heterogeneity (i.e. changes
in their habitat), biological interactions (changes in their relationship to other
organisms), or from the “history” of that particular community (i.e. structure that cam
from earlier relationships in the community (Hughes-Martiny et al. 2006, Prosser et al.
2007, Ramette & Tiedje 2007)). One way to test these controls on the commun
correlate the community similarity matrices to environmental similarity matrices in a
121

milarities the 2000 drogue and the 2004 drogue were uncorrelated to time
nd to environmental similarities (Table 3-6). This is most likely due to the stable
ters
, stable
able 3-6) and 2002
orrelated weakly (0.33, p=0.09) environmental similarity based on sea surface
ommunities are driven by a combination of
ical
Mantel test (Mantel 1967, Legendre & Legendre 1998) the rank correlation coefficients
and p-values determined by random permutation are listed in Table 3-6.

Community si
a
habitats for these bacterioplankton communities which did not show dramatic shifts
(Tables 3-1, 3-3, 3-5). For the 2004 surface water communities, the biotic parame
were also stable (Tables 3-4, 3-5). A scenario that could cause this is if a single
community within a single, stable habitat was sampled (Hughes-Martiny et al. 2006,
Ramette & Tiedje 2007). For the 2004 communities, this is the most likely scenario,
since they were within the center of a water patch which showed little to no spatial
variation (discussed in Chapter 2).

The 2002 and 2003 daily samples were clearly correlated to time (T
c
temperature. This suggests that these c
unmeasured environmental parameters (Ramette & Tiedje, 2007), interactions among
bacteria (Long & Azam 2001), grazing pressure (Simek et al. 2001), and phys
mixing of water masses. Since there was no correlation to virus or bacteria abundance,
it is unlikely that viral predation played a major role.

122

ver, the
ommunity similarities did not correlate to environmental parameters (Table 3-6).
a

it

ARISA taxa, or that the increase in the number of
iruses was not great enough to see an effect over this amount of time (Schwalbach et
ophyll maximum depth below it
igure 3-5, 3-6). However, even the less stable communities sampled would not
change faster than one day. At a single location over 6-7 days, there seemed to be at
Including the assemblages sampled before and after the August 2000 drogue surface
water allowed for the community similarities to correlate with time. Howe
c
There was also a sharp increase in virus abundance (Figure 3-2A, Table 3-1) which also
correlated to time but not significantly correlated (r = 0.51, p=0.24) to bacterial
abundance (fig 2B), or to community similarity (Table 3-6). In fact, the abundant tax
at the isotherm depth, remained steady throughout the 141 hours (Figure 3-3). This
uncoupling of virus abundance from abundant taxa conflicts with a straightforward
“kill the winner” model (Thingstad and Lignell, 1997) where the dominant species are
selectively removed by viruses. These viruses may be targeting less abundant taxa, or
is possible that the increase in viruses was not targeting a specific set of taxa or that
they target a sensitive subset of each
v
al. 2004).

Implications for Sampling Bacterioplankton
Our results show that while there is a detectable difference between sampling within a
water mass and at a single site as the water moves by. In fact in 2004, we
simultaneously sampled a community that remained stable within a water mass at the
isotherm depth and an unstable community in the chlor
(F
123

hat
e a
t
s et

. Changes in more abundant taxa will likely be caught at this level, since
ey are stable on the order of a week (Figures 3-4, 3-8), however, any short term
ked up with a weekly sampling regime.

need at
least a 4 day window where the community was highly similar. In oligotrophic waters
this may extend to slightly more than a week (Figure 3-2). Together, this suggests t
at nearshore and offshore sites, sampling every 3-4 days or, even weekly will captur
representative sample of a bacterial community within a habitat. This is similar to the
weekly rate reported by Hewson et al in 2006 in the oligotrophic ocean.

Many studies which measure the change in bacterioplankton over time are looking a
monthly to seasonal effects (Murray et al. 1998,1999, Moeseneder et al. 1999, Morri
al. 2005, Brown et al. 2005, Fuhrman et al. 2006, Nelson et al. 2008, Carlson et al.
2009, Treusch et al. 2009, Campbell et al. 2009). At the monthly and seasonal scale, the
variability seen in the rare taxa and any response of these taxa to short term changes
will be lost, except in cases where rare taxa increase to become dominant (e.g. Brown
et al. 2005, Fuhrman et al. 2006), or the sampling fortuitously catches one of these taxa
at their peak
th
variations in these abundant taxa might be pic
Based on the response of the rare taxa in this environment, studies which are looking at
a fine-scale interactions or a response to inputs into the environment, would
least a sampling regime that took place every other day.

124

ter variability
ver short timeframes compared to abundant taxa, and these rare taxa are candidates for
g
a
ll

r
r changes in abundant
acterioplankton, while sampling on longer time scales (e.g. bi-weekly to monthly or
reater), is sufficient to capture major changes in the abundant bacterioplankton.

Conclusions

Our results demonstrate a bacterial communities are remarkably stable in offshore and
nearshore waters over short periods (30 hours offshore and 2-3 days nearshore), but
changed over time after 4-7 days. At all sites, rare taxa showed much grea
o
responses on these shorter time scales. With the exception of the 2002 Catalina daily
samples, a lack of correlation with measured environmental variables suggests that the
interactions between bacteria or between bacteria and other organisms may be drivin
the changes on the 4-7 day scale, compared to the environmental changes found to drive
this community on monthly scales (Fuhrman et al. 2006). Virus abundance, including
sharp increase in August 2000, was uncorrelated to community change and bacterial
abundance, suggesting that the response between viruses and bacterioplankton in these
communities is not the straightforward kill the winner hypothesis (Thingstad & Ligne
1997). High community similarity over 4-5 days within a single water patch and in
surface waters suggests that at nearshore and offshore sites, sampling every few days o
even weekly will capture a representative sample of a bacterial community within a
habitat. Finally sampling on short time scales (e.g. every other day to weekly), is
necessary to capture the response of the rare taxa or mino
b
g

125

Chapter 3 References

vaniss-Aghajani E, Jones K, Chapman D, Brunk C (1994) A Molecular Technique For
Identification of Bacteria Using Small Subunit Ribosomal Rna Sequences.

280:694
Azam F, Fenchel T, Field JG, Gray JS, Meyer-Reil LA, Thingstad F (1983) The
Series 10:257-263
Azam F, Malfatti F (2007) Microbial structuring of marine ecosystems. Nat Rev

clone libraries and automated ribosomal intergenic spacer analysis to show
Environ Microbiol 7:1466-1479
Campbell BJ, Yu L, Straza TRA, Kirchman DL (2009) Temporal changes in bacterial
l
57:123-135
Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ, Vergin K (2009)
zones of the northwestern Sargasso Sea. Isme Journal 3:283-295
Cho BC, Azam F (1988) Major role of bacteria in biochemical fluxes in the ocean's

structure. . Australian Journal of Ecology 18:117-143.
Clarke KR, Gorley RN (2006) PRIMER v6: User Manual/Tutorial, Plymouth, UK
Clarke KR, Warwick RM (2001) Change in Marine Communities: an approach to

ased
on 18S rDNA from seawater incubations in the western North Atlantic. J

A
Biotechniques 17:144-149
Azam F (1998) Microbial control of oceanic carbon flux: The plot thickens. Science

ecological role of water-column microbes in the sea. Marine Ecology Progress

Microbiol 5:782-791
Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
marine microbial diversity: development and application to a time series.

rRNA and rRNA genes in Delaware (USA) coastal waters. Aquat Microb Eco

Seasonal dynamics of SAR11 populations in the euphotic and mesopelagic

interior. Nature 332:441-443
Clarke KR (1993) Non-parametric multivariate analyses of changes in community

statistical analysis and interpretation, 2
nd
edn. Plymouth, UK
Countway PD, Gast RJ, Savai P, Caron DA (2005) Protistan diversity estimates b
Eukaryot Microbiol 52:95-106
126

eLong EF, Preston CM, Mincer T, Rich V, Hallam SJ, Frigaard NU, Martinez A,
s
3
Current System. Deep Sea Research II 50:2371-2388
bility

ong CM, McWilliams JC (2007) A numerical study of island wakes in the Southern

ucklow HW (2000) Bacterial production and biomass in the oceans. In: Kirchman DL

Fisher cer
d its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636
Fuhrma
: Falkowski PG, Woodhead AD (eds) Primary productivity and
biogeochemical cycles in the sea. Plenum Press, New York, p 361-383
Fuhrma )
ic
- Evaluation and Field
Results. Marine Biology 66:109-120
uhrman JA, Comeau DE, Hagstrom A, Chan AM (1988) Extraction of DNA suitable
ppl.
Mar.
D
Sullivan MB, Edwards R, Brito BR, Chisholm SW, Karl DM (2006)
Community genomics among stratified microbial assemblages in the ocean'
interior. Science 311:496-50

Di Lorenzo E (2003) Seasonal dynamics of the surface circulation in the Southern
California

Dong CM, Idica EY, McWilliams JC (2009) Circulation and multiple-scale varia
in the Southern California Bight. Prog Oceanogr 82:168-190
D
California Bight. Cont Shelf Res 27:1233-1248
D
(ed) Microbial Ecology of the Oceans. Wiley-Liss, New York, p 85-120
MM, Triplett EW (1999) Automated approach for ribosomal intergenic spa
analysis of microbial diversity an

n JA (1992) Bacterioplankton roles in cycling of organic matter: the microbial
food web. In

n JA (2000) Impact of Viruses on Bacterial Processes. In: Kirchman DL (ed
Microbial Ecology of the Oceans. Wiley-Liss, Inc.

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Azam F (1982) Thymidine Incorporation as a Measure of Heterotroph
Bacterioplankton Production in Marine Surface Waters

F
for molecular biological studies from natural planktonic microorganisms. A
Environ. Microbiol. 54:1426-1429

Fuhrman JA, Eppley RW, Hagstrom A, Azam F (1985) Diel variation in
bacterioplankton, and related parameters in the Southern California Bight.
Ecol. Prog. Ser. 27:9-20

127

n

asol JM, Pinhassi J, Alonso S, xe, ez L, Ducklow H, Herndl GJ, Kobl, xed, zek M,
s a
icrob Ecol
53:21-38
Hewso kton species along
an estuarine gradient in Moreton Bay, Australia. Appl Environ Microb 70:3425-

uber JA, Mark Welch D, Morrison HG, Huse SM, Neal PR, Butterfield DA, Sogin
006) Microbial
biogeography: putting microorganisms on the map. Nat Rev Micro 4:102-112
Hunt D

bacterial diversity. Appl Environ Microb 72:2221-2225
Kirchm tion and Its Potential as a
Measure of Protein-Synthesis by Bacteria in Natural Aquatic Systems. Applied

irchman DL (2000) Uptake and regeneration of inorganic nutrients by marine

ee S, Fuhrman JA (1991) Spatial and temporal variation of natural bacterioplankton

ee S, Kang YC, Fuhrman JA (1995) Imperfect Retention of Natural Bacterioplankton

Legend umerical Ecology, Developments in Environmental
Modelling 20, Elsevier, Amsterdam, p.853

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocea
conditions. P Natl Acad Sci USA 103:13104-13109
G
Labrenz M, Luo Y, Mor, xe, n XAG, Reinthaler T, Simon M (2008) Toward
better understanding of microbial carbon flux in the sea*. Aquat M

n I, Fuhrman JA (2004) Richness and diversity of bacterioplan
3433
H
ML (2007) Microbial population structures in the deep marine biosphere.
Science 318:97-100

Hughes-Martiny JB, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL,
Horner-Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S,
Ovreas L, Reysenbach A-L, Smith VH, Staley JT (2

E, Klepac-Ceraj V, Acinas SG, Gautier C, Bertilsson S, Polz MF (2006)
Evaluation of 23S rRNA PCR primers for use in phylogenetic studies of

an D, Knees E, Hodson R (1985) Leucine Incorpora
and Environmental Microbiology 49:599-607
K
heterotrophic bacteria. In: Kirchman DL (ed) Microbial Ecology of the Oceans.
Wiley, New York, p 261-288
L
assemblages studied by total genomic DNA cross-hybridization. Limnol.
Oceanogr. 36:1277-1287
L
Cells By Glass Fiber Filters. Mar Ecol-Prog Ser 119:285-290
re P, Legendre L (1998) N
128

.

una GM, Dell'Anno A, Giuliano L, Danovaro R (2004) Bacterial diversity in deep

Mantel N (1967) Detection of Disease Clustering and a Generalized Regression
Approach. Cancer Research 27:209-&
Moesen
al-restriction fragment length polymorphism analysis for complex marine
bacterioplankton communities and comparison with denaturing gradient gel

Morris RM, Vergin KL, Cho JC, Rappe MS, Carlson CA, Giovannoni SJ (2005)
Temporal and spatial response of bacterioplankton lineages to annual convective

Murray F
(1999) A time series assessment of planktonic archaeal variability in the Santa

Murray K, DeLong EF (1998)
Seasonal and spatial variability of bacterial and archaeal assemblages in the
5-

elson JD, Boehme SE, Reimers CE, Sherrell RM, Kerkhof LJ (2008) Temporal
s

oble RT, Fuhrman JA (1998) Use of SYBR Green I rapid epifluoresence counts of

O’Donnell J, Allen AA, Murphy DL (1997) An assessment of the errors in Lagrangian
velocity estimates obtained by FGGE drifters in the Labrador Current Journal of

atel A, Noble RT, Steele JA, Schwalbach MS, Hewson I, Fuhrman JA (2007) Virus
en I. Nat Protoc 2:269-276

Long RA, Azam F (2001) Antagonistic interactions among marine pelagic bacteria
Appl Environ Microb 67:4975-4983
L
Mediterranean sediments: relationship with the active bacterial fraction and
substrate availability. Environ Microbiol 6:745-753

eder MM, Arrieta JM, Muyzer G, Winter C, Herndl GJ (1999) Optimization of
termin
electrophoresis. Appl Environ Microbiol 65:3518-3525
overturn at the Bermuda Atlantic Time-series Study site. Limnol Oceanogr
50:1687-1696
AE, Blakis A, Massana R, Strawzewski S, Passow U, Alldredge A, DeLong E
Barbara Channel. Aquat Microb Ecol 20:129-145
AE, Preston CM, Massana R, Taylor LT, Blakis A, Wu
coastal waters near Anvers Island, Antarctica. Appl Environ Microbiol 64:258
2595
N
patterns of microbial community structure in the Mid-Atlantic Bight. Fem
Microbiol Ecol 65:484 - 493
N
marine viruses and bacteria. Aquatic Microbial Ecology 14:113-118
Atmospheric and Oceanic Technology 14:292–307
P
and prokaryote enumeration from planktonic aquatic environments by
epifluorescence microscopy with SYBR Gre
129

,
icrobial ecology. Nat Rev
Micro 5:384-392
Ramett n emerging cornerstone for
understanding prokaryotic diversity, ecology, and evolution. Microb Ecol

Rieman
e NE Monsoon periods
in the Arabian Sea studied by denaturing gradient gel electrophoresis (DGGE)

Ruan Q
lgorithm for binning microbial community profiles.
Bioinformatics 22:1508-1514
Schwal
t Microb Ecol 34:117-127

utrophic reservoir. Appl Environ Microb 67:2723-2733

ogin ML, Morrison HG, Huber JA, Welch DM, Huse SM, Neal PR, Arrieta JM,

Pedros-Alio C (2006) Marine microbial diversity: can it be determined? Trends in
Microbiology 14:257-263

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green
JL, Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ
Young JPW (2007) The role of ecological theory in m

e A, Tiedje JM (2007) Biogeography: a
53:197-207
n L, Steward GF, Fandino LB, Campbell L, Landry MR, Azam F (1999)
Bacterial community composition during two consecutiv
of rRNA genes. Deep-Sea Res Pt Ii 46:1791-1811
, Steele JA, Schwalbach MS, Fuhrman JA, Sun FZ (2006) A dynamic
programming a

bach MS, Hewson I, Fuhrman JA (2004) Viral effects on bacterial community
composition in marine plankton microcosms. Aqua

Simek K, Pernthaler J, Weinbauer MG, Hornak K, Dolan JR, Nedoma J, Masin M,
Amann R (2001) Changes in bacterial community composition and dynamics
and viral mortality rates associated with enhanced flagellate grazing in a
mesoe

Simon M, Azam F (1989) Protein content and protein synthesis rates of planktonic
marine bacteria. Marine Ecology Progress Series 51:201-213
S
Herndl GJ (2006) Microbial diversity in the deep sea and the underexplored
"rare biosphere". P Natl Acad Sci USA 103:12115-12120

Thingstad TF, Lignell R (1997) Theoretical models for the control of bacterial growth
rate, abundance, diversity and carbon demand. Aquat. Microb. Ecol. 13:19-27

130

noni
ertical structure of microbial communities in an
ocean gyre. Isme Journal 3:1148-1163
Winter

Treusch AH, Vergin KL, Finlay LA, Donatz MG, Burton RM, Carlson CA, Giovan
SJ (2009) Seasonality and v

C, Smit A, Herndl GJ, Weinbauer MG (2005) Linking bacterial richness with
viral abundance and prokaryotic activity. Limnol Oceanogr 50:968-977
131

Chapter 4: Bacterial Community Composition in the Sea Surface Microlayer and
Surface Waters in the Southern California Bight.

Chapter 4 Abstract

The sea surface microlayer (SML) is widely reported to be enriched in bacterial
abundance and activity. It is an environment distinct from the surface waters of the
ocean, and may select for a unique bacterial community. We investigated the free living
bacterial community (fraction excluding particles, 0.2 ~ 1.2 m) in the SML nearshore
to Catalina Island in May 2004, May 2005, and October 2006, and offshore of Los
Angeles and San Diego, CA in May 2005. We collected the bacterial community in the
SML using a nylon screen in 2004 and a Teflon coated drum in 2005 and 2006 and
compared them to free-living bacterioplankton communities in the immediate
underlying water (0.5m depth). Our results showed a distinct sea surface microlayer
community in some locations (as little as 50% similarity), and no difference between
the communities at others (90%). The ratio of bacterial abundance in the SML
compared to the ULW ranged from depleted to moderately enriched in the microlayer
(SML:ULW = 0.4-3.5, average=1.9) and bacterial production, measured by
3
H-Leucine
incorporation, was depleted (0.03-0.9 average=0.25). Viral abundance had SML ULW
values closer to 1 (0.7-1.3, average=1.04), i.e. there was little depletion or enrichment.

132

The wide range of similarities found between the SML and ULW communities and the
high number of shared taxa between the communities, suggests a patchy microlayer
community that is likely formed by bacteria being transported from the surface waters
to the microlayer, where they encounter different selection factors.

133

Introduction

The sea surface microlayer is operationally defined as the top millimeter of the ocean
(Hardy 1982, Liss & Duce 1997). Because of its location at the air-sea interface, the sea
surface microlayer is a fundamental part of regulating biogeochemical and geochemical
processes between the ocean and the atmosphere (Liss & Duce 2005). The sea surface
microlayer has been described as a hydrated gelatinous layer (Sieburth 1983, Cunliffe &
Murrell 2009) and it has been shown to be enriched in organic compounds such as
carbohydrates, proteins, amino acids, lipids, humic substances, and transparent
exopolymer particles (Williams et al. 1986, Gasparovic et al. 1998, Zhengbin et al.
1998, Wurl & Holmes 2008). This enrichment would be expected to favor microbial
growth. However, this layer is subject to intense solar radiation, including UV-B which
can be harmful to organisms (Regan et al. 1992), tends to be enriched in heavy metals
and pollutants (Hardy et al. 1987), and it is influenced by wind and wave energy, and
atmospheric deposition (Norkrans 1980, Liss & Duce 2005) – all factors that may
inhibit growth.

These differences in environment could provide a distinct, if transient, habitat for
bacteria. This habitat might select for bacteria with high UV resistance (Agogue et al.
2005b) for those involved in biogeochemical processes such as DMS, DMSP, and
methane cycling (e.g. DMS, DMSP, and methane cycling (Yang et al. 2001, Upstill-
Goddard et al. 2003, Franklin et al. 2005, Yang et al. 2005). Most previous work has
134

reported that in the SML, bacteria range from 2-200 fold of enrichment in abundance
and production compared to the bacterioplankton community (Sieburth 1971, Carty &
Colwell 1975, Carlucci et al. 1985, Carlucci et al. 1991, Agogue et al. 2004,
Obernosterer et al. 2005). Increased rates of respiration in the SML have also been
reported in different coastal and offshore marine environments (Williams et al. 1986,
Garabetian 1990, Obernosterer et al. 2005, Reinthaler et al. 2008). However, the
response is far from uniform, and recent studies at offshore sites in the South Pacific
(Obernosterer et al. 2008), the South Atlantic and Mediterranean (Reinthaler et al.
2008), and in the Baltic Sea (Stolle et al. 2009) found depleted bacterial production in
the SML. It is still an open question if bacteria in the SML are a unique community
thriving at the air-water interface, or a collection of microbes brought up from the ocean
(Bezdek & Carlucci 1972) or deposited from the atmosphere (Norkrans 1980, Liss &
Duce 1997).

The examination of the biodiversity of the bacteria in the SML with modern molecular
techniques has begun only recently (Agogue et al. 2005a, Agogue et al. 2005b, Franklin
et al. 2005) and there are fewer studies which have investigated the community
structure in the SML and the corresponding underlying water (ULW). Using a 16S
clone library, Franklin et al. (2005) reported a distinct bacterial community in the SML
compared to underlying water in the North Sea, with two taxa accounting for 80% of
the clones. Studies in Hawaii have found similar bacterial communities compared to
their corresponding ULW (Cunliffe et al. 2009). Using a molecular fingerprinting
135

method, Agogue et al. (2005) reported communities in the SML that were both distinct
and similar to the ULW communities when sampled one day apart at the same location
in the Mediterranean. Still other studies have found few differences between bacteria in
the SML and the ULW in the Mediterranean (Joux et al. 2006), and in the South Pacific
(Obernosterer et al. 2008).

The method of collection of the SML is as important as the methods used to measure
and detect the bacteria which reside there. There have been few comparative studies
(Van Vleet & Williams 1980, Agogue et al. 2004), and there are benefits and drawbacks
to each technique (Liss & Duce 2005, Stolle et al. 2009). For this study, we used a
nylon mesh screen (as described in (Williams et al. 1986) and a rotating Teflon coated
drum sampler (described in (Hardy et al. 1987). The screen exploits the surface tension
to hold the SML between the spaces in the nylon mesh when lifted slowly parallel to the
ocean surface. The thickness of the layer sampled ranges from 100-400µm (Garrett
1965, Van Vleet & Williams 1980). The Teflon drum rotates in the reverse direction of
its heading and relies on surface tension and to draw a thin layer of water up to a Teflon
squeegee linked to a funnel for collection, and samples a thinner section of the
microlayer: 30-100µm (Hardy et al. 1987, Frew et al. 2002).

In this study our goal was to determine the composition of the bacterial assemblage in
the SML and compare it to the composition of the bacterial assemblage in the
underlying water (taken at 500 cm depth); we did this offshore of Los Angeles and San
136

Diego and nearshore at Santa Catalina Island, California (Figure 4-1). We also
compared bacterial abundance, viral abundance, and heterotrophic bacterial production
through Leucine incorporation between the bacteria in the SML and ULW. We
collected the bacterioneuston using a nylon screen (NS) in May 2004 and a Teflon-
coated rotating drum (RD, Figure 4-2) in May 2005 and October 2006. We found a
depletion of bacterial production in the SML compared to ULW, while bacterial and
viral enrichment and depletion were variable from site to site. While there were distinct
SML communities detected in 2004 and 2005, they tended to share taxa which differed
in abundance, rather than contain an entirely unique set of bacteria. This suggests that
the bacteria which make up the SML community in Southern California waters are
transported from the surface waters rather than an outside source, and the SML is an
interface that interacts with the surface waters and in which the bacteria encounter
different selective factors.

San Diego
Catalina Island
San Pedro Channel
Figure 4-1 Map of sampling locations in the Southern California Bight. Samples were
taken in the San Pedro Channel in May of 2004 and 2005 , at Catalina Island in
May 2004, May 2005, and October 2006, and off the Coast of San Diego in May
2005.

137

C
B
A

Figure 4-2. Photographs of the rotating drum sampler. Pictures show the sampler
attached to a dingy (A), being hoisted out of into the water in 2005 (B), and
attached to a modified catamaran sampling platform in 2006 (C).

138

139

Results And Discussion

Viral abundance and Bacterial Abundance and Productivity
Virus abundance did not show any consistent difference in the SML compared to the
underlying water: the average enrichment factor (EF) was 1.04 with a range of 0.7-1.3
(Figure 4-3 A, E) across all samples. Viruses were slightly more enriched in 2004
(average EF of 1.1), compared to 2005 (average EF of 0.94). The virus to bacteria ratio
(VBR) was slightly lower in the SML averaged across all sites (18 ±8) compared to the
ULW (22 ±7), although their values frequently overlapped. These VBR numbers are
within the range values reported in the Mediterranean (Joux et al. 2006) SML but about
10-fold higher than those reported for a transect in the Atlantic (Kuznetsova et al.
2004). In both studies, there was a similar, but slightly lower VBR for the SML
compared to the VBR for the 0.5m depth.

The similar VBR suggests that the SML is not a refuge from viral predation for bacteria.
However, the VBR may not be an accurate indication of viral infection rates (Fuhrman
1999). UV and visible light radiation can cause a decrease in virus abundance and
activity (Suttle & Chen 1992, Noble & Fuhrman 1997, Wilhelm et al. 1998) although
UV can increase lysis and activate lysogeny (Weinbauer et al. 2003, Weinbauer 2004).

140

A E

B F

G
C

D H

Sites
Sites

Figure 4-3 Total viruses and bacteria, bacterial production and richness in the SML and
the corresponding ULW, along with the SML:ULW enrichment factors. Shown
are Virus Counts (A), Bacteria Counts (B), Bacterial Production (C), and
ARISA richness (D) in the SML and at 0.5m depth and their enrichment or
depletion of these parameters in the SML (E-H). For the enrichments, values <1
show depletion and values >1 show enrichment. Sites are the same as in Table 1.

141

UV resistance has been shown in bacteria from the SML (Agogue et al. 2005b) and this
could serve as an ecological niche with reduced predation by viruses. The low bacterial
production values, however, suggest that perhaps the mechanisms of bacterial hosts and
thus virus production are inhibited by the UV radiation.

Bacterial abundance was not enriched in the SML in the majority of samples (Figure 4-
3 B, F). The average enrichment factor for bacterial abundance was 1.9 and ranged
from 0.4-3.5. Samples 3, 11 and 13, collected at Catalina Island in May 2004 and
October 2006 (Table 4-1), were the most enriched in bacterial abundance (Figure 4-3E).
Even the highest EF of 3.5 was at the low end of the range of 2-200 fold enrichment
reported in earlier studies (Williams et al. 1986, Carlucci et al. 1991). Bacterial
abundance values in the ULW showed little variation with geographic location within
each study, but did show a difference from year to year (Table 4-2). The abundance
values in the SML varied with location (i.e. are patchier), and showed higher abundance
near Catalina Island (Table 4-2).

Bacterial production was depleted in the SML in all samples measured and ranged from
0.03to 0.9 with an average of 0.25 (Figure 4-3 C, D). Production rates ranged from
0.31-9.41 x 10
5
cells ml
-1
d
-1
in the SML and from 1.26- 46.3 cells ml
-1
d
-1
in the ULW.
These rates were uncoupled from bacterial abundance (Table 4-2) and because bacterial
abundance did not differ greatly between the SML and ULW samples (Figure 4-3),
bacteria in the microlayer were growing at a much slower rate. The slowest growth rate

142

Table 4-1 Site number, sampling type, date, location, and measured environmental parameters
Site No
Sampling
Method Date
Latitude
(° N)
Longitude
(°W)
Sea Surface
Temperature
(°C)
Salinity
(ppt)
Wind
speed
(m/s)
-12hr
Wind
Speed
(m/s)
1 N.S. 5/18/2004 33.44 118.47 18.6 33.64 2.67 1.26
2 N.S. 5/20/2004 33.44 118.47 18.7 33.27 4.71 4.23
3 N.S. 5/21/2004 33.44 118.47 18.5 33.29 4.67 6.81
4 N.S. 5/21/2004 33.55 118.40 18.3 33.39 2.99 6.54
5 N.S. 5/25/2004 33.44 118.47 18.0 33.30 5.38 5.30
6 N.S. 5/26/2004 33.55 118.40 18.8 33.31 3.40 5.38
7 T.D 5/9/2005 33.55 118.40 16.4 33.37 5.05 6.00
8 T.D. 5/9/2005 33.55 118.40 16.5 33.35 4.33 6.00
9 T.D. 5/12/2005 32.76 117.65 16.7 33.27 2.86 4.56
10 T.D. 10/28/2006 33.44 118.47 18.1 Nd 1.80 5.41
11 T.D. 10/28/2006 33.44 118.47 18.1 Nd 1.80 5.41
12 T.D. 10/29/2006 33.44 118.47 18.5 Nd 0 5.36
13 T.D. 10/29/2006 33.44 118.47 18.5 Nd 2.34 5.36

Table 4-2 Location, depth and biological parameters measured at the sampling sites. ARISA richness is the number of unique
taxa identified by ARISA length. Bray-Curtis and Sorenson similarities are pairwise similarities between the SML and the 0.5m
depth communities. Biological parameters are reported as averages ± standard deviation.
Site
No Date
Latitude
(° N)
Longitude
(°W) Depth
ARISA
Rich-
ness
Bray-Curtis
Similarity
(%)
Sorenson
Similarity
(%)
Bacterial
Production
(x10
5
cells
ml
-1
d
-1
)
Bacteria
Abundance
(x10
6
cells
ml
-1
)

Virus
Abundance
(x10
7
cells
ml
-1
)
1 5/18/2004 33.44 118.47 SML 51 51.2 71.2 1.43 ± 0.12 3.82 ±0.07
0.5m 50 1.55 ± 0.06 3.97 ±0.26
2 5/20/2004 33.44 118.47 SML 46 59.9 64.7 0.46 ±0.08 1.64 ± 0.16 3.89 ±0.07
0.5m 56 12.4 ±4.00 1.52 ± 0.09 2.97 ±0.24
3 5/21/2004 33.44 118.47 SML 51 62.4 74.2 0.31 ±0.10 4.30 ± 0.93 3.69 ±0.34
0.5m 38 21 ±1.38 1.21 ± 0.13 3.77 ±0.05
4 5/21/2004 33.55 118.40 SML 64 73.2 66.1 9.41 ±1.05 1.70 ± 0.18 4.40 ±0.04
0.5m 45 46.3 ±2.80 1.53 ± 0.06 4.26 ±0.04
5 5/25/2004 33.44 118.47 SML 32 73.7 63.4 0.33 ±0.02 1.36 ± 0.03 3.73 ±0.25
0.5m 50 6.62 ±3.62 1.46 ± 0.11 4.12 ±0.16
6 5/26/2004 33.55 118.40 SML 29 71.6 70.1 1.42 ± 0.13 4.07 ±0.06
0.5m 48 1.38 ± 0.03 3.40 ±0.08
7 5/9/2005 33.55 118.40 SML 59 49.1 56.5 4.55 ±0.04 3.31 ± 0.17 3.54 ±0.26
0.5m 57 12. ±0.14 3.66 ± 0.19 4.65 ±0.90
8 5/9/2005 33.55 118.40 SML 76 53.3 51.8 1.33 ±0.23 3.73 ± 0.37 3.50 ±0.50
0.5m 65 4.3 ±0.12 3.52 ± 0.72 3.82 ±0.63
9 5/12/2005 32.76 117.65 SML 76 75.6 70.1 1.15 ±0.06 1.72 ± 0.09 3.99 ±0.20
0.5m 66 1.26 ±0.01 3.61 ± 0.46 4.21 ±0.12

143

144

Table 4-2 (Cont’d)

Site
No Date
Latitude
(° N)
Longitude
(°W) Depth
ARISA
Rich-
ness
Bray-
Curtis
Similarity
(%)
Sorenson
Similarity
(%)
Bacterial
Production
(x10
5
cells
ml
-1
d
-1
)
Bacteria
Abundance
(x10
6
cells
ml
-1
)

Virus
Abundance
(x10
7
cells
ml
-1
)
10 10/28/2006 33.44 118.47 SML 49 88.7 84.8 2.38 ± 0.16 4.36 ±0.26
0.5m 43 2.05 ± 0.08 4.74 ±0.17
11 10/28/2006 33.44 118.47 SML 48 90.7 82.1 2.37 ± 0.12 3.59 ±0.24

0.5m 47 1.84 ± 0.07 4.91 ±0.11
12 10/29/2006 33.44 118.47 SML 58 90.8 80.7 4.46 ± 0.17 4.06 ±0.08
0.5m 51 2.16 ± 0.09 4.96 ±0.16
13 10/29/2006 33.44 118.47 SML 53 80.5 81.5 3.68 ± 0.04 4.91 ±0.05

0.5m 52 2.29 ± 0.09 5.10 ±0.51

145

was 0.006 d
-1
in the SML compared to 1.7 d
-1
in the ULW in sample 3 near Catalina
Island while sample 9 had a growth rate in the SML (0.07 d
-1
) nearly double that of
those in the surface waters (0.035 d
-1
). The bacterial production rates (Figure 4-3,
Table 4-2) shown are within the recent ranges reported in southern California surface
waters (Fuhrman et al. 2006) and, lower leucine incorporation rates have been shown in
microlayer communities (Obernosterer et al. 2008, Reinthaler et al. 2008, Stolle et al.
2009), especially when the ULW is highly productive (Carlucci et al. 1986).

Recently there have been many reports of no enrichment in bacterial abundance or
production in the SML (Joux et al. 2006, Reinthaler et al. 2008, Stolle et al. 2009) and
in some cases, there has been an inhibition of bacterial production reported for screen
samplers (Stolle et al. 2009). If this is the case here, the rates in samples 3-6 may be
underestimated. It is also worth remembering that bacterial production measurements
are a bulk average rate (Luna et al. 2004), and all members of the bacterial community
may not be acting in concert. Furthermore, collecting the SML with Teflon drum or
nitex screen and placing these samples into bottles may introduce amino acids to the
which would cause an inhibition in the growth rates measured by amino acid
incorporaton.

Differences in bacterial abundance and growth rates in the SML compared to the ULW
must be caused by differences in top-down control (predation by grazers or viruses)
(Thingstad & Lignell 1997), bottom up control (differences in available substrates)
146

(Torsvik et al. 2002), or sideways control (competition with other bacteria) (Long &
Azam 2001, Fuhrman & Hagström 2008). The slower growth rate may be caused from
the bottom-up by a difference in available food sources, or possibly through the
energetic cost of living in the surface microlayer environment. For instance, adaptation
to a high UV environment (shown by Agogue et al. 2005b) may cost the bacteria energy
it could otherwise use for growth. Given this slower growth rate, there should be a
pronounced decrease in the number of bacteria in the SML if the predation rate was the
same as the ULW. Since, this was not the case (i.e. there was little to no difference in
abundance) it is implied that there is reduced pressure from predation or competition or
both. However, it is not possible to determine whether the bacteria in the microlayer
are endemic or deposited from another source by counts and production alone.

Free-living Bacterial Community Composition of the SML and ULW
There are at least three ways that the bacterial communities in the SML would relate to
the bacteria from the ULW. They could have been brought to the microlayer, either
through mixing and bubble-scavenging from the water below (Bezdek & Carlucci 1972,
1974) or through deposition from the atmosphere (Norkrans 1980), or they could be
endemic to the microlayer. Once in the microlayer, the community may vary from the
community in the underlying water due to differential growth or losses. If the
communities are similar to the ULW, there is a good chance that they are arriving from
the water below. Differences in communities can stem from a host of factors, such as
increased or decreased viral predation or predation by eukaryotes in the SML compared
147

to the ULW or differential growth rates due to the UV inhibition or food sources. We
investigated the composition of SML and the ULW bacterial communities by ARISA
and putatively identified the ARISA peaks using 16S-ITS clone libraries (Fisher &
Triplett 1999, Brown et al. 2005, Fuhrman et al. 2006). We then compared whole
communities using the Bray-Curtis index, which compares relative abundance of taxa,
and the Sorenson index, which compares the presence or absence of taxa ignoring
abundance (Legendre & Legendre 1998).

Cluster analysis of the Bray-Curtis similarities for the 2004 bacterial communities
shows that SML communities did not most closely resemble their respective ULW
communities (Figure 4-4). The SML communities showed no consistent spatial pattern
but there seemed to be a broad temporal pattern (e.g. communities from samples 1 & 3
clustered together and from samples 4-6 clustered together, Figure 4-4A). SML
communities showed no clear spatial or temporal pattern (Figure 4-4A). The
communities from samples 2 & 3 clustered together, but other SML communities
showed as much difference from one another as from the ULW communities (Figure 4-
4A). Although some of the communities clustered differently by the Sorenson index,
the patterns were roughly the same (Figure 4-4B). For the ULW communities, there
was no clear spatial pattern and the temporal pattern was weaker (Figure 4-4B).

1
148

4

2
1

Figure 4-4 Dendrograms showing whole community similarities for free living bacteria
from the SML and ULW in 2004. Similarities calculated using Bray-Curtis (A)
and Sorenson (B) indices. Lines are drawn to connect SML communities to their
ULW counterparts. Boxes are drawn around offshore (San Pedro Channel) sites.
Red clusters are statistically identical and black clusters are distinct by
SIMPROF permutation test.

6
3
2
3
5
6
4
5
A
depth grp
sml
0.5m
B
5
1
4
2
6
6
3
4
5
1
2
3
100 90 80 70 60 50
Similarity
149

For the SML communities, the patterns were similar as well. The only major differences
were the SML community in sample 5 clustering out at 58% similarity to all other
samples, and the SML communities in samples 1 and 4 showing slightly greater
similarity to one another (72%) compared to the other samples (Figure 4-4B).

These community similarities provide evidence for a patchy community in the SML that
sometimes changes at a different pace than the ULW. One intriguing cluster that
occurred in both Bray-Curtis and Sorenson index was the high similarity between the
SML community from sample 6 and the ULW community from sample 3 (fig ure 4-4).
The SML sample came from the San Pedro Channel and the ULW sample came from
Catalina and they were separated by 5 days (Table 4-1). This likely stems from mixing
between the surface waters near Catalina and in the San Pedro Channel (Dong &
McWilliams 2007) and provides evidence for interactions between the microlayer
communities and the surface water communities. This pattern of distinct communities in
2004 samples was confirmed with a permutation-based MANOVA procedure
(PERMANOVA, (Anderson 2001, Lear et al. 2008), found significant difference
between the 2004 SML and ULW communities (Table 4-3).

Cluster analysis for the 2005 and 2006 community similarities showed much stronger
spatial and temporal patterns compared to 2004 (Figure 4-5) with all SML communities
clustering with their geographic location. At the 2006 Catalina site (samples 9-13) the
SML and ULW communities were highly similar (Figure 4-5A) and using the Sorenson

Table 4-3 Results from PERMANOVA test. Permutation based multivariate ANOVA of difference between
ULW (0.5m) bacterioplankton community and SML community sampled with nylon screen and with
rotating drum.

Sampling Type Source Df SS MS Pseudo-F P(perm) perms P(MC)
Nylon Screen De 2 1741.9 870.95 1.7526 0.043 999 0.086
Res 14 6957.3 496.95
Total 16 8699.2
Rotating Drum De 1 690.09 690.09 0.31098 0.77 767 0.78
Res 12 26629 2219.1
Total 13 27319

150

151

Figure 4- 5 Dendrograms showing whole community similarities between SML and
ULW communities in May 2005 and October 2006. Bray-Curtis similarity(A)
and Sorenson similarity (B) from ARISA for SML and ULW (0.5m)
communities from May 2005 (samples 7-9) and October 2006 (samples 10-13).
Geographical groups are indicated by lines and labeled SP for San Pedro
Channel, SD for San Diego, and CI for Catalina Island sites.
12
13
11
13
10
12
10
11
9
9
8
7
7
8
100 90807060 5040 302010 0
Similarity
8
7
7
8
13
13
11
10
12
10
11
12
SP
A
CI
SD
9
9
depth grp
sml
0.5m
CI
SD
B
SP
152

index, the communities were statistically indistinguishable (Figure 4-5B). The 2005
samples showed distinct SML communities at the San Pedro Channel location in the
Bray-Curtis cluster, but this relationship was weaker by Sorenson clustering. There was
high temporal variation in the San Pedro channel as well with the ULW assemblages
showing as little as 55% Bray-Curtis similarity to each other, although these temporal
differences were weaker when the Sorenson index was used (Figure 4-5). In general,
the communities showed higher Sorenson similarity indicating they had a greater
number of shared taxa, and the community differences detected were driven by relative
abundances.

Although the San Pedro Channel SML communities in 2005 are distinct from the
corresponding ULW communities, the PERMANOVA procedure was unable to detect
differences between the SML and ULW communities in 2005 and 2006 (Table 4-3).
Clearly, the variation in ULW and SML communities over space and time was so high
that it masked the distinction in the SML communities. This high spatial and temporal
variation in 2005 and much lower variation in 2006, particularly in the 0.5m ULW may
be due to seasonal variation in the Southern California Current system, which changes
from a colder, southward flow to a warmer, northward flow in early Summer and then
strengthens throughout the Summer and Fall (Di Lorenzo 2003).

153

Sampling method may have played a role as well, although comparisons made between
the RD and screen samplers have not shown great differences in community
composition when tested in the field (Agogue et al. 2004) or in the laboratory (Van
Vleet & Williams 1980, Stolle et al. 2009). The samples in 2005 and 2006 were taken
with a rotating drum sampler that was pushed forward as it sampled. This would have
the effect of integrating the SML over a greater distance, compared to the nylon screen
method used in 2004, which is more passive and samples a much smaller area at one
time.

Comparison of individual taxa between SML and ULW
ARISA peaks were identified from the USC Microbial Observatory clone library
database and plotted against location in order to determine differences between the taxa
in the SML and the ULW communities (Figures 4-6, 4-7). In all samples, the most
abundant taxa were shared between the SML and ULW communities (Figures 4-6, 4-7).
In both nearshore and offshore waters in May 2004, SAR11 Surface group 2 688,
SAR11 Surface group 1 666, 668, 663, and 681 had the highest abundance in both the
SML and ULW depths (Figure 4-6). In the 2005 San Pedro Channel samples
Roseobacter 1189, OTU 808, and Synechococcus 1132 were the most abundant in both
SML and ULW communities (Figure 4-7A) while in San Diego Synechococcus 1054
and 1038 , Cytophogales 621, Prochlorococcus 830, and SAR11-S1 681 were abundant
in both SML and ULW communities (Figure 4-7B).

154

Figure 4-6 ARISA-derived rank-abundance of free-living bacterial OTUs from 2004.
Bacterial OTUs sampled from the SML (A) and the 0.5m ULW (B) at Catalina
Island (sites 1-3, 5) and the San Pedro Channel (sites 4 & 6).

ARISA Abundance (%)
0
5
10
15
20
25
123 546
A
15
20
25
B
0
5
10
123546
SAR11‐S2 688 SAR11‐S1 666 SAR11‐S1 668 SAR11‐S1 663 SAR11‐S1 681 OTU 756 OTU 877
OTU 742 Cytoph 727 OTU 618 Synecho 1054 Flavo 856 OTU 575 OTU 623
OTU 572 Prochloro 823 OTU 579 Altero‐4 551 Plast 562

155

Figure 4-7 ARISA-derived rank-abundance of free living bacterial OTUs from 2005
and 2006. Samples of communities from the SML and the 0.5m ULW taken
from the San Pedro Channel in 2005 (A), San Diego in 2005 (B), and Catalina
Island in 2006 (C).

0
5
10
15
20
25
9 SML 9 ULW
Synecho 1054 Cytoph 621
Synecho 1038 Prochloro 830
SAR11‐S1 681 SAR11‐S1 663
SAR11 704 SAR11‐S2 688
Sphingo 727 SAR11‐S1 666
Roseo 1189 β‐Proteo 843
SAR11 S1 670 OTU 618
SAR92 750 OTU 808
0
5
10
15
20
25
30
35
7 SML 8 SML 7 ULW 8 ULW
Roseo 1189 OTU 808 Synecho 1132 OTU 595 SAR11‐S2 688 Synecho 1054
Synecho 1038 Prochloro 823 Bacter 613 SAR92 761 OTU 869 Cytoph 739
Cytoph 969 Cytoph 752 γ‐Proteo 921 SAR11 684 OTU 802 OTU 479
Synecho 1067 SAR11 699 Sphingo 641 OTU 632
A
B
ARISA Abundance (%)
0
5
10
15
20
25
30
35
10 SML 11 SML 12 SML 13 SML 10 ULW 11 ULW 12 ULW 13 ULW
SAR11‐S1 666 Actino 419 Prochloro 830 CHABI‐7 402 Actino 422
SAR11‐S2 688 OTU 406 Actino 434 Synecho 1054 Cytoph 727
Actino 425 SAR11‐S1 681 Cytoph/Plastid 538 SAR11 677
C
156

The SML and ULW communities from Catalina in October 2006 shared nearly all taxa,
and SAR11-Surface 1 666, Actinobacterium 419, Prochloroccocus 830, and CHABI-7
402 were the most abundant (Figure 4-7C).

However, there were some clear differences in the taxa found in the SML communities
in May 2004 and May 2005. The SML community in sample 1 showed the greatest
difference from the ULW with much lower abundance of SAR11 bacteria and high
abundance of Alteromonas-like group 4 551, unidentified OTUs 575, 579, and 572,
Plastid 562 and Prochlorococcus 823(Figure 4-6). In 2005, the SML communities in the
San Pedro channel had much higher ARISA abundances of Cytophogales 969 and 752,
and Sphingobacter 641, and did not have OTU 595 (Figure 4-7A). The San Diego SML
community (sample 9) had higher ARISA abundances of unknown OTU 808 and
Sphingobacter 641 (Figure 4-7B).

It is interesting to note the sharp differences between May 2004 and May 2005. May
2004 was much warmer water (Table 4-1) dominated by SAR11 taxa from Surface
groups 1 and 2, while in May 2005, there was clearly a bloom of Synechococcus and
Roseobacter occurring in the surface waters in the San Pedro Channel. This is the
likely the driver for the high degree of spatial variation shown between the ULW in
samples 7 and 8 (Figure 4-5). The San Diego communities in May 2005 were also
dominated by Synechococcus but was a more even community containing
Cytophogales, Prochlorococcus, SAR11Surface-1 bacteria. We also note that other
157

samples taken in May 2005 nearby at 5m depth, did not show as high an abundance of
Synechococcus and Roseobacter (Chapter 2), which shows an interesting distinction
between 0.5m and 5m depth.

Relationship of SML and ULW community to environmental variables
The relationship of bacterial community patterns to environmental parameters should
provide insight into the environmental mechanisms driving these patterns (Prosser et al.
2007, Fuhrman & Steele 2008, Fuhrman 2009). We compared community similarity to
geographic distance and environmental similarity using the rank correlation Mantel-
type tests from the BIO-ENV and RELATE routines (Clarke 1993, Clarke & Gorley
2006). We also examined univariate correlations between SML vs ULW community
similarity and environmental and biological parameters (e.g. ARISA richness, Bray-
Curtis and Sorenson similarities between the SML and ULW communities, and
enrichment values, Table 4-4).

SML community similarity across all samples was correlated to temperature and virus
abundance (r=0.63, p=0.002). Temperature was also negatively correlated with ARISA
richness from both the SML and ULW communities, and, not surprisingly, it also
correlated to the Sorenson similarity values calculated by the ARISA richness (Table 4-
4).

158
Table 4-4. Correlations between environmental variables, similarity indices, and biological parameters. (n), *p<0.05, **p<0.01 ***p<0.001.
Lat-
itude
Long-
itude
Temp-
erature Salinity
Wind
speed
-12hr
Wind
Speed

SML
Rich-
ness
0.5m
Rich-
ness
Bray-
Curtis
Similarity
Sorenson
Similarity
Bacteria
E-rich-
ment
Virus
Enrich-
ment
Latitude (13)
Longitude (13) -0.92*
Temperature
(13) 0.31 -0.51
Salinity (9) 0.23 -0.27 0.16
Windspeed
(13) 0.14 -0.02 -0.32 -0.50
-12hr Wind-
speed (13) 0.23 -0.08 -0.22 -0.69* 0.17
SML Richness
(13) -0.40 0.52 -0.66* 0.06 -0.15 0.12
ULW Richness
(13) -0.45 0.59* -0.70* -0.13 0.17 -0.20 0.57*
Bray-Curtis
Similarity (13) -0.20 0.03 0.35 -0.49 -0.72** 0.23 -0.17 -0.32
Sorenson
Similarity (13) -0.15 -0.12 0.59* 0.10 -0.76** -0.09 -0.26 -0.56* 0.79**
Bacteria
Enrichment
(13) 0.24 -0.37 0.33 -0.15 0.02 0.42 -0.15 -0.65* 0.11 0.36
Virus
Enrichment
(13) -0.02 -0.01 0.42 -0.28 0.37 -0.23 -0.41 -0.01 -0.27 -0.10 -0.01
Bacterial
Production
Enrichment (7) -0.84* 0.95*** -0.65 -0.11 -0.60 -0.37 0.64 0.64 0.25 0.09 -0.48 -0.30

159

The correlation of the SML communities to temperature and viruses, and the negative
correlation of temperature to richness, suggests that the SML community may be
subject to both top-down control from viruses (Thingstad & Lignell 1997), and bottom-
up control (i.e. resource availability, since temperature can serve as a marker for more
oligotrophic waters in Southern California (Di Lorenzo 2003, Fuhrman et al. 2006)).

ULW community similarities correlated to location (i.e. latitude-longitude) and
bacterial abundance (r=0.78, p=0.001). In the univariate correlations, bacterial
production enrichment was also correlated to location. Some of this relationship may
be due to this was due to the samples collected by rotating drum sample vs the nylon
screen (discussed above, and Agogue et al. 2004, Stolle et al. 2009). The relationship is
also driven by the much higher production measurements at the San Diego location in
2005. While a causal relationship is unclear in this case, both of these correlations
imply a combination of influences on community structure such as interactions between
bacteria (Long & Azam 2001) and resource availability and quality (Torsvik et al.
2002).

There was a surprisingly strong negative correlation SML-ULW similarity values to
wind, i.e. as windspeed increased, the similarity values decreased (table 4). This
relationship was strongly, though not entirely, driven by the high similarity between the
2006 SML and ULW samples. These communities showed high similarity (i.e. no
distinct microlayer community) even at low windspeed, however, the distinct
160

communities in the 2005 San Pedro Channel also contributed to the relationship. One
would expect that increased wind energy would disperse any microlayer that may have
formed, although it would re-form once the wind died down (Hardy 1982, Liss & Duce
1997). However, (Carlson 1983) reported collecting SML on a glass plate in winds up
to 8 m s
-1
(Beaufort Force 5) and suggested that the mixing and turbulence caused by
higher winds may increase transport to the microlayer.

Overall these data do not support a distinct SML community, at least in terms of unique
taxa. However, they do support the theory that bacteria are transported to the microlayer
from the surface waters (Bezdek & Carlucci 1972), as was earlier shown to happen after
blooms of Trichodesmium (Sieburth & Conover 1965). This does allow for a distinct
community in terms of relative abundance, (e.g. SML sample 1 (Figure 4-6) or SML
sample 7 & 8 (Figure 4-7A)). These results markedly differ from the North Sea SML
clone library by which was completely different from the ULW and almost entirely
dominated by Vibrio spp. (68%) and Pseudoalteromonas (21%) related species
(Franklin et al. 2005) and distinct samples found in estuaries and near Hawaii (Cunliffe
et al. 2008, Cunliffe & Murrell 2009). The lack of a distinct microlayer community is
not unprecedented however, and (Agogue et al. 2005a) reported patchiness with time
and space in the Mediterranean. We also acknowledge that this study only considers
the 0.2-1.22 µm fraction of bacteria, excluding particle-attached bacteria. Some studies

161

have suggested that particulate associated bacteria should play a larger role in the SML
(Wurl & Holmes 2008, Cunliffe & Murrell 2009). Although we did not directly test for
it, we note that a high degree of particulate matter was not observed in the bacteria and
virus counts.

In summary, we found depletion of bacterial production in the SML compared to the
ULW, and variable enrichment of bacteria and viruses. Bacterial communities in the
SML were also highly variable over space and time, and did not represent a unique
community. However, we did detect communities which were distinct from the ULW
communities in the San Pedro Channel and Catalina in May 2004 and May 2005, one of
which occurred during a Synechococcus and Roseobacter bloom. Even the bacterial
communities in the SML which showed the greatest difference from the communities in
the ULW shared abundant taxa. This suggests that the bacteria which make up the SML
community in Southern California waters are transported from the surface waters and
that the community in the surface waters have a large impact on the community in the
microlayer.

162

Experimental Procedures

Sample Collection: Samples were collected nearshore in Big Fisherman’s Cove, near
Bird Rock, and from Ship Rock outward no more than 2-3 km from shore at Santa
Catalina Island in May 2004, May 2005, and October 2006. Offshore sampling has
been conducted at SPOTS over in May 2004 and 2005 and off of San Diego in May
2005. Samples were collected in May 2004 using an 5% HCl acid-washed and 70%
ethanol washed 40cm x 80cm nylon screen at 500-micron mesh with 47% open space
which has experimentally determined to collect between 100-400µm from the surface
(Garrett 1965, Van Vleet & Williams 1980). The sampling was performed by small
boat and the screen was lowered into the water vertically and raised through the surface
of the water horizontally. Water samples were collected at the same time at 0.5m depth
into a plastic bottle. In May 2005 and October 2006 samples were collected by a
rotating Teflon drum sampler (after (Hardy et al. 1988) which has been experimentally
determined to collect between 30-100µm from the surface (Hardy et al. 1988, Frew et
al. 2002). Prior to sample collection, the drum, the Teflon squeegee, and the collection
tubing was washed with 5% HCl and 80% ethanol. This sampler was attached to a
small boat in May 2005 and to a catamaran-style sampling platform October 2006. The
rotating drum was designed to turn using a water wheel and required the boat and
platform to push it forward at low speed. Water was collected at the same time at 0.5m
depth into an acid-washed plastic bottle. The collected water was sequentially filtered
through a A/E filter (Gelman, 1.2 µm nominal pore size, 47mm diameter) then a
163

Durapore filter (Millipore, 0.22 µm nominal pore size, 47mm diameter). The filters
were placed on dry ice or frozen at -80°C, transported to USC and stored at -80°C for
further analysis.
DNA Extraction and Amplification:- DNA was extracted from frozen filters following
(Fuhrman et al. 1988). Briefly, frozen filters were crushed, cells were lysed with hot
1% SDS STE buffer, and DNA was purified by phenol:chloroform extraction. DNA
was stored frozen at -80
o
C in TE buffer. Automated rRNA Intergenic Spacer Analysis
(ARISA)(Fisher & Triplett 1999, Hewson & Fuhrman 2004)) was conducted on 2.5 ng
DNA as measured by PICO Green fluorescence. A standard amount of template
genomic DNA was used in each PCR reaction, with the intention of analyzing the same
amount of bacteria from each sample. PCR reactions (50 μl) contained 1X PCR buffer,
2.5 mM MgCl
2
, 250μM of each deoxynucleotide, 200 nM each of universal primer 16S
– 1392F (5’-G[C/T]ACACACCGCCCGT-3’) and a TET labeled bacterial primer 23S –
125R (5’-TET-GGGTT[C/G/T]CCCCATTC(A/G)G-3’), 2.5U Taq polymerase
(Promega), and BSA (Sigma # A-7030; 40 ng/ μl final conc.). This primer set
specifically targets bacteria, and we know of no major group of marine bacteria in
surface waters whose DNA these primers fail to amplify (Brown et al. 2005, Hunt et al.
2006). Thermocycling was preceded by a 3 min heating step at 94
o
C, followed by 30
cycles of denature at 94
o
C for 30 s, anneal at 56
o
C for 30 s, extend at 72
o
C for 45s,
with a final extension step of 7 min at 72
o
C. Amplification products were cleaned using
Clean & Concentrator-5 (Zymo Research), and purified DNA was measured by PICO
Green fluorescence. Purified DNA diluted to a standardized amount (5 ng µl
-1
) was
164

loaded in the fragment analysis. Standardization prevented fragment differences arising
from different amounts of loaded DNA. Products were then run for 5.5 h on an ABI
377XL automated sequencer operating as a fragment analyzer (Avaniss-Aghajani et al.
1994) with a custom-made ROX-labeled 1500bp standards (Bioventures Inc.). The
sequencer electropherograms were then analyzed using ABI Genescan software.
Outputs from the ABI Genescan software were transferred to Microsoft Excel for
subsequent analysis. Peaks less than 5 times the baseline fluorescence intensity were
discarded since they were judged not clearly distinguishable from instrument noise
(Hewson & Fuhrman 2004). With this criterion, the practical detection limit for one
operational taxonomic unit (OTU) is ca. 0.09% of the total amplified DNA (Hewson &
Fuhrman 2004).
Bacterial production:-Bacterial production was estimated by incorporation of
3
H-
leucine into protein as described previously (Kirchman et al. 1985, Simon & Azam
1989). Briefly, triplicate 10 ml seawater samples in sterile, sample rinsed,
polypropylene centrifuge tubes had
3
H- leucine added to 5 nmol L
-1
final concentration.
The samples were then incubated 1 h at ambient seawater temperature in the dark.
Before
3
H- Leucine addition, one replicate sample was killed by adding 5% formalin.
After incubation, samples were filtered onto 25 mm-diameter 0.45 μm Millipore (Type
HA) nitrocellulose filters. Proteins were treated by incubation for 2 min with 2 ml ice-
cold trichloroacetic acid (TCA). TCA was then filtered through, the towers were rinsed
3 times with TCA onto the filters, the filters were then rinsed 3 times with TCA, then
placed into 6 ml scintillation vials containing 5 ml Ultima Gold scintillation fluid. After
165

incubating vials for at least 2 hours at room temperature to allow clarification of filter
membranes in scintillation fluid, radiolabel incorporation was measured in a Beckman-
Coulter LS6500 scintillation counter. We used a conversion factor of 1.5 x 10
17
cells/mol leucine incorporated (Simon & Azam 1989) to estimate bacterial production.
Bacterial and Viral abundance- Bacterial abundance was determined by SYBR Green I
staining and epifluorescence microscopy described in (Noble & Fuhrman 1998) detailed
protocol in (Patel et al. 2007). Briefly, samples (50 ml) were fixed with 0.02 μm-
filtered formaldehyde, to a final concentration of 2%, and kept at 4
o
C in the dark until
processing, which occurred within 24 h of sampling. Aliquots (2-5ml) of the samples
were filtered onto 0.02 μm Anodisc Al
2
O
3
filters, drying the filter on tissue paper, then
staining on a 100 μl, 1:2500-diluted drop of SYBR Green I. After staining, the filters
were re-dried on tissue paper and mounted on a glass slide with a solution of 50:50:0.01
Glycerol: Phosphate Buffered Saline: p-phenylenediamine as mountant. Slides were
observed under blue light excitation at 1,000 X magnification on an Olympus BH-60
microscope. More than 200 viruses and bacteria were counted in 10-20 fields.
Statistical analyses: ARISA peaks were standardized by the total area in each sample
in order to compare relative abundances of ARISA peaks. Prior to analysis, the peaks in
replicate samples were binned according to the dynamic binning protocol described in
(Ruan et al. 2006).
The ARISA peak areas were transformed by ln(x+1), to account for skew in the
distribution. For some analyses, the peak areas were transformed to presence-absence
values. Bacterial counts, virus counts, and bacterial production were also log
166

transformed. The environmental parameters were then normalized according to their
mean in order to remove differences in scale. Euclidean distance was calculated among
the environmental parameters in order for comparisons between similarity matrices.
Similarity matrices on the ARISA samples were created using the Sorenson Index of
Similarity (presence-absence data) and Bray-Curtis similarity (compares relative
abundances) (Legendre & Legendre 1998). These were compared through hierarchical
clustering using unweighted pair group method with arithmetic mean (UPGMA) and
these relationships were visualized through dendrograms using PRIMER (Clarke &
Gorley 2006). Tests for correlation between environmental parameters (e.g.
temperature, salinity, windspeed, bacterial production, bacterial counts and virus
counts) and bacterioplankton assemblage similarity values were carried out using
Mantel-type tests using the BIO-ENV routine in PRIMER v6. Correlations were also
calculated between the environmental parameters and the pairwise similarities of the
SML sites and the corresponding ULW. To test for a difference between communities,
the PERMANOVA routine, a permutation based multi-variate ANOVA was run in
PRIMER v6 (Anderson 2001, Anderson et al. 2008).

167

Chapter 4 References

Agogue H, Casamayor EO, Bourrain M, Obernosterer I, Joux F, Herndl GJ, Lebaron P
(2005a) A survey on bacteria inhabiting the sea surface microlayer of coastal
ecosystems. Fems Microbiology Ecology 54:269-280

Agogue H, Casamayor EO, Joux F, Obernosterer I, Dupuy C, Lantoine F, Catala P,
Weinbauer MG, Reinthaler T, Herndl GJ, Lebaron P (2004) Comparison of
samplers for the biological characterization of the sea surface microlayer.
Limnol Oceanogr-Meth 2:213-225

Agogue H, Joux F, Obernosterer I, Lebaron P (2005b) Resistance of marine
bacterioneuston to solar radiation. Applied and Environmental Microbiology
71:5282-5289

Anderson MJ (2001) A new method for non-parametric multivariate analysis of
variance. Austral Ecol 26:32-46.

Anderson MJ, Gorley RN, Clarke KR (2008) PERMANOVA+ for PRIMER: guide to
software and statistical methods. In. PRIMER-E, Plymouth, UK

Avaniss-Aghajani E, Jones K, Chapman D, Brunk C (1994) A Molecular Technique For
Identification of Bacteria Using Small Subunit Ribosomal Rna Sequences.
Biotechniques 17:144-149

Bezdek HF, Carlucci AF (1972) Surface Concentration of Marine Bacteria. Limnol
Oceanogr 17:566

Bezdek HF, Carlucci AF (1974) Concentration and Removal of Liquid Microlayers
from a Seawater Surface by Bursting Bubbles. Limnol Oceanogr 19:126-132

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show
marine microbial diversity: development and application to a time series.
Environ Microbiol 7:1466-1479

Carlson DJ (1983) Dissolved Organic Materials in Surface Microlayers - Temporal and
Spatial Variability and Relation to Sea State. Limnol Oceanogr 28:415-431

Carlucci AF, Craven DB, Henrichs SM (1985) Surface-Film Microheterotrophs -
Amino-Acid Metabolism and Solar-Radiation Effects on Their Activities.
Marine Biology 85:13-22

168

Carlucci AF, Craven DB, Robertson KJ, Henrichs SM (1986) Microheterotrophic
Utilization of Dissolved Free Amino-Acids in Depth Profiles of Southern-
California Borderland Basin Waters. Oceanol Acta 9:89-96

Carlucci AF, Craven DB, Wolgast DM (1991) Microbial-Populations in Surface-Films
and Subsurface Waters - Amino-Acid-Metabolism and Growth. Marine Biology
108:329-339

Carty C, Colwell RR (1975) A microbiological study of air and surface water
microlayers in the open ocean. J Wash Acad Sci 65:148–152

Clarke KR (1993) Non-parametric multivariate analyses of changes in community
structure. . Australian Journal of Ecology 18:117-143.

Clarke KR, Gorley RN (2006) PRIMER v6: User Manual/Tutorial

Cunliffe M, Harrison E, Salter M, Schafer H, Upstill-Goddard RC, Murrell JC (2009)
Comparison and validation of sampling strategies for the molecular microbial
analysis of surface microlayers. Aquat Microb Ecol 57:69-77

Cunliffe M, Murrell JC (2009) The sea-surface microlayer is a gelatinous biofilm. Isme
Journal 3:1001-1003

Cunliffe M, Schafer H, Harrison E, Cleave S, Upstill-Goddard R, Murrell JC (2008)
Phylogenetic and functional gene analysis of the bacterial and archaeal
communities associated with the surface microlayer of an estuary. Isme Journal
2:776-789

Di Lorenzo E (2003) Seasonal dynamics of the surface circulation in the Southern
California Current System. Deep Sea Research II 50:2371-2388

Dong CM, McWilliams JC (2007) A numerical study of island wakes in the Southern
California Bight. Cont Shelf Res 27:1233-1248

Fisher MM, Triplett EW (1999) Automated approach for ribosomal intergenic spacer
analysis of microbial diversity and its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636

Franklin MP, McDonald IR, Bourne DG, Owens NJP, Upstill-Goddard RC, Murrell JC
(2005) Bacterial diversity in the bacterioneuston (sea surface microlayer): the
bacterioneuston through the looking glass. Environmental Microbiology 7:723-
736

169

Frew NM, Nelson RK, Bock EJ, McGillis WR, Edson JB, Hara T (2002) Spatial
variations in surface microlayer surfactants and their role in modulating air–sea
exchange. In: Donelan MA, Drennan WM, Saltzman ES, Wannickhof R (eds)
Gas Transfer at Water Surfaces. AGU, , Washington, D.C., p 153–159

Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects.
Nature 399:541-548

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Comeau DE, Hagstrom A, Chan AM (1988) Extraction of DNA suitable
for molecular biological studies from natural planktonic microorganisms. Appl.
Environ. Microbiol. 54:1426-1429

Fuhrman JA, Hagström Å (2008) Bacterial and archaeal community structure and its
patterns. In: Kirchman DL (ed) Microbial Ecology of the Oceans. Wiley

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocean
conditions. P Natl Acad Sci USA 103:13104-13109

Fuhrman JA, Steele JA (2008) Community structure of marine bacterioplankton:
patterns, networks, and relationships to function. Aquat Microb Ecol 53:69-81

Garabetian F (1990) Co2 Production at the Sea-Air Interface - an Approach by the
Study of Respiratory Processes in Surface Microlayer. Int Rev Ges Hydrobio
75:219-229

Garrett WD (1965) Collection of slick-forming materials from the sea surface.
Limnology & Oceanography 10:602-605

Gasparovic B, Kozarac Z, Saliot A, Cosovic B, Mobius D (1998) Physicochemical
characterization of natural and ex-situ reconstructed sea-surface microlayers. J
Colloid Interf Sci 208:191-202

Hardy J, Kiesser S, Antrim L, Stubin A, Kocan R, Strand J (1987) The Sea-Surface
Microlayer of Puget-Sound .1. Toxic Effects on Fish Eggs and Larvae. Marine
Environmental Research 23:227-249

Hardy JT (1982) The Sea-Surface Microlayer - Biology, Chemistry and Anthropogenic
Enrichment. Prog Oceanogr 11:307-328

170

Hardy JT, Coley JA, Antrim LD, Kiesser SL (1988) A Hydrophobic Large-Volume
Sampler for Collecting Aquatic Surface Microlayers - Characterization and
Comparison with the Glass Plate Method. Can J Fish Aquat Sci 45:822-826

Hewson I, Fuhrman JA (2004) Richness and diversity of bacterioplankton species along
an estuarine gradient in Moreton Bay, Australia. Appl Environ Microb 70:3425-
3433

Hunt DE, Klepac-Ceraj V, Acinas SG, Gautier C, Bertilsson S, Polz MF (2006)
Evaluation of 23S rRNA PCR primers for use in phylogenetic studies of
bacterial diversity. Appl Environ Microb 72:2221-2225

Joux F, Agogue H, Obernosterer I, Dupuy C, Reinthaler T, Herndl GJ, Lebaron P
(2006) Microbial community structure in the sea surface microlayer at two
contrasting coastal sites in the northwestern Mediterranean Sea. Aquat Microb
Ecol 42:91-104

Kirchman D, Knees E, Hodson R (1985) Leucine Incorporation and Its Potential as a
Measure of Protein-Synthesis by Bacteria in Natural Aquatic Systems. Applied
and Environmental Microbiology 49:599-607

Kuznetsova M, Lee C, Aller J, Frew N (2004) Enrichment of amino acids in the sea
surface microlayer at coastal and open ocean sites in the North Atlantic Ocean.
Limnol Oceanogr 49:1605-1619

Lear G, Anderson MJ, Smith JP, Boxen K, Lewis GD (2008) Spatial and temporal
heterogeneity of the bacterial communities in stream epilithic biofilms. Fems
Microbiol Ecol 65:463-473

Legendre P, Legendre L (1998) Numerical Ecology. In: Developments in
Environmental Modelling 20. Elsevier, Amsterdam, p 853

Liss PS, Duce RA (eds) (1997) The Sea Surface and Global Change, Vol. Cambridge
University Press, Cambridge, UK

Liss PS, Duce RA (eds) (2005) The Sea Surface and Global Change, Vol. Cambridge
University Press, Cambridge, UK

Long RA, Azam F (2001) Antagonistic interactions among marine pelagic bacteria.
Appl Environ Microb 67:4975-4983

171

Luna GM, Dell'Anno A, Giuliano L, Danovaro R (2004) Bacterial diversity in deep
Mediterranean sediments: relationship with the active bacterial fraction and
substrate availability. Environ Microbiol 6:745-753

Noble RT, Fuhrman JA (1997) Virus decay and its causes in coastal waters. Appl.
Environ. Microbiol. 63:77-83

Noble RT, Fuhrman JA (1998) Use of SYBR Green I rapid epifluoresence counts of
marine viruses and bacteria. Aquatic Microbial Ecology 14:113-118

Norkrans B (1980) Surface microlayers in aquatic environments. In: Alexander M (ed)
Advances in microbial ecology. Plenum Press, New York and London, p 51–83

Obernosterer I, Catala P, Lami R, Caparros J, Ras J, Bricaud A, Dupuy C, van
Wambeke F, Lebaron P (2008) Biochemical characteristics and bacterial
community structure of the sea surface microlayer in the South Pacific Ocean.
Biogeosciences 5:693-705

Obernosterer I, Catala P, Reinthaler T, Herndl GJ, Lebaron P (2005) Enhanced
heterotrophic activity in the surface microlayer of the Mediterranean Sea. Aquat
Microb Ecol 39:293-302

Patel A, Noble RT, Steele JA, Schwalbach MS, Hewson I, Fuhrman JA (2007) Virus
and prokaryote enumeration from planktonic aquatic environments by
epifluorescence microscopy with SYBR Green I. Nat Protoc 2:269-276

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green
JL, Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ,
Young JPW (2007) The role of ecological theory in microbial ecology. Nat Rev
Micro 5:384-392

Regan JD, Carrier WL, Gucinski H, Olla BL, Yoshida H, Fujimura RK, Wicklund RI
(1992) DNA as a solar dosimeter in the ocean. Photochem Photobiol 56:35-42

Reinthaler T, Sintes E, Herndl GJ (2008) Dissolved organic matter and bacterial
production and respiration in the sea-surface microlayer of the open Atlantic and
the western Mediterranean Sea. Limnol Oceanogr 53:122-136

Ruan Q, Steele JA, Schwalbach MS, Fuhrman JA, Sun FZ (2006) A dynamic
programming algorithm for binning microbial community profiles.
Bioinformatics 22:1508-1514

172

Sieburth JM (1971) Distribution and activity of oceanic bacteria. Deep Sea Res
18:1111-1121

Sieburth JM (1983) Microbiological and organic-chemical processes in the surface and
mixed layers. In: Liss PS, Slinn WGN (eds) Air-Sea Exchange of Gases and
Particles. Reidel Publishers Co, Hingham, MA, p 121–172

Sieburth JM, Conover JT (1965) Slicks Associated with Trichodesmium Blooms in
Sargasso Sea. Nature 205:830-&

Simon M, Azam F (1989) Protein content and protein synthesis rates of planktonic
marine bacteria. Marine Ecology Progress Series 51:201-213

Stolle C, Nagel K, Labrenz M, Jürgens K (2009) Bacterial activity in the sea-surface
microlayer: in situ investigations in the Baltic Sea and the influence of sampling
devices. Aquat Microb Ecol 58:67–78

Suttle CA, Chen F (1992) Mechanisms and rates of decay of marine viruses in seawater.
Appl. Environ. Microbiol. 58:3721-3729

Thingstad TF, Lignell R (1997) Theoretical models for the control of bacterial growth
rate, abundance, diversity and carbon demand. Aquat. Microb. Ecol. 13:19-27

Torsvik V, Ovreas L, Thingstad TF (2002) Prokaryotic diversity - Magnitude,
dynamics, and controlling factors. Science 296:1064-1066

Upstill-Goddard RC, Frost T, Henry GR, Franklin M, Murrell JC, Owens NJP (2003)
Bacterioneuston control of air-water methane exchange determined with a
laboratory gas exchange tank. Global Biogeochemical Cycles 17:-

Van Vleet ES, Williams PM (1980) Sampling sea surface films: A laboratory evaluation
of techniques and collecting materials. Limnology & Oceanography 25:764-770

Weinbauer MG (2004) Ecology of prokaryotic viruses. Fems Microbiol Rev 28:127-
181

Weinbauer MG, Brettar I, Hofle MG (2003) Lysogeny and virus-induced mortality of
bacterioplankton in surface, deep, and anoxic marine waters. Limnol Oceanogr
48:1457-1465

Wilhelm SW, Weinbauer MG, Suttle CA, Jeffrey WH (1998) The role of sunlight in the
removal and repair of viruses in the sea. Limnol. Oceanogr. 43:586-592

173

Williams PM, Carlucci AF, Henrichs SM, Vanvleet ES, Horrigan SG, Reid FMH,
Robertson KJ (1986) Chemical and Microbiological Studies of Sea-Surface
Films in the Southern Gulf of California and Off the West-Coast of Baja-
California. Marine Chemistry 19:17-98

Wurl O, Holmes M (2008) The gelatinous nature of the sea-surface microlayer. Marine
Chemistry 110:89-97

Yang GP, Tsunogai S, Watanabe S (2005) Biogeochemistry of
dimethylsulfoniopropionate (DMSP) in the surface microlayer and subsurface
seawater of Funka Bay, Japan. Journal of Oceanography 61:69-78

Yang GP, Watanabe S, Tsunogai S (2001) Distribution and cycling of dimethylsulfide
in surface microlayer and subsurface seawater. Marine Chemistry 76:137-153

Zhengbin Z, Liansheng L, Zhijian W, Jun L, Haibing D (1998 ) Physicochemical
studies of the sea surface microlayer. J Colloid Interface Sci 204:294–299
174

Chapter 5: Three-Domain Microbial-Environmental Networks from Ocean Time
Series Data

Chapter 5 Abstract

Microbes are central to ocean food webs (Pomeroy et al. 2007, Sherr et al. 2007, Sherr
& Sherr 2008, Caron 2009, Fuhrman 2009) and global biogeochemical processes
(Ducklow 2000, Azam & Malfatti 2007), yet specific microbial ecological relationships
are largely unknown. This is in part due to the microscopic scale of the microbial
community, which does not allow for interactions between microbes to be directly
observed. Here we use a holistic approach to investigate the interactions of all three
domains of life in a microbial community, described through cultivation–independent
data on community composition of Bacteria, Archaea, and Protists collected monthly
over 3 years at an ocean site off California. By relating the time-dependent changes
among these organisms, we created an interaction network to visualize mathematical
relationships between these organisms, along with viruses and other parameters; this
examined Local Similarity correlations (Ruan et al. 2006) among all parameters with
and without a time lag. The network suggests a succession of microbial communities,
and includes many different ecologically important taxa of groups like the SAR11
cluster, Stramenopiles, Alveolates, Cyanobacteria, and microphagotrophs (e.g. ciliates
and choanoflagellates). Negative correlations, perhaps suggesting competition or
175

predation, were also common. The network is similar to previously described ecological
networks, and in network parlance resembles an organic self-assembling network like
the internet. This promising new approach provides a feel for the natural history of
microbes, should facilitate the inclusion of complex microbial assemblages in
community ecology studies, and also points to biogeochemical roles of “unknown”
microorganisms including possible keystone species.

176

Introduction

The past two decades have seen a revolution in characterizing microbial communities
that make up the majority of global biomass (Pace et al. 1986, Giovannoni et al. 1995,
Diez et al. 2001) reviewed in (Azam & Malfatti 2007, Fuhrman 2009) but translating
that into understanding the actions and interactions of these microbes in complex
systems has been difficult. Cultivation (Giovannoni & Stingl 2007), gene chips (Zhou &
Rocke 2005), and metagenomic or transcriptomic efforts (Tyson et al. 2004, Venter et
al. 2004, Rusch et al. 2007, DeLong 2009) have provided remarkable information on
the actions or potential actions of organisms, but such fundamentally reductionist
approaches do not readily describe the interactions among microbes within a
community or with their environment. Unlike the situation with animals and plants, we
can only rarely observe or infer microscopic or chemical interactions among microbes,
such as grazing by particular protists (Sherr et al. 2007, Sherr & Sherr 2008) or
localized syntrophy (Orphan et al. 2001); whole microbial community studies in most
environments pose major challenges. We have chosen to gain insights into these
interactions by examining co-occurrence patterns of microbes with each other and with
changing environmental conditions over time, and visualizing the relationships as
networks.

Networks have been used in ecology, mostly for food webs, from Elton (Elton 1927)
until today (Bascompte 2009, Ings et al. 2009). Networks are valuable tools to describe
177

and compare systems as far ranging as the connection of pages on the internet (Barabasi
& Albert 1999), the spread of emergent diseases through a population (Pastor-Satorras
& Vespignani 2001), mutualistic (i.e. pollen-pollenator) networks (Olesen et al. 2007),
and the response of populations to disturbances including global change (Montoya et al.
2006, Ledger et al. 2008).

We examined the whole microbial plankton community and changing environmental
conditions by sampling an underwater biological feature, the subsurface chlorophyll
maximum layer, off the Southern California coast from August 2000 – March 2004.
The eukaryotic and bacterial communities were characterized by molecular
fingerprinting and backup clone libraries of ssu rRNA genes, (Brown et al. 2005,
Countway et al. 2005, Fuhrman et al. 2006, Vigil et al. 2009). The archaea were
characterized using quantitative PCR (Beman et al submitted). We also analyzed biotic
parameters (e.g. virus abundance, bacterial abundance, chlorophyll a concentration) and
abiotic parameters (e.g. temperature, salinity, dissolved oxygen) (Fuhrman et al. 2006).
Correlations of all parameters over time used both Pearson’s correlation and local
similarity analysis (LSA) (Ruan et al. 2006), the latter showing both contemporaneous
and time-lagged correlations based on normalized ranked data. In order to sort through,
condense, and visualize the tens of thousands of correlations generated by the analysis,
we used Cytoscape (Shannon et al. 2003) to create visual networks of the various
parameters and to display their statistically significant correlations. Permutation tests
178

were used to determine significance, with a cutoff of p<0.01; the likelihood that a given
reported correlation was false, the q value, was q<0.063 (Storey 2002). We examined
subnetworks to analyze microbial interactions in selected subsets of the community.

179

Results And Discussion

Each subnetwork identified by OTU type showed individuals of the target group as
hubs in “hub and wheel” arrangements (Figures 5-1 to 5-6). These show which
microbes co-occur, or conversely, tend not to occur, together. They also show the
environmental conditions that correlate positively or negatively with these OTUs – i.e.
under which conditions these groupings are found or not. These networks begin to
show us elements of the natural history of the various microbes without the need to
isolate, enrich, or otherwise manipulate the organisms.

One of the biggest questions in microbial ecology, especially regarding bacteria, is
delineation of ecotypes or ecological species, i.e. what level of phylogenetic resolution
corresponds to organisms occupying a unique niche (Cohan 2002, Ward 2002, Achtman
& Wagner 2008, Koeppel et al. 2008). Networks showing unique connections to
particular OTUs are one way to approach that (Fuhrman & Steele 2008, Fuhrman
2009). Our subnetworks identify 10 distinct ecotypes in the SAR11 cluster (Figure 5-1),
4 Cyanobacteria (Figure 5-6b), and many protists. Even among the closely related
members of each subnetwork there is structure among the groups of interconnected
organisms. In the SAR11 subnetwork (Figure 5-1) and the Stramenopile subnetwork
(Figure 5-2) we have 5 blocs (i.e. hubs connected directly or by more than one shared
neighbor) of organisms that correlate with each other over the 3 years of this study.

Figure 5-1 Subnetwork showing SAR11 OTUs as the central nodes and their “nearest
neighbors.” Ten SAR11 OTUs and the organisms and parameters correlated to
the SAR11 OTUs. The red circles are bacteria, the blue diamonds are
eukaryotes, the purple triangles are archaea, the green squares are biotic
environmental variables, the yellow hexagons are abiotic environmental
variables. The relative sizes of the bacteria and eukaryotes indicate the average
abundance of the OTUs as measured by ARISA and TRFLP. Solid lines are a
positive correlation, dashed lines are negative, red arrows indicate a one month
time shift in the covariation.

180

Figure 5-2 Subnetwork with stramenopiles as the central nodes. The symbols and colors
are as in Figure 5-1. Stramenopiles formed 4 disconnected blocs. The first
(clockwise from bottom right) with Stramenopile 490 and Stramenopile 492
directly correlated to one another and Diatom 488 connected through a shared
connection to Choanoflagellate 485. This bloc also contains the only negative
interaction to a bacterial taxon with Diatom 488 negatively correlating with
Alteromonas 642. The second bloc connects Stramenopile 331 to Stramenopile
281 through Eukaryote 265. Stramenopile 276 and Stramenopile 227 center the
last two blocs. Stramenopile 227 and Synechococcus 1051 are positively
correlated with no time lag.

181

182

The Alveolate subnetwork (Figure 5-3a) shows 5 distinct blocs, the Ciliates and
Choanoflagellate subnetwork (Figure 5-6a) shows 3 distinct blocs, and the
Cyanobacteria form two blocs (Figure 5-6b). Niche separation in the cyanobacterium
Prochlorococcus has been shown, (West & Scanlan 1999, Rocap et al. 2002, Johnson et
al. 2006), and ecotypes have been examined in SAR11 (Treusch et al. 2009) and the
eukaryotic alga Ostreococcus (Rodriguez et al. 2005) but with these subnetworks we
have been able to show which other taxa are ecologically connected to these groups.

Blocs may reveal complex community interactions. The alveolate (e.g. dinoflagellates)
subnetwork (Figure 5-3a) includes a highly interconnected group of 13 unknown
eukaryotic taxa that correlate to both Alveolate 562 and Dinoflagellate 198.
Correlations among all 13 range from 0.55 to 0.90 (Figure 5-3b). We speculate that the
high correlations (such as those between Alveolate 562 and EOTU 642 (ls=0.9,
p<0.001), EOTU 622 (ls=0.92, p<0.001), and EOTU 522 (ls=0.88, p<0.001) could
imply direct symbiotic dependence such as mutualism or parasitism. It is also possible
the 13 nodes are not all unique organisms. These most highly correlated nodes may
represent multiple distinct rRNA operons from a small group of highly connected
organisms, maybe from a single taxon.

A
Figure 5-3 Alveolate subnetworks. Subnetwork with Alveolates as central nodes and
their nearest neighbors (A). The alveolates form three blocs with few direct
interactions seen between the central nodes. Each bloc has a direct interaction
with bacteria. The 15
eukaryotic OTUs highly
correlated to Alveolate 562
and Dinoflagellate 198 form a
highly connected cluster
within the network (B). This
cluster accounts for the
eukaryotes with the highest
connectedness in the network
as a whole. They are
predominately connected to
other eukaryotes and illustrate
the value of the network
visualization and the ease with
which patterns can be seen using
B
interaction networks.

183

184

Examining the interactions with the environmental measurements, we can uncover what
conditions tend to favor or disfavor particular collections of organisms. In general, the
interactions between microbes dominated the network, rather than the abiotic and biotic
environmental parameters (Figure 5-4). This may relate to the relatively steady
environment in the deep chlorophyll maximum layer, hence the community
composition tends to change more dramatically than the physicochemical environment.
The correlations with biotic parameters that we do find may be able to provide
hypotheses for the microbes’ niches. For example, in the SAR11 subnetwork, we
observe a negative correlation (ls=-0.576, p<0.001) between SAR11 Surface group 3
719 and Total Bacterial Counts (Figure 5-1). This may reflect superior competition for
growth under low abundance conditions, or may be the result of this OTU being
relatively resistant to losses by grazing or viral lysis compared to other taxa. In contrast,
γ-proteobacteria SAR92 749 (Figure 5-5) is more likely to represent a weedy or
opportunistic species, since it is positively correlated with bacterial production
measured by leucine and thymidine incorporation (ls=0.54, p=0.003,ls=0.495, p=0.005,
respectively).

Correlations with abiotic parameters could indicate the preferred environment for the
connected taxa. For example, the positive correlation between Stramenopile 490 and
density (ls = 0.54, p=0.001), associates this Stramenopile with colder, saltier water
(Figure 5-2). This may indicate a preference for a deeper chlorophyll maximum layer,
physical forces which affect the temperature and salinity of the water (such as
185

upwelling) or the wintertime southward flow of the Southern California current system
(Di Lorenzo 2003). SAR11 S-3 719 is negatively correlated with density at a one
month time lag (ls=-0.5, p=0.004), i.e. a month after the water becomes warmer and less
saline, SAR11 S-3 719 increases (Figure 5-1). Although a direct connection is not
identified by the analysis, this may suggest that SAR11 S-3 719 and Stramenopile 492
are part of a shift in the community or an ecological succession that is tied to seasonal
changes in the water or to physical disruption.

We note that it is possible that a rank correlation may miss “hump shaped” interactions
that are common with ecological gradients and that an alternate technique such as
Canonical Correspondance Analysis may better uncover these relationships (Terbraak
& Verdonschot 1995). However, we feel that it is an adequate method to describe taxa
interactions and note that the environmental correlations that are described through this
method are valid.

Interactions between domains are common in the network. Although the subnetworks
with bacteria at the center (Figures 5-1, 5-5, and 5-7) tend to have more connections to
bacteria, and those with eukaryotes at the center (Figures 5-2, 5-3a, and 5-6) tend to
have more connections to eukaryotes, each subnetwork has multiple blocs in which
bacteria and eukaryotes are connected. Specific Bacteria and Eukaryotes are also
connected to Archaea (e.g. (triangles in Figures 5-2, 5-5, and 5-7). The reasons for

Figure 5-4 Circular network showing all significant correlations (p<= 0.01, q<=0.063)
with nodes sorted by their number of correlations. The red circles are bacteria,
the blue diamonds are eukaryotes, the purple triangles are archaea, the green
squares are biotic environmental variables, the yellow hexagons are abiotic
environmental variables. The relative sizes of the bacteria and eukaryotes
indicate the average abundance of the OTUs as measured by ARISA and
TRFLP. Solid lines are a positive covariation, dashed lines are negative, red
arrows indicate a one month time shift in the covariation.

186

Figure 5-5 Subnetwork with γ-proteobacteria as central nodes. This network identifies
12 γ-proteobacteria ecotypes which form interconnected blocs Altermonas 642
is the only “hub” which is separated from the rest of the subnetwork. The
majority of the connections are to other bacterial ecotypes, only γ-
Proteobacterium 909, Altermonas 648 and SAR92 749 do not connect to at least
one eukaryote. SAR86 525 and SAR86 528 are both positively correlated to
crenarchaea.

187

188

these co-occurrences are probably varied. Some of these co-occurrences may represent
guilds of organisms performing similar or complementary functions to each other, while
others may co-occur because of shared preferred conditions, or they may be providing
each other with complementary functions.

It is tempting to think that positive correlations among “hubs” in a subnetwork may
indicate functional redundancy or niche sharing, e.g. between Sar11 Surface-3 719 with
SAR11 Surface-1 681, (ls=0.54, p=0.003, Figure 5-1, and the correlation between
Stramenopile 492 and Stramenopile 490 (ls= 0.64, p<0.001, Figure 5-2). Yet, if there
were true functional redundancy, other taxa should respond to them in a similar way
and you might expect many shared neighbors. In each of these cases there are distinct,
unshared neighbors: 6 neighbors in SAR11 example (Figure 5-1), and 8 neighbors in
Stramenopile example (Figure 5-2). While redundancy of particular individual functions
must occur, it is unclear to what extent the collective suite of functions of particular
microorganisms overlap – i.e. the distinctiveness of individual microbial niches
(Hutchinson 1961). Fuhrman et al (2006) concluded that the predictability of most
bacterial OTU at this location suggested little functional redundancy among them.

Negative correlations may indicate competition or predation among the taxa. For
example, the negative correlation between Choanoflagellate 485 with Actinobacter 423
(ls= -0.55, p<0.001) and SAR11 693 (ls= -0.52, p=0.003) or the between Ciliate 197
with Flavobacter 854 (ls= -0.54, p=0.003; Figure 5-6) with no time delay could point to
189

a predator-prey relationship, especially considering that these protists are phagotrophic.
A negative correlation (ls= -0.51, p=0.003) with no time delay between Synechococcus
1051 and Ostreococcus 822 (a phototrophic prokaryote and a phototrophic eukaryote of
similar size, Figure 5-7), may reflect competition for a similar niche. These are
obviously not definitive conclusions, but hypothetical possibilities that may be further
examined.

A positive correlation with a time delay may also reflect competition or the succession
of the bacterial community as the environment changes. Prochlorococcus grpII 944 (a
low light-adapted group) and CHABI-7 402 show a positive correlation with a one
month time lag (ls=0.51, p=0.002) and Prochlorococcus grpI 828, (a high light-adapted
group) has a positive correlation with CHABI-7 402 (ls=0.63, p<0.001, Figure 5-7) with
no time delay. The positive correlation without delay suggest the organisms thrive
under similar conditions, and the delayed correlation may reflect a successional shift of
dominance.

For comparison to other organisms or environments, or even other systems, we
analyzed the structure of the entire network (Figure 5-4). This network contains a few
highly connected nodes and many nodes with 1 or 2 connections. The probability
distribution of the node connectivity poorly resembles a random Poisson distribution,

Figure 5-6 Subnetwork with ciliates and flagellates as central nodes. There are three
blocs within this subnetwork with Ciliate 197 and Ciliate 272 connecting
through shared neighbors Eukaryote 495 and Eukaryote 636. Choanoflagellate
485 and Ciliate 196 form their own blocs. Choano 485 has strong negative
interactions with Actinobacter 423 (-0.55) and Sar11 693 (-0.52). Ciliate 197
also has a strong negative correlation with Flavobacter 854 (-0.53).

190

Figure 5-7 Subnetwork with cyanobacteria as central nodes. We are able to identify
two different Prochlorococcus ecotypes and two Synechococcus ecotypes.
Prochlorococcus 944 and Prochlorococcus 828 are connected by shared neighbors
CHAB1-7 402 and Verruco 734. Prochlorococcus 828 is negatively correlated to both
silicate and to phosphate with a one month time lag and with SAR86 534 with no time
lag. Synechococcus 1051 shows a strong negative correlation with a one month time lag
to counts of Euryarchaea and a strong negative correlation with no time lag to
Ostreococcus 322.

191

192

but is best described by an exponential function (y = ae
-b
) and approximated by a non-
random Barabasi-Albert model (Barabasi & Albert 1999). Fitting an exponential
function to the connectivity distribution yields an exponent (b = -0.096, r
2
=0.77). This
distribution is similar to other ecological networks such as those commonly found in
foodwebs and pollinator-plant interactions (Montoya et al. 2006). This suggests we are
not merely looking at random interactions over time, but observing meaningful
relationships. Compared to other systems, the microbial network resembles organic
self-assembling networks such as the internet rather than an evenly distributed planned
network such as the highway system (Barabasi & Albert 1999). This pattern of few
highly connected nodes (which are microbial taxa and environmental parameters), as
opposed to even distribution of connectivity, makes the network more robust to change
(Albert et al. 2000, Montoya et al. 2006) but with an important caveat. Although there
is a lower probability of the network being disrupted by losing a highly connected node,
if highly connected nodes are lost, the network would change dramatically. Such nodes
might be considered microbial “keystone species.”

Although we are at the very beginning of a holistic view of interactions within
microbial communities, an interaction network can provide a useful tool to describe the
natural history of bacteria, eukaryotes, and Archaea with their environment. Creating
networks for different communities will allow us to compare their robustness or
fragility of the community (Albert et al. 2000, Montoya et al. 2006) and can provide
193

insight into how these communities will change with future environmental changes.
These interactions can provide a glimpse into their ecology. We believe this will lead to
creation of hypotheses to test and enable a deeper understanding of the microbial ocean.

194

Methods

Monthly samples were collected from the deep chlorophyll maximum depth (DCM) at
the University of Southern California Microbial Observatory at the San Pedro Ocean
Time Series Site (SPOTS) (Los Angeles, CA, USA). To identify and estimate the
changing abundance of hundreds of different types of Bacteria, Eukarya, and Archaea
we used molecular fingerprinting techniques: the Bacterial community composition was
estimated by automated ribosomal intergenic spacer analysis (ARISA) (Fisher &
Triplett 1999, Fuhrman et al. 2006) and Eukarya were estimated by terminal restriction
fragment length polymorphism (Countway et al. 2005, Vigil et al. 2009) and the
Archaea were estimated by quantitative PCR (Beman et al submitted). We note these
“identifications” are molecular genetic tag data that distinguish organisms but are often
not yet linked to a formal name, and changes in abundance are examined by ranking
only within taxa, not between them. We also measured temperature, salinity, nutrients,
chlorophyll, and abundance of bacteria and viruses, and estimated the bacterial
heterotrophic growth rates following Fuhrman et al 2006. To examine interactions
between the microbial populations and their environment, we analyzed the correlations
of the microbes with each other and with biotic and abiotic conditions over time using
both Pearson’s correlation and local similarity analysis (LSA) (Ruan et al. 2006), a
technique that shows both contemporaneous and time-lagged correlations, and is based
on normalized ranked data. In order to sort through, condense, and visualize the tens of
thousands of correlations generated by the analysis, we used Cytoscape (Shannon et al.
195

2003) to create visual networks of the various parameters and statistically significant
correlations among them. These interaction networks include nodes that consisted of
the operational taxonomic units (OTUs) as a proxy for “species” (Brown et al. 2005,
Countway et al. 2005, Vigil et al. 2009) defined by molecular fingerprinting, as well as
the environmental parameters; the local similarity scores (covariations that may include
a 1-3 month time lag) among the taxa and the environmental parameters constituted the
edges, i.e. connections between nodes. The LSA identified 212 variables with 1005
significant local similarity correlations (p ≤0.01, q≤0.062; Figure 5-4). Examining the
network, we identified ten SAR11 ecotypes with 73 connections to their 61 nearest
neighbors (Figure 5-1), ten Stramenopiles (likely Diatoms or chrysophytes) with 55
connections to their 43 nearest neighbors (Figure 5-2), thirteen alveolates with 96
connections to their 65 nearest neighbors (Figure 5-3), four Mixotrophic and
Heterotrophic Eukaryotes (3 ciliates and 1 choanoflagellate) with 38 connections to
their 34 nearest neighbors (Figure5-6) and four cyanobacteria with 41 connections to
their 34 nearest neighbors (Figure 5-7).

196

Chapter 5 References

Achtman M, Wagner M (2008) Microbial diversity and the genetic nature of microbial
species. Nat Rev Microbiol 6:431-440

Albert R, Jeong H, Barabasi AL (2000) Error and attack tolerance of complex networks.
Nature 406:378-382

Azam F, Malfatti F (2007) Microbial structuring of marine ecosystems. Nat Rev
Microbiol 5:782-791

Barabasi AL, Albert R (1999) Emergence of scaling in random networks. Science
286:509-512

Bascompte J (2009) Disentangling the Web of Life. Science 325:416-419

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show
marine microbial diversity: development and application to a time series.
Environ Microbiol 7:1466-1479

Caron DA (2009) New Accomplishments and Approaches for Assessing Protistan
Diversity and Ecology in Natural Ecosystems. Bioscience 59:287-299

Cohan FM (2002) What are bacterial species? Annu Rev Microbiol 56:457-487

Countway PD, Gast RJ, Savai P, Caron DA (2005) Protistan diversity estimates based
on 18S rDNA from seawater incubations in the western North Atlantic. J
Eukaryot Microbiol 52:95-106

DeLong EF (2009) The microbial ocean from genomes to biomes. Nature 459:200-206

Di Lorenzo E (2003) Seasonal dynamics of the surface circulation in the Southern
California Current System. Deep Sea Research II 50:2371-2388

Diez B, Pedros-Alio C, Massana R (2001) Study of genetic diversity of eukaryotic
picoplankton in different oceanic regions by small-subunit rRNA gene cloning
and sequencing. Appl Environ Microb 67:2932-2941

Ducklow HW (2000) Bacterial production and biomass in the oceans. In: Kirchman DL
(ed) Microbial ecology of the oceans. Wiley-Liss, New York, p 85-120

197

Elton CS (1927) Animal Ecology, Chicago University Press, Chicago.

Fisher MM, Triplett EW (1999) Automated approach for ribosomal intergenic spacer
analysis of microbial diversity and its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocean
conditions. P Natl Acad Sci USA 103:13104-13109

Fuhrman JA, Steele JA (2008) Community structure of marine bacterioplankton:
patterns, networks, and relationships to function. Aquat Microb Ecol 53:69-81

Giovannoni S, Stingl U (2007) The importance of culturing bacterioplankton in the
'omics' age. Nat Rev Microbiol 5:820-826

Giovannoni SJ, Mullins T, Field KG (1995) Microbial diversity in marine systems:
rRNA approaches to the study of unculturable microbes. In: Joint I (ed)
Molecular Ecology of Aquatic Microbes. Springer-Verlag, Berlin-Heidelberg-
New York-Tokyo

Hutchinson GE (1961) The paradox of the plankton. Amer. Nat 45:137-145

Ings TC, Montoya JM, Bascompte J, Bluthgen N, Brown L, Dormann CF, Edwards F,
Figueroa D, Jacob U, Jones JI, Lauridsen RB, Ledger ME, Lewis HM, Olesen
JM, van Veen FJF, Warren PH, Woodward G (2009) Ecological networks -
beyond food webs. Journal of Animal Ecology 78:253-269

Johnson ZI, Zinser ER, Coe A, McNulty NP, Woodward EM, Chisholm SW (2006)
Niche partitioning among Prochlorococcus ecotypes along ocean-scale
environmental gradients. Science 311:1737-1740

Koeppel A, Perry EB, Sikorski J, Krizanc D, Warner A, Ward DM, Rooney AP,
Brambilla E, Connor N, Ratcliff RM, Nevo E, Cohan FM (2008) Identifying the
fundamental units of bacterial diversity: A paradigm shift to incorporate ecology
into bacterial systematics. P Natl Acad Sci USA 105:2504-2509

198

Ledger ME, Harris RML, Armitage PD, Milner AM (2008) Disturbance frequency
influences patch dynamics in stream benthic algal communities. Oecologia
155:809-819

Montoya JM, Pimm SL, Sole RV (2006) Ecological networks and their fragility. Nature
442:259-264

Olesen JM, Bascompte J, Dupont YL, Jordano P (2007) The modularity of pollination
networks. P Natl Acad Sci USA 104:19891-19896

Orphan VJ, House CH, Hinrichs KU, McKeegan KD, DeLong EF (2001) Methane-
consuming archaea revealed by directly coupled isotopic and phylogenetic
analysis. Science 293:484-487

Pace NR, Stahl DA, Lane DL, Olsen GJ (1986) The analysis of natural microbial
populations by rRNA sequences. Adv. Microbiol. Ecol. 9:1-55

Pastor-Satorras R, Vespignani A (2001) Epidemic spreading in scale-free networks.
Physical Review Letters 86:3200-3203

Pomeroy LR, Williams PJI, Azam F, Hobbie JE (2007) The Microbial Loop.
Oceanography 20:28-33

Rocap G, Distel DL, Waterbury JB, Chisholm SW (2002) Resolution of
Prochlorococcus and Synechococcus ecotypes by using 16S-23S ribosomal
DNA internal transcribed spacer sequences. Applied & Environmental
Microbiology 68:1180-1191

Rodriguez F, Derelle E, Guillou L, Le Gall F, Vaulot D, Moreau H (2005) Ecotype
diversity in the marine picoeukaryote Ostreococcus (Chlorophyta,
Prasinophyceae). Environ Microbiol 7:853-859

Ruan QS, Dutta D, Schwalbach MS, Steele JA, Fuhrman JA, Sun FZ (2006) Local
similarity analysis reveals unique associations among marine bacterioplankton
species and environmental factors. Bioinformatics 22:2532-2538

199

Rusch DB, Halpern AL, Sutton G, Heidelberg KB, Williamson S, Yooseph S, Wu DY,
Eisen JA, Hoffman JM, Remington K, Beeson K, Tran B, Smith H, Baden-
Tillson H, Stewart C, Thorpe J, Freeman J, Andrews-Pfannkoch C, Venter JE,
Li K, Kravitz S, Heidelberg JF, Utterback T, Rogers YH, Falcon LI, Souza V,
Bonilla-Rosso G, Eguiarte LE, Karl DM, Sathyendranath S, Platt T,
Bermingham E, Gallardo V, Tamayo-Castillo G, Ferrari MR, Strausberg RL,
Nealson K, Friedman R, Frazier M, Venter JC (2007) The Sorcerer II Global
Ocean Sampling expedition: Northwest Atlantic through Eastern Tropical
Pacific. Plos Biol 5:398-431

Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N,
Schwikowski B, Ideker T (2003) Cytoscape: A software environment for
integrated models of biomolecular interaction networks. Genome Res 13:2498-
2504

Sherr BF, Sherr EB, Caron DA, Vaulot D, Worden AZ (2007) Oceanic Protists.
Oceanography 20:130-134

Sherr EB, Sherr BF (2008) Understanding roles of microbes in marine pelagic food
webs: a brief history. In: Kirchman D (ed) Advances in Microbial Ecology of
the Oceans. Wiley, p 27-44

Storey JD (2002) A direct approach to false discovery rates. J Roy Stat Soc B 64:479-
498

Terbraak CJF, Verdonschot PFM (1995) Canonical Correspondence-Analysis and
Related Multivariate Methods in Aquatic Ecology. Aquat Sci 57:255-289

Treusch AH, Vergin KL, Finlay LA, Donatz MG, Burton RM, Carlson CA, Giovannoni
SJ (2009) Seasonality and vertical structure of microbial communities in an
ocean gyre. Isme J 3:1148-1163

Tyson GW, Chapman J, Hugenholtz P, Allen EE, Ram RJ, Richardson PM, Solovyev
VV, Rubin EM, Rokhsar DS, Banfield JF (2004) Community structure and
metabolism through reconstruction of microbial genomes from the environment.
428:37-43

Venter JC, Remington K, Heidelberg JF, Halpern AL, Rusch D, Eisen JA, Wu DY,
Paulsen I, Nelson KE, Nelson W, Fouts DE, Levy S, Knap AH, Lomas MW,
Nealson K, White O, Peterson J, Hoffman J, Parsons R, Baden-Tillson H,
Pfannkoch C, Rogers YH, Smith HO (2004) Environmental genome shotgun
sequencing of the Sargasso Sea. Science 304:66-74

200

Vigil P, Countway PD, Rose J, Lonsdale DJ, Gobler CJ, Caron DA (2009) Rapid shifts
in dominant taxa among microbial eukaryotes in estuarine ecosystems. Aquat
Microb Ecol 54:83-100

Ward BB (2002) How many species of prokaryotes are there? P Natl Acad Sci USA
99:10234-10236

West NJ, Scanlan DJ (1999) Niche-partitioning of Prochlorococcus populations in a
stratified water column in the eastern North Atlantic Ocean. Applied &
Environmental Microbiology 65:2585-2591

Zhou L, Rocke DM (2005) An expression index for Affymetrix GeneChips based on the
generalized logarithm. Bioinformatics 21:3983-3989

201

Chapter 6: Synthesis and Conclusions

Introduction

Understanding microbial community patterns over space and time and the factors which
influence these communities is vital to understanding ecosystem function and response.
Using molecular techniques we can identify microbes and say how they are distributed
in space and time. Yet, we still are just beginning to expand our focus from looking at
who is there (i.e. what microbes make up the community) to finding out what they are
doing (i.e. how each microorganism interacts and fits within the functioning of the
ecosystem). In the ocean, which is in continuous lateral and vertical motion, these
community patterns and interactions may change on short to medium spatial and
temporal scales. The aim of this study was to determine the community patterns in
bacterioplankton in the ocean surface waters of Southern California and relate them to
environmental factors on kilometer spatial scales (Chapter 2) and hour-day timescales
(Chapter 3), micrometer to meter depth scales (Chapter 4), and interactions from
monthly correlations among bacteria, protists and Archaea over three years (Chapter 5).

202

Summary of Results

In Chapter 2 we found that coastal bacterioplankton assemblages in the Southern
California Bight are often uniform at 2-97 km
2
area scales and at distances of 0-15 km
and the major taxa appeared remarkably consistent throughout. Variations among the
communities appeared to respond to chlorophyll a and virus abundance gradients at 35
km
2
in 2004, but did not show any correlation with distance or with measured
environmental parameters in at 97 km
2
in 2005. At distances from 50-255 km, there
were strong differences in the community similarity values which correlated to
temperature and chlorophyll a which was measured at the sites, though there were still
some remarkable similarities with communities as distant as 45 km. This coherence of
the bacterioplankton community in the San Pedro Channel suggested that despite small
variations between sites, typical collection of water by niskin bottle can provide a
representative sample of the bacterioplankton over tens of square kilometers. At greater
distances (>100 km) environmental parameters, and differences in water mass drove the
variation in community composition. Taxa-area analysis revealed a scale-dependence
for the bacterioplankton communities with z-values near zero at shorter distances but
with values comparable to soil bacteria or diatoms along a 255 km transect. This
suggested that a 10 L sample was representative of an area at a distance between 4 and
15 km and the response of the community to environmental gradients (e.g. temperature,
203

salinity, and chlorophyll a), at least in the coastal Southern California Bight, suggested
that it was possible to measure changes in bacterioplankton habitat, and to predict
locations to sample, while a ship is underway.

In Chapter 3, our results demonstrate a bacterial communities are remarkably stable in
offshore and nearshore waters over short periods (30 hours offshore and 2-3 days
nearshore), but changed over time after 4-7 days. At all sites, rare taxa showed much
greater variability over short timeframes compared to abundant taxa, and these rare taxa
are candidates for responses on these shorter time scales. Virus abundance, including a
sharp increase in August 2000, was uncorrelated to community change or bacterial
abundance, suggesting a complex interaction between viruses and bacterioplankton in
these communities. High community similarity over 4-5 days within a single water
patch and in surface waters suggests that at nearshore and offshore sites, sampling every
few days or even weekly will capture a representative sample of a bacterial community
within a habitat. Sampling on short time scales (e.g. every other day to weekly), is
necessary to capture the response of the rare taxa or minor changes in abundant
bacterioplankton, while sampling on longer time scales (e.g. bi-weekly to monthly or
greater), is sufficient to capture major changes in the abundant bacterioplankton.

In Chapter 4 we found bacterial communities in the sea surface microlayer (SML) were
highly variable over space and time, and did not represent a unique community.
However, we did detect communities which were distinct from the underlying water
204

(ULW) communities in the San Pedro Channel and Catalina in May 2004 and May
2005, one of which occurred during a Synechococcus and Roseobacter bloom. We also
found bacterial production to be consistently depleted in the SML compared to ULW,
and variable enrichment of bacteria and viruses. Even the bacterial communities in the
SML which showed the greatest difference from the communities in the ULW shared
abundant taxa. This suggests that the bacteria which make up the SML community in
Southern California waters are transported from the surface waters and that the
community in the surface waters have a large impact on the community in the
microlayer.

In Chapter 5 we demonstrated ecologically relevant interactions among protists,
bacteria, and Archaea from correlations between the communities derived from holistic
community patterns over three years at an ocean time series. The patterns shown were
suggestive of a succession of microbial communities, and delineated many different
ecotypes of important groups such as the SAR11 cluster, Stramenopiles, Alveolates,
Cyanobacteria, and microphagotrophs (e.g. ciliates and choanoflagellates). Negative
correlations, perhaps suggesting competition or predation, were also common. The
network topology showed relatively few, highly connected nodes (i.e. taxa or
environmental factors which had a high number of significant correlations) with a much
larger number of nodes showing few connections, providing evidence of non-random
205

structure in the network. This promising new approach should facilitate inclusion of
complex microbial assemblages in community ecology studies, and also point to
biogeochemical roles of “unknown” microorganisms.

206

Synthesis and Conclusions

This synthesis will serve to pull some strings that are tied between the chapters, and will
try to avoid repeating the conclusions present in the data chapters and the summary of
results. Probably the most utilitarian, and possibly the most important, conclusion to be
drawn from these studies, and the one which allows for comparison with other studies,
is the implication for what a typical sample represents which are crucial for
extrapolating from a sample to the larger ocean (Fuhrman 2009). When you combine
the results of the spatial and temporal changes from Chapters 2 to 4, you are able to get
an idea of this sample represents in the surface mixed layer, in the chlorophyll
maximum depth, and in the surface waters and in the (SML). Based on the similarities
between hours, days, and across space, described in Chapter 2 and 3 a 10L sample in
Southern California represents a 30-100 km
2
area and is representative anywhere from
2-6 days. This is in nearly perfect agreement with the sampling size of days-weeks and
kilometers to tens of kilometers put forth by earlier studies (Hewson et al. 2006,
Fuhrman & Hagström 2008). As pointed out by both of these studies, this is similar to
the scale of mesoscale eddies in satellite temperature and chlorophyll measurements and
suggests that perhaps bacterioplankton are being controlled by the same environmental
factors as phytoplankton (Fuhrman & Hagström 2008).

207

Although the bacterial community was a little more variable in the chlorophyll
maximum depth over 30 hours, (Chapter 3, Figure 3-6), since that sample was not
within a water mass, and quite possibly could have been moving at a different rate or
even in a different direction, it is difficult to determine that the scale was different at
that depth. As seen in Chapter 5 and the SPOTS dataset in (Brown et al. 2005), there
are clearly repeating patterns in the chlorophyll maximum depth that suggest that it does
not experience a much higher rate of turnover than the 5 m depth. On the other hand the
high degree of variability in the SML and in the 0.5 m depth surface waters (Chapter 4
Figures 4-6 and 4-7) however provides some clear warning signs, particularly for
comparing bucket samples to 5 m CTD casts. Higher variability in the SML and the
0.5m bacterioplankton communities also may explain some of the high variation
previously reported in bacterioplankton in ocean surface waters (e.g. Long & Azam
2001). Although much of the variation reported was likely the result of the small scale
studied, it stands to reason that the patchy nature of the microlayer community and the
surface layer community may have influenced these studies as well.

The agreement between the rates of community change (15-17% d
-1
) between a study in
the oligotrophic ocean where bacteria were growing much slower (Hewson et al. 2006)
and the coastal ocean where we had a higher rate of growth (Chapter 2 Tables 2-1 and
2-2 and Chapter 3 Tables 3-2 and 3-4) indicates a pattern that may be independent of
environmental control. Such a pattern would likely be due to physical processes, such

208

as the 10x10 km water patch, or the local size of physical mesoscale eddies (Fuhrman &
Hagström 2008). These physical processes may represent a scale of functional
microbial mixing in the ocean, which defines the boundaries of a single bacterial
community.

Taking into consideration the appearance of transition communities (Chapter 2 Figure
2-10) and the correlation with temperature and density, further suggests physical mixing
plays a prominent role in determining the bacterioplankton community structure. This
adds to the evidence for spatial structure in oceanic bacterioplankton at larger distance
scales, with mixing driven by physical features, such as eddies, and small-scale
patchiness as suggested by Hewson et al. (2006) and Fuhrman and Hagstrom (2008).
This leaves you with the image of a number of 10 km water patches moving along the
currents with their edges mixing while the bacteria are busily going about their lives
within them.

The stability of bacterioplankton communities over a week at Catalina and in the San
Pedro Channel and the spatial variation between the 0.5 m depth samples in Catalina
bacterioplankton communities 45 km to the South could indicate that there is a variable
scale of microbial habitat in this region and 10 km is actually a low estimate. It could
also indicate the splitting of a habitat into two or more different water masses. Although
this did not seem to affect the bacterial community similarity in this study, the small
scale movement of water masses may certainly move these communities and split or
209

mix them on a small scale. Small scale movement occurs within the hours to days
timeframe, shown by the tendency of drogues to spread during 20-30 hour studies.
When the high degree of physical mixing shown by drifter studies and models of the
small scale currents in the Southern California Bight is considered (Dong et al. 2009,
Noble et al. 2009) it seems likely that the similarity between communities at 45 km
shown is due to just such mixing or splitting.

The taxa-area relationship described in Chapter 3 (Table 3-3) most resembles z-values
for benthic diatoms on scales up to 10
-4
to 10
12
km
2
(Azovsky 2002), the bacteria in
estuarine sediments up to 0.09 km
2
(Horner-Devine et al. 2004), and soil bacteria 10
2
-
10
8
km
2
(Fierer & Jackson 2006) but it’s not entirely clear why this should be the case.
First, a word of caution, although the transect value was higher, it is based on a
distance-decay relationship (similar to Horner-Devine et al 2004) but it is not from a
nested sampling design (i.e. a sample grid that increases in size, but includes the smaller
grid within the larger grids). The advantage of the nested sampling design is the high
coverage of the areas being sampled as well as the increase in number of points of
comparison, compared to a straight-line transect. The caveats with distance decay
apply, but the z-value is likely being underestimated by ARISA which separates
bacteria on a ~98% 16S similarity level (Brown & Fuhrman 2005).

What is striking about the z-value in Chapter 2 is that it increased with distance and that
at its low end it’s comparable to the z value of 0.03 for soil bacteria over 10
2
km
2

210

reported by Fierer and Jackson (2006), but at the high end, it’s comparable to the z
value for estuarine bacteria at 0.1 km
2
, and it’s near, but below the scale-invariant z-
value for diatoms (Azovsky 2002). The scaling of z-values with distance (in plants it
actually shows as a hump shaped relationship (Crawley & Harral 2001) suggests one of
two things: One is that the study undersampled in terms of area (i.e. sampling did not
move beyond the initial set of species) which would produce a low z-value. The other is
that there are different processes acting on the species being studied at the different
scales, creating a different rate of turnover for the species at those scales. The
relationship of the surface waters to soil communities is likely due to the high
environmental heterogeneity in soils (Ramette & Tiedje 2007a, b). In other words, a
patch of ocean water at 5 m over 100 km
2
has as little spatial structure as 100 km
2
of
soil, but once you move out of that patch of water, there is an environmental and spatial
structure present that results in a higher z-value. The physical properties of water
patches in the ocean may provide spatial structure on a global scale. This can provide at
least a partial explanation for the observation of latitude diversity gradient in marine
bacteria (Fuhrman et al. 2008) and the finding of no gradient in soil bacteria (Fierer &
Jackson 2006).

The low production rates reported in the SML (Chapter 4 Figure 4-3, Table 2-2) and
reported elsewhere, lead one to speculate that the bacterioplankton in the SML may in
fact be trapped in the gel, rather than happily making a living or at least struggling
under the exposure and stress. This agrees with the view of Maki (Maki 1993) that the
211

SML represents an extreme environment, because of high pollutant levels, high heavy
metal concentration, and high UV radiation. (Stolle et al. 2009) indicated that Leucine
uptake was more likely to be inhibited by sampling techniques than thymidine uptake.
There were only a few measurements made with Thymidine uptake during the study in
Chapter 4, and the uptake rate was also much lower in the SML compared to the ULW.
Another possibility is that living in the SML provides other advantages that balance out
its challenges, such as a refuge from predation. Although it was not measured here,
(Obernosterer et al. 2008, Reinthaler et al. 2008) found a high rate of bacterial
respiration in the SML even when there was a low rate of Leucine uptake and (Munster
et al. 1998) and (Kuznetsova & Lee 2001) found increased rates of enzymatic activity.
This implies that the bacteria may be adapted to living in the gel even though they are
growing slower. In all, the variability of the microlayer, the difficulty of sampling it,
and the difficulty of comparing those samples across studies will likely be a stumbling
block in getting a much better understanding of the ecology of the organisms in the
microlayer.

In Chapter 5, the increase and decrease of viruses did not show many correlations with
the other nodes in the correlation network. Although the lack of correlation of bacterial
community patterns with viruses is discussed in Chapter 3, it is worth pointing out that
the virus abundance was connected only to bacterial abundance and two bacterial taxa
and no protists. This suggests, along with studies by (Schwalbach et al. 2004), that there
may be a kind of equilibrium with virus populations in situ on a monthly scale as well
212

as a shorter term scale. In these communities this equilibrium may be the result of an
evolutionary and ecological truce reached between the predator (the virus) and the prey
(likely the bacteria) as previously suggested (Fuhrman 1999, 2000, Brüssow 2007).

In Chapter 5, the method of distinguishing ecotypes used in the network by combining
the “strain specific” ITS length (Brown & Fuhrman 2005) and environmental
correlations may be a useful method to determine ecological roles. It has been suggested
that ecotypes may be more appropriate distinction than species in distinguishing
microbes (Hughes-Martiny et al. 2006, Prosser et al. 2007). This same model could
prove useful tool for environmental genomics or even meta-transciptomics (DeLong
2009). It is conceivable that a correlation network between the genes that are activated
and the environmental gradients would be informative as to the ecological role of the
organism, and at the early stage of a systems approach to the network, it doesn’t seem
too far off from early versions of protein networks which were used to infer function of
the proteins within cells (Kitano 2002). In any case, it is a better visualization tool than
a heat map.

It is important to make the distinction that there is no indication from the network itself
whether or not the relationships in the network are direct or indirect and it is difficult to
compare with the classically studied ecological networks, which consider only direct
interactions. However, it may be possible to design experiments to determine the direct
interactions in specific cases. One way would be to study the interactions between
213

specific taxa (such as bacterivory between ciliates and bacteria) or between a group of
taxa and changes in their environment. Another less direct method would be to take
advantage of correlations between unknown organisms and known organisms to help
determine their response to environmental and biological changes. Ideal places to look
for this are the strong positive or negative correlations within the subnetworks, where
symbiosis, competition, or predator-prey relationships are indicated. Even without
direct interactions, it may be possible to compare network statistics to determine the
stability of the interaction networks being studied (Banasek-Richter et al. 2009).

This work is by no means comprehensive and suggests several possible future studies
that would further elucidate the bacterioplankton biogeography or extend the
examination of the concepts addressed within this dissertation. A simple next step to
delve deeper into the spatial ecology of the bacterioplankton would be an investigation
into the spatial variation of different taxonomic groups (e.g. Cyanobacteria, SAR11, γ-
proteobacteria, etc.), and a comparison of the differences between them (including taxa-
area relationships). This could provide insight into each of these taxonomic groups
responses to environmental change, as well as distinguishing the limits of dispersion for
these groups. Suggestions to extend the temporal study include an examination of
weekly scales of bacterioplankton community variation and, provided you had a
sufficient data set, an examination into the taxa-time relationship over days, weeks, and
months. These studies would provide insight into whether the rates of community
change reflect different processes and it would also be able to provide evidence for
214

bacterial community succession. A logical extension over short vertical space would
include sampling the SML and their ULW community within a nested grid design. This
would allow for comparison of the spatial variation of the SML community with the
corresponding ULW community and provide insight into the processes affecting these
communities over different distance scales. Finally, a logical extension of the
interaction networks is to compare the differences in the patterns at different depths or
split among multiple time-frames (e.g. three year periods or season-season), or even to
compare the networks from different environments. These comparisons would provide
insight into the interactions between the bacteria in the different depths and also could
show whether or not the patterns found among bacteria and protists change on longer
timescales.

This study has shown that bacterioplankton community patterns and the relationships
between the bacterioplankton communities and their environment can provide insight
into their basic structure and into microbially relevant habitat in the water column.
Determining the spatial limits of coherent communities on both short depth scales and
kilometer lateral scales will inform sampling of the bacterioplankton as well as
theoretical development of their ecology. While examining the changes in
bacterioplankton communities over time provides insight into the stability of the
communities and (when considering correlations among taxa and environmental
215

factors) may reveal functional roles of the microbes. In any case, the studies in this
dissertation provide a solid addition to a more complete understanding of
bacterioplankton biogeography and their ecological roles.

216

Chapter 6 References
Azovsky AI (2002) Size-dependent species-area relationships in benthos: is the world
more diverse for microbes? Ecography 25:273-282

Banasek-Richter C, Bersier LF, Cattin MF, Baltensperger R, Gabriel JP, Merz Y,
Ulanowicz RE, Tavares AF, Williams DD, De Ruiter PC, Winemiller KO,
Naisbit RE (2009) Complexity in quantitative food webs. Ecology 90:1470-1477

Brown MV, Fuhrman JA (2005) Marine bacterial microdiversity as revealed by internal
transcribed spacer analysis. Aquat Microb Ecol 41:15-23

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show
marine microbial diversity: development and application to a time series.
Environ Microbiol 7:1466-1479

Brüssow H (2007) Bacteria between protists and phages: from antipredation strategies
to the evolution of pathogenicity. Mol Microbiol 65 583–589

Crawley MJ, Harral JE (2001) Scale dependence in plant biodiversity. Science 291:864-
868

DeLong EF (2009) The microbial ocean from genomes to biomes. Nature 459:200-206

Dong CM, Idica EY, McWilliams JC (2009) Circulation and multiple-scale variability
in the Southern California Bight. Prog Oceanogr 82:168-190

Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial
communities. P Natl Acad Sci USA 103:626-631

Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects.
Nature 399:541-548

Fuhrman JA (2000) Impact of Viruses on Bacterial Processes. In: Kirchman DL (ed)
Microbial Ecology of the Oceans. Wiley-Liss, Inc. New York

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Hagström Å (2008) Bacterial and archaeal community structure and its
patterns. In: Kirchman DL (ed) Microbial Ecology of the Oceans.Wiley New
York
217

Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, Brown JH
(2008) A latitudinal diversity gradient in planktonic marine bacteria. P Natl
Acad Sci USA 105:7774-7778

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006) Temporal and spatial scales of
variation in bacterioplankton assemblages of oligotrophic surface waters. Mar
Ecol-Prog Ser 311:67-77

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM (2004) A taxa-area
relationship for bacteria. Nature 432:750-753

Hughes-Martiny JB, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL,
Horner-Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S,
Ovreas L, Reysenbach A-L, Smith VH, Staley JT (2006) Microbial
biogeography: putting microorganisms on the map. Nat Rev Micro 4:102-112

Kitano H (2002) Systems biology: A brief overview. Science 295:1662-1664

Kuznetsova M, Lee C (2001) Enhanced extracellular enzymatic peptide hydrolysis in
the sea-surface microlayer. Marine Chemistry 73:319-332

Long RA, Azam F (2001) Microscale patchiness of bacterioplankton assemblage
richness in seawater. Aquat Microb Ecol 26:103-113

Maki JS (1993) The air-water interface as an extreme environment. In: Ford TE (ed)
Aquatic microbiology—an ecological approach. Blackwell Scientific, Boston, p
409–440

Munster U, Heikkinen E, Knulst J (1998) Nutrient composition, microbial biomass and
activity at the air-water interface of small boreal forest. Hydrobiologia 363:261-
270

Noble M, Jones B, Hamilton P, Xu JP, Robertson G, Rosenfeld L, Largier J (2009)
Cross-shelf transport into nearshore waters due to shoaling internal tides in San
Pedro Bay, CA. Cont Shelf Res 29:1768-1785

Obernosterer I, Catala P, Lami R, Caparros J, Ras J, Bricaud A, Dupuy C, van
Wambeke F, Lebaron P (2008) Biochemical characteristics and bacterial
community structure of the sea surface microlayer in the South Pacific Ocean.
Biogeosciences 5:693-705

218

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green
JL, Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ,
Young JPW (2007) The role of ecological theory in microbial ecology. Nat Rev
Micro 5:384-392

Ramette A, Tiedje JM (2007a) Biogeography: an emerging cornerstone for
understanding prokaryotic diversity, ecology, and evolution. Microb Ecol
53:197-207

Ramette A, Tiedje JM (2007b) Multiscale responses of microbial life to spatial distance
and environmental heterogeneity in a patchy ecosystem. Proc Natl Acad Sci U S
A 104:2761-2766

Reinthaler T, Sintes E, Herndl GJ (2008) Dissolved organic matter and bacterial
production and respiration in the sea-surface microlayer of the open Atlantic and
the western Mediterranean Sea. Limnol Oceanogr 53:122-136

Schwalbach MS, Hewson I, Fuhrman JA (2004) Viral effects on bacterial community
composition in marine plankton microcosms. Aquat Microb Ecol 34:117-127

Stolle C, Nagel K, Labrenz M, Jürgens K (2009) Bacterial activity in the sea-surface
microlayer: in situ investigations in the Baltic Sea and the influence of sampling
devices. Aquat Microb Ecol 58:67–78

219

Alphabetized Bibliography

Achtman M, Wagner M (2008) Microbial diversity and the genetic nature of microbial
species. Nat Rev Microbiol 6:431-440

Acinas SG, Rodriguez-Valera F, Pedros-Alio C (1997) Spatial and temporal variation
in marine bacterioplankton diversity as shown by RFLP fingerprinting of PCR
amplified 16S rDNA. Fems Microbiol Ecol 24:27-40

Agogue H, Casamayor EO, Bourrain M, Obernosterer I, Joux F, Herndl GJ, Lebaron P
(2005a) A survey on bacteria inhabiting the sea surface microlayer of coastal
ecosystems. Fems Microbiology Ecology 54:269-280

Agogue H, Casamayor EO, Joux F, Obernosterer I, Dupuy C, Lantoine F, Catala P,
Weinbauer MG, Reinthaler T, Herndl GJ, Lebaron P (2004) Comparison of
samplers for the biological characterization of the sea surface microlayer.
Limnol Oceanogr-Meth 2:213-225

Agogue H, Joux F, Obernosterer I, Lebaron P (2005b) Resistance of marine
bacterioneuston to solar radiation. Applied and Environmental Microbiology
71:5282-5289

Albert R, Jeong H, Barabasi AL (2000) Error and attack tolerance of complex
networks. Nature 406:378-382

Anderson MJ (2001) A new method for non-parametric multivariate analysis of
variance. Austral Ecol 26:32-46.

Anderson MJ, Gorley RN, Clarke KR (2008) PERMANOVA+ for PRIMER: guide to
software and statistical methods. In. PRIMER-E, Plymouth, UK

Avaniss-Aghajani E, Jones K, Chapman D, Brunk C (1994) A Molecular Technique
For Identification of Bacteria Using Small Subunit Ribosomal Rna Sequences.
Biotechniques 17:144-149

Azam F (1998) Microbial control of oceanic carbon flux: The plot thickens. Science
280:694

Azam F, Fenchel T, Field JG, Gray JS, Meyer-Reil LA, Thingstad F (1983) The
ecological role of water-column microbes in the sea. Marine Ecology Progress
Series 10:257-263

220

Azam F, Malfatti F (2007) Microbial structuring of marine ecosystems. Nat Rev
Microbiol 5:782-791

Azovsky AI (2002) Size-dependent species-area relationships in benthos: is the world
more diverse for microbes? Ecography 25:273-282

Baas-Becking LGM (1934) Geobiologie of Inleiding Tot de Milieukunde The Hague:
Van Stockkum & Zoon

Banasek-Richter C, Bersier LF, Cattin MF, Baltensperger R, Gabriel JP, Merz Y,
Ulanowicz RE, Tavares AF, Williams DD, De Ruiter PC, Winemiller KO,
Naisbit RE (2009) Complexity in quantitative food webs. Ecology 90:1470-
1477

Barabasi AL, Albert R (1999) Emergence of scaling in random networks. Science
286:509-512

Bascompte J (2009) Disentangling the Web of Life. Science 325:416-419

Bell T, Ager D, Song JI, Newman JA, Thompson IP, Lilley AK, van der Gast CJ
(2005) Larger islands house more bacterial taxa. Science 308:1884

Bezdek HF, Carlucci AF (1972) Surface Concentration of Marine Bacteria. Limnol
Oceanogr 17:566

Bezdek HF, Carlucci AF (1974) Concentration and Removal of Liquid Microlayers
from a Seawater Surface by Bursting Bubbles. Limnol Oceanogr 19:126-132

Brown MV, Schwalbach MS, Hewson I, Fuhrman JA (2005) Coupling 16S-ITS rDNA
clone libraries and automated ribosomal intergenic spacer analysis to show
marine microbial diversity: development and application to a time series.
Environ Microbiol 7:1466-1479

Brown MV, Fuhrman JA (2005) Marine bacterial microdiversity as revealed by
internal transcribed spacer analysis. Aquat Microb Ecol 41:15-23

Brüssow H (2007) Bacteria between protists and phages: from antipredation strategies
to the evolution of pathogenicity. Mol Microbiol 65 583–589

Campbell BJ, Yu L, Straza TRA, Kirchman DL (2009) Temporal changes in bacterial
rRNA and rRNA genes in Delaware (USA) coastal waters. Aquat Microb Ecol
57:123-135
221

Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ, Vergin K (2009)
Seasonal dynamics of SAR11 populations in the euphotic and mesopelagic
zones of the northwestern Sargasso Sea. Isme Journal 3:283-295

Carlson DJ (1983) Dissolved Organic Materials in Surface Microlayers - Temporal and
Spatial Variability and Relation to Sea State. Limnol Oceanogr 28:415-431

Carlucci AF, Craven DB, Henrichs SM (1985) Surface-Film Microheterotrophs -
Amino-Acid Metabolism and Solar-Radiation Effects on Their Activities.
Marine Biology 85:13-22

Carlucci AF, Craven DB, Robertson KJ, Henrichs SM (1986) Microheterotrophic
Utilization of Dissolved Free Amino-Acids in Depth Profiles of Southern-
California Borderland Basin Waters. Oceanol Acta 9:89-96

Carlucci AF, Craven DB, Wolgast DM (1991) Microbial-Populations in Surface-Films
and Subsurface Waters - Amino-Acid-Metabolism and Growth. Marine Biology
108:329-339

Caron DA (2009) New Accomplishments and Approaches for Assessing Protistan
Diversity and Ecology in Natural Ecosystems. Bioscience 59:287-299

Carty C, Colwell RR (1975) A microbiological study of air and surface water
microlayers in the open ocean. J Wash Acad Sci 65:148–152

Casamayor EO, Pedros-Alio C, Muyzer G, Amann R (2002) Microheterogeneity in
16S ribosomal DNA-defined bacterial populations from a stratified planktonic
environment is related to temporal changes and to ecological adaptations. Appl
Environ Microb 68:1706-1714

Cho BC, Azam F (1988) Major role of bacteria in biochemical fluxes in the ocean's
interior. Nature 332:441-443

Clarke KR (1993) Non-parametric multivariate analyses of changes in community
structure. . Australian Journal of Ecology 18:117-143.

Clarke KR, Gorley RN (2006) PRIMER v6: User Manual/Tutorial. Plymouth, UK

Clarke KR, Warwick RM (2001) Change in Marine Communities: an approach to
statistical analysis and interpretation, 2
nd
edn., Plymouth, UK

Cohan FM (2002) What are bacterial species? Annu Rev Microbiol 56:457-487
222

Countway PD, Gast RJ, Savai P, Caron DA (2005) Protistan diversity estimates based
on 18S rDNA from seawater incubations in the western North Atlantic. J
Eukaryot Microbiol 52:95-106

Crawley MJ, Harral JE (2001) Scale dependence in plant biodiversity. Science
291:864-868

Cunliffe M, Harrison E, Salter M, Schafer H, Upstill-Goddard RC, Murrell JC (2009)
Comparison and validation of sampling strategies for the molecular microbial
analysis of surface microlayers. Aquat Microb Ecol 57:69-77

Cunliffe M, Murrell JC (2009) The sea-surface microlayer is a gelatinous biofilm. Isme
Journal 3:1001-1003

Cunliffe M, Schafer H, Harrison E, Cleave S, Upstill-Goddard R, Murrell JC (2008)
Phylogenetic and functional gene analysis of the bacterial and archaeal
communities associated with the surface microlayer of an estuary. Isme Journal
2:776-789

DeLong EF (2009) The microbial ocean from genomes to biomes. Nature 459:200-206

DeLong EF, Preston CM, Mincer T, Rich V, Hallam SJ, Frigaard NU, Martinez A,
Sullivan MB, Edwards R, Brito BR, Chisholm SW, Karl DM (2006)
Community genomics among stratified microbial assemblages in the ocean's
interior. Science 311:496-503

Di Lorenzo E (2003) Seasonal dynamics of the surface circulation in the Southern
California Current System. Deep Sea Research II 50:2371-2388

Diez B, Pedros-Alio C, Massana R (2001) Study of genetic diversity of eukaryotic
picoplankton in different oceanic regions by small-subunit rRNA gene cloning
and sequencing. Appl Environ Microb 67:2932-2941

Dong CM, Idica EY, McWilliams JC (2009) Circulation and multiple-scale variability
in the Southern California Bight. Prog Oceanogr 82:168-190

Dong CM, McWilliams JC (2007) A numerical study of island wakes in the Southern
California Bight. Cont Shelf Res 27:1233-1248

Ducklow HW (2000) Bacterial production and biomass in the oceans. In: Kirchman DL
(ed) Microbial ecology of the oceans. Wiley-Liss, New York, p 85-120

223

Elton CS (1927) Animal Ecology. Chicago University Press, Chicago

Fierer N (2008) Microbial biogeography: patterns in microbial diversity across space
and time. In: Zengler K (ed) Accessing Uncultivated Microorganisms: from the
Environment to Organisms and Genomes and Back. ASM Press, Washington
DC, p 95-115

Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial
communities. P Natl Acad Sci USA 103:626-631

Fisher MM, Triplett EW (1999) Automated approach for ribosomal intergenic spacer
analysis of microbial diversity and its application to freshwater bacterial
communities. Appl Environ Microb 65:4630-4636

Franklin MP, McDonald IR, Bourne DG, Owens NJP, Upstill-Goddard RC, Murrell JC
(2005) Bacterial diversity in the bacterioneuston (sea surface microlayer): the
bacterioneuston through the looking glass. Environmental Microbiology 7:723-
736

Frew NM, Nelson RK, Bock EJ, McGillis WR, Edson JB, Hara T (2002) Spatial
variations in surface microlayer surfactants and their role in modulating air–sea
exchange. In: Donelan MA, Drennan WM, Saltzman ES, Wannickhof R (eds)
Gas Transfer at Water Surfaces. AGU, , Washington, D.C., p 153–159

Fuhrman JA (1992) Bacterioplankton roles in cycling of organic matter: the microbial
food web. In: Falkowski PG, Woodhead AD (eds) Primary productivity and
biogeochemical cycles in the sea. Plenum Press, New York, p 361-383

Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects.
Nature 399:541-548

Fuhrman JA (2000) Impact of Viruses on Bacterial Processes. In: Kirchman DL (ed)
Microbial Ecology of the Oceans. Wiley-Liss, Inc.

Fuhrman JA (2009) Microbial community structure and its functional implications.
Nature 459:193-199

Fuhrman JA, Azam F (1982) Thymidine Incorporation as a Measure of Heterotrophic
Bacterioplankton Production in Marine Surface Waters - Evaluation and Field
Results. Marine Biology 66:109-120

224

Fuhrman JA, Hagström Å (2008) Bacterial and archaeal community structure and its
patterns. In: Kirchman DL (ed) Microbial Ecology of the Oceans. Wiley

Fuhrman JA, Steele JA (2008) Community structure of marine bacterioplankton:
patterns, networks, and relationships to function. Aquat Microb Ecol 53:69-81

Fuhrman JA, Comeau DE, Hagstrom A, Chan AM (1988) Extraction of DNA suitable
for molecular biological studies from natural planktonic microorganisms. Appl.
Environ. Microbiol. 54:1426-1429

Fuhrman JA, Eppley RW, Hagstrom A, Azam F (1985) Diel variation in
bacterioplankton, and related parameters in the Southern California Bight. Mar.
Ecol. Prog. Ser. 27:9-20

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, Brown MV, Naeem S (2006)
Annually reoccurring bacterial communities are predictable from ocean
conditions. P Natl Acad Sci USA 103:13104-13109

Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, Brown JH
(2008) A latitudinal diversity gradient in planktonic marine bacteria. P Natl
Acad Sci USA 105:7774-7778

Garabetian F (1990) Co2 Production at the Sea-Air Interface - an Approach by the
Study of Respiratory Processes in Surface Microlayer. Int Rev Ges Hydrobio
75:219-229

Garrett WD (1965) Collection of slick-forming materials from the sea surface.
Limnology & Oceanography 10:602-605

Gasol JM, Pinhassi J, Alonso S, xe, ez L, Ducklow H, Herndl GJ, Kobl, xed, zek M,
Labrenz M, Luo Y, Mor, xe, n XAG, Reinthaler T, Simon M (2008) Towards a
better understanding of microbial carbon flux in the sea*. Aquat Microb Ecol
53:21-38

Gasparovic B, Kozarac Z, Saliot A, Cosovic B, Mobius D (1998) Physicochemical
characterization of natural and ex-situ reconstructed sea-surface microlayers. J
Colloid Interf Sci 208:191-202

Giovannoni S, Stingl U (2007) The importance of culturing bacterioplankton in the
'omics' age. Nat Rev Microbiol 5:820-826

225

Giovannoni SJ, Mullins T, Field KG (1995) Microbial diversity in marine systems:
rRNA approaches to the study of unculturable microbes. In: Joint I (ed)
Molecular Ecology of Aquatic Microbes. Springer-Verlag, Berlin-Heidelberg-
New York-Tokyo

Green J, Bohannan BJ (2006) Spatial scaling of microbial biodiversity. Trends Ecol
Evol 21:501-507

Hardy J, Kiesser S, Antrim L, Stubin A, Kocan R, Strand J (1987) The Sea-Surface
Microlayer of Puget-Sound .1. Toxic Effects on Fish Eggs and Larvae. Marine
Environmental Research 23:227-249

Hardy JT (1982) The Sea-Surface Microlayer - Biology, Chemistry and Anthropogenic
Enrichment. Prog Oceanogr 11:307-328

Hardy JT, Coley JA, Antrim LD, Kiesser SL (1988) A Hydrophobic Large-Volume
Sampler for Collecting Aquatic Surface Microlayers - Characterization and
Comparison with the Glass Plate Method. Can J Fish Aquat Sci 45:822-826

Hewson I, Fuhrman JA (2004) Richness and diversity of bacterioplankton species
along an estuarine gradient in Moreton Bay, Australia. Appl Environ Microb
70:3425-3433

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006a) Remarkable heterogeneity in
meso- and bathypelagic bacterioplankton assemblage composition. Limnol
Oceanogr 51:1274-1283

Hewson I, Steele JA, Capone DG, Fuhrman JA (2006b) Temporal and spatial scales of
variation in bacterioplankton assemblages of oligotrophic surface waters. Mar
Ecol-Prog Ser 311:67-77

Horner-Devine MC, Carney KM, Bohannan BJM (2004a) An ecological perspective
on bacterial biodiversity. P Roy Soc Lond B Bio 271:113-122

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM (2004) A taxa-area
relationship for bacteria. Nature 432:750-753

Huber JA, Mark Welch D, Morrison HG, Huse SM, Neal PR, Butterfield DA, Sogin
ML (2007) Microbial population structures in the deep marine biosphere.
Science 318:97-100

226

Hughes-Martiny JB, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL,
Horner-Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S,
Ovreas L, Reysenbach A-L, Smith VH, Staley JT (2006) Microbial
biogeography: putting microorganisms on the map. Nat Rev Micro 4:102-112

Hunt DE, Klepac-Ceraj V, Acinas SG, Gautier C, Bertilsson S, Polz MF (2006)
Evaluation of 23S rRNA PCR primers for use in phylogenetic studies of
bacterial diversity. Appl Environ Microb 72:2221-2225

Hutchinson GE (1961) The paradox of the plankton. Amer. Nat 45:137-145

Ings TC, Montoya JM, Bascompte J, Bluthgen N, Brown L, Dormann CF, Edwards F,
Figueroa D, Jacob U, Jones JI, Lauridsen RB, Ledger ME, Lewis HM, Olesen
JM, van Veen FJF, Warren PH, Woodward G (2009) Ecological networks -
beyond food webs. Journal of Animal Ecology 78:253-269

Johnson ZI, Zinser ER, Coe A, McNulty NP, Woodward EM, Chisholm SW (2006)
Niche partitioning among Prochlorococcus ecotypes along ocean-scale
environmental gradients. Science 311:1737-1740

Joux F, Agogue H, Obernosterer I, Dupuy C, Reinthaler T, Herndl GJ, Lebaron P
(2006) Microbial community structure in the sea surface microlayer at two
contrasting coastal sites in the northwestern Mediterranean Sea. Aquat Microb
Ecol 42:91-104

Kirchman DL (2000) Uptake and regeneration of inorganic nutrients by marine
heterotrophic bacteria. In: Kirchman DL (ed) Microbial Ecology of the Oceans.
Wiley, New York, p 261-288

Kirchman D, Knees E, Hodson R (1985) Leucine Incorporation and Its Potential as a
Measure of Protein-Synthesis by Bacteria in Natural Aquatic Systems. Applied
and Environmental Microbiology 49:599-607

Kirchman DL, Yu LY, Fuchs BM, Amann R (2001) Structure of bacterial communities
in aquatic systems as revealed by filter PCR. Aquat Microb Ecol 26:13-22

Kitano H (2002) Systems biology: A brief overview. Science 295:1662-1664

Koeppel A, Perry EB, Sikorski J, Krizanc D, Warner A, Ward DM, Rooney AP,
Brambilla E, Connor N, Ratcliff RM, Nevo E, Cohan FM (2008) Identifying
the fundamental units of bacterial diversity: A paradigm shift to incorporate
ecology into bacterial systematics. P Natl Acad Sci USA 105:2504-2509
227

Kuznetsova M, Lee C (2001) Enhanced extracellular enzymatic peptide hydrolysis in
the sea-surface microlayer. Marine Chemistry 73:319-332

Kuznetsova M, Lee C, Aller J, Frew N (2004) Enrichment of amino acids in the sea
surface microlayer at coastal and open ocean sites in the North Atlantic Ocean.
Limnol Oceanogr 49:1605-1619

Lear G, Anderson MJ, Smith JP, Boxen K, Lewis GD (2008) Spatial and temporal
heterogeneity of the bacterial communities in stream epilithic biofilms. Fems
Microbiol Ecol 65:463-473

Ledger ME, Harris RML, Armitage PD, Milner AM (2008) Disturbance frequency
influences patch dynamics in stream benthic algal communities. Oecologia
155:809-819

Lee S, Fuhrman JA (1991) Spatial and temporal variation of natural bacterioplankton
assemblages studied by total genomic DNA cross-hybridization. Limnol.
Oceanogr. 36:1277-1287

Lee S, Kang YC, Fuhrman JA (1995) Imperfect Retention of Natural Bacterioplankton
Cells By Glass Fiber Filters. Mar Ecol-Prog Ser 119:285-290

Legendre P, Legendre L (1998) Numerical Ecology. In: Developments in
Environmental Modelling 20. Elsevier, Amsterdam, p 853

Liss PS, Duce RA (eds) (1997) The Sea Surface and Global Change, Vol. Cambridge
University Press, Cambridge, UK

Liss PS, Duce RA (eds) (2005) The Sea Surface and Global Change, Vol. Cambridge
University Press, Cambridge, UK

Long RA, Azam F (2001a) Antagonistic interactions among marine pelagic bacteria.
Appl Environ Microb 67:4975-4983

Long RA, Azam F (2001b) Microscale patchiness of bacterioplankton assemblage
richness in seawater. Aquat Microb Ecol 26:103-113

Luna GM, Dell'Anno A, Giuliano L, Danovaro R (2004) Bacterial diversity in deep
Mediterranean sediments: relationship with the active bacterial fraction and
substrate availability. Environ Microbiol 6:745-753

228

Maki JS (1993) The air-water interface as an extreme environment. In: Ford TE (ed)
Aquatic microbiology—an ecological approach. Blackwell Scientific, Boston, p
409–440

Mantel N (1967) Detection of Disease Clustering and a Generalized Regression
Approach. Cancer Research 27:209-&

Moeseneder MM, Arrieta JM, Muyzer G, Winter C, Herndl GJ (1999) Optimization of
terminal-restriction fragment length polymorphism analysis for complex marine
bacterioplankton communities and comparison with denaturing gradient gel
electrophoresis. Appl Environ Microbiol 65:3518-3525

Montoya JM, Pimm SL, Sole RV (2006) Ecological networks and their fragility.
Nature 442:259-264

Morris RM, Vergin KL, Cho JC, Rappe MS, Carlson CA, Giovannoni SJ (2005)
Temporal and spatial response of bacterioplankton lineages to annual
convective overturn at the Bermuda Atlantic Time-series Study site. Limnol
Oceanogr 50:1687-1696

Munster U, Heikkinen E, Knulst J (1998) Nutrient composition, microbial biomass and
activity at the air-water interface of small boreal forest. Hydrobiologia 363:261-
270

Murray AE, Blakis A, Massana R, Strawzewski S, Passow U, Alldredge A, DeLong EF
(1999) A time series assessment of planktonic archaeal variability in the Santa
Barbara Channel. Aquat Microb Ecol 20:129-145

Murray AE, Preston CM, Massana R, Taylor LT, Blakis A, Wu K, DeLong EF (1998)
Seasonal and spatial variability of bacterial and archaeal assemblages in the
coastal waters near Anvers Island, Antarctica. Appl Environ Microbiol
64:2585-2595

Nelson JD, Boehme SE, Reimers CE, Sherrell RM, Kerkhof LJ (2008) Temporal
patterns of microbial community structure in the Mid-Atlantic Bight. Fems
Microbiol Ecol 65:484 - 493

Noble M, Jones B, Hamilton P, Xu JP, Robertson G, Rosenfeld L, Largier J (2009)
Cross-shelf transport into nearshore waters due to shoaling internal tides in San
Pedro Bay, CA. Cont Shelf Res 29:1768-1785

229

Noble RT, Fuhrman JA (1997) Virus decay and its causes in coastal waters. Appl.
Environ. Microbiol. 63:77-83

Noble RT, Fuhrman JA (1998) Use of SYBR Green I rapid epifluoresence counts of
marine viruses and bacteria. Aquatic Microbial Ecology 14:113-118

Norkrans B (1980) Surface microlayers in aquatic environments. In: Alexander M (ed)
Advances in microbial ecology. Plenum Press, New York and London, p 51–83

Obernosterer I, Catala P, Lami R, Caparros J, Ras J, Bricaud A, Dupuy C, van
Wambeke F, Lebaron P (2008) Biochemical characteristics and bacterial
community structure of the sea surface microlayer in the South Pacific Ocean.
Biogeosciences 5:693-705

Obernosterer I, Catala P, Reinthaler T, Herndl GJ, Lebaron P (2005) Enhanced
heterotrophic activity in the surface microlayer of the Mediterranean Sea. Aquat
Microb Ecol 39:293-302

O’Donnell J, Allen AA, Murphy DL (1997) An assessment of the errors in Lagrangian
velocity estimates obtained by FGGE drifters in the Labrador Current Journal
of Atmospheric and Oceanic Technology 14:292–307

Olesen JM, Bascompte J, Dupont YL, Jordano P (2007) The modularity of pollination
networks. P Natl Acad Sci USA 104:19891-19896

Orphan VJ, House CH, Hinrichs KU, McKeegan KD, DeLong EF (2001) Methane-
consuming archaea revealed by directly coupled isotopic and phylogenetic
analysis. Science 293:484-487

Pace NR, Stahl DA, Lane DL, Olsen GJ (1986) The analysis of natural microbial
populations by rRNA sequences. Adv. Microbiol. Ecol. 9:1-55

Pastor-Satorras R, Vespignani A (2001) Epidemic spreading in scale-free networks.
Physical Review Letters 86:3200-3203

Patel A, Noble RT, Steele JA, Schwalbach MS, Hewson I, Fuhrman JA (2007) Virus
and prokaryote enumeration from planktonic aquatic environments by
epifluorescence microscopy with SYBR Green I. Nat Protoc 2:269-276

Pedros-Alio C (2006) Marine microbial diversity: can it be determined? Trends in
Microbiology 14:257-263

230

Pomeroy LR, Williams PJI, Azam F, Hobbie JE (2007) The Microbial Loop.
Oceanography 20:28-33

Pommier T, Canback B, Riemann L, Bostrom KH, Simu K, Lundberg P, Tunlid A,
Hagstrom A (2007) Global patterns of diversity and community structure in
marine bacterioplankton. Molecular Ecology 16:867-880

Prosser JI, Bohannan BJM, Curtis TP, Ellis RJ, Firestone MK, Freckleton RP, Green
JL, Green LE, Killham K, Lennon JJ, Osborn AM, Solan M, van der Gast CJ,
Young JPW (2007) The role of ecological theory in microbial ecology. Nat Rev
Micro 5:384-392

Ramette A, Tiedje JM (2007a) Biogeography: an emerging cornerstone for
understanding prokaryotic diversity, ecology, and evolution. Microb Ecol
53:197-207

Ramette A, Tiedje JM (2007b) Multiscale responses of microbial life to spatial distance
and environmental heterogeneity in a patchy ecosystem. Proc Natl Acad Sci U
S A 104:2761-2766

Regan JD, Carrier WL, Gucinski H, Olla BL, Yoshida H, Fujimura RK, Wicklund RI
(1992) DNA as a solar dosimeter in the ocean. Photochem Photobiol 56:35-42

Reinthaler T, Sintes E, Herndl GJ (2008) Dissolved organic matter and bacterial
production and respiration in the sea-surface microlayer of the open Atlantic
and the western Mediterranean Sea. Limnol Oceanogr 53:122-136

Riemann L, Steward GF, Fandino LB, Campbell L, Landry MR, Azam F (1999)
Bacterial community composition during two consecutive NE Monsoon periods
in the Arabian Sea studied by denaturing gradient gel electrophoresis (DGGE)
of rRNA genes. Deep-Sea Res Pt Ii 46:1791-1811

Rocap G, Distel DL, Waterbury JB, Chisholm SW (2002) Resolution of
Prochlorococcus and Synechococcus ecotypes by using 16S-23S ribosomal
DNA internal transcribed spacer sequences. Applied & Environmental
Microbiology 68:1180-1191

Rodriguez F, Derelle E, Guillou L, Le Gall F, Vaulot D, Moreau H (2005) Ecotype
diversity in the marine picoeukaryote Ostreococcus (Chlorophyta,
Prasinophyceae). Environ Microbiol 7:853-859

231

Rosenzweig ML (1995) Species diversity in space and time. Cambridge University
Press. Cambridge, UK.

Ruan QS, Dutta D, Schwalbach MS, Steele JA, Fuhrman JA, Sun FZ (2006) Local
similarity analysis reveals unique associations among marine bacterioplankton
species and environmental factors. Bioinformatics 22:2532-2538

Ruan Q, Steele JA, Schwalbach MS, Fuhrman JA, Sun FZ (2006) A dynamic
programming algorithm for binning microbial community profiles.
Bioinformatics 22:1508-1514

Rusch DB, Halpern AL, Sutton G, Heidelberg KB, Williamson S, Yooseph S, Wu DY,
Eisen JA, Hoffman JM, Remington K, Beeson K, Tran B, Smith H, Baden-
Tillson H, Stewart C, Thorpe J, Freeman J, Andrews-Pfannkoch C, Venter JE,
Li K, Kravitz S, Heidelberg JF, Utterback T, Rogers YH, Falcon LI, Souza V,
Bonilla-Rosso G, Eguiarte LE, Karl DM, Sathyendranath S, Platt T,
Bermingham E, Gallardo V, Tamayo-Castillo G, Ferrari MR, Strausberg RL,
Nealson K, Friedman R, Frazier M, Venter JC (2007) The Sorcerer II Global
Ocean Sampling expedition: Northwest Atlantic through Eastern Tropical
Pacific. Plos Biol 5:398-431

Schauer M, Massana R, Pedros-Alio C (2000) Spatial differences in bacterioplankton
composition along the Catalan coast (NW Mediterranean) assessed by
molecular fingerprinting. Fems Microbiol Ecol 33:51-59

Schwalbach MS, Hewson I, Fuhrman JA (2004) Viral effects on bacterial community
composition in marine plankton microcosms. Aquat Microb Ecol 34:117-127

Seymour JR, Mitchell JG, Seuront L (2004) Microscale heterogeneity in the activity of
coastal bacterioplankton communities. Aquat Microb Ecol 35:1-16

Seymour JR, Patten N, Bourne DG, Mitchell JG (2005) Spatial dynamics of virus-like
particles and heterotrophic bacteria within a shallow coral reef system. Mar
Ecol-Prog Ser 288:1-8

Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N,
Schwikowski B, Ideker T (2003) Cytoscape: A software environment for
integrated models of biomolecular interaction networks. Genome Res 13:2498-
2504

Sherr BF, Sherr EB, Caron DA, Vaulot D, Worden AZ (2007) Oceanic Protists.
Oceanography 20:130-134
232

Sherr EB, Sherr BF (2008) Understanding roles of microbes in marine pelagic food
webs: a brief history. In: Kirchman D (ed) Advances in Microbial Ecology of
the Oceans. Wiley, p 27-44

Sieburth JM (1971) Distribution and activity of oceanic bacteria. Deep Sea Res
18:1111-1121

Sieburth JM (1983) Microbiological and organic-chemical processes in the surface and
mixed layers. In: Liss PS, Slinn WGN (eds) Air-Sea Exchange of Gases and
Particles. Reidel Publishers Co, Hingham, MA, p 121–172

Sieburth JM, Conover JT (1965) Slicks Associated with Trichodesmium Blooms in
Sargasso Sea. Nature 205:830-&

Simek K, Pernthaler J, Weinbauer MG, Hornak K, Dolan JR, Nedoma J, Masin M,
Amann R (2001) Changes in bacterial community composition and dynamics
and viral mortality rates associated with enhanced flagellate grazing in a
mesoeutrophic reservoir. Appl Environ Microb 67:2723-2733

Simon M, Azam F (1989) Protein content and protein synthesis rates of planktonic
marine bacteria. Marine Ecology Progress Series 51:201-213

Sogin ML, Morrison HG, Huber JA, Welch DM, Huse SM, Neal PR, Arrieta JM,
Herndl GJ (2006) Microbial diversity in the deep sea and the underexplored
"rare biosphere". P Natl Acad Sci USA 103:12115-12120

Stolle C, Nagel K, Labrenz M, Jürgens K (2009) Bacterial activity in the sea-surface
microlayer: in situ investigations in the Baltic Sea and the influence of
sampling devices. Aquat Microb Ecol 58:67–78

Storey JD (2002) A direct approach to false discovery rates. J Roy Stat Soc B 64:479-
498

Suttle CA, Chen F (1992) Mechanisms and rates of decay of marine viruses in
seawater. Appl. Environ. Microbiol. 58:3721-3729

Terbraak CJF, Verdonschot PFM (1995) Canonical Correspondence-Analysis and
Related Multivariate Methods in Aquatic Ecology. Aquat Sci 57:255-289

Thingstad TF, Lignell R (1997) Theoretical models for the control of bacterial growth
rate, abundance, diversity and carbon demand. Aquat. Microb. Ecol. 13:19-27

233

Torsvik V, Ovreas L, Thingstad TF (2002) Prokaryotic diversity - Magnitude,
dynamics, and controlling factors. Science 296:1064-1066

Treusch AH, Vergin KL, Finlay LA, Donatz MG, Burton RM, Carlson CA,
Giovannoni SJ (2009) Seasonality and vertical structure of microbial
communities in an ocean gyre. Isme J 3:1148-1163

Tyson GW, Chapman J, Hugenholtz P, Allen EE, Ram RJ, Richardson PM, Solovyev
VV, Rubin EM, Rokhsar DS, Banfield JF (2004) Community structure and
metabolism through reconstruction of microbial genomes from the
environment. 428:37-43

Upstill-Goddard RC, Frost T, Henry GR, Franklin M, Murrell JC, Owens NJP (2003)
Bacterioneuston control of air-water methane exchange determined with a
laboratory gas exchange tank. Global Biogeochemical Cycles 17

van der Gast CJ, Jefferson B, Reid E, Robinson T, Bailey MJ, Judd SJ, Thompson IP
(2006) Bacterial diversity is determined by volume in membrane bioreactors.
Environ Microbiol 8:1048-1055

van der Gast CJ, Lilley AK, Ager D, Thompson IP (2005) Island size and bacterial
diversity in an archipelago of engineering machines. Environ Microbiol 7:1220-
1226

Van Vleet ES, Williams PM (1980) Sampling sea surface films: A laboratory
evaluation of techniques and collecting materials. Limnology & Oceanography
25:764-770

Venter JC, Remington K, Heidelberg JF, Halpern AL, Rusch D, Eisen JA, Wu DY,
Paulsen I, Nelson KE, Nelson W, Fouts DE, Levy S, Knap AH, Lomas MW,
Nealson K, White O, Peterson J, Hoffman J, Parsons R, Baden-Tillson H,
Pfannkoch C, Rogers YH, Smith HO (2004) Environmental genome shotgun
sequencing of the Sargasso Sea. Science 304:66-74

Vigil P, Countway PD, Rose J, Lonsdale DJ, Gobler CJ, Caron DA (2009) Rapid shifts
in dominant taxa among microbial eukaryotes in estuarine ecosystems. Aquat
Microb Ecol 54:83-100

Ward BB (2002) How many species of prokaryotes are there? P Natl Acad Sci USA
99:10234-10236

234

Weinbauer MG (2004) Ecology of prokaryotic viruses. Fems Microbiol Rev 28:127-
181

Weinbauer MG, Brettar I, Hofle MG (2003) Lysogeny and virus-induced mortality of
bacterioplankton in surface, deep, and anoxic marine waters. Limnol Oceanogr
48:1457-1465

West NJ, Scanlan DJ (1999) Niche-partitioning of Prochlorococcus populations in a
stratified water column in the eastern North Atlantic Ocean. Applied &
Environmental Microbiology 65:2585-2591

Wilhelm SW, Weinbauer MG, Suttle CA, Jeffrey WH (1998) The role of sunlight in
the removal and repair of viruses in the sea. Limnol. Oceanogr. 43:586-592

Williams PM, Carlucci AF, Henrichs SM, Vanvleet ES, Horrigan SG, Reid FMH,
Robertson KJ (1986) Chemical and Microbiological Studies of Sea-Surface
Films in the Southern Gulf of California and Off the West-Coast of Baja-
California. Marine Chemistry 19:17-98

Winter C, Smit A, Herndl GJ, Weinbauer MG (2005) Linking bacterial richness with
viral abundance and prokaryotic activity. Limnol Oceanogr 50:968-977

Woodcock S, Curtis TP, Head IM, Lunn M, Sloan WT (2006) Taxa-area relationships
for microbes: the unsampled and the unseen. Ecol Lett 9:805-812

Wurl O, Holmes M (2008) The gelatinous nature of the sea-surface microlayer. Marine
Chemistry 110:89-97

Yang GP, Tsunogai S, Watanabe S (2005) Biogeochemistry of
dimethylsulfoniopropionate (DMSP) in the surface microlayer and subsurface
seawater of Funka Bay, Japan. Journal of Oceanography 61:69-78

Yang GP, Watanabe S, Tsunogai S (2001) Distribution and cycling of dimethylsulfide
in surface microlayer and subsurface seawater. Marine Chemistry 76:137-153

Zhengbin Z, Liansheng L, Zhijian W, Jun L, Haibing D (1998 ) Physicochemical
studies of the sea surface microlayer. J Colloid Interface Sci 204:294–299

Zhou L, Rocke DM (2005) An expression index for Affymetrix GeneChips based on
the generalized logarithm. Bioinformatics 21:3983-3989

Abstract (if available)

Abstract Microbes are central to ocean food webs and global biogeochemical processes. Understanding their distribution and interactions over space and time is fundamental to understanding their ecology. Even with the advent of molecular techniques, it is difficult to observe their interactions with each other and with their environment. We are in the early stages of expanding our focus from simply identifying microbes, to discovering what they are doing (i.e. how each microorganism interacts and fits within the functioning of the ecosystem). In the ocean, which is in constant lateral and vertical motion, these community patterns and interactions may change on short to medium spatial and temporal scales. The aim of this study was to determine the community patterns in bacterioplankton in the ocean surface waters on hour-day temporal scales and kilometer spatial scales, patterns in bacterial communities on µm to m depth scales, and interactions among bacteria, protists and Archaea on monthly scales. Molecular fingerprinting techniques coupled with clone libraries for identification were used to study changes in bacterioplankton communities in the surface waters from 2000-2005, in the sea surface microlayer and surface layer from 2004-2006, and interactions between bacterioplankton and protists from 2000-2004 in the ocean near Southern California. Investigating bacterioplankton community patterns over km spatial scales and hour-day temporal scales, we found coherent communities at 2-97km2 area scales and at distances along transects of 0-15km, and the major taxa appeared remarkably consistent throughout, while at distances from 50-255km, there were strong differences in the communities which correlated to environmental parameters. Temporally, we found stable communities within a water patch and in the surface mixed layer over 20-30 hours, and also high community similarity between days over a week at a single geographic location.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

Patterns of molecular microbial activity across time and biomes

PDF

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

PDF

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

PDF

Plant-microbial interactions in mangrove sediments under different nutrient conditions

PDF

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

PDF

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

PDF

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

PDF

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

PDF

The microbiome of gorgonian octocorals, Muricea, with a description of a novel, photosynthetic protistan symbiont

PDF

Chemical weathering across spatial and temporal scales: from laboratory experiments to global models

PDF

Predictive habitat distribution modeling of sperm whale (Physeter macrocephalus) within the central Gulf of Alaska utilizing passive acoustic monitoring

PDF

Concentration and size partitioning of trace metals in surface waters of the global ocean and storm runoff

PDF

Phytoplankton bloom initiation in the Southern California Bight: a multi-year local and regional analysis

PDF

Germanium and silicon isotope geochemistry in terrestrial and marine low-temperature environments

Asset Metadata

Creator Steele, Joshua Adam (author)

Core Title Marine bacterioplankton biogeography over short to medium spatio-temporal scales

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Marine and Environmental Biology

Publication Date 05/05/2010

Defense Date 01/22/2010

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

Tag association networks,biogeography,biological oceanography,local similarity analysis,marine bacterioplankton,microbial ecology,OAI-PMH Harvest,sea surface microlayer,spatio-temporal patterns,taxa-area relationship,time-series

Place Name bights: Southern California (geographic subject), California (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Fuhrman, Jed A. (committee chair), Capone, Douglas G. (committee member), Devinny, Joseph (committee member), Ziebis, Wiebke (committee member)

Creator Email joshuast@usc.edu,whooshaway@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3002

Unique identifier UC155937

Identifier etd-Steele-3490 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-404506 (legacy record id),usctheses-m3002 (legacy record id)

Legacy Identifier etd-Steele-3490.pdf

Dmrecord 404506

Document Type Dissertation

Rights Steele, Joshua Adam

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu