GENERALIST AND SPECIALIST STRATEGIES OF PHOSPHORUS ACQUISITION
BY AQUATIC BACTERIA
By
Kali Bird

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Microbiology and Molecular Genetics
2012

ABSTRACT
GENERALIST AND SPECIALIST STRATEGIES OF PHOSPHORUS ACQUISITION BY AQUATIC
BACTERIA
By
Kali Bird
Resource heterogeneity increases biological diversity by providing opportunity for
niche partitioning and resource specialization. Organisms which use few of the available
resource forms are considered specialists, while those which use many resource forms
are considered generalists. The relative proportion of specialists and generalists within
a community impacts ecosystem functions, such as total productivity. Being a resource
specialist or generalist may come with a fitness cost or favor performance tradeoffs. For
example, generalists may suffer a fitness cost for maintaining a broad ecological niche.
Heterotrophic microbes and primary producers have the potential to specialize on
different chemical forms of essential nutrients such as nitrogen and phosphorus, yet
there have been few studies of nutrient specialization, limiting our understanding of
associated costs or performance tradeoffs. In the present study, we quantified
phosphorus resource specialization by aquatic bacterial isolates and tested for a
specialization-performance tradeoff, using bacterial growth rate as the measure of
performance. We found evidence for bacterial specialization on phosphorus form and
for an environment-specific specialization-growth rate tradeoff. Our results indicate that
nutrient-based resource specialization can strongly influence an important performance
trait of an organism, but these affects may be environment-specific. Results from this
study improve our understanding of how a species’ niche breadth may impact its
ecological strategies and competitive outcomes.

ACKNOWLEDGEMENTS

As with all accomplishments, completion of this thesis was a collaborative effort.
The research could not have been completed nor the results communicated effectively
without the contributions of dozens of people. First and foremost, I would like to thank
my advisor Jay Lennon, both for his guidance and his patience. I have never been
traditional in any way, and as his first graduate student, I’m sure my fierce
independence was a challenge. Jay has always been understanding, encouraging,
insightful, and a privilege to work with, for which I am truly thankful.
I also give abundant thanks to Steve Hamilton, Todd Barkman, Colin Kremer,
Mridul Thomas, and Kyle Edwards for their many hours of statistical guidance,
thoughtful questions about my study system, and most importantly their friendship. I am
also grateful to the Hamilton lab for the lake nutrient data contributed for this thesis.
Tom Schmidt and Jim Tiedje provided valuable direction and feedback throughout the
development of this project. Brent Lehmkhul was an invaluable resource for molecular
knowledge and lab techniques, and generally a joy to see in the lab every day. Pam
Woodruff, Allyson Hutchins, and Dave Weed provided lab work guidance and
assistance. I also thank the grad students and post-docs of the Lennon lab for their
unique expertise and eager help – Stuart Jones, Zach Aanderud, Megan Larsen, and
Mario Muscarella – thank you!
Finally, I’d like to thank my parents Steve and Cappy Bird, my brother and sister
Tony Hernandez and KC Bird, Jarad Mellard, Dan Sorensen, and the broader KBS

iii

community for their continuous generosity and support. I am who I am because of you,
and this project would not be completed without you.

iv

TABLE OF CONTENTS

LIST OF TABLES……………………………………………………………………………….v
LIST OF FIGURES………….………………………………………………………………….vi
CHAPTER 1
Introduction………….……………………………………………………………………...1
References………….……………………………………………………………………...7
CHAPTER 2
GENERALIST AND SPECIALIST STRATEGIES OF PHOSPHORUS ACQUISITION BY AQUATIC
BACTERIA
Abstract….………………………………………………………………………………..11
Introduction………….…………………………………………………………………….12
Materials and Methods……………………………………………………………………..15
Results………….………………………………………………………………….……..24
Discussion………….……………………………………………………………………..33
References………….…………………………………………………………………….58
CHAPTER 3
FUTURE
DIRECTIONS…………………………………………………………..…………………………69
APPENDICES
Supplementary Figures………….………………………………...………………….……38
Supplementary Methods………….…………………………………..……………………55

v

LIST OF TABLES

Table 1-1: Organophosphate utilization enzymes of bacteria. Shown are many common
bacterial enzymes, primary genes or gene clusters which encode the enzymes, and the
phosphorus resource(s) they target.…………………………………………………………..4
Table 2-1: Lake attributes. Nutrient data was collected several times each year for three
years (2007-2009). Shown are means ± one standard deviation.………….……………25
Table 2-2: Table 2. Phosphorus source abbreviations and properties. …………………29
Table A-1: Influence of phylogenetic history on phosphorus-use traits. ML = Maximum
likelihood. For analysis details and compound abbreviations, see Methods section and
Table 2-2 from text, respectively. Within the software program BayesTraits, one can test
whether a phylogeny correctly predicts species trait covariances by incorporating the
parameter Lambda (λ) in analyses (Pagel 1997, Pagel 1999). Significant P-values for λ=
0 vs. ML comparisons indicate that phylogenetic history at least minimally influences the
trait. Significant P-values for λ= 1 vs. ML comparisons indicate that the trait is not
perfectly correlated with the phylogeny. In this study, we considered traits to be
phylogenetically conserved if >95% of generated phylogenetic trees statistically support
phylogenetic conservatism (α= 0.05). For example, a trait that perfectly correlates with
the phylogeny would yield a ‘100’ in the first column and a ‘0’ in the second; while one
that is minimally influenced by phylogenetic history would yield a 95-100 in the first
column and a value 5 or greater in the second. Values supporting phylogenetic
conservatism are starred.……………….………………………..…………………………..41

vi

LIST OF FIGURES

FIGURE 1-1. Examples of organic phosphorus bonding types. Many phosphorus
resources are bound up in organic forms. Phosphonates contain stable C-P bonds
(circled in blue), while phosphate esters contain more labile C-O-P bonds (circled in
yellow). Monoesters have one C-O-P bond, while di-esters (circled in green) or triesters (not shown) have two or three C-O-P bonds, respectively. Inorganic phosphate
ion (“free phosphate”) and a simple polyphosphate with phosphorus anhydride bonds
(shown with red curves) are included for comparison. For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic
version of this thesis.……………………………………………………………………………5
FIGURE 2-1. Ranked growth rates for each isolate. Each line represents one isolate’s
scaled growth rates in decreasing order of magnitude. Blue and green lines represent
isolates from LL Lake and WG Lake, respectively. Isolates with shallower initial slopes
grow relatively well on many P sources (‘generalists’); while those with steep initial
slopes grow quickly on one or a few P sources and slowly on others
(‘specialists’)..…………………………………………………………….……………….……26
FIGURE 2-2. Relationship between P-niche breadth and bacterial growth rate. Log10transformed maximum growth rates are plotted against the Levins index for each
isolate. Open circles represent isolates from LL Lake. Closed circles represent isolates
from WG Lake. Projected slopes in this figure include phylogenetic correction. R2 of
regression with phylogenetic correction is 0.30. R2 of regression without phylogenetic
correction is 0.38.………………………………………………………………………….…..27
Figure 2-3. Figure 2-3. Heatmap visually displaying isolate maximum growth rates,
scaled from 0-1. Darker colors represent higher values. The dendrogram clusters
isolates and P sources according to these growth rates using the Euclidean method to
calculate distances. Isolates are named according to the genus of each as identified
using the Classifier tool provided by the Ribosomal Database Project (RDP, Wang et al.
2007), numbered when multiple from one genus occur, and are labeled with an “L” or
“W” indicating from which lake they were isolated (LL Lake or WG Lake, respectively). P
compound abbreviations are found in Table 2...………………………………….…….….30
Figure 2-4. P sources clustered according to similarity of P use traits across isolates.
The Manhattan method was used to calculate the distance matrix, and the Group
Average method with 10,000 bootstrapped permutations was used to cluster the
isolates. Green and red numbers indicate the bootstrap probability and the
‘Approximately Unbiased’ (AU) probability that the cluster exists. Red boxes surround
groups for which the AU p-value is less than 0.05, indicating that we should reject the

vii

null hypothesis that the clusters do not exist. Thus the compounds within the red boxes
are not distinguishable from one another using these metrics.…………………….……32
Figure A-1. Phylogeny of bacterial isolates used in this study, with reference
sequences. As in Figure 2-3, isolates from the present study are named according to
the genus of each, as determined from RDP classification, numbered when multiple
from one genus occur, and are labeled with an “L” or “W” indicating from which lake
they were isolated (LL Lake or WG Lake, respectively). They are also color-coded by
lake—LL Lake isolates are blue and WG Lake isolates are green. Reference sequences
(in black) are named according to their RDP or genbank classification, when an RDP
classification was unavailable. All reference sequences are also labeled with their
Genbank identifier.…………………………………………………………………………….39
Figure A-2. Isolate 16S DNA sequences used for phylogenetic analyses. Sequences
are shown in FASTA format.…………….……….………………….……………….………42

viii

CHAPTER 1
INTRODUCTION

An organism’s ecological niche can be described both in terms of its resource
requirements as well as the way its activities influence its environment (Chase & Leibold
2003). In theory, species fill finite quantities of resource space according to their traits.
Some species are considered to be ecological ‘specialists,’ having a narrow niche
breadth and relatively stringent ecological or environmental requirements to satisfy their
resource needs, while ecological ‘generalists,' have a broad niche breadth with respect
to the way they meet their resource requirements. Communities comprised primarily of
specialists may maximize ecosystem resource usage through functional
complementarity (Loreau 2001), while generalists may play important roles in
maintaining ecosystem stability and functions (Richmond et al. 2005; Mou et al. 2008).
Research suggests that neutral processes may dominate species distributions in some
systems (Hubbell 2001), but niche partitioning cannot be ruled out as an important
driver of community composition and ecosystem functions in many systems (Levine &
HilleRisLambers 2009).
Not only macroorganisms, but microorganisms too can be described by their
ecological niche and demonstrate niche specialization. However, microbial interactions
occur at the micrometer scale and smaller, so while 'seed size' may be an ecologically
relevant food preference for a bird, 'molecule structure' may be a more ecologically
relevant food preference for microorganisms. For example, Upton and Nedwell

1

compared the abilities of oligotrophic and copiotrophic bacteria to use a suite of carbon
sources for growth (Upton & Nedwell 1989). They found that oligotrophic bacteria were
able to use more carbon substrates, thus demonstrating a broader niche breadth.
Similarly, Mou et al. compared bacterial communities’ carbon niche breadth in a salt
marsh when supplemented with one of two carbon sources (Mou et al. 2008). Using
DNA-based methods, they found that most bacteria in their study tended to be
generalists. Microbes have also been shown to specialize on other resources such as
light (Stomp et al. 2004).
Microbial consumption and transformation of nutrient resources affect global
processes, such as oceanic primary productivity, biomass transfer, and nitrogen fixation
(Falkowski et al. 2008). As a frequently limiting nutrient in aquatic ecosystems,
phosphorus (P) resources hold a key to ecosystem productivity and functions (Dyhrman
et al. 2007). Inorganic phosphate (Pi) concentrations can be as low as <30 pM in
freshwater environments and <50 nM in the ocean (Karl 2000, Bjórkman & Karl 1994,
Hudson et al. 2000). Organic P (Porg) concentrations are typically much greater, since
many of the compounds that comprise this pool require hydrolytic enzymes for
organisms to access the P. As potentially better competitors for Porg than eukaryotic
phytoplankton, bacteria may control the quantity of P available to eukaryotic
phytoplankton and ultimately primary productivity in some ecosystems (Currie & Kalff
1984, Coveney & Wetzel 1992, Cotner & Biddanda 2002). Excess P release into
surface waters promotes lake eutrophication, which can lead to toxic algal blooms, fish
kills, reduction in recreational value, and decreased drinking water quality (Carpenter et

2

al. 1998). Since microbial communities are essential intermediaries in the uptake and
transformation of these P resources, further research into the processes that control
microbial P transformations is needed as we develop strategies for remediation of
eutrophied waterbodies.
Bacteria employ many strategies to acquire P, such as expressing high- and lowaffinity P-uptake proteins, and secreting and excreting phosphatases. Perhaps the most
well studied mechanisms for accessing Pi are the low-affinity, constitutive Pit system
and the high-affinity, Pi-repressible Pst system expressed in Escherichia coli. To access
P from Porg, bacteria maintain a genetic arsenal of P-acquisition enzymes. Many of
these enzymes and their encoding gene or gene clusters can be found in Table 1-1.
Bacteria commonly use nonspecific acid or alkaline phosphatases, which cleave P from
phosphomonoesters. These enzymes may be attached to the cell membrane, contained
within the cytoplasm or periplasm, or excreted into the environment (Luo et al. 2009;
White A. 2009). Bacteria may also utilize substrate-specific enzymes, such as phytases,
which cleave phosphomonoesters from bulky phytate compounds, or phosphonatases,
which cleave C-P bonds in phosphonate compounds (Table 1-1 and Figure 1-1). In
addition, some bacteria are able to take up certain small molecules in their entirety,
such as adenosine monophosphate (AMP), cyclic adenosine monophosphate (cAMP),
and glycerophosphoric acid (White 2009 and references therein). Figure 1-1 displays
several examples of Porg bond types and indicates enzymes that bacteria frequently
use to cleave P from a variety of resource forms.

3

Table 1-1. Organophosphate utilization enzymes of bacteria. Shown are many common bacterial enzymes,
primary genes or gene clusters which encode the enzymes, and the phosphorus resource(s) they target.
Enzyme

Encoding gene(s)
or gene clusters

Primary substrate(s)
targeted

References

Alkaline
phosphatase

phoA, phoD, phoX

phosphorus esters

Luo et al. 2009

Acidic
Phosphatase

appA

phosphorus esters

Vershinina &
Znamenskaya 2002

Phytases

phy

phytate

Lim et al. 2007

C-P lyase

phn gene cluster

many phosphonates

Huang et al. 2005

Phosphonatase

phnW, phnX

primarily
2-aminoethylphosphonate

Huang et al. 2005

Polyphosphatase

ppK

polyphosphate

Vershinina &
Znamenskaya 2002

Phosphonoacetate
hydrolase

phnA

phosphonoacetate

Gilbert et al. EM 2009

5’-Nucleotidase

nuc

5’ -nucleotides

Vershinina &
Znamenskaya 2002

4

Phosphonate (C-P) bond

Inorganic phosphate ion

(2-aminoethyl) phosphonic acid

Phosphomonoester (R-O-PO3)

Phosphorus anhydride
bonds (PO3-O-PO3)

Short polyphosphate chain

Adenosine triphosphate

Phosphodiester (R-O-PO2-O-R)

Cyclic adenosine monophosphate
Figure 1-1. Examples of organic phosphorus bonding types. Many phosphorus resources are bound up in
organic forms. Phosphonates contain stable C-P bonds (circled in blue), while phosphate esters contain

5

Figure 1-1 (cont’d)
more labile C-O-P bonds (circled in yellow). Monoesters have one C-O-P bond, while di-esters (circled in
green) or tri-esters (not shown) have two or three C-O-P bonds, respectively. Inorganic phosphate ion
(“free phosphate”) and a simple polyphosphate with phosphorus anhydride bonds (shown with red curves)
are included for comparison. For interpretation of the references to color in this and all other figures, the
reader is referred to the electronic version of this thesis.

6

REFERENCES

7

REFERENCES

Bjórkman, K. & Karl, D. (1994). Bioavailability of inorganic and organic phosphorus
compounds to natural assemblages of microorganisms in Hawaiian coastal waters. Mar
Ecol-Prog Ser. 111, 265-273.
Carpenter, S.R., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N., and
Smith, V. H. (1998). Nonpoint pollution of surface waters with phosphorus and nitrogen.
Ecol. Appl. 8, 559-568.
Cotner, J.B. & Biddanda, B.A. (2002). Small Players, Large Role: Microbial Influence on
Biogeochemical Processes in Pelagic Aquatic Ecosystems. Ecosystems 5, 105-121.
Coveney, M.F. & Wetzel, R.G. (1992). Effects of nutrients on specific growth rate of
bacterioplankton in oligotrophic lake water cultures. Appl. Environ. Microbiol., 58, 150156.
Currie, D.J. & Kalff, J. (1984). A comparison of the abilities of freshwater algae and
bacteria to acquire and retain phosphorus. Limnol. Oceanogr. 29, 298-310.
Falkowski, P.G., Fenchel, T. & Delong, E.F. (2008). The microbial engines that drive
Earth’s biogeochemical cycles. Science, 320, 1034-9.
Gilbert, J.A. et al. (2009). Potential for phosphonoacetate utilization by marine bacteria
in temperate coastal waters. Environ. Microbiol., 11, 111-125.
Hubbell, S. P. 2001. The Unified Neutral Theory of Biodiversity and Biogeography.
Princeton University Press, Princeton, NJ. USA.
Hudson, J.J., Taylor, W.D. & Schindler, D.W. (2000). Phosphate concentrations in
lakes. Nature. 406, 54-56.
Levine, M.J. & HilleRisLambers, J. (2009). The importance of niches for the
maintenance of species diversity. Nature, 461, 254-257.

8

Lim, B.L., Yeung, P., Cheng, C. & Hill, J.E. (2007). Distribution and diversity of phytatemineralizing bacteria. ISME J, 1, 321-330.
Loreau, M. (2001). Microbial diversity, producer–decomposer interactions and
ecosystem processes: a theoretical model. Proc. R. Soc. Lond. B, 268, 303-309.
Luo, H., Benner, R. Long, R.A., & Hu, J. (2009). Subcellular localization of marine
bacterial alkaline phosphatases. Proc. Natl. Acad. Sci. USA, 106, 21219-21223.
Mou, X., Sun, S., Edwards, R.A., Hodson, R.E. & Moran, M.A. (2008). Bacterial carbon
processing by generalist species in the coastal ocean. Nature 451, 708-711.
Stomp, M., Huisman, J., deJongh, F., Veraart, A.J., Gerla, D., Rijkeboer, M., et al.
(2004). Adaptive divergence in pigment composition promotes phytoplankton
biodiversity. Nature, 432, 104-107.
Upton, A.C. & Nedwell, D.B. (1989). Nutritional flexibility of oligotrophic and copiotrophic
Antarctic bacteria with respect to organic substrates. FEMS Microbiol Ecol., 62, 1-6.
Vershinina, O.A. & Znanenskaya, L.V. (2002). The pho regulons of bacteria. Microbiol.,
71, 497-511.
White, A.E. (2009). New insights into bacterial acquisition of phosphorus in the surface
ocean. Proc. Natl. Acad. Sci. USA, 106, 21013-21014.

9

CHAPTER 2
GENERALIST AND SPECIALIST STRATEGIES OF PHOSPHORUS ACQUISITION
BY AQUATIC BACTERIA

Abstract
Resource heterogeneity provides opportunity for ecological specialization.
Organisms that use few of the available resources are specialists, while those that are
capable of using many resources are generalists. Theory predicts that there are costs
and tradeoffs with being a specialist or generalist, the magnitude of which may depend
on environmental conditions. For example, specialization is considered to be most
advantageous in homogeneous environments with abundant resources, where
generalists may suffer a large fitness cost for maintaining a broad ecological niche.
Although there is evidence that microorganisms have the potential to specialize
on different forms of an essential nutrient (e.g., phosphorus), there have been few
studies on nutrient specialization, limiting our understanding of associated ecological
strategies or performance tradeoffs. In the present study, we measure bacterial growth
rates, an essential fitness component, for thirty-nine bacterial strains isolated from an
oligotrophic and eutrophic lake on a suite of phosphorus (P) resources. We then
quantified P niche breadth and tested for a specialization-performance tradeoff. We
found that bacterial isolates specialized on a diverse range of P forms, and that there
was a positive linear relationship between P specialization and an isolate’s maximum
growth rate, but only for isolates originating from the more eutrophic lake. These results

10

highlight the potential for P resource heterogeneity and nutrient specialization to drive
ecological strategies and performance tradeoffs in microorganisms.

Introduction
Resource heterogeneity plays a large role in driving and maintaining earth’s
biodiversity. Because resources are typically limited, species have evolved a variety of
ecological strategies to effectively meet their nutritional and energetic needs. For
example, species may evolve ecological tradeoffs that allow them to maximize access
to particular resources, but at a cost; stockpile resources while they are abundant; or
remain dormant until environmental conditions are more favorable (Cáceres 1997,
Caley & Munday 2003, Jones & Lennon 2010). Organisms’ ability to effectively compete
for and acquire resources strongly impacts species distribution, community composition,
and ecosystem functions.
Theory predicts that species’ niche breadth, or the number of different resource
forms that a species can use to meet its growth requirements, should be directly
impacted by resource availability (Futuyma & Moreno 1988, Chow et al. 2004).
Organisms frequently take advantage of resource heterogeneity by partitioning available
resources. Those which use only a small number of the available resource states are
considered niche specialists, while those which use many of the available resource
states are considered niche generalists. The proportion of specialists to generalists in a
community can impact total resource use, productivity, and the relationship between
community diversity and ecosystem function (Finke & Snyder 2008, Gravel et al. 2011).

11

Even resources that seem homogeneous to us may in fact contain multiple
ecologically relevant resource states for certain groups of organisms. For example,
while light is typically considered to be a single resource, photosynthetic pigments only
absorb photons from a portion of the spectrum, allowing for phytoplankton to partition
the light spectrum and ultimately coexist (Stomp et al. 2004). Nutrient resources are
also diverse and have been shown to be an important axis of niche variation. Essential
nutrients are bound up in many chemical forms, which are more or less biologically
available to different organisms. Variation in ability to access nutrient resource forms
can influence species’ resource partitioning, determination of species dominance, and
persistence of less dominant species in communities (McKane et al. 2002, von Felten et
al. 2009).
Ecological and evolutionary constraints limit species’ niche breadth. Ecological
constraints include organisms’ physiological limitations, such as the necessary
allocation of energy to different aims (i.e. fast growth, reproduction, or predator
defense). Maintaining a broad ecological niche may have inherent energetic costs or
favor performance tradeoffs (Futuyma & Moreno 1998). For example, traits that
increase fitness in one environment may decrease it in others (Kassen 2002).
Specialists are theorized to evolve in constant environments with abundant resources,
while generalists should evolve in temporally variable environments with heterogeneous
resources (Futuyma & Moreno 1988, Chow et al. 2004). So while a narrow-niche
specialist may perform better than a broad-niche generalist in its preferred environment,
the generalist may perform less well, but more consistently across environments (Caley
& Munday 2003). However, the shape of performance tradeoffs are highly system-

12

specific and can vary depending on the environmental conditions (Jessup & Bohannan
2008). Perhaps for this reason, ecological and evolutionary performance tradeoffs are
frequently theorized, though only occasionally empirically confirmed. Evolutionary
constraints on niche breadth can include genetic incompatibilities among traits, such as
occurs when there is genetic correlation between a trait subject to directional selection
and one subject to antagonistic selection (Futuyma 2010). Additionally, specialized traits
may evolve rarely, constraining such traits to certain phylogenetic groups. Such a trait is
considered to be ‘historically constrained’ or ‘phylogenetically conserved’ (Prinzing et al.
2001).
Phosphorus (P) is an essential limited resource for all living organisms. It is a
primary component of membranes, nucleic acids, and regulates protein snynthesis.
Heterotrophic microbes and primary producers are crucial for the transformation of
dissolved P resources into biomass. Inorganic phosphate (Pi) is considered to be the
most readily available form of P, being easily taken up by plants and microbes without
the requirement for specialized enzymes (Dyhrman et al. 2007). Yet the dissolved P in
most ecosystems is largely bound in organic forms and requires specialized, microbially
produced enzymes to be accessed. Within this pool of organic phosphorus (Porg), there
is substantial variability in compound lability based on chemical structure. For example,
phosphate esters—Porg compounds with C-O-P bonds—appear to be more biologically
available than phosphonates, which have a more stable C-P bond, possibly because
there are fewer enzymes which facilitate the breaking apart of such compounds (Clark
et al. 1998). The opportunity to acquire P from many resource forms has been shown to

13

be an important driver of genetic diversification for microbes and influences microbial
community composition, including species dominance in low-nutrient areas of the ocean
(Frette et al. 2009, Zubkov et al. 2003, Martiny et al. 2006). Despite the importance of
microbial P transformation in ecosystems and the demonstrated opportunity for nichespecialization, we do not know the extent of variation in microbial P-use niche breadth,
and our understanding of the relative availability of different P resource forms for
microbes remains limited.
Here, we isolated aquatic bacterial strains from different environments to explore
variation in P niche breadth, and test for a tradeoff between growth-rate and P niche
breadth. In this study, we compare bacterial isolates' ability to grow on a suite of
different forms of phosphorus, chosen for their ecological relevance in aquatic
environments and molecular structural diversity. We hypothesized that bacteria would
vary in their P niche breadth and demonstrate a performance tradeoff, such that those
with a wider niche breadth would on average grow slower than those with a narrower
niche breadth in their preferred environment (or P source). We also predicted that
bacterial isolates would grow at different rates on compounds with chemical structures
as similar as ATP and GTP, indicating a fine level of compound recognition among P
resources.

Materials and Methods
Lake Characterization
In fall of 2009, we collected surface water samples (0.5 m) from two southwest
Michigan lakes near the W.K. Kellogg Biological Station, USA: Wintergreen (WG) Lake

14

and Little Long (LL) Lake. WG Lake is an eutrophic waterbody located within Kellogg
Bird Sanctuary, and receives large P inputs from the resident birds. Little Long is an
oligotrophic lake with marl clay sediments and no known point-source P loadings. Both
lakes are sampled for nutrients several times each year as part of a regular monitoring
program. We referenced three years of nutrient data (2007-2009) for this study.
To determine bacterial community similarity between these two lakes, we
analyzed previously collected DNA pyrosequencing data (Jones & Lennon 2010).
Briefly, mixed surface layer samples were collected in summer of 2008. 250mL water
was filtered onto 0.2mm filters, and DNA was extracted using a commercially available
kit (DNA FastPrep purification kit from BIO 101). Using PCR, the DNA was labeled with
barcoded primers targeting the V4 region of the 16S rRNA gene before being
sequenced with an Illumina Genome Analyzer II at the Research Technology Support
Facility (RTSF) at Michigan State University. From 982 total sequences, only unique
(622 total), well-aligned (482 total) sequences were included for analysis.
Only unique tag sequences from the epilimnia communities were included for
analysis, reducing the total number of sequences from 982 to 622. Following sequence
alignment and quality checks for correct position and size, the number of included
sequences was further reduced to 496. These sequences were then binned according
to 97% nucleotide similarity for analysis. Using the libshuff program within the package
Mothur v1.23.1 (Schloss 2009), we calculated the Cramer-von Mises test statistic to test
for bacterial community similarity between the two lakes (Singleton 2001).

Bacterial Enrichment and Isolation

15

We enriched for bacteria in a variety of P environments. Immediately following
sample collection, we spread-plated 50-100 µL water samples from each lake onto
1.5% washed-agar plates containing a modified WC minimal medium based on
Stemberger 1981 (see Appendix B for full recipe). Briefly, the medium contained a
minimal nutrient and trace element mixture with the addition of the vitamins thiamine
(vitamin B1) and biotin (vitamin H), and one source of P. WC agar plates were prepared
at two P concentrations (100 µg P/L and 1 mg P/L), using one of five sources of P
[inorganic phosphate (Pi), (2-amino-ethyl) phosphonic acid (AEP), adenosine
triphosphate (ATP), phytic acid (Phyt) or a combination of all of the compounds], and
were buffered with calcium carbonate. The agar was washed by rinsing with distilled
water until the rinse water remained clear, with a miminum of seven rinses. It was then
rinsed once with Nanopure water, once with 70% ethanol, and finally with acetone
before being aerated at 40˚ C until dry. We allowed these enrichment samples to
incubate at 25˚ C for two to four weeks, to allow enough time for slow-growing bacteria
to form colonies.

Strain isolation
We sought to isolate diverse lake bacteria with different P-utilization strategies,
so we selected colonies for isolation that were morphologically distinct, sampling from
each type of P enrichment. We isolated the bacteria on agar plates using our modified
WC media (mWC) with the addition of Pi as the P source which we assumed would be
readily accessible to all bacteria, 10 mM HEPES buffer, and 50 mg/L cylcohexamide to
prevent fungal contamination. This recipe (mWC) was used to make all subsequent
16

media, with the desired P resource added. Also note that for all media, the stoichiometry
of the compounds was taken into account, and each P resource was added according
to the total P added, rather than the compound concentration. Isolates were re-streaked
multiple times to ensure single-strain isolation, and then grown in 1mg P/L Pi mWC
broth before being cryopreserved (19% glycerol, 81% 720 µg P/L Pi mWC, final
concentration). We preserved 18 isolates from LL Lake and 21 from WG Lake.

P-utilization assays
We tested each isolates' ability to grow on 19 P sources, chosen for their
relevance in aquatic ecosystems and diversity of P-bonding structures (see Tables 1-1
and 2-1). Assays were carried out in 96-well plates, using mWC broth containing one of
each P source at a target concentration of 5 mg P/L. We maintained high P
concentrations in order to ensure that the bacteria were not nutrient limited during
exponential growth. Each treatment was conducted in quadruplicate, with four positive
control wells containing P-free media, and 16 negative controls containing media with
Pi. The negative controls were positioned along either edge of the plate to alleviate
potential edge effects. Prior to initiating an assay, cryopreserved isolates were
inoculated into mWC with Pi [1 mg P/L] and incubated in 10 ml of liquid medium in 125ml shaken flasks (160 rpm) at 25˚ C until turbid, at which time they were diluted 10-fold
with P-free mWC and inoculated at 10% total volume into each of the 80 treatment or
positive control wells in a 96-well plate (20 µL inoculum into 180 µL media). Negative
controls received 20 µL P-free media. The plates were then incubated at 25 ˚C for up to

17

16 days. Each wells' optical density at 600 nm was measured using a Molecular
Devices SpectraMax5 spectrophotometer every 2-24 hours, depending on the speed of
the life cycle of each isolate. We used maximum likelihood (ML) to fit the modified
Gompertz function (Zwietering et al.1990; Lennon et al. 2007), which estimates
ecologically relevant parameters, in particular lag phase, maximum growth rate, and
maximum cell concentration (here, maximum optical density).
In order to account for any growth in the control wells, presumably due to stored
P ('luxury growth,' Bolier et al. 1992) we subtracted the average value of the P-free
control wells from all treatment wells for the isolate. We calculated an average growth
rate from the quadruplicate wells to obtain a single growth rate value for each isolate on
each P source. While, as stated previously, we used maximum likelihood to find the
maximum growth rate of each isolate on each P source, we will refer to these maximum
growth rates (per isolate, per P source) simply as 'growth rates' (GRs). Further, for the
rest of this paper, an isolate's 'maximum growth rate' (max GR) is considered to be its
maximum GR on the single P source on which it grew fastest. We standardized isolate
GRs for nearly all statistical analyses by dividing each isolate’s GR on each P source by
its own max GR. Standardizing in this way accounts for disparate isolate GRs as an
inherent property of each isolate, irrespective of the P source on which it grew. These
standardized growth rates are constrained to a scale from 0-1, with any differences
among isolates representing differences in their relative P-use abilities rather than
absolute differences in growth rates on the different P sources. We also excluded the P
source B12 from further analysis, since no isolates demonstrated detectable positive
growth on it, neither per optical density at 600 nm nor visual inspection. We created a

18

heatmap to visually display all isolate GRs on all P sources using the heatmap.2
function within the gplots package in the R statistical environment (R Development Core
Team 2004).

DNA sequencing and tree construction
We sequenced the isolates’ DNA and constructed phylogenetic trees to test for
the influence of phylogenetic history on P-use traits. We extracted DNA for sequencing
from fresh broth cultures of isolates inoculated from cryopreservation, grown to turbidity.
We used polymerase chain reaction (PCR) to amplify a portion of the encoding for a
region of the16S rRNA gene using the universal bacterial primers 8F (5'AGAGTTTGATCCTGGCTCAG-3') and 1492R (5'-GGTTACCTTGTTACGACTT-3') and
the following thermal cycle conditions: 5 min at 95 ˚C (initial denaturation); 30 cycles of
1 min at 94 ˚C, 1 min at 58 ˚C, 2 min at 72 ˚C; 10 min at 72 ˚C). The amplified DNA was
purified using the Qiagen Quick nucleotide fragment clean-up kit and sequenced on an
ABI PRISM® 3730 Genetic Analyzer at the Research Technology Support Facility at
Michigan State University. To align the sequences, we first used the quick-alignment
tool provided in ARB software (http://www.biol.chemie.tu-muenchen.de), followed by
manual refinement based on known secondary structures (Ludwig et al. 2004). We
assigned each sequence a genus designation, as determined using the Classifier tool
provided by the Ribosomal Database Project (RDP, Wang et al. 2007).
We then constructed a phylogeny from the aligned sequences with a general
time-reversible model of evolution (GTR) using the software package BayesPhylogenies
(Pagel & Meade 2004). Rather than returning a single consensus tree,

19

BayesPhylogenies uses Markov Chain Monte Carlo (MCMC) methods to generate a
suite of trees, whose frequency distributions correspond to the certainty of a given tree.
This is a way of incorporating phylogenetic uncertainty into subsequent analyses. We
chose the GTR method after using the freely available software package jModelTest
(Guindon & Gascuel 2003; Posada, D. 2008) to statistically compare many different
models of evolution to determine which is best for the given data. We specified the
Bacillus subtilis strain as the outgroup and allowed the chain to run for 2,000,000
iterations, with a burn-in period of 10,000 iterations. Thereafter, we sampled every 500
trees to yield a total of 3,981 trees.

Statistical Analyses
We assessed P niche breadth using the Levins index (Levins 1968) with the
standardized GRs. The Levins index incorporates the number of resource states used
(i.e. the number of P sources) as well as the relative frequency with which they are used
(the standardized GR). Higher values indicate a broader niche breadth. We ran a
multiple regression to explain the Levins index as a function of max GR (a continuous
variable) and lake origin (a categorical variable). The max GRs were log10-transformed
to meet the assumption of equal variance. Since we had different numbers of isolates
from each lake, lake origin was an unbalanced covariate. To avoid autocorrelation of the
data, we analyzed the data using type II sums of squares with the ‘car’ package in the R
software environment (Fox & Weisberg 2011). We also determined whether lake origin
significantly influenced isolate growth across all P sources by conducting an analysis of
similarity (ANOSIM) using the vegan package within the R statistical environment

20

(Oksanen et al. 2011, R Development Core Team 2004). We compared results when
using Bray-Curtis, Euclidean, and Manhattan distance matrices. To assess the
significance of the ANOSIM statistic, we ran 10,000 permutations of the data.
The P sources used in this experiment were chosen to maximize our
understanding of how bacterial P-use traits might relate to special enzymes or certain
molecular structures, and to compare P-use abilities among even structurally similar
compounds. In order to quantify similarity among P sources, we subjected the P-use
data to cluster analysis using the R software package ‘pvclust’ (Suzuki & Shimodaira
2009). Multiple methods for computing distance matrices (Manhattan, Euclidean, and
Bray-Curtis) and for conducting cluster analysis (Group Average and Ward’s Method)
were compared to ensure reliability of the results. Approximately Unbiased (AU) pvalues were calculated from multiscale bootstrap resampling for 10,000 iterations. The
hypothesis that "the cluster does not exist" is rejected with significance level at 0.05 for
clusters with AU p-value > 0.95.

Phylogenetic influence
Finding that a trait correlates highly with a clade’s phylogenetic history indicates
that there is some degree of phylogenetic conservatism, i.e. the variance among extant
species is explained well by ancestral relationships. Phylogenetic conservatism
necessitates non-independence of the data and possibly the need for phylogenetic
correction. Therefore, where phylogenetic history correlates with the trait data, we
present results for both traditional analyses without phylogenetic correction and those
including phylogenetic correction. To find maximum likelihood (ML) values for a given

21

trait for each of the 3,981 phylogenetic trees, we used the software program
BayesTraits (Pagel 1997, Pagel 1999). One can test whether a phylogeny correctly
predicts species trait covariances by incorporating the parameter Lambda in analyses.
Likelihood ratio tests can be used to compare models in which Lambda assumes its ML
value, or is forced to be 1 or 0. If the likelihood when Lambda assumes its ML value is
not distinguishable from the likelihood when it is forced to be 1, then the trait has
evolved as expected, given the tree topology and model of evolution. Similarly, if the
likelihood when Lambda assumes its ML value is not distinguishable from the likelihood
when Lambda is forced to be 0, this is evidence that the trait has evolved completely
independently of the phylogeny and phylogenetic correction is unnecessary.
All such analyses are dependent upon the model of evolution used for the
analysis. We used a random walk model of evolution, incorporating the scaling
parameter Kappa for each trait or comparison between traits. The Kappa parameter in
BayesTraits is used to stretch and compress branch lengths, allowing one to test for a
gradual versus a punctuational mode of evolution and incorporate the finding into the
model of evolution used. We allowed the program to first estimate the maximum
likelihood value of Kappa for each trait or comparison between traits, and then
incorporated the mean value to the hundredths place in our analyses. We also allowed
the program to estimate the maximum likelihood value of Delta, a parameter used to
scale total path length in a phylogeny, allowing one to compare models varying the
relative import of earlier versus later trait changes. We found that even though this
parameter was much greater than the Brownian Motion default of 1.0, indicating that
later trait changes correlated better with the phylogeny, incorporating this parameter did

22

not significantly improve the likelihood of the evolutionary model. The Lambda
parameter was incorporated in all analyses with this program, as it is the inclusion and
restriction of this parameter that allows one to conduct phylogenetic conservatism
hypothesis testing. All statistical analyses aside from those involving the
pyrosequencing data were performed with R software (R Development Core Team
2004).

Results
Lake Characterization
Despite their proximity, LL Lake and WG Lake contrast greatly in both nutrient
concentrations and microbial composition. As shown in Table 2-1, the total P content of
WG Lake can be 15 times that of LL Lake. While the Pi content of LL Lake remains at or
near the detection limit, the inorganic nitrogen concentrations are far more abundant
than those of WG Lake. This suggests that LL Lake is likely P-limited, while WG Lake
may be more nitrogen-limited. The lake contrasts are also evident when comparing the
epilimnetic bacterial communities. The pyrosequencing data support the hypothesis that
the epilimnia of LL and WG Lakes contain distinct bacterial communities (p<0.01).

23

Table 2-1. Lake attributes. Nutrient data was collected several times each year for three years (2007-2009).
Shown are means ± one standard deviation.

Lake

Area
(acres)

Little Long
Lake (LL)

170
39

Wintergreen
Lake (WG)

PO4(µg/L)

TDP
(µ/L)

TP
(µg/L)

NO3(mg/L)

NH4+
(µg/L)

0.68±0.54

5.8±2.3

9.9±2.1

1.1±0.60

100±63

14±13

36±15

0.058±0.14

47±55

80±41

Microbial specialization on P compounds
The high mean and narrow variance of the Levins index (mean 11.75 ± 2.59, 1
sd) confirms that many isolates had a broad P-niche breadth, able to use many of the P
sources for growth, while others could only use a few of the P sources. Similarly, while
most isolates demonstrated similar growth rates across their usable P sources, some
isolates grew quickly on one or a few P sources but far more slowly on others. (P-use
‘generalists’ and ‘specialists’, respectively; see Figure 2-1).

24

Figure 2-1. Ranked growth rates for each isolate. Each line represents one isolate’s scaled growth rates in
decreasing order of magnitude. Blue and green lines represent isolates from LL Lake and WG Lake,
respectively. Isolates with shallower initial slopes grow relatively well on many P sources (‘generalists’);
while those with steep initial slopes grow quickly on one or a few P sources and slowly on others
(‘specialists’).
In accordance with our prediction, we found evidence for a tradeoff between max GR
and niche breadth. On average, isolates with a broader niche breadth had lower max
GRs than those with a narrower niche breadth. [log10(max GR) ~ lake origin*levins
index; interaction - F1,35 = 5.48, P = 0.025; levins index - F1,35 = 5.29, P = 0.028; lake
origin - F1,35 = 3.26, P = 0.079]. However, the significant interaction between isolate
lake origin and niche breadth indicates that this tradeoff is present in only one of the two

25

lakes—WG Lake, as shown in Figure 2-1. Isolate P-niche breadth accounted for 3038% of the variation in WG Lake isolates’ maximum GRs (regression coefficients with
and without phylogenetic correction, respectively), with a broader niche breadth
predictive of a lower than average max GR. However, LL Lake isolates’ niche breadths
were independent of their max GRs, therefore lacking the tradeoff found for WG Lake
isolates. This is also supported by analyses that include correction for shared
phylogenetic history (WG Lake tradeoff p<<0.01, R = 0.30; LL Lake tradeoff not
significant; Phylogeny shown in Figure A-1 of Appendix A).

26

Figure 2-2. Relationship between P-niche breadth and bacterial growth rate. Logtransformed maximum growth rates are plotted against the Levins index for each
isolate. Open circles represent isolates from LL Lake. Closed circles represent isolates
from WG Lake. Projected slopes in this figure include phylogenetic correction. R2 of
regression with phylogenetic correction is 0.30. R2 of regression without phylogenetic
correction is 0.38.
We found that the bacterial isolates tended to grow faster on certain P forms.
Isolates grew faster on resources that can be accessed without the need for specialized
enzymes, such as Pi or nucleotides when compared with resources that do, such as the
phosphonate AEP (See Table 2-2 and Figure 2-3; paired t-tests with standardized
growth rates, 38 df, p<<0.05 for both comparisons). Most isolates grew fastest on TPP,
followed by GDP (See Table 2-2 for these and all other P compounds abbreviations).
While few isolates grew fastest on Pi, on average, there was no difference in growth
rates between Pi and TPP (Paired t-test, 38 df, p>0.05). Isolates from both lakes tended
to grow fastest on the same resources and slowest on the same resources (data not
shown, but see Figure 2-3), with one notable exception— Pi. Isolates from WG Lake
had higher standardized growth rates on Pi than those from LL Lake (Two sample t-test
with equal variances, 37 df, p<<0.05).

27

Table 2-2. Phosphorus source abbreviations and properties.
Compound Name

Abbreviation MW

mol P: mol
compound

P bond types

Environmental sources

(2-aminoethyl) phosphonic AEP
acid
phytate
Phyt

125.06

1

C-P

bacteria

660.04

6

inorganic phosphate

Pi

174.18

1

major P storage form
in plants
apatite, Porg hydrolysis

adenosine-3',5'-cyclic
monophosphate
adenosine-5'-triphosphate

cAMP

351.2

1

C-O-P
monoester
phosphoric acid
ester
C-O-P diester

ATP

551.1

3

DNA

phenylphosphonic acid
triphosphate

PhenCP
Poly-P

158.09
367.86

1
3

guanosine-5'-diphosphate

GDP

541.21

2

guanosine-5'-triphosphate

GTP

523.18

3

phospho(enol) pyruvate

PEP

208.04

1

C-O-P
monoester & 2
P anhydrides
C-P
phosphate
esters
C-O-P
monoester & 1
P anhydride
C-O-P
monoester & 2
P anhydrides
C-O-P
monoester

alpha-D-glucose 1phosphate
D-glucose 6-phosphate

G1P

304.1

1

G6P

282.12

1

methylphosphonic acid

MeCP

96.02

1

C-P

beta-glycerophosphate

BGP

216.04

1

deoxyribonucleic acid
L-alphaphosphatidylethanolamine
L-alphaphosphatidylcholine
thiamine pyrophosphate

DNA
Peth

608.93
744.05

4
1

C-O-P
monoester
C-O-P diester
C-O-P diester

Pchol

760.09

1

C-O-P diester

TPP

460.77

2

C-O-P
monoester & 1
P anhydride

28

C-O-P
monoester
C-O-P
monoester

living organisms

herbicide
major P storage form
in bacteria
DNA
DNA
living organisms;
intermediate in
glycolysis
animals; glycogenesis
intermediate
living organisms;
involved in many
metabolic pathways
biological precurser &
degradation product
vertebrates
DNA
bacterial membrane
phospholipids
animal membrane
phospholipids
synthesized by
bacteria, fungi, and
plants; required for all
organisms

Figure 2-3. Heatmap visually displaying isolate maximum growth rates, scaled from 0-1. Darker colors

29

Figure 2-3 (cont’d)
represent higher values. The dendrogram clusters isolates and P sources according to these growth rates
using the Euclidean method to calculate distances. Isolates are named according to the genus of each as
identified using the Classifier tool provided by the Ribosomal Database Project (RDP, Wang et al. 2007),
numbered when multiple from one genus occur, and are labeled with an “L” or “W” indicating from which
lake they were isolated (LL Lake or WG Lake, respectively). P compound abbreviations are found in Table
2.
The isolate trait data confirms some of our expectations for how similar an
isolate’s GRs should be on various P sources, based on similarity of compound
structure. For example, as shown in Figure 2-4, GDP and GTP consistently yield the
most similar GRs for a give isolate and in fact, are indistinguishable from each other in
this analysis. The two phospholipids and two of the phosphonates also cluster together
closely. However, we were surprised to find that PolyP clusters more closely with Porg
compounds than with Pi, the only other inorganic P source in this study. Some
clustering metrics found DNA to be in an indistinguishable cluster from MeCP and
PhenCP. This is likely due to the large variance in DNA GR values, and generally poor
growth of most bacteria on these P sources.

30

Figure 2-4. P sources clustered according to similarity of P use traits across isolates. The Manhattan
method was used to calculate the distance matrix, and the Group Average method with 10,000
bootstrapped permutations was used to cluster the isolates. Green and red numbers indicate the bootstrap
probability and the ‘Approximately Unbiased’ (AU) probability that the cluster exists. Red boxes surround
groups for which the AU p-value is less than 0.05, indicating that we should reject the null hypothesis that
the clusters do not exist. Thus the compounds within the red boxes are not distinguishable from one
another using these metrics.
Importantly, our phylogenetic analyses confirmed that we can consider the
isolates to be independent replicates when comparing most P-use traits. While
phylogenetic history minimally influences P-niche breadth and growth on AEP and Phyt,
bacterial growth rates on cAMP, DNA, and Pchol suggest these P-use traits have

31

evolved as expected for vertical gene transfer, given the tree topology and model of
evolution. We did not find evidence that phylogenetic history affected any other P-use
traits (see Table A-1 in Appendix A). Further, the high estimated delta parameter ML
values for all traits is indicative of a species-specific mode of trait evolution, while the
zero or near-zero estimated maximum value of the scaling parameter Kappa for all traits
is consistent with a punctuational, rather than a gradual mode of evolution. These
results indicate that phylogenetic history constrains few of the tested P-use traits.

Discussion
The phosphorus resource pool is diverse and plays an important role in
determining ecosystem productivity. Recent studies have demonstrated the potential for
niche variation according to nutrient use abilities (Martiny et al. 2009). Here, we have
used physiological assays with environmental isolates to demonstrate that aquatic
bacteria vary in their P-use niche breadth, and that this variation can be the basis for a
specialization-performance tradeoff.
While most bacteria had a broad P-niche breadth, some tended towards
specialization on one or a few P forms. Isolates even specialized on compounds known
to be degraded most efficiently by substrate-specific enzymes, such as AEP and Phyt.
Though typically considered to be less-accessible forms of P, these compounds have
been shown to be readily metabolized by some bacterial groups and have been
suggested to play important roles in P metabolism in both aquatic and terrestrial
ecosystems (Rodríguez & Fraga 1999, Orchard et al. 2009). As the primary P storage
form in plants and a significant pool of P in manure, bacterial Phyt degradation has

32

been suggested to be an important mechanism by which Porg from agricultural lands is
made labile across a landscape, traveling from farms to nearby water bodies where
excess P causes unwanted algal blooms (Hill et al. 2007). The widespread ability of our
isolates to access P from and even specialize on Phyt suggests that P liberation from
this compound may not only be important in P transport across terrestrial environments,
but likely continues in aquatic systems.
Our results support the system-specific nature of performance tradeoffs. WG
Lake isolates with a broader niche grew more slowly than those with a narrower niche
breadth, yet there was no such tradeoff among LL Lake isolates. There are many
possible causes of this disparity. More productive environments may confer the greatest
benefit to niche-specialization by both increasing specialization opportunity on the most
abundant resources and increasing availability of rare resources (Futuyma & Moreno
1988, Chow et al. 2004). Increased availability of a variety of resources may effectively
make WG Lake a more homogeneous environment with respect to the P resource
needs of any given bacterial strain. If this is true, and WG Lake bacteria are frequently
limited or co-limited by resources other than P, then P-use generalists in WG Lake may
be unnecessarily expending more energy to meet their P resource needs than their
specialist counterparts. Alternatively, differences in nutrient acquisition among LL
isolates may be minimized in productive environments like those of the experimental
environment, and any advantage of niche-specialization may only be apparent under
conditions of nutrient scarcity, more closely resembling the environment from which the
strains were isolated (Jessup & Bohannan 2008, Buckling et al. 2007). On the other
hand, LL Lake isolates may trade off traits in favor of niche-generalization that were not

33

measured in this study, such as expending energy to maintain a larger genome size.
Finally, generalists do not necessarily experience a cost for maintaining a broad nichebreadth, and this may be the case for the LL Lake isolates (Buckling et al. 2007).
The isolates’ ecological history primarily constrained the P-niche breadth/ GR
tradeoff, while their phylogenetic history likely constrains only a few P-use traits.
Accounting for phylogenetic history only minimally reduced the effect size of the
ecological tradeoff in WG Lake. This indicates that ecological or environmental factors
such as differences in lake bacterial community composition, nutrient concentrations, or
abundances of specific P forms may primarily influence the strength of this ecological
tradeoff. Consistent with other studies, we found substantial strain-specific variation in
P-use for nearly all traits (Huang et al. 2005, Martiny et al. 2006). That most trait data
did not correspond well with the phylogeny may be a reflection of complex and variable
genetic regulation, repeated independent evolution of many traits, or possibly the lateral
acquisition of P-use genes (Martiny et al. 2006; Martiny et al. 2009). Yet that some Puse traits were at least minimally influenced by phylogenetic history suggests that they
may be phylogenetically constrained. These traits included those enhanced by one or
more specialized enzymes, such as phosphonate, phytate, cAMP and DNA metabolism.
If these traits are phylogenetically constrained, community composition may ultimately
limit P turnover of these compounds in an ecosystem.
Our physiological data support the importance of diverse P sources for meeting
bacterial community P demand. Isolates in our study differentially grew on P
compounds with similar compound structures, such as the nucleotides GTP and ATP.
This measurable physiological response to relatively small differences in P resource

34

form may be an effect of a number of factors. Even small structural differences between
the compounds may affect the ability for enzymes to cleave phosphate from a molecule;
bacteria may be able to access both nitrogen and phosphorus from some compounds;
or bacteria may save different amounts of energy by taking up whole or partial Pcontaining structural components rather than synthesizing organic compounds from
their inorganic building blocks. Though the P source on which the isolates most
commonly grew fastest was TPP, a vitamin with an easy-to-degrade pyrophosphate
group, many also grew fastest given GTP or GDP as their sole P source. Several
studies have indicated that nucleotides may be among the most readily available Porg
sources, particularly in oligotrophic environments, taken up and regenerated up to five
times more quickly than the bulk dissolved Porg pool (Cotner & Wetzel 1991; Siuda &
Chróst 2001, Karl and Barkman 2005, Lennon 2007 and references within). If bacteria
in natural communities gain an advantage for fast growth on particular P sources, then
the dominant P form in aquatic environments may significantly affect microbial
community structure and species dominance.
This study provides evidence that nutrient-based resource specialization can
significantly influence important performance traits of an organism, such as its growth
rate, though these effects may be system-specific. Additionally, organic nutrient forms
may be more important in structuring community and ecosystem dynamics than
previously thought.

35

APPENDICES

36

APPENDIX A
Supplementary Figures

37

Figure A-1. Phylogeny of bacterial isolates used in this study, with reference sequences. As in Figure 2-3,
isolates from the present study are named according to the genus of each, as determined from RDP

38

Figure A-1 (cont’d)
classification, numbered when multiple from one genus occur, and are labeled with an “L” or “W” indicating
from which lake they were isolated (LL Lake or WG Lake, respectively). They are also color-coded by
lake—LL Lake isolates are blue and WG Lake isolates are green. Reference sequences (in black) are
named according to their RDP or genbank classification, when an RDP classification was unavailable. All
reference sequences are also labeled with their Genbank identifier.

39

Table A-1. Influence of phylogenetic history on phosphorus-use traits. ML = Maximum likelihood. For
analysis details and compound abbreviations, see Methods section and Table 2-2 from text, respectively.
Within the software program BayesTraits, one can test whether a phylogeny correctly predicts species trait
covariances by incorporating the parameter Lambda (l) in analyses (Pagel 1997, Pagel 1999). Significant
P-values for l= 0 vs. ML comparisons indicate that phylogenetic history at least minimally influences the
trait. Significant P-values for l= 1 vs. ML comparisons indicate that the trait is not perfectly correlated with
the phylogeny. In this study, we considered traits to be phylogenetically conserved if >95% of generated
phylogenetic trees statistically support phylogenetic conservatism (a= 0.05). For example, a trait that
perfectly correlates with the phylogeny would yield a ‘100’ in the first column and a ‘0’ in the second; while
one that is minimally influenced by phylogenetic history would yield a 95-100 in the first column and a value
5 or greater in the second. Values supporting phylogenetic conservatism are starred.
Phosphorus
compound
AEP
Phyt
Pi
cAMP
ATP
PhenCP
PolyP
GDP
GTP
G1P
G6P
MeCP
BGP
DNA
Peth
Pchol
TPP
Levins

λ = 0 vs. ML value
% of trees with P-values <0.05
27
100*
0
100*
0
0
0.10
0.0
0.0
0.0
0.0
0.0
24
100*
0.0
100*
0.70
0.0

λ = 1 vs. ML value
% of trees with P-values <0.05
32
0.55*
90
2*
100
100
99
100
100
72
94
98
13
13
100
21
86
100

40

Figure A-2. Isolate 16S DNA sequences used for phylogenetic analyses. Sequences are shown in FASTA
format.
>L_Aeromicrobium2
acacgtgagcaatctgcccttctcatcggaataaccattg
gaaacgatggctaatgccgaatacgacctcctttcgcatgatcggaggtggaaagctccg
gcggagaaggatgagctcgcggcctatcagctagttggcggggtaacggcccaccaaggc
gacgacgggtagccggcctgagagggtgaccggccacactgggactgagacacggcccag
actcctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagcaa
cgccgcgtgagggatgacggccttcgggttgtaaacctctttcagcagggacgaagcgaa
agtgacggtacctgcagaagaaggaccggccaactacgtgccagcagccgcggtaatacg
tagggtccgagcgttgtccggaattattgggcgtaaagggctcgtaggcggtttgtcgcg
tcgggagtgaaaactcagggcttaaccctgagcgtgcttccgatacgggcagactagagg
tattcaggggagaacggaattcctggtgtagcggtggaatgcgcagatatcaggaggaac
accggtggcgaaggcggttctctgggaatacctgacgct
>W_Aeromonas
cggcagcgggaagtagcttgctactt
ttgccggcgagcggcggacgggtgagtaatgcctggggatctgcccagtcgagggggata
acagttggaaacgactgctaataccgcatacgccctacgggggaaaggaggggaccttcg
ggcctttcgcgattggatgaacccaggtgggattagctagttggtggggtaatggctcac
caaggcgacgatccctagctggtctgagaggatgatcagccacactggaactgagacacg
gtccagactcctacgggaggcagcagtggggaatattgcacaatgggggaaaccctgatg
cagccatgccgcgtgtgtgaagaaggccttcgggttgtaaagcactttcagcgaggagga
aaggttgacagctaatatctgtcagctgtgacgttactcgcagaagaagcaccggctaac
tccgtgccagcagccgcggtaatacggagggtgcaagcgttaatcggaattactgggcgt
aaagcgcacgcaggcggttggataagttagatgtgaaagccccgggctcaacctgggaat
tgcatttaaaactgttcagctagagtcttgt
>W_Bacillus
cagcggcggacgggtgagtaacacgtgggcaacctgcctgtaagactgggataactccgg
gaaaccggagctaataccggatactatgtcaaaccgcatggtttgacattcaaagacggt
ttcggctgtcacttacagatgggcccgcggcgcattagctagttggtgaggtaatggctc
accaaggcaacgatgcgtagccgacctgagagggtgatcggccacactgggactgagaca
cggcccagactcctacgggaggcagcagtagggaatcttccgcaatggacgaaagtctga
cggagcaacgccgcgtgagtgatgaaggttttcggatcgtaaaactctgttgtcagggaa
gaacaagtgccggagtaactgccggtgccttgacggtacctgaccagaaagccacggcta
actacgtgccagcagccgcggtaatacgtaggtggcaagcgttgtccggaattattgggc
gtaaagcgcgcgcaggcggtttcttaagtctgatgtgaaagcccccggctcaaccgggga
gggtcattggaaactgggaaacttgagtgcagaagaggagagtggaattccacgtgtagc
ggtgaaatgcgtagagatgtggaggaacaccagtggcgaaggcgactctctggtctgtaa
ctgacgctgaggcgcgaaagcgtggggagcgaacaggattagataccctggtagtccacg
ccgtaaacgatgagtgctaagtgttagagggtttccgccctttagtgctgcagctaacgc
attaagcactccgc
>L_Brevundimonas
cttcagagttagtggcggacg
41

Figure A-2 (cont’d)
ggtgagtaacacgtgggaacgtgcctttaggttcggaataactcagggaaacttgtgcta
ataccgaatgtgcccttcgggggaaagatttatcgcctttagagcggcccgcgtctgatt
agctagttggtgaggtaaaggctcaccaaggcgacgatcagtagctggtctgagaggatg
atcagccacattgggactgagacacggcccaaactcctacgggaggcagcagtggggaat
cttgcgcaatgggcgaaagcctgacgcagccatgccgcgtgaatgatgaaggtcttagga
ttgtaaaattctttcaccggggacgataatgacggtacccggagaagaagccccggctaa
cttcgtgccagcagccgcggtaatacgaagggggctagcgttgctcggaattactgggcg
taaagggagcgtaggcggacatttaagtcaggggtgaaatcccggggctcaacctcggaa
ttgcctttgatactgggtgtcttgagtatgagagaggtgtgtggaactccgagtgtagag
gtgaaattcgtagatattcggaagaacaccagtggcgaaggcgacacactggctcattac
tgacgctgaggctcgaaagcgtggggagcaaacaggattagataccctggtagtccacgc
cgtaaacgatgattgctagttgtcgggatgcatgc
>L_Brevundimonas1
acgaactcttcggagttagtggcggacgggtgagtaacacgtgggaacgtgcctttaggt
tcggaataactcagggaaacttgtgctaataccgaatgtgcccttcgggggaaagattta
tcgcctttagagcggcccgcgtctgattagctagttggtgaggtaaaggctcaccaaggc
gacgatcagtagctggtctgagaggatgatcagccacattgggactgagacacggcccaa
actcctacgggaggcagcagtggggaatcttgcgcaatgggcgaaagcctgacgcagcca
tgccgcgtgaatgatgaaggtcttaggattgtaaaattctttcaccggggacgataatga
cggtacccggagaagaagccccggctaacttcgtgccagcagccgcggtaatacgaaggg
ggctagcgttgctcggaattactgggcgtaaagggagcgtaggcggacatttaagtcagg
ggtgaaatcccggggctcaacctcggaattgcctttgatactgggtgtcttgagtatgag
agaggtgtgtggaactccgagtgtagaggtgaaattcgtagatattcggaagaacaccag
tggcgaaggcgacacactggctcattactgacgctgaggctcgaaagcgtggggagcaaa
caggattagataccctggtagtccacgccgtaaacgatgattgctagttgtcgggatgca
tgcatttcggtgacgcagctaacgcattaagcaatccgcctggggagtacggtcgcaaga
ttaaaactcaaaggaattgacgg
>W_Brevundimonas1
tagtggcggacgggtgagtacacgtgggaacgtgcctttaggttcggaataactcaggga
aacttgtgctaataccgaatgtgcccttcgggggaaagatttatcgcctttagagcggcc
cgcgtctgattagctagttggtgaggtaaaggctcaccaaggcgacgatcagtagctggt
ctgagaggatgatcagccacattgggactgagacacggcccaaactcctacgggaggcag
cagtggggaatcttgcgcaatgggcgaaagcctgacgcagccatgccgcgtgaatgatga
aggtcttaggattgtaaaattctttcaccggggacgataatgacggtacccggagaagaa
gccccggctaacttcgtgccagcagccgcggtaatacgaagggggctagcgttgctcgga
attactgggcgtaaagggagcgtaggcggacatttaagtcaggggtgaaatcccggggct
caacctcggaattgcctttgatactgggtgtcttgagtatgagagaggtatgtggaactc
cgagtgtagaggtgaaattcgtagatattcggaagaacaccagtggcgaaggcgacatac
tggctcattactgacgctgaggctcgaaagcgtggggagcaaacaggattagataccctg
gtagtccacgccgtaaacgatgattgctatttgtcgggatgcatgcatttcggt
>L_Brevundimonas2
tggcggacgggtgag
42

Figure A-2 (cont’d)
taacacgtgggaacgtgcctttaggttcggaataactcagggaaacttgtgctaataccg
aatgtgcccttcgggggaaagatttatcgcctttagagcggcccgcgtctgattagctag
ttggtgaggtaaaggctcaccaaggcgacgatcagtagctggtctgagaggatgatcagc
cacattgggactgagacacggcccaaactcctacgggaggcagcagtggggaatcttgcg
caatgggcgaaagcctgacgcagccatgccgcgtgaatgatgaaggtcttaggattgtaa
aattctttcaccggggacgataatgacggtacccggagaagaagccccggctaacttcgt
gccagcagccgcggtaatacgaagggggctagcgttgctcggaattactgggcgtaaagg
gagcgtaggcggacatttaagtcaggggtgaaatcccggggctcaacctcggaattgcct
ttgatactgggtgtcttgagtatgagagaggtgtgtggaactccgagtgtagaggtgaaa
ttcgtagatattcggaagaacaccagtggcgaaggcgacatactggctcattactgacgc
tgaggctcgaaagcgtggggagcaaacaggattagataccctggtagtccacgccgtaaa
cgatgattgctagttgtcgggatgcatgcatttcggtgacgcagctaacgc
>W_Brevundimonas2
tcgacgaactcttcggagttagtggcggacgggtgagtaacacgtgggaacgtgccttta
ggttcggaataactcagggaaacttgtgctaataccgaatgtgcccttcgggggaaagat
ttatcgcctttagagcggcccgcgtctgattagctagttggtgaggtaaaggctcaccaa
ggcgacgatcagtagctggtctgagaggatgatcagccacattgggactgagacacggcc
caaactcctacgggaggcagcagtggggaatcttgcgcaatgggcgaaagcctgacgcag
ccatgccgcgtgaatgatgaaggtcttaggattgtaaaattctttcaccggggacgataa
tgacggtacccggagaagaagccccggctaacttcgtgccagcagccgcggtaatacgaa
gggggctagcgttgctcggaattactgggcgtaaagggagcgtaggcggacatttaagtc
aggggtgaaatcccggggctcaacctcggaattgcctttgatactgggtgtcttgagtat
gagagaggtatgtggaactccgagtgtagaggtgaaattcgtagatattcggaagaacac
cagtggcgaaggcgacatactggctcattactgacgctgangctcgaaagcgtggggagc
aaacaggattagataccctggtagtccacgccgtaaacgatgattgctagttgtcgggat
gcatgcatttcggtg
>L_Dietzia
gtaatctgccctgcacttcgggataa
gcctgggaaaccgggtctaataccggatatgagctcctgccgcatggtgggggttggaaa
gtttttcggtgcaggatgagtccgcggcctatcagcttgttggtggggtaatggcctacc
aaggcgacgacgggtagccggcctgagagggtgatcggccacactgggactgagacacgg
cccagactcctacgggaggcagcagtggggaatattgcacaatgggcgaaagcctgatgc
agcgacgccgcgtgggggatgacggtcttcggattgtaaactcctttcagtagggacgaa
gcgaaagtgacggtacctgcagaagaagcaccggccaactacgtgccagcagccgcggta
atacgtagggtgcaagcgttgtccggaattactgggcgtaaagagctcgtaggcggtttg
tcacgtcgtctgtgaaatcctccagctcaactgggggcgtgcaggcgatacgggcagact
tgagtactacaggggagactggaattcctggtgtagcggtgaaatgcgcagatatcagga
ggaacaccggtggcgaaggcgggtctctgggtagtaactgacgctgaggagcgaaagcat
ggggagcaaacaggattagataccct
>W_Flavobacterium1
agtcgaggggtatatgtcttcggatatagagaccgg
cgcacgggtgcgtaacgcgtatgcaatctaccttttacagagggatagcccagagaaatt
43

Figure A-2 (cont’d)
tggattaatacctcatagtatagtgactcggcatcgagatactattaaagtcacaacggt
aaaagatgagcatgcgtcccattagctagttggtaaggtaacggcttaccaaggctacga
tgggtaggggtcctgagagggagatcccccacactggtactgagacacggaccagactcc
tacgggaggcagcagtgaggaatattggacaatgggcgcaagcctgatccagccatgccg
cgtgcaggatgacggtcctatggattgtaaactgcttttgtacgagaagaaacactccta
cgtgtaggagcttgacggtatcgtaagaataaggatcggctaactccgtgccagcagccg
cggtaatacggaggatccaagcgttatccggaatcattgggtttaaagggtccgtaggcg
gtttagtaagtcagtggtgaaagcccatcgctcaacggtggaacggccattgatactgct
aaacttgaattattaggaagtaactagaatatgtagtgtagcggtgaaatgcttagagat
tacatggaataccaattgcgaaggcaggttactactaatggattgacgctgatggacgaa
agcgtgggtagcgaacaggattagataccctggtagtccacgccgtaaacgatggatact
agctgttggaagcaatttcagtggctaagcgaaagtgataagtatcccacctggggagta
cgttcgcaagaatgaaactcaaaggaattgacgggg
>W_Flavobacterium2
atttagagacc
ggcgcacgggtgcgtaacgcgtatgcaatctgcctttcacagagggatagcccagagaaa
tttggattaatacctcatagcattacgggatggcatcatcctgtaattaaagtcacaacg
gtgaaagatgagcatgcgtcccattagctagttggtaaggtaacggcttaccaaggcaac
gatgggtaggggtcctgagagggagatcccccacactggtactgagacacggaccagact
cctacgggaggcagcagtgaggaatattggtcaatgggcgcaagcctgaaccagccatgc
cgcgtgcaggatgacggtcctatggattgtaaactgcttttgcacaggaagaaacactcc
gacgtgtcggagcttgacggtactgtgagaataaggatcggctaactccgtgccagcagc
cgcggtaatacggaggatccaagcgttatccggaatcattgggtttaaagggtccgtagg
cggtttggtaagtcagtggtgaaagcccatcgctcaacggtggaacggccattgatactg
ctaaacttgaattattgggaagtaactagaatatgtagtgtagcggtgaaatgcttagag
attacatggaataccaattgcgaaggcaggttactacccatcgattgacgctgatggacg
aaagcgtgggtagcgaacaggat
>W_Flavobacterium3
agaccgg
cgcacgggtgcgtaacgcgtatgcaatctaccttgtacagagggatagcccagagaaatt
tggattaatacctcatagtatatagagttggcatcaacactatattaaagtcacaacggt
aaaagatgagcatgcgtcccattagctagttggtaaggtaacggcttaccaaggctacga
tgggtaggggtcctgagagggagatcccccacactggtactgagacacggaccagactcc
tacgggaggcagcagtgaggaatattggacaatgggcgcaagcctgatccagccatgccg
cgtgcaggatgacggtcctatggattgtaaactgcttttatacgagaagaaacactcctt
cgtgaaggaatttgacggtatcgtaagaataaggatcggctaactccgtgccagcagccg
cggtaatacggaggatccaagcgttatccggaatcattgggtttaaagggtccgtaggcg
gtcttgtaagtcagtggtgaaagcccatcgctcaacggtggaacggccattgatactgct
ggacttgaattattaggaagtaactagaatatgtagtgtagcggtgaaatgcttagagat
tacatggaataccaattgcgaaggcaggttactactaattgattgacgctgatggacgaa
agcgtgggtagcgaacaggattagataccctggtagtccacgccgtaaacgatggatact
agctgttgggagcaatttcagtggctaagcgaaagtgataagtatcccacctggggagta
cgttcgc
44

Figure A-2 (cont’d)
>W_Flavobacterium4
agtcgaggggtatgttcttcggaattagagaccggc
gcacgggtgcgtaacgcgtatgcaatctaccttttacagagggatagcccagagaaattt
ggattaatacctcatagtataatgacttggcatcaagacattattaaagtcacaacggta
aaagatgagcatgcgtcccattagctagttggtaaggtaacggcttaccaaggctacgat
gggtaggggtcctgagagggagatcccccacactggtactgagacacggaccagactcct
acgggaggcagcagtgaggaatattggacaatgggcgcaagcctgatccagccatgccgc
gtgcaggatgacggtcctatggattgtaaactgcttttatacgagaagaaacactacttc
gtgaagtagcttgacggtatcgtaagaataaggatcggctaactccgtgccagcagccgc
ggtaatacggaggatccaagcgttatccggaatcattgggtttaaagggtccgtaggcgg
tttagtaagtcagtggtgaaagcccatcgctcaacggtggaacggccattgatactgctg
aacttgaattattaggaagtaactagaatatgtagtgtagcggtgaaatgcttagagatt
acatggaataccaattgcgaaggcaggttactactaattgattgacgctgatgg
>L_Flavobacterium
tcggatagagagaccggc
gcacgggtgcgtaacgcgtatgcaatctaccttttacagagggatagcccagagaaattt
ggattaatacctcatagtataatgagttggcatcaacacattattaaagtcacaacggtg
aaagatgagcatgcgtcccattagctagttggtaaggtaacggcttaccaaggctacgat
gggtaggggtcctgagagggagatcccccacactggtactgagacacggaccagactcct
acgggaggcagcagtgaggaatattggacaatgggcgcaagcctgatccagccatgccgc
gtgcaggatgacggtcctatggattgtaaactgcttttgtacaagaagaaacactcctat
gtataggagcttgacggtatcgtaagaataaggatcggctaactccgtgccagcagccgc
ggtaatacggaggatccaagcgttatccggaatcattgggtttaaagggtccgtaggcgg
tttagtaagtcagtggtgaaagcccatcgctcaacggtggaacggccattgatactgctg
aacttgaattattaggaagtaactagaatatgtagtgtagcggtgaaatgcttagagatt
acatggaataccaattgcgaaggcaggttactactaattgattgacgctgatggacgaaa
gcgtgggtagcgaacaggattaaataccctggtagt
>L_Kocuria
tgctgggc
ggattagtggcgaacgggtgagtaatacgtgagtaacctgcccttgactctgggataagc
ctgggaaactgggtctaatactggatactacttcctgccgcatggtgggtggtggaaagg
gttttactggttttggatgggctcacggcctatcagcttgttggtggggtaatggctcac
caaggcgacgacgggtagccggcctgagagggtgaccggccacactgggactgagacacg
gcccagactcctacgggaggcagcagtggggaatattgcacaatgggcggaagcctgatg
cagcgacgccgcgtgagggatgacggccttcgggttgtaaacctctttcagtagggaaga
agcgagagtgacggtacctgcagaagaagcgccggctaactacgtgccagcagccgcggt
aatacgtagggcgcaagcgttgtccggaattattgggcgtaaagagctcgtaggcggttt
gtcgcgtctgctgtgaaagcccggggctcaaccccgggtctgcagtgggtacgggcagac
taaagtgcagtaggggagactggaattcctggtgtagcggtgaaatgcgcagatatcagg
aggaacaccgatggcg
>L_Pelomonas
ctgacgagtggcgaacggg
45

Figure A-2 (cont’d)
tgagtaatatatcggaacgtgcccagttgtgggggataactgctcgaaagagcagctaat
accgcatacgacctgagggtgaaagcgggggatcgcaagacctcgcgcaattggagcggc
cgatatcagattagctagttggcggggtaaaagcccaccaaggcgacgatctgtagctgg
tctgagaggacgaccagccacactgggactgagacacggcccagactcctacgggaggca
gcagtggggaattttggacaatggacgcaagtctgatccagccatgccgcgtgcgggaag
aaggccttcgggttgtaaaccgcttttgtcagggaagaaacgctctgggctaataccctg
gggtaatgacggtacctgaagaataagcaccggctaactacgtgccagcagccgcggtaa
tacgtagggtgcaagcgttaatcggaattactgggcgtaaagcgtgcgcaggcggttatg
caagacagatgtgaaatccccgggctcaacctgggaactgcatttgtgactgcatggcta
gagtacggtagagggggatggaattccgcgtgtagcagtgaaatgcgtagatatgcg
>L_Pseudomonas1
cttgcttctcttgaga
gcggcggacgggtgagtaatgcctaggaatctgcctggtggtgggggataacgttcggaa
acggacgctaataccgcatacgtcctacgggagaaagcgggggatcttcggacctcgcgc
cattagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggcgacg
atccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactc
ctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgcc
gcgtgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggcagtaa
cctaatacgttattgttttgacgttaccgacagaataagcaccggctaacttcgtgccag
cagccgcggtaatacgaagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcg
taggtggttcagtaagttggaagtgaaatccccgggctcaacctgggaactgctttcaaa
actgctgagctagagtacggtagagggtggtggaatttcctgtgtagcggtgaaatgcgt
aaatataggaaagaacaccagtggcgaaagcgaccacctggactgatactgacact
>L_Pseudomonas3
ttgcttctctt
gagagcggcggacgggtgagtaatgcctaggaatctgcctggtggtgggggataacgttc
ggaaacggacgctaataccgcatacgtcctacgggagaaagcgggggatcttcggacctc
gcgccattagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggc
gacgatccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccag
actcctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagcca
tgccgcgtgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggca
gtaacctaatacgttattgttttgacgttaccgacagaataagcaccggctaacttcgtg
ccagcagccgcggtaatacgaagggtgcaagcgttaatcggaattactgggcgtaaagcg
cgcgtaggtggttcagtaagttggaagtgaaatccccgggctcaacctgggaactgcttt
caaaactgctgagctagagtacggtagagggtggtggaatttcctgtgtagcggtgaaat
gcgtagatataggaaggaacaccagtggcgaaggcgaccacctggactgatactgacact
>W_Pseudomonas6
gtcgagcggatgagtgagcttgctcacggattcagcgg
cggacgggtgagtaatgcctaggaatctgcctggtagtgggggacaacgtttcgaaagga
acgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggccttgcgctatc
agatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggctacgatcc
gtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactcctac
46

Figure A-2 (cont’d)
gggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgccgcgt
gtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggttgtagatta
atactctgcaattttgacgttaccgacagaataagcaccggctaactctgtgccagcagc
cgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcgtagg
tggttcgttaagttggatgtgaaatccccgggctcaacctgggaactgcatccaaaactg
gcgagctagagtatggtagagggtggtggaatttcctgtgtagcggtgaaatgcgtagat
ataggaaggaacaccagtggcgaaggcgaccacctggactgatactgacactgaggtgcg
aaagcgtggggagcaacaggatagataccctggtagtccacgcgtaacgatgtcac
>W_Pseudomonas5
ttgcttct
cttgagagcggcggacgggtgagtaatacctaggaatctgcctgatagtgggggataacg
ttcggaaacggacgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggc
cttgcgctatcagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaa
ggctacgatccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtc
cagactcctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccag
ccatgccgcgtgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagg
gcattaacctaatacgttggtgtcttgacgttaccgacagaataagcaccggctaactct
gtgccagcagccgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaa
gcgcgcgtaggtggtttgttaagttgaatgtgaaatccccgggctcaacctgggaactgc
atccaaaactggcaagctagagtatggtagagggtagtggaatttcctgtgtagcggtga
aatgcgtagatataggaaggaacaccagtggcgaaggcgactacctggactgatactgac
actgaggtgcgaaagcgtggggagcaaacaggattagataccctggtagtccacgccgta
aacgatgtcaactagccgttgggagtcttgaactcttagtggcgcagctaacgcattaaa
gtgaccgcctggggagtacggccgc
>L_Pseudomonas2
gagaagcttgcttct
cttgagagcggcggacgggtgagtaatgcctaggaatctgcctggtggtgggggataacg
ttcggaaacggacgctaataccgcatacgtcctacgggagaaagcgggggatcttcggac
ctcgcgccattagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaa
ggcgacgatccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtc
cagactcctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccag
ccatgccgcgtgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagg
gcagtaacctaatacgttattgttttgacgttaccgacagaataagcaccggctaacttc
gtgccagcagccgcggtaatacgaagggtgcaagcgttaatcggaattactgggcgtaaa
gcgcgcgtaggtggttcagtaagttggaagtgaaatccccgggctcaacctgggaactgc
tttcaaaactgctgagctagagtacggtagagggtggtggaatttcctgtgtagcggtga
aatgcgtagatataggaaggaacaccagtggcgaaggcgaccacctggactgatactgac
actgaggtgcgaaagcgtggggagcaaacaggattagataccctggtagtccacgccgta
aacgatgtcaactagccgttggaatccttgagattttagtggcgcagctaacgcattaag
ttgaccgcctggggagtacggccgcagggtaaaactcaaatgaattgacggggg
>W_Pseudomonas1
gcttgctcctgaattcagcg
47

Figure A-2 (cont’d)
gcggacgggtgagtaatgcctaggaatctgcctggtagtgggggacaacgtttcgaaagg
aacgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggccttgcgctat
cagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggcgacgatc
cgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactccta
cgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgccgcg
tgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggcattaacct
aatacgttagtgttttgacgttaccgacagaataagcaccggctaactctgtgccagcag
ccgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcgtag
gtggtttgttaagttggatgtgaaatccccgggctcaacctgggaactgcattcaaaact
gacaagctagagtatggtagagggtggtggaatttcctgtgtagcggtgaaatgcgtaga
tataggaaggaacaccagtggcgaaggcgaccacctggactgatactgacactga
>W_Pseudomonas2
cttgccctcttgagagc
ggcggacgggtgagtaatacctaggaatctgcctggtagtgggggataacgttcggaaac
ggacgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggccttgcgcta
tcagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggctacgat
ccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactcct
acgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgccgc
gtgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggcattaacc
taatacgttagtgttttgacgttaccgacagaataagcaccggctaactctgtgccagca
gccgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcgta
ggtggtttgttaagttgaatgtgaaatccccgggctcaacctgggaactgcatccaaaac
tggcaagctagagtatggtagagggtagtggaatttcctgtgtagcggtgaaatgcgtag
atataggaaggaacaccagtggcgaaggcgactacctggactgatactgacactgaggtg
cgaaagcgtggggagcaaacaggat
>W_Pseudomonas3
ttgctcttcgattcagcg
gcggacgggtgagtaatgcctaggaatctgcctggtagtgggggacaacgtttcgaaagg
aacgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggccttgcgctat
cagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggctacgatc
cgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactccta
cgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgccgcg
tgtgtgaagaaggtcttcggattgtaaagcactttaagttgggaggaagggcagtaagct
aataccttgctgttttgacgttaccgacagaataagcaccggctaactctgtgccagcag
ccgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcgtag
gtggtttgttaagttggatgtgaaagccccgggctcaacctgggaactgcatccaaaact
ggcaagctagagtatggtagagggtggtggaatttcctgtgtagcggtgaaatgcgtaga
tataggaaggaacaccagtggcgaaggcgaccacctggactgatactgacact
>W_Pseudomonas4
agtcgagcggatgagagagcttgctcttcgattagc
ggcggacgggtgagtaatgcctaggaatctgcctggtagtgggggacaacgtttcgaaag
gaacgctaataccgcatacgtcctacgggagaaagcaggggaccttcgggccttgcgcta
48

Figure A-2 (cont’d)
tcagatgagcctaggtcggattagctagttggtgaggtaatggctcaccaaggcgacgat
ccgtaactggtctgagaggatgatcagtcacactggaactgagacacggtccagactcct
acgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagccatgccgc
gtgtgtgaagaaggtcttcggattgtaaagcactttaaggtgggaggaagggttgtagat
taatactctgcaattttgacgttaccgccagaataagcaccggctaactctgtgccagca
gccgcggtaatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcgcgta
ggtggtttgttaagtcggatgtgaaatccccgggctcaacctgggaactgcatccgaaac
tggcaagctagagtatggtagagggtagtggaatttcctgtgtagcggtgaaatgcgtag
atataggaaggaacaccagtggcgaaggcgactacctggactgatactgacactgaggtg
cgaaagcgtggggagcaaacaggattagataccctggtagtccacgccgtaaacgatgtc
aactagccgttggggtccttgagactttagtggcgcagctaacgcattaagttgaccg
>L_Rheineimera
ggggttttcggacctagcggcggacg
ggtgagtaatgcgtaggaagctacccgacagagggggataccagttggaaacgactgtta
ataccgcataatgtctacggaccaaagtgtgggaccttcgggccacatgctgtcggatgc
gcctacgtgggattagctagttggtgaggtaatggctcaccaaggcgacgatccctagct
ggtttgagaggatgatcagccacactggaactgagacacggtccagactcctacgggagg
cagcagtggggaatattggacaatgggcgcaagcctgatccagccatgccgcgtgtgtga
agaaggccttcgggttgtaaagcactttcagcgaggaggaagggttgtgtgttaatagca
catagccttgacgttactcgcagaagaagcaccggctaactctgtgccagcagccgcggt
aatacagagggtgcaagcgttaatcggaattactgggcgtaaagcgcacgcaggcggttg
gttaagtcagatgtgaaagccccgggctcaacctgggaattgcatttgaaactggccaac
tagagtacgtgagaggggggtagaattccaagtgtagcggtgaaatgcgtagagatttgg
aggaataccagtggcgaaggcggccccctggcacgatactgacgctcaggtgcgaaagcg
tggggagcaaacaggattagataccctggtagtccacgccgtaaacgatg
>L_Rhodococcus
gcggc
gaacgggtgagtaacacgtgggtgatctgccctgcactctgggataagcctgggaaactg
ggtctaatactggatatgacctcagcatgcatgtgctggggtggaaagcttttgtggtgc
aggatgggcccgcggcctatcagcttgttggtggggtaatggcctaccaaggcgacgacg
ggtagccgacctgagagggtgaccggccacactgggactgagacacggcccagactccta
cgggaggcagcagtggggaatattgcacaatgggcggaagcctgatgcagcgacgccgcg
tgagggatgaaggccttcgggttgtaaacctctttcagcagggacgaagcgtgagtgacg
gtacctgcagaagaagcaccggctaactacgtgccagcagccgcggtaatacgtagggtg
cgagcgttgtccggaattactgggcgtaaagagttcgtaggcggtttgtcgcgtcgtttg
tgaaaacccggggctcaacttcgggcttgcaggcgatacgggcagacttgagtgtttcag
gggagactggaattcctggtgtagcggtgaaatgcgcagatatcaggaggaacaccggtg
gcgaaggcgggtctctgggaaacaactgacgctgaggaacgaaagcgtgggtagcaaaca
ggattaaataccctgg
>W_Serratia
acgggagagcttgctc
49

Figure A-2 (cont’d)
tctgggtgacgagcggcggacgggtgagtaatgtctgggaaactgcctgatggaggggga
taactactggaaacggtagctaataccgcatgatgtcgcaagaccaaagtgggggacctt
cgggcctcacgccatcggatgtgcccagatgggattagctagtaggtggggtaatggctc
acctaggcgacgatccctagctggtctgagaggatgaccagccacactggaactgagaca
cggtccagactcctacgggaggcagcagtggggaatattgcacaatgggcgcaagcctga
tgcagccatgccgcgtgtgtgaagaaggccttagggttgtaaagcactttcagcgaggag
gaaggcgttgtagttaatagctgcaacgattgacgttactcgcagaagaagcaccggcta
actccgtgccagcagccgcggtaatacggagggtgcaagcgttaatcggaattactgggc
gtaaagcgcacgcaggcggtttgttaagtcagatgtgaaatccccgagcttaacttggga
actgcatttgaaactggcaagctagagtcttgtagaggggggtagaattccaggtgtagc
ggtgaaatgcgtagagatctggaggaataccggtggcgaaggcggccccctggacaaaga
ctgacgctcaggtgcgaaagcgtggggagcaaacaggattagataccctggtagtccacg
ctgtaaacgatgtcgacttggaggttgtgcccttgaggcgtggct
>W_Shewanella
gggagtttacttctg
aggtggcgagcggcggacgggtgagtaatgcctagggatctgcccagtcgagggggataa
cagttggaaacgactgctaataccgcatacgccctacgggggaaaggaggggaccttcgg
gccttccgcgattggatgaacctaggtgggattagctagttggtgaggtaatggctcacc
aaggcgacgatccctagctgttctgagaggatgatcagccacactgggactgagacacgg
cccagactcctacgggaggcagcagtggggaatattgcacaatgggggaaaccctgatgc
agccatgccgcgtgtgtgaagaaggccttcgggttgtaaagcactttcagtagggaggaa
agggtgtaatttaatacgctatatctgtgacgttacctacagaagaaggaccggctaact
ccgtgccagcagccgcggtaatacggagggtccgagcgttaatcggaattactgggcgta
aagcgtgcgcaggcggtttgttaagcgagatgtgaaagccctgggctcaacctaggaata
gcatttcgaactggcgaactagagtcttgtagaggggggtagaattccaggtgtagcggt
gaaatgcg
>W_Sphingobium
cttcagatctagtggcgcacgggt
gcgtaacgcgtgggaatctgcccttgggttcggaataacttctggaaacggaagctaata
ccggatgatgacgtaagtccaaagatttatcgcccaaggatgagcccgcgtaggattagc
tagttggtggggtaaaggcccaccaaggcgacgatccttagctggtctgagaggatgatc
agccacactgggactgagacacggcccagactcctacgggaggcagcagtagggaatatt
ggacaatgggcgaaagcctgatccagcaatgccgcgtgagtgatgaaggccttagggttg
taaagctcttttacccgggatgataatgacagtaccgggagaataagctccggctaactc
cgtgccagcagccgcggtaatacggagggagctagcgttgttcggaattactgggcgtaa
agcgcacgtaggcggctattcaagtcagaggtgaaagcccggggctcaaccccggaactg
cctttgaaactagatagcttgaatccaggagaggtgagtggaattccgagtgtagaggtg
aaattcgtagatattcggaagaacaccagtggcgaaggcggctcactggactggtattga
cgctgaggtgcgaaagcgtggggagcaaacaggattagataccctggtagtccacgccgt
aaacgatgataactagctgtcagggcacatgg
>L_Sphingobium
tcttcggatctagtggcgcacgggt
50

Figure A-2 (cont’d)
gcgtaacgcgtgggaatctgcccttgggttcggaataacttctggaaacggaagctaata
ccggatgatgacgtaagtccaaagatttatcgcccaaggatgagcccgcgtaggattagc
tagttggtggggtaaaggcccaccaaggcgacgatccttagctggtctgagaggatgatc
agccacactgggactgagacacggcccagactcctacgggaggcagcagtagggaatatt
ggacaatgggcgaaagcctgatccagcaatgccgcgtgagtgatgaaggccttagggttg
taaagctcttttacccgggatgataatgacagtaccgggagaataagctccggctaactc
cgtgccagcagccgcggtaatacggagggagctagcgttgttcggaattactgggcgtaa
agcgcacgtaggcggctattcaagtcagaggtgaaagcccggggctcaaccccggaactg
cctttgaaactagatagcttgaatccaggagaggtgagtggaattccgagtgtagaggtg
aaattcgtagatattcggaagaacaccagtggcgaaggcggctcactggactggtattga
cgc
>W_Sphingomonas
ggcgcacgg
gtgcgtaacgcgtgggaatctgccttggggttcggaataactccccgaaaggggtgctaa
taccggatgatgtcgaaagaccaaagatttatcgccctgagatgagcccgcgtaggatta
gctagttggtgtggtaaaggcgcaccaaggcgacgatccttagctggtctgagaggatga
tcagccacactgggactgagacacggcccagactcctacgggaggcagcagtggggaata
ttggacaatgggcgaaagcctgatccagcaatgccgcgtgagtgatgaaggccttagggt
tgtaaagctcttttacccgggaagataatgactgtaccgggagaataagccccggctaac
tccgtgccagcagccgcggtaatacggagggggctagcgttgttcggaattactgggcgt
aaagcgcacgtaggcggctttgtaagtcagaggtgaaagcctggagctcaactccagaac
tgcctttgagactgcatcgcttgaatccaggagaggtgagtggaattccgagtgtagagg
tgaaattcgtagatattcggaagaacaccagtggcgaaggcggctcactggactggtatt
gacgctgaggtgcgaaagcgtggggagcaaacaggattagataccctggtagtccacgcc
gtaaacgatgataactagctgtccgggtgcttggca
>L_Vogesella
gggagcttgctccgctgacgagtgg
cgaacgggtgagtaatgcgtcggaacgtgccgagtagtgggggataacgcagcgaaagtt
gtgctaataccgcatacgtactgaggtagaaagtgggggaccttcgggcctcacgctatt
cgagcggccgacgtctgattagctagtaggtgaggtaaaggctcacctaggcgacgatca
gtagcgggtctgagaggatgatccgccacactgggactgagacacggcccagactcctac
gggaggcagcagtggggaattttggacaatgggcgaaagcctgatccagccatgccgcgt
gtctgaagaaggccttcgggttgtaaaggacttttgtcagggaggaaatccccagtgtta
ataccgctgggggatgacagtacctgaagaataagcaccggctaactacgtgccagcagc
cgcggtaatacgtagggtgcaagcgttaatcggaattactgggcgtaaagcgtgcgcagg
cggtttgataagccagatgtgaaatccccgagctcaacttgggaactgcgtttggaactg
tcagactagagtgcgtcagaggggggtggaattccgcgtgtagcagtgaaatgcgtagag
atgcggaggaacaccgatggcgaaggcagccccctgggatgacactgacgctcatgcacg
aaagcgtggggagcaaacaggattagataccctggtagtccacgccctaaacgatgtcaa
ttagctgttgggggttagaatccctggtagcgtagctaacgcgtgaaattgaccgcctgg
ggagtacggccgcaaggttaaaa
>L_Williamsia
51

Figure A-2 (cont’d)
cctcctgatgcaacgacgc
cgccagagggatgacggccttcgggttgtaaacctctttcaccagggacgaagcgaaagt
gacggtacctggagaagaagcaccggccaactacgtgccagcagccgcggtaatacgtag
ggtgcgagcgttgtccggaattactgggcgtaaagagctcgtaggcggtttgtcgcgtcg
ttcgtgaaaacttggggcttaactccaagcgtgcgggcgatacgggcagacttgagtact
acaggggagactggaattcctggtgtagcggtgaaatgcgcagatatcaggaggaacacc
ggtggcgaaggcgggtctctgggtagtaactgacgctgaggaccgaaagcgtgggtagcg
aacaggattagataccctggtagtccacgccgtaaacggtgggt
>L_Aeromicrobium1
ttcgggagtacacgag
cggcgaacgggtgagtaacacgtgagcaatctgcccttctcatcggaataaccattggaa
acgatggctaatgccgaatacgacctcctttcgcatgatcggaggtggaaagctccggcg
gagaaggatgagctcgcggcctatcagctagttggcggggtaacggcccaccaaggcgac
gacgggtagccggcctgagagggtgaccggccacactgggactgagacacggcccagact
cctacgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagcaacgc
cgcgtgagggatgacggccttcgggttgtaaacctctttcagcagggacgaagcgaaagt
gacggtacctgcagaagaaggaccggccaactacgtgccagcagccgcggtaatacgtag
ggtccgagcgttgtccggaattattgggcgtaaagggctcgtaggcggtttgtcgcgtcg
ggagtgaaaactcagggcttaaccctgagcgtgcttccgatacgggcagactagaggtat
tcaggggagaacggaattcctggtgtagcggtggaatgcgcagatatcaggaggaacacc
ggtggcgaaggcggttctctgggaatacctgacgct
>W_Rhodococcus
gcgaacgggtgagtaacacgtgggatgatctgccctgcacttcgggataagcccggga
aactgggtctaataccggatatgaccacagcatgcatgtgttgtggtggaaagcttttgc
ggtgtgggatgggcccgcggcctatcagcttgttggtggggtaatggcctaccaaggcga
cgacgggtagccggcctgagagggcgaccggccacactgggactgagacacggcccagac
tcctacgggaggcagcagtggggaatattgcacaatgggcgcaagcctgatgcagcgacg
ccgcgtgagggatgacggccttcgggttgtaaacctctttcagcagggacgaagcgcaag
tgacggtacctgcagaagaagcaccggccaactacgtgccagcagccgcggtaatacgta
gggtgcaagcgttgtccggaattactgggcgtaaagagctcgtaggcggtttgtcgcgtc
gtctgtgaaaaccagcagctcaactgttggcttgcaggcgatacgggcagacttgagtat
ttcaggggagactggaattcctggtgtagcggtgaaatgcgcagatatcaggaggaacac
cggtggcgaaggcgggtctctgggaaataactgacgctgaggagcgaaagcgtgggtagc
gaa
>W_Mycobacterium
ggcgaacgggtgagtaacacgtgggtgatctgccctgcactttgggataagcctgggaa
actgggtctaataccgaatatgaccatgcgcctcctggtgtgtggtggaaagcttttgcg
gtgtgggatgggcccgcggcctatcagcttgttggtggggtaatggcctaccaaggcgac
gacgggtagccggcctgagagggtgaccggccacactgggactgagatacggcccagact
cctacgggaggcagcagtggggaatattgcacaatgggcgcaagcctgatgcagcgacgc
cgcgtgagggatgacggccttcgggttgtaaacctctttcagcacagacgaagcgcaagt
gacggtatgtgcagaagaaggaccggccaactacgtgccagcagccgcggtaatacgtag
52

Figure A-2 (cont’d)
ggtccgagcgttgtccggaattactgggcgtaaagagctcgtaggtggtttgtcgcgttg
ttcgtgaaaactcacagcttaactgtgggcgtgcgggcgatacgggcagacttgagtact
gcaggggagactggaattcctggtgtagcggtggaatgcgcagatatcaggaggaacacc
ggtggcgaaggcgggtctctgggcagtaactgacgctgaggagcgaaagcgtggggagcg
aacaggattagataccctggtagtccacgccgtaa
>L_Aeromicrobium3
tacaggtaccaggctccttcgggagtacacgagcgg
cgaacgggtgagtaacacgtgagcaatctgcccttctcatcggaataaccattggaaacg
atggctaatgccgaatacgacctcctttcgcatgatcggaggtggaaagctccggcggag
aaggatgagctcgcggcctatcagctagttggcggggtaacggcccaccaaggcgacgac
gggtagccggcctgagagggtgaccggccacactgggactgagacacggcccagactcct
acgggaggcagcagtggggaatattggacaatgggcgaaagcctgatccagcaacgccgc
gtgagggatgacggccttcgggttgtaaacctctttcagcagggacgaagcgaaagtgac
ggtacctgcagaagaaggaccggccaactacgtgccagcagccgcggtaatacgtagggt
ccgagcgttgtccggaattattgggcgtaaagggctcgtaggcggtttgtcgcgtcggga
gtgaaaactcagggcttaaccctgagcgtgcttccgatacgggcagactagaggtattca
ggggagaacggaattcctggtgtagcggtggaatgcgcagatatcaggaggaacaccggt
ggcgaaggcggttctctgggaatacctgacgctgaggagcgaaagcatgggtagcgaaca
gga

53

APPENDIX B
Supplementary Methods
Final Contents of Phosphorus-Defined Media
Chemical

Final Concentration (µM)

CaCl2 • 2H2O

250

MgSO4 • 7H2O

150

NaHCO3

150

NH4Cl

250

KNO3

250

CuSO4 • 5H2O

0.04

ZnSO4 • 7H2O

0.08

CoCl2 • 6H2O

0.04

MnCl2 • 4H2O

0.91

NH4Mo7O24 • 4H2O

0.03

FeCl3 • 6H2O

12

H3BO3

2.1

Na2EDTA • H2O

4.36mg/L

H2O3Se
HEPES buffer

0.6
2.38g/L

Defined Media Recipe
Chemical

Final Concentration

1000X Major Elements Working Stock
1000X Trace Elements Working Stock
1000X Vitamin Working Stock
100X Carbon Source Stock*
Phosphorus source (see Table 2-1)*
HEPES buffer
Cyclohexamide*

1X
1X
1X
1X
Variable. See text for details.
2.38g/L
50 mg/L

*Filter-sterilized.

54

WORKING STOCK SOLUTIONS
1000X Major Elements Working Stock
Chemical

Final Concentration (mM)

CaCl2 • 2H2O

250

MgSO4 • 7H2O

150

NaHCO3

150

NH4Cl

250

KNO3

250

H2O3Se

0.6

H3BO3

2.1

1000X Trace Elements Working Stock
Chemical

Final Concentration (mM)

CuSO4 • 5H2O

0.04

FeCl3 • 6H2O

12

CoCl2 • 6H2O

0.04

MnCl2 • 4H2O

0.91

NH4Mo7O24 • 4H2O

0.03

ZnSO4 • 7H2O

0.08

Na2EDTA • H2O

4.36g/L

Vitamin Working Stock
Chemical

Final Concentration (mg/L)

Biotin
Thiamine HCl

1.0
200

Carbon Source Working Stock

55

Chemical

Final Concentration (g/L)

Glycine
Acetate
Dextrose
NaSuccinate • 6H2O

31.25
31.25
93.75
93.75

PRIMARY STOCK SOLUTIONS
Chemical

Stock Concentration (mM)

CaCl2 • 2H2O

250

MgSO4 • 7H2O

150

NaHCO3

150

NH4Cl

500

KNO3

500

H2O3Se

0.6

CuSO4 • 5H2O

40

ZnSO4 • 7H2O

80

CoCl2 • 6H2O

40

MnCl2 • 4H2O

910

NH4Mo7O24 • 4H2O

30

H3BO3
Biotin
Cyclohexamide

2.1
0.10 g/L
25 g/L

56

REERENCES

57

REFERENCES

Beiko, R.G., Harlow, T.J. & Ragan, M.A. (2005). Highways of gene sharing in
prokaryotes. Proc. Natl. Acad. Sci. USA, 102, 14332-14337.
Bolier, G., Koningh, M.C.J., Schmale, J.C. & Donze, M. (1992). Differential luxury
phosphate response of planktonic algae to phosphorus removal. Hydrobiologia 243244, 113-118.
Bonsall, M.B., Hassell, M.P. & Asefa, G. (2002). Ecological trade-offs, resource
partitioning, and coexistence in a host-parasitoid assemblage. Ecology, 83, 925-934.
Buckling, A., Brockhurst, M.A., Travisano, M. & Rainey, P.B. (2007). Experimental
adaptation to high and low quality environments under different scales of temporal
variation. J. Evol. Biol., 20, 296-300.
Cáceres, C.E. Temporal variation, dormancy, and coexistence: a field test of the
storage effect. Proc. Natl. Acad. Sci. USA, 94, 9171-9175 (1997).
Caley, M.J. & Munday, P.L. (2003). Growth trades off with habitat specialization. Proc.
R Soc B., 270, S175-S177.
Cambui, C.A., Svennerstam, H., Gruffman, L., Nordin, A., Ganeteg, U., & Näsholm, T.
(2011). Patterns of Plant Biomass Partitioning Depend on Nitrogen Source. PLoS ONE,
6, 7.
Cardinale, B.J. (2011). Biodiversity improves water quality through niche partitioning.
Nature, 472, 86-89.
Chow, S.S., Wilke, C.O., Ofria, C., Lenski, R.E. & Adami, C. (2004). Adaptive radiation
from resource competition in digital organisms. Science, 305, 84-86.
Clark, L.L., Ingall, E.D., and Benner, R. (1998). Marine phosphorus is selectively
remineralized. Nature, 393, 426.

58

Cotner, J.B. & Wetzel, R.G. (1991). 5′-Nucleotidase Activity in a Eutrophic Lake and an
Oligotrophic Lake. Appl. Environ. Microbiol., 57, 1306-1312.
Dyhrman, S.T., Ammerman, J.W., & Van Mooy, B.A.S. (2007). Microbes and the
Marine Phosphorus Cycle. Oceanogr., 20, 110-116.
Finke, D.L., & Snyder, W.E. (2008). Niche partitioning increases resource exploitation
by diverse communities. Science, 321, 1488-1490.
Fox, J. & Weisberg, S. (2011). An {R} Companion to Applied Regression, Second
Edition. Thousand Oaks CA: Sage. URL:
http://socserv.socsci.mcmaster.ca/jfox/Books/Companion
Frette, L., Jørgensen, N.O.G., Nybroe, O., Del Giorgio, P.A. & Kroer, N. (2009). Effect of
availability of nitrogen compounds on community structure of aquatic bacteria in model
systems. Microbial Ecol., 57, 104-116.
Futuyma, D.J. & Moreno, G. (1988). The Evolution of Ecological Specialization. Annu.
Rev. Ecol. Syst., 19, 207-233.
Futuyma, D.J. (2010). Evolutionary constraint and ecological consequences. Evolution,
64, 1865-1884.
Gravel, D., Bell, T., Barbera, C., Bouvier, T., Pommier, T., Venail, P., et al. (2011).
Experimental niche evolution alters the strength of the diversity–productivity
relationship. Nature, 469, 89-92.
Gudelj, I., Beardmore, R.E., Arkin, S.S. & MacLean, R.C. (2007). Constraints on
microbial metabolism drive evolutionary diversification in homogeneous environments. J
Evol. Biol., 20, 1882-1889.
Guindon, S. & Gascuel, O. (2003). A simple, fast and accurate method to estimate large
phylogenies by maximum-likelihood. Syst. Biol., 52, 696-704.
Hall, A.R. & Colegrave, N. (2007). How does resource supply affect evolutionary
diversification? Proc. R. Soc. Lond. B, 274, 73-78.

59

Hill, J.E., Kysela, D. & Elimelech, M. (2007). Isolation and assessment of phytatehydrolysing bacteria from the DelMarVa Peninsula. Environ. Microbiol., 9, 3100-3107.
Huang, J., Su, Z. & Xu, Y. (2005). The evolution of microbial phosphonate degradative
pathways. J Molec. Evol., 61, 682-690.
Jessup, C.M. & Bohannan, B.J.M. (2008). The shape of an ecological trade-off varies
with environment. Ecol. Lett., 11, 947-959.
Jones, S.E. & Lennon, J.T. (2010). Dormancy contributes to the maintenance of
microbial diversity. Proc. Natl. Acad. Sci. USA, 107, 5881-5886.
Karl, D.M. (2000). Phosphorus, the staff of life. Nature, 406, 31–33.
Kassen, R. (2002). The experimental evolution of specialists, generalists, and the
maintenance of diversity. J Evol. Biol., 15, 173-190.
Kelley, S.T. & Farrell, B.D. (1988). Is specialization a dead end? The Phylogeny of
Dendroctonus bark beetles (Scolytidae). Evolution, 52, 1731-1743.
Lennon, J.T. (2007). Diversity and Metabolism of Marine Bacteria Cultivated on
Dissolved DNA. Appl. Environ. Microbiol., 73, 2799-2805.
Lennon, J.T., Khatana, S.A.M., Marston, M.F. & Martiny, J.B.H. (2007). Is there a cost
of virus resistance in marine cyanobacteria? ISME J, 1, 300-12.
Levins, R. (1968). Evolution in changing environments. Monographs in Population
Biology. Vol 2:2. Princeton University Press, Princeton.
Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar, A., et al.
(2004). ARB: a software environment for sequence data. Nucleic Acids Res., 32, 13631371.
Luo, H., Benner, R. Long, R.A., & Hu, J. (2009). Subcellular localization of marine
bacterial alkaline phosphatases. Proc. Natl. Acad. Sci. USA, 106, 21219-21223.

60

Martiny, A.C., Coleman, M.L., & Chisholm, S.W. (2006). Phosphate acquisition genes in
Prochlorococcus ecotypes: Evidence for genome-wide adaptation. Proc. Natl. Acad.
Sci. USA, 103, 12552-12557.
Martiny, A.C., Huang, Y., & Li, W. (2009). Occurrence of phosphate acquisition genes in
Prochlorococcus cells from different ocean regions. Environ. Microbiol., 11, 1340-1347.
McKane, R.B., Johnson, L.C., Shaver, G.R., Nadelhoffer, K.J., Rastetter, E.B., Fry, B, et
al. (2002). Resource-based niches provide a basis for plant species diversity and
dominance in arctic tundra. Nature, 415, 68-71.
Mou, X., Sun, S., Edwards, R.A., Hodson, R.E. & Moran, M.A. (2008). Bacterial carbon
processing by generalist species in the coastal ocean. Nature, 451, 708-711.
Oakley, B.B., Carbonero, F., Gast, V.D., Christopher, J., Hawkins, R.J., & Purdy, K.J.
(2010). Evolutionary divergence and biogeography of sympatric niche-differentiated
bacterial populations. ISME J, 4, 488-497.
Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., O’Hara, R. B., Simpson, G. L., et
al. (2011). vegan: Community Ecology Package. R package version 1.17-6.
http://CRAN.R-project.org/package=vegan
Orchard, E.D., Webb, E.A. & Dyhrman, S.T. (2009). Molecular analysis of the
phosphorus starvation response in Trichodesmium spp. Environ. Microbiol., 11, 24002411.
Pagel, M. (1997). Inferring evolutionary processes from phylogenies. Zoologica Scripta
26, 331- 348.
Pagel, M. (1999). Inferring the historical patterns of biological evolution. Nature, 401,
877-884.
Pagel, M. & Meade, A. (2004). A Phylogenetic Mixture Model for Detecting PatternHeterogeneity in Gene Sequence or Character-State Data. Syst. Biol., 53, 571-581.
Posada, D. (2008). jModelTest: Phylogenetic Model Averaging. Mol. Biol. Evol., 25,
1253-1256.

61

Prinzing, A., Durka, W., Klotz, S. & Brandl, R. (2001). The niche of higher plants:
evidence for phylogenetic conservatism The niche of higher plants : evidence for
phylogenetic conservatism. Proc. R. Soc. Lond. B, 268, 2383-2389.
R Development Core Team (2010). R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-070, URL http://www.R-project.org.
Richmond, C.E., Breitburg, D.L. & Rose, K.A. (2005). The role of environmental
generalist species in ecosystem function. Ecol. Model.,188, 279-295.
Rodríguez, H. & Fraga, R. (1999). Phosphate solubilizing bacteria and their role in plant
growth promotion. Biotech. Adv., 17, 319-339.
Schloss, P.D., et al. (2009). Introducing mothur: Open-source, platform-independent,
community-supported software for describing and comparing microbial communities.
Appl. Environ. Microbiol., 75(23):7537-41.
Singleton, D. R., M. A. Furlong, S. L. Rathbun, and W. B. Whitman. 2001. Quantitative
comparisons of 16S rDNA sequence libraries from environmental samples. Appl.
Environ. Microbiol., 67:4373-4376.
Siuda, W., & Chrost, R.J. (2001). Utilization of selected dissolved organic phosphorus
compounds by bacteria in lake water under non-limiting orthophosphate conditions. Pol.
J. Environ. Stud., 10, 475-483.
Smith, E. (1982). Niche breadth, resource availability, and inference. Ecology,
63(6):1675-1681.
Smith, R.G., Mortensen, D.A., & Ryan, M.R. (2010). A new hypothesis for the functional
role of diversity in mediating resource pools and weed-crop competition in
agroecosystems. Weed Res., 50, 37-48.
Stomp, M., Huisman, J., deJongh, F., Veraart, A.J., Gerla, D., Rijkeboer, M., et al.
(2004). Adaptive divergence in pigment composition promotes phytoplankton
biodiversity. Nature, 432, 104-107.

62

Suzuki, R. & Shimodaira, H. (2009). pvclust: Hierarchical Clustering with P-Values via
Multiscale Bootstrap Resampling. R package version 1.2-1.
http://www.is.titech.ac.jp/~shimo/prog/pvclust/.
Vershinina, O.A. & Znanenskaya, L.V. (2002). The pho regulons of bacteria. Microbiol.,
71, 497-511.
von Felten, S., Hector, A., Buchmann, N., Niklaus, P.A., Schmid, B., & SchererLorenzen, M. (2009). Belowground nitrogen partitioning in experimental grassland plant
communities of varying species richness. Ecology, 90, 1389-1399.
Waldrop, M.P., Zak, D.R., Blackwood, C.B., Curtis, C.D. & Tilman, D. (2006). Resource
availability controls fungal diversity across a plant diversity gradient. Ecol. Lett., 9, 11271135.
Wang, Q, G. M. Garrity, J. M. Tiedje, and J. R. Cole. (2007). Naïve Bayesian Classifier
for Rapid Assignment of rRNA Sequences into the New Bacterial Taxonomy. Appl.
Environ. Microbiol., 73, 5261-7.
White, A.E. (2009). New insights into bacterial acquisition of phosphorus in the surface
ocean. Proc. Natl. Acad. Sci. USA, 106, 21013-21014.
Zubkov, M.V., Fuchs, B.M., Tarran, G.A., Burkill, P.H. & Amann, R. (2003). High rate of
uptake of organic nitrogen compounds by Prochlorococcus cyanobacteria as a key to
their dominance in oligotrophic oceanic waters. Appl. Environ. Microbiol., 69, 12991304.
Zwietering, M.H., Jongenburger, I., Rombouts, F.M. & Van ’t Riet, K. (1990). Modeling
of the Bacterial Growth Curve. Appl. Environ. Microbiol., 56, 1875-1881.

63

CHAPTER 3
FUTURE DIRECTIONS

P form may have large impacts on community and ecosystem properties, but these
are neither quantified nor well understood. The present study was an important first step
in understanding how P form can impact the life history of bacteria, whose uptake and
transformation of dissolved P is crucial for nutrient cycling. Yet completion of a few
follow-up research studies could greatly increase our understanding of the impacts of P
form in natural systems. Do bacteria demonstrate niche partitioning according to P form
or abilities? Can P form significantly influence community composition and ecosystem
functions? Experimental results from studies addressing these questions would
contribute greatly to increase our understanding of the maintenance of species diversity,
community dynamics, and ecosystem functioning.
Since bacteria vary greatly in their ability to use diverse P forms and demonstrate
the ability to specialize on P forms, it seems reasonable that they may also partition P
resources. Resource partitioning through niche differentiation can facilitate species
coexistence, thereby increasing species diversity (Chesson 2009). One can
experimentally test for the ability to partition resources by conducting a series of
competition experiments in chemostats with different P environments. For these
experiments, two isolates should be chosen such that each performs best on a different
P source and has a clear disadvantage on the P source the other performs best on. For
example, W_Pseudomonas1 and L_Psuedomonas2 would be good candidates, since
W_Pseudomonas1 grows very well on AEP, but poorly on Phyt, while
L_Psuedomonas2 grows well on Phyt, but poorly on AEP. Positive evidence for niche

64

partitioning would be found if isolates competitively exclude each other when grown in
the most advantageous P environment of each, but coexist when grown with both P
sources available.
P form influences bacterial performance traits and ecology, but can it also have
broader-scale impacts? The often substantial effects of inorganic phosphate additions
on community dynamics and ecosystem functions are well documented (Carpenter et
al. 1998, Smith 2003). However, theoretically all members of plant, algal, and bacterial
communities can access Pi, while access to P resources from Porg is likely
predominantly mediated by bacteria. In P-limited environments, this possible shift to a
bacterial-controlled P limitation may yield different effects on communities and
ecosystem functions. Yet few studies have included diverse phosphorus sources when
comparing community or ecosystem impacts, limiting our understanding of the
importance of P form at these scales.
A mesoscale experiment investigating the community and ecosystem impacts of
diverse P forms would be a valuable contribution. For example, cattle tanks could easily
be used as replicated aquatic mesocosms. These mesocosms could be ‘seeded’ with
microbes and macroinvertebrates from local lakes and allowed to reach a stable state
over time. Several different P forms, such as ATP, Phyt, AEP, and Pi could be added. A
mixed P treatment with equal concentrations of each added P form could provide
valuable insight into broad-scale effects of P resource diversity. Ecosystem-level
variables could then be quantified, such as ecosystem respiration and primary
productivity, nutrient concentrations and stoichiometries, and total biomass across
trophic levels (here, microbes and macroinvertebrates). Measured community-level

65

responses could include bacterial respiration and productivity (and thus growth
efficiencies), and bacterial and zooplankton community compositions and diversity.

66

REFERENCES

67

REFERENCES

Carpenter, S.R., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N., and
Smith, V. H. (1998). Nonpoint pollution of surface waters with phosphorus and nitrogen.
Ecol. Appl., 8, 559-568.
Chesson, P. (2000). Mechanisms of maintenance of species diversity. Annu. Rev. Ecol.
Syst., 31, 343-366.
Smith, V.H. (2003). Eutrophication of freshwater and coastal marine ecosystems: a
global problem. Environ. Sci. Pollut. Res. Int., 10, 126-139.

68