RESEARCH ARTICLE
Effects of differences in organic supply on bacterial diversity
subject to viral lysis
Birte Töpper, Tron Frede Thingstad & Ruth-Anne Sandaa
Department of Biology, University of Bergen, Bergen, Norway
Correspondence: Birte Töpper, Department
of Biology, University of Bergen,
Thormøhlensgate 53A, 5006 Bergen,
Norway. Tel.: +47 55584644;
fax: +47 55584450;
e-mail: [email protected]
Received 5 March 2012; revised 27 July
2012; accepted 28 July 2012.
Final version published online 22 August
2012.
DOI: 10.1111/j.1574-6941.2012.01463.x
MICROBIOLOGY ECOLOGY
Editor: Tillmann Lueders
Keywords
bacterial diversity; 16S rRNA gene; organic
carbon composition; virus control;
Thalassiosira sp.; Phaeocystis pouchetii.
Abstract
Bacterial diversity is believed to be controlled both by bottom-up and
top-down mechanisms such as nutrient competition, predation and viral lysis.
We hypothesise that lytic viruses create trophic niches within bacterial communities, and thus primarily control richness and evenness, while substrate
composition primarily controls community composition, that is, the inhabitants of these niches. To investigate this, we studied diversity of mixed bacterial
communities subject to viruses under different regimes of organic matter supply. From a predator-free inoculum, bacterial communities were allowed to
develop in batch cultures where the organic substrate was either a single compound [glucose (G)] or more complex mixtures produced by phytoplankton
[Phaeocystis pouchetii (P) or Thalassiosira sp. (T)]. Throughout the experiment,
c. 98% of the sequences in treatment G belonged to the Gammaproteobacteria
class, which dominated also in the initial phase of the other treatments
[T (c. 87%) and P (62%)]. In treatment T, the composition shifted to a dominance of Alphaproteobacteria (c. 37%), while in P, the proportion of Gammaproteobacteria remained stable. Richness increased with increasing substrate
complexity, while evenness remained similar in the different treatments. The
results suggest that both substrate composition (bottom-up) and viral lysis
(top-down) operate simultaneously in the control of bacterial diversity. Despite
the reduction in factors supposed to influence prokaryote diversity, the system
was still complex if taken into account the potential synergistic interactions
within and between the remaining factors.
Introduction
Central to contemporary microbial ecology is the two-faceted
effort of quantifying microbial biodiversity and understanding how biodiversity is controlled. Classes of environmental factors potentially contributing to biodiversity
include spatial (Venail et al., 2010) and temporal (Connell, 1979) variability. In addition to such characteristics
of the environment, trophic interactions, including competition for different substrates (Tilman, 1977), the role
of selective predators and the role of lytic viruses (Sandaa
et al., 2009; Winter et al., 2010), are believed to influence
the diversity of microbial communities. The trophic factors of substrate, predatory and viral control obviously
work in concert and are difficult to separate both
experimentally and conceptually. To illustrate principles,
models (Thingstad, 2000) have been developed using the
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
simplifying assumption that predation on bacteria is nonselective. This is clearly an approximation, because studies
have shown different combinations of resistance to predation and viral lysis in bacterial isolates (Simek et al.,
2010). To avoid this problem, we here have focused on
the bottom-up control by organic carbon sources of different complexity, in combination with viral control, in
predator-free cultures. A simplified conceptual description
of this situation can be constructed where total abundance of individuals in the community is given by the
amount of organic material available, while lytic viruses
split this community into trophic niches, here termed
‘host groups’. The number of such host groups is thus a
measure of community richness, while the distribution of
abundances of individuals among these host groups represents the evenness of the community. Which organisms
that fill these niches should, however, depend upon the
FEMS Microbiol Ecol 83 (2013) 202–213
203
Control mechanisms for bacterial diversity
hosts’ abilities to use and compete for the substrate(s)
available. Community composition is therefore expected
to be sensitive to the spectrum of compounds present in
the organic supply. On the basis of these concepts, we
wanted to test whether one could experimentally support
the simplifying hypothesis that richness and evenness are
properties primarily top-down controlled by viruses,
while community composition is primarily controlled by
the substrate(s) available.
Heterotrophic bacteria are known to have distinct
carbon uptake patterns, depending on the molecular weight
(Cottrell & Kirchman, 2000; Elifantz et al., 2005) and the
availability of organic carbon compounds (Alonso & Pernthaler, 2006). Glucose, as a low-molecular-weight carbon, is
preferred by bacteria belonging to the Alphaproteobacteria
(Elifantz et al., 2005; Alonso-Saez & Gasol, 2007), but also
bacteria belonging to the Gammaproteobacteria are known
to be easily stimulated by high concentrations of glucose
(Øvreås et al., 2003; Pinhassi & Berman, 2003; Harvey
et al., 2006). Extracellular DOC released by different phytoplankton communities differs in quantity (Wolter, 1982;
Lancelot, 1983) and quality (Myklestad, 1995). Thus, considering the varying carbon preferences of heterotrophic
bacteria, the phytoplankton community will also play an
important role in structuring the bacterial community
(Pinhassi et al., 2004).
A laboratory experiment was preformed to investigate
the effect of substrate composition (bottom-up) and viral
control (top-down) on the bacterial community. In two
treatments, the bacterial community was provided with
presumably complex carbon compounds released by two
different phytoplankton populations – Thalassiosira sp.
and Phaeocystis pouchetii, respectively. In another treatment, only a single carbon compound in form of glucose
was added to the bacterial community. Further, a treatment with no additional carbon compounds was used as
a control. Owing to differences in carbon supply, different species are expected to fill the created host niches in
the four treatments. Host–virus interactions may also be
different because of differences in host communities, and
richness and evenness in host groups may therefore also
change because of changes in the organic matter supply.
As a first approximation, however, we hypothesise that
host–virus interactions lead to the same number and size
of niches. In this case, differences in substrate composition would be expected to give large changes in community composition, but only minor changes in richness and
evenness indices. The dynamics within the bacterial community was determined by analysing 16S ribosomal RNA
gene (rDNA) fragments with denaturing gradient gel
electrophoresis (DGGE), while the bacterial community
structure and diversity were analysed by construction of
clone libraries and analysis of 16S rRNA gene sequences.
FEMS Microbiol Ecol 83 (2013) 202–213
Materials and methods
Experiment setup
A laboratory experiment with four different carbon treatments, each with three replicates, was performed in an incubation cabinet (8 °C, 24-h light) for 3 weeks in June 2010.
On day 0, each flask was filled with 1.3 L seawater medium
(IMR/2, 0.5 lM PO3
4 , 16 lM NO3 ) and 200 mL prefiltered (GF/F filter on top of a 0.8-lm polycarbonate filter)
fresh seawater from Raunefjorden, Norway, (N60°15.587
E005°08.448, 5 m depth) including the ambient bacterial
community. DOC was either provided by glucose (treatment
G) or by one of the two phytoplankton populations; the diatom Thalassiosira sp. (treatment T) isolated from Kongsfjorden, Svalbard (Thingstad et al., 2008), or the flagellate
P. pouchetii (treatment P). Glucose was added in three times
the Redfield ratio in terms of carbon relative to the phosphorus additions (159 lM C). One treatment without any
additional carbon was used as a control (treatment C).
Treatment T received additional silicate (16 lM) into the
dialysis bags. Approximately 30 mL of the cultures (1.5–
8.9 9 106 cells mL1) was filled into dialysis bags (Spectra/
Por Biotech, Regenerated Cellulose Dialysis Membranes,
MWCO: 10 000 Da) before placed in the flasks to avoid disturbance of the test bacterial assemblages by bacteria present
in the phytoplankton cultures. Dialysis bags containing the
glucose solution (treatment G) and empty dialysis bags
(treatment C) were used for the other treatments. Three
hundred millilitres from each culture was collected every
second day for analysis, and the sampled volume was
replaced each time with an equal volume of IMR/2 medium.
Dialysis bags were replaced with new bags containing fresh
algae culture or glucose solution concurrent with sampling.
Total number of bacteria and viruses
Bacteria and virus abundance were determined using a
FACSCalibur flow cytometer (Becton Dickinson)
equipped with an air-cooled laser providing 15 mW at
488 nm and with standard filter set-up. Flow cytometer
instrumentation and methodology followed the recommendations of Marie et al. (1999). The cells were fixed
with glutaraldehyde for 30 min at 4 °C (final concentration, 0.5%), snap frozen in liquid nitrogen and stored at
20 °C until further analysis. Appropriate dilutions of
fixed samples were prepared in 0.2 lm filtered TE buffer
and stained with SYBR Green I (Invitrogen, Carlsbad,
CA) for 10 min in the dark, either at room temperature
for bacteria counts or at 80 °C for virus counts. Fluorescent beads (Molecular Probes) were used as standard.
The samples were analysed for 1 min and the discriminator set on green fluorescence. Discrimination of bacteria
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
204
and virus populations was based on groups observed in
scatter plots of SSC signal versus green DNA dye fluorescence (SYBR Green I). Cell numbers were calculated from
volumetric measurements of instrument flow rate.
DNA isolation, PCR and DGGE
Sample volumes of 150–250 mL were filtered through a
sterile 0.2-lm polycarbonate filter (47 mm) and stored at
20 °C until further processing. Different sample volumes
were a consequence of hampered filtration because of
increasing bacterial biomass. DNA isolation, PCR and
DGGE were performed as described in Töpper et al.
(2010). An internal standard for the DGGE gels was
created using 16S rRNA genes amplicons of four bacteria
isolates: Gelidibacter algens (DSMZ: 12408), Microbacterium maritypicum (DSMZ: 12512), Flexibacter aurantiacus
(DSMZ: 6792) and Sulfitobacter mediterraneus (DSMZ:
12244). PCR amplification of F. aurantiacus and M. maritypicum 16S rRNA genes was performed as described
previously (Töpper et al., 2010), while G. algens and
S. mediterraneus 16S rRNA genes were amplified using the
primers 357f (Escherichia coli position 341–357), with 40bp GC-clamp, and 907rM (E. coli position, 907–926), with
a touchdown PCR program (94 °C for 5 min; 10 cycles
with 94 °C for 1 min, 65 °C for 1 min (1 °C/cycle),
72 °C for 1 min; 25 cycles with 94 °C for 1 min, 55 °C
for 1 min, 72 °C for 1 min; 72 °C for 10 min; 4 °C).
Digitalised DGGE images were analysed using the
program GEL2k (Svein Norland, Department of Biology,
University of Bergen), which determines presence/absence
and intensity of bands based on pixel intensity.
Cloning and sequencing
For generation of clone libraries, partial 16S rRNA genes
were amplified with the same primers used for DGGE but
without the GC-clamp. The PCR products were inserted
into the vector pSC-A-amp/kan and transformed into
competent E. coli cells (StrataCloneTM PCR Cloning Kit,
Stratagene, California). Blue-white screening of transformants was performed by plating on agar plates amended
with 100 mg mL1 ampicillin and 5-bromo-4-chloro3-indolyl-b-D-galactopyranoside (X-Gal) according to kit
instructions. White colonies were picked and sent for
sequencing at LGC Genomics (Berlin, Germany).
The sequences were processed using the Ribosomal
Database Projects (RDPII) Pyrosequencing Pipeline
(http://pyro.cme.msu.edu/) and aligned to the closest relative in the GenBank database using BLAST (Zhang et al.,
2000). The chimera check was performed using the program Black Box Chimera Check B2C2 (Gontcharova
et al., 2010).
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
B. Töpper et al.
Statistical analysis
To capture the different host groups, threshold levels
were set on 99% similarity for calculation of the OTUs.
Shannon index (H′), Chao1 index (chao) and Pielou’s
evenness index (J′) of sequence data sets were calculated
by complete linkage clustering (RDPII) with a 1% dissimilarity cut-off. To assess differences between Shannon
indices of different treatments, adjusted t-tests and
adjusted degrees of freedom (Redfield) were applied
(Magurran, 1988):
H10 H20
t ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VarH10 þ VarH20
df ¼
ðVarH10 þ VarH20 Þ2
½ðVarH10 Þ2 =N1 þ ½ðVarH20 Þ2 =N2 H10 and H20 are the Shannon indices, VarH10 and VarH20 ,
the respective variances and N1 and N2, the respective
number of observations of the two treatments under
comparison.
Results
Total number of bacteria and viruses
The bacterial abundance was 2.6 9 104 cells mL1 for all
treatments in the beginning of the experiment (day 0,
Fig. 1a). During the experiment, highest bacterial abundance was measured in the glucose treatments, while the
control treatments showed lowest cell numbers. Bacteria
abundance increased in the phytoplankton treatments
until day 10 [with a average peak at 1.7 9 107 cells mL1
(T) and at 1.9 9 107 cells mL1 (P)] and until day 12 in
the glucose treatment (with an average peak at
2.8 9 107 cell mL1), while the bacterial abundance of
the control treatments increased only until day 6 (average
peak at 1.1 9 107 cell mL1). After a minor decrease in
cell numbers on day 12, both of the phytoplankton treatments were characterised by another peak on day 18 (T,
2.2 9 107 cells mL1) and day 16 (P, 2.1 9 107 cells mL1),
respectively. Until day 16, the bacterial abundance in the P
treatments was higher than in the T treatments. After day 16,
highest bacteria numbers were observed for the T treatments. At the end of the experiment (day 21), bacteria
abundance decreased in all treatments and varied between
5.2 9 106 cells mL1 (C) and 1.1 9 107 cell mL1 (T).
The number of viruses accounted for 5.5 9 105
viruses mL1 on day 0 and increased continuously in all
treatments until day 21 (Fig. 1b). Lowest numbers were
observed in the control treatments (3.7 9 108
viruses mL1), while highest virus numbers were detected
in the T and G treatments (7.1 9 108 viruses mL1).
FEMS Microbiol Ecol 83 (2013) 202–213
205
Control mechanisms for bacterial diversity
treatments were highly similar. After day 12, the band
patterns changed in all treatments. Two gels, one before
(day 8) and one after the observed changes (day 18), were
digitalised for clarifying this development (Fig. 2). The
Cells mL–1
(a)
(a)
Viruses mL–1
(b)
Day
Fig. 1. Bacterial abundance in cells mL1 (a) and viral abundance in
viruses mL1 (b) determined by flow cytometry within the four
different treatments (T = Thalassiosira, P = Phaeocystis, G = glucose,
C = control).
(b)
In the P treatment, 5.6 9 108 viruses mL1 were detected
at the end of the experiment.
In our predator-free system, the upper bound for total
abundance of individuals in the prokaryote community
should be limited by the amount of organic material
available. In our experiment, we got the highest abundance of bacteria in the glucose treatment, followed by
the two treatments with substrate from different phytoplankton cultures. Assuming a similar yield for bacterial
growth in all treatments, this can be used to loosely estimate the production of organic material to 102 and
121 lM of glucose-equivalent material in the T and P
cultures, respectively, as compared to the 159 lM
glucose-C in treatment G.
DGGE
DGGE results indicate high stability within the bacterial
community composition until day 12, prior to which the
band pattern of the different sample days and different
FEMS Microbiol Ecol 83 (2013) 202–213
Fig. 2. DGGE profiles of day 8 (a) and day 18 (b). The marker on
each side of each gel is labelled with M. The three parallel samples
(1, 2 and 3) are presented for each treatment: Thalassiosira sp. (T),
Phaeocystis pouchetii (P), glucose (G) and the control (C).
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
206
B. Töpper et al.
DGGE profile from day 8 presents eight bands at the
same positions for the different treatments and parallels
with one extra band in the glucose treatments (Fig. 2a).
The bacterial community on day 18 showed a different
DGGE profile with 4–14 bands at 20 different positions
(Fig. 2b).
Bacterial community structure
A total of 96 clones for each treatment were sequenced,
with each 48 clones from day 8 and day 18. Thereof, 81
(T), 86 (P), 88 (G) and 78 (C) sequences, respectively,
were used for phylogenetic analysis. Highest sequence
similarity was found to bacteria belonging to the classes
Bacteroidetes, Actinobacteria, Alpha-, Gamma-, Epsilonand Deltaproteobacteria (Fig. 3 and Supporting information, Fig. S2).
Most sequences for each treatment were assigned to
bacteria belonging to the Gammaproteobacteria class. This
result was most pronounced in the glucose treatment
where 97.7% of the sequences assigned to bacteria
belonging to this class (Fig. 3), with dominance on both
day 8 and day 18. Out of these sequences, 93.0% from
day 8 belonged to the Colwelliaceae family (Figs S1 and
S2c), while 55.6% of the sequences from day 18 were
assigned to bacteria belonging to the Vibrionaceae cluster
(Fig. S1 and S2c). Further, only 2.2% of the sequences
from the glucose treatment were assigned to bacteria
belonging to other bacterial classes as Epsilonproteobacteria and the Bacteroidetes.
100
Actinobacteria
Bacteriodetes
Deltaproteobacteria
Epsilonproteobacteria
Alphaproteobacteria
Gammaproteobacteria
80
Sequences (%)
In the phytoplankton and control treatments, bacteria
belonging to Gammaproteobacteria accounted for 61.7%
(T), 61.6% (P) and 52.6% (C), respectively, of total
amount of sequences (Fig. 3). Further, bacteria belonging
to the Alphaproteobacteria [19.8% (T), 16.3% (P) and
14.1% (C)] and to the Bacteroidetes [7.4% (T), 17.4% (P)
and 24.4% (C)] were the next abundant groups within
these treatments. In addition, bacteria belonging to the
Epsilonproteobacteria [4.9% (T), 2.3% (P), 9.0% (C)],
Deltaproteobacteria [only in T (6.2%)] and Actinobacteria
[only in P (2.3%)] were also detected.
The abundance of bacteria belonging to the Gammaproteobacteria did not change from day 8 to day 18 in
treatment P and G, but a decrease in the numbers of bacteria belonging to this class in the same time period was
detected for treatment T (from 87.0% to 28.9%) and C
(from 63.4% to 40.5%) (Fig. 3). In these two treatments,
bacteria belonging to the Alphaproteobacteria [37.1% (T)]
and Bacteroidetes [43.2% (C)] respectively dominated the
bacterial community in the end of the experiment. For
the two phytoplankton treatments and the control treatment, more than 50% of the sequences were assigned to
bacteria belonging to Colwelliaceae or other families
belonging to the Alteromonadales order (Figs S1 and S2a,
b,d). All sequences assigned to bacteria belonging to the
Alphaproteobacteria belonged to the family of Rhodobacteraceae.
Sequences were checked for chimeras with the program
B2C2 (Gontcharova et al., 2010). Sixty sequences of 333
were identified as chimeras (18%). Performing a second
60
40
20
0
Tt T8 T18
Pt P8 P18
Gt G8G18
Treatment
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
Ct C8 C18
Fig. 3. Taxonomic composition; percentage of
Actinobacteria, Bacteroidetes, Gamma-, Alpha-,
Epsilon- and Deltaproteobacteria in the
treatments Thalassiosira (T), Phaeocystis (P),
Glucose (G) and Control (C). The ratio is
shown for the total amount of sequences (t),
and day 8 sequences (8) and day 18
sequences (18) separately.
FEMS Microbiol Ecol 83 (2013) 202–213
207
Control mechanisms for bacterial diversity
blast search of these chimeric sequences resulted in high
query coverage (98–100%) to sequences in the GenBank
database. We therefore decided to include these sequences
in our analysis. An explanation for the classification of
these sequences as chimeras by B2C2 might be that the
database used by the program consists of bacterial
sequences with full taxonomic information. As know,
there are more unknown bacteria in the marine
environment with no or rare taxonomic information
(Pedros-Alio, 2006). Owing to this, sequences that are
most similar to uncultured bacteria might have been classified as chimera and, however, included in the analysis.
The Shannon index, and thus, diversity of the total
amount of sequences, was significantly lower in the glucose treatment (2.94) than in the phytoplankton treatments [3.33 (T), 3.60 (P)] and the control (3.52)
(Tables 1 and 2). While diversity in treatment T, P and C
was higher on day 18 [3.03 (T), 3.04 (P), 2.97 C)] than
on day 8 [2.45 (T), 2.82 (P), 2.80 (C)], the glucose treatment showed lower diversity on day 18 (2.39) than on
day 8 (2.42). Assessed differences in Shannon indices
showed significant results of the glucose treatment on day
18 but not on day 8 (Table 2). Total evenness in the different treatments varied between 0.82 (treatment G) and
0.92 (treatment C). Evenness decreased from day 8 to day
18 in treatment P [0.90 (day 8), 0.88 (day18)] and G
[0.82 (day 8), 0.77 (day 18)], but increased in treatment
T [0.81 (day 8), 0.94 (day 18)] and C [0.88 (day 8), 0.94
(day 18)]. The Chao1 index (richness) was highest in
treatment P (189) and lowest in treatment C (126). In
general, richness was higher on day 18 [130 (T), 467 (P),
99 (G), 62 (C)] than on day 8 [61 (T), 57 (P), 21 (G),
55 (C)] in all treatments. Most pronounced differences in
richness were detected on day 18 with a Chao1 index of
treatment P being more than four times higher than
of treatment G and more than three times higher than of
treatment T.
Discussion
In our grazing free experiment, we showed that substrate
availability (bottom-up) affected the composition of the
bacterial community. Bacterial richness, measured as different OTUs, increased with increasing carbon complexity, while evenness remained similar in the different
setups. These results indicate that both substrate composition (bottom-up) and viral lysis (top-down) operate
together in the control of bacterial diversity.
Bacterial community composition – substrate
availability
Coastal bacterial communities have been shown to consist
of generalists able to metabolize a wide variety of organic
carbon compounds (Mou et al., 2008), suggesting that
the coastal prokaryotic community is less controlled by
changes in the carbon reservoir than by other factors like
trophic interactions and physical conditions. This is in
contrast to our study where different carbon reservoirs
resulted in major differences in the bacterial community
compositions. However, our results also showed similarities in the bacterial community despite different carbon
Table 1. Calculated values describing diversity in the four different treatments: Shannon index (H′), variation of the Shannon index (VarH′),
evenness (J′) and Chao1 index. Calculations were carried out for both, the total amount of sequences and day 8 and 18 separately. N represents
the amount of sequences used for the respective calculations. Amount of cluster was determined using an identity threshold of 99% within the
Complete Linkage Clustering analysis that was the basis for the calculations
Treatment
Thalassiosira sp.
Total
Day 8
Day 18
Phaeocystis puchetii
Total
Day 8
Day 18
Glucose
Total
Day 8
Day 18
Control
Total
Day 8
Day 18
N
Cluster
H′
varH′
J′
Chao1
81
46
35
43
21
25
3.33
2.45
3.03
0.017
0.034
0.023
0.89
0.81
0.94
142
61
130
86
39
47
52
23
32
3.60
2.82
3.04
0.014
0.027
0.033
0.91
0.90
0.88
189
57
467
88
43
45
37
19
22
2.94
2.42
2.39
0.020
0.034
0.046
0.82
0.82
0.77
132
21
99
78
41
37
46
24
24
3.52
2.80
2.97
0.013
0.031
0.021
0.92
0.88
0.94
126
55
62
FEMS Microbiol Ecol 83 (2013) 202–213
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
208
B. Töpper et al.
Table 2. Results of the t-test, assessing differences between
Shannon indices of the different treatments (T = Thalassiosira,
P = Phaeocystis, G = Glucose, C = Control). The Shannon indices of
Table 1 were used for calculations. Presented are t-values and
degrees of freedom for the different treatment comparisons
Total
P
T
C
Day 8
P
T
C
Day 18
P
T
C
G
C
T
3.58/169**
2.03/169*
3.19/154**
0.49/164
1.1/157
1.53/164
1.62/82
0.12/89
1.49/84
0.08/80
1.37/87
1.50/85
2.31/89*
2.44/77*
2.24/67*
0.30/83
0.29/72
0.04/82
**P 0.01, *P 0.05.
sources, with for example, OTUs similar to Colwellia rossensis (Fig. S2) present in all treatments. We may therefore
suggest that the bacterial community in our experiment
consisted of a mixture of specialists, specified on particular components, but also generalists that are able to
degrade a variety of organic carbon compounds.
The composition of the bacterial community in the
simple carbon (glucose) treatment differed from the bacterial community in the phytoplankton treatments that
are assumed to comprise a more diverse pool of carbon
compounds (Figs 2 and 3, Fig. S2). In both the glucose
and the Phaeocystis treatment, bacteria belonging to the
Gammaproteobacteria dominated throughout the whole
experiment, while Alphaproteobacteria (37.1% in treatment T) and Bacteroidetes (43.2% in treatment C) dominated in the end of the experiment in the Thalassiosira
treatment and in the control (Fig. 3 and Fig. S2). In
accordance with our study, low-molecular-weight carbon
compounds like glucose have been reported to stimulate
growth of bacteria belonging to the Gammaproteobacteria
group, especially when applied in high concentrations
(Øvreås et al., 2003; Pinhassi & Berman, 2003; Harvey
et al., 2006). In addition, the dominance of the families
Colwelliaceae and Vibrionaceae as seen in our study in the
glucose treatment has also been reported in other studies
(Eilers et al., 2000a, b), where these two families
responded rapidly to culturing conditions.
The family Colwelliaceae belongs to the order Alteromonadales, which are typically rare in situ (Eilers et al.,
2000a, b). In our study, a large fraction of Alteromonadales was detected in all four treatments. Reports have
shown that bacteria belonging to this order seem to
react to experimental manipulation (bottle incubation)
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
rather than to nutrient manipulation (Eilers et al.,
2000a, b; Beardsley et al., 2003). The water treatment
during the experiment, including filtration and stirring,
might cause flocculation of colloidal material because of
turbulence (Kepkay, 1994). This might have benefited
bacteria belonging to the Alteromonadales, as they can
thrive on solid surfaces and particles (Acinas et al.,
1999; Dang & Lovell, 2000). On the other hand, the
persistence of Alteromonadales seems to be influenced by
the presence of glucose as a carbon source, as the abundance of bacteria belonging to this order was highest in
this treatment (Fig. S2). This has also been reported
from other experiments with glucose amendments
(Allers et al., 2007).
An explanation for the dominance of the families Vibrionaceae and Alteromonaceae might be linked to similarities in specific bacterial growth rates. In situations of
carbon starvation, it has been demonstrated that bacteria
belonging to Vibrionaceae maintain ribosomes in large
excess that exceed the demand of protein synthesis
(Flärdh et al., 1992). Similar observations have been
reported for bacteria belonging to the Alteromonadales
(Eilers et al., 2000a, b; Pernthaler et al., 2001). As a
result, these bacteria might respond with rapid cell
growth to substrate addition (Nyström et al., 1990) and,
thus, have an advantage in the nutrient-rich treatments.
Another explanation for the dominance of the two families Vibrionaceae and Alteromonaceae might be explained
by the absence of grazers. Bacteria belonging to both
these families have shown to be particularly sensitive to
heterotrophic nanoflagellate grazing (Beardsley et al.,
2003). Both, the grazing and the growth rate effects may
contribute to the observed dominance of these two families under our grazing free conditions despite their low
abundance in situ.
Alphaproteobacteria have previously been shown to
increase in the presence of algal blooms (Riemann et al.,
2000; Pinhassi et al., 2004). Our results show an increase
in Alphaproteobacteria in the Thalassiosira treatment but
not in the Phaeocystis treatment (Fig. 3), supporting the
assumption that phytoplankton plays an important role
in shaping bacterial communities. The detected Alphaproteobacteria belong all to the family Rhodobacteraceae
(Fig. S2a), which are typically associated with phytoplankton blooms (Pinhassi et al., 2004). Some species in
this family have also demonstrated the availability to
utilise phytoplankton-derived DOC (Zubkov et al.,
2001).
Bacterial community dynamics
In the glucose treatment, the dominance of Colwelliaceae
on day 8 and Vibrionaceae on day 18 indicates a shift in
FEMS Microbiol Ecol 83 (2013) 202–213
209
Control mechanisms for bacterial diversity
the bacterial community, which was also detected by
DGGE (Fig. 2).
One explanation for these shifts may be linked to viral
lysis or differences in host sensitivity to phage infection.
A variety of both lytic and lysogenic bacteriophages
belonging to the same viral families have been reported
to infect hosts belonging both to the Alteromonadales and
Vibrionaceae family (Chow & Rouf, 1983; Wichels et al.,
1998, 2002; Petrov et al., 2006; Gibb & Edgell, 2007).
Further, bacteriophage resistance mechanisms are easily
developed in infected hosts (Labrie et al., 2010). Thus,
there is no obvious indication that these two host families
should differ dramatically from each other with respect to
interactions with their specific viruses and/or sensitivity
for phage infections.
An additional effect contributing to the shift from
Alteromonadales to Vibrionaceae might be linked to the
competition success of Vibrionaceae with respect to mineral nutrient uptake. The consumption of glucose in
excess to increase cell size (production of carbon-rich
inclusion bodies) without a corresponding increase in
nitrogen and phosphorous requirements has been
reported for Vibrio splendidus (Thingstad et al., 2005), a
strain which was also detected in the glucose treatment in
the present study (Fig. S2c). This characteristic trait may
allow the V. splendidus cells to be better competitors for
mineral nutrients, which might enabled them to out-compete the dominating Alteromonadales in the first half of
the experiment. Vibrio splendidus was also detected in the
Thalassiosira treatment (Fig. S2a), but at lower abundance
than in the glucose treatment.
From the DGGE result, we could observe a very stable
bacterial community structure for the first 12 days of the
experiment (Fig. 2a). This result seems to be partly contradictory to the sequencing results, which showed differences within the bacterial community on day 8 (Fig. 3).
The discrepancy in these results might be due to different
level of resolution captured by the two methods. The
amount of DGGE bands does not necessarily reflect the
amount of different species and, thus, richness. Owing to
heterogeneity of 16S rRNA gene (Fogel et al., 1999), one
bacterial species might produce more than one band on
the gel (Dahllöf et al., 2000) and bands might migrate
together to the same gel position (Muyzer et al., 1998).
Although, DGGE seems to have its limitations in significance, more bands on day 18 (Fig. 2b) and, hence, a
possibly higher richness is conform to higher richness
on day 18 compared to day 8 within the analysed
sequences (Table 1). Although cloning might have a
higher resolution than DGGE, our results indicate that
the bacterial community still was under-sampled. The calculated ratio of unique sequences (presented as clusters
in Table 1) to the estimated richness (Chao1, Table 1)
FEMS Microbiol Ecol 83 (2013) 202–213
averaged 0.31 for the total amount of sequences, implying
that approximately 70% of the phylotypes remained
undetected. This may be due to the fact that cloning of
PCR amplicons mainly targets the abundant phylotypes
within the community (Acinas et al., 2005).
Bacterial diversity – viral lysis
All treatments had a virus-to-bacteria ratio of 21 on day
0 and 35 on day 18, which is comparable with what is
found under in situ conditions (De Corte et al., 2012).
This indicates that the viral–host interaction in our experiment was comparable with what is described in the marine environment for lytic viruses and their hosts.
Although bacterial diversity of the two phytoplankton
treatments and the control treatment was not significantly
different from each other (Table 2), a higher t-value for
the Phaeocystis–Thalassiosira comparison on day 8 (1.50)
than on day 18 (0.04) showed that the bacterial diversity
in the two phytoplankton treatments became more similar by the end of the experiment. However, Chao1 indices
indicate an increasing divergence in richness towards the
end of the experiment (Table 1). The observed difference
in development of bacterial community structure and
diversity in the two phytoplankton treatments are consistent with results of earlier studies that discussed the
strong impact of phytoplankton communities on bacterial
communities (Pinhassi et al., 2004). Our results showed
that the changes within bacterial diversity and community
composition from day 8 to day 18 changed more within
the Thalassiosira treatment than within the Phaeocystis
treatment, but the total bacterial diversity and richness
were higher in the Phaeocystis treatment than in the
Thalassiosira treatment (Table 1).
Significantly lower bacterial diversity and lowest
evenness in the glucose treatment on day 18 as well as
lower richness in the glucose treatment than in the two
phytoplankton treatments (Table 1) indicate that organic
matter supply might also play a role in defining bacterial
richness and evenness, contrary to our hypothesis. Further, higher viral abundance (Fig. 1b) and decreasing
bacterial abundance (Fig. 1a) at the end of the experiment might indicate a larger impact of virus on the bacterial community in the second part of the experiment.
Differences in evenness and richness are more pronounced at the end of the experiment (day 18), which
strengthen the idea that organic matter is more important
in determining richness and evenness than in our simplifying hypothesis.
The model proposed by Thingstad (2000) conceptually
operates with ‘host groups’, not with ‘species’ or ‘strains’.
Viruses are suggested to control abundance within each
such host group. Richness of host groups is thus the ratio
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
210
between total population size, here determined by the
amount of substrate available, and the average size of the
host groups, determined by the viruses. Thus, models of
this type suggest that richness and evenness are controlled
mainly by lytic viruses. The detected minor differences in
evenness in our samples are thus in accordance with our
hypothesis, while, contrary to the hypothesis, richness
developed differently in the four treatments (Table 1).
Our hypothesis of a clear separation between top-down
control of richness and evenness, and bottom-up control
of community composition, is clearly based on a ‘simplest
possible’ conceptual model. The distinctions between topdown and bottom-up mechanisms become less clear if
different host species have widely different interaction
patterns with their respective viruses. This would create a
feedback where a change in substrates creates a change in
community composition that again leads to potential differences in the size distribution of host groups. With such
a change in the size distribution of host groups, both
richness and evenness could change in response to a
change in substrate composition.
Small variations in the genome of a host may alter the
infection success of a specific virus (Avrani et al., 2011).
Strains from the same species seem to share a core part of
the genome, while there are variable genome regions coding for between-strain differences in their viral resistance
(Rodriguez-Valera et al., 2009). Although this will create
new host groups, the change will probably not be detectable
in the 16S rRNA gene. To increase the sensitivity of our
method, we used 99% similarity for the determination of
OTUs (host groups) rather than 97% normally used for
definition of different species (Kunin et al., 2010). We
decided on 99% similarity for the determination of the
OTUs and not 100% to take into account possible sequencing and PCR biases as discussed by Acinas et al. (2005).
The problems arising in properly comparing data at
species level to a model describing diversity at strain level
can be illustrated by considering two extreme examples: it
would theoretically be possible (1) to fill all the hostgroup niches with different species (one strain present for
each species) or to (2) fill all the niches with different
strains from one species. While richness and evenness at
host-group level could theoretically be identical in the two
cases, species diversity would obviously be very different.
Unfortunately, this weakness in our present ability to
observationally resolve the host community at strain level
seems to create a situation where critical testing of models
for virus-controlled diversity appears rather difficult.
Bottle experiments, as applied in our study, are in general presenting living conditions that are far away from
conditions that occur in natural systems. Changing conditions for the bacterial community because of incubation
might stimulate certain bacteria assemblages more than
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
B. Töpper et al.
others, as it was eventually the case for the dominating
Gammaproteobacteria in our study. By this, some changes
within bacterial community compositions occur possibly
due to rather the bottle incubation itself than certain
treatments. However, bottle experiments are commonly
used to investigate processes in microbial communities
(e.g. Pinhassi & Berman, 2003; Alonso & Pernthaler,
2005; Sapp et al., 2007), and the different development of
certain factors in different treatments provide results that
can contribute to more knowledge about basic principles
in microbial communities.
The Ocean is a complex environmental system with
great fluctuations in biological, physical and chemical factors that will have a direct and synergetic effect on the
prokaryotic community. In this study, we intentionally
made a strong simplification on how the prokaryotic
community is regulated compared with in situ conditions.
One of these simplifications was made by removing the
top-down control mechanisms forced by heterotrophic
grazers. Studies investigating the synergetic effect of the
two top-down control mechanisms (viral lysis and predation by grazers) have shown that grazing may have a
stimulating effect on the prokaryotic growth in addition
to viral proliferation (Simek et al., 2001; Sime-Ngando &
Ram, 2005; Weinbauer et al., 2007) and a reduction in
the bacterial diversity (Weinbauer et al., 2007). Contradictory, a reduction in bacterial production in the presence of both predator groups, accompanied by sustained
prokaryote species richness, has also been reported
(Zhang et al., 2007; Bonilla-Findji et al., 2009).
Conclusion
Our results demonstrate that different carbon sources
resulted in a difference in the community composition,
confirming our hypothesis that present organic supply
and the host competitive abilities will influence which
organisms that fills the different niches. Nevertheless, the
observation of high diversity in the treatment with only a
simple carbon source (glucose) indicates that substrate
diversity does not alone control bacterial diversity. Further, we hypothesised that diversity, number of host
groups (richness) and the number of bacteria within the
different host groups (evenness) are mainly controlled by
viruses. In our experiment, similar evenness indices were
found for the different treatments, supporting our hypothesis, however, the bacterial diversity was higher in the
phytoplankton treatments, suggesting that substrate complexity also had a role in determining both richness and
evenness in these samples. The complexity of microbial
food webs makes the separation of factors determining
bacterial diversity difficult, and in our study, substrate
composition and viral lysis are suggested to act in concert.
FEMS Microbiol Ecol 83 (2013) 202–213
Control mechanisms for bacterial diversity
Acknowledgement
We thank H.M.K. Stabell for her great support in the laboratory, J.L. Ray for helping with the sequence analysis
and helpful comments and J.P. Spindelböck for his great
statistical support. This work was financed by the University of Bergen, Research Council of Norway through the
International Polar Year project 175939/S30 ‘PAME-Nor’
(IPY activity ID no. 71) and EU FP7 through ERC project MINOS ref.nr 250254.
References
Acinas SG, Anton J & Rodriguez-Valera F (1999) Diversity of
free-living and attached bacteria in offshore western
Mediterranean waters as depicted by analysis of genes
encoding 16S rRNA. Appl Environ Microbiol 65: 514–522.
Acinas SG, Sarma-Rupavtarm R, Klepac-Ceraj V & Polz MF
(2005) PCR-induced sequence artifacts and bias: insights
from comparison of two 16S rRNA clone libraries
constructed from the same sample. Appl Environ Microbiol
71: 8966–8969.
Allers E, Gomez-Consarnau L, Pinhassi J, Gasol JM, Simek
K & Pernthaler J (2007) Response of Alteromonadaceae
and Rhodobacteraceae to glucose and phosphorus
manipulation in marine mesocosms. Environ Microbiol 9:
2417–2429.
Alonso C & Pernthaler J (2005) Incorporation of glucose
under anoxic conditions by bacterioplankton from coastal
North Sea surface waters. Appl Environ Microbiol 71:
1709–1716.
Alonso C & Pernthaler J (2006) Roseobacter and SAR11
dominate microbial glucose uptake in coastal North Sea
waters. Environ Microbiol 8: 2022–2030.
Alonso-Saez L & Gasol JM (2007) Seasonal variations in the
contributions of different bacterial groups to the uptake of
low-molecular-weight compounds in northwestern
Mediterranean coastal waters. Appl Environ Microbiol 73:
3528–3535.
Avrani S, Wurtzel O, Sharon I, Sorek R & Lindell D (2011)
Genomic island variability facilitates Prochlorococcus-virus
coexistence. Nature 474: 604–608.
Beardsley C, Pernthaler J, Wosniok W & Amann R (2003) Are
readily culturable bacteria in coastal North Sea waters
suppressed by selective grazing mortality? Appl Environ
Microbiol 69: 2624–2630.
Bonilla-Findji O, Herndl GJ, Gattuso J-P & Weinbauer MG
(2009) Viral and flagellate control of prokaryotic production
and community structure in offshore mediterranean waters.
Appl Environ Microbiol 75: 4801–4812.
Chow MS & Rouf MA (1983) Isolation and partial
characterization of two aeromonas hydrophila
bacteriophages. Appl Environ Microbiol 45: 1670–1676.
Connell JH (1979) Intermediate-disturbance hypothesis.
Science 204: 1345.
FEMS Microbiol Ecol 83 (2013) 202–213
211
De Corte D, Sintes E, Yokokawa T, Reinthaler T & Herndl GJ
(2012) Links between viruses and prokaryotes throughout
the water column along a North Atlantic latitudinal
transect. ISME J 6: 1566–1577.
Cottrell MT & Kirchman DL (2000) Natural assemblages of
marine proteobacteria and members of the CytophagaFlavobacter cluster consuming low- and high-molecularweight dissolved organic matter. Appl Environ Microbiol 66:
1692–1697.
Dahllöf I, Baillie H & Kjelleberg S (2000) rpoB-based
microbial community analysis avoids limitations inherent in
16S rRNA gene intraspecies heterogeneity. Appl Environ
Microbiol 66: 3376–3380.
Dang HY & Lovell CR (2000) Bacterial primary colonization
and early succession on surfaces in marine waters as
determined by amplified rRNA gene restriction analysis and
sequence analysis of 16S rRNA genes. Appl Environ
Microbiol 66: 467–475.
Eilers H, Pernthaler J & Amann R (2000a) Succession of
pelagic marine bacteria during enrichment: a close look at
cultivation-induced shifts. Appl Environ Microbiol 66:
4634–4640.
Eilers H, Pernthaler J, Glockner FO & Amann R (2000b)
Culturability and in situ abundance of pelagic bacteria from
the North Sea. Appl Environ Microbiol 66: 3044–3051.
Elifantz H, Malmstrom RR, Cottrell MT & Kirchman DL
(2005) Assimilation of polysaccharides and glucose by major
bacterial groups in the Delaware Estuary. Appl Environ
Microbiol 71: 7799–7805.
Flärdh K, Cohen PS & Kjelleberg S (1992) Ribosomes exist in
large excess over the apparent demand for protein-synthesis
during carbon starvation in marine Vibrio sp strain
CCUG15956. J Bacteriol 174: 6780–6788.
Fogel GB, Collins CR, Li J & Brunk CF (1999) Prokaryotic
genome size and SSU rDNA copy number: estimation of
microbial relative abundance from a mixed population.
Microb Ecol 38: 93–113.
Gibb EA & Edgell DR (2007) Multiple controls regulate the
expression of mobE, an HNH homing endonuclease gene
embedded within a ribonucleotide reductase gene of Phage
Aeh1. J Bacteriol 189: 4648–4661.
Gontcharova V, Youn E, Wolcott RD, Hollister EB, Gentry TJ
& Dowd SE (2010) Black box chimera check (B2C2): a
windows-based software for batch depletion of chimeras
from bacterial 16S rRNA gene datasets. Open Microbiol J 4:
47–52.
Harvey HR, Dyda RY & Kirchman DL (2006) Impact of DOM
composition on bacterial lipids and community structure in
estuaries. Aquat Microb Ecol 42: 105–117.
Kepkay PE (1994) familyNamePrefix aggregation and the
biological reactivity of colloids. Mar Ecol Prog Ser 109: 293–
304.
Kunin V, Engelbrektson A, Ochman H & Hugenholtz P (2010)
Wrinkles in the rare biosphere: pyrosequencing errors can
lead to artificial inflation of diversity estimates. Environ
Microbiol 12: 118–123.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
212
Labrie SJ, Samson JE & Moineau S (2010) Bacteriophage
resistance mechanisms. Nat Rev Microbiol 8: 317–327.
Lancelot C (1983) Factors affecting phytoplankton extracellular
release in the southern bight of the North-Sea. Mar Ecol
Prog Ser 12: 115–121.
Magurran A (1988) Ecological Diversity and Its Measurement.
Princeton University Press, Cambridge.
Marie D, Brussaard CPD, Thyrhaug R, Bratbak G & Vaulot D
(1999) Enumeration of marine viruses in culture and
natural samples by flow cytometry. Appl Environ Microbiol
65: 45–52.
Mou XZ, Sun SL, Edwards RA, Hodson RE & Moran MA
(2008) Bacterial carbon processing by generalist species in
the coastal ocean. Nature 451: 708–711.
Muyzer G, Brinkhoff T, Nübel U, Santegoeds C, Schäfer H &
Wawer C (1998) Denaturing gradient gel electrophoresis
(DGGE) in microbial ecology. Molecular Microbial Ecology
Manual 3.4.4, (Akkermans ADL, van Elsas JD & de Bruijn
FJ, eds), pp. 1–27. Kluwer Academic Publishers, Dordrecht,
The Netherlands.
Myklestad SM (1995) Release of extracellular products by
phytoplankton with special emphasis on polysaccharides.
Sci Total Environ 165: 155–164.
Nyström T, Albertson NH, Flardh K & Kjelleberg S (1990)
Physiological and molecular adaptation to starvation and
recovery from starvation by the marine Vibrio sp S14.
FEMS Microbiol Ecol 74: 129–140.
Øvreås L, Bourne D, Sandaa RA, Casamayo EO, Benlloch S,
Goddard V, Smerdon G, Heldal M & Thingstad TF (2003)
Response of bacterial and viral communities to nutrient
manipulations in seawater mesocosms. Aquat Microb Ecol
31: 109–121.
Pedros-Alio C (2006) Marine microbial diversity: can it be
determined? Trends Microbiol 14: 257–263.
Pernthaler A, Pernthaler J, Eilers H & Amann R (2001)
Growth patterns of two marine isolates: adaptations to
substrate patchiness? Appl Environ Microbiol 67: 4077–4083.
Petrov VM, Nolan JM, Bertrand C, Levy D, Desplats C, Krisch
HM & Karam JD (2006) Plasticity of the gene functions for
DNA Replication in the T4-like phages. J Mol Biol 361:
46–68.
Pinhassi J & Berman T (2003) Differential growth response of
colony-forming alpha- and gamma-proteobacteria in
dilution culture and nutrient addition experiments from
Lake Kinneret (Israel), the eastern Mediterranean Sea, and
the Gulf of Eilat. Appl Environ Microbiol 69: 199–211.
Pinhassi J, Sala MM, Havskum H, Peters F, Guadayol O,
Malits A & Marrase C (2004) Changes in bacterioplankton
composition under different phytoplankton regimens. Appl
Environ Microbiol 70: 6753–6766.
Riemann L, Steward GF & Azam F (2000) Dynamics of
bacterial community composition and activity during a
mesocosm diatom bloom. Appl Environ Microbiol 66:
578–587.
Rodriguez-Valera F, Martin-Cuadrado AB, Rodriguez-Brito B,
Pasic L, Thingstad TF, Rohwer F & Mira A (2009)
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
B. Töpper et al.
OPINION Explaining microbial population genomics
through phage predation. Nat Rev Microbiol 7: 828–836.
Sandaa RA, Gomez-Consarnau L, Pinhassi J, Riemann L,
Malits A, Weinbauer WG, Gasol JM & Thingstad TF (2009)
Viral control of bacterial biodiversity – evidence from a
nutrient-enriched marine mesocosm experiment. Environ
Microbiol 11: 2585–2597.
Sapp M, Schwaderer AS, Wiltshire KH, Hoppe HG, Gerdts G
& Wichels A (2007) Species-specific bacterial communities
in the phycosphere of microalgae? Microb Ecol 53: 683–699.
Simek K, Pernthaler J, Weinbauer MG, Hornak K, Dolan JR,
Nedoma J, Masin M & Amann R (2001) Changes in
bacterial community composition and dynamics and viral
mortality rates associated with enhanced flagellate grazing
in a mesoeutrophic reservoir. Appl Environ Microbiol 67:
2723–2733.
Simek K, Kasalicky V, Hornak K, Hahn MW & Weinbauer
MG (2010) Assessing niche separation among coexisting
Limnohabitans strains through interactions with a
competitor, viruses, and a bacterivore. Appl Environ
Microbiol 76: 1406–1416.
Sime-Ngando T & Ram ASP (2005) Grazer effects on
prokaryotes and viruses in a freshwater microcosm
experiment. Aquat Microb Ecol 41: 115–124.
Thingstad TF (2000) Elements of a theory for the mechanisms
controlling abundance, diversity, and biogeochemical role of
lytic bacterial viruses in aquatic systems. Limnol Oceanogr
45: 1320–1328.
Thingstad TF, Ovreas L, Egge JK, Lovdal T & Heldal M (2005)
Use of non-limiting substrates to increase size; a generic
strategy to simultaneously optimize uptake and minimize
predation in pelagic osmotrophs? Ecol Lett 8: 675–682.
Thingstad TF, Bellerby RG, Bratbak G et al. (2008)
Counterintuitive carbon-to-nutrient coupling in an Arctic
pelagic ecosystem. Nature 455: 387–390.
Tilman D (1977) Resource competition between
phytoplanktonic algae – experimental and theoretical
approach. Ecology 58: 338–348.
Töpper B, Larsen A, Thingstad TF, Thyrhaug R & Sandaa R-A
(2010) Bacterial community composition in an Arctic
phytoplankton mesocosm bloom: the impact of silicate and
glucose. Polar Biol 33: 1557–1565.
Venail PA, Maclean RC, Meynard CN & Mouquet N (2010)
Dispersal scales up the biodiversity-productivity relationship
in an experimental source-sink metacommunity. Proc Biol
Sci 277: 2339–2345.
Weinbauer MG, Hornak K, Jezbera J, Nedoma J, Dolan JR &
Simek K (2007) Synergistic and antagonistic effects of viral
lysis and protistan grazing on bacterial biomass, production
and diversity. Environ Microbiol 9: 777–788.
Wichels A, Biel SS, Gelderblom HR, Brinkhoff T, Muyzer G &
Schütt C (1998) Bacteriophage diversity in the North Sea.
Appl Environ Microbiol 64: 4128–4133.
Wichels A, Gerdts G & Schütt C (2002) Pseudoalteromonas
spp. phages, a significant group of marine bacteriophages in
the North Sea. Aquat Microb Ecol 27: 233–239.
FEMS Microbiol Ecol 83 (2013) 202–213
213
Control mechanisms for bacterial diversity
Winter C, Bouvier T, Weinbauer MG & Thingstad TF (2010)
Trade-offs between competition and defense specialists
among unicellular planktonic organisms: the “killing the
winner” hypothesis revisited. Microbiol Mol Biol Rev 74:
42–57.
Wolter K (1982) Bacterial incorporation of organic-substances
released by natural phytoplankton populations. Mar Ecol
Prog Ser 7: 287–295.
Zhang Z, Schwartz S, Wagner L & Miller W (2000) A greedy
algorithm for aligning DNA sequences. J Comput Biol 7:
203–214.
Zhang R, Weinbauer MG & Qian PY (2007) Viruses and
flagellates sustain apparent richness and reduce biomass
accumulation of bacterioplankton in coastal marine waters.
Environ Microbiol 9: 3008–3018.
Zubkov MV, Fuchs BM, Archer SD, Kiene RP, Amann R &
Burkill PH (2001) Linking the composition of
bacterioplankton to rapid turnover of dissolved
dimethylsulphoniopropionate in an algal bloom in the
North Sea. Environ Microbiol 3: 304–311.
FEMS Microbiol Ecol 83 (2013) 202–213
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
Fig. S1. Taxonomical trees for the four treatments show
the most abundant bacteria groups that could be assigned
to the sequences based on blast results.
Fig. S2. Phylogenetic tree of the 16S rRNA gene
sequences achieved in our study from the different treatments: Thalassiosira (a), Phaeocystis (b), Glucose (c), and
Control (d).
Please note: Wiley-Blackwell is not responsible for the
content or functionality of any supporting materials supplied by the authors. Any queries (other than missing
material) should be directed to the corresponding author
for the article.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved