The distribution of bacterial doubling times in the wild

Abstract

Generation time varies widely across organisms and is an important factor in the life cycle, life history and evolution of organisms. Although the doubling time (DT) has been estimated for many bacteria in the laboratory, it is nearly impossible to directly measure it in the natural environment. However, an estimate can be obtained by measuring the rate at which bacteria accumulate mutations per year in the wild and the rate at which they mutate per generation in the laboratory. If we assume the mutation rate per generation is the same in the wild and in the laboratory, and that all mutations in the wild are neutral, an assumption that we show is not very important, then an estimate of the DT can be obtained by dividing the latter by the former. We estimate the DT for five species of bacteria for which we have both an accumulation and a mutation rate estimate. We also infer the distribution of DTs across all bacteria from the distribution of the accumulation and mutation rates. Both analyses suggest that DTs for bacteria in the wild are substantially greater than those in the laboratory, that they vary by orders of magnitude between different species of bacteria and that a substantial fraction of bacteria double very slowly in the wild.

Full text

rspb.royalsocietypublishing.org
Research
Cite this article: Gibson B, Wilson DJ, Feil E,
Eyre-Walker A. 2018 The distribution of
bacterial doubling times in the wild.
Proc. R. Soc. B 285: 20180789.
http://dx.doi.org/10.1098/rspb.2018.0789
Received: 9 April 2018
Accepted: 18 May 2018
Subject Category:
Genetics and genomics
Subject Areas:
microbiology, genetics, evolution
Keywords:
generation time, mutation rates, bacteria
Author for correspondence:
Adam Eyre-Walker
e-mail: a.c.eyre-walker@sussex.ac.uk
Electronic supplementary material is available
online at https://dx.doi.org/10.6084/m9.
figshare.c.4114037.
The distribution of bacterial doubling
times in the wild
Beth Gibson1, Daniel J. Wilson2, Edward Feil3and Adam Eyre-Walker1
1
School of Life Sciences, University of Sussex, Brighton BN1 9QG, UK
2
Nuffield Department of Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK
3
The Milner Centre for Evolution, Department of Biology and Biochemistry, University of Bath, Claverton Down,
Bath, BA2 7AY, UK
BG, 0000-0001-7984-6188; DJW, 0000-0002-0940-3311; EF, 0000-0003-1446-6744;
AE-W, 0000-0001-5527-8729
Generation time varies widely across organisms and is an important factor
in the life cycle, life history and evolution of organisms. Although the dou-
bling time (DT) has been estimated for many bacteria in the laboratory, it is
nearly impossible to directly measure it in the natural environment. How-
ever, an estimate can be obtained by measuring the rate at which bacteria
accumulate mutations per year in the wild and the rate at which they
mutate per generation in the laboratory. If we assume the mutation rate
per generation is the same in the wild and in the laboratory, and that all
mutations in the wild are neutral, an assumption that we show is not very
important, then an estimate of the DT can be obtained by dividing the
latter by the former. We estimate the DT for five species of bacteria for
which we have both an accumulation and a mutation rate estimate. We
also infer the distribution of DTs across all bacteria from the distribution
of the accumulation and mutation rates. Both analyses suggest that DTs
for bacteria in the wild are substantially greater than those in the laboratory,
that they vary by orders of magnitude between different species of bacteria
and that a substantial fraction of bacteria double very slowly in the wild.
1. Introduction
The bacterium Escherichia coli can divide every 20 min in the laboratory under
aerobic, nutrient-rich conditions. But how often does it divide in its natural
environment in the gut, under anaerobic conditions where it probably spends
much of its time close to starvation? And what do we make of a bacterium,
such as Syntrophobacter fumaroxidans, which only doubles in the laboratory
every 140 h [1]. Does this reflect a slow doubling time (DT) in the wild, or
our inability to provide the conditions under which it can replicate rapidly?
Estimating the generation time is difficult for most bacteria in their natural
environment and very few estimates are available. The DT for intestinal bacteria
has been estimated in several mammals by assaying the quantity of bacteria in
the gut and faeces. Assuming no cell death Gibbons & Kapsimalis [2] estimate
the DT for all bacteria in the gut to be 48, 17 and 5.8 h in hamster, guinea pig
and mouse, respectively. More recently Yang et al. [3] have shown that the DT of
Pseudomonas aeruginosa is correlated to cellular ribosomal content in vitro and
have used this to estimate the DT in vivo in a cystic fibrosis (CF) patient to be
between 1.9 and 2.4 h.
Although there are very few estimates of the generation time in bacteria, this
quantity is important for understanding bacterial population dynamics. Here
we use an indirect method to estimate the DT that uses two sources of infor-
mation. First, we can measure the rate at which a bacterial species
accumulates mutations in its natural environment through time using tempor-
arily sampled data [4], or concurrent samples from a population with a known
date of origin. We refer to this quantity as the accumulation rate, to differentiate
&2018 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution
License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original
author and source are credited.
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

it from the mutation rate, and the substitution rate, the rate at
which mutations spread through a species to fixation. If we
assume that all mutations in the wild are neutral, an assump-
tion that we show to be relatively unimportant, then the
accumulation rate is an estimate of the mutation rate per
year, u
y
. Second, we can estimate the rate of mutation per
generation, u
g
, in the laboratory using a mutation accumu-
lation experiment and whole-genome sequencing, or
through fluctuation tests. If we assume that the mutation
rate per generation is the same in the wild and in the labora-
tory, an assumption we discuss further below, then if we
divide the accumulation rate per year in the wild by the
mutation rate per generation in the laboratory, we can esti-
mate the number of generations that the bacterium goes
through in the wild and hence the DT (DT ¼8760 u
g
/u
y
,
where 8760 is the number of hours per year).
2. Results
The accumulation rate in the wild and the mutation rate in
the laboratory have been estimated for 34 and 26 bacterial
species, respectively (electronic supplementary material,
tables S1, S2); we only consider mutation rate estimates
from mutation accumulation experiments, because estimates
from fluctuation tests are subject to substantial sampling
error and unknown bias, and we exclude estimates from
hypermutable strains. For five species, E. coli, P. aeruginosa,
Salmonella enterica, Staphylococcus aureus and Vibrio cholerae,
we have both an accumulation and a mutation rate estimate
and hence can estimate the DT. Among these five species
we find our DT estimates vary from 1.1 h in V. cholerae to
25 h in S. enterica (table 1). In all cases the estimated DT in
the wild is greater than that of the bacterium in the labora-
tory. For example, E. coli can double every 20 min in the
laboratory but we estimate that it only doubles every 15 h
in the wild.
In theory, it might be possible to estimate the DT in those
bacteria for which we have either an accumulation or
mutation rate estimate, but not both, by finding factors that
correlate with either rate and using those factors to predict
the rates. Unfortunately, we have been unable to find any
factor that correlates sufficiently well to be usefully predic-
tive. It has been suggested that the mutation rate is
correlated to genome size in microbes [30] but, the current
evidence for this correlation is very weak, and depends
upon the estimate from Mesoplasma florum (r¼20.68, p,
0.001 with M. florum and r¼20.39, p¼0.053 without
M. florum) (electronic supplementary material, figure S1)
[31]. However, we can use the accumulation and mutation
rate estimates to estimate the distribution of DTs across bac-
teria if we assume that there is no phylogenetic non-
independence in the mutation and accumulation data, an
assumption we address below. We can estimate the distri-
bution of DTs by fitting distributions to the accumulation
and mutation rate data, using maximum likelihood, and
then dividing one distribution by the other. We assume
that both variables are lognormally distributed, an assump-
tion that is supported by Q–Qplots with the exception of
the mutation rate per generation in M. florum, which is a
clear outlier (figure 1). We repeated all our analyses with
and without M. florum.
If the accumulation and mutation rate data are lognor-
mally distributed then the distribution of DT is also
lognormally distributed with a mean of log
e
(8760) þm
g
2
m
y
and a variance of v
g
þv
y
22Cov(g,y), where 8760 is the
number of hours per year and m
g
,m
y
,v
g
and v
y
are the
mean and variance of the lognormal distributions fitted to
the mutation (subscript g) and accumulation (subscript y)
rates. Cov(g,y) is the covariance between the accumulation
and mutation rates. We might expect that species with
higher mutation rates also have higher accumulation rates,
because the accumulation rate is expected to depend on the
mutation rate, but the correlation between the two will
depend upon how variable the DT and other factors, such
as the strength of selection, are between bacteria. The
observed correlation between the log accumulation rate and
log mutation rate is 0.077, but there are only five data
points, so the 95% confidence intervals on this estimate
encompass almost all possible values (20.86 to 0.89). We
explore different levels of the correlation between the
accumulation and mutation rates; it should be noted that
Cov(g,y) can be expressed as Sqrt(v
g
v
y
)Corr(g,y) where
Corr(g,y) is the correlation between the two variables.
The distribution of DTs in the wild inferred using our
method is shown in figure 2. We infer the median DT to be
7.04 h, but there is considerable spread around this even
when the accumulation and mutation rates are strongly corre-
lated (figure 2a); as the correlation increases so the variance in
DTs decreases, but the median remains unaffected. The
Table 1. Doubling time estimates (hours) for those species for which we have both an estimate of the accumulation and mutation rate. Accumulation rate (AR)
references—(1) [5]; (2) [6,7]; (3) [8–12]; (4) [13 – 23]; (5) [8,24]. Mutation rate (MR) references—(6) [25]; (7) [26]; (8) [27]; (9) [28]; (10) [29].
species
accumulation rate
per site per year
mutation rate per
site per generation
DT (h)
(s.e.)
laboratory
DT (h) ratio
AR
ref.
MR
ref.
Escherichia coli 1.44 10
27
2.54 10
210
15 (7.7) 0.33 45 1 6
Pseudomonas
aeruginosa
3.03 10
27
7.92 10
211
2.3 (0.77) 0.5 4.6 2 7
Salmonella
enterica
2.50 10
27
7.00 10
210
25 (7.9) 0.5 50 3 8
Staphylococcus
aureus
2.05 10
26
4.38 10
210
1.87 (0.98) 0.4 4.7 4 9
Vibrio cholerae 8.30 10
27
1.07 10
210
1.1 (0.26) 0.66 1.7 5 10
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
2
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

analysis suggests that most bacteria have DTs of between 1
and 100 h but there are substantial numbers with DTs
beyond these limits. For example, even if we assume that
the correlation between the accumulation and mutation rate
is 0.5 we infer that 10% of bacteria have a DT of faster than
1 h in the wild and 4.2% have a DT slower than 100 h in
the wild. If we remove the M. florum mutation rate estimate
from the analysis the median doubling is slightly lower at
6.16 h, but there is almost as much variation as when this bac-
terium is included; at a correlation is 0.5 we infer that 12% of
bacteria have a DT faster than 1 h in the wild and 3.5% have a
DT slower than 100 h.
To investigate how robust these conclusions are to statisti-
cal sampling, we bootstrapped the accumulation and
mutation rate estimates, refit the lognormal distributions
and reinferred the distribution of DT. The 95% confidence
intervals for the median are quite broad at 3.4 to 14.2 h (3.1
to 11.3 h if we exclude M. florum). However, all bootstrapped
distributions show substantial variation in the DT with a sub-
stantial fraction of bacteria with long DTs and also some with
very short DTs (figure 3).
We have assumed that there is no phylogenetic inertia
within the accumulation and mutation rate estimates. To
test whether this is the case we constructed a phylogenetic
tree using 16S rRNA sequences and applied the tests of
Pagel [33] and Blomberg et al. [34]. Both the accumulation
and mutation rate data show some evidence of phylogenetic
signal. For the accumulation data, Pagel’s
l
¼0.68 ( p¼
0.001) and Blomberg et al.’s K¼0.0005 ( p¼0.35); and for
the mutation rate data Pagel’s
l
¼0.88 ( p¼0.026) and
Blomberg et al.’s K¼0.5 ( p¼0.009). We also find some evi-
dence that the data depart from a Brownian motion model
using Pagel’s test (i.e.
l
is significantly less than one) for
the accumulation data ( p,0.001) but not the mutation rate
data ( p¼0.094); i.e. the accumulation rates are more different
than we would expect from their phylogeny and a Brownian
motion model. A visual inspection of the data suggests that
the phylogenetic signal is largely contributed by species
that are closely related, rather than deeper phylogenetic
levels (figure 4a,b) and species for which we have accumu-
lation and mutation rate estimates are interspersed with one
another on the phylogenetic tree ( figure 4c). It, therefore,
seems unlikely that phylogenetic inertia will influence our
results.
It is of interest to compare the distribution of DTs in the
wild to the distribution of laboratory DTs (figure 1). The
−18
−16
−14
−12
sample quantiles
−2 −1 012
−23
−22
−21
−20
−19
theoretical
q
uantiles
−2 −1 012
theoretical
q
uantiles
−2 −1 0 1 2
−23
−22
−21
(b)(a)
Figure 1. Normal Q–Qplots for the log of (a) accumulation and (b) mutation rate data. The main plot in (b) includes all 26 mutation rate estimates and the insert
excludes Mesoplasma florum estimate.
doublin
g
time (h)
density
0.01 0.1 1 10 100 1000 10 000
doublin
g
time (h)
0.01 0.1 1 10 100 1000 10 000
0
0.1
0.3
0.5
0.7
0.9
0
0.1
0.3
0.5
0.7
0.9
Figure 2. The distribution of DTs among bacteria inferred assuming different levels of correlation between the accumulation and mutation rates—orange r¼0,
purple r¼0.5 and red r¼0.75. We also show the distribution of laboratory DTs (green histogram) from a compilation of over 200 species made by Vieira-Silva &
Rocha [32]. In (a) we include all mutation rate estimates and in (b) we exclude the mutation rate estimate for Mesoplasma florum. (Online version in colour.)
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
3
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

distributions are different in two respects. First, the median
laboratory DT of 3 h is less than half the median wild DT
of 7.04 h (6.16 h without M. florum); the two are significantly
different ( p¼0.012 with M. florum and p¼0.016 without M.
florum, inferred by bootstrapping each dataset and recalculat-
ing the medians). Second, many more bacteria are inferred to
have long DTs in the wild than in the laboratory.
3. Discussion
The DT of most bacteria in their natural environment is not
known. We have used estimates of the rate at which bacteria
accumulate mutations in their natural environment and esti-
mates of the rate at which they mutate in the laboratory to
estimate the DT for several bacteria and infer the distribution
of DTs across bacteria. We estimate that DT are generally
longer in the wild than in the laboratory, but critically we
also infer that DTs vary by several orders of magnitude
between bacterial species and that many bacteria have very
slow DTs in their natural environment.
The method, by which we have inferred the DT in the wild,
makes three important assumptions. We assume that the
mutation rate per generation is the same in the laboratory
and in the wild. However, it seems likely that bacteria in the
wild will have a higher mutation rate per generation than
those in the laboratory for two reasons. First, bacteria in the
wild are likely to be stressed and this can be expected to elevate
the mutation rate [35– 39]. Second, if we assume that DTs are
longer in the wild than the laboratory then we expect the
mutation rate per generation to be higher in the wild than in
the laboratory because some mutational processes do not
depend upon DNA replication. The relative contribution of
replication dependent and independent mutational mechan-
isms to the overall mutation rate is unknown. Rates of
substitution are higher in Firmicutes that do not undergo spor-
ulation suggesting that replication is a source of mutations in
this group of bacteria [40]. However, rates of mutation accumu-
lation seem to be similar in latent versus active infections of
Mycobacterium tuberculosis, suggestingthat replication independent
mutations might dominate in this bacterium [41,42].
The second major assumption is that the rate at which
mutations accumulate in the wild is equal to the mutation
rate per year; in effect, we are assuming that all mutations
are effectively neutral, at least over the time frame in which
they are assayed (or that some are inviable, but the same
0.01 0.1 1 10 100 1000 10 000
0.01 0.1 1 10 100 1000 10 000
0.01 0.1 1 10 100 1000 10 000
0.01 0.1 1 10 100 1000 10 000
0.01 0.1 1 10 100 1000 10 000
0.01 0.1 1 10 100 1000 10 000
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
density densitydensity
doublin
g
time (h) doublin
g
time (h)
(e)
(f)
(b)
(a)
(c)
(d)
Figure 3. DT distributions inferred by bootstrapping the accumulation and mutation rate data and refitting the lognormal distributions to both datasets. Eachplotshows
20 bootstrap DT distributions assuming different levels of correlation between the accumulation and mutation rates—orange r¼0, purple r¼0.5 and red r¼0.75.
(a–c) Include all mutation rate estimates and (d–f) show the analysis after removal of the Mesoplasma florum mutation rate estimate. (Online version in colour.)
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
4
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

proportion are inviable in the wild and the laboratory). In
those accumulation rate studies, in which they have been
studied separately, non-synonymous mutations accumulate
more slowly than synonymous mutations; relative rates
vary from 0.13 to 0.8, with a mean of 0.57 (electronic sup-
plementary material, table S3). There is no correlation
between the time frame over which the estimate was made
and the ratio of non-synonymous and synonymous accumu-
lation rates (r¼0.2, p¼0.53). We did not attempt to control
for selection because the relative rates of synonymous and
non-synonymous accumulation are only available for a few
species, and the relative rates vary between species. However,
we can estimate the degree to which more selection against
deleterious non-synonymous accumulations in the wild
causes the DT to be underestimated as follows. The observed
rate at which mutations accumulate in a bacterial lineage is
m
obs ¼
am
true
d
iþð1
a
Þð1
b
Þ
m
true
d
sþ
ð1
a
Þ
bm
true
d
n,ð3:1Þ
where
a
is the proportion of the genome that is non-coding
and
b
is the proportion of mutations in protein coding
Burkholderia cenocepacia
Janthinobacterium lividum
Teredinibacter turnerae
Escherichia coli
Salmonella enterica
Vibrio shilonii
Vibrio fischeri
Vibrio cholerae
Colwellia psychrerythraea
Pseudomonas aeruginosa
Caulobacter crescentus
Agrobacterium tumefaciens
Rhodobacter sphaeroides
Ruegeria pomeroyi
Bacillus subtilis
Staphylococcus aureus
Staphylococcus epidermidis
Mesoplasma florum
Lactococcus lactis
Micrococcus sp.
Arthrobacter sp.
Kineococcus radiotolerans
Mycobacterium smegmatis
Deinococcus radiodurans
10–5
10–6
10–7
10–8
1 × 10–8
2 × 10–9
5 × 10–10
1 × 10–10
Chlamydophila psittaci
Helicobacter pylori
Campylobacter jejuni
Neisseria meningitidis
Neisseria gonorrhoeae
Burkholderia dolosa
Bordetella pertussis
Vibrio cholerae
Buchnera aphidicola
Yersinia pestis
Yersinia pseudotuberculosis
Klebsiella pneumoniae
Shigella sonnei
Escherichia coli
Shigella dysenteriae
Salmonella enterica
Pseudomonas aeruginosa
Acinetobacter baumannii
Legionella pneumophila
Mycoplasma gallisepticum
Clostridium difficile
Streptococcus pneumoniae
Streptococcus equi
Streptococcus pyogenes
Streptococcus agalactiae
Enterococcus faecium
Staphylococcus aureus
Renibacterium salmoninarum
Mycobacterium abscessus
Mycobacterium ulcerans
Mycobacterium leprae
Mycobacterium bovis
Mycobacterium tuberculosis
Treponema pallidum
Campylobacter jejuni
Helicobacter pylori
Caulobacter crescentus
Rhodobacter sphaeroides
Ruegeria pomeroyi
Agrobacterium tumefaciens
Neisseria meningitidis
Neisseria gonorrhoeae
Bordetella pertussis
Janthinobacterium lividum
Burkholderia cenocepacia
Burkholderia dolosa
Legionella pneumophila
Acinetobacter baumannii
Pseudomonas aeruginosa
Teredinibacter turnerae
Colwellia psychrerythraea
Yersinia pestis
Yersinia pseudotuberculosis
Klebsiella pneumoniae
Salmonella enterica
Shigella sonnei
Escherichia coli
Shigella dysenteriae
Buchnera aphidicola
Vibrio cholerae
Vibrio fischeri
Vibrio shilonii
Deinococcus radiodurans
Clostridium difficile
Bacillus subtilis
Staphylococcus epidermidis
Staphylococcus aureus
Enterococcus faecium
Lactococcus lactis
Streptococcus pneumoniae
Streptococcus agalactiae
Streptococcus pyogenes
Streptococcus equi
Mycoplasma gallisepticum
Mesoplasma florum
Mycobacterium abscessus
Mycobacterium smegmatis
Mycobacterium leprae
Mycobacterium ulcerans
Mycobacterium tuberculosis
Mycobacterium bovis
Micrococcus sp.
Renibacterium salmoninarum
Arthrobacter sp.
Kineococcus radiotolerans
Treponema pallidum
Chlamydophila psittaci
accumulation rate (mutations/site/year)
mutation rate (mutations/site/generation)
(b)(a)
(c)
Figure 4. (a) 16S rRNA phylogeny and mutation rate estimates for 24 species of bacteria (two species are excluded because of erroneous positioning on the
phylogeny—see electronic supplementary material, figure S2A,B for details). (b) 16S rRNA phylogeny and accumulation rate estimates for 34 species of bacteria.
(c) 16S rRNA phylogeny combining species for which we have an estimate of the mutation rate and/or accumulation rate. Coloured dots indicate which kind of
information each species provides—red, accumulation rate; green,mutation rate and blue, both a mutation rate and an accumulation rate. (Online version in colour.)
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
5
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

sequence that are non-synonymous.
d
x
is the proportion of
mutations of class x(iis intergenic, sis synonymous and n
is non-synonymous) that are effectively neutral.
a
and
b
are
approximately 0.15 and 0.7, respectively, in our dataset.
Although there is selection on synonymous codon use in
many bacteria [43], selection appears to be weak [44] we,
therefore, assume that
d
s
¼1. This implies, from the rate at
which non-synonymous mutations accumulate relative to
synonymous mutations, that
d
n
¼0.6. A recent analysis of
intergenic regions in several species of bacteria has concluded
that selection is weaker in intergenic regions than at non-
synonymous sites; we, therefore, assume that
d
i
¼0.8 [45].
Using these estimates suggests that selection leads us to
underestimate the true mutation rate per year in the wild
by approximately 27%; this in turn means we have over-
estimated the DT by approximately 37%, a relatively small
effect. To investigate how sensitive this estimate is to the par-
ameters in equation 1, we varied each of them in turn
(electronic supplementary material, table S4). We find that
the observed mutation rate is most sensitive to selection on
synonymous codon use, because if there is selection on
synonymous codon use this also affects our estimates of selec-
tion at non-synonymous sites and in intergenic. For example,
if selection on synonymous codon use depressed the synon-
ymous accumulation rate by 0.5 this would lead to an
underestimate of the mutation rate of 63%, which would in
turn have led to a 2.7-fold over-estimate of the DT.
Finally, although each study attempted to remove single
nucleotide polymorphisms (SNPs) that had arisen by recom-
bination, it is possible that some are still present in the data.
Recombinant SNPs can have two effects. First, if they have
recombined from outside the clade, they inflate the accumu-
lation rate estimate and hence lead to an underestimate of the
DT. Second, if there is recombination within a clade, they
affect the phylogeny and potentially lead to the root of the
tree being estimated as younger than it should be. This will
lead to an overestimate of the DT.
It is important to appreciate that our method estimates an
average DT within a particular environment that the bacteria
were sampled from. The bacterium may go through periods
of quiescence interspersed with periods of growth.
Despite the assumptions we have made in our method,
our estimate of the DT of Pseudomonas aruginosa of 2.3 h in
a CF patient is very similar to that independently estimated
using the ribosomal content of cells of between 1.9 and
2.4 h [3]. There is also independent evidence that there are
some bacteria that divide slowly in their natural environ-
ment. The aphid symbiont Buchnera aphidicola is estimated
to double every 175– 292 h in its host [46,47], and Mycobacter-
ium leprae doubles every 300– 600 h on mouse footpads
[48–50], not its natural environment, but one that is probably
similar to the human skin. Furthermore, in a recent selection
experiment, Avrani et al. [51] found that several E. coli popu-
lations, which were starved of resources, accumulated
mutations in the core RNA polymerase gene. These
mutations caused these strains to divide more slowly than
unmutated strains when resources were plentiful. Interest-
ingly these same mutations are found at high frequency in
unculturable bacteria, suggesting that there is a class of
slow growing bacteria in the environment that are adapted
to starvation.
Korem et al. [52] have recently proposed a general method
by which the DT can be potentially estimated. They note that
actively replicating bacterial cells have two or more copies of
the chromosome near the origin of replication but only one
copy near the terminus, if cell division occurs rapidly after
the completion of DNA replication. Using next-generation
sequencing, they show that it is possible to assay this signal
and that the ratio of sequencing depth near the origin and ter-
minus is correlated to bacterial growth rates in vivo.Brown
et al. [53] have extended the method to bacteria without a
reference genome and/or those without a known origin
and terminus of replication. In principle, these measures of
cells performing DNA replication could be used to estimate
the DT of bacteria in the wild. However, it is unclear how
or whether the methods can be calibrated. Both Korem
et al. (2015) and Brown et al. (2016) find that their replication
measures have a median of approximately 1.3 across bacteria
in the human gut. However, a value of 1.3 translates into
different relative and absolute values of the DT in the two
studies. Brown et al. [53] show that their measure of replica-
tion, iRep, is highly correlated to Korem et al.’s [52]
measure, PTR, for data from Lactobacillus gasseri; the equation
relating the two statistics is iRep ¼20.75 þ2 PTR. Hence,
when PTR ¼1.3, iRep ¼1.85 and when iRep ¼1.3, PTR ¼
1.03. The two methods are not consistent. They also yield
very different estimates for the absolute DT. Korem et al.
[52] show that PTR is highly correlated to the growth rate
of E. coli grown in a chemostat. If we assume that the relation-
ship between PTR and growth rate is the same across bacteria
in vivo and in vitro, then this implies that the median DT for
the human microbiome is approximately 2.5 h. By contrast,
Brown et al. [53] estimate the growth rate of Klebsiella oxytoca
to be 19.7 h in a new-born baby using faecal counts and find
that this population has an iRep value of approximately 1.77.
This value is greater than the vast majority of bacteria in the
human microbiome and bacteria in the Candidate Phyla Radi-
ation, suggesting that most bacteria in these two communities
replicate very slowly. These discrepancies between the two
methods suggest that it may not be easy to calibrate the PTR
and iRep methods to yield estimates of the DT across bacteria.
Finally, how should we interpret our results for the five
focal species in the context of what is known of their ecology?
Vibrio cholerae displays the shortest DT of 1.1 h. Vibrio species
are ubiquitous in estuarine and marine environments [54].
They are known to have very short generation times in cul-
ture, the shortest being Vibrio natriegens of just 9.8 min [55].
In the wild they can exploit a wide range of carbon and
energy sources, and as such have been termed ‘opportuni-
trophs’ [56]. Natural Vibrio communities do not grow at an
accelerated rate continuously, but can exist for long periods
in a semi-dormant state punctuated by rapid pulses of high
growth rates [57], or blooms [58], when conditions are favour-
able. It has also been argued that the unusual division of
Vibrio genomes into two chromosomes facilitates more rapid
growth [59]. By pointing to a very short DT in V. cholerae,
our analysis is, therefore, consistent with what is known of
the ecology of this species.
Staphylococcus aureus is predominantly found on animals
and humans and inhabits various body parts, including the
skin and upper respiratory tract [60]. It can cause infection of
the skin and soft tissue as well as bacteraemia [61]. Staphylococ-
cus aureus exhibits a range of modes of growth, some of which
may to allow it to survive stress and antimicrobials while in its
host. For instance, small subpopulations can adopt a slow-
growing, quasi-dormant lifestyle, either in a multicellular
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
6
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

biofilm or as small colony variants (SCVs) or persister cells
[62]. Our short DT of 1.8 h suggests this is not the typical
state for S. aureus in the wild, which is not surprising consider-
ing the incidence of SCVs in clinical samples is fairly low,
between 1 and 30% [63].
Pseudomonas aeruginosa can inhabit a wide variety of
environments, including soil, water, plants and animals.
Like our other focal species, it is an opportunistic pathogen
and can also infect humans, especially those with compro-
mised immune systems, such as patients with CF. In this
context infection is chronic. Parallel evolution, the differential
regulation of genes which allow it to evade the host immune
system and resist antibiotic treatment during infection [64],
and evidence of positive selection [65] suggests P. aeruginosa
can adapt to the lungs of individuals with CF for its long-
term survival. It is known to actively grow in sputum [3],
where it uses the available nutrition which supports its
growth to high population densities [66]. Its ability to adapt
and actively grow in the CF sputum is consistent with its rela-
tively short DT of 2.3 h, especially considering this is the
environment in which the accumulation rate was measured
and matches that estimated by Yang et al. [3].
Escherichia coli and S. enterica primarily reside in the lower
intestine of humans and animals, but can also survive in the
environment. Although E. coli is commonly recovered from
environmental samples, it is not thought able to grow or sur-
vive for prolonged periods outside of the guts of warm
blooded animals, except in tropical regions where conditions
are more favourable [67], although some phylogenetically
distinct strains appear to reproduce and survive well in the
environment [68]. In contrast, Salmonella is also an enteric
colonizer of cold-blooded animals, in particular reptiles, is
better adapted than E. coli at surviving and growing in
environmental niches. For example, Salmonella can survive
and grow for at least a year in soil [69], whereas E. coli can
survive for only a few days [70]. Although these secondary
niches may play a greater role in Salmonella than in E. coli,
it remains the case the growth rates in the environment will
be much lower than those in a gut. Therefore, the increased
tenacity of Salmonella in non-host environments compared
to E. coli might help to explain the slower DT in this species.
In summary, the availability of accumulation and
mutation rate estimates allows us to infer the DT for bacteria
in the wild, and the distribution of wild DTs across bacterial
species. These DT estimates are likely to be underestimates
because the mutation rate per generation is expected to be
higher in the wild than in the laboratory, and some mutations
are not generated by DNA replication. Our analysis, there-
fore, suggests that DTs in the wild are typically longer than
those in the laboratory, that they vary considerably between
bacterial species and that a substantial proportion of species
have very long DTs in the wild.
4. Material and methods
We compiled estimates of the accumulation and mutation rate of
bacteria from the literature. We only used mutation rate estimates
that came from a mutation accumulation experiment with whole-
genome sequencing. If we had multiple estimates of the mutation
rate we summed the number of mutations across the mutation
accumulation experiments and divided this by the product of
the genome size and the number of generations that were
assayed. We averaged the accumulation rate estimates where
we had multiple estimates from the same species. We recalcu-
lated the accumulation rates in two cases in which the number
of accumulated mutations had been divided by an incorrect
number of years: E. coli [5] and Helicobacter pylori [71]. For
E. coli, we reestimated the accumulation rate using BEAST by
constructing sequences of the SNPs reported in the paper and
the isolation dates. For, Helicobacter pylori the 3-year and 16-
year strains appear to form a clade to the exclusion of the 0-
year strain because they share some differences from the 0-year
strain. If the number of substitutions that have accumulated
between the common ancestor of the 3-year and 16-year strain
and each of the two strains are S
3
and S
16
, respectively, then
the rate of accumulation can be estimated as (S
16
2S
3
)/(13
years genome size). For the isolates NQ1707 and NQ4060
we have estimated the accumulation rate to be 5 10
26
and
for NQ1671 and NQ4191 5.9 10
26
. We excluded some accumu-
lation rate estimates for a variety of reasons. We only considered
accumulation rates sampled over an historical time frame of at
most 1500 years. Most of our estimates of the accumulation
rate are for all sites in the genome, so we excluded cases in
which only the synonymous accumulation rate was given. We
also excluded accumulation rates from hypermutable strains.
Accumulation and mutation rate estimates used in the analysis
are given in electronic supplementary material, tables S1 and
S2, respectively.
The estimate of the standard error associated with our esti-
mate of the DT was obtained using the standard formula for
the variance of a ratio: V(x/y)¼(M(x)/M(y))
2
(V(x)/M(x)
2
þ
V(y)/M(y)
2
) where Mand Vare the mean and variance of x
and y. The variance for the accumulation rate was either the var-
iance between multiple estimates of the accumulation rate if they
were available, or the variance associated with the estimate if
there was only a single estimate. The variance associated with
the mutation rate was derived by assuming that the number of
mutations was Poisson distributed.
We fit lognormal distributions to the accumulation and
mutation rate data by taking the log
e
of the values and then fit-
ting a normal distribution by maximum likelihood using the
FindDistributionParameters in Mathematica. Normal Q–Qplots
for the accumulation and mutation rate data were produced
using the qqnorm function in R version 1.0.143. In fitting these
distributions, we have not taken into account the sampling
error associated with the accumulation and mutation rate esti-
mates. However, these sampling errors are small compared to
the variance between species: for the accumulation rates the var-
iance between species is 3.9 10
211
and the average error
variance is an order of magnitude smaller at 3.6 10
212
; for
the mutation rate data, the variance between species is 7.5 
10
218
and the average variance associated with sampling is
more than two orders of magnitude smaller at 1.8 10
220
.
Note that we cannot perform these comparisons of variances
on a log-scale because we do not have variance estimates for
the log accumulation and mutation rates.
To estimate phylogenetic signal in the accumulation and
mutation rates we generated phylogenetic trees for each set of
species in the two datasets. 16S rRNA sequences were down-
loaded from the NCBI genome database (https://www.ncbi.
nlm.nih.gov/genome/) and aligned using MUSCLE [72] per-
formed in Geneious version 10.0.9 (http://www.geneious.com,
Kearse et al. [73]). From these alignments, maximum-likelihood
trees were constructed in RAxML [74] and integrated into the
tests of Pagel [33] and Blomberg et al. [34] to the log
10
(accumu-
lation rates) and log
10
(mutation rates) implemented in the
phylosig function in the R package phytools v.0.6 [75]. For the
mutation rate dataset two species were excluded because of erro-
neous positioning on the phylogeny. See electronic supplementary
material, figure S2A,B for details.
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
7
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

Data accessibility. All relevant data for this work are available as elec-
tronic supplementary material accompanying this article.
Authors’ contributions. All authors were involved in conceptualizing the
work. B.G. and A.E.W. wrote the manuscript with support from
D.J.W. and E.F. B.G. and A.E.W. carried out data collection and
analysis. All authors provided critical feedback and shaped the
research, analysis and manuscript.
Competing interests. The authors declare that they have no competing
interests.
Funding. D.J.W. is a Sir Henry Dale Fellow, jointly funded by the Well-
come Trust and the Royal Society (grant no. 101237/Z/13/Z).
Acknowledgements. We are very grateful to Michael Lynch for sharing
his mutation rate estimates prior to publication.
References
1. Harmsen HJM, van Kuijk BLM, Plugge CM,
Akkermans ADL, De Vos WM, Stams AJM. 1998
Syntrophobacter furnaroxidans sp. nov., a syntrophic
propionate-degrading sulfate-reducing bacterium.
Int. J. Syst. Bacteriol. 48, 1383– 1388. (doi:10.1099/
00207713-48-4-1383)
2. Gibbons RJ, Kapsimalis B. 1967 Estimates of the
overall rate of growth of the intestinal microflora of
hamsters, guinea pigs, and mice. J. Bacteriol. 93,
510–512.
3. Yang L, Haagensen JAJ, Jelsbak L, Johansen HK,
Sternberg C, Høiby N, Molin S. 2008 In situ growth
rates and biofilm development of Pseudomonas
aeruginosa populations in chronic lung infections.
J. Bacteriol. 190, 2767–2776. (doi:10.1128/JB.
01581-07)
4. Drummond AJ, Pybus OG, Rambaut A, Forsberg R,
Rodrigo AG. 2003 Measurably evolving populations.
Trends Ecol. Evol. 18, 481– 488. (doi:10.1016/
S0169-5347(03)00216-7)
5. Reeves PR et al. 2011 Rates of mutation and host
transmission for an Escherichia coli clone over 3
years. PLoS ONE 6, e26907. (doi:10.1371/journal.
pone.0026907)
6. Markussen T, Marvig RL. 2014 Environmental
heterogeneity drives within-host diversification and
evolution of Pseudomonas aeruginosa.MBio 5,
e01592-14. (doi:10.1128/mBio.01592-14.Editor)
7. Marvig RL, Johansen HK, Molin S, Jelsbak L. 2013
Genome analysis of a transmissible lineage of
Pseudomonas aeruginosa reveals pathoadaptive
mutations and distinct evolutionary paths of
hypermutators. PLoS Genet. 9, e1003741. (doi:10.
1371/journal.pgen.1003741)
8. Duche
ˆne S, Holt KE, Weill F-X, Le Hello S, Hawkey J,
Edwards DJ, Fourment M, Holmes EC. 2016 Genome-
scale rates of evolutionary change in bacteria.
Microb. Genomics 2, e000094. (doi:10.1101/069492)
9. Zhou Z, McCann A, Weill F-X, Blin C, Nair S, Wain J,
Dougan G, Achtman M. 2014 Transient Darwinian
selection in Salmonella enterica serovar Paratyphi A
during 450 years of global spread of enteric fever.
Proc. Natl Acad. Sci. USA 111, 12 199–12 204.
(doi:10.1073/pnas.1411012111)
10. Okoro CK et al. 2012 Intracontinental spread of
human invasive Salmonella typhimurium
pathovariants in sub-Saharan Africa. Nat. Genet. 44,
1215–1221. (doi:10.1038/ng.2423)
11. Hawkey J, Edwards DJ, Dimovski K, Hiley L, Billman-
Jacobe H, Hogg G, Holt KE. 2013 Evidence of
microevolution of Salmonella typhimurium during a
series of egg-associated outbreaks linked to a single
chicken farm. BMC Genomics 14, 800. (doi:10.1186/
1471-2164-14-800)
12. Mather AE. 2013 Distinguishable epidemics of
multidrug-resistant Salmonella typhimurium DT104
in different hosts. Science 341, 1514– 1518.
13. Stinear TP, Holt KE, Chua K, Stepnell J, Tuck KL,
Coombs G, Harrison PF, Seemann T, Howden BP.
2014 Adaptive change inferred from genomic
population analysis of the ST93 epidemic clone of
community-associated methicillin-resistant
Staphylococcus aureus.Genome Biol. Evol. 6, 366–
378. (doi:10.1093/gbe/evu022)
14. Baines SL et al. 2015 Convergent adaptation in the
dominant global hospital clone ST239 of
methicillin-resistant Staphylococcus aureus.MBio 6,
e00080-15. (doi:10.1128/mBio.00080-15)
15. Ward MJ, Gibbons CL, McAdam PR, van Bunnik
BAD, Girvan EK, Edwards GF, Fitzgerald JR,
Woolhouse MEJ. 2014 Time-scaled evolutionary
analysis of the transmission and antibiotic resistance
dynamics of Staphylococcus aureus clonal complex
398. Appl. Environ. Microbiol. 80, 7275– 7282.
(doi:10.1128/AEM.01777-14)
16. Holden MTG et al. 2013 A genomic portrait of the
emerfences, evolution, and global spread of
methicillin-resistant Staphylococcus aureus.Genome
Res. 23, 653– 664. (doi:10.1101/gr.147710.112.
Freely)
17. Uhlemann A-C, Dordel J, Knox JR, Raven KE, Parkhill
J, Holden MTG, Peacock SJ, Lowy FD. 2014
Molecular tracing of the emergence, diversification,
and transmission of S. aureus sequence type 8 in a
New York community. Proc. Natl Acad. Sci. USA 111,
6738–6743. (doi:10.1073/pnas.1401006111)
18. Alam MT, Read TD, Petit RA, Boyle-Vavra S, Miller
LG, Eells SJ, Daum RS, David MZ. 2015 Transmission
and microevolution of USA300 MRSA in U.S.
households: evidence from whole-genome
sequencing. MBio 6, e00054-15. (doi:10.1128/mBio.
00054-15)
19. Smyth DS, McDougal LK, Gran FW, Manoharan A,
Enright MC, Song JH, De Lencastre H, Robinson DA.
2010 Population structure of a hybrid clonal group
of methicillin-resistant Staphylococcus aureus,
ST239-MRSA-III. PLoS ONE 5, e8582. (doi:10.1371/
journal.pone.0008582)
20. Harris SR et al. 2010 Evolution of MRSA during
hospital transmission and intercontinental spread.
Science. 327, 469–474. (doi:10.1126/science.
1182395)
21. Gray RR, Tatem AJ, Johnson JA, Alekseyenko AV,
Pybus OG, Suchard MA, Salemi M. 2011 Testing
spatiotemporal hypothesis of bacterial evolution
using methicillin-resistant Staphylococcus aureus
ST239 genome-wide data within a Bayesian
framework. Mol. Biol. Evol. 28, 1593– 1603. (doi:10.
1093/molbev/msq319)
22. Nu¨bel U et al. 2010 A timescale for evolution,
population expansion, and spatial spread of an
emerging clone of methicillin-resistant
Staphylococcus aureus.PLoS Pathog. 6, e1000855.
(doi:10.1371/journal.ppat.1000855)
23. Young BC et al. 2012 Evolutionary dynamics of
Staphylococcus aureus during progression from
carriage to disease. Proc. Natl Acad. Sci. USA 109,
4550–4555. (doi:10.1073/pnas.1113219109)
24. Mutreja A et al. 2011 Evidence for several waves of
global transmission in the seventh cholera pandemic.
Nature 477, 462– 465. (doi:10.1038/nature10392)
25. Lee H, Popodi E, Tang H, Foster PL. 2012 PNAS Plus:
Rate and molecular spectrum of spontaneous
mutations in the bacterium Escherichia coli as
determined by whole-genome sequencing. Proc.
Natl Acad. Sci. USA 109, E2774–E2783. (doi:10.
1073/pnas.1210309109)
26. Dettman JR, Sztepanacz JL, Kassen R. 2016 The
properties of spontaneous mutations in the
opportunistic pathogen Pseudomonas aeruginosa.BMC
Genomics 17, 1–14. (doi:10.1186/s12864-015-2244-3)
27. Lind PA, Andersson DI. 2008 Whole-genome
mutational biases in bacteria. Proc. Natl Acad. Sci.
USA 105, 17 878– 17 883. (doi:10.1073/pnas.
0804445105)
28. Long H et al. 2018 Evolutionary determinants of
genome-wide nucleotide composition. Nat. Ecol.
Evol. 2, 1– 4. (doi:10.1038/s41559-017-0425-y)
29. Sung W, Ackerman MS, Miller SF, Doak TG, Lynch
M. 2012 Drift-barrier hypothesis and mutation-rate
evolution. Proc. Natl Acad. Sci. USA 109, 18 488 –
18 492. (doi:10.1073/pnas.1216223109)
30. Drake JW. 1991 A constant rate of spontaneous
mutation in DNA-based microbes. Proc. Natl Acad. Sci.
USA 88, 7160– 7164. (doi:10.1073/pnas.88.16.7160)
31. Lynch M, Ackerman MS, Gout J, Long H, Sung W,
Thomas WK, Foster PL. 2016 Genetic drift, selection
and the evolution of the mutation rate. Nat. Publ.
Gr. 17, 704–714. (doi:10.1038/nrg.2016.104)
32. Vieira-Silva S, Rocha EPC. 2010 The systemic imprint
of growth and its uses in ecological
(meta)genomics. PLoS Genet.6, e1000808. (doi:10.
1371/journal.pgen.1000808)
33. Pagel M. 1999 Inferring the historical patterns of
biological evolution. Nature 401, 877– 884. (doi:10.
1038/44766)
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
8
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from

34. Blomberg SP, Garland T, Ives AR. 2003 Testing for
phylogenetic signal in comparative data: behavioral
traits are more labile. Evolution 57, 717– 745.
(doi:10.1111/j.0014-3820.2003.tb00285.x)
35. Bjedov I, Tenaillon O, Gerard B, Souza V, Denamur
E, Radman M, Francois T, Matic I. 2003 Stress-
induced mutagenesis in bacteria. Science 300,
1404–1409.
36. Galhardo RS, Hastings PJ, Rosenberg SM. 2007
Mutation as a stress response and the regulation of
evolvability. Crit. Rev. Biochem. Mol. Biol. 42, 399-
435. (doi:10.1080/10409230701648502)
37. Foster PL. 2007 Stress-induced mutagenesis in
bacteria. Crit. Rev. Biochem. Mol. Biol. 42, 373–
397. (doi:10.1080/10409230701648494)
38. Maclean RC, Torres-barcelo
´C, Moxon R. 2013
Evaluating evolutionary models of stress-induced
mutagenesis in bacteria. Nat. Rev. Genet. 14,
221–227. (doi:10.1038/nrg3415)
39. Shewaramani S, Finn TJ, Leahy SC, Kassen R, Rainey
B, Moon CD. 2017 Anaerobically grown Escherichia
coli has an enhanced mutation rate and distinct
mutational spectra. PLoS Genet. 13, 1– 22. (doi:10.
1371/journal.pgen.1006570)
40. Weller C, Wu M. 2015 A generation-time effect on
the rate of molecular evolution in bacteria.
Evolution. 69, 643– 652. (doi:10.1111/evo.12597)
41. Ford CB et al. 2011 Use of whole genome
sequencing to estimate the mutation rate of
Mycobacterium tuberculosis during latent infection.
Nat. Genet. 43, 482– 486. (doi:10.1038/ng.811)
42. Lillebaek T, Norman A, Rasmussen EM, Marvig RL,
Folkvardsen DB, Andersen A
˚
B, Jelsbak L. 2016
Substantial molecular evolution and mutation rates
in prolonged latent Mycobacterium tuberculosis
infection in humans. Int. J. Med. Microbiol. 306,
580–585. (doi:10.1016/j.ijmm.2016.05.017)
43. Hershberg R, Petrov DA. 2008 Selection on codon
bias. Annu. Rev. Genet. 42, 287–299. (doi:10.1146/
annurev.genet.42.110807.091442)
44. Sharp PM, Bailes E, Grocock RJ, Peden JF, Sockett
RE. 2005 Variation in the strength of selected codon
usage bias among bacteria. Nucleic Acids Res. 33,
1141–1153. (doi:10.1093/nar/gki242)
45. Thorpe HA, Bayliss SC, Hurst LD, Feil EJ. 2017
Comparative analyses of selection operating on non-
translated intergenic regions of diverse bacterial
species. Genetics 206, 363–376. (doi:10.1534/
genetics.116.195784)
46. Ochman H, Elwyn S, Moran NA. 1999 Calibrating
bacterial evolution. Proc. Natl Acad. Sci. USA 96,
12 638–12 643. (doi:10.1073/pnas.96.22.12638)
47. Clark MA, Moran NA, Baumann P. 1999 Sequence
evolution in bacterial endosymbionts having extreme
base compositions. Mol. Biol. Evol. 16, 1586– 1598.
(doi:10.1093/oxfordjournals.molbev.a026071)
48. Shepard CC. 1960 The experimental disease that
follows the injection of human leprosy bacilli into
foot-pads of mice. J. Exp. Med. 112, 445– 454.
(doi:10.1016/j.infsof.2008.09.005)
49. Rees RJ. 1964 Limited multiplication of acid-fast bacilli
in the foot-pads of mice inoculated with Mycobacterium
leprae.Br. J. Exp. Pathol. 45, 207– 218.
50. Levy L. 1976 Bactericidal action of dapsone against
Mycobacterium leprae in mice. Antimicrob. Agents
Chemother. 9, 614–617. (doi:10.1128/AAC.9.4.614)
51. Avrani S, Bolotin E, Katz S, Hershberg R. 2017 Rapid
genetic adaptation during the first four months of
survival under resource exhaustion. Mol. Biol. Evol.
34, 1758–1769. (doi:10.1093/molbev/msx118)
52. Korem T et al. 2015 Growth dynamics of gut
microbiota in health and disease inferred from
single metagenomic samples. Science 349, 1101–
1106. (doi:10.1126/science.aac4812)
53. Brown CT, Olm MR, Thomas BC, Banfield JF. 2016
Measurement of bacterial replication rates in
microbial communities. Nat. Biotechnol. 34, 1256–
1263. (doi:10.1038/nbt.3704)
54. Reidl J, Klose KE. 2002 Vibrio cholerae and cholera:
out of the water and into the host. FEMS Microbiol.
Rev. 26, 125–139. (doi:10.1111/j.1574-6976.2002.
tb00605.x)
55. G. ER. 1961 Pseudomonas natriegens, a marine
bacterium with a generation time of less than
10 min. J. Bacteriol. 83, 736 – 737.
56. Polz MF, Hunt DE, Preheim SP, Weinreich DM. 2006
Patterns and mechanisms of genetic and phenotypic
differentiation in marine microbes. Phil. Trans. R. Soc.
B361, 2009– 2021. (doi:10.1098/rstb.2006.1928)
57. Blokesch M, Schoolnik GK. 2008 The extracellular
nuclease Dns and its role in natural transformation
of Vibrio cholerae.J. Bacteriol. 190, 7232– 7240.
(doi:10.1128/JB.00959-08)
58. Takemura AF, Chien DM, Polz MF. 2014 Associations
and dynamics of Vibrionaceae in the environment,
from the genus to the population level. Front.
Microbiol. 5, 1– 26. (doi:10.3389/fmicb.2014.00038)
59. Yamaichi Y, Iida T, Park KS, Yamamoto K, Honda T.
1999 Physical and genetic map of the genome of
Vibrio parahaemolyticus: presence of two
chromosomes in Vibrio species. Mol. Microbiol. 31,
1513-1521. (doi:10.1046/j.1365-2958.1999.01296.x)
60. Schenck LP, Surette MG, Bowdish DME. 2016
Composition and immunological significance of the
upper respiratory tract microbiota. FEBS Lett. 590,
3705–3720. (doi:10.1002/1873-3468.12455)
61. John JF. 2004 Staphylococcal infection: emerging
clinical syndromes. In Staphylococcus aureus (eds
DAA Ala’Aldeen, K Hiramatsu), pp. 1– 30.
Chichester, UK: Horwood Publishing Limited.
62. Bui LMG, Conlon BP, Kidd SP. 2017 Antibiotic
tolerance and the alternative lifestyles of
Staphylococcus aureus.Essays Biochem. 61, 71– 79.
(doi:10.1042/EBC20160061)
63. Proctor RA, von Eiff C, Kahl BC, Becker K, McNamara
P, Herrmann M, Peters G. 2006 Small colony
variants: a pathogenic form of bacteria that
facilitates persistent and recurrent infections. Nat.
Rev. Microbiol. 4, 295– 305. (doi:10.1038/
nrmicro1384)
64. Huse HK, Kwon T, Zlosnik JEA, Speert DP, Marcotte
EM, Whiteley M. 2010 Parallel evolution in
Pseudomonas aeruginosa over 39 000 generations in
vivo.MBio 1, e00199-10. (doi:10.1128/mBio.00199-
10.Editor)
65. Smith EE et al. 2006 Genetic adaptation by
Pseudomonas aeruginosa to the airways of cystic
fibrosis patients. Proc. Natl Acad. Sci. USA 103,
8487–8492. (doi:10.1073/pnas.0602138103)
66. Palmer KL, Mashburn LM, Singh PK, Whiteley M.
2005 Cystic fibrosis sputum supports growth and
cues key aspects of Pseudomonas aeruginosa
physiology cystic fibrosis sputum supports growth
and cues key aspects of Pseudomonas aeruginosa
physiology. J. Bacteriol. 187, 5267– 5277. (doi:10.
1128/JB.187.15.5267)
67. Winfiel MD, Groisman EA. 2003 Role of nonhost
enviroments in the lifestyles of Salmonella and
E. coli.Appl. Environ. Mcrobiol. 69, 3687– 3694.
(doi:10.1128/AEM.69.7.3687)
68. Oh S, Buddenborg S, Yoder-Himes DR, Tiedje JM,
Konstantinidis KT. 2012 Genomic diversity of
Escherichia isolates from diverse habitats. PLoS ONE
7, e47005. (doi:10.1371/journal.pone.0047005)
69. Davies RH, Wray C. 1996 Seasonal variations in the
isolation of Salmonella typhimurium, Salmonella
enteritidis, Bacillus cereus and Clostridium
perfringens from environmental samples. Zentralbl.
Veterinarmed. B 43, 119–127. (doi:10.1111/j.1439-
0442.1996.tb00435.x)
70. Bogosian G, Sammons L. 1996 Death of the
Escherichia coli K-12 strain W3110 in soil and water.
Appl. Environ. Microbiol. 62, 4114– 4120.
71. Kennemann L et al. 2011 Helicobacter pylori
genome evolution during human infection. Proc.
Natl Acad. Sci. USA 108, 5033– 5038. (doi:10.1073/
pnas.1018444108)
72. Edgar RC. 2004 MUSCLE: Multiple sequence
alignment with high accuracy and high throughput.
Nucleic Acids Res. 32, 1792–1797. (doi:10.1093/
nar/gkh340)
73. Kearse M et al. 2012 Geneious basic: an integrated
and extendable desktop software platform for the
organization and analysis of sequence data.
Bioinformatics 28, 1647– 1649. (doi:10.1093/
bioinformatics/bts199)
74. Stamatakis A. 2014 RAxML version 8: a tool for
phylogenetic analysis and post-analysis of large
phylogenies. Bioinformatics 30, 1312– 1313.
(doi:10.1093/bioinformatics/btu033)
75. Revell LJ. 2012 phytools: An R package for
phylogenetic comparative biology (and other
things). Methods Ecol. Evol. 3, 217–223. (doi:10.
1111/j.2041-210X.2011.00169.x)
rspb.royalsocietypublishing.org Proc. R. Soc. B 285: 20180789
9
on June 13, 2018http://rspb.royalsocietypublishing.org/Downloaded from