genome sequences of evolved clones from each of the 16
populations together with their respective ancestors. Competence
could potentially alter mutation fixation in two ways, both caused
by the fact that competence unites mutations from separate cells
into a common genetic background. First, if recombined
mutations are beneficial either alone or in combination, competence
could increase the fixation rate because recombinant cells
would be predicted to increase in frequency. Second, if
recombined mutations are deleterious, alone or in combination,
competence could reduce the fixation rate because recombinant
cells would be exposed to natural selection and eliminated from
the experimental population. A similar reduction in fixation rate
would be anticipated if competence replaces new mutations in the
host genome with donor DNA containing the wild-type allele,
thereby ‘‘correcting’’ mutations. We furthermore predict that sub-
MIC antibiotic stress will have a general increase on mutation
fixation, owing to potentially mutagenic effects of kanamycin [24].
Sequencing of evolved genomes identified a total of 421
synonymous mutations and 1282 non-synonymous mutations
across all evolved lines. Substitutions were not evenly distributed
across treatments and populations, however. Consistent with the
second possibility outlined above, we found that the total number
of mutations in competent populations was significantly lower than
in non-competent populations (Fig. 2A; GLM: z=29.344, df = 1
p,0.0001). Moreover, the total number of mutations was
significantly higher in populations experiencing periodic stress
for both competent and non-competent populations (Fig. 2A;
GLM: z= 7.379, df = 1 p,0.0001).