>Did the resistant form of the bacteria show up as a result of random
>mutations, or were they already in the population and became dominant when
>the non-resistant individuals were knocked off by the presence of
>penicillin? Is there evidence that will decide this question?
As penicillin (and several other antibiotics) naturally occur in fungi, it
seems likely that bacteria that contact these fungi could have genes for
resistance. Thus, at least some of the appearance of drug resistance is
believed to represent either selection for previously rare genes or
transfer of genes between different types of bacteria. I believe some
studies have tried to look for old bacteria (possibly reviving old spores)
to look for resistance earlier than the discovery of a particular drug, but
do not recall further details.
Also, Bob DeHaan quoted Darwin as saying "Only those variations which are
in some way profitable will be preserved or naturally selected."
Current evolutionary theory, however, recognizes that neutral
variation is also possible and can persist, although there will be no
strong influence to keep a neutral innovation. Even slightly deleterious
variations can survive if the negative impact on fitness is not enough to
drive it to extinction (e.g., genetic diseases that do not cause
significant harm until post-reproduction age). In small populations or in
regions with limited crossing over (Y chromosome, mitochondrial DNA, etc.),
population genetics show that mildly deleterious mutations can become
predominant in a population despite contrary selection. "Random"
catastrophes can also result in the survival of less fit individuals and
death of fitter ones simply because they were at the wrong place at the
wrong time (e.g., near an asteroid impact).
Also, a broad definition of "some way profitable" is necessary. A
classic example is the sickle-cell trait in malarial regions in Africa-the
heterozygous condition was the healthiest, so selection supported
persistance of the gene in the population, even though homozygotes for the
trait would die in childhood. A deleterious trait may also be genetically
linked to one that is more beneficial than it is harmful, and thereby be
able to spread. Another counter-intuitive way in which traits can be
profitable is sexual selection. Females prefer showier males in many
species, yet showier males are also easier for predators to spot. The net
benefit, measured in terms of proportional survival of offspring, is what
is crucial to natural selection.
Thus, just about anything can have an evolutionary explanation.
Whether or not the explanation is correct is another issue.