In article <D70K2q.29A at murdoch.acc.Virginia.EDU> Mahlon G. Kelly,
mgk at darwin.clas.Virginia.EDU writes:
>Rubbish. Evolution via natural selection only occurs as it
>selects certain genomes in large numbers of organisms, i.e.
>populations. The loss of a gene or trait in a single individual
>is unimportant except in the small populations in which genetic
>drift occurs.
>
Why the "large numbers" qualifier? You are either being overly
restrictive in your definition or there is a significant misunder-
standing here. Selection is a *general* principle applicable
to any level of biological organization (gene, "individual" organism,
population, species, whatever) as long as there are units at that
level which reproduce to form offspring units that tend to share
the traits of the parents, *and* when those heritable traits lead
to differences in the reproduction of those units. Thus, there may
be so-called "individual level" selection, driven by the differential
reproduction of individual organisms *within* one population
(one should nod, as you have, to the fact that these "individuals"
are cooperating genes in a *generally* obligate mutualism, but
that does not change the fact that the level of selection is still
at the so-called "individual" organism). Similarly, there may
be selection at the level of the gene, for example in cases of
meiotic drive. Finally, to the extent that species "reproduce"
through speciation, "passing on" traits with some fidelity to
daughter species, and to the extent that those traits tend to
affect speciation rates or extinction probabilities, there may
be something called "species selection". The latter is a bit fuzzier,
but I see no reason to exclude it from consideration.
Natural selection can occur at all of these levels, but it is
generally accepted that traits are most often explained by selection
at the level of the "individual" organism (thus the phrase "selection
is generally strongest at the level of the individual").
dave marshall, dmarshal at umich.edu
