def of evolution

arlin at arlin at
Thu Oct 24 10:47:31 EST 1991

In article <stormo.688250251 at beagle>, stormo at boulder.Colorado.EDU (Gary Stormo) writes:
> Some time ago there was a discussion about the definition of evolution.
> While a consensus did not emerge there was considerable support for any
> change in allelic frequencies.  There have also been discussions of the

I have not seen the earlier parts of this thread, but I can tell you
that the definition of evolution is most definitely not "shifting
allele frequencies."  This canard is propagated in introductory
textbooks, but a brief look at facts familiar to us all will show that
the vast panorama of evolution cannot adequately be explained solely
by the process of replacing one allele with another.

Consider that three or four billion years ago there lived a population
of cells that was the ancestor of all extant cellular life.  These
organisms almost certainly had small circular genomes a few mega-bases
in size and with a thousand (or a few thousand) genetic loci.  Through
several billion years of "evolution," the ancestors of these organisms
have become the 100's of millions of genetically complex species we
see today.

If "evolution" is defined simply as a succession of allele
replacements, we must explain how to get from point A (the ancestor of
all extant cellular life) to point B (all extant cellular life) by a
succession of allele replacements.  Here are some problems:

1.  We started out with a small circular genome, and the modern
organisms have a great variaety of genomes from 2 Mb circles (with
accompanying plasmids) to monstrous arrays of dozens of linear
chromosomes.  Size and form of DNA has obviously changed by processes
other than switching one sequence for another at a single locus.

2.  We started out with one gene pool, and "evolution" resulted in
100's of millions of gene pools.  Thus, evolution must be more than
the shifting of the frequencies of alleles in a single gene pool.
Ernst Mayr argued all of his life for the importance of non-biological
("geographic") isolating mechanisms in evolution.  Can anyone
concieive of a mechanism of splitting one panmictic gene pool into two
by a process of replacing one allele with another (remember to justify
all intermediate steps)?

We can immediately take steps to correct the problem, by including
some other processes in the allele replacements definition of
evolution.   For instance, we can explicitly include the physical
environment to the extent that it may contribute to the isolation of
populations from each other, allowing them the opportunity to build up
incompatibilities.  Along with allele replacement, we can include the
distinct mechanisms of gene duplication and molecular drive to explain
the evolution of gene families and the comings and goings of
repetitive sequences and other genomic parasites.   Thus, the above
two problems can be addressed somewhat by generalizing the allele
replacements definition of evolution to involve "fluctuations in the
genetic composition of the population" (a la Sewall Wright) rather
than just "allele frequency changes," and we can mention the
importance of physical isolating mechanisms.

However, these corrective measures will not be enough.   If you start
to think about evolution in an open-minded way, you should be able to
think of more phenomena that cannot be explained by allele
replacements.  This is especially true if you are not one of those
molecular biology types who can only think of DNA when asked to think
about evolution.  For instance, some viruses have evolved with RNA
genomes-- could they have evolved from DNA by a process of allele
replacement?  Some organisms have evolved by taking up cellular
endosymbionts: was the endosymbiotic incorporation of a mitochondrion
in any sense an "allele replacement"? (isn't a mitochondrial membrane
a non-genetic factor necessary for inheritance of the mitochondrion?).
Perhaps the most important episode in the history of evolution on our
planet involved the oxygen crisis created by oxygenic
photosynthesizers two billion years ago: evolution changed the
environment of the planet, placing strict limits on the course of
future evolution: can we expain this with allele replacements?

More problems could be listed, but it is time to state that the basic
problem with the "allele replacements" definition of evolution is that
it represents a failed attempt at reducing a set of phenomena to a
single mechanism.  I have no quarrel with reductionism:  if we can
reduce the vast panorama of evolution to a set of precisely defined
mechanistic processes, I will applaud the effort (it should be noted
that evolution will never be reduced to a *single* mechanism, for the
same reason that there will never be a unified field theory in
physics).  At present, however, we cannot.  Therefore it is best to
define evolution as a phenomenon, rather than a mechanism.  This
definition would be verbose, and would refer to anagenetic changes,
speciation, adaptation, interactions of organisms with each other and
with the environment, evolutionary trends and other important
phenomena in the history of life.

Arlin Stoltzfus
 Department of Biochemistry
 Dalhousie University

Arlin at

usual disclaimer

More information about the Mol-evol mailing list