Re: Reply to Rindos re genetic drift

mike salovesh (T20MXS1@NIU.BITNET)
Wed, 18 May 1994 01:46:00 CDT

On Wed., 18 May 1993, Dave Rindos said--in part--

==============partial quote, with no internal deletions==============

Of greater importance in terms of the specific question posed here, we
must consider the matter of "random" events and their role in evolution.
This is a complex business, and hard to spin out quickly in this medium.

When the claim is made that an evolutionary change is due to NON-selected
["random"] processes, such as drift or the closely related founder effect,
what is really being claimed?

Clearly, no claim whatsoever is being made that the change OPPOSES
selective forces. Drift in traits cannot occur if the state being
drifted towards has a lower fitness value at any moment in time. Hence,
minimally, the claim is made that the change in states is evolutionarily
neutral. That no *difference* at the level of causation exists.

====================end of partial quote============================

The discussion that follows is NOT about genetic drift, sometimes
called "the Sewall Wright effect".

The whole point to genetic drift is that it CAN work toward a state
that has a lower fitness value. It is, by definition, NOT the same
as selection ("natural" or otherwise).

For drift to occur, it is necessary that the effective breeding
group consist of a small number of individuals. ("Small" is a rela-
tive term, of course. In practical terms, it probably means
n less than 500; below that, the smaller the breeding population,
the more likely the operation of drift.) It is also necessary that
there be some variation in the initial breeding group. (You can't
get drift OR selection if the frequency of an allele is 100% !)
That's it. Given these two conditions, drift WILL occur and its
direction is independent of selection. And, typically, it takes
place in just one generation, not gradually through many.

The next question is what does it matter? Well, it matters if that
once small breeding population expands, e.g., to fill a new terri-
torial niche. Consider: the groups that originally peopled the
Americas appear to have been hunting and gathering bands, with
effective breeding populations around 20 to 40 individuals or less.
It looks like they were in at least partial reproductive isolation
most of each year, if we can make comparative analogies with known
historical groups (e.g., Shoshone, Inuit). But these small bands
expanded to fill a pair of continents. Those are exactly the
circumstances where non-selective genetic drift has made a terrific
evolutionary difference. Native Americans are simply not the same as
Siberians or any other Old World group, nor could they possibly be
expected to be, given their history.

But can drift go in the direction of "less fitness"? This is not a
semantic quibble, where one might assert the fact that the new
gene distribution after drift is obviously more fit because it has
the descendants. Could the genes lost through drift be ones that
would have produced more offspring than the ones that remained if
only they had not been eliminated by random, non-selective chance?
Of course! The chance that that would happen in a very large,
randomly mating breeding population is vanishingly small. The
chance that it would happen is a very small population is, for
all practical purposes, 1 or one hundred per cent or however you want
to express effective certainty.

No very small group can possibly duplicate the genetic distributions
of the larger group from which it is taken. Isolate a very small
group from its parent population and you have already demonstrated
genetic drift. It's VERY simple statistics.

Sorry, Dave. But then, drift as genetic or evolutionary process just
is not like the kind of drift you're talking about in flowers.

Footnote of sorts: In 30 years of teaching introductory anthro, in
places ranging from small community colleges to big ten universities,
genetic drift is one of the things I've had the hardest time teaching
because my students have almost no knowledge of statistics. I can
make Hardy-Weinberg makes sense; I can get across some idea of what
cross-cousin marriage might mean; I can even sometimes get my stu-
dents to accept the fact that their culture(s) can be just as
strange and arbitrary as any others. But genetic drift never seems
to get through to them.

mike salovesh anthropology dept, northern illinois univ
<t20mxs1@niu.bitnet> OR <>

"Anthropologists just don't count in this world. If we could
count, we'd be economists. But if we understood economics we sure
as hell never would have chosen to be anthropologists!"