genetic diversity revisited

Chris Wilson (CWILSON at rsbs-central.anu.edu.au)
Fri, 12 Apr 1996 19:31:21 +0000

Dear All,

Like Joshua Levy, I've been away from my computer for awhile, so
missed whatever discussion there was re: genetic diversity. I
originally sent this reply directly to him, but after writing it was too
far into sermon mode to let it go. Hopefully this will be useful for
someone else out there.

Chris Wilson
Research School of Biological Sciences
Australian National University
Canberra, ACT
Australia 2601

-----------------------------------------------------------
Dear Joshua,

You've probably already received a lot of feedback on this, but
here's more. Peter Unmack is essentially correct that once genetic
diversity is gone, it's gone, period. The counter-argument that
mutation will recreate lost diversity, especially in a vacuum, is
completely false and based on a false premise. There are several steps
involved to show this, so here goes:

(1) Although it is true that certain sites within DNA sequences are
more prone to mutation than others (e.g. noncoding regions such as those
that forensic DNA techniques make use of, or third-base nucleotides in a
DNA codon (trio or triplet of nucleotides)), most of this variation is
random "noise", and is able to go unchecked specifically because that
particular sequence or portion of a sequence doesn't code for anything.
It's basically "filler" between the informative bits that ARE important
and do code for specific products such as proteins - interestingly
enough, despite all the mutation that does occur on DNA strands, these
sequences are often remarkably slow to change.

(2) There are several steps or layers between changes at the DNA
level and detectable variation for the species. For instance, cheetahs
in Africa which have amongst the least genetic diversity of a species
alive today, probably have DNA mutation rates that are comparable to
those in other mammals. For differences in DNA sequence to be expressed
at some higher level, the change in sequence must have an effect on the
product being coded for (i.e. a different variant). Then, if this
variant doesn't behave differently than the protein that's normally
constructed, it is unlikely to become established in the population
which the mutant is present in.

(3) Many visible traits such as size, colouration, mouth shape, etc.
are coded for by many different genes. For observable differences to
occur, there are probably quite a few differences at the genetic level.