Actually, quite an interesting article.
Until it is pointed out that he sets his default
population size to 1000. You won't find an actual
population geneticist in the world who won't agree that
with a "bottle-neck" size cohort, one this tiny,
genetic disaster is quite likely.
But that said, the article is still provocative enough
that I went searching the scholarly literature to find
out how actual populations geneticists had reacted, and
see what further research of this nature Sanford's
article has inspired.
Nothing. Nada. Not one journal article; not even one
letters-page commentary.
Perhaps this total vacuum was due to the fact that, as
the entry on Sanford in Wikipedia succinctly put it:
"Sanford's position is rejected by most geneticists and
biologists".
But looking even further, I finally found quite an
interesting blog -- not a journal, a blog -- where
Sanford's model was in fact being discussed. I will
quote from one blogger, whose nym is "Just a theory":
"[...] I find that they have not taken into
account:
1. Neutral mutations - the program classifies
mutations as having some selection coefficient.
Genes are not free to mutate within boundaries
provided that the selection coefficient is zero.
This is in direct contravention to innumerable
papers on genetics, starting with Kimmura's
original one on neutral mutations. The ability for
random mutation to explore neutral sequence space
has been well documented.
2. Linkage - the program classifies genes as
dominant (+) or recessive (-), there are no other
choices. Gene linkage allows harmful genes to piggy
back on beneficial, successful genes.
3. Sexual selection - the program does not simulate
sexual selection at all. Sexual selection allows
harmful mutations (eg. peacock's tail) to
accumulate because they are favourable for
reproductive success.
4. Duplication - the program does not allow for
gene duplication events. Simple thought
experimentation reveals that a duplicated gene is
free to vary provided that the original gene
maintains function. Thus harmful mutations in the
duplicated gene are less harmful overall but
beneficial mutations are still emphasised.
Furthermore, the duplicated gene may assume a novel
function.
[this one, of course, is something I have
pointed out to Andrew roughly one zillion times now
-- NBL]
I could probably go on, others certainly have, but
those 4 are enough for now. The program is
excessively simplistic in its treatment of
evolutionary mechanisms and excludes several
extremely important factors which favour
accumulation of non-harmful mutations. Exclusion of
those factors erroneously leads one to the
conclusion that the genome is deteriorating by the
accumulation of a weight of deleterious mutations".
About the 1000 default population cap, another blogger
resorts to a basic statistical argument, the Drunkard's
Walk, to examine the effects of using a number that
small:
"The main issue is the population cap indeed. The
results are legit for populations with such a low
cap. Think of this way: at the cap of 1000
individuals, the mean number of offspring never
exceed 2 (to go above 2 there'd have to be more
than 1000 individuals). It can be lower though.
This is a drunkard's walk phenomenon, you have a
chance to go lower, but not higher than 2 and at 0
there's an absorbing boundary - extinction. So
ultimately any population will go extinct.
For larger population sizes there are more
deletious mutations, but that doesn't matter -
selection gets more efficient as well and you can
do a lot of maths there. But the drunkard's walk
argument is rather simple and gets the point
across. BTW, the result is legit: All species will
eventually go extinct because there is a population
cap (if only that theres a finite amount of matter
in the universe). But the mean waiting time to
extinction goes up if population sizes are larger
and itll take longer for all life on earth to go
extinct in that way than for the sun to blow up."
Obviously a valid point (albeit taken to the extremes)
NBL