Population size and molecular clock?

Arlin Stoltzfus arlin at is.dal.ca
Thu Dec 31 12:46:44 EST 1998


On 25 Dec 1998, Ludvig Mörtberg wrote:

> Does population size affect the speed with which the clock ticks?
> Assume that the mutations are neutral and must be fixed or lost in a
> population through genetic drift. Any clues? I read a genetics book
> about this but couldn't figure it out.

Do you want to know about the rate of molecular evolution or
the  expected rate of neutral evolution?  These are two different
issues. The "molecular clock" is 
*phenomenological*-- it is based on the notion that the *observed* 
rate of molecular evolution is approximately constant (not everyone 
agrees that this is truly observed, but that is a separate issue).
So, we could answer "yes" to your first question if it were 
observed that the rate of molecular evolution varies with 
population size.  I have no idea whether such an effect has 
been observed.  

Your suggested assumptions indicate a second, theoretical 
issue, one that concerns neutral evolution and is easier to
address. If the rate of neutral mutations is constant, then the rate
of neutral evolution is constant, regardless of population size 
(assuming as always the steady-state condition).  But, the 
rate of neutral mutation is expected to vary with population 
size, other things being equal, because the *definition of 
"neutral" depends on population size*, that is, two alleles 
are neutral if | w1 - w2 | << 1/N.  Thus, the bigger the population, 
the more difficult it is to escape selection, so the smaller the
proportion of changes that are "neutral" (other things being equal).  
A fitness difference of 1 part in a million would be neutral in 
almost any mammal population, but not in a large microbial 
population such as that of E. coli or B. subtilis. 

Arlin





More information about the Mol-evol mailing list