> As if Behe were the first to 'challenge' the molecular clock
> hypothesis. Most competent molecular biologists know that molecular
> are stochastic in nature and that rates of mutation are not uniform. The
> concept of a 'global' clock -i.e., one which has a steady and constant
> mutation rate across both history and organisms, is all but gone.
I'm not sure exactly how you mean "stochastic" in this case (conjectural?
involving chance and probability?), but according to Wen-Hsiung Li and Dan
Graur (_The Fundamentals of Molecular Evolution_, Sinauer Associates,
Massachusetts, 1991, pg. 67-98) molecular clocks are based on the rate of
nucleotide substitution, which is defined as the number of substitutions per
site per year. Substitution rates are known to vary from one region of the
DNA to another, with pseudogenes, four-fold degenerate sits, introns and 3'
flanking regions having the highest rates, and nondegenerate sites having the
lowest. They explain that the rate of substitution is determined by two
factors: the rate of mutation and the probability of fixation of a mutation.
The latter depends upon whether the mutation is advantageous, neutral or
deleterious, and so can vary enormously, sometimes as much as a 1000-fold.
The former is known to vary little within genes and to vary only as much as
2-fold among different genes (at the time of that writing), so in fact the
rate of substitution appears to be determined more by selection intensity
than by mutation rate. As such, while Li and Graur accept that the rate of
mutation is constant (for any one mutational event, not for all events
throughout the genome), they acknowledge that the rate of fixation is
variable, which makes the rate of substitution variable. However, the
variability is based on the probability of a mutation becoming fixed, so
perhaps by stochastic you mean based on probability.
Because of this the controvery over molecular clocks often involves
disagreement over divergence times. Because of that, most molecular
biologists use what is called the relative-rate test, which does not require
knowledge of divergence times. In this test, the rates of two lineages are
compared to a third that supposedly diverged before they did. The number of
substitutions between species 1 and the reference, between species 2 and the
reference, and between species 1 and 2 can be used to determine the number of
substitutions between species 1 and its hypothetical common ancestor with
species 2, between species 2 and the common ancestor, and between the common
ancestor and the reference species. Since by definition the time that has
passed since species 1 and 2 last shared a common ancestor is the same for
both lineages, according to the molecular clock hypothesis the number of
substitutions between species 1 and the common ancestor and between species 2
and the common ancestor should also be the same; that is, if you substract
one value from the other you should get zero. Since mathematically this
difference is equal to the difference obtained by subtracting the number of
substitutions between species 1 and the reference from the number of
substitutions between species 2 and the reference, you can compare the rates
of substitution in species 1 and 2 directly from the comparison of the
species to the reference without dealing with any hypothetical ancestor.
Li and Graur then provide data that shows that mice and rats have very nearly
equal rates of substitution, that Old World monkeys have twice the rate of
humans, and that rodents have a rate some four to six times higher than that
of primates. They then suggest that the higher rates are due to the
generation-time effect, which in essence means that the shorter the
generation time, the higher the substitution rate. Another possibility is
that rodents have a less efficient DNA repair system than primates.
Their conclusion is I believe significant: "The above results should not be
taken as evidence that no molecular clock exists. We note that differences
in rates of substitution are observed between organisms with very different
generation times. When organisms with similar generation times such as mice
and rats are compared, the rate-constancy holds fairly well (see page 82).
Thus, although there is no global clock for the mammals, local clocks may
exist for many groups of relatively closely related species."
So in fact it seems more likely that most competent molecular biologists know
that molecular clocks are probabilistic in nature, and that while the rates
of mutation are constant but not necessarily uniform from one region of the
genome to another, the rates of fixation are neither constant nor uniform.
Yet for closely related species with similar generation times the rates of
substitution are fairly nearly uniform. Therefore, while the concept of a
'global' clock -- i.e., one which has a steady and constant substitution rate
across both history and organisms -- is highly questionable, the existence of
'local" clocks -- i.e., ones which have steady and constant substitution
rates for a few closely related species recently diverged from a common
ancestor -- is more reasonable.
Kevin L. O'Brien