> At 07:30 AM 7/23/99 -0400, Marcio Pie wrote:
> >Interesting idea. However, current understanding of speciation processes
> >suggests that the most common speciation process (allopatric) involves
> >geographical separation of ancestral populations + differences in
> >selective environment between those populations. Genetic drift may help,
> >but is not necessary. If that is true, the forces generating species
> >formation are *outside* the influence of the genome. How to reconcile this
> >with your scenario?
>
> First, I was dealing with the mathematically and physically real sequences
> space. This is a valid mathematical way to view sequences like DNA. The
> scenario I proposed is not incompatible with selection. Consider this. The
> DNA phase space (which is merely a way of arranging or plotting the DNA of
> an entire population rather than a single individual) is multiply
> connected. Humans have a 3.5 billion dimensional phases space. This means
> that you can move in 3.5 billion different right-angle directions in such a
> space. Say selection pushes the genome in the 2.5 billionth direction and
> there is a path to a new species cavern region along that dimension, then
> the cloud of population points will be taken in that direction totally
> consistent with selection. (I prefer to call the regions of viability
> caverns) and the regions in which the DNA sequence causes death as cavern
> walls.) If there isn't a path to a new species cavern in that direction
> and the selection is strong enough the species might go extinct.
>
> Now, due to certain limitations of population genetics, it is difficult for
> a large population to travel down the small tunnel to the new cavern. Most
> members of a population cloud are not near the tunnel and due to population
> genetics their numerical superiority keep most of the offspring far from
> the tunnel also. This is why it is hard for a large population which
> interbreeds to travel down the tunnel to a new life form.
Glenn
Your scenario has several similarities with Sewall-Wright's "adaptive
landscapes." An adaptive landscape is a representation of a phenotypic
space, where each point has an associated relative fitness value. I don't
know of you are familiar with it, but I'll give an example anyway. Imagine
the relative frequencies of two phenotypes (say, color vision and diurnal
foraging) in a cartesian X-Y plane. Now, plot the relative fitness of each
combination of frequencies of the two phenotypes on a Z axis. Let's say
that when both phenotypes are present (the animal is both diurnal and
has color vision), that combination has a high relative fitness (a peak in
the landscape). However, when both are absent, even though this
combination may have a smaller relative fitness than the first peak, it is
more advantageous that either having only color vision or diurnal
foraging. Now, we can consider the relationships between those two
phenotypes as an adaptive landscape, with two peaks ("color vision +
diurnal foraging" and none), connected by a valley of lower relative
fitness.
Adaptive landscapes have been used to illustrate changes in phenotypes
*within* lineages. It predicts, for instance, that the probability for a
population to go from one peak to another higher peak depends on the
"deepness" of the valley separating them, and can be attained by the
influence of factors such as genetic drift.
However, speciation events are not usually caused by processes such as the
ones depicted in adaptive landscapes. Currently the process believed to be
the most common in generating new species is called allopatric speciation.
What happens is simply a barrier that prevent gene flow between two
populations that previously were interconnected, *regardless* of the
configuration of the adaptive landscapes. Reproductive isolation is simply
the by-product of differences in selective environments, population sizes
(genetic drift) and things like that. The barrier is the most important
factor, otherwise gene flow would homogenize the differences among
populations.
Now, it seems that you are suggesting that the path taken by speciation
events could be previously determined by the genome of the mother species.
It seems to me that speciation events are much more complex than that, and
much of it is determined outside the influence of the DNA itself. If God
has directed speciation events, his influence must be much more complex
than just influencing mutation rates or anything like that. MOreover, if
God did such influences in the lineages leading to us, would it be
possible for us to detect that influence with scientific methods?
Marcio