Re: Haldane's Dilemma and quantitative genetics

ErikW writes:

I still don't get the argument for how population structure induces
stasis (if that's what you're saying roughly).

Yes, that is precisely what I'm saying.


And why calculate the
fate of a neutral allele with markov chains instead of simple frequency
based probabilities? It's well known that a new neutral allele has a
miniscule chance of becoming fixed (depends on population size). And of
course a new neutral allele has a hard time dispersing across a gene
flow barrier. Or are you now talking about gene flow barriers at all?
Do you mean to say that all populations look like tessellated maps and
each copy of an allele at a locus occupies one of these triangles or
squares?

I probably need to back up and explain the rationale behind my
statements more fully in order to make my comments more understandable.
Thirty-five years ago, I was quite interested in understanding the
origins of biological complexity, especially intelligence, and I still
am. It was obvious to me from the outset that I was going to get
nothing of interest from standard population genetics models, thus I
took more an ecological approach.

The most fundamental question in ecology at the time was, "what are the
underlying forces that promote species diversity?," which is a closely
related question, and thus I built my simulations on that premise.

The answers that standard ecology had come to regarding the causes of
species diversity were that it was promoted primarily by area,
stability and environmental heterogeneity.

Moreover, it was obvious that taking a gene's-eye view in the
simulations would have been a fundamental philosophical error to
understanding the evolutionary process. Area and spatial heterogeneity
have virtually no input into standard mathematical genetical models,
yet they are of obvious importance to the process. Moreover, behavior
is the quality sifted by natural selection, not the underlying genetics
per se, thus the models must be primarily behaviorally oriented. I
chose finite-state automata as the model organisms to inhabit a map.

I was surprised by how quickly and easily the simple FSA models that I
was generating recapitulated the same basic ecological conclusions:
species diversity, and by consequence biological complexity, were
governed principally by area, heterogeneity and stability. Increasing
seasonal cyclicities, as occur on the Earth due its planetary
obliquity, greatly suppressed the evolution of both species diversities
and complexities.

Anyone who models any form of evolutionary simulation quickly notices
how rapidly evolution comes to a stop. As optimality is approached (at
whatever level of optimality that might be achieved), further
evolutionary optimization is forestalled. But that's simply the nature
of all optimization algorithms. One of my a priori hopes was that
promoting mechanisms that fostered neutral mutations would help prevent
that stalling, especially as the FSA organisms moved into new
environments, thus I was particularly sensitive to observing and
measuring the presence of neutral mutations.

Rather, it became immediately obvious to me that neutral mutations had
almost no chance on a map, for all of the reasons that I outlined
earlier. The process is Gambler's Ruin on steroids, and that greatly
surprised at the time, at least for a few minutes, until I thought
about it for a bit.

But what did truly surprise me was that there were two more factors
that greatly promoted species diversification and complexification:
episodic catastrophes and environmental ambiguities (conflicting
demands that could not be satisified by any single organismal type).

I've let this work lay for about 20 years now, but I've recently begun
to become interested in starting it up again, in part because so many
people are beginning to say similar things nowadays. I've recently
written two proposals to have the work funded, but so far both have
been rejected, thus I'll probably just fund the work myself.

The most recent proposal was to the NASA Institute of Advanced Concepts
(NIAC). 166 proposals were received in the last cycle, but only 7 could
be funded, thus obviously the chances for any one proposal weren't
good. I've put the proposal up at:

http://67.41.4.238/niac-evol.pdf

if you wish to read more.



I'll have to give up for the moment. Perhaps you can show a biological
or population genetic example where the biogeographic map, whatever it
is, hinders neutral evolution and also say why traditional population
genetics fail to predict the outcome in that case.

We either have a mountain of evidence for the phenomenon, or we have
none. I suspect that will depend a little bit on your definition of
published work. I know of no studies that have specifically looked at
the rate of neutral mutations in highly coevolved, stable populations
vis-a-vis those that are expanding into empty ecological spaces (but I
could easily be wrong in that statement).

As for the mountain of evidence, let me quote again just a few
paragraphs of Olivia Judson's comments of a few days ago. They are a
very succinct description of the mass of indirect evidence we've
accumulated, and she does propose an explicit experiment at the end of
her comments that would answer your question. Of interest, her expected
result in bacteria is exactly the one that I get in my FSA:

================================================

Yesterday, I claimed that a major reason large evolutionary changes
often don't happen is that competition from the creatures around you
stops you from changing. In other words, in environments that are
already rich in different species, natural selection often prevents
large changes. My piece of evidence for this was a claim that, when you
take other organisms away - when you reduce competition for food, or
space - evolution explodes. Today, I want to examine the truth of
these claims more closely...

My prediction rests on the fact that it is typically difficult to
evolve to occupy a niche that is already full. An invader must be
better at exploiting the niche than the current occupant, who has
already evolved to make effective use of it. By contrast, when a niche
is empty - when seeds are falling to the ground and no one is eating
them, say - it doesn't matter if, at first, an animal is a bit
inept at finding and opening the seeds.

Consistent with this idea is the observation that when new niches open
up - perhaps because new islands or lakes or cave systems have
formed, or because an asteroid has hit the earth and eradicated
millions of species - the first organisms to become established in
the new environments evolve quickly and reliably into all sorts of new
species. This phenomenon is known as adaptive radiation.

Newly erupted islands are famous for this. Over and over again,
archipelagos see explosive bursts of evolutionary change and the rapid
appearance of species found nowhere else...

Rapid bursts of evolution can also happen in new lakes - which in
many respects are islands too, just islands of water surrounded by
land. Indeed, right now, the great lakes of tropical Africa are the
backdrop for the fastest known radiation of vertebrates, the cichlid
fishes. Lake Victoria, for example, appears to have dried up and then
refilled around 14,600 years ago. Since then, perhaps 500 different
species of cichlid have evolved within it, with all manner of habits.
Lake Victoria has cichlids that eat algae, cichlids that eat other
cichlids, cichlids that eat fish eggs - cichlids, in short, that have
evolved to eat everything that can be eaten. Some fish live in shallow
water; others prefer the deeps. They have evolved a huge diversity of
sexual behaviors, too...

Ideas about adaptive radiation can also be tested in experiments. As I
said yesterday, many bacteria can whiz through hundreds of generations
in a month. This makes it relatively easy to use bacteria to look at
radiations. Here's what you do. You create two sets of environments,
one simple, and one complex. The complex environment might have several
different places to live, or a variety of sources of carbon. The simple
environment has just one habitat or foodstuff. Then, since bacteria
reproduce asexually, you take genetically identical individuals, and
release them into the two different environments. Sure enough,
mutations happen, and the bacteria rapidly evolve to exploit the
different niches. After a month, you will find that bacteria from the
complicated environment have become genetically diverse. Those from the
simple environment, in contrast, remain unevolved.

In short, empty niches are a license for evolutionary change. Once the
new niches are full, natural selection acts to stop further change, and
the rate of evolutionary change slows. Fossils, islands and test tubes
- they all show the same dynamics.

================================================

Wirt Atmar


.

Relevant Pages

  • Lecture of the Week: Why are We So Smart?
    ... with a single living species. ... Although Rick Potts never uses the ... term in this talk on the evolution of humans, ... adaptation to the broader range of environments ...
    (sci.bio.evolution)
  • Re: On a proposal to make a correction to Wikipedias definition of "Evolution".
    ... evolution is the change in the inherited traits of a ... to re-adapt to new environments, or to environmental changes, or to ... populations and species in the same way that diploid species do. ... Richard Harter, cri@xxxxxxxx ...
    (talk.origins)
  • Re: Darwinian Mechanism of Mutation and Natural Selection Found Lacking
    ... who openly question the theory of evolution. ... problem with evolution is that the fossil record does not honestly ... Gould was referring to was stasis in individual species. ...
    (talk.origins)
  • Re: Macroevolution FAQ 2.1D
    ... Macroevolution - its definition, philosophy and history ... "microevolution" is, and by this they seem to mean that whatever evolution is ... evolutionary change at or above the level of species. ...
    (talk.origins)
  • Re: Macroevolution FAQ 2.1D
    ... > Macroevolution - its definition, ... > "microevolution" is, and by this they seem to mean that whatever evolution is ... > evolutionary change at or above the level of species. ...
    (talk.origins)