john wilkins

Possibly. The term "network" was used, and I immediately thought of Kaufmann's
regulatory networks, which was misleading so I didn't use it here.

--
John S. Wilkins, Postdoctoral Research Fellow, Biohumanities Project
University of Queensland - Blog: evolvethought.blogspot.com
"Darwin's theory has no more to do with philosophy than any other
hypothesis in natural science." Tractatus 4.1122

tim tyler

Stasis would be the main "stuck on the top of an adaptive peak" prediction.

Therefore, I would nominate most of the contents of:

http://www.nwcreation.net/fossilsliving.html

This page includes the coelacanth, alligators, the cockroach
and velvet worms.

Things like ammonites and trilobites would also probably qualify -
though they are now extinct.
--
__________
|im |yler http://timtyler.org/ tim@tt1lock.org Remove lock to reply.

perplexed in peoria

Gavrilets is certainly not the first to point out that Wright's
metaphor of a landscape can be misleading for reasons that our
geometrical intuitions are poorly suited for spaces of many dimensions.

It seems to me that Gavrilets could not possibly have "proved" that all
landscapes of high dimensionality are inevitably 'connected' (assuming
that the valleys are flooded). For a counter-proof, just raise the
'water level' until the connections are broken.

What he has done is to produce a number of random landscape models
which seem to match pretty well with biological intuitions as to what
a landscape should look like in some ways, but which also contradict
our intuitions in other ways. This is valuable in that it shows that
our intuitions are not always as self-supporting as we might think.

Here is an online copy of a TREE review article that provides a nice
introduction to his work.
http://www.tiem.utk.edu/%7Egavrila/PAPS/tree.pdf

And, for the more adventuresome, here are a lot more of his publications: http://www.tiem.utk.edu/~gavrila/Publications.htm


I suspect not. Unless you mean adaptation vs drift as a cause of
speciation. It may be relevant to that question.

carrie diazeaton

I am a student of Mathematics and Evolutionary biology, so I hope I can
clarify. As a couple of other postings pointed out, this result is
connected to "holey adaptive landscapes" or "neutral networks." These
are "nearly neutral networks," and the results used to talk about
neutral fitness landscapes were actually first used in the context of
physics.

The problem addressed is that for most people a visualization of the
adaptive landscape includes many mountains, representing potential fit
genotypes, but a population must go downhill through a valley (work
against selection) in order to get to another mountain (ie.
diverge/speciate). This work against selection has been an argument
used by critics of the adaptive landscape framework.

What he is saying is that you still climb out out valleys at a rate
dictated by mutations-selection-drift balance, and once you are on a
ridge, you will not fall off because of the same M-S-D balance.
However, when dimensionalities are high, (when you consider large
numbers of loci, 0<L<64), there are extensive networks of these ridges,
these are not flat ridges, but ridges whose fitness is within [1-delta,
1]. On that ridge, nearly neutral mutations (they don't have to be
exactly neutral) and drift are "sufficient for explaining continuous
evolutionary changes," which can lead to discussions on the feasibility
of allopatric and allochronic speciation and diversification. This
work supports other work by Ohta and Kimura.

Does that help at all?

Carrie Diaz Eaton

guy hoelzer

in article dj6jvp$ti2$1@darwin.ediacara.org, Larry Moran at


The context of this thread and my post was that of a holey adaptive
landscape, which does not have any peaks.

With regard to an exemplar species, I would say you can pick any species you
want. The metabolic functioning of any cellular organism is so complex and
unlikely from a statistical point of view that I expect them all to exist on
the high plateau of a holey adaptive landscape after billions of years of evolutionary history.


Humans are an extremely social species, which amplifies the kind of
coevolutionary dynamics I described in my post. This leads to the constant
perception of "room for improvement", although it is little more than the
cat making strategic decisions about what moves would maximize its chances
of catching its tail. The "room for improvement" is largely an illusion,
IMHO, as it relates to evolutionary fitness.


You are dichotomizing the options in a way that I can't accept. My view is
consistent with the one argued by Per Bak in "How Nature Works", which he
attributed to all self-organizing systems. My expectation is that such
systems approach perfection, but never reach it. Bak's argument is entirely
consistent with the holey adaptive landscape paradigm of Kaufmann and
Gavrilets.

It seems you are in the unusual position (for you) of defending the
importance of natural selection as a sub-process of evolution. In Bak's
view it is essential that self-organizing systems include stochastic
phenomena (externally derived perturbations and/or internally generated) in
order to explore alternative paths toward perfection. In the context of
genetic evolution we have mutation and drift filling this crucial role. The
irony is that stochasticity is essential for approaching perfection (think
Sewall Wright and shifting balance here), but it also prevents the system from attaining perfection.

You are picturing a landscape topology that differs from the holey adaptive
landscape. IMHO, Kaufmann and Gavrilets have made a big advance in the
theory of adaptive landscapes by considering the evolution of landscape
topology, rather than merely thinking of them as assumptions or mere
snapshots of what might be. There conclusion is that adaptive landscapes
will inevitably evolve toward being like high plateau's pitted with deep
holes. In that absence of an alternative theory predicting other sorts of
landscape topologies, I think it behooves you to argue for the plausibility
of other sorts of topologies before assuming them in the context of this thread.


Well, I certainly haven't tried to explain the theory predicting that
adaptive landscapes will have the topology described above, but the theory
is well-established in the literature. I have not just picked this shape
out of the blue, assumed it to be true, and based my argument on such flimsy
grounds. I think that Gavrilets did a nice job introducing the theory in:

Gavrilets, S. 1999. A Dynamical Theory of Speciation on Holey
Adaptive Landscapes. Am. Nat. 1999. Vol. 154, pp. 1*22.

Note that he won the award for the best paper published in Am Nat in 1999
for this paper. The basis of my arguments is not exactly on the fringe.

You snipped the part of my post where I described the role of "positive
selection" in this scenario. It is derived from co-evolutionary
interactions. Note that the conclusion you drew is fully consistent with
Kimura's Neutral Theory, so I would argue again that this does not place my
argument far from conventional thinking in evolutionary biology.

Cheers,

Guy

john wilkins

Beautifully, and places it in a context (Ohta and Kimura) that I can now see
the relevance of. Thanks. Thanks.

--
John S. Wilkins, Postdoctoral Research Fellow, Biohumanities Project
University of Queensland - Blog: evolvethought.blogspot.com
"Darwin's theory has no more to do with philosophy than any other
hypothesis in natural science." Tractatus 4.1122

william morse

And you can add horseshoe crabs to the list.

But I think perhaps there is a middle ground here, abandoning the concept
of "so well adapted that there is no room for improvement" for the
concept of "well enough adapted that the likelihood of an improvement
appearing in the next million years is not very high, especially since
the environment is highly variable". Guy I believe is talking about the
latter. Tim has given examples of species that may be stuck on an
adaptive peak. That still doesn't mean there is no room for improvement
in them, it only means they are stuck on an adaptive peak. Larry's
question equates "well adapted" with "at an adaptive peak",but there is
no particular reason why the two should be related.

Yours,

Bill Morse

erikw

That's a good point. If anywhere Gavrilets would have had to take this
perspective into consideration in his book. I don't suspect anyone here
has read it?
Myself, I am reluctant to buy it for fear of getting two understandable
chapters and the rest of it being hoplessly difficult.

guy hoelzer

in article dj91vc$1tev$1@darwin.ediacara.org, carrie diazeaton at

Thank you Carrie. I am curious who you are working with and what is the
subject of your graduate research?

Regards,

Guy

guy hoelzer

in article dj91v8$1t92$1@darwin.ediacara.org, William Morse at


I don't see an obvious connection, and I don't recall seeing it addressed by
others. IMHO, greater biomass and greater biodiversity in the tropics is
related to the larger amount of solar energy supplied to those regions.

Guy