Dawkins on Kimura

[John Edser]

Larry Moran <[email protected]> wrote or quoted:
> I can understand your "feeling." Since we don't have very good evidence one way of the other is it
> valid science to rely on "feeling"? Why are you asking for experimental evidence against your
> feeling? I could just as easily state that most features are due to drift and it's up to you to
> prove otherwise. Would that be an example of good science?
>
> Why don't we just say that morphological features can become fixed in a population by adaptation -
> and quote some known expamples. We don't know if *most* morphological features are due to
> adaptation or whether they are accidents of evolution.

TT:- Simon Conway Morris's latest book bears on this question:

``Life's Solution : Inevitable Humans in a Lonely Universe''

It suggests that evolution has designed functionally the *same* structure from *different* starting
points - in a very large number of cases.

He gives quite a few good examples of this convergent evolution.

It is easily explicable on the assumption that "most features" of organisms were adaptations - but
would be puzzling if they were accidents.

JE:- Why is it impossible for gene centric Neo Darwinists to understand that the probable _rate_ of
"accidents" such as sampling error (genetic drift) can be selected at the Darwinian fertile organism
level of selection?

Respectfully,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]

 

[Dkomo]

Jeffrey Turner wrote:
>
> dkomo wrote:
> > Jeffrey Turner wrote:
> >>Tim Tyler wrote:
> >>
> >>
> >>>Drift is most effective when population sizes are small. Selection is most effective when
> >>>population sizes are large. I reckon this fact (in conjunction with nature's population sizes)
> >>>will often limit's drift's usefulness as an explanation for features of organisms.
> >>
> >>What is "drift"? Genes mutate randomly, malign mutations die out, mutations that have no effect
> >>on an organism don't really matter and beneficial mutations propagate and supplant organisms
> >>without that mutation. This generally leads to a new species, in time. But what's "drift"?
> >
> >
> >>From a post to talk.origins on Sept 5, 2002:
> >
> > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> > Subject: Neutral evolution simulation Date: Thu, 05 Sep 2002 13:21:25 -0600 From: dkomo
> > <[email protected]> Newsgroups: talk.orig
> >
> > I ran across an interesting thought experiment that illustrates how neutral genes in a small
> > population can change frequencies due solely to sampling error/genetic drift. In other words,
> > the population can evolve even in the absence of any selection pressure.
> >
> > Consider a jar full of 45 red marbles and 55 blue marbles. These represent the blue and red
> > alleles of a neutral gene. To simulate the fact that only a certain fraction of our marble
> > "organisms" will be able to reproduce (the rest will die or be unable to reproduce because of
> > lack of food or water, disease, etc.) select 50 marbles at random and throw them away. The
> > remaining marbles in the jar are allowed to reproduce by creating a copy of their allele,
> > returning the size of the population back to 100 marbles.
> >
> > For example, let's say that in the 50 marbles selected, 20 were red and 30 were blue. The
> > remaining 25 red marbles in the jar reproduce and so do the remaining 25 blue marbles. The jar
> > now contains 50 red marbles and 50 blue marbles.
> >
> > Repeat this experiment 100 times. At the end of the trials, which of these is correct:
> >
> > 1. The jar will contain 100 red marbles.
> > 2. The jar will contain 100 blue marbles.
> > 3. The jar will contain a mix of red and blue marbles in approximately the same ratio as it had
> > initially.
> > 4. It's not possible to predict the marble ratios.
> > 5. None of the above.
> > 6. I don't know.
> > 7. Who cares?
> >
> > Note that if the number of either the red or blue marbles reaches 100, no further change is
> > possible because "mutations" are disallowed (red->blue, or blue->red during reproduction).
>
> Off-hand, I'd say 3 is probably correct. There is a finite probability of any other distribution,
> however.

The answer is (5) and is given in the thread in which the original post appeared:

http://tinyurl.com/2crm3

To summarize what appeared there, I wrote:

"The answer is (5) above. (1) and (2) are each possible, with probabilities 0.198 and 0.287, as
Chris Ho Stuart pointed out. There is also a small chance of (3). (4) is clearly incorrect because
it *is* possible to make a statistical prediction of the distribution of marbles after 100
experiments.

There was a variety of interesting responses in this thread. The main thing I wanted to point out
with this example is that a statistical process can drive an allele distribution in a population
either to fixation, where the frequency of the allele becomes 100%, or to extinction, where it
disappears from the population. And such a process is independent of any selection pressure -- that
is, the selection coefficient of the allele is near zero.

In particular, if the jar starts out with 50 red and 50 blue marbles, it will *inevitably* end up
with either 100 red marbles or 100 blue after a sufficiently large number of "generations." This can
be proven mathematically. The probability of either case is 0.50. I found this to be a fascinating
counterintuitive result. The intuitive situation would be a aimless drifting around the equilibrium
point of 50 red and 50 blue.

The other interesting result is what happens to a *single* mutation,
e.g. a single red marble in a population of 99 blue marbles. It has only a 1% chance of reaching
fixation and a 99% chance of being eliminated. Thus, only 1 out of every 100 neutral mutations
can stabilize within the population."

Mark VandeWettering computed the results using monte carlo simulation. Chris Ho Stuart computed it
analytically, but the calculation was so laborious he needed to write a computer program to derive
the final probabilities. His 2nd post in the thread contains the C code for his program.

> Is one hundred individuals a realistic population?

According to what people have stated here on sbe, subpopulations this small or smaller are the
*norm* for genetics.

> It seems very small, and the odds of any appreciable "drift" decrease dramatically as population
> size increases.

Yes, and I'd like to know exactly how fast it decreases as population size increases. I may write
my own monte carlo simulation. The discussion of drift on sbe has thus far been too qualitative
and vague.

> So, it seems to me, drift may be conceptualizable but it won't be a significant phenomenon.

Not if small subpopulations are the dominant reality in population genetics.

> As for how any significant "deviant" population would arise in the first place without selection,
> it must either have arisen by mutation or been inherited early in the history of a species _or_
> occurred as a fairly regular mutation in a well established species.
>
> From an evolutionary standpoint, the most realistic answer is 7.
>

That's what the debate here is all about.

[email protected]

 

[Tim Tyler]

John Edser <[email protected]> wrote or quoted:
> Tim Tyler wrote:
>> Larry Moran <[email protected]> wrote or quoted:

>> > Why don't we just say that morphological features can become fixed in a population by
>> > adaptation - and quote some known expamples. We don't know if *most* morphological features are
>> > due to adaptation or whether they are accidents of evolution.
>>
>> Simon Conway Morris's latest book bears on this question:
>>
>> ``Life's Solution : Inevitable Humans in a Lonely Universe''
>>
>> It suggests that evolution has designed functionally the *same* structure from *different*
>> starting points - in a very large number of cases.
>>
>> He gives quite a few good examples of this convergent evolution.
>>
>> It is easily explicable on the assumption that "most features" of organisms were adaptations -
>> but would be puzzling if they were accidents.
>
> Why is it impossible for gene centric Neo Darwinists to understand that the probable _rate_ of
> "accidents" such as sampling error (genetic drift) can be selected at the Darwinian fertile
> organism level of selection?

Organisms can alter their mutation rate by the degree of investment they put into insulation, gene
repair, cell house-keeping - and so on. Normally they try to make mutation levels low - but can't
necessarily afford the budget to do it really effectively.

...but what does that have to do with the issue?

Surely there is no argument that the supposed examples of genetic drift are /really/ adaptations -
because organisms have some control over their own mutation rates ;-)

I do think homoplasy is significant.

It may not allow us to say much about the relative significance of adaptation and drift - but it
does mean that we can point to some specific organism features and say - with a reasonable level of
confidence - this as an adaptation - and so is this - and so is this - without being accused of
concocting "just so" stories.

Morris's purpose is not illumating this debate - but some of the most striking illustrations of
homoplasy I've seen are to be found in his book.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[Tim Tyler]

dkomo <[email protected]> wrote or quoted:
> Jeffrey Turner wrote:
> > dkomo wrote:
> > > Jeffrey Turner wrote:
> > >>Tim Tyler wrote:

> > >>>Drift is most effective when population sizes are small. Selection is most effective when
> > >>>population sizes are large. I reckon this fact (in conjunction with nature's population
> > >>>sizes) will often limit's drift's usefulness as an explanation for features of organisms.

[...]

> > > 1. The jar will contain 100 red marbles.
> > > 2. The jar will contain 100 blue marbles.
> > > 3. The jar will contain a mix of red and blue marbles in approximately the same ratio as it
> > > had initially.
> > > 4. It's not possible to predict the marble ratios.
> > > 5. None of the above.
> > > 6. I don't know.
> > > 7. Who cares?
> > >
> > > Note that if the number of either the red or blue marbles reaches 100, no further change is
> > > possible because "mutations" are disallowed (red->blue, or blue->red during reproduction).
> >
> > Off-hand, I'd say 3 is probably correct. There is a finite probability of any other
> > distribution, however.
>
> The answer is (5) and is given in the thread in which the original post appeared:
>
> http://tinyurl.com/2crm3

[...]

> The main thing I wanted to point out with this example is that a statistical process can drive an
> allele distribution in a population either to fixation, where the frequency of the allele becomes
> 100%, or to extinction, where it disappears from the population. And such a process is independent
> of any selection pressure -- that is, the selection coefficient of the allele is near zero.

I think you mean "*IF* the selection coefficient of the allele is near zero" Large selection
pressures are not irrelevant - and would change the answer to the problem from 5 to 1 or 2.

> > Is one hundred individuals a realistic population?
>
> According to what people have stated here on sbe, subpopulations this small or smaller are the
> *norm* for genetics.

For the genetics of small populations on the verge of extiction?

The quantitly of relevance of the calculaton is known as the "effective populaton size".

An effective population size of 500 is considered to be a threshold for status as an
endangered species.

The figures you are talking about (in isolated populations) would represent effective population
sizes of 25-50 - putting them straight into the red, flashing region of the endangered species list.

It is true that we are in the middle of a mass extinction at the momement
- and so it's inevitable that more than the usual number of species are going through small
population sizes as part of the process of shuffling off this mortal coil.

However, the population sizes you talk about still seem to be very low.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[John Edser]

> JE:- Why is it impossible for gene centric Neo Darwinists to understand that the probable _rate_
> of "accidents" such as sampling error (genetic drift) can be selected at the Darwinian fertile
> organism level of selection?

TT:- Organisms can alter their mutation rate by the degree of investment they put into insulation,
gene repair, cell house-keeping - and so on. Normally they try to make mutation levels low - but
can't necessarily afford the budget to do it really effectively.

JE:- My information is that loci mutation rates are not consistent. The implication is that nature
has selected to reduce mutation rates at _some_ loci but allow it to increase at others, i.e.
selection can favour more or less variation at different loci so that mutation _rates_ are under
selective control.

TT:- ...but what does that have to do with the issue?

JE:- As you may know I disallow drift as valid "evolution" but allow it as "temporal variation"
because evolutionary theory becomes non testable if random gene freq. change is allowed as evolution
but remains testable when allowed as temporal variation.

Just as selection can control the rate of mutation at different loci and thus control the rate of
loci variation, nature can control the rate of temporal variation. Explaining drift as temporal
variation allows drift to be used as a a powerful view for the adaptive argument and not within just
a non adaptive argument. Drift is misused as just a competing force _against_ Darwinian natural
selection.

TT:- Surely there is no argument that the supposed examples of genetic drift are /really/
adaptations - because organisms have some control over their own mutation rates ;-)

JE:- I am arguing that "examples of genetic drift are /really/ adaptations " in the general but not
the specific, sense, i.e. nature cannot predict the outcome of specific random events but she can
increase or decrease them where needed. Forms that optimise variation can have a higher fitness.

TT:- I do think homoplasy is significant. It may not allow us to say much about the relative
significance of adaptation and drift - but it does mean that we can point to some specific organism
features and say - with a reasonable level of confidence - this as an adaptation - and so is this -
and so is this - without being accused of concocting "just so" stories.

JE:- So called "just so" stories are IMO maligned by the "drift is evolution" school because "just
so" stories have been shown in the past to be valuable and testable hypothesis. What remains non
testable is the common "drift is evolution" concept.

Regards,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]

 

[Jeffrey Turner]

Larry Moran wrote:
> William Morse <[email protected]> wrote:
>
>
>>If the trait shows little variation throughout a large population or is strongly correlated with
>>an obvious environmental variable, and has an obvious relation to an aspect of species behavior,
>>one can make the default assumption that it is an adaptation, especially if it has existed for a
>>long period.(Again based on recollection, Wirt did a much better job than I just did of defining
>>what traits could be considered adaptive as a default). Examples would include large ears in
>>elephants, skin color in humans, and almost any morphological feature of horseshoe crabs.
>>
>>If the trait shows wide variation throughout a population regardless of environmental variables,
>>or is confined to isolated subpopulations with no obvious relation to fitness, one can make the
>>default assumption that it is due to drift. Examples include coat color in domestic cats and the
>>Rh- blood type in humans.
>
> Hmmmm .... I would have thought that ear size in elephants and skin color in humans were excellent
> example of drift.

So where are all the small-eared elephants?

> Don't you see how difficult it is to make general rules? Your decision about what the "default"
> hypothesis should be depends to a great extent on your original biases. This is exactly the point
> that Lewontin and Gould made in their original paper. If you tend to emphasize natural selection
> in your thinking about evolution then you will look to adaptive explanations ahead of non-adaptive
> explanations.

Considering genetic diversity, itself, gives a species an advantage, you're going to have to explain
why traits would develop for non-adaptive reasons.

>>Note that I am cheating somewhat, as I am assuming there is at least some data available to make a
>>"default" assumption.
>
> Part of the problem is that even the selection of "data" has built-in biases. If you are an
> adaptionist, and you prefer an adaptionist explanation, then you go looking for data to support
> your assumption. For example, if you think human skin color is adaptive then the only "data" you
> quote is the little bit that might explains why people with white skin lived in Northern Europe.
>
>
>>Also, the above _are_ only default assumptions - either may be proved false by better data, so
>>ultimately I agree with Bob. Also note that there will be a lot of traits - morphological features
>>and otherwise - that will not fall into either category. In this case the adaptationists
>>(including myself) will dream up unsubstantiated Just So stories to which the drifters (sorry -
>>the term has both a nice slightly pejorative ring to my ears as well as reminding me of the 60's
>>rock band) will recoil in horror
>
>
> No offense. I'm a drifter in the sense that it's part of my self-proclaimed "mission" to educate
> people about the importance of random genetic drift

Genetic drift isn't very interesting, changes in phenotype are all that evolution can work with.

> and evolution by accident.

Evolution by accident? Can you explain?

> I'm not having much luck on sci.bio.evolution. This newsgroup is heavily dominated by people who
> reject the very concept of drift or who have deliberately chosen not to understand it. I find
> this very strange in a newsgroup that's supposed to be devoted to discussing evolution at a
> serious level.

Maybe because "drift" doesn't seem to be very well defined.

--Jeff

--
A man, a plan, a cat, a canal - Panama!

Ho, ho, ho, hee, hee, hee and a couple of ha, ha, has; That's how we pass the day away, in the merry
old land of Oz.

 

[Guy Hoelzer]

in article [email protected], R.Schenck at
[email protected] wrote on 2/29/04 10:34 AM:

>> I am saying something roughly like your last alternative. I think that the evidence is now
>> overwhelming, and mostly overlooked by evolutionary biologists, that natural selection is a
>> particular form of more general (thermodynamic) set of optimizing, adaptive processes. Biology
>> specializes on natural selection as an adaptive process, but there is no reason that others kinds
>> of adaptive mechanisms can't also participate in causing adaptive evolution.
> i am curious about this, what other mechanisms lead to adaption besides nat select. ?

"Learning" is an example familiar to biologists. I think that "entrainment" is a more general
physical example.

> I understand that other mechanims are involved in evolution, but not that they necessarily lead to
> adapatations, whereas natural selection 'selects' for adapatations. So what are these other
> mechanisms that select adapations?

To keep the discussion centered on biological evolution, you should consider the "Baldwin Effect",
as suggested by Tim Tyler.

> I guess i don't expect a full explanation here, but what texts go into it at least? I probably
> shouldn't be asking for -more- stuff to read now, but at least i could add it to the list.

I don't think that any of the standard textbooks on evolutionary biology deal with these issues.
IMHO this is a problem. You might want to add "Darwinism Evolving" (Depew and Weber) or "Signs of
Life" (Sole and Goodwin) to your list. Also, you will find tons of material on the Baldwin Effect
with a Google search.

Cheers,

Guy

 

[Dkomo]

Tim Tyler wrote:
>
> dkomo <[email protected]> wrote or quoted:
> > Jeffrey Turner wrote:
> > > dkomo wrote:
> > > > Jeffrey Turner wrote:
> > > >>Tim Tyler wrote:
>
> > > >>>Drift is most effective when population sizes are small. Selection is most effective when
> > > >>>population sizes are large. I reckon this fact (in conjunction with nature's population
> > > >>>sizes) will often limit's drift's usefulness as an explanation for features of organisms.
>
> [...]
>
> > > > 1. The jar will contain 100 red marbles.
> > > > 2. The jar will contain 100 blue marbles.
> > > > 3. The jar will contain a mix of red and blue marbles in approximately the same ratio as it
> > > > had initially.
> > > > 4. It's not possible to predict the marble ratios.
> > > > 5. None of the above.
> > > > 6. I don't know.
> > > > 7. Who cares?
> > > >
> > > > Note that if the number of either the red or blue marbles reaches 100, no further change is
> > > > possible because "mutations" are disallowed (red->blue, or blue->red during reproduction).
> > >
> > > Off-hand, I'd say 3 is probably correct. There is a finite probability of any other
> > > distribution, however.
> >
> > The answer is (5) and is given in the thread in which the original post appeared:
> >
> > http://tinyurl.com/2crm3
>
> [...]
>
> > The main thing I wanted to point out with this example is that a statistical process can drive
> > an allele distribution in a population either to fixation, where the frequency of the allele
> > becomes 100%, or to extinction, where it disappears from the population. And such a process is
> > independent of any selection pressure -- that is, the selection coefficient of the allele is
> > near zero.
>
> I think you mean "*IF* the selection coefficient of the allele is near zero" Large selection
> pressures are not irrelevant - and would change the answer to the problem from 5 to 1 or 2.
>

What I meant is that the statistical process alone can strongly affect the allele frequencies and
that can happen in the absence of any selection pressure, which is when the selection coefficient of
the allele is effectively zero.

Even if there is selection pressure, the drift component still operates independently. The final
allele distribution will be a result of the sum of both processes.

> > > Is one hundred individuals a realistic population?
> >
> > According to what people have stated here on sbe, subpopulations this small or smaller are the
> > *norm* for genetics.
>
> For the genetics of small populations on the verge of extiction?
>
> The quantitly of relevance of the calculaton is known as the "effective populaton size".
>
> An effective population size of 500 is considered to be a threshold for status as an endangered
> species.
>
> The figures you are talking about (in isolated populations) would represent effective
> population sizes of 25-50 - putting them straight into the red, flashing region of the
> endangered species list.
>
> It is true that we are in the middle of a mass extinction at the momement
> - and so it's inevitable that more than the usual number of species are going through small
> population sizes as part of the process of shuffling off this mortal coil.
>
> However, the population sizes you talk about still seem to be very low.
>

What I said about small populations had nothing to do with endangered species. I was referring to
the following statements here in this thread:

dkomo:

It may be that many features of organisms are incidental and play no role in their evolution. That
doesn't necessarily imply that these features are subject to drift. Drift is a phenomenon of small,
isolated populations.

Anon:

And there are a lot of them about, even in organisms like insects.

Larry Moran:

Exactly right. In the real world it's almost impossible to have a large population of randomly
mating individuals (i.e., panmitic). Almost any population that you can think of is subdivided into
a large number of much smaller sub-populations that are more-or-less genetically isolated from each
other. This is obvious in humans but it's true of every other species as well.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++

So if it is common to have genetically isolated small sub-populations of large populations of
animals, the phenomenon of genetic drift is going to be dominant and very widespread. And that is
going to blow a huge hole into the side of theories that most features of organisms are adaptations.

[email protected]

 

[Tim Tyler]

Jeffrey Turner <[email protected]> wrote or quoted:

> Evolution by accident? Can you explain?

Check out the differences between mammals and marsupials.

[what? Marsupials aren't mammals? - JAH]

To /some/ extent they exhibit convergent evolution - with similar forms occupying similar niches.

However in other areas, they show significant differences.

Attributing these differences to selection caused by a different environment is /possible/ - but
mammals seem to do OK in Australia, despite all the unusual and unfamiliar species over there - so
the environment probably isn't /that/ different.

The most obvious explanation of the differences is chance. Marsupials *happened* to do things a
different way - simply since there's more than one way to do things - and the differences became
fixed over time.

It /looks/ like the marsupials were better at *spreading* initailly - but that they were not so good
at defending their turf.

A similar thing - perhaps on an even more dramatic scale - happened in South America 3 million
years ago:

``In the upper Pliocene, 3 million years ago, the isthmus of Panama reappeared as a result of
changes in the earth's crust. This was a disaster for many of the animals that had evolved in
isolation in South America. South America was invaded by deer, camels, raccoons, tapirs, horses,
mastodons, bears, peccaries, rabbits, shrews, cats, dogs, weasels and rodents. For some reason these
animals were able to displace many of the South American species, driving many of them to
extinction.''

- http://darwin.bio.uci.edu/~sustain/bio65/lec02/b65lec02.htm#SouthAm

While divergence can be caused by drift, such large scale displacements of whole species are
examples of selection.

Iu /often/ seems to be the smaller population that suffers when populations join.

Perhaps there are fewer different sorts of competitor in the smaller environment - and so
less pressure.

Or perhaps the smaller environment is likely to be less well furnished with specialised species in
the first place - so when existing specialists do arrive they are simply better equipped than the
natives (who have more recently adopted the roles) to exploit the available niches.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[Larry Moran]

On Tue, 2 Mar 2004 04:48:40 +0000 (UTC),
Jeffrey Turner <[email protected]> wrote:
> Larry Moran wrote:
>> William Morse <[email protected]> wrote:
>>
>>>If the trait shows little variation throughout a large population or is strongly correlated with
>>>an obvious environmental variable, and has an obvious relation to an aspect of species behavior,
>>>one can make the default assumption that it is an adaptation, especially if it has existed for a
>>>long period. (Again based on recollection, Wirt did a much better job than I just did of defining
>>>what traits could be considered adaptive as a default). Examples would include large ears in
>>>elephants, skin color in humans, and almost any morphological feature of horseshoe crabs.
>>>
>>>If the trait shows wide variation throughout a population regardless of environmental variables,
>>>or is confined to isolated subpopulations with no obvious relation to fitness, one can make the
>>>default assumption that it is due to drift. Examples include coat color in domestic cats and the
>>>Rh- blood type in humans.
>>
>> Hmmmm .... I would have thought that ear size in elephants and skin color in humans were
>> excellent example of drift.
>
> So where are all the small-eared elephants?

Hmmmm ... I see where you're coming from. You observe that all modern species of elphants have
bigger ears than their ancestors and their modern cousins such as manatees, dugongs, and hyraxes.
This leads you to the conclusion that large ears are an adaptation. However, if the ancestors of
modern elephants just happened to have large ears, and large ears were not a disadvantage, then the
presence of large ears could just as easily be due to drift. The fact that modern species possess a
certain characteristic feature is not prima facie evidence of adaptation since drift and the founder
effect would achieve the same result.

But that's not what I had in mind. I was referring to the fact that Asian elephants have much
smaller ears than African elephants. If ear size is under strong selection then one has to account
for this fact. On the other hand, if ear size is non-adaptive then the difference between Asian and
African elephants could be due entirely to chance.

>> Don't you see how difficult it is to make general rules? Your decision about what the "default"
>> hypothesis should be depends to a great extent on your original biases. This is exactly the point
>> that Lewontin and Gould made in their original paper. If you tend to emphasize natural selection
>> in your thinking about evolution then you will look to adaptive explanations ahead of non-
>> adaptive explanations.
>
> Considering genetic diversity, itself, gives a species an advantage, you're going to have to
> explain why traits would develop for non-adaptive reasons.

Two points ...

1. Genetic diversity cannot be an adaptation since this requires a form of group selection that has
been thoroughly discredited. If a species accidently possesses more diversity then it will be the
lucky survivor when the environment changes. This is more like evolution by chance that real
adaptation.

2. All kinds of traits can arise by chance. Take the ability of some people to roll their tongues as
a simple example. Why do I have to explain why traits would develop for non-adaptive reasons? Is
it because of your (irrational) belief that everything MUST be an adaptation?

[snip]

>> No offense. I'm a drifter in the sense that it's part of my self-proclaimed "mission" to educate
>> people about the importance of random genetic drift
>
> Genetic drift isn't very interesting, changes in phenotype are all that evolution can work with.

Two points ....

3. The mechanism of random genetic drift is the basic evolutionary mechanism underlying most of
molecular evolution. I find this very interesting. You may not. Furthermore, I'm quite interested
in the organization of genomes and junk DNA. Part of the modern explanation of genome evolution
requries random genetic drift. You may not be interested in this either. Finally, I'm really
interested in correct explanations of the main features of modern species (and speciation in
general). If you aren't interested in random genetic drift then you mustn't be interested in
those things either. It's okay to focus all of your attention on adaptations and to admit that
nothing else about evolution interests you. But please don't assume that others share your bias.

4. The most reasonable minimal definition of evolution is change in the frequency of heritable
characteristics in a population over time. This definition includes fixation by random genetic
drift of almost neutral alleles. You seem to want to change that definition to restrict it to
changes in phenotype - where presumably you mean only observable phenotypes. Furthermore, when
you say that this is "all that evolution can work" with you imply that evolution is synonomous
with natural selection. In other words, you would like to change the definition of evolution so
that the only "real" evolution is adaptation. This is consistent with the fact that you aren't
interested in any other kind of evolution but here you go one step farther. Now you are
suggesting that we re-define evolution so that my interests become illegitimate. If you succeed
(highly unlikely) then what word will you use for all those heritable changes that aren't
adaptations? Will this be pseudoevolution?

>> and evolution by accident.
>
> Evolution by accident? Can you explain?

Yes, I'm referring to evolution that takes place without natural selection playing a significnat
role. It includes things such as the random elimination of some species, and the survival of others,
at the time of mass extinctions. (The so-called "Field of Bullets" senario of David Raup.) It
includes lots of other things as well. When I use the phrase "evolution by accident" I mean it to be
a direct contrast to adaptionism. The phrase is intended to provoke people into thinking outside
their box. (It's also a convenient way to emphasize the contrast between real evolution and
intelligent design.)

>> I'm not having much luck on sci.bio.evolution. This newsgroup is heavily dominated by people who
>> reject the very concept of drift or who have deliberately chosen not to understand it. I find
>> this very strange in a newsgroup that's supposed to be devoted to discussing evolution at a
>> serious level.
>
> Maybe because "drift" doesn't seem to be very well defined.

Not true. It's as well-defined as natural selection. It may be a more difficult concept to grasp and
it may be unfamiliar to those who have been brought up on a steady diet of adaptionist dogma, but
that's a different kettle of fish. You can't dismiss random genetic drift just because you don't
understand it.

Larry Moran

 

[Anon.]

Tim Tyler wrote:
> dkomo <[email protected]> wrote or quoted:
>
>>Jeffrey Turner wrote:
>>
>>>dkomo wrote:
>>>
>>>>Jeffrey Turner wrote:
>>>>
>>>>>Tim Tyler wrote:
>>>>
>
>>>>>>Drift is most effective when population sizes are small. Selection is most effective when
>>>>>>population sizes are large. I reckon this fact (in conjunction with nature's population sizes)
>>>>>>will often limit's drift's usefulness as an explanation for features of organisms.
>>>>>
>
> [...]
>
>
>>>>1. The jar will contain 100 red marbles.
>>>>2. The jar will contain 100 blue marbles.
>>>>3. The jar will contain a mix of red and blue marbles in approximately the same ratio as it had
>>>> initially.
>>>>4. It's not possible to predict the marble ratios.
>>>>5. None of the above.
>>>>6. I don't know.
>>>>7. Who cares?
>>>>
>>>>Note that if the number of either the red or blue marbles reaches 100, no further change is
>>>>possible because "mutations" are disallowed (red->blue, or blue->red during reproduction).
>>>
>>>Off-hand, I'd say 3 is probably correct. There is a finite probability of any other distribution,
>>>however.
>>
>>The answer is (5) and is given in the thread in which the original post appeared:
>>
>>http://tinyurl.com/2crm3
>
>
> [...]
>
>
>>The main thing I wanted to point out with this example is that a statistical process can drive an
>>allele distribution in a population either to fixation, where the frequency of the allele becomes
>>100%, or to extinction, where it disappears from the population. And such a process is independent
>>of any selection pressure -- that is, the selection coefficient of the allele is near zero.
>
>
> I think you mean "*IF* the selection coefficient of the allele is near zero" Large selection
> pressures are not irrelevant - and would change the answer to the problem from 5 to 1 or 2.
>
>
>>>Is one hundred individuals a realistic population?
>>
>>According to what people have stated here on sbe, subpopulations this small or smaller are the
>>*norm* for genetics.
>
>
> For the genetics of small populations on the verge of extiction?
>
> The quantitly of relevance of the calculaton is known as the "effective populaton size".
>
> An effective population size of 500 is considered to be a threshold for status as an endangered
> species.
>
> The figures you are talking about (in isolated populations) would represent effective
> population sizes of 25-50 - putting them straight into the red, flashing region of the
> endangered species list.
>
No, Ne would be 100 - the population described is behaving as an ideal population.

Bob

--
Bob O'Hara

Dept. of Mathematics and Statistics
P.O. Box 4 (Yliopistonkatu 5) FIN-00014 University of Helsinki Finland Telephone: +358-9-191 23743
Mobile: +358 50 599 0540 Fax: +358-9-191 22 779 WWW: http://www.RNI.Helsinki.FI/~boh/ Journal
of Negative Results - EEB: http://www.jnr-eeb.org

 

[Tim Tyler]

dkomo <[email protected]> wrote or quoted:
> Tim Tyler wrote:
> > dkomo <[email protected]> wrote or quoted:

> > For the genetics of small populations on the verge of extiction?
> >
> > The quantitly of relevance of the calculaton is known as the "effective populaton size".
> >
> > An effective population size of 500 is considered to be a threshold for status as an endangered
> > species.
>
> What I said about small populations had nothing to do with endangered species. I was referring to
> the following statements here in this thread:

[...]

> Larry Moran:
>
> [...] In the real world it's almost impossible to have a large population of randomly mating
> individuals (i.e., panmitic). Almost any population that you can think of is subdivided into a
> large number of much smaller sub-populations that are more-or-less genetically isolated from each
> other. This is obvious in humans but it's true of every other species as well.
>
> +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> So if it is common to have genetically isolated small sub-populations of large populations of
> animals, the phenomenon of genetic drift is going to be dominant and very widespread. And that is
> going to blow a huge hole into the side of theories that most features of organisms are
> adaptations.

Humans are "subdivided into a large number of much smaller sub-populations that are more-or-less
genetically isolated from each other"?

Give me a break. These populations are extremely unlikely to /actually/ be genetically isolated from
one another. The number of migrants between groups needed to destroy the effect of drift is
notorious for being low - of the order of one migrant per generation - and not many populations are
*that* isolated.

Most of humanity is one big genetic melting pot. There are a /few/ areas where small populations
have invaded other cultures and mate mostly with their neighbours - and take things to the point of
inbreeding depression - but this isn't common and doesn't last long.

Humans deliberately avoid inbreeding - and are usually successful at doing it.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[William Morse]

[email protected] (Larry Moran) wrote in
news:[email protected]:

> On Sat, 28 Feb 2004 18:18:20 +0000 (UTC), William Morse <[email protected]> wrote:
>> [email protected] (Larry Moran) wrote in news:[email protected]:

> Is there a rule about not bringing up the same topic again? It seems to me that there are a some
> topics that seem to be "debated" over and over again on this newsgroup so I assumed that this was
> okay.

> Seriously, I think it's more interesting to discuss adaptionism and random genetic drift than some
> of the other things that are taking up time on sbe. Do you agree?

Yes, entirely. In fact I was hoping to encourage continued discussion with my follow. You may think
that drift is given short shrift on sbe, but I think if you look at number of posters (vs. volume of
posts by a few individuals) you will find that most of us think drift is significant in evolution -
the interesting question is how significant, and this is what I was trying to draw a bead on.

>> If the trait shows little variation throughout a large population or is strongly correlated with
>> an obvious environmental variable, and has an obvious relation to an aspect of species behavior,
>> one can make the default assumption that it is an adaptation, especially if it has existed for a
>> long period.(Again based on recollection, Wirt did a much better job than I just did of defining
>> what traits could be considered adaptive as a default). Examples would include large ears in
>> elephants, skin color in humans, and almost any morphological feature of horseshoe crabs.
>>
>> If the trait shows wide variation throughout a population regardless of environmental variables,
>> or is confined to isolated subpopulations with no obvious relation to fitness, one can make the
>> default assumption that it is due to drift. Examples include coat color in domestic cats and the
>> Rh- blood type in humans.
>
> Hmmmm .... I would have thought that ear size in elephants and skin color in humans were excellent
> example of drift. Don't you see how difficult it is to make general rules? Your decision about
> what the "default" hypothesis should be depends to a great extent on your original biases. This is
> exactly the point that Lewontin and Gould made in their original paper. If you tend to emphasize
> natural selection in your thinking about evolution then you will look to adaptive explanations
> ahead of non-adaptive explanations.

I will stand by my example of ear size in elephants - my reference is a book entitled "Why Elephants
Have Big Ears" - which despite its title was written by Chris Lavers, not Rudyard Kipling In
another follow on this thread you questioned why Indian elephants have smaller ears. Lavers notes
that both Indian elephants and African forest elephants - both of which have smaller ears than
African savanna elephants - inhabit forests with lower temperatures than the savanna. He also notes
that woolly mammoths had tiny ears.

Skin color in humans is obviously controversial, but is a very interesting example. AFAIK skin
pigmentation is fairly well correlated with latitude in the Old World, even among populations with
rather different origins, e.g. Indo-Europeans in India and Bantu speakers in central Africa. And
there is a plausible explanation for this - the balance between Vitamin D production and folate
production. However there is little variation with latitude in the New World - probably an example
of founder effect - while there is considerable variation among relatively isolated island
populations in New Guinea - probably an example of sampling error.

>> Note that I am cheating somewhat, as I am assuming there is at least some data available to make
>> a "default" assumption.
>
> Part of the problem is that even the selection of "data" has built-in biases. If you are an
> adaptionist, and you prefer an adaptionist explanation, then you go looking for data to support
> your assumption.

Of course - that's how science works. Fortunately for me there are people like you that prefer
stochastic explanations and try to poke holes in my data while finding data that support your
assumptions. As long as we are polite in our discourse, honest in our data, and don't exclude from
the table those with ideas out of the mainstream, the system works pretty well.

> No offense. I'm a drifter in the sense that it's part of my self-proclaimed "mission" to educate
> people about the importance of random genetic drift and evolution by accident. I'm not having much
> luck on sci.bio.evolution. This newsgroup is heavily dominated by people who reject the very
> concept of drift or who have deliberately chosen not to understand it. I find this very strange in
> a newsgroup that's supposed to be devoted to discussing evolution at a serious level.

See my comment above. In fact, while I still (as you have undoubtedly noted) have a strong
adaptationist bias, I have learned a lot about drift from discussions on this newsgroup, which have
often lead me to do further reading. You noted in another follow that you are interested in the
mechanism of speciation. While I can keep hoping, I have not seen any convincing explanations based
on adaptation for the extent and pace of speciation. Adaptive explanations may explain _why_ there
should be two different niches available for sticklebacks in Canadian lakes, but not _how_ one
species could rapidly diversify to fill both niches. But it is getting late, so perhaps I will leave
further discussion of that topic for another day.

Yours,

Bill Morse

 

[William Morse]

Tim Tyler <[email protected]> wrote in
news:[email protected]:

> William Morse <[email protected]> wrote or quoted:
>
>> If the trait shows little variation throughout a large population or is strongly correlated with
>> an obvious environmental variable, and has an obvious relation to an aspect of species behavior,
>> one can make the default assumption that it is an adaptation, especially if it has existed for a
>> long period.(Again based on recollection, Wirt did a much better job than I just did of defining
>> what traits could be considered adaptive as a default). Examples would include large ears in
>> elephants, skin color in humans, and almost any morphological feature of horseshoe crabs.
>>
>> If the trait shows wide variation throughout a population regardless of environmental variables,
>> or is confined to isolated subpopulations with no obvious relation to fitness, one can make the
>> default assumption that it is due to drift. Examples include coat color in domestic cats and the
>> Rh- blood type in humans.
>
> Genes for traits can reach fixation by accident in small populations.

Surely. This is why I left the caveat that the default assumption is only an assumption - and why I
did not include traits with little variation in small populations.

> They can remain fixed due to selective lock-in if other adaptations come to depend on their
> presence.

In which case they will still show little variation in large populations
- but those populations will also give more opportunity for competing suites of adaptations to
develop and overcome the lock-in.

> Old, non-variable traits (of any sort) need not /necessarily/ represent adaptations.

I disagree strongly. If they are not being fixed by stabilizing selection, they will drift.

> Incidentally, are you /really/ suggesting a human blood type is neutral? Are not blood types often
> critical in disease resistance? That is surely a case of parasite-driven selection favouring
> diversity.

The Rh- example comes from Cavalli-Sforza. But yes he is really suggesting that Rh- is due to
drift.It apparently comes from a relatively small (historically) population in central Europe - and
given the extreme effects on reproduction (you generally only get one child when mated with an Rh+)
it is unlikely to be parasite-driven. In fact he notes that the absence of B blood types in
Amerindians could also be due to drift - although that one is open to question since it could also
be parasite- driven by syphilis.

Yours,

Bill Morse

 

[Dkomo]

Tim Tyler wrote:
>
> Guy Hoelzer <[email protected]> wrote or quoted:
> > in article [email protected], R.Schenck at
> > > Guy Hoelzer <[email protected]> wrote in message
>
> > >> My point was that natural selection is not the only possible adaptive process.
> > >
> > > what other adpative process is there? i understand that drift is useful for 'adaptive peaks
> > > and valleys' and all, but even then its selection that is the over-riding factor. Or are you
> > > saying that selection is merely not the only mechanism/force/player whatever involved?
> >
> > I am saying something roughly like your last alternative. I think that the evidence is now
> > overwhelming, and mostly overlooked by evolutionary biologists, that natural selection is a
> > particular form of more general (thermodynamic) set of optimizing, adaptive processes.
>
> I.e. self-organising systems.
>
> > Biology specializes on natural selection as an adaptive process, but there is no reason that
> > others kinds of adaptive mechanisms can't also participate in causing adaptive evolution.
>
> I'm sure they do.
>
> To take a simple example, I think the human brain is /heavily/ influenced by "self-organising"
> mechanisms during its development.
>
> Speaking /very/ crudely, I think the genome says something like: make lots of these neuron things,
> according to these rules, wire them together with these sensors and these motor units - and then
> let them get on with things.
>
> The resulting adult brain thus owes a great deal to self-organising processes.
>
> Such processes can influence gene frequencies via selection and the Baldwin effect.
>
> Having said that, natural selection is /still/ pretty fundamental.
>
> Since it gets to decide which self-organising systems exist and which don't, it pretty-much has
> the final word.
>
> In the future we will see completely new ways for adaptions to arise in organisms - namely they
> will be designed and engineered to be there - the products of "mutations" directed by intelligent
> design - but even then, natural selection will still preside over everything.

Natural selection does not and will not preside over everything. It doesn't have any influence over
those phenotypical features that are neutral and don't affect the ability of an organism to
reproduce. Some of us believe that this constitutes the majority of features observed, whether those
features were produced by genetic drift or are accidental side effects like spandrels, or are
results of self-organizing processes.

The new ways you're talking about also won't necessarily be affected by natural selection either.
Consider genetic engineering to boost the intelligence of humans. Humans with IQ's of 250 are not
necessarily going to produce more children than those with IQ's of
95. In fact they may think that the intelligent thing is not to have children at all. Natural
selection can only act where there is differential success in producing offspring. Yet here we
can have a case where there can be a constant, or perhaps even increasing number of IQ 250
humans independent of the reproductive success of those humans. How? By genetic engineering of
the sperm or eggs of those ordinary IQ humans who wish to have enhanced children.

What do you call someone who sees natural selection and adaptation behind every rock? An Ultra-
Darwinist.

[email protected]

 

[Tim Tyler]

Larry Moran <[email protected]> wrote or quoted:
> On Tue, 2 Mar 2004 04:48:40 +0000 (UTC), Jeffrey Turner <[email protected]> wrote:
> > Larry Moran wrote:

> >> Hmmmm .... I would have thought that ear size in elephants and skin color in humans were
> >> excellent example of drift.
> >
> > So where are all the small-eared elephants?
>
> Hmmmm ... I see where you're coming from. You observe that all modern species of elphants have
> bigger ears than their ancestors and their modern cousins such as manatees, dugongs, and hyraxes.
> This leads you to the conclusion that large ears are an adaptation. [...]

A couple of "just so" stories - about why Asian elephants might not /need/ such large ears:

Cooling off is a major challenge to a large creature such as an african elephant. Their ears - and
the skin behind them - are rich with blood vessels - and the African elephant can fan its large ears
backward and forward to circulate cool air over these blood vessels and control its body
temperature.

``Sometimes they will even stand with their face to the wind and spread their ears wide open to
allow the wind to cool the blood in the arteries of their ears.''

- http://ladybug.xs4all.nl/saint/pages/ele/

``Asian elephants are much smaller, weighing between 6,615 and 11,020 pounds at a height of about 7
to 12 feet compared to the 8,820 to 15,430 at 10 to 13 feet of the African elephant.''

- http://www.thewildones.org/Animals/elephant.html

A less extreme volume to surface area ratio will mean less need for alternative thermoregulation
strategies - since the bigger you are, the harder it is to cool off.

Also, the Asian elephant lives further north than the African one - so its typical environment may
be cooler.

In the light of this, I doubt whether the resulting ear size is an accidental feature.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[Tim Tyler]

Larry Moran <[email protected]> wrote or quoted:

>1. Genetic diversity cannot be an adaptation since this requires a form
> of group selection that has been thoroughly discredited. [...]

a different genotype from your neighbours offers some protection from their parasites - and "having
a different genotype from your neighbours" is another way of saying "genetic diversity".

is generating genetic diversity. The resulting diversity acts to slow the spread of parasites
through the population - and increases the number of individuals who are not infected. This benefits
the population - but it /also/ benefits the individuals - and therefore does not need much in the
way of group selection to explain its existence.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[Tim Tyler]

Larry Moran <[email protected]> wrote or quoted:

>2. All kinds of traits can arise by chance. Take the ability of some
> people to roll their tongues as a simple example. [...]

Simple - but /not/ good - it seems:

I found an essay on whether this trait is genetic:

``Everybody who stayed awake through the first nine minutes of high school biology knows the ability
to roll your tongue into the shape of a tube is hereditary.

"We know it's a dominant trait," Wayne Carley, executive director of the National Association of
Biology Teachers, told me.

"We teach it because it's an easy thing to test: You either can roll your tongue, or you can't." .

A public-affairs person at the National Center for Genome Research confidently tells me the same
thing. "If a child can, then at least one of her parents can."

But she's wrong. All the biology teachers are wrong, too.

[...]

Back in 1952 "Matlock" (scientists, like detectives, go by last name) concluded that identical
twins don't always share the tongue-rolling trait. In 1975 "Martin" demonstrated that identical
twins are no more likely to share tongue-rolling than are fraternal twins. In 1983 a Hungarian
named Forrai found no genetic basis for tongue-curling -- or hand- clasping or arm-folding. And
studying the Greeks of Thessaloniki in 1982, Cruz-Gonzalez established that while dry ear wax and
attached ear lobes are recessive traits, which means you need the gene from both your parents, the
genetic basis for tongue-rolling was less clear. [...]''

- http://www.discovery.com/area/skinnyon/skinnyon970226/skinny1.html

I tracked down some of the references - and found:

``TONGUE CURLING, FOLDING, OR ROLLING''

- http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?cmd=entry&id=189300

...which draws much the same conclusion - and has a comprehensive bibliography on the subject.

There is no good evidence that the trait is genetic - and much evidence that it is not.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.

 

[Jeffrey Turner]

Larry Moran wrote:
> Jeffrey Turner <[email protected]> wrote:
>>Larry Moran wrote:
>>>William Morse <[email protected]> wrote:
>>>
>>>
>>>>If the trait shows little variation throughout a large population or is strongly correlated with
>>>>an obvious environmental variable, and has an obvious relation to an aspect of species behavior,
>>>>one can make the default assumption that it is an adaptation, especially if it has existed for a
>>>>long period. (Again based on recollection, Wirt did a much better job than I just did of
>>>>defining what traits could be considered adaptive as a default). Examples would include large
>>>>ears in elephants, skin color in humans, and almost any morphological feature of horseshoe
>>>>crabs.
>>>>
>>>>If the trait shows wide variation throughout a population regardless of environmental variables,
>>>>or is confined to isolated subpopulations with no obvious relation to fitness, one can make the
>>>>default assumption that it is due to drift. Examples include coat color in domestic cats and the
>>>>Rh- blood type in humans.
>>>
>>>Hmmmm .... I would have thought that ear size in elephants and skin color in humans were
>>>excellent example of drift.
>>
>>So where are all the small-eared elephants?
>
> Hmmmm ... I see where you're coming from. You observe that all modern species of elphants have
> bigger ears than their ancestors and their modern cousins such as manatees, dugongs, and hyraxes.
> This leads you to the conclusion that large ears are an adaptation. However, if the ancestors of
> modern elephants just happened to have large ears, and large ears were not a disadvantage, then
> the presence of large ears could just as easily be due to drift. The fact that modern species
> possess a certain characteristic feature is not prima facie evidence of adaptation since drift and
> the founder effect would achieve the same result.

True, but the advantage of large ears in a large mammal that can't really stay out of the tropical
sun is obvious. Why would you think that big ears would be a random occurrence? Drift is probably a
minor effect, while founders are interesting because of the advantage that made them a new species
not because of their random minor traits.

> But that's not what I had in mind. I was referring to the fact that Asian elephants have much
> smaller ears than African elephants. If ear size is under strong selection then one has to account
> for this fact. On the other hand, if ear size is non-adaptive then the difference between Asian
> and African elephants could be due entirely to chance.

Gee, why don't humans have large ears if they are just by chance? Ever think that climate
differences account for the smaller ears in Indian elephants? There's no reason to believe that non-
adaptive mutations lead to speciation - and that's where all the marbles are.

>>>Don't you see how difficult it is to make general rules? Your decision about what the "default"
>>>hypothesis should be depends to a great extent on your original biases. This is exactly the point
>>>that Lewontin and Gould made in their original paper. If you tend to emphasize natural selection
>>>in your thinking about evolution then you will look to adaptive explanations ahead of non-
>>>adaptive explanations.
>>
>>Considering genetic diversity, itself, gives a species an advantage, you're going to have to
>>explain why traits would develop for non-adaptive reasons.
>
>
> Two points ...
>
> 1. Genetic diversity cannot be an adaptation since this requires a form of group selection that
> has been thoroughly discredited. If a species accidently possesses more diversity then it will
> be the lucky survivor when the environment changes. This is more like evolution by chance that
> real adaptation.

Discredited? We all know that in many species that have familial groups, say lions, the males leave
or are expelled from the group to find mates elsewhere. That type of behavior occurs across many
species with the obvious effect of diversifying the genetics. I seem to recall research that showed
that pheromones may play a role in selecting a mate with different characteristics even in humans.

> 2. All kinds of traits can arise by chance. Take the ability of some people to roll their tongues
> as a simple example. Why do I have to explain why traits would develop for non-adaptive
> reasons? Is it because of your (irrational) belief that everything MUST be an adaptation?

Not everything is an adaptation, but adaptation plays a much more important role than chance.

> [snip]
>
>
>>>No offense. I'm a drifter in the sense that it's part of my self-proclaimed "mission" to educate
>>>people about the importance of random genetic drift
>>
>>Genetic drift isn't very interesting, changes in phenotype are all that evolution can work with.
>
>
> Two points ....
>
> 1. The mechanism of random genetic drift is the basic evolutionary mechanism underlying most of
> molecular evolution. I find this very interesting. You may not. Furthermore, I'm quite
> interested in the organization of genomes and junk DNA. Part of the modern explanation of
> genome evolution requries random genetic drift. You may not be interested in this either.
> Finally, I'm really interested in correct explanations of the main features of modern species
> (and speciation in general). If you aren't interested in random genetic drift then you mustn't
> be interested in those things either. It's okay to focus all of your attention on adaptations
> and to admit that nothing else about evolution interests you. But please don't assume that
> others share your bias.

Adaptations ARE evolution. Evolution doesn't care about junk DNA or mutations that are neutral from
the perspective of natural selection.

> 2. The most reasonable minimal definition of evolution is change in the frequency of heritable
> characteristics in a population over time. This definition includes fixation by random genetic
> drift of almost neutral alleles. You seem to want to change that definition to restrict it to
> changes in phenotype - where presumably you mean only observable phenotypes. Furthermore, when
> you say that this is "all that evolution can work" with you imply that evolution is synonomous
> with natural selection. In other words, you would like to change the definition of evolution so
> that the only "real" evolution is adaptation. This is consistent with the fact that you aren't
> interested in any other kind of evolution but here you go one step farther. Now you are
> suggesting that we re-define evolution so that my interests become illegitimate. If you succeed
> (highly unlikely) then what word will you use for all those heritable changes that aren't
> adaptations? Will this be pseudoevolution?

Your interests aren't "illegitimate," they are other aspects of biology. But genetic drift isn't
important to evolution, junk DNA and other unexpressed genes aren't either. Biology and genetics are
areas that hold other intrinsic interest. Why are you trying to shoehorn all of biology into the
study of evolution?

>>>and evolution by accident.
>>
>>Evolution by accident? Can you explain?
>
> Yes, I'm referring to evolution that takes place without natural selection playing a significnat
> role. It includes things such as the random elimination of some species, and the survival of
> others, at the time of mass extinctions. (The so-called "Field of Bullets" senario of David Raup.)

Something usually causes mass extinctions. A significant stressor. It may or may not be transitory.
But I suspect that, for the most part, we can find common characteristics among the surviving
species that would explain why they survived. I have my doubts that things are totally random.

> It includes lots of other things as well. When I use the phrase "evolution by accident" I mean it
> to be a direct contrast to adaptionism. The phrase is intended to provoke people into thinking
> outside their box. (It's also a convenient way to emphasize the contrast between real evolution
> and intelligent design.)

Ah, curiosities for arguing with Luddites. Can you actually convince them of anything?

>>>I'm not having much luck on sci.bio.evolution. This newsgroup is heavily dominated by people who
>>>reject the very concept of drift or who have deliberately chosen not to understand it. I find
>>>this very strange in a newsgroup that's supposed to be devoted to discussing evolution at a
>>>serious level.
>>
>>Maybe because "drift" doesn't seem to be very well defined.
>
> Not true. It's as well-defined as natural selection. It may be a more difficult concept to grasp
> and it may be unfamiliar to those who have been brought up on a steady diet of adaptionist dogma,
> but that's a different kettle of fish. You can't dismiss random genetic drift just because you
> don't understand it.

I can dismiss it because it's not significant in evolution. You haven't shown any method by which it
would cause speciation.

--Jeff

--
A man, a plan, a cat, a canal - Panama!

Ho, ho, ho, hee, hee, hee and a couple of ha, ha, has; That's how we pass the day away, in the merry
old land of Oz.

 

[Jeffrey Turner]

Tim Tyler wrote:
> Jeffrey Turner <[email protected]> wrote or quoted:
>
>
>>Evolution by accident? Can you explain?
>
> Check out the differences between mammals and marsupials.
>
> [what? Marsupials aren't mammals? - JAH]
>
> To /some/ extent they exhibit convergent evolution - with similar forms occupying similar niches.
>
> However in other areas, they show significant differences.
>
> Attributing these differences to selection caused by a different environment is /possible/ - but
> mammals seem to do OK in Australia, despite all the unusual and unfamiliar species over there - so
> the environment probably isn't /that/ different.
>
> The most obvious explanation of the differences is chance. Marsupials *happened* to do things a
> different way - simply since there's more than one way to do things - and the differences became
> fixed over time.

Sure. Species aren't perfect creations, just the best that could be arranged for with the
tools on hand.

> It /looks/ like the marsupials were better at *spreading* initailly - but that they were not so
> good at defending their turf.

Well, when you introduce a new species into an ecosystem it will either be killed off or thrive and
probably kill off or force out other species. Oh, it could also find an empty niche and blend in.

> A similar thing - perhaps on an even more dramatic scale - happened in South America 3 million
> years ago:
>
> ``In the upper Pliocene, 3 million years ago, the isthmus of Panama reappeared as a result of
> changes in the earth's crust. This was a disaster for many of the animals that had evolved in
> isolation in South America. South America was invaded by deer, camels, raccoons, tapirs, horses,
> mastodons, bears, peccaries, rabbits, shrews, cats, dogs, weasels and rodents. For some reason
> these animals were able to displace many of the South American species, driving many of them to
> extinction.''
>
> - http://darwin.bio.uci.edu/~sustain/bio65/lec02/b65lec02.htm#SouthAm
>
> While divergence can be caused by drift, such large scale displacements of whole species are
> examples of selection.
>
> It /often/ seems to be the smaller population that suffers when populations join.
>
> Perhaps there are fewer different sorts of competitor in the smaller environment - and so less
> pressure.
>
> Or perhaps the smaller environment is likely to be less well furnished with specialised species in
> the first place - so when existing specialists do arrive they are simply better equipped than the
> natives (who have more recently adopted the roles) to exploit the available niches.

Hrm, changes in ecosystems happen - sometimes rapidly - due to geological processes and species must
either adapt or die out. That's evolution. Evolution isn't always driven by advantageous mutations
of species within a stable ecosystem. You're not proposing any significant change in the theory,
just applying it to species in extremis.

--Jeff

--
A man, a plan, a cat, a canal - Panama!

Ho, ho, ho, hee, hee, hee and a couple of ha, ha, has; That's how we pass the day away, in the merry
old land of Oz.