Evolution is gradual ?

[CharonZ]

Well actually according to most models those recessive allels won't be eliminated at all, as there is no selection against the heterozygote form. Of course the percentage will be considerably small, but be remain stable at that level.

In addition according to a paper I read recently the so-called puctuated equilibrium problem has been basically settled, in the context of modern synthesis. I have to admit that I didn't look and read the relevant papers, though.

I do think that it might have been still problem in the old neo-Darwinian model, though.

Ref:
Kutschera U, Niklas KJ.The modern theory of biological evolution: an expanded synthesis.

 

[QuarkHead]

Well actually according to most models those recessive allels won't be eliminated at all, as there is no selection against the heterozygote form. Of course the percentage will be considerably small, but be remain stable at that level.

I just came back to correct an error in my last, and here you are! I said populations are alway "in" H-W equilibrium. This, of course cannot be the case where the recessive homozygote is lethal. What I should have said is populations always tend to H-W equilibrium.
Anyway, you're right that full selection against the recessive homozygote by itself will not completely eliminate the heterozygote (any more than repeatedly dividing by 2 ever gets you zero), but once it is below a certain frequency we expect drift to finish the job (usually)

 

[Thersites]

Everybodt speaks of "perfectly adapted" organisms: isn't it more often the case that living beings are adequately adapted to their environment? Often when a foreign competitor is introduced it proves to be function better in an environment than the native species which have been there for a very long time.

 

[CharonZ]

Well this was only a thought experiment (or at least I understood it as that). In this hypothetical model there is a perfect genotype and every aberration results in lowered fitness. In other words, identical clones of this assumed perfect model would have the highest overall fitness, and it is clear that this model does not follow nature.
Of course there is never a state a perfect equilibrium (including perfect adaptation).
And in fact, overspecialization quite often lead to extinction in reality.

Well and yes, genetic drift could eliminate allels with low frequency, but this is of course a purely stochastic process. Depending on the process, it could by chance also locally increase the occurence of detrimental allels. That is, there is no direction of elimination those allels that did not by themselves lead to detrimental phenotypes.

 

[QuarkHead]

Well and yes, genetic drift could eliminate allels with low frequency, but this is of course a purely stochastic process. Depending on the process, it could by chance also locally increase the occurence of detrimental allels.

Didn't I tell you mine is a perfectly outbreeding population? No? silly me! Anyway if it is, here are no local effects. You are of course right though in general, that's why I said "(usually)"

That is, there is no direction of elimination those allels that did not by themselves lead to detrimental phenotypes.

But we would not expect drift to be a significant factor on an allele whose frequency is not very low, for which I'll let selection take the credit.

It is in consideration of this that bottle-necks and founder effects are so interesting, because they can be thought of as special cases of fixation by stochastic means.

 

[QuarkHead]

Everybodt speaks of "perfectly adapted" organisms: isn't it more often the case that living beings are adequately adapted to their environment? Often when a foreign competitor is introduced it proves to be function better in an environment than the native species which have been there for a very long time.

I like that term - adequately adapted. I'm going to use it myself, if you don't mind.
You are right. But I'm modelling. What we do, in all sciences, is we knowingly make unrealistic simplifing assumptions, get our model, see if it fits the real world, then see which of our assumptions needs to be fiddled with so the model matches reality.
Think of the perfect blackbody in physics, the ideal gas in chemistry etc.

 

[CharonZ]

Hum well, I partly agree. Though in this case of course so many basal factors are neglected or not specified that it is not really usable to derive hypotheses. Especially in biology simplification is a big problem, as we always deal with open, complex systems here, as compared to other natural sciences.

 

[QuarkHead]

Hum well, I partly agree. Though in this case of course so many basal factors are neglected or not specified that it is not really usable to derive hypotheses. Especially in biology simplification is a big problem, as we always deal with open, complex systems here, as compared to other natural sciences.

Well, my Ferryman friend, you seem determined to disagree with me about something! But this isn't important or interesting enough to argue about. But let me just say that it is precisely because of the complexity of biological systems that simplified models are so useful as a guide.

If you want a real argument, try this: the key to understanding punctuated equilibria (and possibly a lot more besides) is gene families. Think about it, then draw your sword!!

 

[CharonZ]

Well, if you want, I'll disagree

the key to understanding punctuated equilibria (and possibly a lot more besides) is gene families

Actually I think I don't quite get what you mean, care to elaborate? I suppose I missed some important points here.

I suppose I mainly don't see how the proposed models do relate to the PE theory (as as far as i I see it) try to envision a situation in which PE probably won't happen. But as they are not reflected in nature, the conclusions cannot be tested.

But more to PE:

If I recall the papers correctly, in which PE were stated (that is the papers and book sections by Gould somewhere in the 70s), it was clearly stated that PE was clearly a consequence of the allopatric speciation model.
This is, at best a further addition to the (neo)-Darwinian model (and has been integrated into the synthetic model). As such, there is no serious problem of PE per se.
So, my main point is that I don't clearly see PE as something special within evolutionary theories. The major problems at the time when the PE thory was formulated were mainly some deductions that Gould and colleague proposed as a direct effect of PE, but PE itself is well funded within the allopatric speciation model.

 

[QuarkHead]

Actually I think I don't quite get what you mean, care to elaborate? I suppose I missed some important points here.

Later. Let's deal with this first.

....PE was clearly a consequence of the allopatric speciation model.

Well yes. That allopatry may lead to speciation is self evident. But look - let's remember we may not be alone. In case there are people out there who think we are a couple of experts using funny words, let me put some flesh on these bones.
A population where all possible matings are potentially realizable is called sympatric. A population where this is not the case is called allopatric - there may be any number of reasons for this, let's assume for now it is geographical (which makes it easier to think of local environmental conditions i.e. different selection criteria for allopatric populations).

This is, at best a further addition to the (neo)-Darwinian model (and has been integrated into the synthetic model). As such, there is no serious problem of PE per se.

I suppose if two allopatric and genetically different populations merge, then anything might happen, all in a rush, resulting in some sort of punctuation. But my point would be that, while allopatry self-evidently may lead to speciation according to the gradualism of the classical model, it cannot, without further assumptions, result in a punctuation. And the evidence that links one sub-species to another is not there. That's what PE means!

Anyway, more on gene families later, if you are still interested.

 

[CharonZ]

I suppose if two allopatric and genetically different populations merge, then anything might happen, all in a rush, resulting in some sort of punctuation. But my point would be that, while allopatry self-evidently may lead to speciation according to the gradualism of the classical model, it cannot, without further assumptions, result in a punctuation.

Actually it was argued the other way round. Isolation and genetic drift can, as some phylogenetic studies imply, lead to allopatric speciation. As these effects do not need to be gradual, it may in fact lead to punctuation. As such subspecies are a non-issue here.
So, what Gould and Eldredge simply proposed is that PE is a special case of allopatric speciation. In their view large populations are usually stable (and do not even need to be perfectly adapted as in the model you proposed).
However small isolated groups can evolve more rapidly, due to smaller genetic variation (in conjuncture with the given environmental effects). That is, basically due to the founder effect.
So, small populations are usually evolutionarily more unstable and evolution rates increase, until the conditions for stasis are met again.

This was more or less what I remembered regarding PE. As I am working mostly on microbial genetics (where slightly different rules apply), and not that much in evolutionary biology, I may have overseen some stuff, though.

Regarding gene families, sure, I am eager to hear about that.

 

[QuarkHead]

In their view large populations are usually stable .......
However small isolated groups can evolve more rapidly, due to smaller genetic variation ......basically due to the founder effect. So, small populations are usually evolutionarily more unstable and evolution rates increase, until the conditions for stasis are met again..

What?? Smaller gene pools allow for more rapid evolutionary change? How so, Mr. Boatman? The founder effect severely restricts the genetic variation in a population. It's a simple equation: variation + selection = evolution. Reduce either term on the LHS and the RHS correspondingly decreases!

Regarding gene families, sure, I am eager to hear about that.

Yes...I'm thinking of starting a new thread. But I wonder who else would be interested (other than you) ? In spite of my argumentative tone, I am enjoying "talking" to you.

 

[CharonZ]

Yes! In smaller population evolutionary effects act faster (that is, in less generations).

Evolution is nothing but a change in allel frequency in a population, right? Now imagine a limited population size. Here stochastic events (genetic drift for example) have a much higher influence on the allel frequency than in large population. As such, large populations have a far more stable distribution of allel frequencies than small ones.

In addition, if a smaller variation exists selection can act quicker. In a large population the frequency reduction of a given detrimental allel is far slower than in a slow one, as one can easily imagine (e.g. if in a smaller population only one individual is carrying the allel, the loss of it is far higher compared to a population where 100 are carrying the given allel).
A higher variation only means that there is a higher potential of adaptation, not a difference in the evolutionary path.

Well and if you think that the given talk is not of interest to anyone else (wouldn't be the first thread more or less highjacked by two peeps) we can always continue in pm.

 

[QuarkHead]

Evolution is nothing but a change in allel frequency in a population, right?

Yes - nicely stated.

Now imagine a limited population size. Here stochastic events (genetic drift for example) have a much higher influence on the allel frequency than in large population.

Mmmm...not sure. I'll say yes for now

As such, large populations have a far more stable distribution of allel frequencies than small ones.

Why no!! Remember Hardy-Weinberg. p² + 2pq + q² always.

...if a smaller variation exists selection can act quicker.

This is where we part company. I'm not going to be too dogmatic, because I have been wrong before, but I simply do not see this. Take an extreme example - no variation. No selection, right? How do you translate that into less variation, more selection

In a large population the frequency reduction of a given detrimental allel is far slower than in a small one, as one can easily imagine (e.g. if in a smaller population only one individual is carrying the allel, the loss of it is far higher compared to a population where 100 are carrying the given allel).

ah! Are you confusing "variation" in the general sense of "genetic diversity" with "variation about the mean"? I need to get back to you on that, but not tonight.

A higher variation only means that there is a higher potential of adaptation, not a difference in the evolutionary path.

I agree with the first part of that (who would not), but simply do not understand the second!

Well and if you think that the given talk is not of interest to anyone else (wouldn't be the first thread more or less highjacked by two peeps) we can always continue in pm.

High-jackers?? Pfft! I seem quite capable of giving offence unintentionally. See me when I mean it!

 

[QuarkHead]

High-jackers? Mmm..maybe hi-jackers.

 

[CharonZ]

Why no!! Remember Hardy-Weinberg. p2 + 2pq + q2 always.

Ah, no. Hardy Weinberg explicitly does not apply to small populations!
One of the basic tenets of H-W is that the population is of inifinte size.
What it states is basically that in the given large population even genes with no present selective values will be retained.

These are the basic prerequisites for H-W to apply:


- mutation is not occurring
- natural selection is not occurring
- the population is infinitely (or very) large
- all members of the population breed
- all mating is totally random
- everyone produces the same number of offspring
- there is no migration in or out of the population
- no change in selective pressure


As one can easily see the above given effects, especially genetic drift and possibly natural selection act harsher on the gene pool of a small population.
Again the example, imagine a given allel is present in 1% of the population. Now envision a population of 100,000 individuals and one of 100. In the first case 1000 individuals carry that allel, in the second only one.
Now due to a randome event in both cases one of the carriers of the allel dies. in the first case there are still 99 and given something like h-w equlibrium it will come near to the basic distribution again. In the second case the allel is completely lost, resulting in a change of allel frequency for this allel from 1% to 0%, with no chance of going to equilibrium conditions again (barring gene influx).

 

[QuarkHead]

Ah, no. Hardy Weinberg explicitly does not apply to small populations!
One of the basic tenets of H-W is that the population is of inifinte size.

As I thought, we are at cross-purpose. Note that when I used the term Hardy-Weinberg, I deliberately did not include the term "equilibrium". Why? Because if p and q are allele frequencies, I am free to interpret (p + q)² = 1 as a probability distribution. Which means I can apply H-W probability distribution to a population of any size ( given the other constraints you mentioned).

As one can easily see the above given effects, especially genetic drift and possibly natural selection act harsher on the gene pool of a small population.

Sure, but that's not all you said earlier. You said

In addition, if a smaller variation exists selection can act quicker.

Which is clearly not true. I got confused whether you meant "variation about the mean" or "genetic variation".
Look. Suppose I have two similar but not identical objects which I can measure in some way. From those measurements I can extract a "mean" (oh dear!) and two tails. If I take one of those objects at random and ask what is the probability that it lies within a certain arbitrary (small) distance from the mean, the answer is of course zero, because each object is itself a tail, and thus is as far from the mean as it's possible to be.
So, for these two objects, variation about the mean is about as big as it gets. But if I then ask - what is the actual variation within my sample, the answer is, of course, either one of two values. In other words, the variation defined this way is small.

Oh well. I'm sure we agree, really. And in any case, it is hardly a controversial subject.
Pax!

 

[CharonZ]

Ah yes, I see what you mean. I suppose I should have been more precise about relative and absolute numbers. When I said natural selection acts faster I meant of course that given a stable selective pressure the given time frame in which it will lead to a change in allel frequency in a smaller population will be smaller than in a large one.

And yes, it ain't controversial at all. as the whole PE-discussion. It was only controversial when it was proposed because both, proponents and opponents of that time initially misunderstood each other (hey, what does it remind me of?), but as of now it has been clearly resolved.

Nonetheless, it was somewhat fun trying to remember old stuff that one learned when one was student, wasn't it? Ah the olden days ;P

 

[spuriousmonkey]

We are all talking gene frequencies again. Evolution is also (or totally?) about form though.

Is the evolution of form gradual? I would say yes in principal. No true new form can arise overnight.

Is the speed of evolution variable? Yes, so form can remain constant for a long time and then change quickly.

 

[QuarkHead]

We are all talking gene frequencies again.

Yes, we are. As Charon so elegantly put it, evolution is all about about changing allele frequencies. C'est tout.

Is the evolution of form gradual? I would say yes in principal. No true new form can arise overnight.

Or would you rather...

Yes, so form can remain constant for a long time and then change quickly.

Huh?