Cornelius Hunter seems very confused.
…This brings us back to the UC Berkeley “Understanding Evolution” website. It abuses science in its utterly unfounded claim that “natural selection can produce amazing adaptations.”
In fact natural selection, even at its best, does not “produce” anything. Natural selection does not and cannot influence the construction of any adaptations, amazing or not. If a mutation occurs which improves differential reproduction, then it propagates into future generations. Natural selection is simply the name given to that process. It selects for survival that which already exists. Natural selection has no role in the mutation event. It does not induce mutations, helpful or otherwise, to occur. According to evolutionary theory every single mutation, leading to every single species, is a random event with respect to need.
He has forgotten what “adaptation” means. Of course he is correct that “Natural selection is simply the name given to [differential reproduction]”. And that (as far as we know), “every single mutation …is a random event with respect to need”.
And “adaptation” is the name we give to variants that are preferentially reproduced. So while he would be correct to say that “natural selection” is NOT the name we give to “mutation” (duh); it IS the name we give to the very process that SELECTS those mutations that promote reproduction. i.e. the process that produces adaptation.
Cornelius should spend more time at the Understanding Evolution website.
ETA: CharlieM points out below that…
When CH says that natural selection does not produce adaptations he is talking about individual organisms. He is discussing mutations in individuals and adaptations in individuals. Natural selection has nothing to do with the first appearance of an adaptation in an individual.
And that of course is the confusion – I hadn’t seen just where Cornelius’s confusion lay. Because, of course, the term “adaptation” is a population-level concept. At the level of the individual, the equivalent would be “advantageous mutation”. And that makes the Understanding Evolution website absolutely correct.
Keiths, which is greater:
1,2,3,4…
2,4,8,16…
petrushka,
It’s almost indistinguishable in effect from an advantage that operates through increased offspring, survival or sexiness, at any rate.
Hang about. Shouldn’t you ask why the B’s are more fecund? Producing more offspring (presumably viable ones) carries the implication that B is more efficient at exploiting environmental resources. In other words, the environment is indeed favoring B over A, and selection is occurring. If B was simply more fecund, but the offspring were less able to survive, the allele would decrease in frequency, surely.
The mode of increased fecundity is not specified. It could involve tail feathers.
Anyway, when “without death or limit” is specified, you are talking about infinities, and one arithmetic or geometric infinity is the same as another.
I didn’t hear the word horseshoes.
petrushka,
No, it is not a Cantor-like thing.
You are repeating JoeG’s elementary mistake. A population that is increasing without limit is not infinite. At any given time it is finite, period.
The percentage of the faster-reproducing allele in that finite population increases with time, and it has nothing to do with Cantor.
keiths, Right. It’s like Locke’s “as good and enough to go around.” It need not mean that the land mass is infinite–it’s finitude just isn’t affecting anything…yet.
petrushka,
I thought you meant the halving-generation-time one.
I missed that in the discussion. I am not repeating JoeG’s mistake. I simply didn’t see the time limit condition. But that is a limit. and the more fecund allele is still obeying the rule of differential reproductive success. In this case, faster is more successful.
I don’t see that the specifics make any difference. Faster reproduction without side effects is differential reproductive success.
petrushka,
Yes, that was my prior point.
Yes, I know. What I’m saying is that I don’t see any evidence he thought about the case in which death is not occurring. I see no reason to think he considered this kind of situation and decided that a definition of natural selection that excluded it was superior to one that did. He simply seems to be following Darwin.
I agree that biologists have stated definitions that disagree with mine. What I am still quite skeptical about is whether they would still hold to their definition if they tried applying it to your case.
As I said, this is not theoretical. Biologists use “natural selection” to describe situations very much like yours with some regularity, and they never use any other definition. They’re not generally population geneticists. This suggests to me that the definitions actually used by biologists do not coincide with the ones you’ve stated.
I try very hard to avoid words like fitness and selection, and use differential reproductive success instead.
To me it appears compatible with change in allele frequency.
If an allele spreads at greater than the expected neutral rate, we infer an advantage.
But it might be adventitious. The bean jar game shows wild swings without any “cause”.
petrushka,
We were talking about how the faster-reproducing allele comes to dominate the population, and that happens long before t equals infinity.
petrushka,
In my scenario, yes, but that’s not the topic of debate.
The question is whether natural selection is happening in my scenario given that all offspring survive and reproduce. By the pop gen definition, selection is happening. By the Darwin/Mayr definition, it isn’t.
I think the Darwin/Mayr definition is more appropriate for the purposes of this thread, and I’ll elaborate on this later. For now I’m concentrating on disabusing Steve and Joe of their “One True Scientific Definition” misconception.
petrushka,
Sure you are. You wrote:
That’s not right. The population is growing without limit, but it is not infinite.
timothya,
It isn’t the environment that is favoring B over A — it’s B’s relative fecundity. The environment treats each A exactly the same as each B: they get all the resources they want, they all survive, and they all reproduce.
The pop gen definition of selection includes environmental selection and the effects of differential fecundity. The Darwin/Mayr definition excludes the latter, restricting selection to the picking and choosing done by the environment, in which some variants are eliminated and others are preserved.
I prefer the latter definition in this context because it separates environmental selection from differential fecundity, which is a property of the organism itself and not of its environment.
It could go either way, depending on whether the fecundity advantage outweighed the decrease in viability or vice-versa. If strain B doubles its fecundity but loses only 5% of each offspring’s viability, then it will dominate strain A. If strain B gains 5% in fecundity but loses half of each offspring’s viability, then strain A will dominate instead.
keiths:
Steve:
It doesn’t matter. His definition unambiguously requires elimination:
No elimination means no selection by Mayr’s definition.
Mayr was an evolutionary theorist, and a renowned one. You and Joe are simply wrong about there being One True Scientific Definition of natural selection.
As boring as it is to admit, I agree with ‘keiths’.
Steve’s quasi-atheist ideology is simply presumptive of general communication. Do you biologists/geneticists really wish to act as if you’re at a department meeting? You sound like KN with his dozens of flavours of -isms.
Above I quoted Darwin’s letter (1860) about ‘natural selection’ & ‘natural preservation’ to Lyell. Darwinists might possibly pay attention. But hey, why not just better ignore it?
Anything else than to face ‘agency’ questions in ‘nature/Nature’, right TSZ atheists? Anything than to acknowledge the Malthusianism in your Darwinian evolutionism. Steve shows his allegiance to ideological naturalism far more often than to theism. But hey, isn’t he too supposed to act entirely as a ‘skeptic’ here?
I agree with walto, but then, I’m always wrong.
Mung,
ALMOST always. 🙂
My disagreement with Joe and Steve is reminiscent of the debate over sexual selection: is it a kind of natural selection, or is it distinct?
The answer, of course, is definitional. If you consider potential mating partners to be part of the environment, even in their role as potential mating partners, then you’ll regard sexual selection as a subset of natural selection. (I’m in that camp.)
If you see mating as a distinct from normal organism/environment interactions, you’ll tend to regard sexual selection as separate from natural selection.
Is one of the two positions “scientifically invalid”, to use Joe’s phrase? Of course not.
Likewise with differential fecundity. Some people regard it as part of selection, others don’t, but neither camp is thereby being unscientific.
As I wrote to timothya:
I prefer to think of natural selection as something the environment does, just as artificial selection is something the breeder does. But differential fecundity is a property of the organism, not the environment. The distinction between organism and environment is a useful one, and so is the distinction between differential fecundity and natural selection, for the same reason.
I reiterate that I am suggesting neither that the pop gen definition is wrong nor that it should be changed. I simply think the Darwin/Mayr definition is more appropriate for the purposes of this thread.
I am utterly opposed to the position that it’s “natural selection” even if there is no selection. That’s just silly talk. But that’s not all.
Natural selection is offered as an explanation for how the very improbable can actually occur. If someone is going to argue that it’s still selection even if there is no change in probability then you may as well concede the argument to the CREATIONISTS.
If natural selection didn’t increase the probability of an eye, what use is it?
Actually, I agree with Joe F.
There’s only one scientific version of “natural selection” and it isn’t the one common here at TSZ, upon which programs like Avida are based.
I see your recursive trick, Mung!
No, you’re not. 🙂
Mung,
Having a hard time knowing which is which here. Could you separately state the ‘one scientific version’ Joe favours, and the one common here at TSZ and used in competitive algorithms?
keiths,
That’s fine, but (as of course you know) it turns out that they all reduce to the same: a difference in the rate-of-increase between carriers and non-carriers of a trait. If one needs to separate them out between ‘internal’ and ‘external’ factors, one can do so.
I think the disease case is persuasive. If there is differential output between two strains in a body, and it is simply due to greater efficiency in DNA replication by one, is that environmental or internal? One could argue for both. It isn’t anything outside that makes the faster strain faster. But when the two strains are in the same body, they are each other’s environment. The faster strain is faster by reference to the slower, and vice versa. The bugs may both be growing without limit, so you would say there is no selection. Then a mosquito samples the population. Suddenly, there is a difference between the two, a difference in competition for the exit.
The same happens in sperm competition. And it would be eccentric to argue that streamlined genomes in prokaryotes are not due to selection because ‘the environment’ does not care. The environment is stuffed with other genomes, including your relatives. Representation in the population is a ‘resource’ for which one can compete, even if all other resources are not competed for. Faster reproduction is a causal factor in streamlining prokaryotes.
This was what persuaded me to abandon your ship earlier in the thread – when something comes along that randomly samples the population, taking a finite segment of it, and the frequencies depend significantly on heritable factors, that has the same effect as if the sampler itself were discarding some preferentially. The mosquito is a selective pressure, even though it is having no direct effect.
One can certainly separate out natural selection, fecundity selection and sexual selection. And I think this debate turns on that. When population geneticists talk of ‘selection’, they mean all of ’em. One can reserve ‘Natural’ selection only for the survival part if one wishes. The first word tends to be dropped anyway. But it’s easy to read ‘selection’ and imagine that ‘natural’ has been stuck at the beginning when it hasn’t.
Did you read that in those books about evolution you have claimed to have read?
Allan,
Yes, and tomorrow I’ll give my reasons for wanting, in the context of this thread, to keep fecundity separate from natural selection. I’m off to bed.
As to whether sexual selection is different from natural selection, that is a matter of the definitions you want to make (I prefer to say that the latter includes the former, but your mileage may vary).
My point is that the definition used by keiths, which was apparently used by Mayr in the passage cites by keiths, is silly. However eminent Mayr was, it was silly.
My point was not that all possible definitions other than mine are invalid. Just the one used by keiths.
OK, I thought I had got my head around that bit. What I am perhaps still unconvinced rather than sceptical about is not what drift is, but what drift achieves. In the absence of new alleles turning up and in the absence of advantageous or deleterious alleles, drift will eventually result in a population having a set of fixed alleles. OM’s model demonstrates the process admirably.
Absolutely no problem with the model. I’m sold. Also I can’t recall thanking and congratulating OM for taking the time to develop it. So belated thanks to OM.
I have an image of trying to balance a pencil on its point. The slightest perturbation, an air current, say, will disturb the unstable equilibrium and it will fall, one way or another.
So what is the significance of drift to the process of evolution? Do we need a process that eliminates too much variation?
timothya,
Yes, but in Keiths’s pointless thought experiment, he proposes there are limitless resources. Population A and Population B grow at optimal rates. Keiths’s TE is not wrong: it is just pointless and trivial. If we are interested in modelling reality, finite resources might be something worth including in the model.
Personally, (though I don’t expect anyone to care particularly what I think!) I tend to call all cases where a trait assists or hinders its own prevalence ‘natural selection’, regardless how implemented.
If we allow some latitude in what consists of a ‘being’ – a genetic segment, a chromosome, an organelle, a cell, an organism, a colony, a species – and allow ‘survival’ to extend to multi-generational persistence of the heritable part, I’d say CD had it covered.
Ernst Mayr as quoted by Keiths
The conclusion that these favored individuals had been selected to survive requires an answer to the question, Who does the selecting? In the case of artificial selection, it is indeed the animal or plant breeder who selects certain superior individuals to serve as the breeding stock of the next generation. But, strictly speaking, there is no such agent involved in natural selection.
[I would disagree with Mayr here. The environment is the selecting “engine”. Adaptations that benefit an organism in its niche are selected by the niche.]
What Darwin called natural selection is actually a process of elimination.
[As is artificial selection and sexual selection. The denial of opportunity to breed is eliminative for “deleterious alleles” whether demonstrated so by the niche, decided so by the breeder, or not though appealing by potential mates. This is an opportunity for “beneficial alleles”.]
The progenitors of the next generation are those individuals among their parents’ offspring who survived owing to luck or the possession of characteristics that made them particularly well adapted for the prevailing environmental conditions.
[And who were able to produce offspring]
All their siblings were eliminated by the process of natural selection.
I just hope you are more amused than offended by Keiths’s chutzpah.
Indeed. Why do we need the qualifier “natural”? Potential mates and human plant and animal breeders are part of the environment, surely? Why not say “environmental selection” to cover natural, artificial and sexual selection?
So…
Drift?
What does it do for us?
Alan Fox,
Its fundamental significance is that it is the thing that makes evolution inevitable. Populations are always in the process of losing variation and always in the process of gaining more, regardless of selective advantage. Things don’t just stop when a ‘colour’ is fixed – new variants can be promoted by the same process which eliminated the old. That gained is not the same as that lost. The primary engine of this is actually drift – loss of variation due to sampling. You can’t turn it off.
Drift does not just apply to neutral alleles. If OM’s simulation were furnished with a knob by which you could control selective advantage for one colour, in the current simulation it would be set to zero – there is no advantage. If you turn it up a fraction, you will see increase more frequently for that colour, but you would still see decrease as well, just with less frequency. Those decreases are due to drift – to the fact that the process is stochastic, not deterministic. Each click of the knob gives a more sure arrow to your chosen colour, but fluctuation still occurs.
Deleterious alleles can increase in frequency due to it (potentially to fixation), and beneficial alleles can reduce in frequency due to it (potentially to extinction).
It doesn’t just eliminate variation, it helps generate it too, in tandem with fresh mutation. Selection acts more rapidly and strongly, and so tends to squeeze variation out of a population, both at its locus and its physical neighbours in the genome. But as fast as selection fixes an allele neighbourhood, variants neutral or nearly so with respect to each other can arise and drift. So with drift, a population can have a greater proportion of standing variation than without.
This allows responsiveness to environmental change – it is not necessary to wait for the stress before a mutation can arise; in a varied population, it will frequently already be there. In conjunction with recombination, it also allows advantageous double-mutations to arise with greater frequency from non-advantageous singletons, because two separate subpopulations can arise through drift and recombine. And it shakes populations out of local maxima in an adaptive landscape, potentially allowing adaptation to higher fitness peaks.
It is also the only force that can promote a recessive allele to frequency in a population, until homozygotes start being formed at least.
And it is likely to be a significant factor in divergence of reproductively isolated species, because it never stops, is genome-wide and does not depend on the environment to drive it.
OK. So without drift, new variation would not find its way into the genome? As quickly? Too quickly? It’s undeniable that without existing and new variation there can be no phenotypic change. I’m still unsure what drift brings to the table.
ETA should have read to the end of your comment. Will address in next comment.
Alan Fox,
Apart from the inevitability, the speciation, the variation, the local maxima, the exploration, the recessive alleles, the recombinational pools, the evolutionary resilience … what does drift do for us? 🙂
I’m not quite grasping this. I think I understand how allele frequency will shift to fixation of a (any) particular allele in the absence of any selective bias but the bit I’m missing is how this generates (or helps generate) variation.
Allan Miller,
You left out the sanitation, the medicine, education, wine, public order, irrigation, roads, the fresh-water system, and public health.
Yeah, well, the Romans had pointy swords and slaves. How does drift do it, when the mechanism is the tendency for one allele to move to fixation?
Patrick,
Don’t you bloody start! Mathematicians!
On doing some further reading I see Mayr was not overly impressed with Kimura’s neutral theory. But then the phenomenon of drift was proposed by Kimura to explain the “molecular clock” observation. I’m still struggling to see what drift achieves (or why it is a necessary ingredient) in a context of phenotypic change over time. How is it necessary to explain adaptation?
Glad to help. If there are similar simulations that may also be illustrative by all means let me know. I don’t have much time at the moment, but I can slot stuff in as and when.
Joe,
As I said:
So far we’re in agreeement.
keiths:
Joe:
Your argument would a bit more convincing if it were an actual argument, Joe. “It’s silly and it’s invalid” is mere assertion, not an argument.
If it’s scientifically legitimate to disentangle sexual selection from natural selection — and it clearly is — then why can’t differential fecundity be similarly disentangled?
Please be specific.
Alan, to Joe:
My chutzpah at what, disagreeing with him?
Alan, you say the silliest things sometimes.
I came across John H.Gillespie who seems to have built his career on arguing with Kimura.
Example (PDF)
Well, not exactly that. Joe is an established and universally respected professional geneticist discussing matters within his field of expertise. You are an engineer, I believe.